139
Air-water carbon dioxide exchange in relation to chemical and physical characteristics of the Churchill River and estuary, southwestern Hudson Bay region by Emmelia C. Stainton A Thesis submitted to the Faculty of Graduate Studies of The University of Manitoba in partial fulfilment of the requirements of the degree of MASTER OF SCIENCE Department of Environment and Geography University of Manitoba Winnipeg Copyright © 2009 by Emmelia C. Stainton

Air-water carbon dioxide exchange in relation to chemical

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Air-water carbon dioxide exchange in relation to chemical

Air-water carbon dioxide exchange in relation to chemical and

physical characteristics of the Churchill River and estuary,

southwestern Hudson Bay region

by

Emmelia C. Stainton

A Thesis submitted to the Faculty of Graduate Studies of

The University of Manitoba

in partial fulfilment of the requirements of the degree of

MASTER OF SCIENCE

Department of Environment and Geography

University of Manitoba

Winnipeg

Copyright © 2009 by Emmelia C. Stainton

Page 2: Air-water carbon dioxide exchange in relation to chemical

1*1 Library and Archives Canada

Published Heritage Branch

395 Wellington Street Ottawa ON K1A 0N4 Canada

Bibliotheque et Archives Canada

Direction du Patrimoine de I'edition

395, rue Wellington Ottawa ON K1A 0N4 Canada

Your file Votre r6f6rence ISBN: 978-0-494-63937-5 Our file Notre reference ISBN: 978-0-494-63937-5

NOTICE: AVIS:

The author has granted a non­exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or non­commercial purposes, in microform, paper, electronic and/or any other formats.

The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduce, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'Internet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats.

L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation.

In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis.

While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis.

Conformement a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these.

Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant.

1+1

Canada

Page 3: Air-water carbon dioxide exchange in relation to chemical

THE UNIVERSITY OF MANITOBA

FACULTY OF GRADUATE STUDIES

COPYRIGHT PERMISSION

Air-Water Carbon Dioxide Exchange in Relation to Chemical

and Physical Characteristics of the Churchill River and Estuary,

Southwestern Hudson Bay Region

By

Emmelia C. Stainton

A Thesis/Practicum submitted to the Faculty of Graduate Studies of The University of

Manitoba in partial fulfillment of the requirement of the degree

Of

Master of Science

Emmelia C. Stainton©2009

Permission has been granted to the University of Manitoba Libraries to lend a copy of this thesis/practicum, to Library and Archives Canada (LAC) to lend a copy of this thesis/practicum, and to LAC's agent (UMI/ProQuest) to microfilm, sell copies and to publish an abstract of this

thesis/practicum.

This reproduction or copy of this thesis has been made available by authority of the copyright owner solely for the purpose of private study and research, and may only be reproduced and copied

as permitted by copyright laws or with express written authorization from the copyright owner.

Page 4: Air-water carbon dioxide exchange in relation to chemical

i

Abstract

Work described in this thesis has increased our understanding of high

latitude river/estuarine air-water carbon dioxide (CO2) exchange through

examination of the Churchill River and estuary system of southwestern Hudson Bay,

which was studied throughout the 2007 open-water season. An automated

monitoring instrument was used to measure the concentrations of CO2 in air and

water of the Churchill River and estuary. River and estuarine waters were sampled

and analyzed to determine differences in CO2 flux with respect to season, space, and

source (river or marine). Riverine CO2 flux ranged from -51.4 to +30.8 mmol nr2

day1 (mean flux of-5.8 mmol nr2 day -1), while estuarine flux ranged from -11.04 to

+22.76 mmol nr2 day1 (mean fluxes of -1.07 at low tide, +0.06 at high tide, and -0.37

mmol nr2 day1 overall). These results indicate that the Churchill River and estuary

are net CO2 sinks throughout the 2007 open-water season. Analyses of water

samples collected from the Churchill River (surface waters) and estuary (surface

and bottom waters) indicate that the river is a source of chlorophyll a, dissolved

organic carbon, coloured dissolved organic matter, and particulate organic carbon

and nitrogen to the estuary basin, while marine waters are a source of dissolved

inorganic carbon, salinity, and alkalinity. A multi-sensor probe was used to

determine estuarine stratification and seasonal mixing patterns. The Churchill River

estuary was stratified by water source, with fresh river water flowing over brackish

marine water. The nature of the stratification varied seasonally with temperature

and changing river discharge and tidal amplitude.

Page 5: Air-water carbon dioxide exchange in relation to chemical

11

Acknowledgments

I thank my supervisors, Drs. Tim Papakyriakou and David Barber, and my committee members, Drs. Ray Hesslein and Fei Wang for their guidance and support.

I also thank Drs. Gordon Robinson (Botany) and Brenda Hann (Zoology) for being excellent teachers throughout my undergraduate degree, and encouraging me to pursue aquatic research.

I would also like to thank several organizations for their support: the Natural Sciences and Engineering Research Council of Canada, the Northern Scientific Training Program, the ArcticNet program of the Network of Centers of Excellence, and the Churchill Northern Studies Centre.

I am very thankful for the continued support and encouragement from Simon and my family. To my father, Michael, I especially thank for his continued interest and helpful discussion with respect to this project.

Page 6: Air-water carbon dioxide exchange in relation to chemical

Table of Contents

1 - Introduction 1

Motivation 1

Objectives 3

Thesis outline 4

2 - High latitude river-estuarine-shelf CO2 exchange 6

Introduction 6

Role of the surface waters in determining CO2 flux 8

Processes that affect [CCbJw 9

2.3.1 Chemical Processes 9

2.3.2 Physical processes 12

2.3.3 Biological processes 13

2.3.4 Summary 18

Gas exchange at water surfaces 19

2.4.1 Solubility 20

2.4.2 Stagnant Film Model 21

2.4.3 Determination of gas transfer velocity, k 24

2.4.4 Other considerations for gas transfer 26

Carbon exchange at the land-ocean interface 28

2.5.1 Rivers 29

2.5.2 Estuaries 32

2.5.3 Continental Shelves 35

Page 7: Air-water carbon dioxide exchange in relation to chemical

2.6 CO2 transport from coastal systems

2.7 Climate implications 37

2.8 Summary 39

Chapter 3 - Summer air-water carbon dioxide exchange and water

chemistry of the Churchill River as it enters Hudson Bay 46

3.1 Introduction 47

3.2 Methods 51

3.2.1 Discrete water sampling 52

3.2.2 Continuous CO2, O2, and water sampling 54

3.2.3 Other sources of data 56

3.2.4 Laboratory procedures 57

3.2.5 Calculations 58

3.3 Results 61

3.3.1 Environmental conditions 61

3.3.2 Water chemistry 64

3.3.3 Continuous CO2,02, and water sampling 66

3.3.4 C02flux 68

3.3.5 Production and respiration 70

3.4 Discussion 72

Chapter 4 - Air-water carbon dioxide exchange and physiochemical

processes of the Churchill River estuary 82

4.1 Introduction 83

4.2 Methods 86

Page 8: Air-water carbon dioxide exchange in relation to chemical

4.2.1 Study Area 86

4.2.2 Estuary survey sampling 87

4.2.3 Estuary basin CTD profiling 90

4.2.4 Other sources of data 91

4.2.5 Laboratory and analytical procedures 92

4.2.6 Calculations 93

Results 94

4.3.1 Hydrographic conditions 94

4.3.2 Physical conditions 94

4.3.3 Water chemistry 97

4.3.4 Air/water pC02 and C02 flux 104

4.3.5 pCC»2 concentration at depth 106

Discussion 109

4.4.1 Influence of season, location, depth, and tidal state on water chemistry 109

4.4.2 Influence of water source and physiochemical

conditions on pCCh and CO2 flux 110

4.4.3 Comparison to other high latitude estuaries 97

4.3.4 Air/water pC02 and C02 flux 114

Summary 115

5 - Conclusions 119

Summary and Conclusions 119

Future Directions 120

Page 9: Air-water carbon dioxide exchange in relation to chemical

A - Latitude and longitude coordinates of all stations sampled within the Churchill River and estuary, June-October 2007. 122

B - Figure 1: Relationship between gas trasfer velocity, k, and wind speed from four different models: Wanninkhof (1992), Wanninkhof and McGillis (1999), Nightingale et al. (2000), and Liss and Merlivat (1986). The Wanninkhof (1992) model for calculating gas transfer was used for flux calculations in this study. 123

Table 1: Mean, standard deviation, and range of CO2 flux measurements based on four models for calculating gas transfer, k. The Wanninkhof (1992) model for calculating gas transfer was used for flux calculations in this study. 123

Page 10: Air-water carbon dioxide exchange in relation to chemical

List of Tables

1 Commonly used gas transfer velocity parameterizations. 25

1 Mean and standard deviation of Churchill River annual flow and precipitation before and after the Churchill River Diversion (CRD, created in 1976) and before and after the Churchill River rock weir construction (10km south of the Town of Churchill, built in 1998). 61

2 Summary of water chemistry data from Churchill River site CH7 collected June-October 2007. 64

3 Mean, standard deviation and range of (i) all measured gas monitor data (n = 224), (ii) flux (n = 224), and (iii) net production, respiration, and gross production at 2.5m (Pnet/down-slope peak, n = 27; R/upslope peak, n = 26; Pgross/complete peak, n = 24) from Churchill River station CH7. 67

4 Summary of flux estimates made by Tremblay et al. (2006) for a Manitoba River and some northern Quebec hydroelectric reservoirs. Samples were collected from a 20L chamber connected to a non-dispersive infrared gas analyzer, and data was stored every 20s over a 5-10 minute sample period. 71

1 Water chemistry of Churchill River estuary waters collected from stations along a fresh (CH5) to marine (CHI) gradient, at both high tide (HT) and low tide (LT), from surface (0.5m) and bottom waters (3 and 10m), from June to October 2007. Surface water samples were also collected from the inflowing Churchill River (CH7) upstream of the weir (Stainton et al., in prep. 2009). 100

2 Pearson's Product Moment Correlation (r) between C02w (umol l1) and water chemistry variables. 113

3 Range of flux estimates and water pC02 from several mid- to high-latitude estuaries/coastal seas. 114

Page 11: Air-water carbon dioxide exchange in relation to chemical

List of Figures

Carbon cycling and processes in aquatic systems (Ducklow and McCallister, 2004). pC02 in surface waters can increase/decrease through gas exchange, photosynthesis/respiration from aquatic organisms, precipitation/dissolution reactions with CaC03, and decomposition of organic material by heterotrophs. Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA - IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.287, Cambridge, Mass.: Harvard University Press, Copyright (c) 2006 by the President and Fellows of Harvard University. 15

Biogeochemical processes modifying Alk and DIC in aquatic systems (Zeebe and Wolf-Gladrow, 2001). This article was published in "CO2 in seawater, equilibrium, kinetics, isotopes", Zeebe and Wolf-Gladrow, Equilibrium, p. 7, Copyright Elsevier, 2001. 19

Relationship between solubility of CO2 and water temperature and salinity. Lines S=0, S=10, and S = 25 represent the relationship between water temperature (left y-axis) and solubility at salinities of 0,10, and 25 PSU. Lines T=273, T=281, and T=296 represent the relationship between salinity (right y-axis) and solubility at water temperatures of 273, 281, and 296 °K. 21

Stagnant film model (Sarmiento and Gruber, 2006). Gas transfer at the air-water interface is a relatively slow process, which results in a concentration gradient. CO2 is well mixed in the turbulent layer, uniform concentration of [CO2V and [C02]atm in the air. [C02]w and [C02]atm in the stagnant layers changes linearly to saturation levels [C02]w° and [C02]atm°. Gas transfer across the air-sea interface is governed by the rate of molecular diffusion of gas CO2 through the two stagnant films. 23

The lower Churchill River from Southern Indian Lake to Hudson Bay. Discharge for the Churchill River is gauged at Red Head Rapids. 31

6 Carbon and nutrient cycling between the ocean, atmosphere and land {Ducklow and McCallister, 2004). All three carbon

Page 12: Air-water carbon dioxide exchange in relation to chemical

pumps - biological, solubility, and coastal shelf- are show. Arrows between organic carbon and DIC indicate biological production and respiration. Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA -IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.273, Cambridge, Mass.: Harvard University Press, Copyright © 2006 by the President and Fellows of Harvard University. 37

a) Map of the Churchill River as it enters Hudson Bay. b) CO2/O2 air/water gas monitor moored on float at Churchill River station CH7. Two 20W solar panels [right side) charge two 12V batteries that power the gas monitor system (inside white box). 52

Schematic of air/water gas monitor: wet box consisting of a peristaltic water pump, a thermocouple housing, a gas exchanger, and two solenoid valves (SV); dry box consisting of an air pump, drying compartment, O2 sensor, IR gas analyzer, and data logger. 55

a) Churchill daily mean air temperature and daily precipitation (Environment Canada), and river discharge (combination of Red Head Rapids and the Deer River, Water Survey of Canada) throughout the 2007 open-water season, b) precipitation and air temperature monthly averages from 2007 and historical climate norms (1971-2000) (Environment Canada). 62

Historical flow data from the Churchill River at Red Head Rapids (Water Survey of Canada) and monthly mean precipitation (Environment Canada) between 1971-2008. 63

Churchill River summer water chemistry data collected from 1995-2005 (data from Bezte, 2006) and from 2006-2007 (this study). 65

Diurnal fluctuations of CO2W, CO2A, O2W, O2A, and Tw at Churchill River site CH7 throughout the 2007 open-water season. This data was measured and recorded at 2-hour intervals using a continuous air/water CO2/O2 gas monitor. 68

a) and b) Turbidity, salinity, water temperature and pH collected at Churchill River station CH7 (data measured and recorded using an RBR XR-620) c) Churchill River discharge (combination of Red Head Rapids and the Deer. River, Water

Page 13: Air-water carbon dioxide exchange in relation to chemical

X

Survey of Canada] and water level (recorded at the Churchill pump house station, approximately 3km upstream of station CH7). 69

3.8 Estimated CO2 flux from measured values at Churchill River site CH7 throughout the 2007 open-water season. A positive flux indicates a supersaturation of CO2 in water with respect to air (source), while a negative flux indicates an undersaturation of CO2 in water with respect to air (sink). The Churchill River is a CO2 sink throughout most of the monitored open-water season. 70

3.9 Estimated Pnet and R throughout the 2007 open water-season at Churchill River station CH7. Both Pnet and R were calculated from gas data measured using a continuous air/water gas monitor. 72

3.10 Delivery rate of C02w to the Churchill River estuary and 24-hour upstream flow distance upstream from station CH7. 73

4.1 a) Churchill River (CH7) and estuary (CH1-5) sampling stations at a) high tide and b) low tide (Landsat 7 satellite image, Natural Resources Canada). 87

4.2 Air/water gas monitoring instrument. Dry box: 1-air pump, 2-drying compartment, 3-O2 sensor, 4-gas analyzer, and 5-data logger; wet box: 6-peristaltic water pump, 7-thermocouple housing, 8- air/water gas exchanger, and 9- solenoid valves. 89

4.3 a) Churchill daily mean air temperature and precipitation (Environment Canada), b) Churchill River discharge (combination of Red Head Rapids and the Deer River, Water Survey of Canada) and water temperature (Manitoba Hydro pumphouse station), and c) Churchill harbour tide height (Fisheries and Oceans Canada), with sampling periods indicated by dashed vertical lines. 96

4.4 Depth, temperature, salinity cross-sections of Churchill River estuary CTD transects at high tide (left column) and low tide (right column) in June, July, August, and October 2007. Temperature is represented in colour (note all colour legends on the right side of figures are different), salinity is denoted by numbered contour intervals (ranging from 2-30), and individual CTD casts are shown as bold black vertical lines within the cross-sections. White circles on the high/low tide maps (bottom panels) are the CTD sampling locations. 99

Page 14: Air-water carbon dioxide exchange in relation to chemical

XI

4.5 Temperature-salinity scatter plot of all CTD casts taken within the Churchill River estuary throughout the 2007 open-water season. Data is grouped by sampling month: red = June, blue = July, green = August (light green for high tide casts, dark green for low tide casts), black = September/October. 101

4.6 Water chemistry of the Churchill River estuary between June and October, 2007. Samples were taken at 0.5m from the surface, and 0.5m from the bottom at each of 5 stations throughout the length of the estuary (ranging from fresh to brackish water), at both high and low tide. The x-axis is salinity in all four panels and the right y-axis colour legend is reflected on the data points in each separate panel. Point distribution reflects the left y-axis parameters versus salinity. 103

4.7 Water-air difference in pCCh (top panels) and air/water CO2 flux from Churchill River estuary waters (both using average air pCC»2 = 374 uatm). Samples were collected using a continuous automated CO2 monitor that averaged air and water CO2 concentrations over a 30-minute period. Stations range from fresh (CH5) to brackish/marine (CHI) waters and were sampled at high and low tide, June to October, 2007. Overall mean CO2 flux is -0.62 mmol m2 day1. 105

4.8 pC02 concentration, salinity, and temperature variation with depth (surface, mid-depth, and bottom waters) at Churchill River estuary stations CH3 (top panels) and CH4 (bottom panels). Samples were collected between June and October at high tide using a continuous automated CO2 monitor. Water was pumped from each depth for 10 minutes, data logging every 10 seconds. Data points represent an average of pC02 over the 10-minute sampling period. 107

4.9 PCO2 concentration, salinity, and temperature variation with depth (surface, mid-depth, and bottom waters) at Churchill River estuary stations CH3 (top panels) and CH4 (bottom panels). Samples were collected between June and August at low tide using a continuous automated CO2 monitor. Water was pumped from each depth for 10 minutes, data logging every 10 seconds. Data points represent an average of pC02 over the 10-minute sampling period. 108

4.10 Churchill River CO2 flux and ApC02. Water was sampled using a continuous air/water CO2 gas monitor. The dashed line indicates August 1. I l l

Page 15: Air-water carbon dioxide exchange in relation to chemical

Xll

5.1 Summary of processes within the Churchill River and estuary system. 120

Page 16: Air-water carbon dioxide exchange in relation to chemical

xiii

List of Copyright Permissions

Figure 2.1 Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA - IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.287, Cambridge, Mass.: Harvard University Press, Copyright © 2006 by the President and Fellows of Harvard University. 15

Figure 2.2 This article was published in "CO2 in seawater, equilibrium, kinetics, isotopes", Zeebe and Wolf-Gladrow, Equilibrium, p. 7, Copyright Elsevier, 2001. 19

Figure 2.6 Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA - IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.273, Cambridge, Mass.: Harvard University Press, Copyright © 2006 by the President and Fellows of Harvard University. 37

Page 17: Air-water carbon dioxide exchange in relation to chemical

1

Chapter 1: Introduction

1.1 - Motivation

Atmospheric carbon dioxide (CO2) concentrations have risen to 379 ppm [in

2005, IPCC, 2007] from the pre-industrial revolution level of 280 ± 10 ppm [IPCC,

2007]. Antarctic ice cores, which agree very well with atmospheric measurements of

CO2, have shown that the current atmospheric CO2 concentration has not been

exceeded in the past 420,000 years [IPCC, 2007]. This dramatic increase in

atmospheric CO2 is a result of anthropogenic CO2 production, primarily from fossil

fuel burning, although cement production and land use changes (e.g. biomass

clearing for agriculture, predominantly in the tropics) also release substantial

amounts of CO2 [IPCC, 2007]. Interestingly, the rates of atmospheric CO2 increase

and the estimated anthropogenic CO2 emission do not balance [IPCC, 2007]. A

"missing" 50-60% of anthropogenic CO2 emission is thought to be captured by the

global oceans and terrestrial ecosystems [Millero, 2000].

Increases in atmospheric CO2 and other greenhouse gases also linked to

industrialization (CH4, N2O, CFCs, HFCs, etc., not discussed in detail here) are driving

climate warming by their ability to absorb a portion of the IR radiation emitted from

Earth. Although greenhouse gasses are a critical regulator of the Earth's

temperature, increasing concentrations have the potential to regulate Earth's

temperature at levels above historic. The resulting linear global surface temperature

increase throughout the past 50 years is 0.1-0.16 °C per decade and global

temperatures are projected to increase by another 1.4 - 5.8°C by 2100 [IPCC, 2007].

Page 18: Air-water carbon dioxide exchange in relation to chemical

2

Estimates of temperature change for high latitude regions are quite different than

global averages, with surface temperatures having increased by an average of 2-3°C

and up to 4°C in the winter since 1950 [ACIA, 2005].

Climate warming has lead to ice cover reduction (spatial and temporal),

water temperature increases, and altered chemical and biological cycles in northern

coastal marine environments [Chen, etal, 2003]. Human activities also affect the

amount - damming, irrigation, hydroelectric power generation, river diversions -

and composition (nutrients, carbon, suspended sediments) of freshwater runoff to

northern estuarine-shelf environments, which could impact their biogeochemical

cycling and overall metabolism [Borges, 2005; Chen, etal, 2003]. All of these

changes from normal are likely to alter the rates of processes governing the

production and sequestration of CO2 in estuarine-shelf systems.

An understanding of current carbon pools and fluxes, especially CO2, within

the biosphere and within/between smaller scale environments will enhance our

ability to predict changes under higher atmospheric CO2 conditions. CO2 is

intimately connected with all subsets of the biosphere, and thus knowledge of

carbon budgets and chemical species/phase transformations within specific

environments is important for understanding both current and future cycling and

possible feedback mechanisms.

Global warming is resulting in physical, chemical, and biological

transformations in many global environments [IPCC, 2007]. This is especially of

concern for high latitude polar regions, though a dearth of scientific observations

Page 19: Air-water carbon dioxide exchange in relation to chemical

3

renders uncertain the extent to which these changes will affect high latitude

environments - and the influence this will have globally.

ArcticNet, a multi-year collaborative research project, was designed to fill

some of these knowledge gaps in the Canadian Arctic. This study is one of hundreds

undertaken as part of ArcticNet, as we attempt to understand the main pathways of

carbon and nutrients into and through Hudson Bay. Prior to this research, only a

handful of studies existed for the Churchill River - none pertaining to gas exchange.

1.2 - Objectives

The objective of this thesis is to better understand current carbon and

nutrient cycling of the high-latitude Churchill River and estuary system, to

determine whether these environments are sources or sinks for atmospheric CO2

during the open-water season, and to determine what physical, chemical, and

biological factors govern these conditions. The work and its interpretation are

presented in two parts. Part 1 determines what carbon stocks exist and how they

seasonally vary between the components of the carbon pool of the Churchill River

flowing into Hudson Bay. Part 2 determines whether the Churchill River estuary is a

source or sink for atmospheric CO2, and how seasonal physiochemical changes affect

the carbon pool within the Churchill River estuary. The methods and results of this

thesis are described separately in each of these parts.

Page 20: Air-water carbon dioxide exchange in relation to chemical

4

1.3 - Thesis Outline

Chapter two of this thesis consists of a detailed overview of carbon cycling in

aquatic systems by discussing the importance of the oceanic carbon cycle, and

provides an explanation and synthesis of 1) processes that affect CO2 in water, 2)

carbon sources and transformations in aquatic systems, and 3) how the Churchill

river and estuarine system fits within the context of these processes and

transformations, and how it compares with similar systems globally. This chapter

will explain gas exchange at water surfaces and provide a review of the currently

accepted models, determination of a gas transfer velocity for estimation of gas flux,

and mention other factors that should be considered for gas transfer estimations.

This chapter also outlines air-water CO2 exchange in rivers, estuaries, and coastal

marine systems and identifies the processes by which CO2 is transported to the deep

ocean.

Chapter three will discuss the results from the paper: Stainton, E.C., R.H.

Hesslein, T.N. Papakyriakou, and D.G. Barber. Summer air-water carbon dioxide

exchange and water chemistry of the Churchill River as it enters Hudson Bay.

Chapter four will discuss the results from paper: Stainton, E.C., T.N.

Papakyriakou, and D.G. Barber. Air-water carbon dioxide exchange and

physiochemical processes of the Churchill River estuary.

The concluding chapter of this thesis will summarize the results and outline

suggestions for future research in the Churchill River estuary and other high-

Page 21: Air-water carbon dioxide exchange in relation to chemical

latitude estuaries with the purpose of expanding our knowledge of CO2 transfer and

cycling, and influences of estuarine processes on the global oceans.

Page 22: Air-water carbon dioxide exchange in relation to chemical

6

Chapter 2: High latitude river-estuarine-shelf CO2 exchange

2.1 - Introduction

Coastal marine environments are defined as the ocean extending from

shoreline to the continental shelf break, and are among the most productive

ecosystems in the biosphere. Representing only 7% of ocean surface area and less

than 0.5% of the ocean volume, the global marine shelf environments produce 20%

of the total oceanic organic matter (OM), 80-90% of new oceanic production, 80% of

oceanic OM burial, up to 50% of denitrification, 90% of oceanic sedimentary

mineralization, 30% of the oceanic production of particulate organic carbon (POC],

and 50% of the oceanic accumulation of POC [Borges, 2005; Chen, etal, 2003]. The

high rate of production on marine continental shelves is due to riverine inputs and

coastal upwelling, both of which supply nutrients and influence water retention

times [Chen, etal, 2003]. Continental shelves have considerably higher air-water

CO2 exchange and higher respiration and photosynthetic rates (2x) than the open

ocean and as a result, accumulate more organic carbon and calcium carbonate

(CaC03) [Chen, etal, 2003].

The Arctic Ocean is quite different compared to other global oceans because

of its large proportion of shelf relative to ocean area (30% shelf), cooler water

temperatures, its significant freshwater inputs from terrestrial runoff, and the

presence of sea ice. Although several similarities exist between arctic shelves and

others globally - higher primary production, terrestrial influence, shelf stratification

Page 23: Air-water carbon dioxide exchange in relation to chemical

7

from freshwater runoff, and wind-induced upwelling - many processes are unique

to these northern environments [Chen, etal, 2003]. There is a strong seasonality in

northern marine environments due to ice cover, snow deposition, and riverine

input, which causes water column stratification in the summer from freshwater

input via snow- and ice-melt and freshet, and stratification breakdown in the winter

from salt rejection during ice formation [Chen, et al, 2003]. The difference in

temperature and salinity between the converging fresh river [source of heat) and

saline marine waters also contribute to seasonal stratification in many high latitude

estuaries [Macdonald and Yu, 2006].

In the Arctic, marine shelf-estuarine systems receive large amounts of

terrestrial freshwater input - the Arctic Ocean receives approximately 10% of the

global river discharge - that carry in high volumes of dissolved and particulate

matter and nutrients from predominantly permafrost regions rich in organic carbon

[Dittmar and Kattner, 2003]. As a result, Arctic and sub-arctic estuarine-shelf

systems are highly dynamic with increased production, respiration/degradation,

resuspension, and sedimentation [Borges, 2005]. These processes all affect the

transport and sequestration of carbon and the air-water flux of CO2 in estuarine-

shelf systems.

Hudson Bay receives a freshwater load of 713.6 km3 year1, [Dery and Wood,

2005] about 18% of the total discharge to the Arctic Ocean. The majority of this

freshwater input is controlled as a result of hydroelectric development. The large

amount of riverine water delivered results in freshening of Hudson Bay coastal

Page 24: Air-water carbon dioxide exchange in relation to chemical

8

surface waters (salinities <28 PSU relative to ~34 for peripheral seas and bottom

waters) [Prinsenberg, 1986). A dynamic boundary exists where the inflowing rivers

meet Hudson Bay waters and mix under the influence of wind and tides. Here,

physical, chemical, and biological responses to the mixing waters influence gas

exchange. Little work has been done that explains how estuarine systems affect air-

water CO2 exchange relative to the respective river and marine environments.

Though the Churchill River comprises only a small proportion (2.8%) of the

freshwater delivered to Hudson Bay [Dery and Wood, 2005], its location (adjacent to

a town, with access to air, rail, and marine transportation), past hydroelectric

development (background data from impact assessments and current hydrological

monitoring), and future significance as a high latitude marine port make the

Churchill River and estuary system an ideal location to conduct a process-based

river-estuarine study, and a site for monitoring of climate induced changes to the

carbon cycle. Understanding the physiochemical transformations of the subarctic

Churchill River and estuary system will provide some insight into local carbon

budgets and may lead to a better understanding of how high latitude estuaries cycle

carbon.

2.2 - Role of the surface waters in determining CO2 flux

The exchange of CO2 between surface waters and the atmosphere governs

whether a water body will be either a source or sink for atmospheric CO2. This

exchange, or flux of CO2 (fco2, equation 2.2), is equal to the concentration gradient of

Page 25: Air-water carbon dioxide exchange in relation to chemical

9

CO2 between the surface water and the atmosphere (A[C02], equation 2.1)

multiplied by the gas transfer velocity, k [Liss, 1983; Liss and Merlivat, 1986; Millero,

2000].

A[C02] = [ C 0 2 ] w - [ C 0 2 ] A [ 2 < 1 ]

^ * t f«> 2 ] w " ^ A ) [ 2 2 ]

The resulting flux can either be positive, indicating a source of CO2 (flux from

a water body to atmosphere) or can be negative, indicating a sink for CO2 (flux into a

water body from atmosphere). A concentration of [CC>2]w greater than [C02]A

indicates a supersaturation of water with respect to air, whereas if [C02]w is smaller

than [C02]A, water would be undersaturated.

2.3 - Processes that affect [C02]w

2.3.1 - Chemical Processes

The carbonate system in oceans, which contains the majority of oceanic

carbon, is extremely important because it controls water pH and the exchange of

CO2 between the atmosphere, lithosphere, and biosphere [Millero, 2000].

Dissolved carbonate equilibria can be described with a set of equations, as

follows:

Page 26: Air-water carbon dioxide exchange in relation to chemical

10

C°2(g)

PC02(w)

* H2C03(aq)

HC03(aq)

< •

K0

K1

K2 < > •

PC02(w)

* H2C03(aq)

H+ + HC03(aq)

H + + co3;aq)

[2.3]

[2.4]

[2.5]

[2.6]

[H2C03] [2.7]

pco2

_ [H+][HCO;] [ 2 .8 ] 1 [H2C03]

K = [^HCof] [2.9] 2 [Hccg

H2CO3* represents the analytical sum of CC^wand H2C03faqj [DOE, 1994]

[Sarmiento and Gruber, 2006; Stumm and Morgan, 1996a]. From these equations, we

can see that CCbwcan be increased by 1) air/water exchange (equation 2.3) 2)

shifting the reactions in equations 2.4, 2.5, or 2.6 to the left or 3) changing the

equilibrium constants in equations 2.7, 2.8, or 2.9.

The carbonate system can be described for aquatic systems through

measurement of two or more of the following variables: total [C02]w [or dissolved

inorganic carbon, DIC) (equation 2.10), partial pressure of CO2 (pCChw) (as seen in

equation 2.4), alkalinity (Alk, equation 2.11), and pH (equation 2.12) [Millero, 2000;

Stumm and Morgan, 1996b]. Alk represents the "acid-neutralizing capacity of an

Page 27: Air-water carbon dioxide exchange in relation to chemical

11

aqueous system" [Stumm and Morgan, 1996a] or "the concentration of all the bases

that accept H+ when a titration is made with H+ to the carbonic acid endpoint",

where the principal bases include HCO3-, CO32-, and B(0H)4- [Millero, 2000].

DIC = [H2C03*] + [HCO3- ] + [C032 • ]

[2.10]

Alk = [HC03"] + 2[C032"] + [B(OH)4"] + [OH"]-[H+]

+ [SiO(OH)3- ] + [MgOH+ ] + [HP042' ] + [P04

3" ] 1

pH = - l o g { H + } [2.12]

pH is the activity of H+ ions in an aqueous solution, and therefore is affected

by carbonate equilibria that can either produce or consume H+ with an

increase/decrease in [CCbJw [Stumm and Morgan, 1996a]. For example, increases in

[C02]w by way of equations 2.3 and 2.4 will increase both the acidity (lower pH) and

the total concentration of dissolved carbon species in water. Millero [2000] suggests

that DIC, pC02w, and pH are the three best variables to measure (in that order)

because they undergo the largest amount of change over time (diurnally and

seasonally).

Alk is unaffected by changes in [C02]w because carbonic acid is the reference

proton level used to determine proton deficiency [Stumm and Morgan, 1996a]. Alk is

however, affected by reactions consuming or producing H+ and OH", such as calcium

carbonate (CaCOs) precipitation and dissolution, and photosynthesis and

Page 28: Air-water carbon dioxide exchange in relation to chemical

12

respiration/decomposition (e.g. NO3" assimilation from photosynthesis increases Alk

while decomposition decreases it) [Stumm and Morgan, 1996a].

H2CO3* only contributes slightly to the DIC pool, and therefore carbonate and

bicarbonate ions are ultimately controlling pCC^w [Sarmiento and Gruber, 2006]. By

rearranging and combining equations 2.5 to 2.9 and replacing carbonate and

bicarbonate with DIC and Alk, we arrive at the following equation expressing pC02w

[Sarmiento and Gruber, 2006]:

pC02

K, (2 • DIC - Alk) '2

*0*1 Mk-D,C [2.13]

From equation 2.13 we can see that there are three main parts to what

controls surface water pCCh: 1) the ratio of equilibrium constants (K2/K0K1) 2) DIC,

and 3) Alk [Sarmiento and Gruber, 2006; Stumm and Morgan, 1996a]. It is easiest to

separate these pCCh controls into two sections: 1) those that are affected by physical

processes and 2) those that are affected by biological processes.

2.3.2 - Physical processes

Gas exchange at the air-water interface, controlled by k and A[C02], will

result in DIC concentration changes (Figure 2.1). Looking at equation 2.3, we can see

that [CO2V will increase or decrease in response to CO2A in order to maintain

equilibrium. The transformation of CO2W to H2CO3* translates into only a small

Page 29: Air-water carbon dioxide exchange in relation to chemical

13

increase in [C02]w [Sarmiento and Gruber, 2006], and is quite slow and less

important compared other physical and biological processes influencing [C02]w.

2.3.3 - Biological processes

Photosynthetic Production and Respiration

Plankton growth and death in the euphotic zone has a strong influence on

[CO2V and carbonate equilibria, as well as carbon transport to the sediments

[Figure 2.1) [Portielje and Lijklema, 1995; Wetzel, 2001]. Biological production and

respiration affects [CChJwin aquatic systems through the following reaction

[Redfield, etal, 1963]:

106CCL + 16NO / + HPO.2" + 78hLO + 18H+ <—• C.^H.^O.-N^P + 150CL 2 3 4 2 106 175 42 16 2 >

[2.14]

where the forward reaction is photosynthesis and the reverse reaction is

respiration. Surface waters become undersaturated in CO2 when photosynthesis is

high; CO2 uptake rate is higher than transfer from the atmosphere across the air-

water interface or that released through respiration; while waters may become

saturated in CO2 at night when photosynthetic metabolism stops, and where

bacterial respiration is large (e.g. DOC rich fresh waters) [Millero, 2000; Stumm and

Morgan, 1996a; Wetzel, 2001]. Should this occur, the DIC pool in surface waters will

try to return to equilibrium conditions by CO2 diffusion at the air-water interface.

Page 30: Air-water carbon dioxide exchange in relation to chemical

14

Net primary production is the difference between photosynthetic production

(daytime] and respiration (both day and night), and is determined by nutrient

concentrations (primarily N and P, though Si may be limiting to diatoms) and light

conditions (and to a degree, water temperature which controls respiration rate)

[Sakshaug, 2004]. Schindler has shown in several whole-lake experiments that

atmospheric carbon is generally sufficient to fulfill phytoplankton carbon needs for

growth when nitrogen and phosphorus concentrations are ample [Schindler, 1974;

Schindler, 1977], though DIC may become rate limiting for portions of the day when

primary production is very high in surface waters [Portielje and Lijklema, 1995].

Increasing concentrations of the limiting nutrient will shift the species composition

and thus alter the uptake of other compounds, including dissolved CO2 [Schindler,

1974; Schindler, 1977]. Redfield et al. [1963] described aquatic systems as N limiting

if the N:P ratio is < 16, while P is limiting if N:P >16. Sakshaug [2004] reports an N:P

ratio ranging from 11-16 (mol mol-1) for Arctic marine waters, making them

predominantly N limited.

Another indicator of algal production is chlorophyll a, a pigment in all algal

cells, though its relationship with carbon fixation depends on acclimation and thus

may not be a useful indicator of carbon biomass. The chla:C ratio in arctic waters

has been reported by Sakshaug [2004] to range 0.003 to 0.8 (weight:weight), and is

highest for high nutrient/low light conditions (e.g. under ice) and lowest for

nutrient poor/high light conditions.

Page 31: Air-water carbon dioxide exchange in relation to chemical

Figure 2.1: Carbon cycling and processes in aquatic systems [Ducklow and McCallister, 2004). pCC»2 in surface waters can increase/decrease through gas exchange, photosynthesis/respiration from aquatic organisms, precipitation/dissolution reactions with CaC03, and decomposition of organic material by heterotrophs. Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA - IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.287, Cambridge, Mass.: Harvard University Press, Copyright (c) 2006 by the President and Fellows of Harvard University.

Atmosphere pC02 - 370 fiatm

£s?

I f £ F-kVApCO,

1* Ocean pCO, 100-1000 patm

Ca , t + 2HCOi<=>CaCOJi+H,0+C02 <=> H.CO, <=> H*+HCO; ** 2H'+COj-

\t7 Zooplankton « Phytoptankton

DOM i

Net downward transport of particles and dissolved organic matter

(Biological pump) Y

u

I i

Bacterial respiration CO,

Production can be sustained by nutrients either released by organisms in the

photic zone [regenerative production) or supplied externally to the photic zone via

coastal upwelling of nutrient-rich deep water, riverine input, etc. (new production)

[Sakshaug, 2004]. The incidence of new production is highest in areas with almost

continual upwelling, such as along shelf breaks and where wind-driven vertical

Page 32: Air-water carbon dioxide exchange in relation to chemical

16

mixing is sufficiently deep, and smallest where vertical mixing does not reach the

nutrient-rich deep layer even in winter, such as in the deep Arctic Ocean [Sakshaug,

2004]. The combination of seasonal ice cover and light intensity in arctic marine

systems influence the algal community structure and thus primary production.

Phytoplankton growth in arctic and sub-arctic marine systems generally begins

under ice as light intensity ad penetration through the ice increases, and continues

as surface blooms as ice retracts from the nutrient rich water. In stratified open

arctic waters, production will decrease throughout the summer with new

production occurring just above the pycnocline (30-35m), where nutrients from

deeper waters can be introduced [Sakshaug, 2004]. In estuarine-shelf environments,

however, primary production remains high throughout the open water season due

to nutrient input from rivers, tidal current mixing, and coastal upwelling, with

growth generally ending in the fall with the formation of sea ice [Sakshaug, 2004].

Borges [2005] proposes that primary production may increase [for some

species) in marine coastal systems as a result of CO2 accumulation in surface waters.

Primary production may also be enhanced by an extended growth season due to

shorter annual ice cover and earlier spring melt, which is predicted by most climate

warming models to occur throughout the next 100 years [IPCC, 2007]. Other factors

such as riverine freshwater supply, coastal upwelling, and wind- and tidal-induced

mixing also influence coastal primary production and will all likely contribute to

variability in primary production as they are individually affected by climatic

change.

Page 33: Air-water carbon dioxide exchange in relation to chemical

Remineralization is the process by which organic matter, particulate or

dissolved, is returned to an inorganic state through decomposition and respiration.

These reactions cause an increase in [CO2V through the breakdown of particulate

organic matter (POM) and dissolved organic matter (DOM), which can either be

recycled by autotrophs in the photic zone, or accumulate in deep waters (Figure

2.1). Since many heterotrophic organisms can function in both anaerobic conditions

and below the photic zone, accumulations of POM and DOM from sedimentation and

sinking cells/organisms can result in high concentrations oiDIC in deep waters

[Wetzel, 2001]. This accumulation can contribute significantly to the surface water

DIC concentrations in upwelling regions, as we will see in subsequent sections.

Precipitation and Dissolution

Interaction with solid CaC03 also influences [C02]w and Alk:

Ca2+ + 2HC0 3 - ^ -> CaC03 + H20 + C02 [ 2 l 5 ]

The forward reaction (precipitation) produces CO2, while the reverse

reaction (dissolution) consumes gaseous CO2 (2C02W can be produced from 2HCO3"

for ICO2A consumed) [Stumm and Morgan, 1996a]. Carbonate precipitation is an

important process for transporting CO2 to deep oceanic waters (Figure 2.1, 2.2); a

vast amount of oceanic carbon is stored as carbonate in rocks and sediment, much

larger than the amount of cycled CO2 [Millero, 2000]. Some biota (e.g.

coccolithophores-algae, and foraminifera-amoeboid protists) produce CaC03 scales

Page 34: Air-water carbon dioxide exchange in relation to chemical

18

and shells, which upon death, fall to the sediments carrying the carbon stored as

CaC03 with them.

2.3.4 - Summary

The main processes that modify pCCb in water can be summarized in Figure

2.2. Photosynthesis, CO2 efflux, and to some extent CaC03 formation all reduce DIC,

while respiration, CO2 influx, and CaC03 dissolution increase DIC. Only processes

that consume or release H+ ions lead to changes in Alk, with CaC03 dissolution

increasing Alk while CaCCh formation decreases it [Stumm and Morgan, 1996a;

Wetzel, 2001; Zeebe and Wolf-Gladrow, 2001]. Alk is also affected indirectly by

photosynthesis when nutrients (NO3", PO43) are consumed (increased Alk] or

released when cells die (decreased Alk). Precipitation, dissolution, photosynthesis,

and respiration reactions, as well as the hydration or dehydration of CO2 are all

relatively fast reactions producing/consuming CO2W when compared to gas transfer

at the air-water interface [Stumm and Morgan, 1996a].

Page 35: Air-water carbon dioxide exchange in relation to chemical

19

Figure 2.2: Biogeochemical processes modifying Alk and DIC in aquatic systems [Zeebe and Wolf-Gladrow, 2001). This article was published in "CO2 in seawater, equilibrium, kinetics, isotopes", Zeebe and Wolf-Gladrow, Equilibrium, p. 7, Copyright Elsevier, 2001.

DIC (mmol kg 1)

2.4 - Gas Exchange at Water Surfaces

The flux of CO2 to or from surface waters is balanced by biogeochemical

processes [photosynthesis, respiration, variation in the carbonate system) and

varies diurnally, seasonally, and globally with latitude [Borges, 2005; Ducklow and

McCallister, 2004; Millero, 2000; Sarmiento and Gruber, 2006]. The gas transfer

velocity, k, strongly influences the flux of CO2 across the air-water interface, and this

transport is driven primarily by the concentration gradient of CO2 in the boundary

Page 36: Air-water carbon dioxide exchange in relation to chemical

20

layer [Borges, et al, 2004; Portielje and Lijklema, 1995]. The gas transfer reaction at

the air-water interface is relatively slow and as a result, CO2 in many surface waters

is not at equilibrium with atmospheric concentrations [Stumm and Morgan, 1996a].

2.4.1 - Solubility

Air-water CO2 transfer is driven by the concentration gradient of CO2 at the

air-water interface:

ApC02 =PC02W - pC02A [216]

A[C02] = [ C 0 2 ] W - [ C 0 2 ] A [ 2 1 7 ]

where w and A refer to water and air. For a system at equilibrium, A[C02] and ApC02

will equal zero, while if ApC02 is positive, the flux will be from the water to the

atmosphere. The reverse is true for a negative ApCCh.

The concentration of CChw is determined by the solubility and partial

pressure of CO2 in water, following Henry's law [Stumm and Morgan, 1996a]:

[ C 0^ • ° • P C C \ [2.18]

where

a = EXP ( /A1 +A2 (100/ 7) + A3 LN (77 100) +A^ (7/ 100)2 + S• [8,+ S2 (T 1100) + S3 (7/ 100)2] ) t

[2.19]

a is the solubility coefficient (mol I/1 atnr1), Ai, A2, A3, A4, Bi, B2 and B3 are the

empirically-derived coefficients for molar solubility [Weiss, 197'A; Weiss and Price,

Page 37: Air-water carbon dioxide exchange in relation to chemical

21

1980], and 7 is the water/air temperature (K). The solubility of C02, and thus [C02],

are also affected by both salinity and temperature (Figure 2.3).

Figure 2.3: Relationship between solubility of CO2 and water temperature and salinity. Lines S=0, S=10, and S = 25 represent the relationship between water temperature (lefty-axis) and solubility at salinities of 0,10, and 25 PSU. Lines T=273, T=281, and T=296 represent the relationship between salinity (right y-axis) and solubility at water temperatures of 273, 281, and 296 °K.

305

°_ 300

g 295 -•

t 290 +

£ 285 + £ 280 +

5 275 -•

270 -! 20

•S = 0

30 40 50 60 70

Solubility (mol m3 atm1)

T 30

•- 25

• • 2 0 w CL.

• - 1 5 ^

"S

• • 5

4- 0

•S = 10 •S = 25 •T = 273 •T = 281

80

•T = 296

2.4.2 - Stagnant Film Model

From the previous section, we established that the concentration gradient of

CO2 at the boundary layer occurs as a result of ApCCh from the air-water interface

equilibrium CO2 concentration difference. This gradient is steepest just above or

below the air-water interface (see Figure 2.4) [Portielje and Lijklema, 1995].

Page 38: Air-water carbon dioxide exchange in relation to chemical

22

The stagnant film model assumes that two "stable stagnant films of finite

thickness", one in water and one in the atmosphere, are formed in the absence of

turbulence, and that gas diffusion through these films occurs by molecular

diffusivity, e [Whitman, 1923]. The flux of CO2 through a stagnant film can be

calculated from Fick's first law:

4> = -£ A z [2.20]

where § is flux (mmol m2 s-1), E is molecular diffusivity (air or water), [A] is the

concentration of gas A (air or water), and z is the thickness of the stagnant film layer

[Sarmiento and Gruber, 2006].

Using figure 2.4, we can rewrite equation 2.20 for both the atmosphere and

water stagnant film layers:

4>a = - "7T * (KOJa ~ [COJ?) a [2.21]

and

d>w = -V([co 2 ] w - [co 2 U 0

- rrn.i ^ [2.22]

where a is solubility (equation 2.19), ka = £a/Aza, kw = EW/AZW, and ka and /fw are

referred to as the gas transfer or piston velocity for air and water [Sarmiento and

Gruber, 2006]. Since actual measurements of CO2 concentration or partial pressures

are always made below or above the stagnant film layers, we can remove the terms

[C02]A° and [CO2V0 from equations 2.21 and 2.22 and calculate flux from:

Page 39: Air-water carbon dioxide exchange in relation to chemical

23

* = -*-([co2]a-[co2]j [223]

where

J_ . k

= J- + i ^w ^a [2.24]

Liss and Slater [1974] have compared the gas transfer velocities of the air

and water film layers, and discovered that the water film layer, with a transfer 100

times slower than the air film layer, controls CO2 at the air-water interface. We can

therefore ignore the a/ka term, so k = kw [Sarmiento and Gruber, 2006].

Figure 2.4: Stagnant film model [found as Figure 3.3.1 in "Air-Sea Interface", Sarmiento and Gruber, 2006, p. 82]. Gas transfer at the air-water interface is a relatively slow process, which results in a concentration gradient. CO2 is well mixed in the turbulent layers, uniform concentrations of [CChJw and [C02]atm occur in the air. [CChJw and [C02]atm in the stagnant layers changes linearly to saturation levels [C02]w° and [CCbJatm0- Gas transfer across the air-sea interface is governed by the rate of molecular diffusion of gas CO2 through the two stagnant films.

Page 40: Air-water carbon dioxide exchange in relation to chemical

2.4.3 - Determination of Gas Transfer Velocity, k

The gas transfer velocity is a function of turbulence in the mixed water layer

(due to primarily wind stress], and by the thickness of the molecular diffusive layer,

determined by the Schmidt number:

c £ [2.25]

where v is the kinematic viscosity of water.

The parameters to calculate the Schmidt number of CO2 between 0-30 °C have been

determined by Jahne et al. [1987b]:

S C =A-BT + CT2-DT3 ^

where A = 2073.1, B = 125.62, C = 3.6276, D = 0.043219, and T is water temperature

in °C \Jahne, etal, 1987b].

Measurements of k have been made by several techniques: isotope studies,

synthetic dye evasion methods, wind tunnel experiments, and through direct

observation using eddy-correlation (Table 1) [Liss and Merlivat, 1986; Wanninkhof,

1992; Wanninkhof and McGillis, 1999, Nightingale etal, 2000].

Since the gas transfer velocity is the primary source of error in calculating

CO2 flux, it is recommended that k be determined for specific locations until

empirical models are developed that can be applied to a range of aquatic systems.

Experimental estimation of k, however, requires extensive monitoring, and

therefore it is much more common for studies to use k parameterizations based on

wind speed. Some of the most common k parameterizations in use today are shown

Page 41: Air-water carbon dioxide exchange in relation to chemical

in Table 1 [Liss and Merlivat, 1986; Wanninkhof, 1992; Wanninkhofand McGillis,

1999, Nightingale etal, 2000].

Table 1: Commonly used k parameterizations estimated for some freshwater and marine environments under different wind conditions.

Source Basis Gas transfer velocity parameterization (cm hour-1) Nightingale et al. [2000] dual deliberate tracer k = (0.333-U + 0.222'U2)'(Sc/600)"0-5

k = 0.17«O(Sc/600)"2/3, U < 3.6 ms'1

Liss & Merlivat [1986] 'S?^*^ * = ( U ' 3-4>-2.8.(Stf600)*s, 3-6 ms"1 < U * 13 ms1

k = (U - 8.4)'5.9-(Sc/600)"0-5, U > 13 ms"1

Wanninkhof & McGillis [1999] d i r e c t o b s e r v a t i o n s k = o.0283«U3-(Sc/600)-°'5

using eddy covanance v ;

Wanninkhof [1992] isotope studies k = 0,31-U2'(Sc/600) " 5

However, relationships using only wind speed to predict gas transfer

velocities can be flawed since some of the factors regulating gas transfer are not

related to wind speed [Wanninkhof, 1992]. Based on this premise, Raymond and

Cole [2001] have found that predetermined wind speed-parameterized /rvalues

may be significantly different in estuarine environments than for adjacent marine

waters under the same wind conditions. Borges et al. [2004] suggest that this is due

to the highly turbulent nature of estuarine environments compared to the open

ocean and that other factors such as tidal currents and fetch limitation may increase

k. Since depth, tidal velocity, and wind strength and direction is variable between

estuaries, Borges et al. [2004] suggest that k parameterized as a function of wind

speed is site specific for estuarine environments. This implies that flux calculations

using pre-determined equations may be subject to significant error.

Page 42: Air-water carbon dioxide exchange in relation to chemical

26

Through various studies it has been shown that k is also dependent on other

variables: waves, bubble formation, surface films and surfactants, air-water

interface turbulence, and tidal currents, all of which can be highly site specific

[Borges, etal, 2004; Portielje and Lijklema, 1995; Wanninkhof, 1992].

2.4.4 - Other considerations for gas transfer

Air-water interface CO2 fluxes are best and most robustly measured by

collecting in situ pCCb data, by equilibrating a water sample with air or nitrogen and

later measuring the equilibrated gas with an infrared gas analyzer, or by making

continuous measurements of surface water by passing water through an

equilibrator and infrared gas analyzer [Borges, 2005; Millero, 2000]. Though pQOi

can be measured accurately, the calculation of fluxes are influenced by the gas

transfer velocity parameterizations which are site specific and may not be

accurately scaled to certain environments [Borges, 2005].

Bubble enhancement of gas transfer

Few studies have attempted to quantify the effect that bubbles have on air-

water gas transfer [Atkinson, 1973; Thorpe, 1982]. It seems near impossible to

estimate the number and size of bubbles as they are a function of water body

conditions and depth from the air-water interface [Liss, 1983]. Other factors making

the quantitative assessment of bubbles challenging include: differential gas

absorption/desorption, pressure changes (as a function of depth], and films

Page 43: Air-water carbon dioxide exchange in relation to chemical

27

surrounding bubbles [Liss, 1983]. Atkinson [1973] and Thorpe [1982] have

estimated that bubble contribution to gas flux resulting form breaking waves is

likely only significant at wind speeds over 10 (12) m s1.

Chemical enhancement of gas transfer

The chemical enhancement of CO2 exchange will increase CO2 flux at the air-

water interface at low wind speeds, though this factor has been ignored in most

wind speed-parameterized gas transfer equations [Wanninkhof, 1992]. CO2

enhancement occurs at the air-water interface by the reaction between CO2A and

H2O or OH- ions, which increases the concentration difference of CO2 between the

air and water turbulent layers, and thus gas transfer [Wanninkhof, 1992]. An

enhancement factor E(t, pH, k) (enhancement as a function of temperature, pH, and

gas transfer velocity) can be used to account for the chemical enhancement of CO2

[Wanninkhof, 1992], which may yield more accurate flux calculations at wind speeds

below 5 ms1 [Wanninkhof, 1992].

Fugacity of CO? It is a common oversight in aquatic chemistry use the terms pC02 and/C02

interchangeably, though in fact they describe relationship between CO2 in two

different hydration states:

[C02] / c ° 2 = "F"

0 [2.27]

Page 44: Air-water carbon dioxide exchange in relation to chemical

28

where Ko is the solubility coefficient of CO2 in water from the reaction CO2A = CO2W.

This is different from pC02, the partial pressure of CO2, which is calculated

using Henry's Law constant, Kh, in pCCh = [CO2] / Kh, where Kh is a function of

temperature, the gas in question (CO2), and the solvent (water, CO2W + H2OQ) =

H2CO3).

Fugacity is the tendency for a substance to "prefer" one state, the most

thermodynamically stable (i.e. lowest Gibbs free energy of solid, liquid, and gas

phases), to another. Fugacity may also be defined as the pressure change necessary

for a real gas to behave as an ideal gas, which for low pressure gases in accordance

with the ideal gas law, fugacity * pressure. The fugacity coefficient, §, represents the

ratio of fugacity to pressure, which for an ideal gas equals to 1. For CO2, a non-ideal

gas, fugacity is not equal to partial pressure but values are very close in magnitude

[McGillisand Wanninkhof, 2006].

2.5 - Carbon exchange at the land-ocean interface

Arctic and subarctic environments, including terrestrial, freshwater, and

marine systems, demonstrate seasonality due to a wide variation in solar radiation,

temperature, and hydrological influences between seasons. Coastal oceans and

estuaries are highly dynamic systems that undergo marked seasonal changes, and

are affected by the seasonal evolution of both terrestrial and marine environments.

Page 45: Air-water carbon dioxide exchange in relation to chemical

29

2.5.1 - Rivers

The physical, chemical, and biological properties of estuarine-shelf

environments are governed by both riverine and oceanic processes; watershed size,

flow volume, and the physiochemical composition of the inflowing freshwater from

rivers into the ocean significantly influences the estuarine-shelf environment.

Many arctic and sub-arctic rivers have expansive watersheds covering rich

permafrost soils, that currently store over half the global organic carbon [Dittmar

and Kattner, 2003], and this soil composition contributes directly to the amount and

types of carbon collected in rivers. As water flows over the land towards streams

and rivers during the spring thaw and ice-free months, it collects particles rich in

organic C. Due to the seasonality of river water discharge in high latitude

environments, riverine organic carbon concentrations are highest in summer and

lowest in winter (highest when discharge is highest) while nutrient concentrations

are generally highest in the winter and lowest in the summer (highest at minimum

discharge) [Dittmar and Kattner, 2003].

Approximately 80% of the total organic carbon total organic carbon (TOC)

input to the Arctic Ocean via rivers is in the form of dissolved organic carbon (DOC)

(discharge of 18-26 x 1012 g DOC/year), originating primarily from terrestrial

sources (60-70% humic compounds) [Dittmar and Kattner, 2003]. This abundance

of labile carbon is available to decomposers making most large rivers net

heterotrophic systems [Cole and Caraco, 2001]. Initially high DIC and [C02]w

production from heterotrophy make most rivers net sources of CO2 [Raymond, et ah,

Page 46: Air-water carbon dioxide exchange in relation to chemical

30

2000]. Several studies have also found high latitude lakes to be net sources of CO2

[Cole, etal, 1994].

High latitude rivers supply the Arctic Ocean with large quantities of

freshwater and nutrients. These vast systems have been used more extensively for

their power generating capabilities over the past few decades, but not without

significant impact to both upstream and downstream ecosystems. Hydroelectric

development relies on a relatively constant annual water flow in order to be

profitable. Since the hydrograph of high latitude rivers vary considerably with

season, a significant amount of water must be impounded and reserved for periods

of natural low flow (winter months). This often results in major flooding and

particle accumulation upstream, altered flow and changes to chemical, physical, and

biological water characteristics downstream [Rosenberg, etal, 1997], and can lead

to an altered water budget for the affected catchment area due to enhanced

evaporation [Vorosmarty and Sahagian, 2000]. Hydroelectric development also has

potential to alter ocean currents (through changes in freshwater load, and thus

stratification stability) although transformations of this intensity have not yet been

documented [Rosenberg, etal, 1997].

The work undertaken here relates to the Churchill River and estuary. The

Churchill River passes through the interior Great Plains, the Canadian Boreal Shield,

and the Subarctic Taiga before reaching Hudson Bay. In 1976, Manitoba Hydro

undertook the Churchill River Diversion project that diverted approximately two

thirds of the Churchill River's flow from Hudson Bay to the Nelson River through

Page 47: Air-water carbon dioxide exchange in relation to chemical

31

Southern Indian Lake (Figure 2.5) [Baker et a/., 1994]. Since this diversion, the

majority of flow in the lower Churchill River is from inflowing rivers and streams

downstream of Southern Indian Lake, and this is reflected in the river's water

chemistry [Bezte, 2006].

Figure 2.5: The lower Churchill River from Southern Indian Lake to Hudson Bay (Landsat 7 satellite image, Natural Resources Canada). Discharge for the Churchill River is gauged at Red Head Rapids.

The flow of the Churchill River (gauged at Red Head Rapids, Figure 2.5) is

lowest in April at 105 m3s_1 and peaks in late May/early June to 589 m3s-1 (mean

Page 48: Air-water carbon dioxide exchange in relation to chemical

32

monthly discharges, 1981-1993) [Bezte, 2006]. Water chemistry of the Churchill

River was determined from samples collected between 1999-2005 (just upstream of

the weir) and have the following ranges: pH - 7.8 to 8.6, TSS - <2 to 21 mgL-1,

nitrate/nitrite N - <0.005 to 0.01 mgL1, total phosphorus - 0.007 to 0.056 mgL-1,

DOC - 8 to 14 mgL1, chlorophyll a - 0.9 to 7.3 ugL1 [Bezte, 2006].

2.5.2 - Estuaries

Estuarine systems have been studied for decades although due to their

dynamic nature, only a small proportion of studies have examined gas transfer, and

an even smaller proportion have taken place at high latitudes. Estuarine

environments vary in carbon cycling and CO2 exchange depending on the size of

inflowing river and the width of the continental shelf. Processes in inner and outer

estuaries are quite different and terrestrial matter carried through estuaries

undergo significant changes before being transported to the coastal shelf and

adjacent sea [Borges, 2005].

Inner estuaries are diverse and dynamic systems described by Borges

[2005] as "characterized by strong gradients of biogeochemical compounds,

enhanced organic matter production and degradation processes, and intense

sedimentation and resuspension". Physical characteristics of inner estuaries

(watershed area, freshwater discharge, geomorphology, water chemistry, and tides)

influence both nutrient and carbon exchanges and biogeochemical cycling as well as

water column stratification in the estuary, freshwater residence time, and spatial

Page 49: Air-water carbon dioxide exchange in relation to chemical

33

extent of the plume [Borges, 2005]. The geographic boundaries of the inner estuary

are the river mouth [lower limit) and the upstream limit of tidal influence (upper

limit). Inner estuaries are identified by Borges [2005] as heterotrophic systems

(respiration, R > gross primary production, GPP), though this may shift for some

estuaries in the future due to increased nutrient loading. CO2 exchange from inner

estuaries depends on DIC inputs, ecosystem metabolism and the physical

characteristics mentioned above, though inner estuaries are generally a source of

nutrients and CO2 to oceanic waters, a possible source of CO2 to the atmosphere, and

a sink for organic matter [Borges, 2005].

Outer estuaries are defined by fresh water plumes floating upon dense sea

water and extending beyond the mouth of the river. Water temperature, depth, and

tidal currents govern the presence and degree of stratification (thermal or haline) in

outer estuaries, and the extent of plumes can be determined by both temperature

and salinity gradients, as well as gradients in turbidity, nutrients, total alkalinity,

and chlorophyll a [Borges and Frankignoulle, 1999]. Unlike inner estuaries, outer

estuaries have frequently been reported to be undersaturated in CO2 [Borges, 2005]

and have smaller air-water CO2 fluxes and reduced carbon and nutrient exchange

with the ocean [Borges and Frankignoulle, 1999]. Though outer estuary air-water

CO2 fluxes are often much lower than for inner estuaries (one order of magnitude

lower) and can vary both spatially and temporally, the surface area of the outer

estuary is quite large comparatively and thus contributes significantly to the CO2

budget of estuaries [Borges, 2005].

Page 50: Air-water carbon dioxide exchange in relation to chemical

34

The inner part of the Churchill River estuary is enclosed within a basin that

extends approximately 13.2 km in length (latitude 58.793°, longitude -94.2048°), at

low tide is up to 1.7 km wide, and has an average depth of 3m at low tide (except for

a dredged, narrow shipping channel of 21m) (Baker et al. 1994). The outer estuary

of the Churchill River consists of a brackish plume extending into the Hudson Bay at

varying distances depending on tide (plume greatest at low tide) and river

discharge (plume extends furthest in the spring at freshet).

A recent study by Kuzyk et al. (2008) examined biophysical processes within

the Churchill River estuary during the spring melt period. Their findings explain the

relationship between river load and sea ice within the Churchill River estuary, and

demonstrate the importance of river input to the local marine system through ice

melt and the supply of nutrients. Aside from a few technical reports prepared for

Manitoba Hydro (impact assessments following the Churchill River diversion, and

prior to and following construction of a rock weir) little other physiochemical and

biological data exist for the Churchill River and estuary. (Data presented in these

reports were from samples collected upstream of the weir, and were outlined above

in the "River" section).

Significant physical, chemical, and biological transformations occur within

estuaries throughout the year, and it is crucial that we understand how these

processes currently vary in estuaries with respect to season in order to forecast how

climate warming may affect these systems in the future. What these transformations

Page 51: Air-water carbon dioxide exchange in relation to chemical

35

represent to the global carbon cycle is largely unknown, and therefore further work

is necessary in order to understand these processes.

2.5.3 - Continental Shelves

Shelf environments can behave very differently from the open ocean with

respect to carbon cycling and CO2 exchange depending on shelf width, depth and

shelf-edge depth, river input, water retention time on the shelf, and heat balance

[Chen, et ah, 2003]. An extensive literature review on air-sea CO2 exchange by

Borges [2005] found that both high and temperate latitude continental shelves tend

to be sinks for atmospheric CO2, while subtropical and tropical shelves tend to be

sources [Borges, etal, 2005]. A similar review of coastal margins arrived at

comparable results [Cai, etal, 2006].

2.6 - CO2 transport from coastal systems

There are three main methods by which the oceans can take up atmospheric

CO2: the solubility pump, the biological pump, and the coastal shelf pump [Millero,

2000]. The solubility pump is driven by the difference in CO2 concentration between

the atmosphere and surface water, where higher concentrations in the surface

water cause a flux of CO2 from the ocean and a higher concentration in the air will

cause a flux of CO2 to the ocean surface water [Millero, 2000; Ducklow and

McCallister, 2004]. The biological pump is driven by primary production in the

surface mixed layer of oceanic waters. During the open water months,

Page 52: Air-water carbon dioxide exchange in relation to chemical

36

phytoplankton will grow and die or be consumed by zooplankton and other

secondary producers, all of which will eventually settle down to deep oceanic

waters taking the carbon (CO2) sequestered over their lifetime with them [Millero,

2000; Ducklow and McCallister, 2004]. The coastal shelf pump transports carbon and

nutrients from rivers and coastal oceans to the deep ocean through particle settling

[Bates, 2006; Tsunogai, etal, 1999]. In high latitudes, this process is based on

surface water cooling on coastal shelves that produces dense waters, which along

with primary production increase the influx of atmospheric CO2 [Tsunogai, etal,

1999]. This denser, C enriched water from the shelf is transported below the

pycnocline along the shelf slope by isopycnal mixing [Tsunogai, etal, 1999]. This

process, if confirmed for coastal shelf environments globally, may have a significant

effect on our understanding of oceanic uptake of atmospheric CO2 [Borges, 2005].

Figure 2.6 highlights the main carbon and nutrient transformations at the

land-ocean interface and the pathways by which carbon can be buried (solubility,

biological, and coastal pumps). Many reactions seen in Figure 2.6 were covered in

previous sections and will not be discussed in detail here.

Page 53: Air-water carbon dioxide exchange in relation to chemical

37

Figure 2.6: Carbon and nutrient cycling between the ocean, atmosphere and land [Ducklow and McCallister, 2004). All three carbon pumps - biological, solubility, and coastal shelf - are show. Arrows between organic carbon and DIC indicate biological production and respiration. Reprinted by permission of the publisher from THE GLOBAL COASTAL OCEAN: THE SEA - IDEAS AND OBSERVATIONS ON PROGRESS IN THE STUDY OF THE SEAS, VOLUME 13, edited by Kenneth H. Brink and Allan R. Robinson, pp.273, Cambridge, Mass.: Harvard University Press, Copyright (c) 2006 by the President and Fellows of Harvard University.

Atmosphere C0 2

2.7 - Climate implications

The estuarine-shelf environment in arctic and sub-arctic regions is

biogeochemically very active and is influenced by both freshwater and marine

processes. Estuaries integrate climate change impacts from large watershed areas

and may undergo significant changes due to the combined climate change effects on

both marine and terrestrial systems. It is important to understand carbon cycling

Page 54: Air-water carbon dioxide exchange in relation to chemical

38

and water/atmosphere interactions in northern coastal environments as they may

play a significant role in global CO2 flux estimates either as sources or sinks for CO2,

though their current significance in the global carbon cycle is unknown.

Increasing atmospheric CO2 concentrations and climate warming have the

potential to affect sea water temperatures, the extent and duration of sea ice cover,

photosynthetic production and respiration, and overall carbonate equilibria, which

will all influence the ability of high latitude coastal areas to sequester CO2. In arctic

and sub-arctic coastal systems, changes to ice conditions will likely have a profound

effect on the light regime and primary production, stratification stability, and

transport of particles (and some nutrients) to the open seas from the coasts and

estuaries. As sea ice forms, salt is rejected forming a saline, nutrient rich layer under

the ice. This process is reversed as ice melts forming a freshwater layer on top of

marine water and influences thermohaline circulation in high latitude seas [Chen, et

ah, 2003].

Shortening the duration of sea ice cover and extending the open water season

would enhance total primary production and lengthen the period of tidal and wind-

induced mixing (which could increase coastal erosion [Macdonald, etal, 1998] and

introduce a higher volume of sediment to the coastal oceans). Where sea ice does

not effectively transport nutrients and salt, sediments and organic carbon are

carried from coastal margins to the deep ocean via ice rafting [Eicken, et ah, 2000].

Sea ice changes may therefore affect the biogeochemistry of both the arctic/sub­

arctic coastal and open ocean environments [Chen, etal, 2003].

Page 55: Air-water carbon dioxide exchange in relation to chemical

39

The affect of climate change on land-based processes will also influence the

delivery of water [Deryand Wood, 2005; Rouse, etal, 1997], and thus nutrients and

sediment, to estuarine-shelf systems. Changes in the amount and timing of

precipitation will affect the rate at which rivers deliver nutrients and reduced

carbon to marine environments. Changes in the delivery of freshwater to marine

environments also has the potential to alter stratification: surface freshening from

increased precipitation could enhance ocean stratification, thus reducing the

downward transport of carbon via the continental shelf pump [Sarmiento, etal,

1998]. This, along with oceanic surface water warming and sea ice reduction, has

the potential to alter the function of the Arctic and sub-arctic coastal regions as a

source or sink for CO2.

It is unknown whether climate change will increase or decrease the extent to

which coastal environments can sequester carbon. This is especially of concern for

arctic and subarctic coastal shelves where little existing CO2 monitoring has been

done [Borges, 2005]. Currently coastal oceans are not included in global budgets for

CO2 and it is unknown what their inclusion will mean for future modeling/budgeting

and remediation strategies.

2.8 - Summary

Much is known about aquatic carbon chemistry, gas exchange at water surfaces,

and carbon exchange in rivers, lakes, and coastal oceans. However, little is known

about the relationships between gas exchange and biological, physical, and chemical

Page 56: Air-water carbon dioxide exchange in relation to chemical

40

processes within river-estuarine systems, and even less is known for arctic and

subarctic systems. High latitude estuaries are challenging environments to study as

a result of tides, strong currents, and ice floes during the spring and early summer.

The lack of temporal-spatial data for arctic and subarctic river-estuarine systems

makes this research both important and unique.

Page 57: Air-water carbon dioxide exchange in relation to chemical

References

ACIA (2005), Arctic Climate Impact Assessment, Cambridge University Press, Cambridge, U.K.

Atkinson, L. P. (1973), Effect of air bubble solution on air-sea gas exchange, Journal of Geophysical Research, 78,962-968.

Bates, N. R. (2006), Air-sea C02 fluxes and the continental shelf pump of carbon in the Chukchi Sea adjacent to the Arctic Ocean, Journal of Geophysical Research, 111, doi:10.1029/2005JC003083.

Borges, A. V. (2005), Do we have enough pieces of the jigsaw to integrate CO2 fluxes in the coastal ocean?, Estuaries, 28, 3-27.

Borges, A. V., et al. (2005), Budgeting sinks and sources of CO2 in the coastal ocean: Diversity of ecosystems count, Geophysical Research Letters, 32, doi:10.1029/2005GL023053.

Borges, A. V., and M. Frankignoulle (1999), Daily and seasonal variations of the partial pressure of C02 in surface seawater along Belgian and southern Dutch coastal seas. Journal of Marine Systems, Journal of Marine Systems, 19, 251-266.

Borges, A. V., et al. (2004), Variability of the gas transfer velocity of C02 in a macrotidal estuary (the Scheldt), Estuaries, 27, 593-603.

Cai, W.-J., et al. (2006), Air-sea exchange of carbon dioxide in ocean margins: a province-based synthesis, Geophysical Research Letters, 33, doi:10.1029/2006GL026219.

Chen, C, et al. (2003), Continental Margin Exchanges in Ocean Biogeochemistry: The Role of the Ocean Carbon Cycle in Global Change, edited by Fasham, pp. 53-94, Springer.

Cole, J. J., and N. F. Caraco (2001), Carbon in catchments: connecting terrestrial carbon losses with aquatic metabolism, Marine and Freshwater Research, 51,101-110.

Cole, J. J., et al. (1994), Carbon dioxide supersaturation in the surface waters of lakes, Science, 265,1568-1570.

Page 58: Air-water carbon dioxide exchange in relation to chemical

42

Dery, S. J., and E. F. Wood (2005], Decreasing river discharge in northern Canada, Geophysical Research Letters, 32, doi:10.1029/2005GL022845. Dittmar, T., and G. Kattner (2003), The biogeochemistry of the river and shelf ecosystem of the Arctic Ocean: a review, Marine Chemistry, 83,103-120.

DOE (1994), Handbook of methods for the analysis of the various parameters of the carbon dioxide system in sea water. Version 2, in ORNL/CDIAC-74, edited by A. G. Dickson and C. Goyet.

Ducklow, H. W., and S. L. McCallister (2004), The biogeochemistry of carbon dioxide in the coastal oceans, in The Sea, Volume 13, edited by A. R. Robinson, Brink, K.H., Harvard University Press.

Eicken, H., et al. (2000), A key source area and constraints on entrainment for basin scale sediment transport by Arctic sea ice, Geophysical Research Letters, 26,1919-1922.

Guo, L., and R. W. Macdonald (2006), Source and transport of terrigenous organic matter in the upper Yukon River: evidence from isotope (613C, Al4C, and 6l5N) composition of dissolved, colloidal, and particulate phases, Global Biogeochemical Cycles, 20.

IPCC (2007), Summary for Policymakers, Cambridge University Press, Cambridge, U.K. New York, U.S.A.

Jahne, B., et al. (1987b), Measuremen of the diffusion coefficients of sparingly soluble gases in water, Journal of Geophysical Research, 92,10767-10776.

Liss, P. S. (1983), Gas Transfer: experiments and geochemical implications, in Air-Sea exchange of gase and particles, edited by P. S. Liss and W. G. N. Slinn, pp. 241-298, D. Reidel Publishing Company.

Liss, P. S., and L. Merlivat (1986), Air-sea gas exchange rates: introduction and synthesis, in The role of air-sea exchange in geochemical cycling, edited by P. Buat-Menard, pp. 113-127, D. Reidel

Macdonald, R. W., et al. (1998), A sediment and organic carbon budget for the Canadian Beaufort Shelf, Marine Geology, 144, 255-273.

Page 59: Air-water carbon dioxide exchange in relation to chemical

43

Macdonald, R. W., and Y. Yu (2006), The Mackenzie Estuary of the Arctic Ocean in Handbook of Environmental Chemistry (Estuaries), edited by Springer-Verlag, pp. 91-120, Berlin Heidelberg.

McGillis, W. R., and R. Wanninkhof (2006), Aqueous C02 gradients for air-sea flux estimates, Marine Chemistry, 98,100-108.

Millero, F. J. (2000), The carbonate system in marine environments in Chemical Processes in Marine Environments, edited by A. Gianguzza, Pelizzetti, E., Sammartano, S., pp. 9-41, Springer.

Nightingale, P.D., et al. (2000), In situ evaluation air-sea gas exchange parameterizations using novel conservative and volatile tracers, Global Biogeochemical Cycles, 14,373-387.

Portielje, R., and L. Lijklema (1995), Carbon dioxide fluxes across the air-water interface and its impact on carbon availability in aquatic systems, Limnology and oceanography, 40, 690-699.

Prinsenberg, S.J. 1986. Salinity and temperature distribution of Hudson Bay and James Bay. Elsevier Oceanographic series, 44,163-186.

Raymond, P. A., et al. (2000), Atmospheric C02 evation, dissolved inorganic carbon production, and net heterotrophy in the York River estuary, Limnology and Oceanography, 45,1707-1717.

Redfield, A. C, et al. (1963), The influence of organisms on the composition of sea-water, in The sea, edited by M. N. Hill, pp. 26-77, Wiley.

Rosenberg, D. M., et al. (1997), Large-scale impacts of hydroelectirc development, Environmental Reviews, 27-54.

Rouse, W. R., et al. (1997), Effects of climate change on the freshwaters of arctic and subarctic North America, Hydrological Processes, 11.

Sakshaug, E. (2004), Primary and secondary production in the Arctic seas, in The oganic carbon cycle in the Arctic Ocean, edited by R. Stein, Macdonald, R., pp. 57-81, Springer.

Sarmiento, J. L., and N. Gruber (2006), Ocean Biogeochemical Dynamics, 503 pp., Princeton University Press, Princeton.

Page 60: Air-water carbon dioxide exchange in relation to chemical

44

Sarmiento, J. L, et al. (1998), Simulated response of the ocean carbon cycle to anthropogenic climate warming, Nature, 393, 245-249.

Schindler, D. W. (1974), Eutrophication and Recovery in Experimental Lakes: Implications for Lake Management, Science, 184, 897-899. Schindler, D. W. (1977), Evolution of phosphorus limitation in lakes, Science, 195, 260-262.

Stumm, W., and J. J. Morgan (1996a), Dissolved carbon dioxide in Aquatic Chemistry, edited by W. Stumm, Morgan, J.J., pp. 148-205, Wiley-Interscience.

Stumm, W., and J. J. Morgan (1996b), Atmosphere-Water interactions in Aquatic Chemistry, edited by W. Stumm, Morgan, J.J., pp. 206-249, Wiley-Interscience.

Takahashi, T., et al. (1993), Seasonal variation of C02 and nutrients in the high-latitude surface oceans: a comparative study, Global Biogeochemical Cycles, 7,843-878.

Thorpe, S. A. (1982), On the clouds of bubbles formed by breaking wind-waves in deep water, and their role in air-sea gas transfer, 304,155-210.

Tsunogai, S., et al. (1999), Is there a "continental shelf pump" for the absorption of atmospheric C02?, Tellus, 51B, 701-712.

Vorosmarty, C, and D. Sahagian (2000), Anthropogenic disturbance of the terrestrial water cycle, BioScience, 50, 753-765.

Wanninkhof, R. (1992), Relationship between wind speed and gas exchange over the ocean, Journal of Geophysical Research, 97, 7373-7382.

Wanninkhof, R., et al. (1985), Gas exchange-wind speed relation measured with sulfur hexafluoride on a lake, Science, 227,1224-1226.

Wanninkhof, R., and W. R. McGillis (1999), A cubic relationship between air-sea C02

exchange and wind speed, Geophysical Research, 26,1889-1892.

Weiss, R. F. (1974), Carbon dioxide in water and seawater: the solubility of a non-ideal gas, Marine Chemistry, 2, 203-215.

Weiss, R. F., and B. A. Price (1980), Nitrous oxide solubility in water and seawater, Marine Chemistry, 8, 347-359.

Page 61: Air-water carbon dioxide exchange in relation to chemical

45

Wetzel, R. G. (2001), Limnology: Lake and River Ecosystems, 3 ed., Academic Press.

Whitman, W. C. [1923), The two-film theory of gas absorption, Chemical metallic engineering 29,146-148.

Zeebe, R. E., and D. Wolf-Gladrow (2001), C02 in seawater: equilibrium, kinetics, isotopes, 346 pp., Elsevier Ltd., Amsterdam.

Page 62: Air-water carbon dioxide exchange in relation to chemical

46

Chapter 3: Summer air-water carbon dioxide exchange and water chemistry of the Churchill River as it enters Hudson Bay

Emmelia C. Stainton1, Raymond H. Hesslein2, Tim N. Papakyriakou1, and David G. Barber1

1Centre for Earth Observation Science Department of Environment and Geography, University of Manitoba, Winnipeg, Manitoba R3T2N2, Canada. fisheries and Oceans Canada Freshwater Institute, Department of Fisheries and Oceans, 501 University Crescent, Winnipeg, Manitoba R3T2N6, Canada.

Abstract

A continuous CO2 monitor was moored in the central channel of the Churchill

River with the purpose of measuring gas concentrations in water and air allowing

the calculation of the CO2 exchange of the river just before it enters the estuarine

system of Hudson Bay. The CO2 flux ranged from -51.4 to 30.8 with a mean flux of

-5.8 mmol nr2 day-1. The delivery rate of C02wto Hudson Bay via the Churchill River

estuary was calculated based on morphometry and flow, and ranges from 39 to

1335 kmol day1, with a mean rate of 355 ± 328 kmol day1. Net production (Pnet)

and respiration (R) for the Churchill River were estimated directly from diurnal

peaks measured using the gas monitor, resulting in a mean Pnet of 82.6 and a mean R

of 77.0 mmol m2 day1 respectively. These results indicate that the Churchill River is

a net CO2 sink throughout much of the 2007 open-water season, which is striking

Page 63: Air-water carbon dioxide exchange in relation to chemical

47

given that high latitude rivers and aquatic systems in general are considered strong

sources of atmospheric CO2.

3.1 - Introduction

Arctic and sub-arctic rivers, many with expansive watersheds covering

several biomes, pass over carbon rich permafrost soils that currently store over

one-third of the global terrestrial soil carbon (Gorham 1991). This process results in

high concentrations of organic carbon in rivers, which are available for

decomposition to dissolved CO2 (CChw). C02w is determined by photosynthetic

uptake, decomposition of organic carbon (respiration), air-water gas exchange, and

precipitation/dissolution reactions (Wetzel 2001). These factors are in turn

controlled independently by physical, chemical, and biological processes, resulting

in many variables influencing the aquatic carbon cycling.

An abundance of organic carbon in most high latitude rivers and lakes is the

reason that many are net heterotrophic systems, typically with CO2 levels in surface

waters well in excess of the overlying atmosphere (Kling et al. 1992; Cole et al.

1994). This excess of CO2 in surface waters evades into the atmosphere. However, a

recent study by Tank et al. (2008) found several shallow lakes of the Mackenzie

Delta (with varying connectedness to the Mackenzie River in western Canada) to be

sinks for CO2 throughout the summer. With dissolved organic carbon (DOC)

concentrations similar to those reported for other high latitude rivers (Guo and

Macdonald 2006; Wetzel 2001), the partial pressure of CO2 (PCO2) in surface waters

Page 64: Air-water carbon dioxide exchange in relation to chemical

48

reported for most of the Mackenzie Delta lakes investigated was less than

atmospheric pC02, and interestingly was lowest when DOC was highest (Tank et al.

2008].

Biogeochemical processes and carbon dynamics in rivers are complicated by

the fact that they reflect physical, chemical, and biological conditions integrated

throughout the watershed. As a result only a few studies have investigated the

relationships among riverine biogeochemistry, flow dynamic and delivery

constituents, and air-riverine CO2 exchange (Telmer and Veizer 1999). Most

freshwater CO2 flux estimates have been made for mid-latitude lakes (Kelly et al.

2001; Sobek et al. 2005) and rivers (e.g. Jones et al. 2003), though some stream

studies have been published in recent years (Hope et al. 2001, Jones and Mulholland

1998). Few studies exist for high latitude lakes and river systems (Kling et al. 1992;

Hamilton et al. 1994; Tank et al. 2008), where baseline estimates are important,

especially since climate scale temperature changes are expected to influence DOC

mobilization (Guo and Macdonald 2006; Guo et al. 2007), seasonal hydrological

processes, and metabolic rates (Rouse et al, 1997). By comparison several studies

have reported on the nature of gaseous carbon cycling from northern wetlands and

bogs, with a strong regional representation within the Hudson Bay Lowland (HBL)

area of sub-arctic central Canada (Macrae et al. 2004).

Hudson Bay drains a vast area (~3.7xl06 km2, approximately one third the

area of Canada) and receives a huge amount of freshwater —710 km3 year1 (18%

of the annual discharge to the Arctic Ocean) - from several major river systems

Page 65: Air-water carbon dioxide exchange in relation to chemical

49

(Dery et al. 2005). The natural flow and annual discharge cycle of many of these

rivers has been altered by hydroelectric development (Dery et al. 2005).

The lower reaches of inflowing rivers pass through the HBLs along the south

and south-western fringe of Hudson Bay over soils that are rich in organic carbon

(OC) and nutrients, resulting in high input of DOC in the coastal marine waters

(Granskog et al. 2007). The prevalence of high DOC associated with freshwater

plumes within the Bay differentiates this water body from other coastal seas within

the Canadian Arctic, and as a result Hudson Bay is far less effective in taking in

atmospheric CO2 relative to other peripheral seas (Else et al. 2008), which

characteristically are strong CO2 sinks (e.g. Miller et al. 2002).

To date, studies have reported on the OC delivery into coastal environments

by rivers, but none have examined the nature of the air-water CO2 exchange

upstream of the estuarine environment. This information would provide insight into

possible differences in carbon cycling that may exist between the riverine and

estuarine environments. Despite some recent advances in our understanding of the

physiochemical and gas exchange characteristics of the estuary/shelf system in

Hudson Bay (Else et al. 2008), little is known of the air-water CO2 exchange and

water chemistry of the inflowing rivers.

The primary objective of this study is to describe air-water CO2 exchange at

the lower reaches of the Churchill River waters as it enters the Churchill River

estuary and to relate observed fluxes to basic flow dynamics and characteristics. In

the following we present an overview of 1) the variability of dissolved/atmospheric

Page 66: Air-water carbon dioxide exchange in relation to chemical

50

CO2/O2 and general water chemistry variables measured in the Churchill River

throughout the 2007 open-water season, 2) the flux of CO2 from collected field data

estimated using the thin boundary-layer (TBL) model (Wanninkhof 1992), 3)

production and respiration estimates for the lower reaches of the Churchill River,

and 4) the delivery rate of dissolved CO2 from the Churchill River to Hudson Bay.

This represents part 1 of a 2-part study examining the relationships among

biogeochemistry, flow dynamic and delivery constituents, and air-water CO2

exchange in both a river and estuarine system. It is important that we understand

the processes that define the transport and transformation of carbon from the

terrestrial landscape, through the aquatic network and into the marine system. A

better understanding of the inflowing Churchill River will allow us to gauge the

relative contributions of marine and riverine (terrestrial] influences in the mixing

estuarine waters, and ultimately identify the Churchill River watershed's input to

Hudson Bay. This research is part of a larger study that investigates the coupling

between freshwater quality and quantity and marine processes within Hudson Bay,

and how Hudson Bay may be impacted by the combined effects of climate change

and hydroelectric development (Freshwater-Marine Coupling in the Hudson Bay

IRIS, ArcticNet, http://www.arcticnet.ulaval.ca).

Page 67: Air-water carbon dioxide exchange in relation to chemical

51

3.2 - Methods

All sampling associated with this work took place during the summer of 2007

at the lower reaches of the Churchill River (station CH7, Lat: 58.6693°, Long: -

94.1883°, Figure 3.1a]. The Churchill River enters into the coastal waters of

southwestern Hudson Bay (Lat: 58.7910°, Long: -94.2082°; Figure 3.1a) draining an

area of 281,300 km2 across the Canadian Shield, the HBLs, and the northern interior

prairies where the river's headwaters are situated. Approximately two-thirds of the

lower Churchill River's historical flow was diverted in 1976 for hydroelectric power

generation on the Nelson River, and hence the Churchill River's discharge is

controlled. The resulting reduction in water level negatively impacted use of the

river by residents of Churchill, Manitoba, and as a result a rock weir extending the

width of the river (~ 2km) was constructed approximately 10km upstream and

south of the town site in 1998. This raised the local water level of the Churchill River

by approximately 2m, flooding an area of approximately 6.5 km2 extending 10km

upstream (south) from the weir (Bezte, 2006). Our sampling location was situated

approximately 0.5 km upstream of the weir (Figure 3.1a). Sampling was done in two

ways: through discrete sample collection, and though continuous monitoring.

Page 68: Air-water carbon dioxide exchange in relation to chemical

52

Figure 3.1. a) Map of the Churchill River as it enters Hudson Bay (Landsat 7 satellite image, Natural Resources Canada] b) CO2/O2 air/water gas monitor moored on float at Churchill River station CH7. Two 20W solar panels [right side) charge two 12V batteries that power the gas monitor system (inside white box).

3.2.1 - Discrete water sampling

A Van Dorn style water sampler was used to collect surface water samples

(0.25m) at station CH7 1-2 times per month. Dissolved inorganic carbon (DIC) was

collected from the water sampler first by piercing the outlet tubing and drawing 4ml

of water into a syringe triple-rinsed with sample. This was expelled through a

septum cap into Milli-Q triple-rinsed 12 ml borosilicate glass vials containing 20 u.1

of concentrated phosphoric acid. These vials had been evacuated and filled three

times (20 s cycles) with ultra high purity (UHP) helium containing 1% neon and 1%

acetylene. Neon and acetylene had been added as tracers to determine whether

sample contamination from atmosphere had occurred and to validate the

performance of the gas chromatograph used. Using this approach, neon, a very

Page 69: Air-water carbon dioxide exchange in relation to chemical

53

insoluble gas, should be the same concentration in all samples unless the sample is

contaminated with air. Acetylene is similar to CO2 in solubility. The ratio of

acetylene/neon changes relative to sample volume making the addition of this gas a

useful check for sampling technique. Vials were pressurized to 6 PSI such that

leakage, if it occurred, would be out and not into the vials. In the field, vials were de-

pressurized to atmospheric pressure by piercing the septum with a 20-gauge needle

prior to sample collection. Addition of 4 mL of sample liquid to the sealed vial raised

the internal pressure to again insure that any leakage would be out of the vials.

Added acid both preserved the sample and converted the DIC contents to CO2.

A proportion of the water sample for other variables was transferred into 2L

amber polyethylene bottles that were acid-cleaned with 10% hydrochloric acid

(HCL) solution and triple rinsed with Milli-Q water. Upon sample collection, bottles

were triple rinsed with a small aliquot of sample water, filled to the shoulders,

tightly capped, and kept cool until processed. Laboratory processing took place

immediately upon return from the field. Sub-samples were collected for DOC,

particulate organic carbon and nitrogen (POC/N), chromophoric dissolved organic

matter (CDOM), alkalinity, chlorophyll a (Chla), and total suspended solids (TSS)

analyses.

Background water chemistry data (pH, DOC, POC/N, and nutrients) had been

collected in the vicinity of station CH7, and was made available to this study by the

provincial hydroelectric provider, Manitoba Hydro. These samples were collected

and analyzed during 1995-2005. Ancillary information on Chla and POC/N were

Page 70: Air-water carbon dioxide exchange in relation to chemical

54

also available to this study from June and July 2006 as part of another study by

Centre for Earth Observation Science (Department of Environment and Geography,

University of Manitoba) personnel.

3.2.2 - Continuous CCh, 02, and water sampling

In situ measurement of dissolved gasses has been a successful technique for

determining air/water concentrations and flux estimates (Sellers et al. 1995). Here,

a continuous automated air-water CO2 monitoring instrument (following Carignan,

1998) was used to measure CO2 and O2 concentration in both air and water. This

instrument was moored on a 4 by 6 foot float and powered by two 20-watt solar

panels connected to two 12-volt deep cycle batteries. The water intake tube was

covered with 100-micron copper mesh and positioned at 25cm below the water

surface, while the air intake was affixed at approximately 1 meter above the water

surface (Figure 3.1b).

The gas monitoring instrument (GM) consists of two main parts: 1) a "dry"

box containing a gas pump (Spectrex Corp.® AS-200), an infrared CO2 gas analyzer

(LiCor® LI 820), a programmable data logger controller (Campbell Scientific®

CR10X), an oxygen sensor (Qubit® S101 diffusion based O2 sensor), and in-line gas

drying Nafion tubing (Perma Pure® LLC ME-060-24BB) surrounded by silica

desiccant 2) a "wet" box containing a peristaltic pump (Autoclude® M500), a gas

exchanger (Celgard Inc.®), and a thermistor (Omega Engineering®) (Figure 3.2).

Page 71: Air-water carbon dioxide exchange in relation to chemical

Figure 3.2. Schematic of air/water gas monitor: wet box consisting of a peristaltic water pump, a thermocouple housing, a gas exchanger, and two solenoid valves (SV); dry box consisting of an air pump, drying compartment, O2 sensor, IR gas analyzer, and data logger.

DRY BOX

drying compartment

infrared gas

analyzer

Air IN -(gas loop)

WET BOX

pump *~1 I I

data logger

+ r\.u

3 Air OUT

' gas loop)J ^SV

4 air/water gas exchanger

4 = 3 1 r m-ling I X

thermocouple n n

H20 peristaltic

pump

Air OUT (gas loop)

H20 OUT • 4

H2OIN

Atmospheric CO2 and O2 was measured by drawing air into the system with

the gas pump, drying it with the in-line Nafion tubing, and passing it through the

oxygen sensor and CO2 analyzer before leaving the unit. Dissolved gas in water was

measured by drawing water into the wet box via the peristaltic pump, through the

in-line thermistor, and then through the gas permeable tubing in the gas exchanger

before leaving the system. An air pump ran concurrently with the peristaltic pump

Page 72: Air-water carbon dioxide exchange in relation to chemical

56

in a closed loop that was connected to the outer chamber of the gas exchanger. Gas

in this closed loop equilibrates with the dissolved gasses in the water flowing

through the gas exchanger, and is then pumped through the air loop (drying

compartment, O2 sensor, CO2 analyzer) (Figure 3.2).

The data logger was programmed to run the instrument for 30-minute cycles

every 2 hours throughout the 2007 open-water season. The CO2 analyzer was

calibrated monthly with gas standard (span CO2 = 380 and 777 parts per million

(ppm)) and zeroed with UHP nitrogen.

For part of the 2007 open water season, a conductivity, temperature, depth,

pH, and turbidity probe (RBR XR-620) was moored in surface waters adjacent to the

GM's intake and programmed to log hourly.

3.2.3 - Other sources of data

River discharge at station CH7 was estimated by summing flows from the

Churchill River at Red Head Rapids (Lat: 58.1208°, Long: -94.625°) and the

inflowing Deer River (Lat: 58.015°, Long: -94.1956°) using daily discharge data

obtained from the Water Survey of Canada

(http://www.wsc.ec.gc.ca/hydat/H20/index_e.cfm).

Water and air temperature, wind speed, and water level data are routinely

monitored by Manitoba Hydro at the Churchill River pump-house station

(approximately 6km upstream of the weir), and the data were made available to this

study. Meteorological data (wind speed, air temperature, precipitation) were also

Page 73: Air-water carbon dioxide exchange in relation to chemical

57

collected at the Churchill weather station by Environment Canada

[www.weatheroffice.gc.ca/city/pages/mb-42_metric_e.html) and hourly data were

downloaded and used here.

3.2.4 - Laboratory procedures

A pre-combusted 25 mm glass fiber (GF/F) filter was positioned on an acid-

soaked [24 hours) and Milli-Q water triple-rinsed polypropylene filter holder and

tightly sealed with a Milli-Q water rinsed o-ring. An acid-soaked and Milli-Q water

rinsed polypropylene syringe was triple rinsed with a small volume of sample, and

then filled with 60 mL of sample. This sample was then pushed through the filter [in

filter holder), discarding the first 30 mL of sample, and collected into 1) a pre-

combusted 5mL glass vial containing 25 uL of 50% high pressure liquid

chromatography grade phosphoric acid [DOC), and then 2) a pre-combusted 50 mL

amber borosilicate glass vial [CDOM). Samples were stored at 4°C until analysis.

Another 60 mL of sample was collected and pushed through the same filter, filter

holder and syringe [total volume of 120 mL). This filter was folded in half, carefully

wrapped in pre-combusted aluminum foil, and frozen for analysis of POC/N.

Nutrient samples [NO2VNO3", PO43") were prepared after Simpson et al.

(2008) and concentrations were measured on a continuous flow Bran + Luebbe

Auto-Analyzer 3 [AA3). DOC samples were analyzed on a Shimadzu TOC-5000A.

CDOM absorbance was measured using a Hewlett-Packard 845 2A diode array

spectrophotometer in an acid-cleaned 10-cm quartz cell and scanned at 2 nm

Page 74: Air-water carbon dioxide exchange in relation to chemical

58

intervals between 250 and 820 nm against Milli-Q water. Absorption (X) was

calculated after Stedmon et al. (2000]. POC/N was measured on a Carlo Erba NA-

1500 Elemental Analyzer following acid fuming of sample filters to remove

carbonate (following Verardo et al, 1990).

Using a vacuum filtration system, particulate matter from a known volume of

sample (ranging from 250 to 500 mL) was collected onto 42.5mm GF/F filters for

the analysis of TSS (pre- combusted and weighed filters) and Chla. These were

stored in Petri dishes and kept frozen and dark (to reduce Chla degradation) until

analysis. TSS was determined after Stainton et al. (1977). Chlorophyll a

concentrations were measured on a Turner Designs fluorometer (10-AU) following

24-hour extraction in 90% acetone at 4°C (JGOFS 1994).

DIC samples were analyzed on a Varian CP4900 MicroGC. Gas was introduced

into the GC by piercing the septum cap of the vials with a needle and allowing gas to

flow (vials pressurized in the field at time of sampling) through Nafion drying tubing

to the GC sample loop. The sample line was flushed with UHP helium between

samples. Alkalinity samples were analyzed on a PC TitratlON Plus following the

Gran alkalinity titration method (Stainton et al. 1977).

3.2.5 - Calculations

C02w data measured by the GM was recorded as parts per million (ppm) and

was converted to concentration (umol L1) using:

[C02] = a • PC02

Page 75: Air-water carbon dioxide exchange in relation to chemical

59

where a is the solubility coefficient calculated for CO2 (mol L_1 atnr1) (Weiss and

Price 1980].

O2 measurements were converted from ppm to concentration (umol L1)

using:

a 0.2095 x 106

[OJ = x x pO0water 1 22.4136 /?02air

where a is the solubility coefficient calculated for O2 (mol I/1 atnr1) (Weiss 1970].

The flux of C02 was calculated after Wanninkhof (1992]:

Flux = a x kC02 x ApC02

where

^°2 = °-31Ul20mXf^)"1/2

1600/ f ( 4 ]

ApCCh is the air-water difference in pCOi, Uiom is the wind speed at 10m from the

water surface, and Sc is the Schmidt number (a dimensionless coefficient] (Jahne et

al. 1987). Fluxes are sensitive to the parameterization of gas transfer with wind

(see Appendix B). The Wanninkhof (1992) model was chosen because it lies in

between the highest and lowest k values calculated at high wind speeds. At low wind

speeds, little difference exists between the various models to calculate k.

Net production (Pnet) and respiration (R) rates were estimated using the

diurnal C02w concentration data as:

ACO. P^tCmmolm^day"1) = — x Z x 1000 - Flux

At (5)

Page 76: Air-water carbon dioxide exchange in relation to chemical

60

and

A CO. R (mmol m'2 day"1) = — x Z x 1000 + Flux

where ACChw is in mmol I/1, At is time in hours, Z is a measure of river depth (2.5m),

and Flux is in mmol nr2 hour1. ACChw/ At was determined by plotting CO2 w versus

time for each diurnal peak (upslope for R = increase in CChw, and down slope for

Pnet= decrease in CChw). Since CO2 air/water flux was calculated for each data point,

the average flux between the bottom and top (for R, or top and bottom for Pnet) of a

diurnal curve was either subtracted (equation 5) or added (equation 6) to arrive at

an estimate of Pnet and R. Gross production (Pgross) is the sum of Pnet and R from one

diurnal curve.

We estimated the approximate upstream distance from station CH7 that a

parcel of river water travels within a 24-hour period in order to identify the possible

reaches of the river where Pnet and R influence dissolved gas concentrations. This

was done by dividing the flow rate by the wetted perimeter (channel depth x

channel width), where flow is the sum of Churchill River flow at Red Head Rapids

and the inflowing Deer River, channel width was assumed to be 2000m (the

approximate width of the Churchill River at CH7), and channel depth was assumed

to be 2.5m (depth at station CH7). Flow data was also used to estimate the delivery

rate of CChw from the Churchill River to Hudson Bay. This was calculated by

multiplying CChw concentration (mmol L1 from the GM, averaged to match daily

flow data) by river discharge (L day1) to get kmol day1.

Page 77: Air-water carbon dioxide exchange in relation to chemical

61

3.3 - Results

3.3.1 - Environmental conditions

Climate data collected by the Environment Canada meteorological station in

Churchill indicate daily average air temperature peaks in late July and August, while

precipitation amount and frequency peaks in August and September (Figure 3.3a).

Discharge rates are initially high following spring ice breakup (June 5), decrease

steadily throughout July, and peak once again in early August coinciding with

several substantial rainfall events. On average, Churchill was warmer and dryer

(until September) in 2007 compared to climate norms (1971-2000) (Figure 3.3b).

Table 3.1: Mean and standard deviation of Churchill River annual flow and precipitation before and after the Churchill River Diversion (CRD, created in 1976) and before and after the Churchill River rock weir construction (10km south of the Town of Churchill, built in 1998).

Period Pre-CRD Post-CRD Pre-weir Post-weir

Year 1971-1976 1976-2008 1976-1998 1998-2008

Flow (mV) Mean S.D. 1361 280 380 435 373 386 396 521

Precipitation (mm) Mean S.D. 30.3 25.9 38.0 29.8 36.5 28.4 41.4 32.4

Historical discharge indicates the mean annual flow of the Churchill River

(metered at Red Head Rapids) has decreased by approximately 70% between pre-

(1971 to 1976) and post-diversion (1976 to 2008) flow regimes (Table 3.1, Figure

3.4). Prior to the diversion, the mean annual flow of the Churchill River was 1361 ±

280 m3s1, though in subsequent years mean annual flow was substantially reduced

Page 78: Air-water carbon dioxide exchange in relation to chemical

and was much more variable (380 ± 435 m3s1j (Table 3.1, Figure 3.4). Mean annual

precipitation was greater and more variable following the diversion (38 ± 29.8 mm,

compared to 30.3 ± 25.9 mm Table 3.1).

Figure 3.3. a) Churchill daily mean air temperature and daily precipitation (Environment Canada), and river discharge (combination of Red Head Rapids and the Deer River, Water Survey of Canada) throughout the 2007 open-water season, b) precipitation and air temperature monthly averages from 2007 and historical climate norms (1971-2000) (Environment Canada).

1000

5/1 5/15 5/29 6/12 6/26 7/10 7/24 8/7 8/21 9/4 9/18 10/2 10/16 10/30 Date

U l 2 O

I

31971-2000 climate norms precip

-1971-2000 climate norms temp

12007 precip -2007 temp

J A Month

Page 79: Air-water carbon dioxide exchange in relation to chemical

63

Following weir construction, the mean annual flow in the lower reaches of

the Churchill River increased slightly from 373 (pre-weir) to 396 m3s-1 (post-weir)

and was more variable (standard deviation (S.D.) of 385.9 for pre-weir to 521 m3s'1

for post-weir, Table 3.1), which can be attributed partly to the large peak flow in

2005 (Figure 3.4). Mean annual precipitation was also greater throughout the post-

weir period (41.4 mm compared to 36.5 pre-weir) and more variable (S.D. of 28.4

for pre-weir to 32.4 for post weir, Table 3.1)

Figure 3.4. Historical flow data from the Churchill River at Red Head Rapids (Water Survey of Canada) and monthly mean precipitation (Environment Canada) between 1971-2008.

180 dischage monthly mean precip

86 91 Year

3000

Page 80: Air-water carbon dioxide exchange in relation to chemical

64

3.3.2 - Water chemistry

Water chemistry collected from station CH7 between June and October 2007

is summarized in Table 3.2. TSS, chlorophyll a, and DIC values all peak in August

while alkalinity is highest from July to mid-August. DOC, CDOM, and NO2" + NO3" are

highest in late August and October. PON and POC follow a similar pattern where

concentrations are higher in June and August, and lower in early July and October.

PO4 is exceptionally high after ice-out, however concentrations decrease throughout

the open-water season to levels that are consistent with previous years data for the

Churchill River (Table 3.2, Figure 3.5). pH deviates most from its mean of 8.2 in July

(Table 3.2).

Table 3.2: Summary of water chemistry data from Churchill River site CH7, collected June-October 2007.

Parameter

pH

Alkalinity (ueq L"')

DIC (nmol L1)

DOC (mg I/1)

PON(mgL')

POC(mgL')

N0 2 +N0 3 (mgL ' ) PO^mgL-') Chlorophyll a (\ig L"1)

CDOM (abs 350 nm) TSS(mgL')

11-Jun

8.2

1306

1031

8.79

0.10

0.70

0.05 0.18 1.3

12.2 7.7

17-Jun

8.2

1361

1244

8.68

0.07

0.58

0.11 0.02 2.0

10.8 4.4

12-Jul

7.5

1423

1162

8.34

0.05

0.36

0.03 ~

2.7

8.7 2.3

Sampling Date 28-Jul

8.7

1531

1290

8.39

0.07

0.50

0.23 0.04 2.7

8.6 1.0

16-Aug

8.2

1448

1325

12.06

0.08

0.54

0.06 -

6.3

20.5 10.4

20-Aug

1375

1325

11.12

0.08

0.62

0.06 —

3.4

18.3 8.2

2-Oct

8.1

1312

1017

14.56

0.04

0.33

0.09 0.06 1.4

30.2 2.1

Page 81: Air-water carbon dioxide exchange in relation to chemical

Figure 3.5. Churchill River summer water chemistry data collected from 1995-2005 (data from Bezte, 2006) and from 2006-2007 (this study).

700

•J 600 o--3 500 .£> .S 400

300

200

100

T 0.2 J e» E, O 0.15 o-

O

§, 0 1

o z

n • Conductivity ApH

A

A

A A

A

tt

A

AA

A

*

A

AA

A

A

A A A A

• \

A

A

* A

L

. . . . K «NOVNO.

DPO;

'

a

D

w •

<fas a

D

• n •

D

t

25

20

o D

z o o. u o

8.6

8.4

8.2

a c

8

7.8

7.6

7.4

C ODOC + CHLA

«>

. °» « O +

• /

j oPOC BTSS U XPON

0

. X

^ X

• y 0

$ +

+ +

« *« +

— t H

0

/ ,

+

+

o. o

0

0

1

+

0

0

*

+

o

* > •

* «

+ '

++

'

/ 0°

°o m

,*

7

6 ^

._] 5 w>

3 .

4 I 3 g

2 U

1

0 25

20

10*-

0.5

Apr-95 Apr-97 Apr-99 Apr-01 Apr-03 Apr-05 Apr-07

Date

Page 82: Air-water carbon dioxide exchange in relation to chemical

66

Data indicate that little variation occurred in each of pH, conductivity,

chlorophyll a, DOC, POC, TSS or dissolved nutrients between 1994 and 2007, though

PON decreased throughout this sampling period (R2 of 0.68) [Figure 3.5). NO2/NO3

concentrations were more variable for samples collected in 2006 and 2007

compared to those between 1994 and 2005 [Figure 3.5).

3.3.3 - Continuous CO2,02, and water sampling

Air and water concentrations of CO2 and O2, water temperature, and wind

speed were measured at station CH7 for consecutive days in June [4), July (16), and

August (6 and 4). Mean and standard deviation of C02w, C02a, 02w, and 02a

concentrations were 11.7 ± 6.6,15.8 ± 1.7, 315.5 ± 16.3, and 309.5 ± 17.7 umol L1

respectively, while mean water temperature and wind speeds at CH7 were 16.8 ±

2.8 °C and 4.1 ± 2.4 m s-1 (Table 3.3). Interestingly, our C02w mean value is similar to

those reported by Hamilton et al. (1994) from several bogs in the HBLs during the

early summer. Tank et al. (2008) report similar C02w (published as PCO2) in four out

of six intensively studied shallow lakes of the Mackenzie Delta (western Canadian

arctic) from July to September 2006.

Dissolved and atmospheric gas concentrations and water temperature were

plotted over time and vary diurnally throughout most of the monitoring period.

From the three periods of near-continuous operation, C02Wvaries asynchronously to

02w and water temperature most obviously in July (Figure 3.6). C02w drops during

the daylight hours while 02w levels increase, both of these as a result of primary

Page 83: Air-water carbon dioxide exchange in relation to chemical

67

production. Not surprisingly water temperature follows a diurnal pattern of daytime

warming and nighttime cooling, although the signal is not as strong as for the

dissolved gasses. Atmospheric gas concentrations also fluctuate diurnally, following

a similar pattern to their dissolved counterparts (i.e. atmospheric CO2 tends to

increase at the same time as a decrease in dissolved CO2, and vice versa).

Table 3.3: Mean, standard deviation and range of (i) all measured gas monitor data (n = 224), (ii) flux (n = 224), and (iii) net production, respiration, and gross production at 2.5m (Pnet/down-slope peak, n = 27; R/upslope peak, n = 26; Pgross/complete peak, n = 24) from Churchill River station CH7.

Parameter Mean S.D. Max Min

TW(°C)

Wind (m s"1)

(^ (umolL 1 )

02w (umol L"1)

C02a (umol L"1)

C02w (y,mol L"1)

Flux (mmol m"2 day"1)

Pnet (mmol m"2 day"1)

R (mmol m"2 day"1)

PBross (mmol m"2 day"1)

16.8

4.1

309.5

315.5

15.8

11.7

-5.8

82.6

77.0

155.3

2.8

2.4

17.7

16.3

1.7

6.6

11.9

58.9

44.1

95.5

21.3

13.2

353.9

357.4

20.1

28.5

30.8

200.7

168.2

355.6

10.3

0.1

283.2

285.9

13.3

3.4

-51.4

5.8

18.5

31.3

At the beginning of August the concentrations of C02w and 02w increase and

their diurnal signals no longer follow the trend seen throughout much of July

(Figure 3.6). This takes place simultaneously with increases in river discharge,

water level, and CTDP measured turbidity, and decreases in pH, salinity, and water

temperature (Figure 3.7).

Page 84: Air-water carbon dioxide exchange in relation to chemical

68

Figure 3.6. Diurnal fluctuations of CO2W, CO2A, O2W, O2A, and Tw at Churchill River site CH7 throughout the 2007 open-water season. Data were measured and recorded at 2-hour intervals using a continuous air/water CO2/O2 gas monitor.

v V

340 •

+ 320 S

3 300 a.

8/9 8/11

3.3.4 - C02 flux

Air-river flux estimates range between -51.4 to +30.8 mmol nr2 day-1 (mean -

5.8 mmol nr2 day1, Table 3.3) and indicate that this system is a net sink for CO2

throughout most of the 2007 open-water season. Fluxes tended to be more negative

during the longest sampling period (July) when river discharge was declining and

most water chemistry variables and water temperature were increasing (Figure

3.8). Days are also longer during this period. Fluxes shift to being mostly positive at

the very end of July, concurrent with increasing river discharge, decreasing water

temperature, and increasing turbidity (Figure 3.7). The transformation from a CO2

sink to a source at the end of July is likely due in part to light limitation of primary

production within the river.

Page 85: Air-water carbon dioxide exchange in relation to chemical

69

Figure 3.7. a) and b) Turbidity, salinity, water temperature and pH collected at Churchill River station CH7 (data measured and recorded using an RBR XR-620) c) Churchill River discharge (combination of Red Head Rapids and the Deer River, Water Survey of Canada) and water level (recorded at the Churchill pump house station, approximately 3km upstream of station CH7)

7/27 7/29 7/31 8/10 8/12 8/14 8/16 Date

Tremblay et al. (2006) report CO2 fluxes from the Lee River in southern

Manitoba and some northern Quebec rivers and reservoirs that are well within the

range we estimate for the Churchill River throughout the 2007 open water season

(Table 3.4). Our results are also consistent with those reported by Tank et al.

(2008), who found that several Mackenzie delta lakes were net CO2 sinks

throughout the summer growing season. However, the results from this study differ

from those mentioned above in three key ways: 1) our fluxes were determined from

Page 86: Air-water carbon dioxide exchange in relation to chemical

70

continuous measurements, and therefore our mean reflects the net carbon

movement and is a true indicator of net flux, 2) we collected a higher frequency of

data (n = 224), and 3) our results are over a much tighter range [82 mmol nr2 day1).

Figure 3.8. Estimated CO2 flux from measured values at Churchill River site CH7 throughout the 2007 open-water season. A positive flux indicates a supersaturation of CO2 in water with respect to air (source), while a negative flux indicates an undersaturation of CO2 in water with respect to air (sink). The Churchill River is a CO2 sink throughout most of the monitored open-water season.

7/28 7/30 8/1

Date

3.3.5 - Production and respiration

Estimated Pnet, R, and Pgross for the Churchill River average 82.6 ± 58.9, 77.0 ±

44.1, and 155.3 ± 95.5 mmol nr2 day1 respectively for the water depth observed at

CH7 (2.5m) (Table 3.3 and Figure 3.9). Kling et al. (1992) report primary production

values (based on 24hr incubation of discrete samples with 14C) from Toolik Lake

(northern Alaska, United States of America) surface waters ranging from 0 to ~125

mgC nr3 day1, which is at the lower end of our estimated Pnet range (101 to 3530

mgC nr3 day1, converted from mmol nr2 day1).

Page 87: Air-water carbon dioxide exchange in relation to chemical

71

Table 3.4: Summary of flux estimates made by Tremblay et al. (2006) for a Manitoba River and some northern Quebec hydroelectric reservoirs. Samples were collected from a 20L chamber connected to a non-dispersive infrared gas analyzer, and data was stored every 20s over a 5-10 minute sample period.

Site Caniapiscau River Laforge 2 La Grande 3 La Grande 4 Lee River

Site type reservoir reservoir reservoir reservoir

river

Location lat 54.95°, long 69.49° lat 54.57°, long 70.15° lat 53.72°, long 75.68° lat 54.36°, long 73.13° lat 50.33°, lone 95.84°

Year sampled 2003 2003 2003 2003 2002

N 28 11 35 27 5

co2

Mean 15.20 18.93 38.79 26.77 7.50

(mmol

S.D. 23.38 29.06 29.04 15.38 24.45

m"2 day1)

Min Max -0.27 95.71 -6.29 95.37 -5.70 114.63

-27.18 58.94 -32.65 247.51

The Churchill River varies in channel width, depth, and substrate

composition, which will influence physical and chemical water properties, gas

exchange, and Pnet and R throughout its length. Pnet is the rate of drawdown of CCbw,

per unit time, per unit area, at a given depth, though this drawdown occurs

simultaneous to air-water interface gas exchange. The surface flux of CO2 is thus

subtracted from the rate of CO2 drawdown when flux is positive, and added when

flux is negative so that we do not overestimate the rate of change attributed to

biological production. R is thus the increase of CChw (per unit time, per unit area, at a

given depth), and is corrected by subtracting the flux when negative, and adding the

flux when positive. Since both A[C02]w and CO2 flux vary throughout the day, the

relative correction of CO2 flux to Pnet and R estimates also varies (mean corrections

of 11% and 10.8% respectively).

Page 88: Air-water carbon dioxide exchange in relation to chemical

72

Figure 3.9. Estimated Pnet and R throughout the 2007 open water-season at Churchill River station CH7. Both Pnet and R were calculated from gas data measured using a continuous air/water gas monitor.

6/17 6/19

The Churchill River delivers between 39 to 1335 kmol CO2W day-1 to the Churchill

River Estuary and Hudson Bay (Figure 3.10), and represents a slight dilution based

on surface water CO2W concentration measurements made in the estuary basin and

Hudson Bay from June to October, 2007 (range 14.5 to 24.2 umol L1, mean 17.7 ±

2.8 |imol I/1, n = 34; E.C. Stainton, unpublished data). Clearly a function of river

discharge, the delivery rate of C02w declines throughout July (as discharge and

water level decline) and sharply increases in late July/early August as water levels

and river flow rate rises.

3.4 - Discussion

Relating an observed air-surface gas exchange to river properties and

processes is complicated because, depending on flow rate, conditions being

measured do not necessarily reflect biogeochemical processes at the site, but rather

Page 89: Air-water carbon dioxide exchange in relation to chemical

73

the integration of processes over a stretch of the river and period of time. For part of

the monitoring period, our GM-measured air/water gas data has a time lag [up to 12

hours) from the diurnal signal expected for the light/dark periods of Churchill, and

we believe this to be dependent on the Churchill River's discharge. A faster flow rate

brings water that has traveled a greater distance in 24-hours, and thus represents

gas exchange and the influence of Pnet and R that have occurred over a greater

distance and at an earlier time upstream. Conversely, with slower flow the water

being sampled is presumably influenced to a greater extent by local primary

production and air-water gas exchange.

Figure 3.10. Delivery rate of CC>2w to the Churchill River estuary and 24-hour upstream flow distance upstream from station CH7.

16001 r 14

1400 +

O 400-U

200 f — 24-hour flow distance upstream

o Delivery rate 0 -I 1 ^

8 ° ^ o o >

12 £ cr o

3

o

C </> >-t a> P3

+ 6

+ 4

6/17 6/24 7/1 7/8 7/15 7/22 7/29 8/5 8/12

Date

Page 90: Air-water carbon dioxide exchange in relation to chemical

74

For this reason, we estimated the upstream distance that a parcel of river

water travels within a 24-hour period. Based on flow and a 24-hour travel time, the

furthest upstream source is ~14 km (maximum flow] while the closest is ~3.5 km

(minimum flow) (Figure 3.10). The reaches of the river influencing the dissolved gas

measurements, CO2 flux, and Pnet and R are local at either end of the recorded flow

range and within the area flooded by the Churchill River weir construction

(approximately 10 km upstream of the weir). The relatively uniform basin

characteristics of the stretch of the river influencing the water we monitored are

consistent with the assumption to use the wetted perimeter area when estimating

Pnet and R.

Our Pnet and R estimates are unique in that they represent: 1) the integral of

riverine Pnet and R. This is a significant improvement over approaches that focus on

estimating pelagic or benthic Pnet and R, 2) have used methods that usually don't

measure R, and 3) have historically used discrete samples collected in very narrow

windows of space and time, and as such have huge variance.

Estimation of the C02w delivery rate is important because it provides a

reference point for estimating estuarine CO2 exchange. Implications of an altered

hydrological cycle and temperature regime resulting from climate warming are

largely unknown for high latitude watersheds, and therefore further changes in flow

of the Churchill River and riverine biogeochemical cycling with respect to CCtew will

affect the load delivered to Hudson Bay. It is thus necessary to establish baseline

conditions as climate change has the potential to influence the delivery of nutrients

Page 91: Air-water carbon dioxide exchange in relation to chemical

75

and carbon from this and other large high latitude rivers. Regional variation in both

the quality and quantity of riverine inputs is expected, which will affect coastal

regions as the fresh water delivered integrates climate change impacts over entire

watersheds.

The interrelationship between nutrients, respiration, and temperature will

determine the fate of high latitude aquatic systems as sources or sinks for CO2 (Kling

et al. 1992). It is generally accepted that nutrient concentrations and light, rather

than temperature, limit primary production in high latitude aquatic and terrestrial

environments (Billings et al. 1984). However decomposition and respiration rates

are temperature dependent (Billings et al. 1982) making these high latitude

environments potential sources of CO2 with increasing mean annual temperature. In

spite of this, sufficient nutrient concentrations and increasing water temperatures

could enhance aquatic primary production and thus provide a greater sink for CO2.

This complex interplay of variables, most uncertain for a changing climate, makes

projections for the future role of high latitude aquatic systems as CO2 sources or

sinks speculative.

Using a continuous GM was a successful method for tracking changes in

air/water CO2 flux, Pnet, and R in the Churchill River, and could provide a means to

monitor climate feedbacks (water flow and temperature changes) in the future, as

climate change has the potential to change Pnet, R, and runoff water quality in

response to changes in hydrology and temperature. Little is known about the impact

of changes in hydrology and temperature in high latitude lakes and rivers, however

Page 92: Air-water carbon dioxide exchange in relation to chemical

76

this data establishes a background for further studies/monitoring on a river that is

well positioned (location, large area integrated, ease of access, local infrastructure,

etc.) to be a sentinel watershed in tracking climate change.

Production and respiration estimates here indicate that the Churchill River is

a net productive system as it enters Hudson Bay, which is contrary to much of the

existing CO2 exchange literature for lakes and rivers (Kling et al. 1991; Cole et al.

1994; Sobek et al. 2005). The view that, globally, aquatic ecosystems are sources of

CO2 to the atmosphere has been perpetuated in the literature over the past >10

years, though recent findings reported here and by Tank et al. (2008) suggest

otherwise.

This paper represents one of few accounts of carbon cycling in a high latitude

river throughout the peak growing season, and as a result further studies examining

high latitude air-water CO2 exchange are required to better understand this issue.

Identifying current conditions may lead to more confident predictions of future

nutrient and carbon cycling in high latitude rivers, and may provide some insight

into how rivers will affect upstream estuarine and marine coastal environments.

Page 93: Air-water carbon dioxide exchange in relation to chemical

77

Acknowledgements

We gratefully acknowledge the field support and assistance from K. Swystun,

R. Whitten, P. Fitzpatrick, J. Batstone, and the Churchill Northern Studies Centre. We

thank Manitoba Hydro, Water Survey of Canada, and Environment Canada for

providing river discharge, water chemistry, and meteorological data. We also thank

M. Stainton for helpful discussion and comments. E. Stainton is funded by awards

from the Natural Sciences and Engineering Research Council of Canada (NSERC,

PGS-M) and the Northern Scientific Training Program (NSTP, Indian and Northern

Affairs). This research was conducted within the umbrella of Arctic-Net, a Network

of Centers of Excellence (http:/www.arcticnet-ulaval.ca). Support for this research

was provided by NSERC grants (Barber & Papakyriakou) and the Canada Research

Chairs program (Barber).

Page 94: Air-water carbon dioxide exchange in relation to chemical

78

References

Bezte, C.L. 2006. Lower Churchill River water level enhancement weir project post-project monitoring: assessment of water chemistry and phytoplankton responses to operation of the Project - Year VII, 2005, A report prepared for Manitoba Hydro by North/South Consultants Inc.

Billings, W.D., J.O. Luken, DA Mortensen, and K.M. Peterson. 1982. Arctic tundra: A source or sink for atmospheric carbon dioxide in a changing environment? Oecologia 53: 7-11, doi:10.1007/BF00377129.

Billings, W. D., K.M. Peterson, J.O. Luken, and D.A. Mortensen. 1984. Interaction of increasing atmospheric carbon dioxide and soil nitrogen on the carbon balance of tundra microcosms. Oecologia 65: 26-29, doi:10.1007/BF00384458.

Carignan, R. 1998. Automated determination of carbon dioxide, oxygen, and nitrogen partial pressures in surface waters. Limnol. and Oceanogr. 43: 969-975.

Cole, J. J., N.F. Caraco, G.W. Kling, and T.K. Kratz. 1994. Carbon dioxide supersaturation in the surface waters of lakes. Science 265:1568-1570.

Dery, S.J., M. Stieglitz, E.C. McKenna, and E.F. Wood. 2005. Characteristics and trends of river discharge into Hudson, James, and Ungava Bays, 1964-2000. J. of Climate 18: 2540-2557.

Else, B.G.T, T. Papakyriakou, M.A. Granskog, and J.J. Yackel. 2008. Observations of sea surface/CO2 distributions and estimated air-sea CO2 fluxes in the Hudson Bay region [Canada) during the open-water season. J. Geophys. Res. 113: C08026, doi: 10.1029/2007JC004389.

Gorham E. 1991. Northern peatlands: role in the carbon cycle and probable responses to climatic warning. Ecological Applications 1: 182-195.

Granskog, M.A., R.W. Macdonald, C.-J. Mundy, D.G. Barber. 2007. Distribution, characteristics and potential impacts of chromophoric dissolved organic matter (CDOM) in Hudson Strait and Hudson Bay, Canada. Cont. Shelf Res. 27: 2032-2050.

Guo, L. and R. W. Macdonald. 2006. Source and transport of terrigenous organic matter in the upper Yukon River: evidence from isotope (S13C, A14C, and 615N) composition of dissolved, colloidal, and particulate phases. Global Biogeochem. Cycles 20: GB2011, doi:10.1029/2005GB002593.

Page 95: Air-water carbon dioxide exchange in relation to chemical

79

Guo, L., C.-L. Ping, and R. W. Macdonald. 2007. Mobilization pathways of organic carbon from permafrost to arctic rivers in a changing climate. Geophys. Res. Lett. 34: L13603, doi:10.1029/2007GL030689.

Hamilton, J.D., C.A. Kelly, J.W.M Rudd, R.H. Hesslein, and N.T. Roulet. 1994. Flux to the atmosphere of CH4 and CO2 from wetland ponds on the Hudson Bay lowlands (HBLs). J. Geophys. Res. 99: 1495-1510.

Hope, D., S.M. Palmer, M.F. Billet, and J.J.C. Dawson. 2001. Carbon dioxide and methane evasion from a temperate peatland stream. Limnology and Oceanography 46: 847-857.

Jahne, B., G. Heinz, and W. Dietrich. 1987. Measurement of the diffusion coefficients of sparingly soluble gases in water. Journal of Geophysical Research 92:10767-10776, doi:10.1029/JC092iC10pl0767.

JGOFS. 1994. JGOFS Report No. 19-Protocols for the Joint Global Ocean Flux Study fJGOFS) core measurements.

Jones, J.B., JR and P.J. Mulholland. 1998. Carbon dioxide variation in a hardwood forest stream: an integrative measure of whole catchment soil respiration. Ecosystems 1:183-196.

Jones, J.B., JR, E.H. Stanley, and P.J. Mulholland. 2003. Long-term decline in carbon dioxide supersaturation in rivers across the contiguous United States. Geophysical Research Letters 30:1495, doi: 10.1029/2003GL017056.

Kelly, C.A., E. Fee, P.S. Ramlal, J.W.M. Rudd, R.H. Hesslein, C. Anema, and E.U. Schindler. 2001. Natural variability of carbon dioxide and net epilimnetic production in the surface waters of boreal lakes of different sizes. Limnology and Oceanography 46:1054-1064.

Kling, G.W., G.W. Kipphut, and M.C. Miller. 1991. Arctic lakes and streams as gas conduits to the atmosphere: implications for tundra carbon budgets. Science 251: 298-301.

Kling, G.W., G.W. Kipphut, and M.C. Miller. 1992. The flux of C02 and CH4 from lakes and rivers in arctic Alaska. Hydrobiologia 240: 23-36.

Macrae, M. L., R. L. Bello, and L.A. Molot. 2004. Long-term carbon storage and hydrological control of CO2 exchange in tundra ponds in the Hudson Bay Lowland. Hydrological processes 18: 2051-2069.

Page 96: Air-water carbon dioxide exchange in relation to chemical

80

Miller, LA, P.L. Yager, K.A. Erickson, D. Amiel, J. Bade, J. Kirk Cochran, M.-E. Garneau, M. Gosselin, D.J. Hirschberg, B. Klein, B. LeBlanc, and W.L. Miller. 2002. Carbon distributions and fluxes in the North Water, 1998 and 1999. Deep-Sea Research Part II49: 5151-5170.

Rouse, W. R., M.V. Douglas, R.E. Hecky, A.E. Hershey, G.W. Kling, L. Lesack, P. Marsh, M. McDonald, B.J. Nicholson, N.T. Roulet, and J.P. Smol. 1997. Effects of climate change on the freshwaters of arctic and subarctic North America. Hydrological Processes 11:873-902.

Sellers, P., R.H. Hesslein, and C.A. Kelly. 1995. Continuous measurement of CO2 for estimation of air-water fluxes in lakes: an in situ technique. Limnology and Oceanography 40: 575-581.

Simpson, K. G., J.-E. Tremblay, Y. Gratton, and N. M. Price. 2008. An annual study of inorganic and organic nitrogen and phosphorus and silicic acid in the southeastern Beaufort Sea. Journal of Geophysical Research. 113: C07016, doi: 10.1029/2007 JC004462.

Sobek, S., L.J. Tranvik, and J.J. Cole. 2005. Temperature independence of carbon dioxide supersaturation in global lakes. Global Biogeochemical Cycles 19: GB2003, doi:10.1029/2004GB002264.

Stainton, M.P., M.J. Capel, and F.A.J. Armstrong. 1977. The chemical analysis of fresh water. Canadian Fisheries Marine Service, Miscellaneous Special Publication 25,180 pp.

Stedmon, C.A., S. Markager, and H. Kaas. 2000. Optical properties and signatures of chromophoric dissolved organic matter (CDOM) in Danish coastal waters. Estuarine, Coastal, and Shelf Science 52: 67-278.

Tank, S.E., F.W. Lesack, and R.H. Hesslein. In press. Northern delta lakes as summertime CO2 absorbers within the arctic landscape. Ecosystems, DOI: 10.1007/S10021-008-9213-5.

Telmer, K. and J. Veizer. 1999. Carbon fluxes, pC02 and substrate weathering in a large northern river basin, Canada: carbon isotope perspectives. Chemical Geology 159: 61-86.

Tremblay, A., J. Therrien, B. Hamlin, E. Wichmann, and L.J. LeDrew. 2005. GHG emissions from boreal reservoirs and natural aquatic ecosystems, p. 209-232. In A. Tremblay, L. Varfalvy, C. Roehm, and M. Garneau [eds.], Greenhouse gas emissions -fluxes and processes. Springer.

Page 97: Air-water carbon dioxide exchange in relation to chemical

81

Wanninkhof, R. 1992. Relationship between wind speed and gas exchange over the ocean. Journal of Geophysical Research 97: 7373-7382.

Weiss, R.F. 1970. The solubility of nitrogen, oxygen, and argon in water and seawater. Deep Sea Research 17: 721-735.

Weiss, R.F. and B.A. Price. 1980. Nitrous oxide solubility in water and seawater. Marine Chemistry 8: 347-359.

Wetzel, R.W. 2001. Limnology: lake and river ecosystems. Academic Press.

Page 98: Air-water carbon dioxide exchange in relation to chemical

82

Chapter 4: Air-water carbon dioxide exchange and physiochemical processes of the Churchill River estuary

Emmelia C. Stainton, Tim N. Papakyriakou, and David G. Barber

Centre for Earth Observation Science Department of Environment and Geography, University of Manitoba, Winnipeg, Manitoba R3T2N2, Canada.

Abstract

A continuous CO2 monitor was used to measure pCQi in water and air along a

fresh to marine transect in the Churchill River estuary allowing the calculation of

air-water CO2 exchange. A multiparameter probe was used to profile conductivity,

temperature, depth, and pH at stations along the transect, and water was sampled

for chemical analysis of various parameters. Estuarine stratification and seasonal

mixing patterns were determined using the integrated data set. The Churchill River

estuary was stratified by water source, with fresh river water flowing over brackish

marine water. The nature of the stratification varied seasonally with temperature

and changing river discharge and tidal amplitude. Conservative mixing of several

water chemistry parameters was observed. The river was a source of chlorophyll a,

dissolved organic carbon, coloured dissolved organic matter, and particulate organic

carbon and nitrogen to the estuary basin, while the marine water was a source of

salinity, alkalinity, and dissolved inorganic carbon to the estuary basin. Air-water

flux estimates ranged between -11.04 to +22.76 mmol nr2 day-1, with mean fluxes

being -1.07 ± 1.90 mmol nr2 day1 at low tide, 0.06 ± 5.93 mmol nr2 day1 at high

Page 99: Air-water carbon dioxide exchange in relation to chemical

83

tide, and -0.37 ± 4.79 mmol nr2 day-1 overall. Both between and within month

variation in CO2 flux exists. These results indicate that the Churchill River estuary is

a net CO2 sink throughout much of the 2007 open-water season, which is striking

given that high latitude estuaries are generally considered sources of atmospheric

CO2. These results are consistent with other data collected from the Churchill River

[upstream of the estuary) and with flux estimates made for Hudson Bay.

4.1 - Introduction

Growing evidence linking increased levels of atmospheric anthropogenic

carbon dioxide (CO2) to climate change has heightened interest in regional

budgeting and modeling of atmospheric CO2 concentrations to identify current

trends and understand carbon pathways among various interconnected carbon

stocks at or near to the Earth's surface. Model and budget accuracy are still quite

limited by major gaps in understanding of important components of the global

carbon cycle, in particular within the productive coastal marine environments of the

high latitudes in both hemispheres (Borges, 2005). Coastal shelves and estuaries,

though only making up 7% of oceanic surface area, are not included in oceanic

carbon budgets (Borges, 2005). The role of these regions in the global CO2 cycle is

largely unknown, with uncertainties of up to 75% (Takahashi et al., 2002), and it is

not known whether the inclusion of polar shelves and estuaries will increase the

efficiency of high latitude oceans as a carbon sink, or alter the region's role from a

strong sink to source of atmospheric CO2 (Borges, 2005). An understanding of the

Page 100: Air-water carbon dioxide exchange in relation to chemical

84

underlying dynamics of carbon cycling in these environments is particularly

important as there is a general lack of good observational data on the time/space

characteristics of CO2 cycling (Borges, 2005). The estuarine environment in the

northern high latitudes is generally unknown, and their complexity is hard to study

because it involves processes at a spatially and temporally dynamic boundary.

The surface waters of Hudson Bay are slightly fresher relative to other

peripheral seas (Prinsenberg 1986) resulting from a large freshwater input from

several large rivers (713.6 km3 year1, Dery et al. 2005) that drain the Northern

Great Plains and the boreal belt of eastern Canada. In 1976 Churchill River flow was

reduced by approximately two-thirds when water was diverted to the Nelson River

to enhance hydroelectric power generation (Bezte 2006). Although the Churchill

River comprises only 2.8% of the freshwater input to Hudson Bay (20.57 km3 year1,

Dery et al. 2005), it's watershed is similar in characteristic to other larger rivers

entering Hudson Bay, like the Nelson and La Grande Rivers of northern Manitoba

and Quebec, Canada. The lower Churchill River passes through the Hudson Bay

Lowlands, a vast area with organic carbon-rich peatlands (Gorham 1991) dotted

with thousands of small ponds, bogs, and fens.

To date, research conducted within the Churchill River and estuary system

has focused largely on water quality/chemistry in relation hydroelectric

development (Technical Reports by North-South consultants prepared for Manitoba

Hydro), however a recent study by Kuzyk et al. (2008) investigated the

Page 101: Air-water carbon dioxide exchange in relation to chemical

85

physical/chemical/biological coupling within the Churchill estuary system through

the spring/melt period.

The objective of this research is to describe 1) the current physical and

chemical conditions of the Churchill River estuary, 2] the air-water exchange of CO2,

and how physiochemical conditions influence the CO2 exchange, and 3) how these

conditions and processes vary with sampling location, season, depth, and tidal

amplitude. We relate CO2 flux to water column properties and processes along the

estuarine salinity gradient in order to determine the extent to which this area is a

source or sink for CO2, and the conditions (chemical, biological, and physical) that

govern CO2 flux.

An understanding of the CO2 flux between estuarine waters and the

atmosphere in polar regions is important for assessing potential climatic feedbacks,

especially with the increasing influence of climate change in these areas. This work

also ties in with other CO2 budgeting studies in the Churchill area, and will lead to a

better understanding of carbon transformations at the land/freshwater/marine/air

interface. The accessibility of the Churchill region compared to other high latitude

estuaries makes it a useful study area for modeling the impact of climate change in

an arctic coastal area over a range of potential changes. This work at this locale

suggests that there may be significant value in developing a long-term monitoring

program in the Churchill estuary that would serve to track the influence of climate

change on northern estuaries over time.

Page 102: Air-water carbon dioxide exchange in relation to chemical

86

4.2 - Methods

4.2.1 - Study area

The Churchill River enters into the coastal waters of southwestern Hudson

Bay (Lat: 58.791°, Long: -94.2082°; Figure 4.1] draining an area of 281,300 km2

across the Canadian Shield, the Hudson Bay Lowlands, and the northern interior

prairies where the river's headwaters are situated. During the summer months, the

Churchill River flows over a rock weir and into an enclosed estuary basin (~13 km

long by 3 km wide) before entering Hudson Bay (Figure 4.1). The aerial extent of

estuarine waters varies significantly with tidal cycle (Figure 4.1), with exposed

rocks and mudflats visible at low tide. At high tide seawater flows into the estuary

basin raising the water level by 2-4 meters. This causes the inflowing river water to

pile up downstream of the weir and flow over the dense salt water, forming a "salt

wedge" (Kuzyk et al., 2008). From late November to May the Churchill River and

Hudson Bay are ice-covered (Kuzyk et al., 2008).

All sampling associated with this work took place between June-October

2007, and was conducted in two ways: 1) estuary survey sampling (CTD casts,

discrete water sampling, and discrete dissolved/atmospheric CO2 sampling), and 2)

estuary basin quick CTD casting. Information on the inflowing Churchill River water

was also collected throughout the 2007 open-water season (Stainton et al., in prep.

2009).

Page 103: Air-water carbon dioxide exchange in relation to chemical

87

Figure 4.1. Churchill River (CH7) and estuary (CH1-5) sampling stations at a] high tide and b) low tide (Landsat 7 satellite image, Natural Resources Canada).

4.2.2 - Estuary survey sampling

Sample locations were situated within the estuary basin and extend 13 km

along a transect from fresh to marine water (Figure 4.1). Discrete water samples

were collected from each of 5 stations at high tide (CH5 to CHI), and each of four

stations at low tide (CH4 to CHI) once each month throughout the sampling period.

Water samples

A Van Dorn style water sampler was used to collect surface water (0.5m from

top) and bottom water samples (0.5m from bottom, stations CH4-CH1 only) once

per month, at both high and low tide. Sample for dissolved inorganic carbon (DIC)

analysis was collected from the water sampler first by piercing the outlet tubing and

Page 104: Air-water carbon dioxide exchange in relation to chemical

88

drawing 4mL of water into a syringe triple-rinsed with sample. Water was expelled

through a septum cap into 12 mL borosilicate glass vials containing 20 iL of

concentrated phosphoric acid following Stainton et al. (in prep. 2009).

Sample for other variables was transferred into acid-cleaned (10% HCL

solution) and Milli-Q and sample water rinsed 2L amber polyethylene bottles, filled

to the shoulders, tightly capped, and kept cool until processed. Laboratory

processing took place in Churchill immediately upon return from the field. Sub-

samples were collected for alkalinity, chlorophyll a, total suspended solids (TSS),

dissolved organic carbon (DOC), chromophoric dissolved organic matter (CDOM),

and particulate organic carbon and nitrogen (POC/N) analyses.

Air-water CO2 sampling

Concurrent with water sample collection, dissolved CO2 concentrations were

measured in the surface waters of each station (CH5 to CHI) using a continuous

flow automated air-water CO2 monitoring instrument (Figure 4.2). The system is

described in detail by Stainton et al. (in prep. 2009), and is only briefly reviewed

here. The gas monitor (GM) consists of two main parts: 1) a "dry" box containing an

infrared CO2 gas analyzer (LiCor® LI 820), a programmable data logger controller

(Campbell Scientific® CR10X), a gas pump (Spectrex Corp.® AS-200), and in-line gas

drying Nafion tubing (Perma Pure® LLC ME-060-24BB) surrounded by silica

desiccant 2) a "wet" box containing a gas exchanger (Celgard Inc.®), a peristaltic

pump (Autoclude® M500), and a thermistor (Omega Engineering®) (Figure 4.2).

Page 105: Air-water carbon dioxide exchange in relation to chemical

89

Figure 4.2. Air/water gas monitoring instrument. Dry box: 1-air pump, 2-drying compartment, 3-O2 sensor, 4-gas analyzer, and 5-data logger; wet box: 6-peristaltic water pump, 7-thermocouple housing, 8- air/water gas exchanger, and 9- solenoid valves.

Atmospheric CO2 was measured by drawing air into the system with the gas

pump, drying it with the in-line Nafion tubing, and passing it through the CO2

analyzer before leaving the unit. Dissolved gas in water was measured by drawing

water into the wet box via the peristaltic pump, through the in-line thermistor, and

then through the gas permeable tubing in the gas exchanger before leaving the

system. An air pump ran concurrently with the peristaltic pump in a closed loop that

was connected to the outer chamber of the gas exchanger. The water intake tube

was covered with 100-micron copper mesh and positioned at 0.25 m below the

Page 106: Air-water carbon dioxide exchange in relation to chemical

90

water surface, while the air intake was affixed at approximately 0.5 m above the

water surface. The GM was run for 30 minutes at each station resulting in one air

and one water pCOi measurement.

Following surface air/water pCCh measurement, water was also sampled

from mid-column and bottom waters (0.5m from bottom) at stations CH3 and CH4

using the GM to determine whether variation existed with depth. Water was

pumped through the gas exchanger for 10 minutes at each depth (with data

recorded every 10 seconds), resulting in one average water pCOi measurement.

The CO2 analyzer was calibrated bi-weekly with a certified gas standard

(span CO2 = 380 parts per million (ppm)) and zeroed with ultra high purity

nitrogen. Drift was less than 2 ppm between calibrations.

CTD profiling

At each sampling station, a conductivity, temperature, depth, and pH probe

(model RBR XR-620) was used for water-column profiling. Data were collected and

recorded 6 times per second.

4.2.3 - Estuary basin CTD profiling

Water column profiling (using RBR XR-620) was also done by visiting

stations within the estuary basin during a two-hour window around high or low

tide. Here, more stations were sampled along a transect between stations CH5 and

Page 107: Air-water carbon dioxide exchange in relation to chemical

91

CH2 each month at both high and low tide in order to get a "snapshot" of the

estuary's stratification and mixing patterns.

4.2.4 - Other sources of data

Discharge of the inflowing Churchill River was estimated by summing flows

from the Churchill River at Red Head Rapids (Lat: 58.1208°, Long: -94.625°) and the

inflowing Deer River (Lat: 58.015°, Long: -94.1956°) using daily discharge data

obtained from the Water Survey of Canada

(http://www.wsc.ec.gc.ca/hydat/H20/index_e.cfm).

Meteorological data (precipitation and air temperature) were collected at the

Churchill weather station by Environment Canada

(www.weatheroffice.gc.ca/city/pages/mb-42_metric_e.html) and hourly data were

downloaded and used here. River water temperature data are routinely monitored

at the Churchill River pump-house station by Manitoba Hydro (~6km upstream of

the weir), and the data were made available to this study.

Water chemistry and air/water CO2 concentrations were collected from

Churchill River station CH7 throughout much of the 2007 open-water season. Water

chemistry samples were collected and processed as outlined above. Air/water CO2

concentrations were measured on a second GM (identical to that used for estuary

surveying) that collected continuous data (sampling every 2 hours). Details on this

system appear in Stainton et al. (in prep. 2009).

Page 108: Air-water carbon dioxide exchange in relation to chemical

92

4.2.5 - Laboratory and analytical procedures

DIC samples were analyzed on a Varian CP4900 MicroGC. Gas was introduced

into the GC by piercing the septum cap of the vials with a needle and allowing gas to

flow [vials pressurized in the field at time of sampling) through Nafion drying tubing

to the GC sample loop following Stainton et al. (in prep. 2009). The sample line was

flushed with UHP helium between samples. Alkalinity samples were analyzed on a

PC TitratlON Plus following the Gran alkalinity titration method (Stainton et al.

1977).

Using a vacuum filtration system, particulate matter from a known volume of

sample (ranging from 250 to 500 mL) was collected onto 42.5mm GF/F filters for

the analysis of chlorophyll a and TSS (pre- combusted and weighed filters). These

were kept frozen and dark (to reduce chlorophyll a degradation) until analysis.

Chlorophyll a concentrations were measured on a Turner Designs fluorometer (10-

AU) following 24-hour extraction in 90% acetone at 4°C (JGOFS 1994). TSS was

determined after Stainton et al. (1977).

Samples for DOC and CDOM were filtered through a 25mm glass fiber (GF/F)

filter contained in a polypropylene filter holder previously rinsed with Milli-Q and

sample, and collected into 1) a pre-combusted 5mL glass vial containing 25 uL of

50% HPLC grade phosphoric acid (DOC), and 2) a pre-combusted 50 mL amber

borosilicate glass vial (CDOM). Samples were stored at 4°C until analysis. A total

sample volume of 120mL was filtered through this GF/F filter, which was wrapped

in pre-combusted aluminum foil and frozen for analysis of POC/N.

Page 109: Air-water carbon dioxide exchange in relation to chemical

93

DOC samples were analyzed on a Shimadzu TOC-5000A. CDOM absorbance

was measured using a Hewlett-Packard 8452A diode array spectrophotometer in an

acid-cleaned, Milli-Q water rinsed, 10-cm quartz cell and scanned at 2 nm intervals

between 250 and 820 nm against Milli-Q water, and absorption [X) was calculated

after Stedmon et al. (2000). POC/N was measured on a Carlo Erba NA-1500

Elemental Analyzer following acid fuming to remove carbonate (following Verardo

et al., 1990). Nutrient samples (NO2/NO3", PO43) were analyzed following Simpson

et al. (2008) and concentrations were measured on a continuous flow Bran +

Luebbe Auto-Analyzer 3 (AA3).

4.2.6 - Calculations

Dissolved CO2 data measured by the GM was converted to concentration

(umol Lr1) from molar fraction using:

[C02] = a • PC02 w

where a is the solubility coefficient calculated for CO2 (mol L1 atnr1) (Weiss and

Price 1980).

CO2 flux was calculated after Wanninkhof (1992):

Flux = a x kCO x A/CO, 2 2, (2)

where

*CO2 = 0.31U12

0m*[_ScV''2

Page 110: Air-water carbon dioxide exchange in relation to chemical

94

Uiom is the wind speed at 10m from the water surface, Sc is the Schmidt number (a

dimensionless coefficient) (Jaime et al. 1987), and ApCCh is the water-air difference

inpC02.

4.3 - Results

4.3.1 - Hydrographic conditions

Both the amount and frequency of precipitation peaked in late August and

throughout September, while air temperature peaked in mid July (Figure 4.3a).

Water temperature also peaked in mid July (Figure 4.3b) and variation is in line

with the air temperature record throughout the entire open-water season. River

discharge was initially high following ice breakup, decreased to a minimum in late

July, and peaked once again in early August coinciding with several substantial

rainfall events (Figure 4.3b).

Tidal amplitude was greatest around September 1 and at the beginning/end

of October, and lowest around the 20th day of each month. Much of our sampling

took place when tidal amplitude was high, except for in August where amplitude

decreased from 4 meters to 2 meters during the sampling trip (Figure 4.3c).

4.3.2 - Physical conditions

From the estuary basin water column profiles, we produced depth-distance

plots of salinity and temperature conditions within the Churchill River estuary at

Page 111: Air-water carbon dioxide exchange in relation to chemical

95

both high and low tide, from June to October (Figure 4.4). Inflowing Churchill River

water was warmer than Hudson Bay marine waters during all sampling trips except

October. This caused thermohaline stratification within the estuary basin with

warm, river water flowing into and over the cooler, denser marine waters. In the

fall, the slightly cooler inflowing river water led to inverse temperature

stratification within the estuary basin.

In June, high river discharge volume and melting ice floes from the break-up

of Hudson Bay resulted in both the river plume extending further towards the

estuary basin mouth, and the presence of a thicker fresh surface-water layer at both

high and low tide compared to the other sampling months. The effect of river

discharge volume on estuary basin thermohaline stratification can also be seen in

the high tide panels of July and August. Reduced discharge volume in July led to a

sharp thermocline (and halocline) when marine waters intrude up the estuary

basin. In August, high river discharge (Figure 4.3b) caused river water to flow over

the marine water at high tide, resulting in more gradual stratification and river

water extending further north in the estuary basin compared to that measured in

July. At low tide (June, July, August) river water reached the estuary mouth, creating

a brackish mixture with the underlying marine water in the estuary basin. Cold

temperature and windy conditions in October are a possible explanation for the

pocket of cold water sampled in the estuary at low tide. Strong south winds during

sampling may have enhanced mixing in this uniformly shallow (<5m) area of the

estuary basin.

Page 112: Air-water carbon dioxide exchange in relation to chemical

Figure 4.3. a) Churchill daily mean air temperature and precipitation (Environment Canada), b) Churchill River discharge (combination of Red Head Rapids and the Deer River, Water Survey of Canada) and water temperature (Manitoba Hydro pumphouse station), and c) Churchill harbour tide height (Fisheries and Oceans Canada), with sampling periods indicated by dashed vertical lines.

Page 113: Air-water carbon dioxide exchange in relation to chemical

97

Temperature-salinity data (collected during estuary survey sampling and

basin CTD casting) is shown in Figure 4.5. All June casts at either tidal amplitude

were stratified in both temperature and salinity. This was also true for the July CTD

casts, though river and marine waters were warmer. The slight separation in the

July point cluster between 20-30 psu and ~7-12 °C was a result of separate CTD

casting days, where the cooler data were collected on July 12, and the warmer

portion was collected between July 15-19. In August there was a clear separation

between the low tide and high tide CTD casts, and for this reason the data was

divided into two groups (all low tide casts in dark green, and all high tide casts in

light green) (Figure 4.5). Thermohaline stratification persisted in August, however

at low tide the basin was largely isothermal with observed vertical variation only in

salinity. In the fall (September and early October) the estuary was typically

isothermal, as both marine and river water undergoes seasonal cooling. Haline

stratification was present in the fall, both at high and low tide.

4.3.3 - Water chemistry

Results from the water chemistry analyses at estuary stations CHI to CH5

and river station CH7 from June to October are summarized in Table 4.1. Mean

dissolved oxygen, chlorophyll a, PON, POC, DOC, CDOM, Si, and PO4 concentrations

and pH were generally highest for river station CH7 and estuary station CH5, and

showed a decreasing trend from the river/upstream estuary stations towards

Hudson Bay (Table 4.1). Conversely, mean alkalinity concentration and salinity were

Page 114: Air-water carbon dioxide exchange in relation to chemical

98

greatest for estuary station CHI, and values decreased moving upstream towards

river station CH7. DIC and TSS, and NO3+NO2 however, did not follow a decreasing

pattern from the river to marine station. TSS and NO3+NO2 were greatest for

stations CH4 and CHI, though variability in TSS was also greatest at these stations.

Mean DIC concentration is highest at estuary stations CH2 and CH4 (Table 4.1).

Stations CH2 and CH4 also appeared (based on visual observation in the field) to

experience significant turbulence due to fresh-marine water confluence (CH4), and

tidal movement (CH2). When sampling at high tide, the surface waters near CH4

would show a delineation between fresh and marine water in the form of ripples

(calm days) and wave fronts (windy days). Station CH2 was located in the mouth of

the estuary basin where tidal currents were quite strong, especially during the

receding tide.

Within station variability also existed and is evident in Table 4.1. Comparing

surface and bottom water means for each station, chlorophyll a, DOC, CDOM, and to

some extent PON, POC, and Si were all greater for surface waters. Conversely, DIC,

TSS (except station CH2), salinity and alkalinity were all larger at the bottom

compared to surface waters. Consistent vertical variation in DO, NO3+NO2 and PO4

was not observed. NO3+NO2 concentration was observed to be higher in surface

waters for stations CHI and CH2, though higher in bottom waters for stations CH3

and CH4, while PO4 concentration is greater for CHI and CH3 bottom waters and for

CH2 and CH4 surface waters (Table 4.1).

Page 115: Air-water carbon dioxide exchange in relation to chemical

99

Figure 4.4. Depth, temperature, salinity cross-sections of Churchill River estuary CTD transects at high tide [left column) and low tide [right column) in June, July, August, and October 2007. Temperature is represented in colour (note all colour legends on the right side of figures are different), salinity is denoted by numbered contour intervals (ranging from 2-30), and individual CTD casts are shown as bold black vertical lines within the cross-sections. White circles on the high/low tide maps (bottom panels) are the CTD sampling locations. Temperature-salinity plots were created using Ocean Data View (Schlitzer, 1998).

High Tide

12.5

10

7.5

5

, 5

0 June

- rn^^v •-

' • « . - .

. _

, ^ _

^^^^^^^^^^^^^^i j j j

'

X 9 1

^ 7

56

18

16

14

12

10

8

L ^ * j | ^ ^ _ | ' .

\ «• July

W\ 17 ?

•I 13

12

August

^^^^MPT^^ 1 -7

VSs 7

| -----f V>

iff

n

13.5

13

II" f ^ 10.5

Page 116: Air-water carbon dioxide exchange in relation to chemical

100

Table 4.1: Water chemistry of Churchill River estuary waters collected from stations along a fresh (CH5) to marine (CHI) gradient, at both high tide (HT) and low tide (LT), from surface (0.5m) and bottom waters (3 and 10m), from June to October 2007. Surface water samples were also collected from the inflowing Churchill River (CH7) upstream of the weir (Stainton et al., in prep. 2009).

Station

CH3 CH3 CH4 CH4 CH4

g CH4 5 CH5

CH5 CH7 CH7

CHI CHI CHI CH2 CH2 CH3 CH3

>. CH3 •3 CH3

CH4 CH4 CH5 CH7 CH7

CHI CHI CHI CHI CH2 CH2 CH2 CH2 CH3

1 CH3 | CH3

CH3 CH4 CH4 CH4 CH5 CH7 CH7

CH2 CH3 CH3 CH3

_ CH3

1 CH4 3 CH4

CH4 CH5 CH7

Depth Tide height m 0.5 10 0.5 0.5 3 3

0.5 0.5 0.5 0.5

0.5 0.5 10 0.5 10 0.5 0.5 10 10 0.5 0.5 0.5 0.5 0.5

0.5 0.5 10 10 0.5 0.5 10 10 0.5 0.5 10 10 0.5 0.5 3

0.5 0.5 0.5

0.5 0.5 0.5 10 10 0.5 0.5 3

0.5 0.5

LT LT HT HT HT HT HT HT

--

Mean SD. HT HT HT HT HT LT LT LT LT HT LT HT

--

Mean S.D. LT LT LT LT HT LT HT LT HT LT HT LT HT LT HT HT

--

Mean S.D. LT LT HT LT HT LT HT HT HT

-Mean S.D.

DO umol L'1

488 631 474 455 504 497 492 435 488 521 499 53 350 306 394 389 400 500 499 531 335

-330 480 439 399 412 73 326 415 357 412 374 420 365 417 436 399 417 324 429 441 437 424 416 430 402 37 357 361 385 346 373 379 428 369 436 395 383 30

DIC Umol L"'

1114 1012 883 1087 1292 1595 1098 943 1031 1244 1130 205 385 360 948 890 1213 596 552 887 562

-136 1354 1162 1290 795 393 1227 1292 1795 1904 1408 1278 1780 1914 1466 1268 2001 1335 1359 1355 2011 1381 1325 1325 1524 285 1140 899 1460 1073 1743 1241 1148 1521 1106 1017 1235 260

TSS mgL-1

6.5 13.0 4.9 6.8 4.6 10.2 7.4 5.5 7.7 4.4 7.1 2.7 3.0

0.6 7.2 0.4 2.8 1.2 4.4 3.4 1.3 4.0 2.5 2.3 1.0 2.6 1.9 4.6 4.4 6.2 18.6 2.6 5.8 2.4 3.8 4.2 3.8 2.7 6.6 5.6 16.3 2.8 7.2 10.4 8.2 6.5 4.5 5.7 4.0 2.7 4.2 8.0 5.6 4.4 45.2 5.2 2.11 8.7 12.9

Chla ftpL-1

1.03 0.80 1.54 1.59 0.42 0.44 1.84 1.96 1.33 1.97 1.29 0.59 0.53 0.06 0.38 0.25 0.28 1.10 1.42 0.40 0.52 2.29 2.46 2.66 2.71 2.67 1.27 1.06 1.66 2.77 1.17 0.58 1.85 2.86 0.26 0.55 2.13 2.51 0.20 1.60 3.00 4.68 0.55 3.93 6.27 3.43 2.22 1.65 1.10 0.94 0.70 1.03 0.67 1.05 0.94 0.70 1.33 1.41 0.99 0.25

PON ugN mL"1

-0.085

-0.034

-0.047

--

0.097 0.069 0.066 0.026

-0.045 0.031 0.082

-0.050 0.050 0.032 0.049

---

0.049 0.070 0.051 0.016

----------

0.052 0.069 0.023 0.076 0.065 0.122 0.036 0.088 0.081 0.075 0.069 0.028

-0.044 0.052 0.051 0.056 0.048 0.034 0.037 0.046 0.045 0.046 0.007

POC HgC mL''

-0.621

-0.551

-0.267

---

0.698 0.578 0.543 0.164

-0.128 0.120 0.160 0.491 0.297 0.348 0.221 0.281

---

0.355 0.499 0.290 0.137

--------

0.391 0.531 0.100 0.426 0.506 0.991 0.220 0.608 0.538 0.623 0.493 0.242

-0.230 0.353 0.192 0.415 0.315 0.280 0.218 0.292 0.329 0.291 0.071

DOC HM 691 579

-725

--

737

-732 723 698 61

-197 171 222 215 454 538 174 279

--

708 695 699 396 228

---------

846 958 114 523 903 825 248 921 1005 927 727 318

-1148 583 1010 206 881 1175 416 1194 1214 870 378

Salinity PSU

--10.43 22.14

---

0.08

---

10.88 11.04

-25.58 27.69 24.18 25.34 14.41 9.80 27.82 22.01

-9.17 0.08

--

18.61 9.61 12.44 1.97

28.73 27.34 8.58 1.37

30.75 27.70 4.24 1.31

30.51 15.72 4.10 1.68 1.52 0.08

--

12.38 12.35 5.56 7.91 18.77 5.48

27.72 8.97 3.11 22.77

--

12.54 9.22

CDOM abs 350

12.6 10.3 11.7 11.4 6.8 3.6 12.1 11.1 12.2 10.8 10.3 2.9 3.7 2.1 1.5 2.5 2.3

-6.9 1.7 3.6 8.2 7.5 8.8 8.7 8.6 5.1 3.0 10.7 16.2 1.4 2.5 14.5 16.9 0.6 1.9 16.7 18.9 0.7 8.8 17.7 15.4 3.6 17.6 20.5 18.3 11.3 7.4 25.2 34.8 15.0 27.4 3.6 22.8 32.4 10.5 30.6 30.2 23.3 10.3

Alk ^ q L"1

1445 1631 1320 1357 1858 1997 1330 1342 1306 1361 1495 249 1740 1891 2043 1938 1924 1741 1554 2061 1705 1513 1940 1435 1423 1531 1746 224 1664 1492 2114 2082 1634 1460 2227 2083 1504 1525 2222 1850 1413 1352 2099 1376 1448 1375 1718 329 1250 1517 1815 1577 2118 1550 1369 1958 1320 1312 1578 295

SiuM HM 3.88 1.34

-9.10 13.92 6.62

-6.09 16.58 20.89 9.80 6.73 2.53 2.24 1.74 3.60 2.95 3.76 2.11 2.08 5.27 2.08 0.21 0.33

--

2.41 1.40 5.89 4.98 1.93 5.10 6.94 7.02 2.44 4.22 9.94 10.66 1.90 6.71 6.04 7.53 5.19 7.40

--

5.87 2.52 11.90 8.08 6.16 7.57 4.32 10.51 5.30 6.40 21.84 20.65 10.27 6.22

N03

mgL-' 0.008 0.040 0.019 0.009 0.109 0.031 0.047 0.020 0.048 0.114 0.045 0.038 0.106 0.049 0.202 0.035 0.046 0.052 0.041 0.072 0.085 0.074 0.039 0.060 0.033 0.232 0.080 0.062 0.447 0.255 0.076 0.065 0.167 0.011 0.038 0.037 0.108 0.058 0.190 0.457 0.017 0.162 0.069 0.321 0.061 0.062 0.144 0.140 0.465 0.056 0.182 0.079 0.250 0.224 0.719 0.446 0.268 0.089 0.278 0.210

P04 pH mgL1

0.020 8.04 0.002 7.89

- 8.51 0.010 8.43 0.051 7.81 0.047 7.85 0.076 --0.109 8.13 0.177 8.17 0.018 8.19 0.057 8.11 0.057 0.25 0.024 8.20

-- 8.29 0.044 8.23 0.044 8.30 0.039 8.27 0.002 8.21

- 8.49 0.035 8.13 0.043 8.24 0.011 8.86

-- 8.51 0.105 8.69

-- 7.50 0.036 8.73 0.038 8.33 0.028 0.33 0.017 8.60 0.016 8.71 0.034 8.44 0.059 7.96 0.012 8.58 0.013 8.83 0.053 8.44 0.054 8.42 0.025 8.76 0.020 8.74 0.051 8.42 0.016 8.35 0.020 8.86 0.032 8.78 0.045 8.37 0.063 8.59

- 8.25

.. 0.033 8.54 0.018 0.24 0.149 8.64 0.147 8.63 0.031 8.61 0.215 8.58 0.116 8.56 0.220 8.64 0.070 8.88 0.028 8.52 0.003 -0.058 8.14 0.104 8.58 0.078 0.20

Page 117: Air-water carbon dioxide exchange in relation to chemical

101

Figure 4.5. Temperature-salinity scatterplot of all CTD casts taken within the Churchill River estuary throughout the 2007 open-water season. Data is grouped by sampling month: red = June, blue = July, green = August (light green for high tide casts, dark green for low tide casts), black = September/October.

20 • June • July

August (HT) • August (LT) • Sept/Oct

0 + V-v

X 10 20

Salinity (psu) 30

When partitioned by sampling month, DO decreased from June to October,

while NO3+NO2 concentration increased from June to a maximum in the fall (Table

4.1). All of chlorophyll a, PON, and POC concentrations were greatest in June and

August, while TSS, DOC, and CDOM were all highest in October, and lowest in July.

DIC concentration was greatest in August, while alkalinity was higher in the summer

Page 118: Air-water carbon dioxide exchange in relation to chemical

102

months relative to the spring and fall. Both PO4 and Si concentrations were highest

in October, while pH was slightly higher in July (Table 4.1).

Figure 4.6 shows the relationship between the water variables (DOC, CDOM,

POC, PON, Alk, DIC, chlorophyll a, and pH) and salinity. Here, point location is

determined by the left y-axis and x-axis variables, while the point colour reflects the

right y-axis variable. DOC and CDOM concentrations were both greatest for the

lower salinity samples. This trend is also evident for POC and PON, however there is

more variability in concentration for samples with salinity ranging 0 to 10 psu.

Chlorophyll a and pH follow a similar trend, decreasing with increasing salinity,

though there are a few data points with a pH of less than 8.2 with salinities between

0-10 psu. In general, alkalinity and DIC concentrations increase with increasing

salinity (Figure 4.6).

Some water chemistry variables were also sampled for the Churchill River

estuary region between March and May 2005 by Kuzyk et al. (2008), DOC, NO3+NO2,

and chlorophyll a, which are comparable to values presented here.

Page 119: Air-water carbon dioxide exchange in relation to chemical

Figure 4.6. Water chemistry of the Churchill River estuary between June and October, 2007. Samples were taken at 0.5m from the surface, and 0.5m from the bottom at each of 5 stations throughout the length of the estuary [ranging from fresh to brackish water), at both high and low tide. The x-axis is salinity in all four panels and the right y-axis colour legend is reflected on the data points in each separate panel. Point distribution reflects the left y-axis parameters versus salinity. Plots were created using Ocean Data View (Schlitzer, 2008).

1200

1000

^ 800

o O 600 Q

400

200

2200

^.2000 u

•f 1800

I < 1600

1400

1200

o

o

° o o °

o

o

0 o

G

0

. P-

35

bo

25

o

©

o

" .' O ^

° r.o c o"

. o o

o

o

S3 9,

°o. ®

-C)

0

10

5

0

12000

1.0

0.8

o » r > , — i

20 | ^0 .6

1 5 o y

§ 20-4

0.2

-7 ao 1500^ ^

It 'I n g, O

IOOO^ 8

U

50

o o o

Sb

Q D

O o

O O

o ° 0 O o o

0.12

0.1

0.08 j ?

o 0.06 **

0.04

0.02

8.8

8.6

8.4 ^

8.2

10 20 30 Salinity [psu]

10 20 30 Salinity [psu]

II

Page 120: Air-water carbon dioxide exchange in relation to chemical

104

4.3.4 - Air/water pC02 and C02 flux

Air/water pC02

Estuary surface water was in general undersaturated in CO2 with respect to

the atmosphere, with a mean difference in partial pressure (ApCCh) of-4.30 ± 29.33

uatm and a ApCC>2 range of-64.26 to +86.65 uatm. At high tide (i.e. marine water)

ApC02 averaged -0.86 ± 29.55 uatm, while at low tide (i.e. river water) the average

was -9.86 ± 29.26 uatm. Estuarine surface waters were generally undersaturated

with respect to the atmosphere at low tide, with the exception of station CH3 in June

and CH4 and CHI in July. At high tide, ApC02 was much more variable, though the

following trends exist: 1) surface waters are undersaturated in June, 2) marine

dominated stations CHI and CH2 are undersaturated, while river-dominated

stations CH5, CH4, and CH3 are supersaturated in July, 3) there is no obvious trend

in the August samples, however waters are supersaturated at station CH5 (86.65

uatm) and undersaturated at station CHI, 4) in October, ApCC»2 decreases from

station CH5 to CH3 (Figure 4.7).

C02 flux

The air-water CO2 flux ranged between -11.04 to +22.76 mmol nr2 day1, with

mean values of -1.07 ± 1.90 mmol nr2 day1 at low tide and +0.06 ± 5.93 mmol nr2

day_1 at high tide. At low tide, CO2 flux estimates are predominantly negative with

the exception of station CH4 and CHI in July, and station CH3 in June (Figure 4.7). At

high tide, flux magnitude is greatest at stations CH5 and CH4, with waters being a

Page 121: Air-water carbon dioxide exchange in relation to chemical

105

sink for CO2 in June and a source in July and October [Figure 4.7). Overall, June flux

estimates were predominantly negative while July fluxes were mostly positive, with

the exception of station CH3 at low tide (June) and stations CH2 and CHI at high tide

(July) (Figure 4.7). Flux estimates for August were all negative at low tide, however

much variability exists among the high tide samples, where a positive flux of 22.76

mmol nr2 day _1 was observed at station CH5. Total mean CO2 flux for the estuarine

surface waters was -0.37 ± 4.79 mmol nr2 day-1 and indicate that this system is a net

sink for CO2 throughout the 2007 open-water season.

Figure 4.7. Water-air difference in pC02 (top panels) and air/water CO2 flux from Churchill River estuary waters. Samples were collected using a continuous automated CO2 monitor that averaged air and water CO2 concentrations over a 30-minute period. Stations range from fresh (CH5) to brackish/marine (CHI) waters and were sampled at high and low tide, June to October 2007. Overall mean CO2 flux and ApC02 was -0.62 mmol nr2 day1 and -4.30 uatm respectively.

High Tide Low Tide

-a

(N o u

4 3 2 1 0

- 1 - 2 - 3 - 4 --5

+22.76 BJune °July a August • October

1 - 1 1 . 0 4

J 1 n 1 wm

I

CH5 CH4 CH3 CH2 CHI CH5 CH4 CH3 CH2 CHI

Page 122: Air-water carbon dioxide exchange in relation to chemical

106

4.3.5 - pC02 concentration at depth

At high tide, stations CH3 and CH4 were stratified with warmer, fresher

water overlying cooler, marine water (Figure 4.8]. Higher pCCh was generally

associated with the marine bottom water, which is consistent with information

presented in Figure 4.7, and indicates that fresher surface water was typically more

undersaturated in dissolved CO2 compared to marine water. In July however, little

variation in pCCb with depth was observed at CH3.

At low tide, station CH4 was well mixed with respect to temperature and

salinity, resulting in little variation in pCCh with depth (Figure 4.9). Station CH3 was

stratified, although the pattern differed depending on date. In June, water column

pCCh at CH3 was greater than for the other station (Figure 4.9), which is different

than what we measured at high tide where water pCCh was lowest in June (CH4)

(Figure 4.8). Little variation existed in pCOz with depth on July 14, indicative of a

well-mixed water column (Figure 4.9). CTD and pCCh profiles for station CH3 on

August 26 varied the most in pCOi and salinity, though temperature decreased by

less than 2 °C with depth.

Page 123: Air-water carbon dioxide exchange in relation to chemical

107

Figure 4.8. pCCh concentration, salinity, and temperature variation with depth (surface, mid-depth, and bottom waters) at Churchill River estuary stations CH3 (top panels) and CH4 (bottom panels). Samples were collected between June and October at high tide using a continuous automated CO2 monitor. Water was pumped from each depth for 10 minutes, data logging every 10 seconds. Data points represent an average of pCOi over the 10-minute sampling period. Temperature-salinity plots were created using Ocean Data View (Schlitzer, 2008).

0

2

?4

a6

8f

10 0

- B - 17-Jul •-*•- 21-Aug -a— 1-Oct

CH3

2 +

o. Q

4 +

CH4

o 16-Jun -•^-21-Aug —e— 1-Oct

6 300 320 340 360 380 400 420 440 0

pC02 (|iatm) 10 20 30

Salinity (psu)

Page 124: Air-water carbon dioxide exchange in relation to chemical

108

Figure 4.9. pCCh concentration, salinity, and temperature variation with depth [surface, mid-depth, and bottom waters) at Churchill River estuary stations CH3 [top panels) and CH4 (bottom panels). Samples were collected between June and August at low tide using a continuous automated CO2 monitor. Water was pumped from each depth for 10 minutes, data logging every 10 seconds. Data points represent an average of pCC>2 over the 10-minute sampling period. Temperature-salinity plots were created using Ocean Data View (Schlitzer, 2008).

0

2

1 4 Si o, a> Q u 6

8 1

10 0

2 +

Q 4 +

•13-Jun -14-Jul -26-Aug

P CH3

*

CH4

26-Aug

300 320 340 360 380 400 420 440 0 pC02 (uatm)

10 20 30 Salinity (psu)

Information from Figures 4.8 and 4.9 indicate that freshwater and marine

water sources (and thus water temperature, salinity, and PCO2) are better separated

at high tide compared to low tide. Although tidal amplitude varied with season, this

did not appear to affect the extent of upstream marine water intrusion at high tide

Page 125: Air-water carbon dioxide exchange in relation to chemical

109

(see Figure 4.8 bottom right panel). River discharge, which was greatest at the

beginning of June and middle of August, is the cause for estuarine stratification and

this is reflected in the CTD measurements at CH3 on June 13 and August 26 where

both surface and bottom waters were fresher, compared to the July 14 profile.

Thermoclines were observed for the June 13 and August 26 CTD profiles, where

only one thermocline was present on July 14 (Figure 4.9 top right panel).

4.4. Discussion

4.4.1 - Influence of season, location, depth, and tidal state on water chemistry

Water samples corresponding to high tide had greater DIC concentrations

relative to those collected at low tide. It appears that DIC is supplied to the estuary

basin by marine waters (which make up the majority of bottom waters); DIC

concentrations were elevated at station CH4, where the incoming tide causes

turbulent mixing at the confluence of the river and marine waters, and at station

CH2 where incoming/receding tide flows through a narrow channel creating

turbulent waters. TSS and NO3+NO2 followed a similar pattern, with highest

concentrations at stations CH4 and CHI. NO3+NO2 concentration were also higher at

high tide compared to low tide, although there was no obvious relationship between

TSS and tide height.

Chlorophyll a, PON, POC, DOC, CDOM, and P04 all had higher concentrations

at the river station and in surface waters, and follow a decreasing trend towards

Page 126: Air-water carbon dioxide exchange in relation to chemical

110

Hudson Bay. This trend is also evident if the data are discriminated by tide height,

that is, all parameters had higher concentrations at low tide (fresher) compared to

high tide (more saline). It is clear that 1) the concentration of these variables in the

Churchill River estuary was determined by river load, and 2) in general the

concentration of these variables was negatively correlated with salinity. Recall that

chlorophyll a, PON, and POC concentrations were all greatest in June and August,

which correspond to periods of maximum river discharge.

Alkalinity and salinity decreased in concentration from station CHI to river

station CH7. When samples are divided by tide height, the same pattern emerges:

alkalinity and salinity were higher at high tide compared to samples collected at low

tide. DO concentration was highest for the freshwater influenced stations and

decreased towards Hudson Bay, however there was little difference in DO between

surface and bottom water samples, or between samples at the far range of tidal

heights.

4.4.2 - Influence of water source and physiochemical conditions on pCQz and C02 flux

ApC02 was continuously monitored for the river water entering the estuary

in June, July, and August 2007 (Stainton et al., in prep. 2009), and the associated CO2

flux averaged -5.8 mmol nr2 day1 (Figure 4.10). ApC02 and CO2 flux measured at

station CH7 were all negative in July, however river waters were supersaturated at

times in mid-June and throughout most of August (Stainton et al., in prep. 2009).

Page 127: Air-water carbon dioxide exchange in relation to chemical

I l l

This is contrary to what was measured in the estuary, where ApCCb and the CO2 flux

were in fact negative at most stations sampled in June, and positive at most stations

sampled in July. The river monitoring dates in August [1-12) unfortunately do not

match up to the estuary survey sampling dates (21-26), so it is difficult to interpret

the difference in ApCCh and CO2 flux between river (mostly positive) and estuary

(mostly negative, with the exception of station CH5 at high tide) during this time

period without the collection of further data.

Figure 4.10. Churchill River CO2 flux and ApCCb. Water was sampled using a continuous air/water CO2 gas monitor. The dashed line indicates August 1.

E

d"

O

2 4 6 8 10 12 14 August

The average ApCCh observed here for the estuary is -4.39 uatm, which is in

between results associated with an open-water Churchill River mean A/9CO2

estimate of-103.9 uatm (Stainton et al., in prep. 2009), and measurements made by

Page 128: Air-water carbon dioxide exchange in relation to chemical

112

Else et al. (2008), where summer Hudson Bay ApCC>2 was +3.3 uatm (sampled 25 km

north-east of the Churchill River estuary mouth). Else et al. (2008) estimate the

mean open-water flux for all of Hudson Bay to be -0.73 mmol nr2 day-1, which is also

in agreement with our estuary flux estimate of-0.37 mmol nr2 day-1.

It appears that estuarine mixing changes the fluxes measured in the river

waters such that fluxes were more negative at stations where river and marine

water converge (station CH4 at high tide, and stations CH3 and CH2 at low tide, see

Figure 4.4). For example, of the surface waters measured at high tide in June,

convergence between the fresh river water and the intruding marine waters may

account for the more negative flux at station CH4 compared to station CH5, where

waters are fresh and isothermal (Figures 4.4 and 4.7). Similarly, fresh and marine

waters converge at stations CH3 and CH2 in August which may be attributed to the

lower fluxes of-4.85 and -4.05 mmol nr2 day1 compared to station CH4 (1.23 mmol

nr2 day-1) that was mostly isothermal (Figures 4.4 and 4.7). In August, flux at high

tide is positive for station CH5, but negative at station CH4. There does not appear to

be an explanation for this based on the profile cross-sections or from the water

chemistry collected at these sites.

Pearson's product moment correlation (r2) between CO2 flux and other

measured variables is very small. However, since CO2A varies by only a small

amount, CC»2w is ultimately dictating ACO2 and CO2 flux; therefore it is appropriate to

relate C02w to other water chemistry variables in order to decipher what may

influence the distribution of C02w in the estuary. CChw correlates negatively with

Page 129: Air-water carbon dioxide exchange in relation to chemical

water temperature, pH, chlorophyll a, and alkalinity, and correlates positively with

CDOM, DOC, P04, NO3+NO2, and Si (Table 4.2). This indicates that processes other

than mixing account for the distribution of pCChw in the Churchill River estuary. The

negative correlation between temperature and CO2W indicates that the distribution

of CO2W within the estuary is determined primarily by the relationship between CO2

solubility and temperature (higher solubility at lower temperatures, see Figure 2.3).

Table 4.2: Pearson's Product Moment Correlation (r) between CChw (umol 1_1) and several water chemistry variables.

Variables R^_

CCL vsCDOM 0.42 2w

CO., vsALK -0.11 2w

CO. vsDOC 0.21 2w

CO. vsPO. 0.16 2w 4

CO, vsNCL+NCL 0.26 2w 3 2

CO. vsSi 0.25 2w

C02 wvspH -0.50 C0 2 %sH 2 0 temp -0.66 CO„WvsCHLA -0.12 Why is there a difference in the mean ApCCh and CO2 flux between the river

station and the estuary? One possibility is that gas exchange is enhanced once the

river waters flow over the weir into the shallow estuary basin. Stainton et al. (2009,

in prep.) found that the diurnal signal for CO2 in the river was in fact a result of net

production and respiration upstream from the monitoring station. It is possible that

the riverine phytoplankton community ceases to photosynthesize (and thus be a

drawdown of CO2) once they mix with saline water in the estuary basin. This could

Page 130: Air-water carbon dioxide exchange in relation to chemical

114

cause an increase in estuarine CCW Another possibility is that waters undergo

enhanced gas exchange as they flow turbulently over the weir and into the estuary

basin. Where the river is fairly uniform in depth (mean ApCCb of 103.9 uatm), water

flows over the rock weir into a shallow basin where air-water exchange would take

place on a greater surface area (mean ApCC»2 of -4.39 uatm].

4.4.3 - Comparison to other high latitude estuaries

Relatively few studies have examined the relationship between air/water

CO2 exchange and physiochemical processes in estuaries, with fewer still

investigating high latitude systems. Table 4.3 lists flux and pCC»2 ranges for several

mid- and high-latitude estuaries where data does exist. Churchill River estuary flux

estimates are most similar in range to those from Hudson Bay (Else et al. 2008) and

the Beaufort Sea (Murata et al. 2008), though no similar high latitude estuarine

studies exist for comparison.

Table 4.3: Range of flux estimates and water pCC»2 from several mid- to high-latitude estuaries/coastal seas.

Location Churchill River estuary (Canada) Churchill River (Canada) Hudson Bay (Canada) Beaufort Sea (Canada) Lena delta/Laptevs Sea (Russia) Randers Fjord (Denmark) Rhine (Netherlands) Sheldt (Belgium/Netherlands) Thames (UK) Gironde (France) Hudson (US) York River (US)

Flux (mmol m'2 day"1) -11.04 to+22.76

-51 to+ 30.8 -19.6 to+16.5 -17.3 to-15

-36 to +290 70 to 160

-21 to+2028 0 to 1728 50 to 110 16 to 36

6.1 to 36.1

pC02 (uatm) 308 to 441 87 to 626

259 to 425 (/C02) 252.6 to 490 522 to 1171 355 to 400

570 to 1870 349 to 460 400 to 460 440 to 2860 515 to 1795

113to3467(ppmv)

Reference present study Stainton et al. 2009 (unpublished) Else et al. 2008 Murata et al. 2008 Semiletov 1999 Borges et al. 2004 Frankignoulle et al. 1998 Borges et al. 2004 Borges et al. 2004 Frankignoulle et al. 1998 Raymond et al. 1997 Raymond et al. 2000

Page 131: Air-water carbon dioxide exchange in relation to chemical

115

Water pCC>2 estimates are also similar to those from Hudson Bay (Else et al.,

2008) and the Beaufort (Murata et al., 2008), though are also within the same range

as data collected by Borges et al. (2004) for the Thames, Sheldt, and Randers Fjord

estuaries. However, the upper limit of flux estimates for these European estuaries is

much greater than that of the Churchill (Table 4.3). The watersheds of the lower

Churchill River and that of the Lena River are similar (both are composed of carbon-

rich soils and peatlands), however the pCC>2 range for the Lena River delta and

adjacent Laptevs Sea is much larger than that of the Churchill and Hudson Bay.

4.5 - Summary

We found the Churchill River estuary to be a net sink for CO2 throughout the

2007 open-water season. Comparing this data to that collected from the Churchill

River indicates that transformations within the estuary basin supply dissolved CO2

to the estuarine waters. Future research should include monitoring of the diurnal

variability in air-water CO2 exchange in both the estuary basin and the outer

estuarine waters through the summer growing season, as data presented here

represents only a snapshot of air-water CO2 exchange in the estuary basin.

Continuous monitoring of the inflowing river, estuary basin (at station CH5 or CH4),

and the outer plume waters would build on this work to create a complete picture of

air-water CO2 exchange and cycling in the Churchill River and estuary system.

Page 132: Air-water carbon dioxide exchange in relation to chemical

116

Acknowledgements

We gratefully acknowledge the field support and assistance from K. Swystun,

E. Chmelnitsky, T. Kelly, P. Fitzpatrick, J. Batstone, the Port of Churchill, and the

Churchill Northern Studies Centre. We thank Manitoba Hydro, Water Survey of

Canada, Environment Canada, and Natural Resources Canada for providing water

chemistry, river discharge, meteorological data, and satellite images. We also thank

M. Stainton for helpful discussion and comments. E. Stainton is funded by awards

from the Natural Sciences and Engineering Research Council of Canada (NSERC,

PGS-M) and the Northern Scientific Training Program (NSTP, Indian and Northern

Affairs). This research was conducted within the umbrella of Arctic-Net, a Network

of Centers of Excellence (http:/www.arcticnet-ulaval.ca). Support for this research

was provided by NSERC grants [Barber & Papakyriakou) and the Canada Research

Chairs program (Barber).

Page 133: Air-water carbon dioxide exchange in relation to chemical

References

Bezte, C.L. 2006. Lower Churchill River water level enhancement weir project post-project monitoring: assessment of water chemistry and phytoplankton responses to operation of the Project - Year VII, 2005, A report prepared for Manitoba Hydro by North/South Consultants Inc

Borges, A. V. (2005), Do we have enough pieces of the jigsaw to integrate C02 fluxes in the coastal ocean?, Estuaries, 28, 3-27.

Borges, A.V, J.P Vanderborght, et al. 2004. Variability of the gas transfer velocity of CO2 in a macrotidal estuary (the Scheldt). Estuaries. 27:593-603.

Carignan, R. 1998. Automated determination of carbon dioxide, oxygen, and nitrogen partial pressures in surface waters. Limnol. and Oceanogr. 43: 969-975.

Else, B.G.T, T. Papakyriakou, M.A. Granskog, and J.J. Yackel. 2008. Observations of sea surface/CO2 distributions and estimated air-sea CO2 fluxes in the Hudson Bay region (Canada) during the open-water season. J. Geophys. Res. 113: C08026, doi: 10.1029/2007JC004389.

Gorham E. 1991. Northern peatlands: role in the carbon cycle and probable responses to climatic warning. Ecological Applications 1: 182-195.

Jahne, B., G. Heinz, and W. Dietrich. 1987. Measurement of the diffusion coefficients of sparingly soluble gases in water. Journal of Geophysical Research 92:10767-10776.

JGOFS. 1994. JGOFS Report No. 19-Protocols for the Joint Global Ocean Flux Study (JGOFS) core measurements.

Kuzyk Z.A., R. Macdonald, et al. 2008. Sea ice, hydrological, and biological processes in the Churchill River estuary region, Hudson Bay. Estuarine, Coastal and Shelf Science.

Murata A., K. Shimada, S. Nishino, and M. Itoh. 2008. Distributions of surface water CO2 and air-sea flux of C02 in coastal regions of the Canadian Beaufort Sea in late summer. Biogeosciences Discussions. 5: 5093-5132.

Schlitzer, R. 2008. Ocean Data View, http://odv.awi.de.

Simpson, K. G., J.-E. Tremblay, Y. Gratton, andN. M. Price. 2008. An annual study of inorganic and organic nitrogen and phosphorus and silicic acid in the southeastern

Page 134: Air-water carbon dioxide exchange in relation to chemical

118

Beaufort Sea. Journal of Geophysical Research. 113: C07016, doi: 10.1029/2007 JC004462.

Stainton, E.C., R. Hesslein, T.N. Papakyriakou, and D.G. Barber. 2009. Summer air-water carbon dioxide exchange and water chemistry of the Churchill River as it enters Hudson Bay. In prep.

Stainton, M.P., M.J. Capel, and F.A.J. Armstrong. 1977. The chemical analysis of fresh water. Canadian Fisheries Marine Service, Miscellaneous Special Publication 25,180 pp.

Stedmon, C.A., S. Markager, and H. Kaas. 2000. Optical properties and signatures of chromophoric dissolved organic matter [CDOM] in Danish coastal waters. Estuarine, Coastal, and Shelf Science 52: 67-278.

Verardo, David J., P.N. Froelich and A. Mclntyre [1990). Determination of organic carbon and nitrogen in marine sediments using the Carlo Erba NA-1500 Analyzer. Deep Sea Research, vol.37, no.l pp 157-165.

Takahashi, T., S.C. Sutherland, C. Sweeney, et al. 2002. Global sea-air C02 flux based on climatological surface ocean pC02, and seasonal biological and temperature effects. Deep Sea Research Part II: Topical Studies in Oceanography. 49:1601-1622.

Wanninkhof, R. 1992. Relationship between wind speed and gas exchange over the ocean. Journal of Geophysical Research 97: 7373-7382.

Weiss, R.F. and B.A. Price. 1980. Nitrous oxide solubility in water and seawater. Marine Chemistry 8: 347-359.

Wetzel, R.W. 2001. Limnology: lake and river ecosystems. Academic Press.

Page 135: Air-water carbon dioxide exchange in relation to chemical

Chapter 5: Conclusions

5.1 - Summary and Conclusions

This research provides baseline water chemistry and air-water CO2 flux

estimates of the Churchill River and estuary.

Chapter 1 introduces the rationale behind this work, and outlines the main

research objectives.

Chapter 2 provides a literature review of: processes that affect aqueous CO2,

gas exchange at water surfaces, carbon exchange at the land-ocean interface, CO2

transport from coastal systems, and of climate implications with respect to carbon

cycling.

Chapter 3 presents the findings of data collected from the Churchill River.

Through air-water gas monitoring, the Churchill River was found to be a sink for CO2

throughout the 2007 monitoring period (Figure 5.1). Changes to the diurnal signal

of dissolved gases, water temperature, and turbidity were measured at times of

increased river flow. Production and respiration were calculated from the

uptake/release of CO2, and indicate that production outweighs respiration

throughout most of the monitoring period.

Chapter 4 presents the findings of data collected from the Churchill River

estuary throughout the 2007 open-water season. Unlike many high latitude rivers,

the Churchill was found to be a net productive system throughout the monitoring

period, however the riverine pCChw trends were not observed in estuarine waters,

Page 136: Air-water carbon dioxide exchange in relation to chemical

120

likely due to microbial respiration of river waters, and enhanced gas exchange as a

result of shallower water depths and turbulence within the estuary (Figure 5.1).

Most parameters measured for Churchill estuary waters mix conservatively within

the estuary basin (based on water source, river or marine). Estimated CO2 flux was

more negative at low tide compared to high tide. Variation in flux magnitude was

greatest at sampling locations with turbulence where fresh and marine water

masses meet and where the tide recedes through a narrow channel.

ApC02-103.9 uatm flux C02 - 5.4 mmol m 2 day1

HB ApC02 + 3.3 uatm flux C03 - 0.73 mmol nrr2 day1

Figure 5.1: Summary of processes within the Churchill River and estuary system.

5.2 - Future Directions

The Churchill River and estuary are an ideal location to study climate change

impacts on fresh-marine water physiochemistry and air-water carbon exchange.

The proximity of the Town of Churchill to the estuary, and the boat launching and

Page 137: Air-water carbon dioxide exchange in relation to chemical

121

docking facilities accessing the river make water-based sampling relatively easy

compared to other northern estuaries (e.g. the Nelson River).

In the beginning, this thesis was meant to include air-water pCC»2 and water

chemistry monitoring of Hudson Bay waters using a marine buoy (Environment

Canada) as a platform. Due to logistical issues, the buoy was not deployed. Future

research should include sampling of Hudson Bay waters throughout the open-water

season using this infrastructure in order to obtain a marine reference for ApCC»2. On

a larger scale, further measurements of air-water CO2 exchange should be taken

from other river-estuarine systems in Hudson Bay with the purpose of

understanding the collective importance of arctic and subarctic estuaries in

sequestering carbon.

Page 138: Air-water carbon dioxide exchange in relation to chemical

122

Appendix A

Summary of all Churchill River Estuary 2007 stations with latitude and longitude positions.

Station CH9 CH10 CH11 CH12 CH13 CH14 CH15 CH16 CH17 CH18 CH19 CH20 CH21 CH22 CH26 CH27 CH28 CH29 CH30 CH31 CH32 CH33 CH34 CH35 CH36 CH37 CH38 CH39 CH1 CH2

CH3A CH3B CH4 CH5 CH7

Long [deg. east] -94.1978 -94.2012 -94.2079 -94.2107 -94.2203 -94.2152 -94.2098 -94.2049 -94.1997 -94.1940 -94.1951 -94.2089 -94.2252 -94.1845 -94.2159 -94.2067 -94.2132 -94.2183 -94.2094 -94.1991 -94.1899 -94.2016 -94.1878 -94.1951 -94.1913 -94.2170 -94.2051 -94.2056 -94.1865 -94.2082 -94.2098 -94.2039 -94.2054 -94.1911 -94.1875

Lat [deg. north] 58.7104 58.7253 58.7595 58.7864 58.7717 58.7715 58.7720 58.7722 58.7719 58.7717 58.7044 58.7927 58.7719 58.7732 58.7880 58.7866 58.7874 58.7819 58.7803 58.7595 58.7604 58.7402 58.7002 58.6978 58.7267 58.7262 58.7489 58.7352 58.8098 58.7910 58.7756 58.7798 58.7406 58.6983 58.7904

Page 139: Air-water carbon dioxide exchange in relation to chemical

123

Appendix B

Figure 1: Relationship between gas trasfer velocity, k, and wind speed from four different models: Wanninkhof [1992), Wanninkhof and McGillis (1999], Nightingale et al. (2000), and Liss and Merlivat (1986). The Wanninkhof (1992) model for calculating gas transfer was used for flux calculations in this study.

/—"\

hour

S u o

100 -

90 -

80 -

70 -

60 •

50 •

40 -

• W92

• WM9 9

°N00

30

20 --

10 •-

0

6 8 10

Wind speed (m s 1 )

12 14 16

Table 1: mean, standard deviation, and range of C02 flux measurements based on four models for calculating gas transfer, k. The Wanninkhof (1992) model for calculating gas transfer was used for flux calculations in this study.

Flux C02 (mmol m"2 day"1)

AVG STDEV MAX MIN

W92 -5.8 11.9 30.8

-51.4

WM99 -3.2 9.0

20.7 -57.8

N00 -5.4 10.4 26.5

-43.5

LM86 -3.0 7.5

20.3 -33.1