Flux of Nutrients from Russian Rivers to the Arctic Ocean: Can We Establish a Baseline Against Which to Judge Future Changes? R

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by UNH Scholars' Repository

University of New Hampshire University of New Hampshire Scholars' Repository

Faculty Publications

8-1-2000 Flux of nutrients from Russian rivers to the Arctic Ocean: Can we establish a baseline against which to judge future changes? R. M. Holmes

B. J. Peterson

V. V. Gordeev

A. V. Zhulidov

M. Meybeck Université de Paris

See next page for additional authors

Follow this and additional works at: https://scholars.unh.edu/faculty_pubs

Recommended Citation Holmes, R.M., B.J. Peterson, V.V. Gordeev, A.V. Zhulidov, M. Meybeck, R.B. Lammers, and C.J. Vorosmarty (2000) Flux of nutrients from Russian rivers to the Arctic Ocean: Can we establish a baseline against which to judge future changes? Water Resources Research, 36:2309-2320.

This Article is brought to you for free and open access by University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Faculty Publications by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. Authors R. M. Holmes, B. J. Peterson, V. V. Gordeev, A. V. Zhulidov, M. Meybeck, Richard B. Lammers, and Charles J. Vorosmarty

This article is available at University of New Hampshire Scholars' Repository: https://scholars.unh.edu/faculty_pubs/150 WATER RESOURCES RESEARCH, VOL. 36, NO. 8, PAGES 2309-2320, AUGUST 2000

Flux of nutrients from Russian rivers to the Arctic Ocean: Can we establish a baseline against which to judge future changes? R. M. Holmes,• B. J. Peterson,• V. V. Gordeev,2 A. V. Zhulidov,3 M. Meybeck,4 R. B. Lammers? and C. J. V6r6smartys

Abstract. Climate modelspredict significantwarming in the Arctic in the 21st century, whichwill impact the functioningof terrestrialand aquaticecosystems as well as alter land-oceaninteractions in the Arctic. Becauseriver dischargeand nutrient flux integrate large-scaleprocesses, they shouldbe sensitiveindicators of change,but detectionof future changesrequires knowledge of currentconditions. Our objectivein this paper is to evaluatethe current state of affairswith respectto estimatingnutrient flux to the Arctic Oceanfrom Russianrivers. To this end we provideestimates of contemporary(1970s- 1990s)nitrate, ammonium,and phosphatefluxes to the Arctic Oceanfor 15 large Russian rivers.We rely primarily on the extensivedata archivesof the former SovietUnion and current RussianFederation and comparethese values to other estimatesand to model predictions.Large discrepanciesexist among the variousestimates. These uncertainties must be resolvedso that the scientificcommunity will have reliable data with which to calibrateArctic biogeochemicalmodels and so that we will have a baselineagainst which to judge future changes(either naturalor anthropogenic)in the Arctic watershed.

1. Introduction M. Meybeck and A. Ragu, Rivers Dischargesto the Oceans: An Assessmentof SuspendedSolids, Major Ions and Nutri- Earth's temperatureis predictedto rise 1ø-3.5øCin the next ents, book draft, United Nations Environment Programme, century,with evengreater increases in the Arctic [Houghtonet 1995] (hereinafter referred to as Meybeck and Ragu, book al., 1996]. This temperature increaseis expectedto impact draft, 1995) (see also Global EnvironmentalMonitoring Sys- numerousaspects of the Arctic system,including the extent of tem (GEMS-Water), United Nations Environment Pro- permafrostand ice-coveredregions, the amount and distribu- gramme,www.cciw.ca/gems/), there hasbeen little criticaleval- tion of precipitation,and the productivityand biogeochemistry uation of the published values. We will derive nitrate, of terrestrialand aquaticecosystems [Chapin et al., 1995;Anisi- ammonium,and phosphateflux estimatesfor 15 Russianrivers mov and Nelson,1996; Hobble et al., 1998;Serreze et al., 2000]. that enter the Arctic Oceanusing a previouslyunavailable data All of thesechanges will affectriver dischargeand nutrientflux set and compareour estimateto model predictions[Seitzinger to the Arctic Ocean, which in turn may impact Arctic Ocean and Kroeze,1998] and to other data. We will concludethat in processes[Aagaard and Carmack,1989; Broecker, 1997; Ander- spite of the extensivedata set, it is currentlynot possibleto son et al., 1998]. Becauseriver dischargeand nutrient flux quantifyriverine nutrient flux to the Arctic Ocean with suffi- integratelarge-scale watershed processes, they shouldbe early cient confidenceto establisha contemporarybaseline. We will and accurateindicators of climate changein the Arctic. arguethat the scientificcommunity must soon resolve the remain- Detectionof future changesrequires knowledge of current ing uncertaintiesso that we do not squandera powerfuloppor- conditions.In this paper,we assesscurrent (1970s-1990s) nu- tunityto detectthe impactof dimatechange on theArctic system. trient flux from Eurasia to the Arctic Ocean. We focus on Russianrivers becausethe majority of riverine input to the Arctic Ocean comes from Russia.Although several sources 2. Description of Data Set report nutrient concentrationsand fluxes for RussianArctic During the Sovietera the Russianwater qualitymonitoring rivers[Alekin and Brazhnikova, 1964; Tarasov et al., 1988;Smir- systemwas among the mostextensive on Earth. However,prior nov, 1994; Gordeev et al., 1996; Gordeevand Tsirkunov, 1998; to the 1990s,scientists (Russian and otherwise)were unableto access,analyze, or publishthe officialwater qualitydata of the •The EcosystemsCenter, Marine BiologicalLaboratory, Woods former SovietUnion (FSU), largelybecause of political and Hole, Massachusetts. ideologicalreasons [Zhulidov et al., 1998].Such restrictions no 2p.p. ShirshovInstitute of Oceanology,Russian Academy of Sci- ences, Moscow. longer exist, but many of the data remain inaccessible.For 3Centrefor Preparationand Implementationof International example,data are often storedin notebooksinstead of digital Projectson TechnicalAssistance, Rostov-on-Don, Russia. form, and thesenotebooks are not necessarilycentrally located 4Laboratoriede G6ologieAppliqu6e, CNRS, Paris. but insteadmay residein regionallaboratories. SComplexSystems Research Center, University of NewHampshire, Durham. Owing to these complications,use of the Russiannutrient data has been limited, and their fate has been uncertain be- Copyright2000 by the American GeophysicalUnion. causerecent economicand political instabilityin Russia has Paper number2000WR900099. lead to closureof laboratoriesand the potentialloss of data. In 0043-1397/00/2000WR900099 $09.00 order to help preservethe data set of the FSU and to estimate

23O9 2310 HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

Arctic Ocean East Siberian Sweden Laptev•• Sea Barents.• Sea Se•a• KaraSeaTaymyr Peninsula

Sea of Okhotsk

China Kazakstan Mongoli___•Japan Figure 1. Map of Eurasia showingrivers in the United Federal Servicefor Observationand Control of EnvironmentalPollution (OGSNK/GSN) data set and their approximatewatershed boundaries. Discharge stationsare shownby opencircles, and nutrientstations are givenby crosses.Consult Table 1 for stationnames and coordinates.As is apparent, rivers in the OGSNK/GSN data set encompassmost of the Eurasian watersheddraining into the Arctic Ocean,with notableexceptions in the RussianFar East and on the Taymyr Peninsula.River codesare asfollows: 1, Onega;2, SevernayaDvina; 3, Mezen'; 4, Pechora;5, Ob'; 6, Nadym; 7, Pur; 8, Taz; 9, Yenisey;10, Anabar; 11, Olenek; 12, Lena; 13, Yana; 14, Indigirka; and 15, Kolyma. nutrient flux to the Arctic Ocean from Russian rivers, we have whereas rivers in eastern Siberia typically have about 50 compiled and digitizedthe data archivesof the FSU and the monthswith data entries.Sample concentrations reported as current RussianFederation for 15 Russianrivers entering the below detectionlimit (BDL) were treated as zero, whichmay Arctic Ocean.The data comefrom samplesthat were collected lead to a slight underestimationof nutrient flux. and analyzedas part of the Unified Federal Servicefor Ob- Mean monthlyriver discharge(Figure 2), obtainedfrom the servationand Control of EnvironmentalPollution (OGSNK R-ArcticNet database (www.R-arcticnet.sr.unh.edu)(R. B. prior to 1992 and GSN from 1992 onward). We will refer to Lammers et al., An assessmentof the contemporarygauged these data as the OGSNK/GSN data set. river dischargeand runoff in the Pan-Arctic region, submitted The 15 river basinsrepresented in the OGSNK/GSN data to Journalof GeophysicalResearch, 2000) (hereinafterreferred set nearlyspan the >5000 km width of Russia(Figure 1) and to as Lammerset al., submittedmanuscript, 2000), is usedto include three of the world's 13 largest rivers by discharge compute monthly nutrient fluxes.In many but not all cases, [Shiklomanov, 1993]. Watershed areas range from -50- R-ArcticNet dischargedata are availablefor the nutrient sam- 3000 X 103 km2 (Table1). The nutrientdata set consists of pling stations(Table 1). When dischargedata are not available time'series of ammonium,nitrate, and phosphateconcentra- at the nutrient station,data from the closestdischarge station tions for the 15 Russianrivers, with samplesgenerally being are used.Annual river dischargeat the R-ArcticNet stations, collected at the downstreammost station that was free of tidal for the rivers representedin the OGSNK/GSN nutrient data influence.The periodsof recordfor nutrientconcentrations in set,ranges from 13.3to 577.3km3/yr (Table 1). individualrivers typically are 10-20 yearswithin the mid-1970s to the mid-1990speriod, with rivers in the Siberian far east 3. Data havingthe shortestrecords. Data were compiledon a monthly basis,but not all months in all years were represented,and 3.1. Mean Monthly Nutrient Concentrations means for individual months might representone to several and Fluxes samples.More than 100 monthsof data are availablefor sev- Mean monthlynitrate and ammoniumconcentrations varied eral of the rivers, particularly those around the Ob' Estuary, temporallywithin individualrivers as well as spatiallyacross HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN 2311

Table 1. Locationsof OGSNK/GSN Nutrient SamplingStations and R-ArcticNet DischargeStations and Their Respective Periods of Record

Nutrient Discharge Drainage Mean Annual Q at OGSNK/GSN Period of R-ArcticNet Period of Area,a DischargeStation, River NutrientStation Record DischargeStation Record 103km 2 km3/yr

Yeniseyb Dudinka (69.2øN,86.1øE) 1985-1995 Igarka (67.4øN,86.5øE) 1936-1995 2440 577.3 Lena Kyusyur(70.7øN, 127.4øE) 1984-1995 Kyusyur(70.7øN, 127.4øE) 1936-1994 2430 532.5 Ob' Salekhard(66.6øN, 66.6øE) 1986-1995 Salekhard(66.6øN, 66.6øE) 1936-1994 2950 404.1 Pechora Oksino (67.6øN,52.2øE) 1979-1995 Oksino (67.6øN,52.2øE) 1916-1993 312 135.1 Severnaya Arkhangel'sk(64.3øN, 40.3øE) 1976-1995 Ust' Pinega(64.1øN, 41.9øE) 1881-1993 348 105.6 Dvina Kolyma Cherskiy(68.4øN, 161.2øE) 1984-1994 Kolymskoye(68.7øN, 158.7øE) 1936-1988 526 70.8 Indigirka Chokurdakh(70.4øN, 147.6øE) 1984-1995 Vorontsovo(69.6øN, 147.5øE) 1936-1994 305 50.4 Taz Tazovskiy(67.3øN, 78.4øE) 1975-1995 Sidorovsk(66.6øN, 82.3øE) 1962-1994 100 33.1 Yana Yubileynaya(70.8øN, 136.0øE) 1984-1995 Yubileynaya(70.8øN, 136.0øE) 1972-1994 224 32.2 Olenek Taymylyr(71.6øN, 123.3øE) 1984-1995 (71.8øN,123.6øE) 1965-1985 198 31.5 Pur Samburg(67.0øN, 78.2øE) 1975-1992 Samburg(67.0øN, 78.2øE) 1961-1985 95 28.3 mezen' Dorogorskoye(64.4øN, 44.3øE) 1978-1994 Malonisogorskaya(65.0øN, 45.6øE) 1920-1988 56 20.4 Onega Porog (63.8øN,38.5øE) 1977-1995 Porog (63.8øN,38.5øE) 1943-1993 56 15.7 Nadym Nadym (65.6øN,72.7øE) 1978-1995 Nadym (65.6øN,72.7øE) 1955-1990 48 14.6 Anabar Saskylakh(72.0øN, 114.1øE) 1975-1995 Saskylakh(72.0øN, 114.1øE) 1954-1988 79 13.3

Annual dischargeestimates are calculatedby summingmonthly dischargedata, as given in the R-ArcticNet database.The rivers are ranked by discharge.The accessaddress for R-ArcticNet is www.R-arcticnet.sr.unh.edu (R. B. Lammers,submitted manuscript, 2000). aDrainageareas represent watershed areas above R-ArcticNet dischargestations, as reported in the R-ArcticNet database. bNumerousEnglish spellings exist for manyRussian place names. For consistency,we havefollowed the spellingsgiven by the Geographic NamesInformation System(GNIS) of the United Statesgovernment. One name (Yubileynaya)could not be verifiedby the GNIS.

the RussianArctic (Figure3). Highestconcentrations for both cent months,since exclusion of these high-dischargemonths ammoniumand nitrate often occurredin spring,whereas low- from our annual estimateswould lead to significantunderes- est concentrationswere frequently observedduring summer. timation of annual nutrient flux. Surprisinglyhigh ammoniumconcentrations were measuredin A strikingfeature of the annual flux estimatesis the high the riversentering the Ob' Estuary(Ob', Nadym,Pur, and Taz ammoniumflux in the Ob' and Yenisey rivers. Although the rivers),although nitrate values for theserivers were moremod- Amazon is more than 10 timesbigger (by discharge),ammo- erate. Maximum mean monthly ammonium concentration niumflux in the Ob' river is morethan twice as great (Table 2). reached almost 3 mg N/L in the Pur River, whereas nitrate In fact, if the data are correct,it seemslikely that the Ob' River seldomexceeded 0.5 mg N/L in any of the rivers.Nitrite con- transportsmore ammoniumthan any other river on Earth. In centrationswere very low, and the nitrate valuesreported in contrastto ammonium,nitrate fluxesin the Ob', Yenisey,and the OGSNK/GSN data set representnitrate alone, not nitrate other Russianrivers are much smallerthan other large rivers plus nitrite. suchas the Amazon and Mississippi(Table 2). Nitrate and ammoniumfluxes were generallyhighest in May When standardizedby catchment area, ammonium and or June (Figure 4), correspondingto the period of highest phosphateflux ratesappear to be correlatedand are greatestin discharge(Figure 2). For example,approximately 60% of the central Siberianrivers (Figures7a and 7c), whereasspecific annualammonium flux in the YeniseyRiver occursin a single nitrate flux is generallyhigher in the western Russianrivers month (June). Phosphateconcentrations often showedless (Figure 7b). Although it has been estimatedthat on average clear seasonal trends than did nitrate and ammonium but, as 85% of dissolvedinorganic nitrogen (DIN) transportin rivers with ammonium,tended to be highestin rivers entering the is as nitrate [Meybeck,1982], the OGSNK/GSN data set sug- Ob' Estuary(Figure 5). Phosphateflux in all of the riversin our geststhat there is a strong spatial componentto the ammo- data set was stronglyregulated by discharge,since discharge nium to nitrateflux ratio in RussianArctic rivers(Figure 8a). varied muchmore annuallythan did phosphateconcentration All of the riversin the OGSNK/GSN data set transportmore (Figure 6). ammoniumthan nitrate, and those drainingthe Siberianlow- land regiontransport more than 10 times as much ammonium 3.2. Annual Flux Estimates, Flux Ratios, as nitrate annually.In contrast,there is no clear spatialpattern and Specific Fluxes of DIN to phosphateflux (Figure 8b). Annual nutrient fluxeswere calculatedby summingflux estimates determined for individual months.In general, fluxes 4. Discussion were only calculatedfor months when both dischargeand nutrient data were available. Since rivers in eastern Siberia Our objectivein this paper is to determinewhether we can often are missingnutrient data for winter months(Figures 3 accuratelydetermine the contemporarydissolved inorganic ni- and 5), this protocolwill tend to underestimatetheir annual trogen and phosphateflux from Russianrivers to the Arctic nutrient fluxes,but the magnitudeof error is probablysmall Ocean and thus whether we will be able to detect changesin giventheir very low dischargeduring winter (Figure 2). In a the future. To this end, we have assembled an extensive data few cases,nutrient data were missingfor a springor summer set from the archives of the FSU and current Russian Feder- monthwhen dischargewas substantial.In thesecases, nutrient ation,which, if reliable,will be sufficientfor this purpose.Thus concentrationswere estimatedby interpolationbetween adja- our next taskis to evaluatethe reliabilityof thesedata. We will 2312 HOLMES ET AL.' NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

2500 16oo 3OOO

Onega 2500 • 2000 • Sever.naya 1200 E 2000 1500 8oo 1500 O3 1000 1000 4oo (/) 500 5O0

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 18000 40000 2000 Ob' • 15000 1600 E 30000 ,•- 12000 1200 03 9000 20000 8OO r- 6000 10000 .•_ 4OO I• 3000 o 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

50OO 5OOO 100000 Pur Yenisey • 4000 4OOO 80000 E 3000 3000 60000

O3 2000 2000 40000

(/) 1000 1000 2000O

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

3500 8000 10000

3000 8000 2500 6000 2000 4000 6000 1500 4000

1000500 2000 2000

0•0 0 0• Feb Mar Feb Mar

5000 6000 12000

4000 4000 100008000 3000 2000 6000 2000 4000 1000 2000 0 • • .... 0 ...... 0 Jan Feb Mar •r May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 2. Mean monthlydischarge for the 15 rivers in the OGSNK/GSN nutrient data set. Dischargedata are from the R-ArcticNET database(www. R-arcticnet.sr.unh.edu) (Lammers et al., submittedmanuscript, 2000). Error bars(many smaller than symbolsize) represent standard errors. In Figures2-6, riversare arrayed from westernmost(top left) to easternmost(bottom right).

do this by criticallyexamining trends in the data, by comparing not (Table 2). In fact, severalRussian Arctic riversapparently the OGSNK/GSN data set to other nutrient data that are avail- transportmore than 10 times as much ammoniumas nitrate, ablefor Russianrivers, and by comparisonto modelpredictions. comparedto rivers such as the Mississippiand Amazon that

4.1. Are the Data Reasonable? transportmuch more nitrate than ammonium.Exceptions to the generalrule are typicallyheavily polluted rivers, but the The OGSNK/GSN data set contains some unusual features rivers in our data set have relatively low human population and thuswarrants close scrutiny. First, althoughthe generally densities,fertilizer use, and atmosphericN depositionrates. acceptedparadigm is that rivers transport more nitrate than Thus the ammoniumto nitrate flux ratio calculatedusing the ammonium, all of the rivers in the OGSNK/GSN data set do OGSNK/GSN data set (Figure 7a) is anomalous. HOLMES ET AL.' NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN 2313

0.3 0.6 ! Severnaya 0.5 •, Onega • Dvina 0.2 0.4

0.3

0.1 0.2

0.1 NH4 • 0.0 0.0 I Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2.0 a Ob' Nadym 1.5

1.0

0.1 0.5

0.0 ...... t•-4•.a•1••••.• 0.0 , , , Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2.0

Pur 0.60 isey 1.5 2 0.45

1.0 0.30 1 0.5 0.15

0-•-• 0.0 0.00 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0.6 ar O!enek Lena

0.4

0.2

0.0 , , Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec Jan Feb Mar Apr May Jun Jul Aug $ep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0.8 0.5 0.6 • Yana Indigirka Kolyma z 0.4 c• 0.6 0.4 0.3 o 0.4 0.2 0.2

0.1

0.0 -, ß, 0.0 0.0 , , Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 3. Mean monthlyammonium (circles) and nitrate (triangles)concentrations for the 15 riversin the OGSNK/GSN data set.

Second, not only are the ammonium to nitrate flux ratios Third, althoughthe Sovietwater qualitymonitoring program suspicious,but the reported ammoniumconcentrations are was extensiveand centrallyorganized, it has been noted that also high (Figure 3). Whereas the estimatedglobal average instrumentation,materials, and supplieswere often of ques- river DIN concentrationis about0.12 mg/L (only 15% of which tionablereliability and that qualityassurance and control (QA/ is ammonium)[Meybeck, 1982], many of the riversin our data QC) procedureswere poorly executed[Tsirkunov, 1998]. For set have ammonium concentrationsthat greatly exceed this example,a studyof the distributionof organochlorineinsecti- mean DIN value. Average ammonium concentrationsin the cidesin Russianrivers indicated widespread discrepancies be- Ob', Nadym, Pur, and Taz rivers often exceed 1 mg N/L, in tween OGSNK/GSN laboratoriesand data collectedby inde- contrastto even heavily polluted rivers elsewheresuch as the pendentspecialists [Zhulidov et al., 1998].The explanationfor Thames and Rhine rivers where mean ammonium concentrations the differences was not clear, but it was noted that lack of are lessthan 1 mg N/L (Meybeckand Ragu,book draft 1995). supplies,equipment issues, and inexperiencedpersonnel may 2314 HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

8OO ooo lOOO Mezen' ooo • Sever.naya 8oo c- 600 ooo z NH4AOnega ooo 6oo ,,,,, 400 ooo 4oo x ooo -- 200 2oo ooo

o o o Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 10000 10000 ooo

ooo c- 8ooo 8000 Nadym z ooo O3 6000 6000 ooo X 4000 4000 ooo

2000 2000 ooo

o Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

12000 8oo0 00000

75000 c- 10o0o 5000 sey 50000 z 8000 2000 25000

v 6000 9000 00000 x 75000 ::3 4000 6000 50000 2000 3000 25000

o 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2000 3000 oooo 1750 r ek 2500 Lena oooo

2000 • 15001250 oooo • 1000 1500 oooo •X 750 1000 oooo • 500 500 250

0 , 0 ..... o , Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

5000 3000 40O Yana Indigirka a c- 4000 2500 30O z 2000 03 3000 1500 200 X 2000 1000 100 1000 500

, , 0 .... , Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 4. Mean monthlyammonium (circles) and nitrate (triangles)fluxes for the 15 riversin the OGSNK/ GSN data set.

have contributed.Similarly, an analysisof OGSNK/GSN data at., submittedmanuscript, 2000). Moreover,while all regional for the lower Don River found that a high percentage of laboratoriestheoretically used identical analyticalmethods, it nutrient data were unreliable, and the authorsspeculated that now seemsthat in reality a variety of methodswere usedwith analyticalproblems may have been widespreadin the federal little intercalibration(Zhutidov et at., submittedmanuscript, monitoring system [Boeva et al., 1999]. Although official 2000). At this point it is difficultor impossibleto reconstruct QA/QC proceduresfor the OGSNK/GSN monitoringsystem detailsof past QA/QC procedures,but it is clear that signifi- may have been adequate,it appearsthat in practiceminimal cantproblems existed. This lackof QA/QC informationcreates QA/QC actuallyoccurred (A. V. Zhutidov et at., The State obviousproblems for evaluatingdata reliability, since well- Service of Observation and Control of Environmental Pollu- documentedQA/QC proceduresshould be consideredvital to tion (OGSNK) in the former SovietUnion: A concisecritical anywater qualitymonitoring program. analysis, submitted to Canadian Journal of Fisheries and Althoughwe havejust outlined severalreasons to be suspi- AquaticSciences, 2000) (hereinafterreferred to as Zhutidovet ciousof the reliability of the OGSNK/GSN data set, there are HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN 2315

0.024 0.07 Onega 0.06 Mezen' 0.020 T Sever.naya 0.05 0.016 0.04 0.012 0.03 0.008 0.02

0.004 0.01

0.000 0.00 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec 0.06 0.5 Ob' 0.05 Nadym 0.4 0.04 0.3 0.03 0.2 0.02

0.01 0.1

0.00 0.0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec

0.5 0.3 0.025 Pur 0.020 Yenisey

0.2 0.3 0.015

0.2 0.010 0.1

0.1 0.005

0.0 0.0 0.000 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec

• 0.010 0.03 0.0 Anabar Olenek Lena • 0.008 0.0 E 0.02 ,-- 0.006 0.0 o

0• 0.004 0.0 0.01

o 0.002 0.0 o O 0.000 0.00 0.0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0.025 0.025 Yana 0.03 Indigirka Kolyma 0.020 0.020

0.015 0.02 0.015

0.010 0.010 0.01 0.005 0.005

0.000 0.00 0.000 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep OCt Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 5. Mean monthlyphosphate concentration for the 15 riversin the OGSNK/GSN data set. intriguingpatterns in the data (both seasonallyand spatially) and other spatialpatterns, there are alsoclear seasonaltrends that might not be expectedfrom erroneousdata. For example, in the data. For example, nutrient concentrationsare fre- specificfluxes of ammoniumand phosphate peak in the vicinity quentlyhighest during the springrunoff period and are lowest of the Ob' Estuary and decreaseto the east and west,whereas during summerlow-flow conditions.It is difficult to imagine specificnitrate flux generallyincreases to the west (Figure 7). how poor quality data could exhibit such clear seasonaland Thesetrends combine to give a distinctpattern of ammonium spatialpatterns. to nitrate flux that is highestin the watershedsof the western As we have noted, a particularlysurprising aspect of the Siberianlowlands (Figure 8). The riverswhich are most sus- OGSNK/GSN data set is the high ammoniumconcentration pect with respectto ammoniumconcentrations (Ob', Nadym, reported for severalof the RussianArctic rivers.In order to Pur, and Taz) are all in the westernSiberian lowlands, which accumulateammonium in rivers,nitrification (the microbial have low-lying,marshy soils that might through natural pro- conversionof ammoniumto nitrate) mustbe blockedor satu- cesseslead to the patternsthat we have observed[Neischtadt, rated, or at least nitrification and ammoniumuptake must 1971;Stairnov, 1994; Zhulidov et al., 1997].In additionto these proceedmore slowlythat ammoniumproduction. This is rarely 2316 HOLMES ET AL.- NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

70 8OO 2OO 60 • Seve.rnaya 6OO 150

4OO 100 • 30 x • 20 2OO 5O lO

o 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 15oo 8000 8OO ! , 1200 6000 6OO • 900 4000 4OO v • 600 •-. 300 2000 2OO

0 0 0 Jan Feb Mar Apr May Jun Jul Aug $ep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

1600 150 5OO Taz ey 120 4OO c.- 1200

90 3OO ¾ 800 v 60 2OO

-- 400 30 100

, , , Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

24 2400 r Olenek • 20 2000 a

Q_ 16 1600

..• 12 1200 v

x 8 8O0

u_ 4 400

...... ,•, , •, ...... 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 250 2OO

Re me 200 Indigirka 150

150 IO0 100

5O 50

, , 0 , , , ,• ...... Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 6. Mean monthlyphosphate flux for the 15 rivers in the OGSNK/GSN data set. observedin nature, particularlyin relativelypristine ecosys- hypoxiawere a commonfeature of thesesystems, nitrification tems.However, some possibilities for blockingnitrification ex- would be blocked and ammonium could accumulate to the ist, includinglimitation by dissolvedoxygen availability, limi- levels we have reported. More data will be needed to ade- tation by cold temperatures,and inhibitionby specificorganic quatelytest this hypothesis. compounds[Focht and Verstraete,1977; Sprent, 1987; Dyreborg andArvin, 1995].Interestingly, the Ob' River hashigh levelsof 4.2. Comparison With Other Estimates ferrous iron, and its oxidation periodically causeshypoxia The best method for testingthe reliability of the OGSNK/ [Telanget al., 1991].Although data are limited, it seemslikely GSN data set is to compareit to independentnutrient concen- that otherrivers in the centralSiberian lowlands (including the tration and flux estimates.Unfortunately, many of the publi- Pur, Taz, and Nadym)may also have elevated iron levels,given cationsthat at first appear promising[Tarasov et al., 1988; their similar catchment characteristics.Moreover, high dis- Smirnov,1994; Gordeevet al., 1996; Tsirkunovet al., 1998] solvedorganic matter concentrationsin theserivers might also actuallyuse values from the officialstate data (i.e., part of the contribute to oxygenconsumption and hypoxia. If, in fact, OGSNK/GSN data set) to derive their estimates.Similarly, HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN 2317

Table 2. Annual Nutrient Flux Estimates Derived From not universal.For example,data set II givesa much higher the OGSNK/GSN Nutrient Data Set and the R-ArcticNet DIN flux estimate for the Lena River than does the OGSNK/ DischargeData, Ordered by DecreasingAmmonium Flux GSN data set.In thiscase the major discrepancyis not with the ammoniumflux estimate (39.4 versus 40.4 x 103t/yr, OGSNK/ Flux to Ocean, 1000 t/yr GSN data setversus data set II, respectively)but insteadwith River NH4-N NO3-N PO4-P thenitrate flux estimate (19.5 versus 137.8 x 103t/yr, OGSNK/ GSN data setversus data set II, respectively).Therefore there Ob' 287.4 34.8 23.5 may not only be a problem with ammonium but also with Yenisey 207.8 18.4 6.2 Lena 39.4 19.5 3.5 nitrate in at least some rivers. Taz 30.5 0.75 2.8 If Russianrivers behaved as would be expectedbased on a Pur 24.3 0.74 3.0 recentlypublished model [Seitzingerand Kroeze,1998; Caraco Pechora 17.8 7.1 4.2 and Cole, 1999],they would be transportingless DIN than the SevernayaDvina 14.9 6.7 2.0 Nadym 12.2 0.55 2.0 OGSNK/GSN data suggest(Table 3). We shouldnote that the Yana 6.8 1.2 0.36 model is not specificfor Arctic rivers but, instead,was cali- Kolyma 5.2 2.5 0.76 brated largelyusing data from temperateand tropical rivers. Indigirka 3.8 2.3 0.35 Overall, the OGSNK/GSN data give an estimate of annual Olenek 2.7 0.78 0.23 Mezen' 2.2 0.71 0.44 DIN flux 2.3 timesgreater than the model estimatefor the nine Anabar 1.9 0.09 0.03 rivers in common. However, for some rivers such as the Pur Onega 1.6 0.99 0.15 and Taz Rivers, the estimatesdiffer much more, by about an

Amazon 131.8 1021.5 160.7 order of magnitude. In some casesthe model estimatesare Mississippi 21.2 740.6 71.4 closerto the valuesreported in data set II, but in other cases Yukon 9.1 21.7 2.1 they are not (e.g., nitrate in the Lena River). Mackenzie ... 23.6 1.5 The Lena River in particularhas receivedconsiderable in- ternationalattention during the 1990s,largely becauseof in- Note that for somerivers, monthly discharge data were not available at the OGSNK/GSN nutrient station, so the closest R-ArcticNet dis- terest in the role of freshwaterinput on sea-iceformation in charge station was used to compute fluxes. Flux estimatesfor the the LaptevSea (into whichthe Lena flows)and its impacton Amazon, Mississippi,Yukon, and Mackenzie riverswere made using global climate [Aagaardand Carmack, 1989; Cauwetand Si- data contained in the GEMS-GLORI database[Meybeck and Ragu, dorov, 1996; Saliot et al., 1996; Kassenset al., 1998; Lara et al., 1997]. Ellipsisindicates data are not availablein GEMS-GLORI database. 1998;RachoM and Hubberten,1998]. Lara et al. [1998] report resultsfrom a river expeditionin July 1994, where nutrient concentrationsat Kyusyurnear the mouth of the Lena River were 0.0077 and 0.0014 mg/L for NO3-N and NH4-N , respec- large United Nations-sponsoreddatabases, GEMS/Global tively.Comparable numbers (same month and year) from the Registerof River Inputs (GLORI) (Meybeckand Ragu, book OGSNK/GSN data set are 0.02 mg/L NO3-N and 0.01 mg/L draft, 1995)and GEMS/Water (www.cciw.ca/gems/),also sum- NH4-N , whereasdata set II lists 0.28 mg/L NO3-N and 0.04 marize parts of the official Russian data. Although the mg/L NH4-N. Thus nitrate and ammoniumconcentration mea- OGSNK/GSN data set we presentis far more extensivethan surementsfor the Lena River in July 1994 vary by more than these other reports, the data sets ultimately come from the 2500%. At this point it is not possibleto determine which samesource and thereforecannot justifiably be usedfor crit- valuesare more reliable,although those reported by Lara et al. ical comparisons. [1998] are closer to what might be expectedfor a relatively Fortunately, other sourcesof information exist. The most pristineArctic watershed. extensiveindependent data set comesfrom Russianscientists working outsideof the OGSNK/GSN framework.Their results havebeen compiledfor a 10-yearperiod (1986-1995) for 10 Arctic rivers, all of which are included in the 15-river OGSNK/ 5. Conclusions GSN data set that we have alreadypresented. From this data Nutrient flux from land to ocean integrateschanges in ter- set(which we shallrefer to asdata set II) we calculatedannual restrial ecosystems,in land use, and in other human activities. DIN (nitrate plus ammonium)flux, usingdischarge from the As globalchange due to greenhousewarming and humanpop- R-ArcticNet database,and comparedit to the OGSNK/GSN ulation growth accelerates,a record of water quality is one data and to model estimatesderived from Seitzingerand Kroeze metric of that change. On the ocean side the coastal deltas, [1998] (Table 3). estuaries,and seasrespond to changingnutrient fluxeswith DIN flux estimatesvary greatly dependingon the data set changesin the intensityand distributionof primaryproductiv- used(Table 3). For example,annual DIN flux estimatesfor the ity, which in turn impact coastalfisheries. Thus monitoringof Ob' Riverrange from less than 50 x 103 t N/yr usingdata set nutrient fluxescan provideessential information for watershed II to greaterthan 300 x 103 t/yrusing the OGSNK/GSNdata management,for coastalfisheries management, and for detec- set, with the model estimate of Seitzingerand Kroeze [1998] tion of regional aspectsof global changein the Arctic. being intermediate.Moreover, it is not simply a systematic In the title of the paper we ask whether it is currently offset between the two data sets; in the case of ammonium in possibleto establisha reliablebaseline against which to judge the Ob' River, there appearsto be almost no relationship future changesin nutrient export to the Arctic Ocean from betweenthe two data sets(Figure 9). Obviously,at least one, Russia. It is not. if not both, of the data setsis grosslyin error. The most extensive data set available, the OGSNK/GSN Although the OGSNK/GSN data set generally yields a data set of the FSU and current Russian Federation, is of higherDIN flux estimatethan doesdata set II, the pattern is questionablereliability. It is possiblethat the data are accurate, 2318 HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

- euJ,•lO}l

- e>lJ!6!pul

- eueA

- eue-I

- >leUelO - Jeqeuv e/ •es!ueA .• o Zel J

Jnd

- w•peN

- ,qo

eJoqoed

,uezelAI

eU!AO 'S

- e6euO

[ t i ] [ o o o o o o o • o • o •

(JI•IOW)0!•.1• xnl=l •ON 0• •FIN I•nuuv (Jl•lOW)o!•i xnl=l•Ocl o• NIO I•nuuv

i t i i i , , i i [ 0 0 0 0 0 0 0 0 • 0 • 0 • 0 0 • 0 • 0 • 0 • 0 • 0 14) 0 • 0 •

(•/•LWl/Nõ•1) xnl_-I uJn!uouJuJV (•/•LWl/Nõ•1)xnl_-I el"J•!N (•/•LU•l/dõ•1)xnl_-I el.LldsOLId HOLMES ET AL.' NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN 2319

Table 3. Comparisonof Annual DIN (Nitrate Plus nutrient flux to the Arctic Ocean from these Russian rivers. Ammonium) Flux Estimatesfor SelectedRussian Arctic Perhapsmore exciting,we will have identifiedan unusualbio- Rivers geochemical phenomenon (high ammonium export from Nitrate Plus Ammonium Flux, 1000 t N/yr sparselyinhabited catchments) that we currentlydo not under- stand.Although further researchwould be needed to under- River OGSNK/GSN Data Set II Model a stand the observedpatterns, at least the existingdata would

Ob' 322.3 45.4 161.3 allow us to detect future changes. Yenisey 226.2 134.9 73.98 If, however, the nutrient concentration data are not reliable, Lena 58.9 169.5 33.62 then the flux estimatesthat we and othershave providedare Taz 31.3 4.87b wrong. This will lead to incorrectconclusions concerning fu- Put 25.1 4.87 b Pechora 25.0 15.4 10.24 ture changesand a faultyunderstanding of the currentbiogeo- SevernayaDvina 21.6 7.5 chemicalfunctioning of these catchments. Yana 8.0 2.2 1.15 As is the casefor hydrologicand sedimentflux modeling, Kolyma 7.6 6.1 4.64 one of the major challengesfor the constructionand calibra- Indigirka 6.0 2.8 1.95 Olenek 3.4 1.3 1.72 tion of large-scalebiogeochemistry models is data availability Onega 2.6 2.70 [Milliman and Syvitski,1992; VOrOsmartyet al., 1996, 2000]. Anabar 2.0 0.5 0.72 Thus every effort must be made to augment the available Total c 659.4 378.1 289.3 discharge,suspended sediment, and nutrient chemistrydata- R-ArcticNet monthly dischargedata (Lammers et al., submitted bases,both regionallyin the Arctic aswell asglobally, either by manuscript,2000) were usedto computefluxes for the OGSNK/GSN collectingnew dataor makingavailable previously collected data. and data set II, whereasthe model estimatesused annual discharge estimates. In augmentingnutrient databases, however, data qualitymust be aFromSeitzinger and Kroeze[1998], based on model of Caracoand closelyscrutinized. For RussianArctic rivers it is unclearwhether Cole [1999].Estimates for most of the individualRussian rivers were any of the currentlyavailable long-term data setsare reliable. not reported in the Seitzinger and Kroeze paper but were kindly The clarificationof this puzzle requiresindependently col- providedto us by the authors. bThemodel estimate (4.87 x 103 tons N/yr) is for thesum of thePur lected and analyzednutrient samples.Only in this way will we and Taz rivers. be able to fully assessthe qualityof current data setsand state CThe total is for rivers common to both data sets and the model. with any confidencethe magnitudeof contemporarynutrient flux to the Arctic Ocean.Until new samplesare collectedand a contemporarynutrient flux baselineis established,we will be but other data setsoften give far different values,and we are squanderingone of our better chancesfor early detectionof not able to determine which of the available data sets are most globalchange in the Arctic, and our understandingof Arctic- reliable. If the OGSNK/GSN data set that we have presented wide biogeochemicalcycling and land-oceaninteractions in the can be proven reliable, then we have accuratelyquantified Arctic will remain uncertain.

0.20 •I Jan y-int = 0.022 O Feb slope = 0.029 ß Mar r2=0.13 V Apr 0.15 ß May I--I Jun •I* Jul • Aug & Sep /• Oct 0.10 I•I Nov O Dec

0.05

0.00 i 0.0 0.5 1.0 1.5 2.0 2.5 OGSNK/GSN Ammonium(mg N/L) Figure 9. Comparisonof ammoniumconcentration estimates for the Ob' River (1988-1994) using the OGSNK/GSN data set and data set II. 2320 HOLMES ET AL.: NUTRIENT FLUX FROM RUSSIAN RIVERS TO THE ARCTIC OCEAN

Acknowledgments. This researchwas funded by the National Sci- Saliot, A., G. Cauwet, G. Cahet, D. Mazaudier, and R. Daumas, ence Foundation (NSF-OPP-9524740and NSF-OPP-9818199).We Microbial activitiesin the Lena River delta and Laptev Sea, Mar. gratefullyacknowledge David Clow, Bob Meade, and two anonymous Chem., 53, 247-254, 1996. reviewersfor commentson the manuscript.We also thank Rainer Seitzinger,S. P., and C. Kroeze, Global distributionof nitrous oxide Amon, Ludmila Boeva, Ed Landa, Joerg Lobbes, Gerhard Kattner, productionand N inputsin freshwaterand coastalmarine ecosys- Vladimir Khlobystov,Bob Meade, Richard Robarts, and Dmitry tems, Global Biogeochem.Cycles, 12, 93-113, 1998. Varlyguin for insightfuldiscussions and assistanceduring the prepa- Serreze, M. C., J. E. Walsh, F. S. Chapin, T. Osterkamp, ration of this manuscriptand Sybil Seitzingerfor sharingher model M. Dyurgerov,V. Romanovsky,W. C. Oechel,J. Morison,T. Zhang, results with us. and R. G. Barry, Observationalevidence of recent changein the northernhigh-latitude environment, Clim. Change,irt press,2000. Shiklomanov, I. A., World freshwater resources,in Water in Crisis:A References Guide to the World'sFresh Water Resources, edited by P. H. Gleick, Aagaard, K., and E. C. Carmack,The role of sea ice and other fresh pp. 13-24, Oxford Univ. Press,New York, 1993. Smirnov,M.P., Assessmentof the dischargeof nutrientsinto seasof waterin the arcticcirculation, J. Geophys.Res., 94, 14,485-14,498,1989. the Arctic and Pacificoceans and of the anthropogeniccomponent of Alekin, O. A., and L. V. Brazhnikova,Flux of DissolvedSubstances From the Territoryof the USSR, 144 pp., Nauka, Moscow, 1964. thisdischarge (in Russian),Hydrochem. Mater., 113, 121-147,1994. Anderson,L. G., K. Olsson,and M. Chierici, A carbonbudget for the Sprent,J. I., The Ecologyof the NitrogenCycle, 151 pp., Cambridge Univ. Press, New York, 1987. Arctic Ocean, Global Biogeochem.Cycles, 12, 455-465, 1998. Tarasov, M. N., M.P. Smirnov, I. A. Kruchkov, and G. I. Laki, The Anisimov, O. A., and F. E. Nelson, Permafrost distribution in the northern hemisphereunder scenariosof climate change,Global river dischargeof nutrientsin the USSR and its temporal changes Planet. Change,14, 59-72, 1996. (1936-1980)(in Russian),Hydrochem. Mater., 103, 49-66, 1988. Boeva, L. V., Y. Y. Vinnikov, A. V. Zhulidov, G. S. Volovik, V. V. Telang, S. A., R. Pocklington,A. S. Naidu, E. A. Romankevich,I. I. Khlobystov,and S. R. Brown, An assessmentof the quality of Gitelson,and M. I. Gladyshev,Carbon and mineral transportin major North American, Russianarctic, and Siberianrivers: The St. OGSNK/GSN data relatingto nutrient compounds,organochlorine Lawrence, the Mackenzie, the Yukon, the arctic Alaskan rivers, the pesticidesand biochemicaloxygen demand (BODs) in the water of the Lower Don, SIL News, 28, 4-5, 1999. arctic basin rivers in the Soviet Union, and the Yenisei, in Biogeo- Broecker, W. S., Thermohaline circulation, the Achilles heel of our chemistryof Major WorldRivers, edited by E. T. Degens,S. Kempe, and J. E. Richey, pp. 75-104, John Wiley, New York, 1991. climate system:Will man.made CO2 upset the current balance?, Science,278, 1582-1588, 1997. Tsirkunov,V. V., Water qualitymonitoring systems, inA WaterQuality Assessmentof the Former Soviet Union, edited by V. Kimstach, Caraco, N. F., and J. J. Cole, Human impact on nitrate export: An M. Meybeck, and E. Baroudy,pp. 95-112, Routledge,New York, analysisusing major world rivers,Ambio, 28, 167-170, 1999. 1998. Cauwet, G., and I. Sidorov,The biogeochemistryof Lena River: Or- Tsirkunov, V. V., M.P. Polkanov, and V. G. Drabkova, Natural com- ganic carbon and nutrientsdistribution, Mar. Chem., 53, 211-227, 1996. positionof surfacewater and groundwaters,in A Water Quality Assessmentof the Former SovietUnion, edited by V. Kimstach,M. Chapin,F. S., G. R. Shaver,A. E. Giblin, K. J. Nadelhoffer,and J. A. Meybeck, and E. Baroudy,pp. 25-68, Routledge,New York, 1998. Laundre, Responsesof arctictundra to experimentaland observed V6r6smarty,C. J., C. J. Willmott, B. J. Choudhury,A. L. Schloss,T. K. changesin climate,Ecology, 76, 694-711, 1995. Stearns,S. M. Robeson,and T. J. Dorman, Analyzingthe discharge Dyreborg, S., and E. Arvin, Inhibition of nitrificationby creosote- contaminated water, Water Res., 29, 1603-1606, 1995. regime of a large tropical river through remote sensing,ground- based climatic data, and modeling, Water Resour.Res., 32, 3137- Focht, D. D., and W. Verstraete, Biochemicalecology of nitrification 3150, 1996. and denitrification, Adv. Microbial Ecol., 1, 135-214, 1977. Gordeev, V. V., and V. V. Tsirkunov, River fluxes of dissolvedand V6r6smarty, C. J., B. Fekete, M. Meybeck, and R. B. Lammers,The suspendedsubstances, in A WaterQuality Assessment of the Former global systemof rivers:Its role in organizingcontinental land mass and definingland-to-ocean linkages, Global Biogeochem. Cycles, 14, SovietUnion, edited by V. Kimstach,M. Meybeck,and E. Baroudy, 599-621, 2000. pp. 311-350, Routledge,New York, 1998. Gordeev, V. V., J. M. Martin, I. S. Sidorov, and M. V. Sidorova, A Zhulidov, A. V., J. V. Headley, D. F. Pavlov, R. D. Robarts, L. G. Korotova, V. V. Fadeev, O. V. Zhulidova, Y. Volovik, and reassessmentof the Eurasian river input of water, sediment,major elements, and nutrients to the Arctic Ocean, Am. J. Sci., 296, 664- V. Khlobystov,Distribution of organochlorineinsecticides in rivers of the RussianFederation, J. Environ. Qual., 27, 1356-1366, 1998. 691, 1996. Hobble, J. E., B. L. Kwiatkowski, E. B. Rastetter, D. A. Walker, and Zhulidov, A. V., J. V. Headley, R. D. Robarts,A.M. Nikanorov, and A. A. Ischenko,Atlas of RussianWetlands: Biogeography and Metal R. B. McKane, Carboncycling in the Kuparukbasin: Plant produc- Concentrations,309 pp., Natl. Hydrol. Res. Inst., Saskatoon, tion, carbon storage,and sensitivityto future changes,J. Geophys. Saskatchewan,Canada, 1997. Res., 103, 29,065-29,073, 1998. Houghton, J. T., L. J. Meira Filho, B. A. Callander, N. Harris, A. Kattenberg,and K. Maskell,Climate Change 1995: The Science of V. V. Gordeev and P. P. ShirshovInstitute of Oceanology,Russian ClimateChange, 584 pp., CambridgeUniv. Press,New York, 1996. Academy of Sciences, Moscow 117218, Russia. (gordeev@geo. Kassens,H., I. Dmitrenko, V. Rachold, J. Thiede, and L. Tomokhov, sio.rssi.ru) R. M. Holmes and B. J. Peterson,The EcosystemsCenter, Marine Russianand German scientistsexplore the Arctic'sLaptev Sea and its climate, Eos Trans.AGU, 79, 317, 322-323, 1998. BiologicalLaboratory, Woods Hole, MA 02543. ([email protected]; Lara, R. J., V. Rachold, G. Kattner, H. W. Hubberten, G. Guggen- [email protected]) R. B. Lammers and C. J. V6r6smarty, Complex SystemsResearch berger,A. Skoog,and D. N. Thomas,Dissolved organic matter and nutrients in the Lena River, Siberian Arctic: Characteristics and Center, University of New Hampshire, Durham, NH 03824. distribution, Mar. Chem., 59, 301-309, 1998. ([email protected];[email protected]) M. Meybeck, Laboratorie de G6ologie Appliqu6e, CNRS, Place Meybeck, M., Carbon, nitrogen, and phosphorustransport by world rivers, Am. J. Sci., 282, 401-450, 1982. Jussieu,75257 Paris, France. ([email protected]) A. V. Zhulidov, Centre for Preparation and Implementation of Milliman, J. D., and J.P. M. Syvitski,Geomorphic/tectonic control of International Projects on Technical Assistance, Rostov-on-Don sedimentdischarge to the ocean:The importanceof small moun- tainous rivers, J. Geol., 100, 524-544, 1992. 344104,Russia. ([email protected]) Neischtadt,M. I., World naturalphenomenon (bogginess of the West SiberiaPlain), Izv. USSRAS, Ser.Geogr., 1, 21-34, 1971. Rachold, V., and H. W. Hubberten, Carbon isotopecomposition of particulateorganic material in eastSiberian rivers, in Land-Ocean Systemsin the SiberianArctic: Dynamics and History, edited by (ReceivedNovember 15, 1999;revised April 7, 2000; H. Kassenset al., pp. 223-238, Springer-Verlag,New York, 1998. acceptedApril 7, 2000.)