Direct evidence of the feedback between climate and ... · Direct evidence of the feedback between...

Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

ScienceDirect

Geochimica et Cosmochimica Acta 166 (2015) 249–266

Direct evidence of the feedback between climate andnutrient, major, and trace element transport to the oceans

Eydis Salome Eiriksdottir a,⇑, Sigurður Reynir Gislason a, Eric H. Oelkers a,b,c

a Institute of Earth Sciences, Sturlugata 7, 101 Reykjavık, Icelandb GET, CNRS/URM 5563 – Universite Paul Sabatier, 14 rue Edouard Belin, 31400 Toulouse, France

c Earth Sciences, University College London, Gower Street, London WC1E 6BT, United Kingdom

Received 23 December 2013; accepted in revised form 8 June 2015; Available online 15 June 2015

Abstract

Climate changes affect weathering, denudation and riverine runoff, and therefore elemental fluxes to the ocean. This studypresents the climate effect on annual fluxes of 28 dissolved elements, and organic and inorganic particulate fluxes, determinedover 26–42 year period in three glacial and three non-glacial river catchments located in Eastern Iceland. Annual riverinefluxes were determined by generating robust correlations between dissolved element concentrations measured from 1998 to2003 and suspended inorganic matter concentrations measured from 1962 to 2002 with instantaneous discharge measuredat the time of sampling in each of these rivers. These correlations were used together with measured average daily dischargeto compute daily elemental fluxes. Integration of these daily fluxes yielded the corresponding annual fluxes.

As the topography and lithology of the studied glacial and non-glacial river catchments are similar, we used the records ofaverage annual temperature and annual runoff to examine how these parameters and glacier melting influenced individualelement fluxes to the oceans. Significant variations were found between the individual elements. The dissolved fluxes of themore soluble elements, such as Mo, Sr, and Na are less affected by increasing temperature and runoff than the insoluble nutri-ents and trace elements including Fe, P, and Al. This variation between the elements tends to be more pronounced for theglacial compared to the non-glacial rivers. These observations are interpreted to stem from the stronger solubility controlon the concentrations of the insoluble elements such that they are less affected by dilution. The dilution of the soluble elementsby increasing discharge in the glacial rivers is enhanced by a relatively low amount of water–rock interaction; increased runoffdue to glacial melting tend to be collected rapidly into river channels limiting water–rock interaction. It was found that theclimate effect on particle transport from the glacial rivers is far higher than all other measured fluxes. This observation,together with the finding that the flux to the oceans of biolimiting elements such as P and Fe is dominated by particulates,suggests that particulate transport by melting glaciers have a relatively strong effect on the feedback between continentalweathering, atmospheric chemistry, and climate regulation over geologic time.� 2015 Elsevier Ltd. All rights reserved.

1. INTRODUCTION

This study is based on a set of fluid and suspended par-ticulate samples collected from three sampling sites in twoglacial rivers and three sampling sites in non-glacial rivers

http://dx.doi.org/10.1016/j.gca.2015.06.005

0016-7037/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: [email protected] (E.S. Eiriksdottir).

located in the basalts of Eastern Iceland. In total, 220 watersamples were collected from these rivers from 1998 to 2003.A total of 28 distinct dissolved element concentrations andthe suspended inorganic and organic material concentra-tion of these samples were measured, for a total of 6600 dis-tinct concentration data. Elemental concentration andsuspended inorganic material – discharge rating curves cre-ated for each river was used together with daily measuredriver discharge to calculate the daily elemental fluxes


mailto:[email protected]


http://crossmark.crossref.org/dialog/?doi=10.1016/j.gca.2015.06.005&domain=pdf

250 E.S. Eiriksdottir et al. / Geochimica et Cosmochimica Acta 166 (2015) 249–266

towards the oceans over a 26–42 year period. This extensivedataset was used in four previous papers to illuminate var-ious aspects of chemical transport to the oceans. Gislasonet al., 2006 calculated the daily dissolved and sedimentfluxes of Ca towards the ocean and concluded that the lat-ter flux dominated the feedback between climate and conti-nental weathering. Gislason et al. (2009) used this dataset tocalculate the annual fluxes of Ca and bicarbonate to theoceans and showed that these annual fluxes have increasedsubstantially in response to increased temperature over thepast 40 years. Eiriksdottir et al. (2008) used this dataset toassess the validity of steady-state weathering models.Eiriksdottir et al. (2013) used instantaneous dissolved Nafluxes from these rivers to determine the distinct effects oftemperature versus runoff on chemical weathering. Thisstudy aims to use this dataset to illuminate the distincteffects of (1) climate (temperature and runoff), and (2) gla-cial cover, on the riverine fluxes of the elements to theoceans, highlighting the distinct behavior of trace elementsand nutrients.

Riverine transport of elements to the oceans plays amajor role in moderating global climate on a geologicaltime scale. Increased flux of major elements and alkalinitydue to increasing temperature promotes drawdown ofCO2 from the atmosphere through the formation of car-bonate minerals in the oceans, providing a negative feed-back between climate and chemical weathering (Walkeret al., 1981; Amiotte-Suchet and Probst, 1993; White andBlum, 1995; Berner and Caldeira, 1997; Dessert et al.,2001, 2003; Gislason et al., 2006, 2009; Gaillardet andGaly, 2008; Ferrier et al., 2012; Eiriksdottir et al., 2013).An additional feedback between continental weathering,riverine transport, and climate is the transport and subse-quent burial of organic carbon (e.g. Berner, 1982, 2004;Berner and Raiswell, 1983; Arthur et al., 1988;France-Lanord and Derry, 1997; Burdige, 2005; Galyet al., 2007; Hilton et al., 2012; Smith, 2013). Critical to thisfeedback between organic material transport to the oceansand global climate is that it is buried prior to its degrada-tion (Raven and Falkowski, 1999; Berner, 2004).

Rivers transport a large quantity of essential nutrients tothe oceans. Of particular interest are the macronutrients N,P, and Si which may limit primary production in variousEarth surface environments (Egge and Aksnes, 1992;Jickells, 1998; Falkowski, 2004; Galloway, 2004;Ruttenberg, 2004; Olafsson et al., 2008; Statham, 2012;Hartmann et al., 2014), and the trace elements Fe, B, Mn,Zn, Cu, Ni, Mo and V which are needed either to buildorganic cells or to catalyze biochemical transformations(e.g. Martin et al., 1994; Falkowski, 1997; Cullen et al.,1999; White, 1999; Lane and Morel, 2000a,b; Morel andPrice, 2003; Raiswell and Canfield, 2012). Variations innutrient transport to the oceans could also provide a feed-back between weathering and climate if an increase in nutri-ents leads to increased primary productivity and organiccarbon burial (e,g, Falkowski et al., 1998; Bains et al.,2000; Jickells et al., 2005; Jeandel et al., 2011; Pearceet al., 2013).

Weathering has been observed to be greatly effected bythe presence of glaciers (e.g. Gislason et al., 1996; Hallet

et al., 1996; Tranter, 2003; Anderson, 2005, 2007;Gislason, 2008; Li et al., 2012; Opfergelt et al., 2013).Evidence shows that sediment yields increase with the tem-perate glacier coverage of catchments in Alaska (Halletet al., 1996) and in Iceland (Gislason et al., 1996). The aver-age mechanical denudation rate in Iceland is 650 t/km2/yr,excluding bedload and major floods (Tomasson, 1990); thisis equivalent to 0.24 mm/yr assuming rock density of2.7 g/cm3 and considerably higher than the global averagemechanical denudation rate of 230 t/km2/yr (Millimanand Syvitski, 1992) equal to 0.085 mm/yr. Mechanicaldenudation in the glaciated areas of Iceland can be twoorders of magnitude higher than in non-glaciated areas(Tomasson, 1990). This agrees with studies summarizedby Hallet et al. (1996) showing that river basins with greaterthan about 30% glacier cover yield about an order of mag-nitude more suspended sediments than glacier-free basins.In the extreme, mechanical erosion rates greater than27,000 t/km2/yr or 10 mm/yr have been measured in theglaciated catchments of southern Alaska (Hallet et al.,1996). The suspended material in glacial rivers inNE-Iceland are highly reactive basaltic glass (Eiriksdottiret al., 2008), and play a major role in ocean chemistry(Oelkers et al., 2012; Jones et al., 2014; Morin et al.,2015). The high adsorption potential of the sediment sur-faces contribute to ocean chemistry due to desorption inthe coastal waters (Ruttenberg, 2004; Sinkko et al., 2013).This desorption process is particularly significant for phos-phorus, an essential nutrient for primary productivity(Ruttenberg, 2004; Raiswell and Canfield, 2012). Glaciershave also been observed to increase chemical weatheringrates; the average chemical denudation rates in Iceland is35 t/km2/yr (Gislason, 2008) compared to the 24 t/km2/yron the continents (Gaillardet et al., 1999), which are dom-inated by carbonate weathering.

In this study, which builds upon that of Gislason et al.(2009), the annual elemental fluxes of suspended organicand inorganic material and 28 dissolved elements in the riv-ers are used to determine the distinct behavior of each inresponse to climate change and the presence of glaciers.Focus will be placed on the role of nutrients essential forthe primary productivity in the oceans including Si, P,and Fe. The purpose of this paper is to report the resultsof this study, which illuminates how climate change andglacial melting might affect the chemistry and primary pro-ductivity in the oceans.

2. STUDY AREA AND METHODS

Samples of water and suspended inorganic particulatematter were collected from six river catchments located inEastern Iceland (Fig. 1). Samples were collected 44 timesfrom each sampling site, from 1998 to 2003, exceptFjardara, which was sampled 20 times from 1998 to 2000,and Jokulsa a Dal at Bru, which was sampled 24 times from2000 to 2003. Discharge, air, and water temperature weremeasured at the time of sampling. Three of the studiedcatchments, Jokulsa a Dal at Bru, Jokulsa a Dal atHjardarhagi, and Jokulsa ı Fljotsdal at Holl are glaciated,and three are non-glaciated, the direct runoff rivers,

Fig. 1. A simplified geological map of the study area in NE Iceland (The Icelandic Institute of Natural History, 1977). Boundaries of rivercatchments are superimposed on the geological map (The Hydrological Services of the National Energy Authority, 2005a).

E.S. Eiriksdottir et al. / Geochimica et Cosmochimica Acta 166 (2015) 249–266 251

Fellsa, Grımsa, and Fjardara. Jokulsa a Dal at Bru is asub-catchment of Jokulsa a Dal at Hjardarhagi. TheHydrological Service of the National Energy Authorityhas continuously monitored the discharge of these riversand the suspended load of the glacial rivers over the pastthree to four decades (Tomasson, 1990; Palsson andVigfusson, 1996; Tomasson et al., 1996; Adalsteinsson,2000; The Hydrological Services of the National EnergyAuthority, 2005b; Gislason et al., 2006, 2009; Eiriksdottiret al., 2008). The concentration of suspended inorganicmatter measured in the glacial rivers for the years prior to1998 reported in the literature was used in this study,together with the data collected in 1998–2002, to build rat-ing curves for the suspended inorganic matter. The bedrockin these river catchments consists predominantly of basalticlavas and hyaloclastite (glassy basalt) with some acidicintrusions associated with extinct central volcanoes(Wood, 1976, 1978; Johannesson and S�mundsson, 1998;Hards et al., 2000; Eiriksdottir et al., 2008). The age ofthe bedrock increases from the west to the east and theaverage area-weighted age of the river catchments rangesfrom 1.3 to 11.2 Ma (Gislason et al., 2006; Kardjilov,2008; Table 1 and Fig. 1). Vegetation also increases fromwest to east as the bedrock becomes older and less perme-able. A large part of the study area is located in the high-lands and is poorly vegetated (Kardjilov et al., 2006;Kardjilov, 2008). Wetlands are sparsely present in theBru, Hjardarhagi, and Holl catchments (Kardjilov, 2008).The Grımsa catchment has the most abundant vegetation,with some forests. The Icelandic soil is rich in air borneglassy volcanic material and the main secondary minerals

present are ferrihydrate and allophane (Wada et al.,1992). Air temperature has been monitored in the studyarea for the past several decades (The IcelandicMeteorological Office, 2007; Gislason et al., 2009); therange of average annual temperature in each studies catch-ment is provided in Table 1.

The collection and analysis of samples in this study havebeen previously described by Gislason et al. (2009) andEiriksdottir et al. (2013). A total of 220 water samples werecollected. The samples were filtered on site through 0.2 lmMillipore cellulose acetate filters using a peristaltic pump,silicone tubing, and a 142 mm Sartorius filter holder. Thesamples were not ultra-filterated so the use of 0.2 lm filtersmay allow some colloidal sized particles to remain in thefluid phase after sampling and filtering. Water samples col-lected for major cations and trace element analyses wereacidified with suprapure HNO3. Water samples for majoranions were non-acidified. Total suspended inorganic andorganic matter (SIM and POC) samples were collected fromthe main channel of the rivers using either a DH48 or a S49sampler (Guy and Norman, 1970). The S49 sampler wasused in all rivers except Fjardara and Fellsa where theDH48 sampler was used.

Samples for pH measurement were collected in amberglass bottles, filling the bottles until it was overflowing, toavoid air in the samples. The samples were kept cold dur-ing until the pH was measured in the laboratory, using aCole-Palmer combined glass electrode together with anOrion pH meter. The electrodes were calibrated using4.01 and 7.00 NBS standards prior to all analyses; theuncertainty of the analyses is estimated to be ±0.02.

Tab

le1

Ph

ysic

alch

arac

teri

stic

so

fth

eri

ver

catc

hm

ents

stu

die

d(m

od

ified

afte

rG

isla

son

etal

.,20

09).

Sam

ple

site

Siz

ek

m2

Gla

cial

cove

r%

Ro

ckag

eaM

yrS

tud

yp

erio

db

yrR

un

off

ran

gec

m/y

rR

un

off

incr

ease

%M

ean

ann

ual

air

Tra

nge

�CR

elat

ive

Tch

ange

�C%

Ru

no

ffch

ange

pr.

�C

JDB

ru20

8968

1.32

1971

–200

41.

14–2

.47

117

�1.

3–3.

214.

5112

.2JD H

jard

arh

agi

3338

431.

7119

70–2

004

0.94

5–1.

7686

�1.

3–3.

214.

517.

8

JFH

oll

560

272.

1419

63–2

004

1.12

–2.3

430

1.7–

5.67

3.97

17.3

Fel

lsa

124

05.

919

77–2

003

1.25

–2.7

712

21.

2–4.

683.

4815

.3G

rım

sa50

70

6.45

1961

–199

70.

910–

2.29

152

1.2–

4.5

3.3

16.5

Fja

rdar

a56

011

.219

62–2

004

1.30

–2.5

495

2.1–

5.3

3.2

aA

rea

wei

ghte

dav

erag

ero

ckag

e.b

Tim

ep

erio

dw

her

ed

isch

arge

mea

sure

men

tsw

ere

avai

lab

lefo

rfl

ux

calc

ula

tio

ns.

cR

un

off

ran

ged

uri

ng

the

stu

dy

per

iod

.


The composition of the major and trace cations in col-lected water samples was measured using either anOptima 4300 DV Series Inductively Coupled PlasmaAtomic Emission Spectroscopy (ICP-AES) or anElement 1 model Inductively Coupled Plasma SectorField Mass Spectroscopy (ICP-SFMS) located at ALSScandinavia, Lulea in Sweden. These analyses were per-formed according to EPA-method USP 200.7(mod) forICP-AES and EPA 200.8(mod) for ICP-SFMS; theNational Research Council of Canada SLRS-4 riverwater reference material was used for external calibration.The uncertainties of all measurements are less than 10%.Filtrated samples for alkalinity measurements were col-lected in amber glass bottles and kept cold until theywere determined by Gran titration (c.f. Stumm andMorgan, 1996). Concentrations of Cl, SO4, and F weredetermined using a Dionex anion chromatograph. Thenutrients PO4, NO3, NO2, and NH4 were determinedusing segmented flow analysis on an Alpkem colorimetricsystem with an uncertainty of ±10%. Total nitrogen wasobtained by digesting the water sample for 2 h in aMetrohm 705 UV digestor after adding H2O2 to thewater sample, and then analyzing the fluid for NO3. Toeliminate the effects of sea spray input on calculatedfluxes, all fluid concentrations were corrected assumingthat all the dissolved Cl present in the water samplesoriginated from rainwater and using the element/Cl ratioof Icelandic rainwater (Gislason et al., 1996). The totaldissolved inorganic carbon (DIC) flux was calculatedfrom the measured alkalinity, pH, temperature, total dis-solved silica, and discharge at the time of sampling(Gislason et al., 2004). The average charge balance ofthe 220 samples was 3%.

The concentration of the suspended inorganic particu-late matter (SIM) was measured at the HydrologicalService of the National Energy Authority from 1960 topresent. The organic fraction of the suspended materialwas removed prior to analysis by boiling the sample inH2O2, (Palsson and Vigfusson, 1996). The chemical com-position of the SIM samples collected from 1998 to 2003was measured as described by Eiriksdottir et al. (2008).Analysis of particulate organic carbon and nitrogen onglass fiber filters (Whatman GF/F 0.7 lmpre-combusted for 4 h at 450 �C) was performed with aCarlo Erba model 1108 high temperature combustion ele-mental analyzer, using standard procedures and a com-bustion temperature of 1030 �C. Acetanilide was usedfor standardization, and results were corrected for blankfilter carbon content. A further description of dischargemeasurements, sampling, and analyses is given inGislason et al. (2009) and Eiriksdottir et al. (2013).

Riverine element fluxes of dissolved and suspendedmatter were calculated by multiplying the chemical con-centration of each element by the corresponding dis-charge at the time of sampling. The propagation ofvariance of the elemental flux calculations were estimatedto be 11%, given that the discharge has a 5% uncertainty(Milliman and Farnsworth, 2011) and analytical error ondissolved element and suspended matter concentration is10%.


3. RESULTS

The concentrations of dissolved elements and suspendedmaterial, pH of the river water, alkalinity, instantaneousdischarge, and water- and air temperature in all samplesare presented in the electronic supplement (Table A1).

The discharge dependence of the concentration of sus-pended inorganic particulate matter (SIM) and dissolvedNa and Fe in the representative glacial river Jokulsa aDal at Hjardarhagi is illustrated in Fig. 2. Sodium and Feconcentrations were chosen for this figure as they are repre-sentative of the behavior of soluble and insoluble elements,respectively. In this representative river, the SIM concentra-tion increases in response to increasing discharge, whereasthe dissolved concentration of the soluble element Nadecreases with increasing discharge consistent with fluiddilution. Similar trends have been previously reported(e.g. Tipper et al., 2006; Calmels et al., 2011). In contrast,the dissolved concentration of the insoluble Fe appears tobe nearly independent of discharge and exhibits significantscatter; likely reasons for this scatter are outlined below.The logarithmic plots in Fig. 2 demonstrate that the SIM

1

SIM

(mg/

l)

1

Na

(mm

ol/k

g)

1

Fe (μ

mol

/kg)

Na = 1.24 Q-0.42

R² = 0.81

0.00

0.10

0.20

0.30

0.40

0.50

0 200 400 600

Na

(mm

ol/k

g)

Fe = 0.06 Q0.17

R² = 0.05

0.00

0.30

0.600.90

1.20

1.50

1.80

2.10

0 200 400 600

Fe (μ

mol

/kg)

Discharge at JD Hjardarhagi (m3/s)

SIM = 0.07 Q1.69

R² = 0.75

0

2000

4000

6000

8000

10000

0 500 1000

SIM

(mg/

l)

Fig. 2. The discharge dependence of the instantaneous concentration of sNa and Fe on linear–linear and a log–log scales in Jokulsa a Dal at Hjardperiod 1998–2003, whereas data for SIM was collected over the periodshowing how concentrations would evolve if they were controlled solely bto a least squares fit of the data, consistent with the equations given inuncertainty is estimated for the measured concentrations.

and dissolved Na concentrations are closely consistent withan exponential dependence on discharge. Concentrationversus discharge plots, such as shown in Fig. 2, providesthe means to estimate the instantaneous concentration ofthe dissolved elements and SIM if the instantaneous dis-charge is known.

The dissolved element concentration versus dischargecorrelations in samples collected from 1998 to 2003 andfrom 1962 to 2003 for SIM concentration, as shown inFig. 2, were used together with the average daily dischargeof the individual rivers, as reported by The HydrologicalService (2005b) to calculate the daily dissolved elementand SIM fluxes in the studied rivers over the past 26–42 years. The flux calculations end in 2004 because the gla-cial rivers were dammed for hydropower production thefollowing year. The effect of the dam construction on theriverine fluxes will be described in a follow-up paper.Note, however, that although the damming of the glacialrivers changed discharge dramatically, the concentration–discharge correlation of dissolved major elements and mosttrace elements has not changed much (Eiriksdottir et al.,2014). Therefore, it seems reasonable to assume that the

log SIM = log 0.07 + 1.69 log QR² = 0.75

1

10

100

1000

0000

1 10 100 1000

log Na = log 1.24 -0.42 log QR² = 0.81

0.01

0.10

1.00

0.00

1 10 100 1000

log Fe = log 0.06 + 0.17log QR² = 0.05

0.01

0.10

1.00

0.00

1 10 100 1000log discharge at JD Hjardarhagi (m3/s)

uspended inorganic material (SIM), and concentrations of dissolvedarhagi. Note, data for the dissolved samples were collected over the1961–2003. The dashed lines are the conservative mixing line andy the dilution of the catchment waters. The solid curves correspond

the each plot, where Q stands for the discharge in m3/s. A 10%

0

1

2

3

4

5

6

7

8

9

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

log

flux

(kg/

day)

SIM Na dissolved Jökulsá á Dal at Hjardarhagi

Fig. 3. Average daily flux of suspended inorganic material (SIM) and dissolved Na in the glacial river Jokulsa a Dal at Hjardarhagi. Thecurves were calculated using the fits from Fig. 2 together with the average daily discharge data of the river.


dissolved element concentration–discharge correlations arestable over long time periods and can be used to calculatethe dissolved fluxes back in time. An example of calculateddaily fluxes is shown in Fig. 3. Several observations areapparent in this figure. First, the daily fluxes exhibit a sig-nificant seasonal variation, where the SIM fluxes are upto 4 orders of magnitude higher in the summer than thewinter. Second, in this glacial river, the SIM varies far moreseasonally than the corresponding dissolved fluxes. Third,the daily SIM flux in this glacial river exceeds the dissolvedNa fluxes throughout the year.

The daily fluxes, such as shown in Fig. 3, were integratedto generate the corresponding annual fluxes over 26–42 years of SIM and 28 distinct dissolved elements in thestudied rivers. Examples of the variation of the calculatedfluxes of annual SIM, annual dissolved Na, and annual dis-solved Fe as a function of annual runoff in all studied riversare illustrated in Fig. 4. Annual runoff was obtained bysumming the average daily discharge throughout the year,and dividing this result by the geographic surface area ofthe river catchment; these surface areas are listed inTable 1. The annual fluxes shown in Fig. 4 have been nor-malized to the lowest annual flux of the individual dissolvedelement or SIM of each river to allow their runoff depen-dence to be directly compared. These normalized fluxes willbe referred to as “relative fluxes” in the following text, whilenon-normalized fluxes are referred to as “absolute fluxes”.The calculated annual fluxes in Fig. 4 are approximately alinear function of their corresponding annual runoff. Inall the glacial river sites, the impact of increasing runoff isstrongest on the annual SIM flux and least on the annualdissolved Na flux. In contrast, the impact of increasing run-off is stronger on the dissolved Fe flux than on the SIM fluxin Fellsa, one of the non-glacial rivers. Notably, the runoffdependence of the dissolved Na flux is less than that of theFe flux in all the rivers. This latter observation stems fromthe distinct effect of dilution on soluble elements such asNa; the effect of dilution on insoluble elements such as Feis low because its concentration in the river waters is likelycontrolled by the low solubility of fine-grained secondaryminerals, such as ferrihydride, in the river catchments.

The variation of the relative riverine annual fluxes of thedissolved elements with annual runoff is shown in Fig. 5.Also shown in this figure, are dashed lines having a slope

of 0 and 1. If the annual dissolved flux of an element is con-trolled solely by water dilution, the slope of the flux versusrunoff curve would equal 0. If the river water concentrationis unaffected by dilution and the dissolved concentration ofthe element is independent of discharge, the slope of theflux versus runoff curve would equal 1. In general, the lesssoluble an element, the more its’ flux is affected by changesin annual runoff. The dissolved elemental fluxes mostaffected by runoff are those of Zn, Ti, and Fe, whereasthe fluxes least affected by runoff are those of Mo, Sr,Mg, and K. Interestingly, in numerous cases, the slope ofrelative annual dissolved elemental flux versus annual run-off exceeds 1, implying that the concentration of the ele-ment increases with increasing discharge (Fig. 2). Thisbehavior is seen for the insoluble elements Fe, Ti, and Znin several rivers, and for Al in the Grımsa river. These ele-ments tend to precipitate or co-precipitate as colloidal sizedsecondary phases. Their elevated runoff dependencies maytherefore stem from our use of 0.2 lm filters, which mayallow colloidal sized particles to remain in the fluid phaseafter sampling and filtering (c.f. Raiswell and Canfield,2012).

A summary of the effect of annual runoff on the annualdissolved element and SIM fluxes is listed in Table 2 andillustrated in Fig. 6. Table 2 provides the slopes and inter-cepts of the linear correlations between the runoff and theriverine fluxes. The values in Table 2 were used to calculatethe percent change in dissolved element and SIM flux inresponse to a 1% increase in runoff in each river. Severalobservations are apparent in Fig. 6. First, the fluxes ofthe soluble element in the non-glacial rivers tend to increaseby �0.8% increase for every 1% increase in runoff. Theannual fluxes of the insoluble elements, Mn, Al, Ni, Co,Fe, and Ti, tend to increase more than 1% for each percentincrease in annual runoff, but do not to exceed 2%. Incontrast, the annual fluxes of each element tend to show amore distinct behavior in the glacial rivers. The annualfluxes of the most soluble elements (e.g. Mo, Sr, and Mg)tend to increase by no more than 0.5% for each percentincrease in annual runoff, and far fewer elements haveannual fluxes that increase more than 1% per percentannual runoff increase. The increased variation in theannual fluxes of the soluble elements in the glacial riverscompared to the non-glacial rivers suggests that dilution

0

1

2

3

4

0,0 1,0 2,0 3,0

Fellsá

SIM = 0.97 R - 0.24Fe = 1.22 R + 0.55Na = 0.66 R + 0.22

0

1

2

3

4

5

0 1 2 3

Grímsá

SIM = 1.85 R - 0.77Fe = 1.18 R + 0.07Na = 0.86 R + 0.21

0

1

2

3

0 1 2 3 4

Fjardará

SIM = 0.93 R - 0.56Fe = 0.62 R + 0.01Na = 0.47 R + 0.26

012345678

0 1 2 3

JD Brú

NaFeSIM

SIM = 3.71 R - 3.67Fe = 1.17 R - 0.48Na = 0.39 R + 0.59

0

1

2

3

4

5

6

7

0 0,5 1 1,5 2

JD Hjardarhagi

SIM = 6.14 R - 5.34Fe = 1.37 R - 0.33Na = 0.47 R + 0.61

0

1

2

3

4

5

0,0 1,0 2,0 3,0

JF Hóll

SIM = 1.97 R -1.12Fe = 0.84 R + 0.05Na = 0.57 R + 0.32

Rela

�ve

annu

al fl

ux (X

/Xm

in)

Annual runoff (m/yr)

Glacial rivers Non-glacial rivers

Fig. 4. Variations of the relative annual suspended inorganic material (SIM) and relative annual dissolved Na and Fe fluxes as a function ofannual runoff in all studied rivers. Annual elemental fluxes were generated by integrating the daily fluxes shown in Fig. 3. The relative annualfluxes have been normalized to the lowest annual flux of the individual element or SIM of each river to allow direct comparison of their runoffdependence. Least squares fits of these data are shown in the plots.


is more significant in the former. An example of the dis-charge dependency of SiO2 concentration for Fellsa, anon-glacial river, and Jokulsa a Dal at Hjardarhagi, a gla-cial river, is presented in Figure A1 in the electronic supple-ment. The lower significance of discharge on dissolvedelemental fluxes in non-glacial rivers may be due to moreextensive water–rock interaction in these catchments. Theincrease in runoff in glacial rivers tends to be dominatedby glacial melting; glacial melt waters are collected rapidlyinto river channels and flushed through the system. In con-trast, the increasing runoff in the non-glacial rivers, whichhave thicker soils, originates from increased rain- andsnowmelt much of which passes through catchment soilsbefore reaching the rivers.

Previous models of the feedback between climate andchemical weathering suggest a 2–10% increase in chemicalweathering rates for each degree temperature increase

(Berner and Kothavala, 2001; Wallmann, 2001). Fig. 7,which is analogous to Fig. 5, shows how the relative annualdissolved element fluxes in the studied rivers vary withannual air temperature in five of the six studied river catch-ments. Fjardara river is not shown due to lack of accurateair temperature data. Globally, a 4.9% increase in humidityis found in the surface layers of the atmosphere for each1 �C warming; 5.7% per 1 �C over the oceans and 4.3%per 1 �C over the continents (IPCC, 2007). The runoff inthe studied catchments in Eastern Iceland increased by 8–17% due to each 1 �C temperature increase (see Table 1).As there is a correlation between temperature and runoff(Labat et al., 2004; IPCC, 2007; Trenberth, 2007;Bjornsson et al., 2008), there is also a dependence of dis-solved element fluxes on average annual air temperature.Nevertheless, the scatter of the annual dissolved elementfluxes versus average annual air temperature plot is far

0

1

2

3

4

1,0 1,5 2,0 2,5 3,0

JD Brúno dilu�on

pure dilu�on

0

1

2

3

0,5 1,0 1,5 2,0

JD Hjardarhagi

0

1

2

3

1,0 1,5 2,0 2,5

JF Hóll

0

1

2

3

4

1,0 1,5 2,0 2,5 3,0

Fellsá

0

1

2

3

4

5

6

7

0,5 1 1,5 2 2,5 3

Grímsá

0

1

2

3

4

1,0 1,5 2,0 2,5 3,0 3,5

Fjardará

Annual runoff(m/yr)

Rela

�ve

annu

al e

lem

enta

l flux

(X/X

min

)


Fig. 5. Variations of relative annual fluxes of dissolved elements with annual runoff in all rivers considered in this study. The dashed line has aslope of 0 consistent with dissolved concentrations controlled solely by water dilution and the diagonal dashed line with a slope of 1 consistentwith elemental concentration which is unaffected by dilution.


greater than on the corresponding annual runoff plotsshown in Fig. 5. There are several reasons for this increasedscatter. First, there is significant scatter in annual rainfallversus temperature plots (see Fig. 7f). Second, the averageannual air temperature is based on measurements from asingle weather station located near each sampling site.This temperature may not be representative of the wholecatchment. Third, the calculated annual elemental fluxesare based on a fit of instantaneous river water concentra-tions versus instantaneous discharge (see Fig. 2), whichdampens scatter on the correlations shown in Fig. 5. Notethat some of the climate response variation between indi-vidual catchments could stem from different vegetation

cover and properties of the basement rock. Despite thisscatter and caveats, the slopes of the flux versus tempera-ture plots shown in Fig. 7 provide insight into the feedbackbetween air temperature and chemical transport to theoceans.

A summary of the effect of average annual air temper-ature on the annual dissolved element- and SIM fluxes islisted in Table 3 and illustrated in Fig. 8. Table 3 pro-vides both the intercept and slopes of the linear correla-tions shown in Fig. 7 describing the effect of annualaverage temperature on annual fluxes. These values wereused to calculate the percent change in dissolved elemen-tal fluxes and SIM flux in response to a 1 �C temperature

Table 2Intercepts, slopes and coefficients of determination (R2) of the linear fits through the correlation of absolute element and SIM fluxes and runoff (m/yr) of the studied rivers. These fluxes are shownas normalized fluxes in Fig. 5. Also provided in this table is the calculated flux change of dissolved elements, POC, and SIM flux due to 1% runoff change. Fluxes are in units of: Si to S mol/m2/yr,F to Sr mmol/m2/yr, Ba to Ti lmol/m2/yr, SIM is in g/m2/yr and POC is mg/m2/yr.

JD Bru JD Hjardarhagi JD Holl Fellsa Grımsa Fjardara

Interc. Slope R2 % change

pr. 1%

runoff

change


pr. 1%

runoff

change


pr. 1%

runoff

change


pr. 1%

runoff

change


pr. 1%

runoff

change


pr. 1%

runoff

change

Particulate and major dissolved elements

Si 0.058 0.0405 0.90 0.45 0.061 0.0504 0.85 0.46 0.041 0.0710 0.96 0.64 0.030 0.0901 0.98 0.82 0.022 0.135 1.00 0.85 0.032 0.0893 0.98 0.79

Na 0.093 0.0613 0.89 0.44 0.073 0.0564 0.84 0.44 0.038 0.0589 0.95 0.62 0.015 0.0253 0.95 0.72 0.012 0.0403 0.99 0.76 0.017 0.0372 0.98 0.75

K 0.001 0.0013 0.93 0.53 0.002 0.0018 0.88 0.50 0.001 0.00300 0.97 0.71 0.000 0.0020 1.00 0.92 0.001 0.0025 1.00 0.80 0.000 0.0016 1.00 0.94

Ca 0.042 0.0446 0.94 0.56 0.044 0.0563 0.91 0.57 0.056 0.126 0.97 0.70 0.015 0.0356 0.97 0.78 0.032 0.0874 0.99 0.72 0.015 0.0331 0.98 0.76

Mg 0.011 0.0042 0.81 0.31 0.026 0.0096 0.70 0.27 0.024 0.0240 0.93 0.53 0.010 0.0225 0.97 0.77 0.014 0.0381 0.99 0.72 0.009 0.0179 0.97 0.72

DIC 0.190 0.164 0.92 0.51 0.212 0.198 0.87 0.49 0.169 0.363 0.97 0.69 0.068 0.17 0.97 0.79 0.094 0.287 0.99 0.75 0.069 0.156 0.98 0.76

S 0.009 0.00340 0.80 0.31 0.006 0.0028 0.74 0.31 0.018 0.0148 0.91 0.48 0.001 0.0021 0.97 0.78 0.009 0.0178 0.98 0.64 0.001 0.0053 0.99 0.84

F 1.543 0.730 0.84 0.36 1.269 0.835 0.81 0.40 0.760 1.70 0.97 0.70 0.000 0.0007 0.99 0.85 0.227 1.30 1.00 0.85 0.035 0.762 1.00 0.97

Nutrients

P-total 0.162 0.415 0.99 0.76 0.083 0.354 0.99 0.85 �0.014 0.333 1.00 1.04 0.016 0.0405 0.97 0.80

PO4 0.158 0.418 0.99 0.76 0.021 0.469 1.00 0.96 �0.035 0.348 1.00 1.11 0.018 0.0978 0.99 0.89 0.017 0.100 1.00 0.85 0.014 0.0824 1.00 0.89

NO3 1.324 2.19 0.97 0.61 0.553 1.51 0.97 0.79 0.352 1.54 0.99 0.78 0.441 0.762 0.95 0.81 0.375 0.734 0.98 0.86 0.459 1.56 0.99 0.71

N-total 2.470 3.20 0.96 0.67 0.893 3.29 0.98 0.74 0.797 2.72 0.99 0.82 0.904 2.48 0.98 0.72 0.44 2.80 1.00 0.65 1.26 2.19 0.97 0.83

Trace elements

Al 0.106 0.519 1.00 0.86 0.001 0.577 1.00 1.00 �0.078 0.477 0.99 1.19 �0.008 0.151 1.00 1.03 �0.111 0.283 0.97 1.63 �0.022 0.182 1.00 1.10

Fe �0.061 0.160 0.98 1.43 �0.041 0.170 0.98 1.29 0.009 0.135 1.00 0.94 �0.097 0.214 0.94 1.53 �0.022 0.275 1.00 1.09 0.002 0.169 1.00 0.99

B 0.069 0.0411 0.88 0.42 0.048 0.065 0.92 0.58 0.026 0.0867 0.99 0.78 0.009 0.0466 0.99 0.89 0.011 0.0473 1.00 0.81 0.037 0.0810 0.98 0.75

Mn 0.005 0.0155 0.99 0.79 0.002 0.0340 1.00 0.94 �0.029 0.137 0.98 1.27 �0.006 0.0120 0.93 1.59 �0.001 0.0142 0.96 2.12 �0.002 0.0175 1.00 1.07

Sr 0.004 0.0014 0.80 0.30 0.011 0.004 0.68 0.25 0.023 0.0401 0.96 0.64 0.005 0.0129 0.98 0.81 0.020 0.0450 0.99 0.68 0.008 0.0137 0.97 0.71

Ba 0.000 0.163 1.00 1.00 0.013 0.160 1.00 0.93 0.026 0.205 1.00 0.89 �0.030 0.237 1.00 1.09 0.109 0.609 1.00 0.84 0.014 0.398 1.00 0.98

Cd 0.009 0.00590 0.89 0.45 0.002 0.0288 1.00 0.93 0.002 0.0242 1.00 0.93 0.004 0.0233 0.99 0.89 �0.001 0.0290 1.00 1.02 0.010 0.0190 0.97 0.73

Co �0.008 0.126 1.00 1.05 0.017 0.205 1.00 0.93 �0.006 0.240 1.00 1.03 �0.013 0.126 1.00 1.07 0.016 0.181 1.00 0.92 �0.044 0.214 0.99 1.17

Cr 0.847 0.370 0.83 0.34 0.789 0.525 0.81 0.40 0.172 0.641 0.99 0.79 0.068 0.634 1.00 0.94 �0.014 1.11 1.00 1.01 0.321 1.13 0.99 0.83

Cu 1.319 2.40 0.98 0.69 1.525 3.98 0.97 0.73 0.691 3.65 0.99 0.84 0.811 3.68 0.99 0.88 0.106 4.11 1.00 0.97 0.695 2.16 0.99 0.81

Ni 0.474 3.08 1.00 0.89 �0.236 4.32 1.00 1.05 �0.059 3.35 1.00 1.02 �1.222 4.62 0.98 1.23 �0.702 4.05 1.00 1.21 �0.852 4.27 0.99 1.16

Pb �0.018 0.0849 0.99 1.20 0.001 0.0826 1.00 0.99 �0.004 0.0936 1.00 1.04 0.003 0.0848 1.00 0.98 �0.017 0.100 1.00 1.20 0.009 0.107 1.00 0.94

Zn �5.230 9.68 0.96 1.75 �0.446 6.07 1.00 1.07 �0.646 7.04 1.00 1.10 �2.27 9.60 0.98 1.20 �0.158 7.90 1.00 1.02 0.849 9.22 1.00 0.94

Mo 1.434 0.312 0.72 0.20 1.155 0.304 0.63 0.21 1.728 1.48 0.91 0.49 0.169 0.145 0.88 0.55 0.268 0.623 0.99 0.69 �0.091 0.405 0.99 1.18

Ti �9.427 20.2 0.97 1.59 �4.460 14.56 0.97 1.40 �2.938 16.93 0.99 1.21 �1.02 4.75 0.99 1.17 �4.74 8.64 0.94 2.10 �2.01 5.06 0.96 1.37

SIM* �2045 2110 0.83 3.32 �1648 2000 0.74 4.22 �372 587 0.77 1.58 �2.82 11.0 0.98 1.23 �4.32 10.46 0.97 1.38 �5.84 12.8 0.95 1.45

POC** �0.153 0.430 0.98 1.32 104 146 0.93 0.59 �196 642 0.96 1.21 �41.5 204 0.99 1.19 �5 174.31 1.00 0.98 �101 321 0.98 1.27

* SIM = Suspended inorganic Material.** POC = Particulate organic carbon.

E.S

.E

iriksd

ottir

etal./

Geo

chim

icaet

Co

smo

chim

icaA

cta166

(2015)249–266

257

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Mo

Sr

S M

g Cr F B N

aCd

SiDI

C K CaN

O3

N- t

otal

Cu

P-

tota

lPO

4M

n

Al

Ni

Ba

Co

Pb

POC

Fe

Ti

Zn

SIM

FellsáGrímsáFjarðará

Non-glacial rivers

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Mo

Sr

S M

g Cr

F B Na

Cd Si

DIC K Ca

NO

3N

-tot

al

Cu

P-to

tal

PO4

Mn

Al

N

i Ba

Co

Pb

PO

CFe

Ti

Zn

SI

M

JD Brú

JD Hjardarhagi

JF Hóll

Glacial rivers

% fl

ux c

hang

e pe

r 1 %

runo

ff ch

ange

Fig. 6. Percent change in the flux of the indicated dissolved element or suspended inorganic material (SIM) and particulate organic material(POC) per 1% increase in runoff. The error bars show the propagation of variance of the fluxes, which is 11%. The order of elements in thisfigure was set by the increasing values of this flux dependence for Jokulsa a Dal at Bru to best show this variation among the elements. Thedashed horizontal line corresponds to a 1% flux increase for a 1% increase in runoff; equal to that when element concentrations are runoffinvariant. If a flux plots above this line, the concentration tends to increase with increasing runoff; those elements that are above the dashedline tend to form colloids – see text.


increase in each river. These latter values are provided inTable 3 and plotted in Fig. 8. The order of the elementsin Fig. 8 is the same as that of Fig. 6. Data on annualaverage air temperature is not available for as manyyears as the data for annual runoff. Nevertheless, Fig. 8demonstrates that the temperature dependence of theannual dissolved fluxes vary far more among the elementsin the glacial rivers than the non-glacial rivers. For exam-ple the annual dissolved fluxes of Mo and Sr increase byless than 5% per degree increase in average annual airtemperature in Jokulsa a Dal at Bru and atHjardarhagi. The annual dissolved fluxes of the least sol-uble elements Zn, Ti, and Fe increase �3 times fasterwith increasing temperature than the most soluble ele-ments in these rivers. In contrast, the annual elementalfluxes in the non-glacial rivers vary only slightly withthe identity of the element, these annual fluxes increaseby 15 ± 4% for each �C increase in average annual airtemperature which is somewhat higher than that sug-gested by Berner and Kothavala (2001) and Wallmann

(2001). The increase in annual SIM flux per degree aver-age air temperature increase in the glacial rivers is signif-icantly greater than that of the dissolved elements.

4. DISCUSSION

4.1. Increasing dissolved element concentrations with

increasing runoff

A significant observation presented above is that theconcentrations of a number of dissolved insoluble ele-ments appear to increase with increasing runoff, such thatthe slope of annual flux versus annual runoff in Fig. 5exceeds one. Elements exhibiting this behavior includeFe in Jokulsa a Dal at Bru, Jokulsa a Dal atHjardarhagi, and Grımsa, and Al in Grımsa. There areseveral possible explanations for these observations.First, the concentrations of these sparingly soluble metalsare likely influenced by the rapid dissolution and precip-itation of minute phases, such that their concentrations

0

0.5

1

1.5

2

2.5

3

-2 -1 0 1 2 3 4 Average annual air temperature (°C)

(a) JD Brú

0

0.5

1

1.5

2

2.5

-2 -1 0 1 2 3 4 Average annual air temperature (°C)

(c) JD Hjardarhagi

0

0.5

1

1.5

2

2.5

3

0 1 2 3 4 5 6

(e) JF Hóll

0

0.5

1

1.5

2

2.5

3

3.5

0 1 2 3 4 5 Average annual air temperature (°C)

(b) Fellsá

0

0.5

1

1.5

2

2.5

3

3.5

4

0 1 2 3 4 5 Average annual air temperature (°C)

(d) Grímsá

Average annual air temperature (°C)

Rela

�ve

annu

al e

lem

enta

l flux

(X/X

min

)


0

0.5

1

1.5

2

2.5

3

-2 -1 0 1 2 3 4 5 6 7

Annu

al ru

noff

(m/y

r)JD HjardarhagiJD BrúJF HóllFellsáGrímsá

(f)

Fig. 7. (a–e) Relative flux of dissolved elements and annual runoff variation as a function of average annual air temperature for five of the sixstudied rivers. (f) Variation of annual runoff in the five rivers as a function of temperature.


are controlled by mineral solubility. As mentioned above,the precipitation of some of these elements could haveformed or co-precipitated as colloids that passed throughthe filters used during sampling. The solubilities of thesesparingly soluble elements are also highly sensitive to thepresence of organic ligands and other complexingdissolved anions. Just a small increase in the concentra-tion of such anions can alter significantly the totaldissolved concentration of a sparingly soluble element.Note that high runoff is associated with higher tempera-tures and thus enhanced biotic activity, which couldrelease additional organic ligands to surface waters,consistent with the observed increase in insoluble metalconcentration with runoff. The effect of these ligandsmay also be in part responsible for the scatter observedin the dissolved Fe versus discharge plots in Fig. 3. Wenevertheless chose to use Fe as an illustrative examplein this report due to its significance to primary productiv-ity in the oceans.

4.2. Nutrient fluxes and climate change from 1961 to 2004

The results presented in the previous section demon-strate a direct relationship between the annual fluxes of sus-pended and dissolved elements from the six studiedcatchments towards the ocean as a function of runoff andtemperature. As emphasized by Gislason et al. (2009), thestudied river catchments have experienced a significant cli-mate change over the 40-year study period. This climatevariation is illustrated for the six studied catchments inFig 9. Despite scatter, the average annual air temperaturehas increased in all of the six studied catchments consideredin this study. Based on the least square fit of the data shownin Fig. 9, the average annual air temperature has increasedmost in the Fellsa catchment, by an average of 0.06 �Cannually, and least in Grımsa, by an average of 0.02 �Cannually. Consistent with the close correspondence betweenthe average annual temperature and runoff both in EasternIceland and worldwide (Labat et al., 2004; IPCC, 2007;

Table 3Intercepts, slopes and coefficients of determination (R2) of linear fits through the correlation of absolute fluxes of dissolved elements, POC, and SIM, and average annual air temperature (�C) in thestudied river catchments. These fluxes are shown as normalized fluxes in Fig. 7. Also provided is the calculated change of dissolved elemental, POC, and SIM flux due to one �C air temperaturechange. Fluxes are in units of: Si to S mol/m2/yr, F to Sr mmol/m2/yr, Ba to Ti lmol/m2/yr, SIM is in g/m2/yr and POC is mg/m2/yr.

Bru Hjardarhagi Holl Fellsa Grımsa

Interc. Slope R2 % change pr.

one �C


one �C


one �C


one �C


one �C

Particulate and major dissolved elements

Si 0.121 0.00924 0.40 7.5 0.123 0.00626 0.31 5.0 0.083 0.0215 0.62 14.0 0.105 0.0269 0.417 14.0 0.146 0.0302 0.354 15.5

Na 0.188 0.0141 0.40 7.5 0.143 0.00706 0.31 5.0 0.072 0.0179 0.63 13.5 0.034 0.00796 0.445 13.0 0.0480 0.00931 0.370 15.2

K 0.00347 0.00030 0.40 8.0 0.004 0.00022 0.31 5.5 0.003 0.000892 0.61 14.5 0.002 0.000577 0.388 14.5 0.00290 0.000579 0.363 15.0

Ca 0.111 0.0099 0.40 8.5 0.115 0.00671 0.30 5.5 0.132 0.0374 0.61 14.5 0.044 0.0109 0.429 13.5 0.109 0.0205 0.377 14.5

Mg 0.0176 0.00102 0.39 5.5 0.038 0.00129 0.30 3.5 0.037 0.00751 0.65 12.0 0.028 0.00688 0.431 13.5 0.0476 0.00895 0.377 14.5

DIC 0.444 0.03669 0.40 8.0 0.457 0.0243 0.31 5.0 0.387 0.1084 0.61 14.5 0.205 0.0513 0.425 13.5 0.348 0.0668 0.373 15.0

S 0.0144 0.00082 0.38 5.5 0.0097 0.000371 0.30 3.5 0.026 0.00471 0.66 11.0 0.00255 0.000629 0.430 13.5 0.0246 0.00431 0.392 14.0

F 2.67 0.172 0.39 6.5 2.30 0.106 0.31 4.5 1.787 0.507 0.61 14.5 0.00073 0.000193 0.409 14.0 1.413 0.291 0.355 16.0

Nutrients

P-total 0.810 0.08726 0.39 10.5 0.530 0.0385 0.28 7.0 0.216 0.0913 0.53 17.5 0.049 0.0122 0.424 13.5

PO4 0.812 0.08784 0.39 10.5 0.615 0.0487 0.26 7.5 0.210 0.0940 0.51 18.0 0.102 0.0281 0.396 14.5 0.108 0.0223 0.36 15.5

NO3 4.75 0.471 0.40 9.5 2.45 0.17 0.29 6.5 1.33 0.444 0.58 16.0 1.03 0.239 0.445 13.0 1.01 0.177 0.39 14.0

N-total 7.46 0.698 0.40 9.0 5.04 0.36 0.28 7.0 2.50 0.7944 0.59 15.5 2.93 0.744 0.421 14.0 3.01 0.62 0.35 16.0

Trace elements

Al 0.919 0.107 0.39 11.0 0.733 0.0591 0.25 8.0 0.266 0.127 0.49 18.5 0.132 0.0407 0.356 15.5 0.188 0.0505 0.26 18.5

Fe 0.193 0.03052 0.35 15.0 0.176 0.0160 0.21 9.0 0.099 0.0379 0.55 17.0 0.121 0.0516 0.272 18.5 0.244 0.0569 0.32 17.0

B 0.132 0.00949 0.40 7.0 0.129 0.00766 0.30 6.0 0.080 0.0253 0.59 15.5 0.049 0.0134 0.398 14.5 0.053 0.0107 0.36 15.5

Mn 0.029 0.00323 0.39 10.5 0.045 0.0036 0.26 7.5 0.071 0.0358 0.47 19.0 0.006 0.00287 0.265 19.0 0.0131 0.00280 0.29 16.0

Sr 0.006 0.00034 0.38 5.5 0.015 0.000499 0.30 3.0 0.047 0.0121 0.62 14.0 0.015 0.00386 0.420 14.0 0.0593 0.0108 0.38 14.5

Ba 0.256 0.03235 0.37 12.0 0.215 0.0167 0.26 7.5 0.160 0.0582 0.56 16.5 0.193 0.0625 0.341 16.0 0.665 0.137 0.36 15.5

Cd 0.018 0.00135 0.40 7.5 0.039 0.0030 0.26 7.5 0.018 0.00680 0.55 17.0 0.024 0.00671 0.396 14.5 0.0270 0.0061 0.33 16.5

Co 0.191 0.02515 0.37 12.5 0.276 0.0215 0.26 7.5 0.158 0.0660 0.53 17.5 0.105 0.0335 0.346 16.0 0.185 0.0396 0.34 16.0

Cr 1.418 0.08785 0.39 6.0 1.44 0.0667 0.31 4.0 0.576 0.186 0.59 15.5 0.627 0.179 0.384 15.0 1.04 0.234 0.33 16.5

Cu 5.072 0.514 0.40 10.0 6.53 0.446 0.29 6.5 3.04 1.05 0.57 16.0 3.95 1.07 0.401 14.5 3.98 0.881 0.34 16.5

Ni 5.308 0.631 0.38 11.5 5.26 0.435 0.24 8.0 2.23 0.923 0.53 17.5 3.29 1.17 0.313 17.0 3.31 0.807 0.30 17.5

Pb 0.116 0.0166 0.36 13.5 0.106 0.008 0.25 8.0 0.061 0.0256 0.53 17.5 0.079 0.0234 0.372 15.0 0.082 0.0200 0.30 17.5

Zn 10.09 1.79 0.33 17.0 7.28 0.608 0.24 8.0 4.29 1.90 0.51 18.0 7.03 2.45 0.319 16.5 7.37 1.67 0.33 16.5

Mo 1.91 0.0785 0.37 4.0 1.53 0.0420 0.29 2.5 2.52 0.468 0.65 11.5 0.265 0.0502 0.490 11.5 0.811 0.148 0.38 14.5

Ti 22.5 3.79 0.34 16.0 14.2 1.33 0.20 9.0 9.29 4.48 0.49 18.5 3.56 1.22 0.323 16.5 4.87 1.38 0.22 19.0

SIM* 1313 363 0.27 27.0 978 127 0.07 13.0 126.4 134 0.30 23.0 7.85 2.78 0.314 16.7 6.78 1.84 0.25 19.0

POC** 1312 363 0.27 14.5 286 17 0.30 6.0 290.1 163 0.44 20.0 154.9 53 0.326 16.5 162 37 0.33 16.5

* SIM = suspended inorganic material.** POC = Particulate organic material.

260E

.S.

Eirik

sdo

ttiret

al./G

eoch

imica

etC

osm

och

imica

Acta

166(2015)

249–266

Fig. 8. Percent change in the flux of the indicated dissolved element or suspended inorganic material (SIM) and particulate organic material(POC) per �C temperature change. The error bars show the propagation of variance of the fluxes which is 11%. The order of elements is thesame as in Fig. 6, the increasing values of the flux dependence for Jokulsa a Dal at Bru to best show this variation among the elements.


Trenberth, 2007; Gislason et al., 2009), this temporalincrease in average annual air temperature has led to a cor-responding temporal increase in the annual runoff in thesecatchments. As can be seen in Fig. 9, the average runoffhas increased in all six studied catchments. Overall, theannual runoff increases over the 26–42 year study periodrange from 0.2 to 2.0 cm/yr, and as noted in Table 1, runoffchanges by 8–17% for each degree C change.

As illustrated in Figs. 5 and 7 there is a direct correlationbetween the dissolved chemical fluxes towards the oceanand both temperature and runoff. As such one would expectto see an increase in these fluxes due to climate change overthe 40-year study period. Such appears to be the case as canbe seen in Fig. 9, where the annual dissolved fluxes of thenutrients Si, PO4, and Fe are illustrated for the six studiedcatchments from �1965 to 2004. These fluxes were

calculated using daily discharge from �1965 to 2004 andthe rating curves for discharge – elemental flux relationshipsbased on samples collected from 1998 to 2003. The fluxes ofall but Si in Jokulsa a Dal at Hjardarhagi and Bru haveincreased during this time concurrent with climate change.A large variation is observed in the temporal variation ofdissolved fluxes towards the oceans in the distinct catch-ments. The largest temporal increases are found for thecatchments of Fellsa and Jokulsa ı Fljotsdal at Holl.These rivers have experienced approximately a 30–75% risein annual dissolved Si, PO4, and Fe fluxes over the 26–42 year study period. The rise is larger for Fe and PO4 thanSi, consistent with the relative climate dependence of thesefluxes shown in Figs. 6 and 8. This increase, however, isnegligible for the case of Si in the Bru and Hjardarhagicatchments. This latter observation is likely due to the

y = 0.0046x - 7.45R² = 0.03

0.00.51.01.52.02.53.0

1950 1960 1970 1980 1990 2000 2010Year

y = 0.0202x - 36.99R² = 0.076

0.0

1.0

2.0

3.0

4.0

5.0

1950 1960 1970 1980 1990 2000 2010Year

y = 0.0037x - 4.87R² = 0.020.0

1.0

2.0

3.0

4.0

1950 1960 1970 1980 1990 2000 2010

Annu

al a

vera

ge ru

noff

(m/y

r)

y = 0.0006x - 0.97R² = 0.033

0.0

0.1

0.2

0.3

0.4

0.5

1950 1960 1970 1980 1990 2000 2010Year

y = 0.0005x - 0.72R² = 0.032

0.0

0.1

0.2

0.3

1950 1960 1970 1980 1990 2000 2010

y = 0.0002x - 0.33R² = 0.032

0.00

0.04

0.08

0.12

0.16

1950 1960 1970 1980 1990 2000 2010

Year

y = 0.0003x - 0.383R² = 0.0223

0.0

0.1

0.2

0.3

1950 1960 1970 1980 1990 2000 2010

SiO

2 flu

x (m

ol/k

m2/

yr)

Year

y = 0.0003x - 0.34R² = 0.020

0.0

0.1

0.2

0.3

1950 1960 1970 1980 1990 2000 2010PO4

flux

(mm

ol/k

m2/

yr)

Year

y = 0.0005x - 0.68R² = 0.019

0.00.10.20.30.40.5

1950 1960 1970 1980 1990 2000 2010

Year

y = 0.0184x - 34.9R² = 0.17

0.0

1.0

2.0

3.0

1970 1980 1990 2000 2010Year

y = 0.0655x - 127R² = 0.46

0.0

1.0

2.0

3.0

4.0

5.0

1970 1980 1990 2000 2010Year

y = 0.0017x - 3.21R² = 0.18

0.0

0.1

0.2

0.3

1970 1980 1990 2000 2010Year

y = 0.0018x - 3.44R² = 0.18

0.0

0.1

0.2

0.3

1970 1980 1990 2000 2010Year

y = 0.004x - 7.77R² = 0.17

0.0

0.2

0.4

0.6

1970 1980 1990 2000 2010

Year

y = 0.0193x - 36.5R² = 0.57

0.00.51.01.52.02.5

1960 1980 2000 2020Year

y = 0.0318x - 59.4R² = 0.21

0.01.02.03.04.05.06.0

1960 1970 1980 1990 2000 2010Year

y = 0.0014x - 2.58R² = 0.55

0.0

0.1

0.2

0.3

1960 1970 1980 1990 2000 2010

SiO

2 flu

x (m

ol/k

m2/

yr)

Year

y = 0.0067x - 12.7R² = 0.56

0.0

0.2

0.4

0.6

0.8

1.0

1960 1970 1980 1990 2000 2010Year

y = 0.0026x - 5.00R² = 0.57

0.0

0.1

0.2

0.3

0.4

1960 1970 1980 1990 2000 2010

Year

y = 0.0021x - 2.82R² = 0.02

0.0

0.5

1.0

1.5

2.0

1960 1970 1980 1990 2000 2010Year

y = 0.0419x - 82R² = 0.20

-2.0-1.00.01.02.03.04.0

1960 1970 1980 1990 2000 2010Year

y = -7E-06x + 0.14R² = 6E-05

0.00

0.04

0.08

0.12

0.16

1960 1970 1980 1990 2000 2010Year

y = 0.0009x - 1.20R² = 0.02

0.00

0.20

0.40

0.60

0.80

1.00

1960 1970 1980 1990 2000 2010

y = 0.0005x - 0.74R² = 0.03

0.00

0.10

0.20

0.30

1960 1970 1980 1990 2000 2010

Fe fl

ux (m

mol

/km

2/yr

)

Year

y = 0.0053x - 8.67R² = 0.02

0.00.51.01.52.02.53.0

1960 1970 1980 1990 2000 2010

Annu

al a

vera

ge

runo

ff (m

/yr)

Year

y = 0.0386x - 75.7R² = 0.16

-2.0-1.00.01.02.03.04.0

1960 1970 1980 1990 2000 2010

Annu

al a

vera

ge ai

r te

mpe

ratu

re (°

C)

Year

y = 1E-05x + 0.11R² = 5E-05

0.00

0.05

0.10

0.15

0.20

1960 1970 1980 1990 2000 2010

SiO

2 (m

ol/k

m2 /

yr)

Year

y = 0.0015x - 2.11R² = 0.01

0.000.200.400.600.801.001.201.40

1960 1970 1980 1990 2000 2010

PO4

mm

ol/k

m2 /

yr)

y = 0.0011x - 2.04R² = 0.04

0.00

0.10

0.20

0.30

0.40

1960 1970 1980 1990 2000 2010

Fe (m

mol

/km

2 /yr

)

Year

Jökulsá á Dal at Brú Jökulsá á Dal at Hjardarhagi Jökulsá í Fljótsdal at Hóll Fellsá Fjardará Grímsá

Fig. 9. Temporal variations in runoff, annual average air temperature, and absolute fluxes of dissolved SiO2, PO4 and Fe in all six catchmentsconsidered in this study. All the parameters increase at the time of the study demonstrating the link between climate change and elementalfluxes to the oceans. The flux calculation is based on discharge – flux relationship of the elements in samples collected in 1998–2003 andaverage daily discharge of the rivers since monitoring of the rivers began (The Hydrological Service, 2005b).


strong dilution of dissolved Si concentration by increasingrunoff – see above and relatively small runoff increase withtemperature (see Table 1). Note also the dissolved PO4 fluxin the glacial rivers is significantly higher than in thenon-glacial rivers. Glacial catchments have been found tohave a relatively high P concentration due to weatheringof fine grained glacial sediments and limited vegetationcover (Gislason et al., 1996; Hodson et al., 2004; Hoodand Berner, 2009). This leads to a higher mobility of phos-phorus in glacial rivers than in non-glacial direct runoffrivers.

4.3. The relative role of dissolved versus particle fluxes in

nutrient transport to the oceans

The temporal variation of SIM in the six river catch-ments considered in this study over a 40-year period hasbeen reported by Gislason et al. (2009). Concentration ofSIM, particle organic carbon (POC), and dissolved ele-ments measured in the samples collected from 1998 to2003 are given in Table A1 in the electronic supplement.Collectively, these results illuminate the relative role of sus-pended particle versus dissolved transport towards theoceans. A comparison of the relative and absolute total dis-solved and particle fluxes towards the ocean of Si, PO4 andFe in the Holl and Fellsa catchments are shown in Fig. 10.It can be seen in this figure that particle transport domi-nates the fluxes of Si, PO4, and Fe. In each case, the particleflux exceeds the dissolved flux by 0.5–4 orders of magni-tude. The importance of particle transport on elementfluxes has been noted on a global scale (Oelkers et al.,2011, 2012); these previous studies also noted that the lesssoluble the element, the more the element tends to be

transported by the suspended particulate material. It canalso be seen in Fig. 10 that suspended material is moreimportant in the glacial than the non-glacial rivers; theabsolute suspended particulate fluxes to the oceans of Si,PO4 and Fe in the glacial river Jokulsa ı Fljotsdal at Hollexceeds their corresponding absolute dissolved fluxes by�2, 1.5, and 3.8 orders of magnitude, respectively, whilethese particulate fluxes exceed their corresponding dis-solved fluxes by 0, 0.5, and 2 orders of magnitude, respec-tively for the non-glacial Fellsa river.

The degree to which the particulate flux is bioavailable isrelated to the release rate of the particulate bound elementsto seawater once they arrive in the ocean. A number ofrecent studies suggest that a substantial portion of the ele-ments transported in particulate material towards theoceans is rapidly released when this material arrives in theestuaries (Stefansdottir and Gislason, 2006; Jeandel et al.,2011; Jones et al., 2012a,b; Pearce et al., 2013; Joneset al., 2014; Jeandel and Oelkers, 2015; Morin et al.,2015). Although the exact percentage of the elements pre-sent in the particulates that is eventually released to seawa-ter is currently unknown, it is clear that if only a fewpercent of the particulate bound PO4 and Fe are released,particulate transport will dominate the input of potentiallybio-available PO4 and Fe to the oceans from the continents.Assuming this is the case, the variation with climate of thebioavailable PO4 and Fe fluxes to the ocean may bestrongly linked to the particulate fluxes of these elementstowards the oceans. The degree to which these fluxes areaffected by climate depends on glacial cover. It can be seenin Fig. 10 that the climate dependence of the suspended par-ticulate Si, PO4 and Fe fluxes to the ocean is approximatelytwice that of the corresponding dissolved fluxes for the

0.0

1.0

2.0

3.0

4.0

5.0

1960 1970 1980 1990 2000 2010Rela

�ve

flux

of S

iO2

(X/X

min

)

Year

Si suspended

Si dissolved

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

1960 1970 1980 1990 2000 2010

Log

SiO

2 flu

x (m

ol/m

2/yr

)

Year

Si suspendedSi dissolved

0.0

1.0

2.0

3.0

4.0

5.0

1960 1970 1980 1990 2000 2010Rela

�ve

flux

of P

O4

(X/X

min

)

Year

PO4 suspended

PO4 dissolved

-4.0-3.5-3.0-2.5-2.0-1.5-1.0-0.50.0

1960 1970 1980 1990 2000 2010Log

PO4

flux

(mol

/m2/

yr)

Year

PO4 suspendedPO4 dissolved

0.0

1.0

2.0

3.0

4.0

5.0

1960 1970 1980 1990 2000 2010Rela

�ve

flux

of F

e (X

/Xm

in)

Year

Fe suspended

Fe dissolved

-4.0

-3.0

-2.0

-1.0

0.0

1.0

1960 1970 1980 1990 2000 2010Log

Fe fl

ux (m

ol/m

2/yr

)

Year

Fe suspendedFe dissolved

Glacial river; Jökulsá í Fljótsdal at Hóll

log

elem

enta

lflux

(mol

/m2 /

yr)

Rela

�ve

annu

al e

lem

enta

lflux

(X/X

min

)

0.00.51.01.52.02.53.03.5

1970 1980 1990 2000 2010Rela

�ve

flux

of S

iO2

(X/X

min

)

Year

Si suspended

Si dissolved

-0.8

-0.6

-0.4

-0.2

0.0

1970 1980 1990 2000 2010

Log

SiO

2 flu

x (m

ol/m

2/yr

)

Year

Si suspendedSi dissolved

0.0

0.5

1.0

1.5

2.0

2.5

3.0

1970 1980 1990 2000 2010

Rela

�ve

flux

of P

O4 (X

/Xm

in)

Year

PO4 suspended

PO4 dissolved

-4.0

-3.5

-3.0

-2.5

-2.0

1970 1980 1990 2000 2010

Log

PO4

flux (

mol

/m2/

yr)

Year

PO4 suspendedPO4 dissolved

0.00.51.01.52.02.53.03.5

1970 1980 1990 2000 2010

Rela

�ve

flux o

f Fe

(X/X

min

)

Year


-4.0

-3.0

-2.0

-1.0

0.0

1970 1980 1990 2000 2010

Log

Fe fl

ux (m

ol/m

2/yr

)

Year


log

elem

enta

lflux

(mol

/m2 /

yr)

Rela

�ve

annu

al e

lem

enta

lflux

(X/X

min

)

Non-glacial river; Fellsá

Fig. 10. Temporal variation of relative annual dissolved and suspended Si, PO4 and Fe flux, normalized to its lowest corresponding value, andthe logarithm of these fluxes in the glacial river Jokulsa ı Fljotsdal and the non-glacial river Fellsa.


glacial Jokulsa ı Fljotsdal at Holl. In contrast, the climatedependence of the particulate fluxes of these elements inthe non-glacial Fellsa river is similar to that of the dissolvedfluxes.

4.4. The effect of glaciers on chemical transport to the oceans

This study provides further insight into the effect of gla-ciers on weathering. The three glacial and three non-glacialriver catchments considered in this study havenear-identical lithologies, and climatic conditions. As such,these rivers are ideally suited to constrain the effects of gla-cial melting on chemical transport to the oceans. A numberof substantial differences between the fluxes of dissolvedand particulate elements of the studied glacial- versusnon-glacial rivers are evident. First, the climate effect onthe dissolved element fluxes of individual elements variesfar more in glacial compared to non-glacial rivers. Thiseffect stems mainly from the dilution of glacial river watersdue to the lower extent of water–rock interaction (Gislasonet al., 2009; see also Figs. A1 and A2 in the electronic sup-plement) – glacial melt waters tends to enter the river chan-nel with little soil interaction whereas water from rain andsnowfall experiences substantial water–rock interaction asit passes through the soil on its way to the river channel.The overall effect of this increased dilution by glacial runoffis to attenuate the climate dependence of the dissolvedfluxes of the more soluble elements compared to the insol-uble elements. As a consequence, the climate dependence ofthe dissolved fluxes of the more soluble elements tends to belower in glacial rivers than in non-glacial rivers, whereas theclimate dependence of the dissolved fluxes of the insolubleelements, including most of the limiting nutrients, is similarin the glacial and non-glacial rivers. Second, consistent withpast studies (e.g. Hallet et al., 1996; Gislason et al., 1996; Liet al., 2012) this study illustrates that the fraction of the

total elemental flux transported to the oceans in suspendedparticulate material is substantially greater in the glacialversus non-glacial rivers. Third, the climate dependence ofthe particulate flux in glacial rivers is far greater than thatof non-glacial rivers.

Taken together, these observations reveal the stronginfluence of glacial melting on the global cycles of the ele-ments and the feedback between climate and Earth surfaceweathering. The strongest feedback between climate andthe elemental fluxes will be observed for the insoluble ele-ments, which are primarily transported by suspended parti-cles in the glacial rivers. The feedback between climate andthe suspended particle flux of insoluble elements, includingmany biolimiting nutrients, appears to be 2–4 times greaterin glacial rivers than their corresponding dissolved flux innon-glacial rivers. It follows that the suspended particletransport of insoluble nutrients and trace elements frommelting glaciers at the end of glacial epochs may play a crit-ical role in the long-term moderation of global climate.

5. CONCLUSION

The results of this study demonstrate directly the linkbetween climate and the fluxes of nutrients, and trace andmajor element towards the ocean in six river catchmentsin Eastern Iceland. Climate change affects the various ele-mental fluxes differently. In general, the more soluble theelement, the more it is diluted as runoff increases.Therefore, the dissolved fluxes of trace elements, which tendto be insoluble, are affected to a larger degree by changes intemperature and runoff than those of the soluble major ele-ments. This effect tends to be more pronounced in the gla-cial rivers, where a substantial fraction of the fluid flowsdirectly into the river channels, limiting the duration ofwater–rock interaction. Similarly, the glacial rivers showthe greatest temperature and runoff dependence of the


particulate flux towards the oceans. The increase in particlefluxes to the oceans with increasing average air temperatureand runoff is approximately twice that observed for the cor-responding dissolved fluxes in the glacial rivers. In contrast,the climate dependence of the dissolved and particulatefluxes is similar for the non-glacial rivers. The results of thisstudy thus demonstrates the major role of glacial meltingon the feedback between climate and nutrient fluxes tothe ocean via primary productivity followed by organic car-bon burial.

The results of the study also demonstrate that the fluxesof nutrients, trace, and major elements towards the oceanhave increased substantially in the rivers of EasternIceland over the 40-year study period in response to globalwarming and increased runoff. These rivers were pristinebefore the construction of dams in the glacier fed riversafter the year 2004 and their catchments sparingly inhabitedand cultivated. The results reported here are, therefore,likely representative of natural Earth surface processes.As such, these results may prove to be an effective baselinefor assessing the effect of climate on dissolved riverine fluxesover geological timescales.

ACKNOWLEDGMENTS

This study was funded by the Icelandic Power Company,Landsvirkjun, The Ministry for the Environment and NaturalResources and the Research Fund of the University of Iceland.In particular we would like to thank Hakon Adalsteinsson,Sigmundur Freysteinsson, Oli Gretar Blondal Sveinsson, HelgiJensson and Gunnar Steinn Jonsson for keeping this project goingfor such a long time. We would like to thank our collaborators atthe Hydrological Service at the Meteorological Office in Iceland,the Institute of Earth Sciences and CNRS in Toulouse, especiallyArni Snorrason, Jorunn Harðardottir, Sverrir Elefsen, BergurSigfusson, Svanur Palsson, Ingvi Gunnarsson, Svava B. Þorlaksdottir, Iwona M. Galeczka and Rebecca A. Neely for their help andsupport during the course of this study.

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.gca.2015.06.005.

REFERENCES

Adalsteinsson H. (2000) Aurframburdur a Eyjabokkum. OS-2000/071, National Energy authority. In Icelandic.

Amiotte-Suchet P. and Probst J. L. (1993) Modelling of atmo-spheric CO2 consumption by chemical weathering of rocks:application to the Garonne, Congo, and Amazon Basins. Chem.

Geol. 107, 205–210.Anderson S. P. (2005) Glaciers show direct linkage between erosion

rate and chemical weathering fluxes. Geomorphology 67, 147–157.

Anderson S. P. (2007) Biogeochemistry of glacial landscapesystems. Annu. Rev. Earth Planet. Sci. 35, 375–399.

Arthur M. A., Dean W. E. and Pratt L. M. (1988) Geochemicaland climatic effects of increased marine organic-carbon burialat the Cenomanian Turonian boundary. Nature 335, 714–777.

Bains S., Norris R. D., Cornfield R. M. and Faul K. L. (2000)Termination of global warmth at the Palaeocene/Eoceneboundary through productivity feedback. Nature 407, 171–174.

Berner R. A. (1982) Burial of organic-carbon and pyrite sulphur inthe modern ocean – its geochemical and environmental signif-icance. Am. J. Sci. 282, 451–473.

Berner R. A. (2004) The Phanerozoic Carbon Cycle. OxfordUniversity Press; New York, NY.

Berner R. A. and Caldeira K. (1997) The need for mass balanceand feedback in the geochemical carbon cycle. Geology 25, 955–956.

Berner R. A. and Kothavala Z. (2001) GEOCARB III. A revisedmodel of atmospheric CO2 over Phanerozoic time. Am. J. Sci.

301, 182–204.Berner R. A. and Raiswell R. (1983) Burial of organic-carbon and

pyrite sulphur in sediments over phanerozoic time – a newtheory. Geochim. Cosmochim. Acta 47, 855–862.

Bjornsson H., Sveinbjornsdottir A. E., Danıelsdottir A. K.,Snorrason A., Sigurðsson B. D., Sveinbjornsson E.,Viggosson G., Sigurjonsson J., Baldursson S., ÞorvaldsdottirS. and Jonsson T. (2008) Hnattr�nar loftslagsbreytingar ogahrif þeirra a Islandi – Skyrsla vısindanefndar um loftslags-breytingar. Umhverfisraðuneytið 118 pp.

Burdige D. J. (2005) Burial of terrestrial organic matter in marinesediments: a reassessment. Global Biogeochem. Cycles 19,GB4011.

Calmels D., Gali A., Hovius N., Bickle M., West A. J., Chen M.-C.and Chapman H. (2011) Contribution of deep groundwater tothe weathering budget in a rapidly eroding mountain belt,Taiwan. Earth Planet. Sci. Lett. 202, 48–58.

Cullen J. T., Lane T. W., Morel F. M. M. and Sherrell R. M.(1999) Modulation of cadmium uptake in phytoplankton byseawater CO2 concentration. Nature 402, 165–167.

Dessert C., Dupre B., Francois L. M., Schott J., Gaillardet J.,Chakrapani G. and Bajpai S. (2001) Erosion of Deccan Trapsdetermined by river geochemistry: Impact on the global climateand the 87Sr/86Sr ratio of seawater. Earth Planet. Sci. Lett. 188,459–474.

Dessert C., Dupre B., Gaillardet J., Francois L. and Allegre C. J.(2003) Basalt weathering laws and the impact of basaltweathering on the global carbon cycle. Chem. Geol. 202, 257–273.

Egge J. K. and Aksnes D. L. (1992) Silicate as regulating nutrientin phytoplankton competition. Mar. Ecol. Prog. Ser. 83, 281–289.

Eiriksdottir E. S., Louvat P., Gislason S. R., Oskarsson N. O. andHardardottir J. (2008) Temporal variation of the chemical andmechanical weathering in NE-Iceland, evaluation of a steady-state model of erosion. Earth Planet. Sci. Lett. 272, 78–88.

Eiriksdottir E. S., Gislason S. R. and Oelkers E. H. (2013) Doestemperature or runoff control the feedback between chemicaldenudation and climate? Insights from NE Iceland. Geochim.

Cosmochim. Acta 107, 65–81.Eiriksdottir E. S., Gislason S. R., Snorrason A., Hardardottir J.,

Thorlaksdottir S. B., Sveinbjornsdottir A. E. and Neely, R. A.(2014) Chemical composition, discharge and suspended sedi-ments in rivers in Eastern Iceland. Database of the Institute ofEarth Science and the Meteorological Office, Iceland. RH-05-2014. A report in Icelandic, 126 pp.

Falkowski P. G. (1997) Evolution of the nitrogen cycle and itsinfluence on the biological sequestration of CO2 in the ocean.Nature 387, 272–275.

Falkowski P. G. (2004) Biogeochemistry of primary production inthe sea. In Biogeochemistry (ed. W. H. Schlesinger) vol. 8.Treatise on Geochemistry (eds. H. D. Holland and K. K.Turekian). Elsevier – Pergamon, Oxford. pp. 185–213.



http://refhub.elsevier.com/S0016-7037(15)00388-9/h0010





































































Falkowski P. G., Barber R. T. and Smetacek V. (1998)Biogeochemical controls and feedbacks on ocean primaryproduction. Science 281, 200–206.

Ferrier K. L., Kirchner J. W. and Finkel R. C. (2012) Weakinfluence of climate and mineral supply rates on chemicalerosion rates: measurements along two altitudinal transects inthe Idaho Batholith. J. Geophys, Res. Earth Surf. 117, F02026.

France-Lanord C. and Derry L. A. (1997) Organic carbon burialand he forcing of the carbon cycle from Himalayan erosion.Nature 390, 65–67.

Gaillardet J. and Galy A. (2008) Himalaya – carbon sink orsource?. Science 320 1727–1728.

Gaillardet J., Dupre B., Louvat P., Allegre C. J. and Negrel P.(1999) Global silicate weathering and CO2 consumption ratesdeduced from the chemistry of large rivers. Chem. Geol. 159, 3–30.

Galloway J. N. (2004). The global nitrogen cycle. InBiogeochemistry (ed. W. H. Schlesinger) vol. 8. Treatise on

Geochemistry (eds. H. D. Holland and K. K. Turekian).Elsevier – Pergamon, Oxford. pp. 557–583.

Galy A., France-Lanord C., Beyssac O., Faure P., Kudrass H. andPalhol F. (2007) Efficient organic carbon burial in the Bengalfan sustained by the Himalayan erosional system. Nature 450,407–410.

Gislason S. R. (2008) Weathering in Iceland. Jokull 58, 387–408.Gislason S. R., Arnorsson S. and Armannsson H. (1996) Chemical

weathering of basalt as deduced from the composition ofprecipitation, rivers, and rocks in SW Iceland. Am. J. Sci. 296,837–907.

Gislason S. R., Snorrason A., Sigfusson B., Eiriksdottir E. S.,Elefsen S. O., Hardardottir J., Gunnarsson A., Hreinsson E. O.,Torssander P., Oskarsson N. and Oelkers E. (2004) Chemicalcomposition, discharge and suspended sediments in rivers inEastern Iceland. Database of the Institute of Earth Science andthe Meteorological Office, Iceland. RH-05-2004. A report inIcelandic, 101 pp.

Gislason S. R., Oelkers E. H. and Snorrason A. (2006) The role ofriver suspended material in the global carbon cycle. Geology 34,49–52.

Gislason S. R., Oelkers E. H., Eiriksdottir E. S., Kardjilov M. I.,Gisladottir G., Sigfusson B., Snorrason A., Elefsen S.,Hardardottir J., Torssander P. and Oskarsson N. (2009)Direct evidence of the feedback between climate and weather-ing. Earth Planet. Sci. Lett. 277, 213–222.

Guy H. P. and Norman V. W. (1970) Techniques of Water-

Resources Investigations of the United States Geological Survey.

Field Methods for Measurement of Fluvial Sediment 3 (C2).United States Government Printing Office, Washington, DC.

Hallet B., Hunter L. and Bogen J. (1996) Rates of erosion andsediment evacuation by glaciers: a review of field data and theirimplications. Global Planet. Change 12, 213–235.

Hards V. L., Kempton P. D., Thompson R. N. and Greenwood P.B. (2000) The magmatic evolution of the Snaefell volcaniccentre; an example of volcanism during incipient rifting inIceland. J. Volcanol. Geother. Res. 99, 97–121.

Hartmann J., Moosdorf N., Lauerwald R., Hinderer M. and WestA. J. (2014) Global chemical weathering and associated P-release — The role of lithology, temperature and soil properties.Chem. Geol. 363, 145–163.

Hilton R. G., Galy A., Hovius N., Kao S.-J., Horng M.-J. andChen H. (2012) Climatic and geomorphic controls on theerosion of terrestrial biomass from subtropical mountain forest.Global Biogeochem. Cycles 26, GB3014.

Hodson A., Mumford P. and Lister D. (2004) Suspended sedimentand phosphorus in proglacial rivers: bioavailability and

potential impacts upon the P status of ice-marginal receivingwaters. Hydrol. Process. 18, 2409–2422.

Hood E. and Berner L. (2009) Effects of changing glacial coverageon the physical and biogeochemical properties of coastalstreams in southeastern Alaska. J. Geophys. Res. 114, GO3001.

IPCC 2007: Summary for Policymakers. In Climate Change 2007:

The Physical Science Basis. Contribution of Working Group I to

the Fourth Assessment Report of the Intergovernmental Panel on

Climate Change (eds. S. Solomon, D. Qin, M. Manning, Z.Chen, M. Marquis, K. B. Averyt, M. Tignor and H. L. Miller).Cambridge University Press, Cambridge, United Kingdom andNew York, NY, USA. http://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_report_wg1_report_the_physical_science_basis.htm.

Jeandel C. and Oelkers E. H. (2015) The influence of terrigenousparticulate material dissolution on ocean chemistry and globalelement cycles. Chem. Geol. 395, 50–66.

Jeandel C., Peucker-Ehrenbrink B., Jones M. T., Pearce C. R.,Oelkers E. H., Godderis Y., Lacan F., Aumont O. and ArsouzeT. (2011) Ocean margins: the missing term for oceanic elementbudgets? EOS Trans. Am. Geophys. Union 92, 217–224.

Jickells T. D. (1998) Nutrient biogeochemistry of the coastal zone.Science 10, 217–222.

Jickells T. D., An Z. S., Andersen K. K., Baker A. R., BergamettiG., Brooks N., Cao J. J., Boyd P. W., Duce R. A., Hunter K.A., Kawahata H., Kubilay N., LaRoche J., Liss P. S.,Mahowald N. M., Prospero J., Ridgwell A. J., Tegen I. andTorres R. (2005) Global iron Connections between desert dust,ocean biogeochemistry, and climate. Science 308, 67–71.

Johannesson H. and S�mundsson K. (1998) Geological Map of

Iceland – Bedrock Geology. The Icelandic Institute of NaturalHistory, Reykjavik, Iceland.

Jones M. T., Pearce C. R. and Oelkers E. H. (2012a) Anexperimental study of the interaction of basaltic riverineparticulate material and seawater. Geochim. Cosmochim. Acta

77, 108–120.Jones M., Pearce C. R., Jeandel C., Gislason S. R., Eiriksdottir E.

S., Mavromatis V. and Oelkers E. H. (2012b) Riverineparticulate material dissolution as a significant flux of strontiumto the oceans. Earth Planet. Sci. Lett. 355–356, 51–59.

Jones M. T., Gislason S. R., Burton K. W., Pearce C. R.,Mavromatis V., Pogge von Strandmann P. A. E. and Oelkers E.H. (2014) Quantifying the impact of riverine particulatedissolution in seawater on ocean chemistry. Earth Planet. Sci.

Lett. 395, 91–100.Kardjilov M. I. (2008) Riverine and terrestrial carbon fluxes in

Iceland. Doctoral Thesis from the Department of Geographyand Tourism, University of Iceland, Reykjavik. 83 pp.

Kardjilov M. I., Gisladottir G. and Gislason S. R. (2006) Landdegradation in northeastern Iceland: present and past carbonfluxes. Land Degrad. Dev. 17, 401–417.

Labat D., Godderis Y., Probst J. L. and Guyot J. L. (2004)Evidence for global runoff increase related to climate warming.Adv. Water Res. 27, 631–642.

Lane T. W. and Morel F. M. M. (2000a) Regulation of Carbonicanhydrase expression by zinc, cobalt, and carbon dioxide in themarine diatom Thalassiosira weissflogii. Plant Physiol. 123, 345–352.

Lane T. W. and Morel F. M. M. (2000b) A biological function forcadmium in marine diatoms. Proc. Natl. Acad. Sci. 97, 4627–4631.

Li Z., Gao W., Zhang M. and Gao W. (2012) Variations insuspended and dissolved matter fluxes from glacial and non-glacial catchments during a melt season at Urumqi River,eastern Tianshan, central Asia. Catena 95, 42–49.





























































http://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_report_wg1_report_the_physical_science_basis.htm





















































Martin J. H., Coale K. H., Johnson K. S., Fitzwater S. E., GordonR. M., Tanner S. J., Hunter C. N., Elrod V. A., Nowicki J. L.,Coley T. L., Barber R. T., Lindley S., Watson A. J., Van ScoyK., Law C. S., Liddicoat M. I., Ling R., Stanton T., Stockel J.,Collins C., Anderson A., Bidigare T., Ondrusek M., Latasa M.,Millero F. J., Lee K., Yao W., Zhang J. Z., Friederich G.,Sakamoto C., Chavez F., Buck K., Kolber Z., Greene T.,Falkowski P., Chisholm S. W., Hoge F., Swift R., Yungel J.,Turner S., Nightingale P., Hatton A., Liss P. and Tindale N. W.(1994) Testing the iron hypothesis in ecosystems of theequatorial Pacific Ocean. Nature 371, 123–129.

Milliman J. D. and Farnsworth K. L. (2011) River Discharge to the

Coastal Ocean: A Global Synthesis. Cambridge University, UK,384 pp.

Milliman J. D. and Syvitski J. P. M. (1992) Geomorphic/tectoniccontrol of sediment discharge to the ocean: the importance ofsmall mountainous rivers. J. Geol. 100, 525–544.

Morel R. M. M. and Price N. M. (2003) The biogeochemical cyclesof trace metals in the oceans. Science 300, 944–947.

Morin G. P., Vigier N. and Verney-Carron A. (2015) Enhanceddissolution of basaltic glass in brackish waters: impact onbiogeochemical cycles. Earth Planet. Sci. Lett. 417, 1–8.

Oelkers E. H., Gislason S. R., Eiriksdottir E. S., Jones M., PearceC. R. and Jeandal C. (2011) The role of riverine particulatemateriel on the global cycles of the elements. Appl. Geochem.

26, S365–S369.Oelkers E. H., Jones M., Pearce C. R., Jeandal C., Eiriksdottir E.

S. and Gislason S. R. (2012) Riverine particulate materialdissolution in seawater and its implications for the global cyclesof the elements. C.R. Geosci. 344, 646–651.

Olafsson J., Olafsdottir S. R. and Briem J. (2008) Vatnsfoll ogvistkerfi strandsjavar. Natturufr�ðingurinn 76, 95–108. InIcelandic.

Opfergelt S., Burton K. W., Pogge von Strandmann P. A. E.,Gislason S. R. and Halliday A. N. (2013) Riverine siliconisotope variations in glaciated basaltic terrains: implications forthe Si delivery to the ocean over glacial–interglacial intervals.Earth Planet. Sci. Lett. 369–370, 211–2019.

Palsson S. and Vigfusson G. H. (1996) Results of suspended loadand discharge measurements 1963–1995. Reykjavik, NationalEnergy Authority, OS-96032/VOD-05 B.

Pearce C. R., Jones M. T., Oelkers E. H., Pradoux C. and JeandelC. (2013) The effect of particulate dissolution on theneodymium (Nd) isotope and Rare Earth Element (REE)composition of seawater. Earth Planet. Sci. Lett. 369–370, 138–147.

Raiswell R. and Canfield D. E. (2012) The iron biogeochemicalcycle past and present. Geochem. Perspect. 1, 1–220.

Raven J. A. and Falkowski P. G. (1999) Oceanic sinks foratmospheric CO2. Plant, Cell Environ. 22, 741–755.

Ruttenberg K. C. (2004) The global phosphorus cycle. InBiogeochemistry (ed. W. H. Schlesinger) vol. 8. Treatise on

Geochemistry (eds. H. D. Holland and K. K. Turekian).Elsevier – Pergamon, Oxford. pp. 585–643.

Sinkko H., Lukkari K., Sihvonen L. M., Sivonen K., Leivuori M.,Rantanen M., Paulin L. and Lyra C. (2013) Bacteria contributeto sediment nutrient release and reflect progressed eutrophica-tion-driven hypoxia in an organic-rich continental sea. PLoS

ONE 8, 1–14.Smith S. V. (2013) Parsing the Oceanic Calcium Carbonate Cycle: A

Net Atmospheric Carbon Dioxide Source, or a Sink? L&O e-

books. Association for the Sciences of Limnology and

oceanography (ASLO), Waco, TX. http://dx.doi.org/10.4319/svsmith.2013.978-0-984591-2-1.

Statham P. J. (2012) Nutrients in estuaries – an overview and thepotential impacts of climate change. Sci. Total Environ. 434,213–227.

Stefansdottir M. and Gislason S. R. (2006) Suspended basalticglass–seawater interactions. J. Geochem. Explor. 88, 332–335.

Stumm W. and Morgan J. J. (1996) Aquatic Chemistry. Chemical

Equilibria and Rates in Natural Waters, 3rd ed. John Wiley &Sons Inc, 1022 p.

The Hydrological Services of the National Energy Authority(2005a) Location and boundaries of river catchments in NE-Iceland.

The Hydrological Services of the National Energy Authority(2005b) Database of the Hydrological Service 2005/26.

The Icelandic Institute of Natural History (1977) Geological map,pages 1–9. Vector 1:250.000.

The Icelandic Meteorological Office (2007) Annual Average Climate

Information 1961–2005. The Icelandic Meteorological Office,Reykjavık, http://andvari.vedur.is/vedurfar/yfirlit/medaltal-stoflur/Arsgildi.html2007.

Tipper E. T., Bickle M. J., Galy A., West A. J., Pomies andChapaman H. J. (2006) The short term climatic sensitivity ofcarbonate and silicate weathering fluxes: insight from seasonalvariations in river chemistry. Geochim. Cosmochim. Acta 70,2737–2754.

Tomasson H. (1990) Suspended material in Icelandic rivers. In (ed.S. Guttormur) Vatnid og Landid: Reykjavik, Orkustofnun. pp.169–174.

Tomasson H., Palsson S., Vigfusson G. H. and Hafstad T. (1996)Framburður Þjorsar við Þjorsarver. Botnskrið og svifaur 13646Orkustofnun OS-96010/VOD-03 B. 29 pp.

Tranter M. (2003) Geochemical weathering in glacial andproglacial environments. In Treatise on Geochemistry (eds.Heinrich D. Holland and Karl K. Turekian). vol. 5, pp. 189–205.

Trenberth K. E. (2007) Changes in precipitation with climatechange. Climate Res. 47, 123–138.

Wada K., Arnalds O., Kakuto Y., Wilding L. P. and Hallmark C.T. (1992) Clay minerals of four soils formed in eolian andtephra materials in Iceland. Geoderma 52, 351–365.

Walker J. C. G., Hays P. B. and Kasting J. R. (1981) A negativefeedback mechanism for the long-term stabilization of Earth’ssurface temperature. J. Geophys. Res. 86, 9776–9782.

Wallmann K. (2001) Controls on the Cretaceous and Cenozoicevolution of seawater composition, atmospheric CO2 andclimate. Geochim. Cosmochim. Acta 65, 3005–3025.

White D. (1999) The Physiology & Biogeochemistry of Prokaryotes,2nd ed. Oxford University Press, Oxford.

White A. F. and Blum A. E. (1995) Effects of climate on chemicalweathering rates in watersheds. Geochim. Cosmochim. Acta 59,1729–1747.

Wood D. A. (1976) Spatial and temporal variation in the traceelement geochemistry of the Eastern Iceland flood basaltsuccession. J. Geophys. Res. 81, 4353–4360.

Wood D. A. (1978) Major and trace element variations in theTertiary lavas of Eastern Iceland and their significance withrespect to the Iceland geochemical anomaly. J. Petrol. 19, 393–436.

Associate editor: Jerome Gaillardet



















































http://dx.doi.org/10.4319/svsmith.2013.978-0-984591-2-1

http://dx.doi.org/10.4319/svsmith.2013.978-0-984591-2-1









http://andvari.vedur.is/vedurfar/yfirlit/medaltalstoflur/Arsgildi.html2007

http://andvari.vedur.is/vedurfar/yfirlit/medaltalstoflur/Arsgildi.html2007






























Direct evidence of the feedback between climate and ... · Direct evidence of the feedback between...

Documents

Transcript of Direct evidence of the feedback between climate and ... · Direct evidence of the feedback between...