Assessing the Mechanisms and Implications of Altered Carbon Cycling in Arctic and Boreal Lakes Rachel Fowler University of Maine, [email protected]

Assessing the Mechanisms and Implications of Altered Carbon Cycling in Arctic and Boreal Lakes Rachel Fowler University of Maine, Rachel.Dicker@Maine.Edu

The University of Maine DigitalCommons@UMaine

Electronic Theses and Dissertations Fogler Library

Spring 5-11-2019 Assessing the Mechanisms and Implications of Altered Carbon Cycling in Arctic and Boreal Lakes Rachel Fowler University of Maine, [email protected]

Follow this and additional works at: https://digitalcommons.library.umaine.edu/etd Part of the Terrestrial and Aquatic Ecology Commons

Recommended Citation Fowler, Rachel, "Assessing the Mechanisms and Implications of Altered Carbon Cycling in Arctic and Boreal Lakes" (2019). Electronic Theses and Dissertations. 2971. https://digitalcommons.library.umaine.edu/etd/2971

This Open-Access Thesis is brought to you for free and open access by DigitalCommons@UMaine. It has been accepted for inclusion in Electronic Theses and Dissertations by an authorized administrator of DigitalCommons@UMaine. For more information, please contact [email protected].

ASSESSING THE MECHANISMS AND IMPLICATIONS OF ALTERED

CARBON CYCLING IN ARCTIC AND BOREAL LAKES

Rachel A. Fowler

B.A. Bowdoin College, May 2009

M.S. University of Massachusetts Boston, August 2011

A DISSERTATION

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

(in Ecology and Environmental Science)

The Graduate School

University of Maine

May 2019

Advisory Committee:

Jasmine Saros, Professor of Paleoecology and Biological Sciences, Advisor

Hamish Greig, Associate Professor of Stream Ecology

Michael Kinnison, Professor of Evolutionary Applications

Christopher Osburn, Associate Professor of Marine, Earth, and Atmospheric Sciences

Jeffery Stone, Associate Professor of Earth and Environmental Systems

ASSESSING THE MECHANISMS AND IMPLICATIONS OF ALTERED

CARBON CYCLING IN ARCTIC AND BOREAL LAKES

By Rachel A. Fowler

Dissertation Advisor: Dr. Jasmine Saros

An Abstract of the Dissertation Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy (in Ecology and Environmental Science)

May 2019

Dissolved organic carbon (DOC) is an important component of lake ecology, as it contributes to light attenuation and carbon cycling. In recent years, DOC declined in a suite of lakes in Greenland. I performed experiments to test potential mechanisms of

DOC loss. The tested mechanisms did not reduce DOC concentration, but DOC composition was affected. I also paired water quality data with meteorological observations to evaluate effects of climate drivers on lake variables. The lake variables were temporally coherent and associated with patterns of mean annual precipitation.

In the northeastern U.S., recovery from acidification and climate change have contributed to lake brownification. I used paleolimnological techniques to compare algal responses of a clear lake versus a brown lake to multiple drivers in Acadia National Park.

My results suggested that algae in the clear lake were more sensitive to light, while the primary control in the brown lake was loss of nutrient subsidies from allochthonous

DOC. Both lakes exhibited signs of recovery toward pre-acidification conditions, although this response appeared dampened in the clear lake.

My dissertation research highlighted complex controls of DOC in Arctic lakes and clarified associations between climate drivers and lake variables important for carbon cycling and light attenuation. I also found key differences in algal dynamics in a clear versus a brown lake in Maine in response to multiple environmental drivers. With abrupt climate change in the Arctic and increasing brownification of boreal lakes, my research introduces timely new insights on the effects of these drivers on carbon cycling and algal ecology in Arctic and boreal lakes.

As part of the Integrative Graduate Education and Research Traineeship (IGERT)

Collaborative Immersion Project (CIP), I studied temperate Australian lakes in contrast to my focal work in Arctic and boreal systems. Ongoing drought threatens water resources in southeast Australia, many of which experience harmful algal blooms (HABs) rendering them unusable. I used paleolimnological methods to investigate HAB occurrence in a century-long record in two deep lakes. HAB development was likely driven by drought, so seasonal HABs may become a permanent fixture in the HAB- affected lake, as drought intensifies in coming decades.

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Jasmine Saros, for her support and guidance during this endeavor. I have benefited from her exceptional mentoring skills, and am grateful for the opportunities for field work, travel to conferences and workshops, and funding that she facilitated. I also thank my committee members, Dr. Hamish Greig, Dr.

Michael Kinnison, Dr. Christopher Osburn, and Dr. Jeffery Stone, for providing thoughtful input and expertise that helped strengthen each area of this dissertation. Thank you to lab members past and present who generously gave their time and support to contribute to field work, sample analysis, study group discussion, and fun diversions (like intramural volleyball): Kate Warner, Carl Tugend, Ben Burpee, Nora Theodore, Amy

Kireta, Heera Malik, Robert Northington, Edna Pedraza Garzón, Joe Mohan, Vendy

Hazuková, and Simona Lukasik. Classmates, friends, and colleagues in the Climate

Change Institute never hesitated to talk through experiment planning, econ homework, or rankings of Orono pizza places, and for that I am grateful.

I am grateful for assistance with algal pigment analysis and interpretation from

Suzanne McGowan from the University of Nottingham, as well as assistance from

Amanda Klemmer, Tamara Levitsky, and the lab of L. Brian Perkins. Bill Gawley and

Shannon Wiggin from the National Park Service helped with data collection and development of ideas for work in Acadia National Park, and Peter Gell and Phuong Doan from Federation University in Ballarat, Australia were generous in their support of logistics, equipment usage, and data interpretation. I thank Andrea Nurse and Dennis

Anderson for help with data collection and diatom analysis, as well as Kristin Strock,

Helen Schlimm, and Max Egener for helping with Greenland field work, and Jeff Auger

for providing thoughtful assistance with meteorological analysis. Dirk van As and Emma

Johansson kindly shared meteorological data from the GRASP project, and William

Halteman and Adam Heathcote provided statistical advice. I thank Ivan Fernandez for thoughtful discussions on soil organic matter. Jo Bullard shared advice on dust sampling in Greenland, Mike Handley conducted elemental analysis of this dust, and C.T. Hess and

John Cummings in the Department of Physics and Astronomy at the University of Maine performed dating of sediment cores. Becky Addessi and Betty Lee in the Climate Change

Institute were incredibly helpful with all things administrative. This research was funded by the U.S. National Science Foundation Grants #1144423 and #1203434, as well as a

Schoodic Institute Research Fellowship and University of Maine Graduate Student

Government grants. I am also grateful to Friends of Acadia for funding my Aquatic

Scientist position and the School of Biology & Ecology for teaching assistantships.

Thank you to my friends and family for your patience and support as I pursued this research. Your unconditional encouragement and your own persistence through challenges provided a continual source of inspiration. Thank you especially to my husband, Matt. You’ve helped me polish figures, paddled in the pouring rain to collect water samples, cooked extra meals, endured long periods apart, and through it all, have been my happy place to come home to. And finally, thanks to our baby girl, born April

2019, who provided her own motivational “kicks” as I completed this journey!

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... ii

LIST OF TABLES ...... viii

LIST OF FIGURES ...... ix

1. CHAPTER ONE: INTRODUCTION ...... 1

2. CHAPTER TWO: SHIFTING DOC CONCENTRATION AND QUALITY IN

FRESHWATER LAKES OF THE KANGERLUSSUAQ REGION: AN

EXPERIMENTAL ASSESSMENT OF POSSIBLE MECHANISMS ...... 6

Abstract ...... 6

Introduction ...... 7

Methods...... 13

Study site ...... 13

Experimental design...... 16

Dust collection ...... 19

DOC analysis ...... 20

Data analysis ...... 20

Results ...... 21

Soil water experiment ...... 27

Dust addition + light exposure experiment ...... 27

Bacterial activity + light exposure experiment ...... 29

Dust input rates and elemental analysis ...... 30

Discussion ...... 30

Dust ...... 31

Bacterial activity ...... 32

Solar radiation ...... 34

Future work ...... 37

Conclusions ...... 38

3. CHAPTER THREE: COHERENT CLIMATE-DRIVEN CHANGES IN

DISSOLVED ORGANIC CARBON AND WATER CLARITY IN ARCTIC

LAKES OF WEST GREENLAND ...... 40

Abstract ...... 40

Introduction ...... 41

Methods...... 44

Study area...... 44

Lake variables ...... 45

Climate variables ...... 49

Statistical analyses ...... 49

Results ...... 50

Climate and lake variables ...... 50

Within- and between-region lake variable coherence ...... 54

Effects of climate on lake responses ...... 55

Discussion ...... 59

4. CHAPTER FOUR: PALEOLIMNOLOGICAL COMPARISON OF ALGAL

CHANGES IN A CLEAR LAKE VERSUS A BROWN LAKES OVER THE

LAST TWO CENTURIES IN THE NORTHEASTERN U.S.A...... 68

Abstract ...... 68

Introduction ...... 69

Methods...... 72

Site history and description of study lakes ...... 72

Core collection and analysis ...... 78

Data analysis ...... 80

Results ...... 81

Core chronologies ...... 81

Diatom profiles ...... 83

Algal pigments ...... 86

Discussion ...... 90

Conclusions ...... 95

5. CHAPTER FIVE: DECIPHERING DIVERGENT HARMFUL ALGAL

BLOOM OCCURRENCE IN TWO SOUTHEAST AUSTRALIAN DEEP

LAKES: IMPLICATIONS FOR LAKE CONSERVATION IN THE 21ST

CENTURY ...... 97

Abstract ...... 97

Introduction ...... 98

Methods ...... 102

Site descriptions ...... 102

Lake sampling ...... 105

Laboratory methods ...... 105

Core collection and analysis ...... 106

Data analysis ...... 108

Results ...... 109

Pelagic variables ...... 109

Core chronologies ...... 110

Diatom profiles ...... 112

Algal pigments ...... 115

Discussion ...... 118

6. CHAPTER SIX: CONCLUSION ...... 128

DOC dynamics in Arctic and boreal systems ...... 128

IGERT Collaborative Immersion Project (CIP) experience ...... 132

REFERENCES ...... 134

APPENDIX: RESULTS OF CHAPTER TWO DUST + LIGHT AND

BACTERIA + LIGHT EXPERIMENTS ...... 157

BIOGRAPHY OF THE AUTHOR ...... 159

vii

LIST OF TABLES

Table 2.1. Characteristics of the three study lakes ...... 15

Table 2.2. Mean percent change in DOC concentration and quality metrics for

dust + light and bacteria + light experiments ...... 22

Table 2.3. p-values for dust addition-light exposure and bacterial activity-light

exposure experiments...... 23

Table 3.1. Characteristics of study lakes...... 48

Table 3.2. Climate characteristics of the two study regions from 2013-2017 ...... 51

Table 4.1. Select characteristics of Jordan Pond and Seal Cove Pond ...... 76

Table 4.2. PCA Axis 1 loadings for Jordan Pond and Seal Cove Pond pigments ...... 87

Table 5.1. Select characteristics of Lakes Bullen Merri and Purrumbete ...... 104

Table 5.2. Pelagic characteristics of Lakes Bullen Merri and Purrumbete ...... 109

Table 5.3. PCA Axis 1 loadings for Lake Bullen Merri and Lake Purrumbete

pigments ...... 116

Table A.1. Dust + light and bacteria + light experimental results from Chapter 2 ...... 157

viii

LIST OF FIGURES

Figure 2.1. Floodplain dust entrained by wind, Kangerlussuaq ...... 9

Figure 2.2. Conceptual diagram illustrating potential chemical (photodegradation),

biological (bacterial consumption), and physical (flocculation and

precipitation) pathways for DOC loss in surface waters ...... 11

Figure 2.3. Study area...... 16

Figure 2.4. Results of Lake SS2 soil water exposure to dust addition on

SUVA254, S275–295, and [DOC], represented by percent change (n=3) ...... 24

Figure 2.5. Results of SS2, SS1381, and SS901 lake water exposure to dust

addition and light on SUVA254, S275–295, and [DOC], represented by

percent change (n=3)...... 25

Figure 2.6. Results of SS2, SS1381, and SS901 lake water exposure to bacterial

activity and light on SUVA254, S275–295, and [DOC], represented by

percent change (n=3)...... 26

Figure 2.7. S versus a*375 for dust + light (a-c) and bacteria + light (d-f)

experiments (n=3) ...... 27

Figure 3.1. Map of study site ...... 45

Figure 3.2. Climate trends in WD and CW regions from 2013-2017 ...... 51

Figure 3.3. Absolute values of lake responses in WD and CW regions from

2013-2018 ...... 53

Figure 3.4. Mean temporal coherence of lake responses within and between WD

and CW regions from 2013-2018 ...... 55

Figure 3.5. RDA triplot illustrating differences between climate drivers (black

arrows), lake responses (gray arrows), and regions (WD = black

circles, CW = gray circles) from 2013-2017 ...... 56

Figure 3.6. Comparison of interannual variability of climate drivers and lake

responses between WD and CW regions ...... 58

Figure 3.7. Climate exposure index (CEI) of WD and CW lakes ...... 60

Figure 4.1. Map highlighting Jordan Pond and Seal Cove Pond in Acadia National

Park ...... 73

2- Figure 4.2. SO4 in wet deposition at ME-98 station from 1981-2017, data

acquired from NADP/NTN ...... 74

Figure 4.3. Mean minimum temperature and total annual precipitation at ME-98

station from 1940-2013, data acquired from NCDC ...... 75

Figure 4.4. Changes in DOC and Secchi depth in Jordan Pond (a, c) and Seal

Cove Pond (b, d) from 1995-2008 ...... 78

Figure 4.5. Stratigraphic plots of 210Pb activities and sediment chronologies in

Jordan Pond (a, b) and Seal Cove Pond (c, d) ...... 82

Figure 4.6. Relative abundances of selected diatom taxa, mixing depth index

(MDI), planktic:benthic diatom ratio (P:B), and DCA Axis 1 in SD

units for Jordan Pond (a) and Seal Cove Pond (b) sediment cores ...... 84

Figure 4.7. Extracted pigment concentrations, percent LOI, and PCA Axis 1 in

Jordan Pond (a) and Seal Cove Pond (b) sediment cores ...... 88

Figure 5.1. Map of deep crater lake study sites in SE Australia ...... 103

Figure 5.2. Stratigraphic plots of 210Pb activities and sediment chronologies in

Lakes Bullen Merri (a, b) and Purrumbete (c, d) ...... 111

Figure 5.3. Relative abundances of selected diatom taxa, DCA Axis 1 in SD

units, and diatom-inferred salinity (DI-S) for Lake Bullen Merri

sediment core ...... 113

Figure 5.4. Relative abundances of selected diatom taxa, planktic:benthic diatom

ratio (P:B), DCA Axis 1 in SD units, and diatom-inferred salinity

(DI-S) for Lake Purrumbete sediment core ...... 115

Figure 5.5. Extracted pigment concentrations, percent LOI, and PCA Axis 1

scores in SD units in Lake Bullen Merri sediment core ...... 117

Figure 5.6. Extracted pigment concentrations, percent LOI, and PCA Axis 1

scores in SD units in Lake Purrumbete sediment core ...... 118

CHAPTER 1

INTRODUCTION

Mechanisms of dissolved organic carbon (DOC) change in northern hemisphere surface waters have been largely understood through study of mid-latitude lakes.

Documented increases in DOC concentration in mid-latitude lakes in recent decades have been attributed to changes in atmospheric deposition (Evans et al. 2006; Monteith et al.

2007; Clark et al. 2010) and climatic changes such as temperature or precipitation (Reche

& Pace 2002; Weyhenmeyer et al. 2004; Jennings et al. 2012). For example,

Weyhenmeyer & Karlsson (2009) showed a strong temporal and spatial response of DOC concentration to increased temperature in boreal lakes. These increases in DOC can cause subsequent changes in water clarity and thermal structure, resulting in biological implications for primary production (Jennings et al. 2012; Solomon et al. 2015) and relative abundance of zooplankton groups (Cooke et al. 2015). While mechanisms behind changes in lakewater DOC concentration and quality are becoming better understood at mid-latitudes, DOC trends are not as consistent in Arctic lakes, and work in this area remains in the early stages.

DOC concentration has changed variably in Arctic lakes, and DOC export at high latitudes has been predicted to both increase (Finlay et al. 2006) and decrease (Striegl et al. 2005) with climate warming and thawing of permafrost. Saros et al. (2015) described a synchronous decrease in DOC concentration over the last fifteen years in a suite of lakes in west Greenland. While the mechanisms behind shifts in DOC are more clearly understood for lakes at mid-latitudes, much less is known about causes of variation in

Arctic lakes. Lessons learned from mid-latitude lake studies will be a useful guide in the search for mechanisms behind Arctic DOC change. However, climate change in the

Arctic is occurring more quickly than at mid-latitudes, altering properties of landscape variables (e.g. deepening of permafrost active layer) that could contribute to DOC changes via mechanisms not encountered at mid-latitudes.

Over the past two decades, the Arctic has experienced more rapid, non-linear environmental change than any other place on Earth. The most striking example of this change occurs in Greenland, where 2007-2012 temperatures were almost 3̊ C greater than the 1979-2000 average (Mayewski et al. 2014). In addition to increased temperatures, abrupt climate change (ACC) can influence climate variables such as precipitation frequency and severity, cloud cover, and wind speed, inducing a number of cascading processes within landscape and hydrological systems. Temperature, precipitation, wind, and cloud cover can contribute to geographically broad, unidirectional changes in permafrost stability, snowmelt across the landscape, ice-out timing in lakes, glacial meltwater flux, lake level, vegetation cover, and a host of other variables. Furthermore, each of these landscape-scale variables may influence lake characteristics, such as UV attenuation depth, DOC concentration, nutrients, and thermal stratification, which may exhibit varied temporal coherence across a region.

Lakes are important study systems because they integrate climate change effects occurring in their catchments—and despite their comparatively small areas within landscapes, present these changes in measurable ways, acting as sentinels of environmental change (Williamson et al. 2009). Recent work shows that surface temperatures in lakes are rising faster than both ocean and air temperatures (O’Reilly et al. 2015), so with the rapid elevation in air temperature occurring in the Arctic (AMAP

2011; Mayewski et al. 2014), Arctic lakes are especially at risk of abrupt changes. The

Arctic and sub-arctic contain approximately 24% of the area for the world’s 300+ million lakes (Williamson et al. 2009; Lundin et al. 2015). As sites of intense carbon (C) cycling,

Arctic lakes play an important role in the regulation of carbon dioxide (CO2) emissions to the global carbon cycle (Kling et al. 1991; Tranvik et al. 2009) and also bury carbon in their sediments (Cole et al. 2007). Because of the rapid warming occurring within these lakes, it is important to refine our understanding of the ways that ACC can influence cascading landscape and hydrological processes that impact carbon cycling within Arctic lakes, primarily through improving our understanding of factors that influence shifts in

DOC quantity and quality.

DOC represents a component of the dissolved organic matter (DOM) pool that is easy to quantify, and is therefore often used to estimate DOM within a system. DOC is produced during autochthonous and allochthonous decomposition and/or exudation of animal, plant, and microbial biomass. As relatively simple-to-measure metrics that are highly sensitive to environmental change, DOC concentration and quality play an important role in the sentinel function of lakes. In addition, DOC contributes to within- lake carbon cycling and burial (Williamson et al. 2009). DOC production and consumption can be influenced by hydrological factors, soil chemistry, microbial activity, and vegetation cover (Peacock et al. 2014), and the composition or quality of DOC is controlled by its transport, reactivity, and sources of origin (Williamson et al. 1999;

Sachse et al. 2001). Within a lake system, DOC can contribute to both carbon burial in sediments and release of CO2 via respiration and/or photochemical degradation (Osburn et al. 2011). DOC also plays a role in attenuating ultraviolet (UV) radiation that may be harmful to lake organisms, as well as attenuation of visible light important for lake

thermal stratification, with consequences for lake structure and function (Williamson et al. 2009). For example, changes in mixing depth can have ramifications for nutrient availability, which then impacts algal community structure (Williamson et al. 1999).

Lakes in eastern North America have undergone increasing surface water DOC concentration since the implementation of the 1990 Clean Air Act Amendments (Driscoll et al. 2003; Jeffries et al. 2003). A major research objective of lake ecology since this time has been to understand how recovery from acidification (Keller et al. 1999; Stoddard et al. 1999) and subsequent DOC change (Donahue et al. 1998; Jeffries et al. 2003) affect aquatic systems. Not all lakes in New England, and specifically Maine, responded similarly to acid deposition (Davis et al. 1983). In addition, clear and brown lakes exhibit different responses to recovery from acidification and other external pressures

(Williamson et al. 2015). Acadia National Park (ANP) in Maine has experienced reductions in acid deposition, as well as increases in temperature (Strock et al. 2017) and precipitation for several decades. As ANP contains both clear and brown lakes, it is an ideal study location to compare algal responses of a clear versus a brown lake to changes in DOC that result from acidification recovery and climate change effects.

The objectives of this dissertation research were to understand drivers of recent shifts in DOC quantity and quality in Arctic lakes, and to decipher effects of DOC shifts on algal responses in boreal lakes experiencing recovery from acid deposition and increased temperature and precipitation. In Chapter 2, I experimentally evaluated possible mechanisms responsible for decreasing DOC concentration in lakes of west Greenland.

To assess climate-mediated terrestrial-aquatic linkages in Arctic lakes and potential impacts on light attenuation and C cycling in the lakes, I analyzed the coherence of lake

responses in two regions of west Greenland with different climate patterns in Chapter 3.

In Chapter 4, I explored ecological responses of a clear and a brown lake in ANP to reduced acid deposition in the context of a 140-year paleolimnological record, during which time multiple external drivers influenced DOC in both lakes.

Additionally, as a fellow of the NSF Integrative Graduate Education and Research

Traineeship (IGERT), I was part of an interdisciplinary collaborative immersion project

(CIP) that examined the intensification of harmful algal blooms (HABs) in water resources of southeast Australia as a result of ACC. My role in this project was to assess

HAB incidence in two deep lakes over the last century and better understand why HABs currently develop in only one of the lakes, as described in Chapter 5.

CHAPTER 2

SHIFTING DOC CONCENTRATION AND QUALITY IN FRESHWATER

LAKES OF THE KANGERLUSSUAQ REGION: AN EXPERIMENTAL

ASSESSMENT OF POSSIBLE MECHANISMS

Abstract

Since 2003, concentrations of dissolved organic carbon (DOC) in lakes in the

Kangerlussuaq region declined by 14-55%, with these decreasing DOC concentrations potentially altering lake ecology and reflecting changes in regional carbon (C) cycling.

To evaluate possible mechanisms responsible for this shift, we performed experiments to test the effects of dust addition, bacterial activity, or photodegradation on DOC concentration and two DOC quality metrics: the specific ultraviolet absorbance

(SUVA254) and the chromophoric DOC spectral slope coefficient (S275-295). Lake water

DOC concentrations did not decline in any treatments, but there were changes in DOC quality. Dust addition increased SUVA254 and decreased the magnitude of S275-295 in one lake, the impacts of bacterial activity were variable, and sunlight exposure elicited a decline in SUVA254 and an increase the magnitude of S275-295 in all lakes. These results suggest that DOC pools in the study lakes are photoreactive, even though the lakes are characterized by long residence times, but that declining DOC concentration did not result from this mechanism. While the tested mechanisms did not explain the decline in

DOC concentration observed in recent years, they did yield new information about how dust, bacterial activity, or light can influence DOC quality in lakes of the Kangerlussuaq region.

Introduction

Shifting trends in lake water DOC may signal broader climate-mediated biogeochemical changes within lake catchments (Williamson et al. 2009). As DOC in surface water originates from both terrestrial and aquatic sources, it can reflect changes in within-lake processes as well as broader environmental and atmospheric conditions. With climate warming and accelerated permafrost thaw, DOC has been predicted to increase

(Frey and Smith 2005), or decrease (Striegl et al. 2005) in aquatic surface waters of the

Arctic. Because DOC is a primary regulator of lake ecosystem function (Williamson et al. 1999), it is important to identify the drivers behind substantial changes in lake water

DOC concentration and quality.

In freshwater lakes of the Kangerlussuaq area, Anderson & Stedmon (2007) collected DOC concentration data from 2001-2003, and Saros et al. (2015) compared these earlier results to DOC data from twelve of the same lakes from 2011-2014. In this period of about a decade, DOC concentration substantially declined by 14-55% (1-24 mg

C L-1) in eleven of the lakes. The mechanism(s) leading to this decline in DOC, however, remains unclear. Possible mechanisms leading to DOC decline may include flocculation with iron (Fe) in dust and subsequent precipitation from surface waters (Maloney et al.

2005), bacterial degradation (Guillemette & del Giorgio 2011), or photodegradation from light exposure (Cory et al. 2014). The lack of understanding about the drivers of the downward DOC trend must be resolved to better quantify the role of lakes in Arctic terrestrial C budgets as climate warming progresses in this region.

Autochthonous and allochthonous sources contribute to the DOC pool of lakes.

Phytoplankton and macrophytes generate DOC during growth and senescence, and

terrestrial DOC produced in the catchment can be transported across the landscape and into lake basins. In this study, we used controlled experiments to better understand mechanisms responsible for losses of DOC in lakes.

One possible mechanism behind declining lake water DOC concentrations is climate-driven change in dust production that may be occurring in Kangerlussuaq, as in other parts of the world. Between the periods of 1979-2000 and 2007-2012, surface air temperatures in Greenland increased by almost 3̊ C (Mayewski et al. 2014). This rapid surge in temperature is responsible for increased melting of the Greenland Ice Sheet

(Khan et al. 2015) and elevated flows of glacial meltwater (Van As et al. 2014). The

Russell Glacier, flowing west from the Greenland Ice Sheet, feeds into the Watson River that runs through the town of Kangerlussuaq and meets the head of the fjord

Kangerlussuaq. Rising temperatures cause meltwater to accumulate beneath the glacier until enough pressure has developed for growing subglacial lakes to explode into a glacial outburst flood, or jökulhlaup (Mernild et al. 2008). During jökulhlaups and other seasonal flooding events, the flood waters surge over the glacial outwash plain, carrying high concentrations of glacial flour across the landscape. These dust deposits can easily be picked up by wind and transported to other features across the landscape (Figure 2.1;

Bullard 2013). Because of the rapid increases in Arctic temperatures, Prospero et al.

(2012) predict that proglacial dust activity will intensify and become geographically more extensive. Interestingly, the timing of the return of jökulhlaups to Kangerlussuaq (from

2007 to present) is synchronous with the decline in regional lakewater DOC observed between 2003 and 2011 sampling campaigns. Aeolian-driven dust inputs can be rich in iron (Jickells et al. 2005; Mahowald et al. 2005), and humic-Fe associations are well-

documented (Shapiro 1966; Maloney et al. 2005; Francko & Heath 1982; Jones et al.

1993; Dillon & Molot 1997). Some fractions of humic DOC are known to adsorb to Fe oxides, resulting in flocculation of this humic DOC (Tipping 1981; Fox & Wofsy 1983), and hence, decline in lake water DOC concentration (Figure 2.2). It should be noted, however, that dust can also contain organic carbon, and thus be a source of DOC to lakes as well (Mladenov et al. 2012). Furthermore, phosphorus (P) can stimulate bacterial growth in lakes (Carlsson & Caron 2001), so the P content of the dust could influence bacterial contributions to the DOC pool.

Figure 2.1. Floodplain dust entrained by wind, Kangerlussuaq. Photo credit: Rachel Fowler.

Additionally, exposure to sunlight may influence both DOC quantity and composition in lakes. Osburn et al. (2001) and Spencer et al. (2009) demonstrated that natural sunlight causes photodegradation of DOC, resulting in decreases in DOC

concentration and of SUVA254, and Lapierre & del Giorgio (2014) showed high rates of

DOC loss caused by photodegradation in irradiation experiments. Photochemical degradation breaks down DOC into dissolved inorganic carbon (DIC) and carbon monoxide (CO), and promotes photobleaching, or loss of DOC color (Figure 2.2; Moran et al. 2000; Morris & Hargreaves 1997). Cory et al. (2014) found that photooxidation of

DOC by sunlight accounted for 70-95% of all water column C processing in lakes and streams in and near the Kuparuk River Basin in Alaska. In other studies, loss of DOC with light exposure has been significant, as well (i.e. 31% of initial concentration; Moran et al. 2000). In addition, sunlight may interact with increased Fe from dust inputs, with photoflocculation effectively reducing the amount of humic DOC in lake water (Helms et al. 2013; Figure 2.2). However, cloud cover has increased in the Arctic (Eastman and

Warren 2013), and there has been no apparent change in the timing of ice-out in the

Kangerlussuaq region between 2003-2011 (Saros et al. 2015), indicating that exposure to solar radiation has not likely increased in this area. One pathway by which photodegradation may become a more important mechanism is via amplified permafrost thaw that loads proportionally more photochemically-labile fractions into the DOC pool.

Figure 2.2. Conceptual diagram illustrating potential chemical (photodegradation), biological (bacterial consumption), and physical (flocculation and precipitation) pathways for DOC loss in surface waters. Interactions between dust and solar radiation may lead to DOC loss via photoflocculation, and interactions between solar radiation and bacterial activity could have variable effects. Note that pathways are not exclusive of each other (i.e. DOC loss could occur via multiple pathways).

Interactions between photochemical degradation and biological processing have also been demonstrated, in which bacteria readily assimilate labile, low molecular weight

(LMW) DOM photoproducts. As a result, exposure to sunlight can increase DOM processing through chemical and biological pathways, and can also boost bacterial production and respiration (Guillemette & del Giorgio 2012; Miller & Moran 1997).

Ultraviolet light can also inhibit bacterial activity, and could have varying effects on 11

bacterial processing rates of DOC (Granéli et al. 1996). While bacterial activity in natural waters is typically thought to degrade DOC (Lu et al. 2013), variables like temperature, initial DOC composition, nutrients, and light exposure can influence bioavailability of

DOC and the simultaneous production and consumption of DOM by bacteria in aquatic systems (Guillemette & del Giorgio 2012).

There are multiple potential drivers of declining DOC in West Greenland lakes, with the physical, chemical, and biological effects of sunlight, dust, and bacterial transformation possibly contributing to this loss to varying degrees. To investigate possible mechanisms behind these recent changes, we experimentally tested the effects of dust inputs, light, or bacterial activity on DOC concentration and quality (SUVA254 and

S275-295) in lake surface water. We performed two experiments, both factorial in design, based on the mechanisms described above and the possible interactions (dust addition × photodegradation and photodegradation × bacterial activity) depicted in Figure 2.2. We hypothesized that all three pathways (dust addition, sunlight exposure, or bacterial processing) would each lead to reductions in DOC concentration, with potential interactions between light exposure with dust addition or bacterial activity. SUVA254 is an index of source and composition of DOC, with higher values corresponding to higher aromaticity (Williams et al. 2010; Weishaar et al. 2003). We predicted that light would decrease SUVA254, bacterial activity would increase SUVA254, and that dust would have little effect. The magnitude of S275-295 increases (becomes more negative) with photobleaching extent and is inversely proportional to the molecular weight (MW) of

DOM (Helms et al. 2008; Fichot & Benner 2012), and so we predicted that sunlight

exposure would increase the magnitude of S275-295, bacterial activity would decrease the magnitude of S275-295, and that dust inputs would cause no change.

Methods

Study site

The Kangerlussuaq region of West Greenland, situated just north of the Arctic

Circle between 66-68°N and 50-53°W, makes up part of a 150-km ice-free margin between the coast and the Greenland Ice Sheet, which is drained by a large fjord. The region contains approximately 20,000 lakes that cover ~14% of the dry tundra landscape

(Anderson et al. 2009). Most these lakes are oligotrophic and chemically dilute

(Anderson et al. 2001). Betula and Salix woody shrubs are the dominant vegetation, and the continuous permafrost soil active layer ranges from 20-90 cm (Johansson et al. 2015).

The mean summer temperature and annual precipitation are 10.2°C and 173 mm yr-1, respectively, with most of the precipitation occurring as rain between May and

September. Annual evapotranspiration is 200-300 mm yr-1, mostly occurring in the summer (Bosson et al. 2013). The wind regime is dominated by strong easterly katabatic winds that are driven by the Greenland Ice Sheet, as well as smaller-scale westerly winds resulting from Atlantic storms (Bullard & Austin 2011). The lakes in this region are characterized as having long residence times (Leng & Anderson 2003; Anderson &

Stedmon 2007), with run-off from the active layer (Leng & Anderson 2003) and precipitation as the primary water inputs, with minimal surface inflow or outflow

(Bosson et al. 2013).

The average summer 2015 DOC concentration across lakes in this region was 15

-1 -1 -1 -1 mg L , while the average SUVA254 and S275-295 were 1.5 mg L m and -0.03 nm ,

respectively (Saros, unpublished data). Figure 2.3 depicts the lakes selected as experimental subjects for this study on the Kangerlussuaq landscape—two are high-DOC lakes with differing DOC quality and one is a low-DOC lake. Lakes with different initial

DOC properties were selected to provide insight about how these properties may alter the response to experimental drivers. Lake SS2 and Lake SS1381 are high-DOC (22-27 mg

L-1) lakes in the Kellyville region downwind of the proglacial floodplain. In contrast,

DOC concentrations in the surface waters of Lake SS901, located near the Greenland Ice

-1 -1 -1 Sheet, are 7-9 mg L . Lake SS901 has the highest SUVA254 (1.76 L mg C m ) and

-1 lowest-magnitude S275-295 (-0.024 nm ), while Lake SS1381 has the lowest SUVA254

-1 -1 -1 (1.03 L mg C m ) and highest-magnitude S275-295 (-0.034 nm ; Table 2.1). Because water in the study lakes has already been exposed to some quantity of aeolian-transported dust, we also conducted an experiment with soil water extracts near Lake SS2 to determine how dust addition affects DOC in water with minimal prior exposure to dust.

The initial DOC concentration of the soil water extracts was 68 mg L-1, while the initial

-1 -1 -1 SUVA254 was 8.5 L mg C m , and the initial S275-295 was -0.013 nm .

Table 2.1. Characteristics of the three study lakes. DOC, pH, and temperature measurements were taken from lake epilimnia in June 2015. Nutrient data are from June-July 2013. ‘a.s.l.’ = above sea level.

Lake [DOC] SUVA254 S275-295 pH Temperature DIN TP Maximum Surface Elevation (mg L-1) (L mg C-1 m-1) (nm-1) (°C) (µg L-1) (µg L-1) depth (m) area (ha) (m a.s.l.) SS2 23.0 1.42 -0.030 8.9 12.0 15 3 12 41 187 SS1381 24.0 1.03 -0.034 8.7 17.6 5 6 21 22 171 SS901 7.3 1.76 -0.024 8.0 8.3 5 4 15 11 399

Figure 2.3. Study area. SS2, SS1381, and SS901 denote the locations of experimental lakes.

Experimental design

We used factorial designs in two experiments with water from each experimental lake. We tested dust (none or added) × light (light or dark) in the first experiment, and bacterial activity (filtered or inoculated) × light (light or dark) in the second experiment.

“Light” includes wavelengths for both UV and photosynthetically active radiation (PAR).

The dust + light experiment was conducted in July 2015 and the bacterial activity + light experiment in July 2016. Three replicates of each treatment for each lake were created.

For both experiments, two liters of water were collected from each lake at 2-3 m depth and filtered through 100 µm Nitex mesh to remove large zooplankton. A vacuum 16

pump was used to gently pull the water through 0.7 µm GF/F filters, and this filtered water was used for experiments. For the dust + light experiment, fine-grained dust was collected from the floodplain and added at a proportion of 2.5 mg dust to 250 ml lake water (comparable to summer dust deposition measured by Engels 2003). Lake water for the bacteria + light experiment was filtered through Nucleopore 0.2 µm track-edged membranes to remove bacteria. The 0.2 µm-filtered water served as the filtered treatment, while samples for bacterial inoculation treatment were inoculated with 3 ml of the 0.7

µm-filtered lake water. Treatments (250-ml volume) were incubated in dark and light conditions for seven days. The dust + light experiment was carried out in 2015, and the bacteria + light experiment in 2016. “Dark” samples in the dust + light experiment were incubated in acid-washed 250-ml HDPE bottles in a dark laboratory cupboard at 22°C.

“Light” samples were incubated outside in pre-leached and dried 500-ml liquid-tight bags

(Bitran S Series) to allow for unobstructed penetration of light into the water, at 17°C average temperature. Light samples were randomly placed in sunlight-exposed clear plastic bins containing 3 cm water for temperature buffering, and HOBO loggers inside the bins recorded daily temperature and light intensity fluctuations. For the bacteria + light experiment, incubation conditions were modified so that both light and dark treatments were incubated in the same location and at the same temperature. Both dark and light samples were incubated in Bitran bags, and the dark samples were wrapped in aluminum foil to block light. As in the dust + light experiment, the samples were then randomly placed in water-filled clear plastic bins for the duration of the experiment.

Water for DOC analysis was sampled at the onset of each experiment and again on day 7, then filtered through 0.7 µm GF/F filters, and immediately frozen at -20°C. It has been

noted that freezing water samples may reduce DOC concentration (Fellman et al. 2008) or aromaticity (Fellman et al. 2008; Peacock et al. 2015) in surface water samples via flocculation. However, we were most interested in DOC comparisons among treatments, which were all preserved with the same method. We assumed that treatment type would not affect DOC response to freezing, and chose freezing as a preservation method for the samples in this study.

To ensure that the Bitran bags were not leaching compounds that could influence

DOC concentration and quality, we developed a Bitran bag vs. quartz flask comparison with the bacteria + light experiment. For the light treatment, filtered and inoculated replicates were incubated in quartz flasks along with the replicates in the Bitran bags, and their DOC concentration and quality metrics were compared. SUVA254 declined an average of 11.2% in the quartz flasks and 12% in the Bitran bags. The magnitude of S275-

295 increased 12.9% and 14.0%, respectively, in the flasks and bags, and DOC concentration increased 0.2% and 0.06% in the flasks and bags, respectively. This test suggested that the Bitran bags were not leaching chromophoric compounds that affected the outcome of the light experiments.

For the soil experiment, soil samples were collected from below 2-3 cm depth near Lake SS2 and sealed field-moist in Whirl-Pak bags. Deionized (DI) water was added to the soil in an approximately 1:7 soil to water ratio to extract DOM in the soil

(Jones & Willett 2006). The solution was mixed to break up clumps of soil, and stored in the dark at 22°C for seven days. After this time, the soil water solution was filtered through 0.7 µm GF/F filters, the filtrate was added to 250-ml HDPE bottles, and the pH of the filtrate was measured with a Hydrolab ® DataSonde 5a (OTT Hydromet,

Loveland, Colorado). For the dust addition treatments, 2.5 mg dust was added to 250 ml of the soil water solution. Triplicate dust and no-dust treatments were incubated in the dark for seven days at 22°C. Water was collected for DOC analysis as described above.

Dust collection

We also determined the elemental composition and flux of dust to two of the study lakes. To collect aeolian-transported dust for elemental analysis, we set up Big

Spring Number Eight (BSNE) dust samplers during the summer of 2015 at high elevation points alongside Lake SS2 and Lake SS901. The BSNE samplers can trap up to 95% of airborne dust, regardless of wind velocity or direction. Sampling containers were mounted at heights of 40 cm, 65 cm, and 90 cm, and the samplers rotated so as to face into the wind for maximum dust collection. The BSNE dust samplers were left to collect dust for seven days. We also used a fine-bristled brush to collect dust directly from the outwash floodplain. To determine Fe and P content that could be leached into the lakes during a one-month period from this dust, the dust samples were soaked in DI water (pH

5.6) for one month, and filtered through 0.7 µm GF/F filters to remove residual sediment.

A subset of floodplain dust samples was also acidified with 0.02 N nitric acid to pH 3 for estimation of total elemental composition. Elemental analysis of the filtrates was conducted with an ICP/MS. In July 2016, we set up marble dust collectors, simple traps that collect dust in the vertical dimension, to allow us to make estimates of how much dust enters the lakes per unit area over time. The traps were set up for fourteen days on the shores of Lake SS2 and Lake SS901. At the end of the fourteen days, the Lake SS901 collector was found dismantled, presumably by wildlife, and so estimates of dust input were only calculated for Lake SS2. To measure the amount of dust collected in the

marble trap, the marbles were rinsed with DI water, which was then filtered through pre- weighed 0.7 µm GF/F filters. The filters were dried and re-weighed, and the total dust mass was divided by the area of the dust sampler and exposure time to calculate dust flux.

DOC analysis

Upon return to the University of Maine, after being frozen for approximately eight weeks, the experimental water samples were analyzed for DOC concentration

([DOC], in mg L-1) using a Shimadzu TOC-5000 total organic carbon (TOC) analyzer by high-temperature catalytic oxidation. Spectral absorbance of each sample was measured on a Varian Cary 300UV spectrophotometer in 1 cm quartz cells. Absorption coefficients from 220 nm to 700 nm were calculated using the equation:

2.303퐴(휆) 푎(휆) = 퐿

In this equation, a represents the Napierian absorption coefficient (m-1) at wavelength λ,

A is absorbance value at wavelength λ, and L is the cuvette path length, 0.01 m. SUVA254 was calculated following the methods of Helms et al. (2008). Soil water filtrate samples were diluted 1:10 for absorbance measurements.

Data analysis

Two-way ANOVA was used to examine the effects of dust input and light exposure or bacterial activity and light exposure on [DOC], SUVA254, and S275-295 for lake water samples. One-way ANOVA was used to analyze the impact of dust addition on soil water samples incubated in the dark. The significance level chosen was p < 0.05.

Data were square root or log transformed when necessary to better meet assumptions of equal variance and normality. S, the spectral slope coefficient (a measure of the decline in

absorption with wavelength), was plotted against a*375, the absorbance at 375 nm normalized to [DOC] (a qualitative measure of DOM color), as in Anderson & Stedmon

(2007) for interpretation of experimental DOC degradation. Analyses were performed in

R 3.3.1 (R Development Core Team 2016).

Results

The results showed that dust addition, bacterial activity, or natural sunlight did not cause losses in [DOC] in any of the experiments. DOC quality metrics SUVA254, S275-295, a*375, and S were also relatively unaffected by dust addition (with the exception of SS2) and bacterial activity, but experienced unexpectedly large effects from exposure to natural sunlight. These outcomes will be described below by summarizing effects of the treatment (dust or bacterial activity), light exposure, and the interaction of the treatment and light exposure on [DOC], SUVA254, and S275-295 for each experiment. Percent changes and p-values for treatments are given in Tables 2.2 and 2.3, respectively, and Figures 2.4-

2.6 illustrate these results graphically. Finally, we describe shifts in S vs. a*375 to aid interpretation of DOC degradation in both the dust + light and bacterial activity + light experiments, illustrated in Figure 2.7.

Table 2.2. Mean percent change in DOC concentration and quality metrics of for dust + light and bacteria + light experiments. The dust addition experiment for soil water was incubated in dark conditions only.

[DOC] (% change) SUVA254 (% change) S275-295 (% change)

Soil Soil Soil SS2 SS1381 SS901 SS2 SS1381 SS901 SS2 SS1381 SS901 water water water Dark + -0.08 -0.91 4.90 -25.88 -14.09 0.44 -5.66 36.21 -20.56 -0.38 -4.08 -3.53 no dust

Dark + 0.98 0.39 11.96 -21.59 -0.65 -0.15 -12.25 29.73 -1.53 -0.27 -4.82 -1.66

dust

light

+ Light + 6.47 11.17 33.66 - -20.19 -14.97 -36.32 - -20.34 -6.51 -23.70 -

no dust Dust Light + 22 5.38 12.04 41.31 - -6.29 -19.16 -37.76 - -0.81 -7.70 -28.02 -

dust Dark + 2.17 1.19 5.07 - -1.28 -2.90 -9.13 - -1.91 -1.02 -4.96 -

filtered

Dark +

1.81 0.69 10.19 - -1.91 0.34 -19.63 - -1.92 1.00 -1.59 - light

inoculated

Light + 0.13 2.59 9.79 - -10.73 -11.30 -25.05 - -11.49 -12.71 -19.86 -

filtered Bacteria Light + -0.01 1.75 6.17 - -13.07 -9.46 -16.68 - -14.31 -9.55 -11.61 - inoculated

Table 2.3. p-values for dust addition-light exposure and bacterial activity-light exposure experiments. Bolded values represent a significance level < 0.05. Nonsignificant (p ≥ 0.10) interaction terms were removed from statistical models. The dust addition experiment for soil water was incubated in dark conditions only.

[DOC] SUVA254 S275-295 Soil Soil Soil SS2 SS1381 SS901 SS2 SS1381 SS901 SS2 SS1381 SS901 water water water

Treatment 0.985 0.313 0.110 0.230 <0.001 0.329 0.240 0.237 <0.001 0.878 0.111 0.146

+ +

Light <0.001 <0.001 <0.001 - <0.001 <0.001 <0.001 - 0.809 0.045 <0.001 - light Dust Treatment 0.076 - - - - 0.017 ------× light

Treatment 0.670 0.177 0.887 - 0.026 0.009 0.889 - 0.216 0.080 0.122 -

+ +

Light 0.008 0.026 0.401 - <0.001 <0.001 0.254 - <0.001 <0.001 0.005 - light

Bacteria Treatment ------× light

Figure 2.4. Results of Lake SS2 soil water exposure to dust addition on SUVA254, S275- 295, and [DOC], represented by percent change (n = 3).

Figure 2.5. Results of SS2, SS1381, and SS901 lake water exposure to dust addition and light on SUVA254, S275-295, and [DOC], represented by percent change (n = 3).

Figure 2.6. Results of SS2, SS1381, and SS901 lake water exposure to bacterial activity and light on SUVA254, S275-295, and [DOC], represented by percent change (n = 3). F = filtered treatments, I = inoculated treatments.

Figure 2.7. S vs. a*375 for dust + light (a-c) and bacteria + light (d-f) experiments (n = 3).

Soil water experiment

The pH of the soil water filtrate was 6.38. Dust had no effect on any of the response metrics in the soil water experiment. [DOC] declined 26% in the no-dust treatment and 22% in the dust treatment, a difference which was not significant (p =

0.230; Figure 2.4; Tables 2.2 & 2.3). SUVA254 increased 36% in the no-dust treatment and 30% in the dust treatment (p = 0.237 dust-no dust treatment difference). The magnitude of S275-295 increased 4% in no-dust treatments and 2% in dust treatments (p =

0.146).

Dust addition + light exposure experiment

Over the course of the dust + light experiment, [DOC] increased in light treatments for all lakes relative to initial [DOC] (6-38%, 1.1-3.8 mg L-1; p < 0.001), while dust addition had no effect on [DOC] for any of the lakes (p = 0.985 SS2; p = 0.313

SS1381; p = 0.110 SS901; Figure 2.5; Tables 2.2 & 2.3). There were no interaction

effects between dust and light. The increases in [DOC] were all outside the quantification limit of the TOC analyzer (0.1 mg L-1).

SUVA254 decreased more in light treatments (-6 to -20%) than in dark treatments

(-0.7 to -14%) for all three lakes (p < 0.001). The decline in SUVA254 was not different between dust and no-dust treatments, except for Lake SS2, in which SUVA254 declined more in the no-dust treatments than in the dust treatments, regardless of light (17% vs.

3%, respectively; p < 0.001). Interaction effects between dust and light exposure occurred in the Lake SS1381 treatments, in which SUVA254 declined more in light-exposed dust treatments than in dark + dust treatments (p = 0.017).

The magnitude of S275-295 changed little in dark treatments for Lake SS1381 and

Lake SS901, and increased (7 to 28%, respectively) in light treatments for these lakes (p

= 0.045 SS1381; p < 0.001 SS901). Light did not affect the magnitude of S275-295 response for Lake SS2, but dust addition did—the magnitude of S275-295 increased about

1% in the dust treatments, compared to a 20% increase in the no-dust treatments (p <

0.001). Dust addition had no effect on S275-295 for Lake SS1381 or Lake SS901 (p = 0.878

SS1381; p = 0.111 SS901). Interactions between dust addition and light exposure had no effect on S275-295.

For Lake SS2, S values were higher and a*375 values were lower in no-dust treatments than in dust treatments, while light exposure had no effect (Figure 2.7a). Light treatments resulted in lower a*375 and S values for Lake SS1381, while dust had no effect

(Figure 2.7b). Light treatments resulted in lower a*375 and higher S values for Lake

SS901, while dust had no effect (Figure 2.7c).

Bacterial activity + light exposure experiment

[DOC] response to light exposure was variable among the lakes. [DOC] increased more in dark compared to light treatments for Lake SS2 (0.6 mg L-1 vs. 0 mg L-

1; p = 0.008), while it increased in light treatments relative to dark treatments for lake

SS1381 (0.7 mg L-1 vs. 0.3 mg L-1; p = 0.026), and was not influenced by light for Lake

SS901 (p = 0.401; Figure 2.6; Tables 2.2 & 2.3). There were no differences in [DOC] between filtered and inoculated treatments (p = 0.670 SS2; p = 0.177 SS1381; p = 0.887

SS901), and no interaction effects between bacterial activity and light exposure.

SUVA254 declined slightly or did not change in dark treatments for Lake SS2 and

Lake SS1381, while it declined in light treatments (10-12%; p < 0.001 for both lakes), and was not affected by light for Lake SS901 (p = 0.254). The effects of bacterial activity were variable among the lakes. SUVA254 in inoculated treatments was lower than in filtered treatments for Lake SS2 (p = 0.026), higher for Lake SS1381 (p = 0.009), and experienced no change for Lake SS901 (p = 0.889). There were no interaction effects between bacterial activity and light exposure.

For all lakes, light exposure caused the magnitude of S275-295 to increase 10-20% compared to dark treatments, in which it increased slightly or did not change (p < 0.001

SS2 and SS1381, p = 0.005 SS901). Bacterial activity had no effect on S275-295 (p = 0.216

SS2; p = 0.080 SS1381; p = 0.122 SS901). There were no interaction effects between light and bacterial activity on S275-295.

For Lakes SS2 and SS1381, light treatments had lower a*375 values than dark treatments (Figures 2.7d and 2.7e), while a*375 showed little difference between light and dark treatments for Lake SS901 (Figure 2.7f). S values were greater in light treatments

than dark treatments for SS1381 (Figure 2.7e), while showing no difference for Lakes

SS2 and SS901 (Figures 2.7d and 2.7f). Differences between inoculated and filtered treatments were small, except for SS2, in which light + inoculated treatments had higher

S values than light + filtered treatments (Figure 2.7d).

See Appendix 1 for average [DOC], SUVA254, and S275-295 data for the dust + light and bacteria + light experiments.

Dust input rates and elemental analysis

The areal rate of dust deposition near Lake SS2 was 0.931 g m-2 d-1. The Fe content was 1.8 mg kg-1 from dust collected near Lake SS2 and 3.5 mg kg-1 from dust collected near Lake SS901. Dust collected from the glacial outwash floodplain and soaked in DI water had an Fe content of 3 mg kg-1, while the Fe content of dust soaked at pH 3 was 90 mg kg-1. The P content of dust collected near Lake SS2 was 3 mg kg-1 and 6 mg kg-1 for dust collected near Lake SS901, while the P content of floodplain dust soaked in DI water was 3 mg kg-1 and 80 mg kg-1 for floodplain dust soaked at pH 3.

Discussion

Overall, we found that none of the experimental mechanisms tested (dust addition, bacterial activity, or solar radiation) were responsible for declines in [DOC] in experimental lake water. However, exposure to sunlight caused large changes in DOC quality, including declines in SUVA254 and increases in the magnitude of the spectral slope, S275-295. These results indicate that exposure to natural sunlight caused photobleaching of DOC in lake water, reducing aromaticity and degrading HMW DOC into LMW DOC. Because the study lakes have limited hydrological connectivity and low chromophoric DOM (CDOM) compared to boreal systems, it might be expected that

prior sunlight exposure would have reduced their susceptibility to photochemical transformation over the course of the experiment. However, the shifts in DOC quality during the experiment indicate that the DOC of the study lakes is highly photoreactive.

We discuss the implications of this finding below, along with interpretations of the effects of each of the experimental treatments on DOC quantity and quality metrics.

Dust

We predicted that dust addition would cause a decline in [DOC] in the three experimental lakes. We found that dust addition caused no difference in [DOC] and that photoflocculation did not occur. We expected Fe in the dust to bind to the humic portion of DOC in the water column, and then flocculate and precipitate out of the surface water, effectively reducing [DOC]. Fe-humic DOC bonds may be of less consequence than we predicted—the DOC in the lakes near Kangerlussuaq has low color, and may not have high humic content as a result of low precipitation levels and limited catchment inputs.

Additionally, the dust addition may have added DOC, thereby offsetting any Fe-driven losses. Mladenov et al. (2011) hypothesized that remote high-latitude alpine lakes within the Saharan dust belt received inputs of CDOM from atmospheric dust. In the same study, polar lakes at lower elevations and outside the dust source did not reflect changes in DOM. However, many Kangerlussuaq lakes are proximal to the dusty floodplain.

Future analysis of the organic content of dust from this region could help clarify its potential for influencing lake water DOC.

As expected, DOC composition did not change in response to dust inputs in water from Lakes SS1381 and SS901. However, in water from Lake SS2, SUVA254 was higher and the magnitude of S275-295 was smaller in dust treatments than in no-dust treatments.

The magnitude of S275-295 is thought to be inversely proportional to molecular weight

(Helms et al. 2008), and SUVA254 can be used as an indicator of aromaticity. Therefore, the dust additions may have contributed DOC with a more terrestrial signature (greater aromaticity and higher MW). The lower a*375 and higher S values for Lake SS2 dark treatments of this experiment (Figure 2.7a-c) indicate that in 2015, the DOC in Lake SS2 was more processed and less fresh, with possibly greater autochthonous than allochthonous inputs than the DOC in Lakes SS1381 and SS901. For this reason, the dust inputs caused a greater response in DOC quality in Lake SS2 than the other two lakes, which were more influenced by sunlight exposure.

These interpretations are supported by results depicted in Figure 2.7a, which show that Lake SS2 dust treatments have higher a*375 and lower S values. Like SUVA254, larger a*375 values typically correspond to a greater humic content, because humic substances are enriched in aromatic C, whereas larger S values may represent extent of photochemical degradation (Anderson & Stedmon 2007). DOC quality in the dust treatments was fresher and less degraded, with a more terrestrial signature, while DOC quality in the no-dust treatments appeared more processed and degraded. Figures 2.7b and 2.7c illustrate the lack of dust effects on a*375 and S for Lake SS1381 and Lake

SS901, which are instead affected by sunlight exposure.

In the soil water experiment, no significant differences occurred between dust and no-dust treatments. However, in both treatments, SUVA254 increased while [DOC] declined, possibly indicating bacterial degradation—bacteria in the soil solution may have been preferentially consuming aromatic-poor DOC. Kalbitz et al. (2003) found that higher-aromaticity DOM in marsh and forest soils in Germany was less readily degraded

by bacteria than more labile DOM, although it should be noted that lability of terrigenous

DOC in Alaskan rivers has been estimated to be relatively high (Holmes et al. 2008).

Bacterial activity

Contrary to our hypothesis, the bacteria + light treatment did not cause a decline in [DOC] in the experiment, however, the experimental design used here may have strongly limited the potential magnitude of effects of bacteria on [DOC]. To remain consistent in the handling of both the filtered and inoculated treatments, we filtered lake water through a 0.2 µm filter to remove microbes and then to these “inoculated” treatments, we added 3 ml of 0.7 µm-filtered lake water to reintroduce bacteria, while leaving the “filtered” treatments un-inoculated. Thus, via dilution in the inoculated treatments, reduced bacterial biomass may have been unable to induce meaningful effects on [DOC] over the seven-day experiment.

Recognizing this limitation in our experimental design, we do note that other studies have also found a weaker effect of bacterial activity compared to photochemical degradation. Cory et al. (2014) found that in Arctic inland surface waters, the effects of bacterial C processing are overshadowed by photochemical degradation, which could explain lack of a decline in [DOC] in response to inoculation in our experiment.

Obernosterer & Benner (2004) also described competition between photochemical and bacterial degradation of DOC in a lake with a similar quantity of CDOM as the lakes in this study, in which a greater proportion of reactive DOC was susceptible to photodegradation rather than bacterial degradation. In a long-term study comprising hundreds of boreal lakes, wetlands, and rivers in Canada, Lapierre et al. (2013) showed that compared to photochemical degradation rates, bacterial degradation rates were too

small to be meaningful. Photochemical degradation increased with terrestrial inputs of

DOC, while rates of bacterial degradation remained constant. Osburn et al. (2001) showed that wetland (bog) DOC had far greater photochemical susceptibility than adjacent lake water DOC, implying the highly photoreactive nature of soil and wetland sources of DOM.

S275-295 was not affected in inoculated treatments in water from any of the lakes, contradicting our prediction that magnitude of S275-295 values would decline as a result of bacterial degradation, as in Helms et al. (2008). This lack of response could suggest that

S275-295 is more impacted by sunlight than by bacterial processing in these experiments, but the bacterial dilution in the experimental design and resulting low bacterial biomass may have restricted our ability to infer effects of bacterial degradation on DOC quality.

We surmise that the dilution effect may also have inhibited any meaningful change to

SUVA254 values.

Solar radiation

Of the three DOC removal pathways tested in this study, solar radiation caused the most notable responses in DOC concentration and quality. Sunlight exposure increased [DOC] in water from all three lakes in the dust + light experiment. In the bacteria + light experiment, [DOC] changes were small and variable among the lakes. We were surprised that these experiments did not result in loss of [DOC], as photooxidation of DOC into CO2 is widely accepted as a major driver of carbon processing in Arctic surface waters (Cory et al. 2014). One possible factor contributing to [DOC] increases in light treatments of the dust + light experiment may be primary production by picophytoplankton that passed through the 0.7 µm GF/F filters. Chlorophytes and

cyanobacteria are two important phytoplankton classes that contribute to picophytoplankton (Stockner & Antia 1986). While densities of cyanobacteria were low or nonexistent, chlorophytes were present in water samples collected in July 2013 from each of the three study lakes, at densities of 11 cells ml-1, 126 cells ml-1, and 98 cells ml-1 for Lakes SS2, SS1381, and SS901, respectively (Saros unpublished data). It is interesting to note that the smallest experimental [DOC] increase (1.3 mg L-1) occurred in water from Lake SS2, which had the lowest density of chlorophytes in July 2013 samples, while [DOC] increases were greater in water from Lakes SS1381 and SS901

(3.2 and 3.0 mg L-1, respectively), which had higher chlorophyte densities. However, without primary productivity measurements for picophytoplankton in the experimental water, we are unable to estimate the potential contribution of picophytoplankton to the observed [DOC] increase in the dust + light experiment. Even assuming high growth rates throughout the experiment, and a large proportion of primary production excreted as

DOC, it is unlikely that primary production by chlorophytes could have solely been responsible for the magnitude of increase in [DOC] measured in these experiments.

Measurements of picophytoplankton primary production could better elucidate the role of these small autotrophs in carbon cycling in Kangerlussuaq surface waters, but understanding of the primary mechanism responsible for [DOC] increases in the light conditions of these experiments will require further research.

Sunlight exposure caused SUVA254 to decline and the magnitude of S275-295 to increase for lakes in both experiments, as expected (except Lake SS2 in the dust + light experiment and Lake SS901 in the bacteria + light experiment, which showed no change). It has been demonstrated that UV exposure breaks down HMW DOM and

reduces aromaticity (Osburn et al. 2001), which is supported by these results, as well as the reductions in a*375 and increases in S values in light treatments. Our findings are consistent with values for these parameters reported in prior surveys of lakes across West

Greenland, which include DOM quality metrics indicating high DOM susceptibility to photochemical transformation (Anderson & Stedmon 2007; Osburn et al. 2017). These results are especially of interest because they show that the DOC in these comparatively

CDOM-poor lakes remains susceptible to photochemical changes in quality under constant sunlight exposure during the polar summer.

Hydrological connectivity is weak within the Kangerlussuaq landscape and precipitation is low. The presence of continuous permafrost also currently limits connectivity between soils and lakes (Anderson et al. 2017), so we might expect terrestrial DOC loading to these lakes to be minimal. In addition, the low SUVA254 values

(< 2.0 L mg C-1 m-1) of these lakes might suggest limited susceptibility to photochemical transformation. However, mean [DOC] in the Kangerlussuaq lakes is far greater than the global mean and rivals that of darkly-colored boreal lakes (Sobek et al. 2007). The results of our short-term experiments, however, indicate that the quality of DOC in

Kangerlussuaq lakes remains highly responsive to sunlight exposure. This key finding is relevant in the context of the rapid, nonlinear shift in climate presently occurring in the

Arctic—if low-color DOC previously exposed to abundant sunlight remains responsive to photochemical transformation, then future changes in light availability, terrestrial-aquatic linkages, and temperature will likely elicit continued and substantial shifts in the DOC quality of these lakes. While cloud cover may be increasing in the Arctic (Eastman &

Warren 2013), and there has been no clear trend in ice-off dates in limited data sets for

this region (Saros et al. 2015), ice-free duration is lengthening for many lakes in the

Arctic in response to climate warming (Magnuson et al. 2000). If ice-free duration were to increase for Kangerlussuaq lakes, duration of sunlight exposure would also increase, and the seasonal pulse of meltwater inputs may shift temporally and in magnitude

(Anderson et al. 2017). This shift could influence the composition of the DOC pool and its susceptibility to photochemical transformation. Furthermore, continued warming may enhance permafrost thaw, which could introduce fresher, more labile DOC to lake surface waters. The proportion of substrate available for bacterial activity may increase, while photodegradation and photobleaching of this soil-derived DOM could further stimulate bacterial processing.

Future work

While our experiments confirm effects of sunlight exposure on lake water DOC quality suggested by Anderson & Stedmon (2007), they did not elucidate mechanisms responsible for the widespread and coherent decline in DOC concentration in lakes of the

Kangerlussuaq region. Future studies should explore external factors that may be responsible for this pattern. One such mechanism may be increased delivery of atmospheric sulfur to lakes, which is supported by increasing sulfate levels measured in the study lakes (Saros et al. 2015). In North America, declines in atmospheric sulfur deposition have been linked to increasing DOC concentrations in lakes (Strock et al.

2016). Perhaps increased marine primary production in the Arctic since 2003 has enhanced sulfate deposition to lakes via oxidation of dimethyl sulfide (DMS; Arrigo et al.

2008). Additionally, more frequent volcanic eruptions, like that of the Icelandic volcano

Bárðarbunga that released massive clouds of sulfur dioxide in 2014, could be responsible

for the delivery of sulfate to lakes in the Kangerlussuaq region (Gauthier et al. 2016).

Shifts in terrestrial C inputs may also play a role in reduced DOC. With climate warming, herbivory pressure by large mammals is likely to change, with implications for plant community composition and terrestrial DOC pools (Anderson et al. 2017). While it is currently unclear whether hydrological linkages between terrestrial and aquatic systems will strengthen or weaken in the future, the DOC pools of lakes will undoubtedly respond to terrestrial processes, and further study is warranted.

Conclusions

The results described herein represent the complexities involved in deciphering drivers of changes in DOC of West Greenland surface waters. This study demonstrates how several mechanisms can contribute to lake water DOC composition, and that these mechanisms may work in opposite directions and at different magnitudes depending on lake properties and initial characterization of DOC in the lakes. While our experimental examination of potential pathways of DOC changes did not illuminate the cause for widespread DOC concentration decline observed in lakes of the Kangerlussuaq area in recent years, we were able to advance our understanding of the characterization of DOC in this region and how it can fluctuate in response to sunlight exposure, dust inputs, or bacterial activity. The quality of DOC in the study lakes was most susceptible to photochemical transformation, despite the low amounts of CDOM and annual periods of continuous exposure to sunlight during the polar summer (Osburn et al. 2009). These results suggest the potential for continued shifts in DOC quality with rapid and nonlinear climate change in this region, as well as the need for further study of mechanisms

external to lakes that may be responsible for widespread and coherent regional decline in

DOC concentration.

CHAPTER 3

COHERENT CLIMATE-DRIVEN CHANGES IN DISSOLVED ORGANIC

CARBON AND WATER CLARITY IN ARCTIC LAKES OF WEST

GREENLAND

Abstract

To assess climate-mediated terrestrial-aquatic linkages in Arctic lakes and potential impacts on light attenuation and carbon cycling, we evaluated coherence of lake responses in two regions of west Greenland with differing climate patterns. We selected four lakes in a warmer/drier (WD) region to compare with four lakes from a cooler/wetter

(CW) region proximal to the Greenland Ice Sheet (GrIS). In June from 2013-2018, we measured epilimnetic water temperature, 1% depth of photosynthetically active radiation

(PAR), dissolved organic carbon (DOC), specific ultraviolet absorbance (SUVA254),

DOC-normalized absorbance at 380 nm (a*380), and chlorophyll a (Chl a). Coherence of lake variables within and between the two regions was high, and the interannual coherence of 1% PAR and DOC was particularly high for lakes within the WD region.

This coherence suggests strong forcing of Arctic lake features by a large-scale driver, likely climate. Redundancy analysis showed that monthly average precipitation (MAP), winter NAO index (NAOW), spring average air temperature (SAT), and spring average precipitation (SAP) influenced the lake variables (p = 0.003, adj. R2 = 0.58). In particular,

MAP contributed to increases in soil-derived DOC quality metrics and Chl a, and decreased 1% PAR. Interannual changes in lake responses to climate drivers were more apparent in the WD region than the CW region. The high coherence of interannual lake responses within and between regions, associated with climate trends, suggests that with

ongoing rapid climate change in the Arctic, there will be widespread impacts on key lake responses important for light attenuation and carbon cycling.

Introduction

While climate effects on lake ecosystems at mid-latitudes have been extensively investigated (Magnuson et al. 1990; Benson et al. 2000; and others), less is understood about climate forcing of lake responses in the Arctic. The remote location of many Arctic lakes results in relatively lower levels of human disturbance than at mid-latitudes (i.e. land-use change, atmospheric deposition; Smol et al. 2005). With ongoing rapid and nonlinear climate change occurring in the Arctic (Mayewski et al. 2014), it is likely that lakes here are highly responsive to climate forcing and that responses are not confounded by disturbances more common at mid-latitudes. Therefore, Arctic limnology presents a strong model for exploring links between climate drivers and lake responses. Exploration of these links is particularly important, as Arctic lakes are sites of intense carbon cycling and are involved in regulation of carbon dioxide and methane to the global C cycle and C burial in their sediments (Cole et al. 2007).

The coherence of lake responses, defined as the shared variance between time series, can provide information about the effects of climate change and variability on lake characteristics (Magnuson et al. 2005). Large-scale climatic processes (e.g. North

Atlantic Oscillation; NAO) and weather patterns drive coherent responses in lakes, and these climate signals are modified by external processes and internal lake characteristics

(Blenckner 2005; Magnuson et al. 2005). Widespread spatial coherence can occur when climatic “footprints” span large areas, with variability in weather patterns influencing regional coherence patterns on scales of tens to hundreds of kilometers (Magnuson et al.

2005). These climatic footprints contribute to the “energy-mass flux framework” of lakes, in which climate variability regulates lake structure and function via direct and indirect inputs of energy (E) in the form of solar irradiance, heat, and wind, and mass (m) as precipitation, particles, and atmospheric solutes (Leavitt et al. 2009). While E inputs contribute to seasonal and annual coherence (Leavitt et al. 2009), morphometric features can cause differences in individual lake responses to climatic signals, creating a

“filtering” effect that may alter spatial coherence (Magnuson et al. 1990). Changes in climate patterns that impact m influx, such as increased springtime rainfall, may also reduce coherence among lakes (Pham et al. 2008). For northern temperate lakes at mid- latitudes, physical variables typically have the greatest coherence (r = 0.45), followed by chemical (r = 0.20) and biological (r = 0.10) variables (Kratz et al. 1998; Magnuson et al.

2004).

Climatic effects on coherence of lake responses have been broadly explored in mid-latitude lakes in northern temperate regions (Magnuson et al. 1990; Kratz et al. 1997;

Baines et al. 2000; Baron & Caine 2000; Benson et al. 2000; George et al. 2000; Webster et al. 2000). Furthermore, much of this work was focused on lake chains, by testing the importance of landscape position in the hydrological flow regime on interannual variability (Kratz et al. 1997; Soranno et al. 1999; Kling et al. 2000; Webster et al. 2000).

Lakes with low hydrological connectivity in these regions tend to exhibit reduced coherence (Webster et al. 2000; Magnuson et al. 2005). There is currently a lack of understanding about the effects of climate on interannual variability of Arctic lakes, particularly those fed primarily by snowmelt, with limited hydrological connectivity.

The lakes of west Greenland provide a rich testing ground for these questions, as the area contains approximately 20,000 lakes, many of which are characterized by low hydrological connectivity. Deciphering how climate factors affect the coherence of lake responses in this region will be important for understanding how rapid changes in climate may contribute to shifts in Arctic lake ecology and carbon cycling. Precipitation and temperature anomalies have increased in west Greenland in recent years, and these climatic shifts likely alter terrestrial subsidies to lakes and internal lake processing

(Anderson et al. 2017). As Arctic permafrost thaws, the role of soil-derived dissolved organic carbon (DOC) for C cycling in Arctic surface fresh waters may intensify

(Anderson et al. 2017). DOC quality, along with chlorophyll a (Chl a), controls the depth of 1% photosynthetically active radiation (PAR) in Arctic lakes (Saros et al. 2016). The depth of 1% PAR has implications for primary production in Arctic lakes, as different phytoplankton taxa have different requirements for light (Malik & Saros 2016).

Photomineralization and bacterial respiration of soil-derived DOC could amplify release of CO2 from Arctic lakes (Cory et al. 2014). In addition, increases in soil-derived DOC may contribute to a less bioavailable, but more photochemically-labile (Lapierre & del

Giorgio 2014), dissolved organic matter (DOM) pool, which could affect microbial community structure and in-lake C cycling and storage (SanClements et al. 2012).

Therefore, climate-mediated shifts in terrestrial-aquatic linkages will likely elicit responses in key lake variables linked to reduced light attenuation and altered C cycling in Arctic lakes.

To better understand how climate forcing drives these lake responses, we assessed relationships among climate variables and interannual changes in physical, chemical, and

biological lake variables: water temperature, 1% PAR, DOC concentration and quality, and Chl a. We selected four lakes from a warmer, drier (WD) region to compare to four lakes from a cooler, wetter (CW) region. Lakes were sampled in early summer every year from 2013 to 2018. Our objectives were to a) evaluate the degree of coherence of these lake variables within and between lakes in the WD and CW regions, and b) pair meteorological records with lake survey data to explore possible climatic signals that may have driven recent changes in these lake responses.

Methods

Study area

We conducted this study in the Kangerlussuaq region of west Greenland, which is located north of the Arctic Circle between 66-68°N and 50-53°W in the ice-free margin between the Greenland Ice Sheet and the coast. This low Arctic, continental region contains approximately 20,000 lakes (Anderson et al. 2009), most of which are oligotrophic and chemically dilute (Anderson et al. 2001). Ice-off ranges from late-May to late-June, followed by rapid thermal stratification (Broderson & Anderson 2000). The continuous permafrost soil active layer ranges from 20-90 cm (Johansson et al. 2015), and the mean summer temperature and annual precipitation are 10.2°C and 173 mm yr-1.

Lakes in this region have long residence times (Leng & Anderson 2003) and limited hydrological connectivity (Anderson & Stedmon 2007), with run-off from the active layer (Leng & Anderson 2003) and precipitation as the primary water inputs, and minimal surface inflow or outflow (Johansson et al. 2015).

We evaluated meteorological data and a suite of physical, chemical, and biological variables from lakes in two distinct regions of the broader Kangerlussuaq area

located ~40 km apart (Figure 3.1). While temperature and precipitation in both regions fall within a typical range for Arctic systems, the Kangerlussuaq area can be divided into a “warmer, dryer (WD)” region centrally located between the GrIS and the coast, and a

“cooler, wetter (CW)” region adjacent to the GrIS. As in the WD region, lakes in the CW region are fed by precipitation inputs, and not by the GrIS. The woody shrubs Betula and

Salix are the dominant vegetation in the WD region, while the CW region is characterized by heath-graminoid vegetation. The landscape of the WD region is generally flatter with some rolling hills, while catchments in the CW region have steeper slopes and areas of permafrost slumping.

Figure 3.1. Map of study site. White circles denote study lakes in WD and CW regions.

Lake variables

After ice-out, in mid-late June (or early July) from 2013-2018, we sampled a suite of variables from four lakes in each region (SS2, SS85, SS1381, and SS1590 in WD; SS901,

SS904, SS905, and SS906 in CW; Table 3.1). Water temperature was measured at 2 m

depth with a Hydrolab® DataSonde 5a (OTT Hydromet, Loveland, Colorado) from

2013-2015 and a YSI EXO2 multiparameter sonde (YSI Incorporated, Yellow Springs,

Ohio) from 2016-2018. The depth of 1% PAR was calculated using PAR profiles from a

BIC submersible profiling radiometer coupled with a deck radiometer (Biospherical

Instruments, San Diego, California). In 2013-2014, PAR profiles were not collected for

SS904 and SS905, so 1% PAR was estimated from Secchi depth using the equation:

1% PAR = 1.472(Secchi depth) + 0.891

This relationship between Secchi depth and 1% PAR was based on measurements of both parameters across 22 lakes in the Kangerlussuaq region in 2013 (R2 = 0.79; Saros et al.

2016). Epilimnetic water samples were collected with a van Dorn bottle, and were then filtered through pre-combusted 0.7 µm Whatman GF/F filters. For 2013 and 2014 measurements, DOC was analyzed using an OI Analytical Aurora 1030D total organic carbon (TOC) analyzer (Xylem Inc., Rye Brook, NY) using wet chemical oxidation. The procedure was the same for the 2015-2018 samples, except samples were analyzed with a

Shimadzu TOC-5000 analyzer (Shimadzu Corporation, Kyoto, Japan) by high- temperature catalytic oxidation. We compared measurements from 9 samples collected in

2014 using both methods, and the mean variability between the two TOC analyzers was

2.0%, or 0.11 mg C L-1. Spectral absorbance was measured on a Varian Cary-300

Ultraviolet-Visible spectrophotometer (Agilent Technologies, Santa Clara, California) in

1 cm quartz cells, and absorption coefficients at 254 nm and 380 nm were calculated using the equation:

2.303퐴(휆) 푎(휆) = 퐿

In this equation, a is the Napierian absorption coefficient (m-1) at wavelength λ, A is the

Milli-Q-corrected absorbance at wavelength λ, and L is the cuvette pathlength, 0.01 m.

The specific ultraviolet absorption (SUVA254), which is the DOC absorbance at 254 nm normalized to DOC concentration, describes degree of DOC aromaticity (Weishaar et al.

2003), and can be used to infer changes in soil-derived humics. Absorption at 380 nm normalized to DOC (a*380), is used as a reference for absorption by chromophoric DOM

(CDOM; Markager & Vincent 2000). Chl a was determined by filtration of water samples through 0.7 µm Whatman GF/F filters, which were frozen until analysis.

Epilimnion, metalimnion, and hypolimnion Chl a values were averaged to determine mean water column Chl a. Chl a analyses were performed according to standard methods

(American Public Health Association 2000) within 3 weeks of filtration. Chl a was extracted using 90% acetone, clarified via centrifugation, and then concentration was measured on a Varian Cary-300 Ultraviolet-Visible spectrophotometer (Agilent

Technologies, Santa Clara, California).

Table 3.1. Characteristics of study lakes. CEI = Climate exposure index, the ratio of lake surface area to mean depth (Magnuson et al. 1990); C:L = ratio of catchment area to lake area; ‘a.s.l’ = above sea level.

Region Lake Latitude Longitude Maximum Mean Surface CEI C:L Elevation depth (m) depth (m) area (m) (m a.s.l.) (km2)

WD SS2 66.99 -50.96 12 6 0.368 68,478 9.9 187 SS85 66.98 -51.06 11 4 0.246 74,729 6.6 191 SS1381 67.02 -51.12 19 6 0.215 37,153 4.9 171 SS1590 67.01 -50.98 18 5 0.243 50,882 53.3 199 CW SS901 67.13 -50.24 15 8 0.106 14,038 4.4 399 SS904 67.16 -50.28 18 8 0.125 15,256 5.3 405 SS905 67.16 -50.29 20 7 0.132 18,182 5.8 414 SS906 67.12 -50.25 18 7 0.085 11,435 7.3 406

Climate variables

Climate data obtained from weather stations located near lakes in the WD and

CW regions from 2013 to 2017 were used as a proxy for climate trends occurring at lakes in those respective regions. For the WD region, we obtained data from the Danish

Meteorological Institute (DMI) weather station at the Kangerlussuaq airport (Cappelen in press), and we used data from the Greenland Analogue Surface Project (GRASP)

KAN_B weather station for the CW region (Johansson et al. 2015; Johansson et al. in press). Complete weather station data were not available for 2018, so we limited our analysis of climate variables from 2013 to 2017. Climate variables considered for analysis were mean annual and seasonal precipitation, seasonal temperature, wind speed, humidity, air pressure, growing degree days, and the NAO winter index. Collinear variables with a variance inflation factor (VIF) > 5 were identified and excluded. The resulting variables included in analysis were monthly average precipitation over the water year (MAP), monthly average spring precipitation (MAMJ; SAP), average spring air temperature (MAMJ; SAT), and the NAO winter index (Hurrell DJFM index, PC-based;

(NAOW).

Statistical analyses

Prior to all coherence calculations and multivariate analyses, we Z-transformed climate and lake data from the WD and CW regions to standardize differences in values between the two regions. For example, a higher “Z-score” indicates that for a particular year, the absolute value of a response metric was greater than the 6-year mean. The data were not normally distributed, so we used Spearman’s rank correlation coefficient (ρ), a nonparametric approach, as a measure of coherence (Patoine & Leavitt 2006; Pham et al.

2009). With only six years of data, we did not perform significance tests, but instead looked at relative differences in coherence among variables. For each lake variable, we calculated within- and between-region mean coherence.

We used redundancy analysis (RDA) to evaluate effects of the climate variables on lake variables in the WD and CW regions from 2013 to 2017. We matched lake responses with the most influential climate drivers as indicated by the RDA triplot. All variables were centered and standardized prior to analysis. Monte Carlo permutation tests with 999 permutations were used to evaluate the significance of the model and its terms.

To evaluate how climate drivers differentially affected coherence of lake responses across the WD and CW regions from year-to-year, we matched Z-transformed lake variables with Z-transformed climate variables from both regions and compared interannual trends in the climate drivers and lake responses between the regions. All analyses were performed in R Version 3.4.2.

Results

Climate and lake variables

From 2013-2017, both MAP and SAP were lower in the WD region than the CW region, while SAT was higher in the WD region (Figure 3.2). For the WD region, MAP ranged from 6.2-11.9 mm and SAP ranged from 3.4-6.8 mm, while MAP and SAP ranged from 20.8-34.8 mm and 11.5-32.8 mm, respectively, in the CW region (Table

3.2). In the WD region, SAT ranged from -4.9-2.2°C, while in the CW region, SAT ranged from -6.5-0.1°C. NAOW ranged from -1.68 to 1.87 from 2013-2017.

Table 3.2. Climate characteristics of the two study regions from 2013 to 2017.

Region Warmer, Drier Cooler, Wetter (WD) (CW) Mean Range Mean Range Monthly Average Precipitation (MAP, mm) 9.3 6.2-11.9 24.6 20.8-34.8 Spring Average Precipitation (SAP, mm) 5.7 3.4-6.8 21.0 11.5-32.8 Spring Average Air Temperature (SAT, °C) -1.9 -4.9-2.2 -3.7 -6.5-0.1

Figure 3.2. Climate trends in WD and CW regions from 2013-2017. a) Average monthly precipitation over the water year (MAP), b) average spring air temperature (SAT; MAMJ), c) average monthly spring precipitation (SAP; MAMJ), and d) NAO winter index (NAOW; DJFM).

From 2013-2018, water temperature at 2 m depth ranged from 8.4-15.5°C in the

WD region and 6.7-13.3°C in the CW region while the depth of 1% PAR was 8.6-13.1 m in the WD region and 10.8-14.3 m in the CW region (Figure 3.3). DOC was much higher in the WD region compared to the CW region, with ranges from 24-31 mg L-1 and 5.3-5.9

-1 -1 -1 mg L , respectively. SUVA254 and a*380 ranged from 2.8-3.1 L mg m and 0.10-0.17 L mg-1 m-1, respectively, in the WD region, and 3.5-4.1 L mg-1 m-1 and 0.24-0.38 L mg-1 m-

1, respectively, in the CW region. Mean water column Chl a was generally greater in the

WD region than in the CW region, ranging from 1.2-2.7 µg L-1 compared to 1.1-1.8 µg L-

Figure 3.3. Absolute values of lake responses in WD and CW regions from 2013-2018. a) Water temperature, b) 1% PAR, c) DOC, d) SUVA254, e) a*380, and f) Chl a.

Within- and between-region lake variable coherence

Mean coherence of the lake variables was generally high within and between the

WD and CW regions (Figure 3.4). On average, the between-region mean coherence of the physical variables, water temperature and depth of 1% PAR, was highest (ρ = 0.57 and

0.69, respectively), while the chemical variables, DOC, SUVA254, and a*380, had similar between-region mean coherence as Chl a (ρ = 0.40, 0.63, 0.40, and 0.51, respectively).

Notably, mean coherence of DOC in the WD region was much higher than in the CW region, ρ = 0.87 compared to ρ = 0.10, respectively. Similarly, the mean coherence of 1%

PAR was also greater in the WD region (ρ = 0.90) than in the CW region (ρ = 0.55).

Within-region coherence of epilimnetic water temperature and Chl a was ρ = 0.74 and

0.48, respectively, for the WD region and ρ = 0.90 and 0.61 for the CW region. The within-region coherence of a*380 was similar for both regions, with ρ = 0.43 for the WD region and ρ = 0.44 for the CW region.

Figure 3.4. Mean temporal coherence of lake responses within and between WD and CW regions from 2013-2018.

Effects of climate on lake responses

Redundancy analysis of effects of climate drivers on lake responses was significant (p = 0.004), and the constrained variance was 80%, suggesting that the climate drivers selected explained lake responses reasonably well (Figure 3.5). The first RDA axis explained 67% of the variance (p = 0.004), while the second axis explained an additional 23% (p = 0.118). Across both regions, the RDA results suggest that SUVA254, a*380, and Chl a responded positively to increases in MAP, while 1% PAR was negatively associated with MAP. These variables correlate with RDA axis 1. Water temperature was closely related to NAOW, while DOC may be driven by SAT, as well as

SAP, to a lesser extent.

Figure 3.5. RDA triplot illustrating relationships between climate drivers (black arrows), lake responses (gray arrows), and regions (WD = black circles, CW = gray circles) from 2013-2017.

The effect of MAP on lake responses was more apparent in the WD region than the CW region (Figure 3.6). For example, in 2016, the MAP Z-score in the WD region decreased and 1% PAR increased, while a*380, SUVA254, and Chl a all decreased.

However, in the CW region, the MAP Z-score slightly increased, while 1% PAR increased, and a*380, SUVA254, and Chl a decreased. The lake response variables in the

CW region were likely affected by an additional driver(s) than MAP in 2016. In 2017, the association between MAP and lake responses was clearer in both regions, however. The

SAT Z-score pattern matched that of DOC in the WD region, while the effects of SAT on

DOC in the CW region were less apparent. Epilimnetic water temperature in the WD region closely resembled inverse SAP Z-scores, as well at the NAOW Z-scores, which were inversely related to each other in that region. Water temperature trends in the CW

region, however, did not match Z-score trends for SAP in that region or the NAOW. In addition, the Z-scores of SAP in the CW region were not related to those of the NAOW.

Overall, the effects of MAP, SAT, SAP, and NAOW elicited stronger lake responses in the WD region, while the CW region experienced relatively more muted responses to the climate drivers in the model.

Figure 3.6. Comparison of interannual variability of climate drivers and lake responses between WD and CW regions. Gray boxes represent climate drivers included in RDA: a) MAP, f) SAT, h) SAP, and j) NAOW, and white boxes represent lake responses: b) 1% PAR, c) a*380, d) SUVA254, e) Chl a, g) DOC, and i) water temperature. The boxes are stacked vertically according to the associations between climate drivers and lake responses identified by the RDA. For example, there was an inverse association between MAP and 1% PAR, so 1% PAR is vertically adjacent to MAP.

Discussion

Interannual coherence of the lake variables was generally high within and between the WD and CW regions, especially in the context of similar studies from mid- latitudes focused on lakes positioned within lake chains, as the WD and CW lakes do not sit within lake chains. Within the WD region, coherence of DOC and 1% PAR was particularly high. Increased MAP was associated with higher values of SUVA254, a*380, and Chl a, and shallower 1% PAR, while DOC was positively associated with SAT and

SAP, and epilimnetic water temperature was higher with a positive NAOW. Based on these results, regional comparisons of interannual trends in climate and lake variables revealed stronger climate forcing of lakes in the WD region than in the CW region. Lakes in the WD region have higher climate exposure indices (CEI; Figure 3.7; Magnuson et al.

1990) and catchment:lake area (C:L) ratios (Table 3.1; Kling et al. 2000), which may explain why these lakes are more sensitive to climate forcing. The associations of MAP with soil-derived DOC quality metrics, Chl a, and 1% PAR, with the link between SAT and DOC, suggest that temperature and precipitation variables influence terrestrial- aquatic linkages that directly affect in-lake C stocks and processing, as well as water transparency. This study demonstrates that Arctic lakes with limited hydrological connectivity exhibit high within- and between-region coherence of lake responses to climate forcing. Results suggest that soil-derived DOC metrics and water transparency in these lakes will be highly responsive to rapid climate change.

Figure 3.7. Climate exposure index (CEI) values of WD and CW lakes.

Lake variables within and between the WD and CW regions displayed high levels of coherence, even though the absolute values of lake variables differed considerably between the two regions. In the WD region, the interannual coherence among lakes for both 1% PAR (0.90) and DOC (0.87) was very high. For comparison in lake districts at mid-latitudes, which generally include lakes situated within lake chains, median coherence values tend to be greatest for physical lake variables (0.45), followed by chemical variables (0.20) and biological variables (0.10; Kratz et al. 1998). In a group of seven Wisconsin lakes, mean coherence of summer water temperature was 0.42, while mean coherence of Secchi depth and Chl a was 0.35 and 0.19, respectively (Magnuson et al. 1990). Benson et al. (2000) found that epilimnetic summer water temperature mean coherence between western Great Lakes regions ranged from 0.60 to 0.94, while within- region coherence ranged from 0.82 to 0.97. For physical and chemical variables, mean

coherence of five lakes in an English lake district was 0.81, while mean coherence of biological variables was 0.11 (George et al. 2000). As in our study, lakes in remote regions generally exhibit high coherence, with exceptions for some lake variables. In a study of eight alpine lakes in two basins in the Rocky Mountains, Baron & Caine (2000) found that coherence of between-region mean surface water temperature coherence was

0.34 and within-region coherence was 0.53 and 0.28 for the two basins. For chemical variables between the two regions, mean coherence was 0.06 to 0.49, and within the regions, coherence ranged from 0.00 to 0.70 and 0.38 to 0.76. In a series of surface- drainage lakes and streams in Arctic Alaska, mean coherence of chemical and biological variables ranged from 0.18 to 0.90, and the mean for all variables was 0.50 (Kling et al.

2000). The within- and between-region coherence of the lake variables in our study are generally similar to or greater than those found in these studies—a significant finding as the lakes in the WD and CW regions have limited hydrological connectivity. Earlier work showed that lakes positioned within chains or lowlands had stronger temporal coherence than upland lakes or those not connected by streams (Webster et al. 2000; Järvinen et al.

2002; Magnuson et al. 2005).

In the energy-mass flux framework for lakes, climate forcing directly and indirectly affects lake structure and function (Leavitt et al. 2009). Energy inputs such as solar irradiation can increase temporal coherence of lake variables (Magnuson et al.

2000), while mass inputs such as precipitation reduce coherence (Pham et al. 2008). SAT is higher in the WD region, but coherent with year-to-year trends in the CW region, while

MAP and SAP are much higher in the CW region and less coherent with precipitation trends in the WD region. Catchment topography and vegetation, which filter precipitation

signals across the landscape prior to entry into lakes and further processing by in-lake filters (Blenckner 2005), also differ between the two regions. In our study, temporal coherence of lake variables within regions was generally high; however, coherence of 1%

PAR and DOC was lower in the CW region than the WD region (1% PAR: 0.55 vs 0.90;

DOC: 0.10 vs. 0.87). From 2013-2017, the CW region received over twice as much precipitation throughout the water year than the WD region. These higher levels of mass inputs in the CW region may have decreased coherence of 1% PAR and DOC.

Furthermore, lakes in the WD region are part of a topographically flatter landscape with some rolling hills, while some catchments in the CW region are characterized by steep slopes, a geographical feature that may have contributed to reduced coherence by increasing spatial variability.

In the northern Great Plains of North America, Pham et al. (2008) found that coherence among lakes was influenced more strongly by energy transfer to the lake surface than by mass inputs, and so lakes with higher CEI were more sensitive to climate forcing from energy inputs. These findings substantiate the idea by Magnuson et al.

(1990) that the ratio of lake area to mean depth (CEI) provides a strong estimate of lake exposure to climate, as CEI acts as a primary filter of climatic signals prior to their entry to lakes. In our study, the WD lakes had CEI values of 37,000-75,000 m, while CEI of lakes in the CW region were much lower, ranging from 11,000-18,000 m (Figure 3.7).

The greater climatic exposure of lakes in the WD region may have contributed to high coherence of 1% PAR and DOC in that region. Interannual coherence of epilimnetic water temperature and Chl a was greater in the CW region than the WD region. In the context of the lower CEI values for the CW region, along with greater mass inputs from

precipitation, we expected these variables to exhibit reduced coherence in the CW region.

However, ice-out happens later in the CW region than the WD region; as a result, our

June sampling events occurred closer to ice-out in the CW region. This difference in lag time between ice-out and sampling between the two regions may have resulted in the higher-than-expected coherence values for epilimnetic water temperature and Chl a in the

CW region, as the lakes here had less time to filter climate signals (i.e. solar irradiance, heat, wind) that could differentially influence epilimnetic water temperature and Chl a.

Throughout the water year, monthly precipitation inputs in the WD and CW regions connect catchments and their lakes—evidenced by the positive association with soil-derived CDOM metrics (a*380 and SUVA254) and MAP. The coastal region of southwest Greenland receives significantly higher mean annual precipitation than the WD and CW regions, and as a result, lakes there have higher values of SUVA254, as well as other markers of terrestrial DOC (Osburn et al. 2017). Our results follow those of Osburn et al. (2017) that regions with greater mean annual precipitation have more terrestrial- derived DOC inputs to lakes, regardless of DOC concentration. Run-off from the landscape with precipitation also delivers nutrients to the lakes, which may contribute to the increase in Chl a with MAP. With increases in a*380, SUVA254, and Chl a from MAP, the depth of 1% PAR declines as transparency is reduced. These findings are supported by Saros et al. (2016), who found that CDOM metrics like a380, with Chl a, were more important controls of 1% PAR than DOC concentration. In this part of the Arctic, many lakes have high DOC and low color, hence measures of CDOM are more important for models of light attenuation than DOC concentration alone (Saros et al. 2016). Broadly, in semi-arid tundra and grasslands, there are hundreds of thousands of small lakes

characterized by high DOC and low CDOM (Curtis & Adams 1995; Anderson &

Stedmon 2007; Osburn et al. 2011). These lakes have different optical properties than lakes in forested watersheds with high CDOM; models of light attenuation for these lakes should prioritize CDOM metrics above DOC concentration (Saros et al. 2016). In the

Kangerlussuaq region, Anderson et al. (2017) described a strong link between terrestrial and aquatic DOC with spring snowmelt early in the growing season that weakens later in the growing season as soils dry with limited recharge from precipitation. Our findings that DOC was responsive to SAT and SAP follow this idea that with higher spring temperature and precipitation, more DOC is delivered to the lakes. This DOC may have both terrestrial and in-lake sources, as snowmelt transports allochthonous material to lakes, and autochthonous production of DOC is stimulated by incoming nutrients from snowmelt.

Breaking down the RDA results by matching climate drivers with lake responses highlighted differences in relationships between climate and lake responses between the

WD and CW regions. Lake responses between WD and CW regions reacted to climate drivers differently; the influences of all four climate drivers in the model were more apparent in the WD region than the CW region. The lake responses to climate drivers in the CW region were more muted, potentially affected by contributions from other variables not included in the study, or by interactions with climate variables not evident in the RDA results because they were overshadowed by strong relationships between climate drivers and the WD lake responses.

While WD coherence was higher for some lake variables (1% PAR and DOC), in general, year-to-year trends of all lake responses in the WD region matched those of the

climate factors that the RDA identified as important drivers. Conversely, year-to-year trends of lake responses in the CW region more poorly matched those of the climate drivers suggested by the RDA to be important. These findings corroborate those of Pham et al. (2008) and Magnuson et al. (1990) that lakes with lower CEI and more mass inputs via precipitation appear to be less connected to the influence of forcing by energy inputs.

We suggest that higher MAP and SAP and lower CEI of lakes in the CW region contributed to reduced coherence of 1% PAR and DOC and weaker relationships between climate drivers and lake responses. The disparity between climate sensitivity in lakes of the WD and CW regions indicates that different landscape and/or lake “filters” may have contributed to varied modification of climate signals between the regions. For example, some catchments in the CW region have steep slopes and areas of slumping permafrost, which may have variably altered timing or magnitude of run-off inputs from precipitation to lakes in this region. MAP was inversely associated with 1% PAR for lakes in this study, but the relationship was clearer for WD lakes. The variability in landscape features of the CW region may have contributed to reduced coherence of 1% PAR in CW lakes and masked its relationship to MAP.

We found strong evidence for the influence of MAP on soil-derived CDOM metrics, Chl a, and depth of 1% PAR in Arctic lakes with limited hydrological connectivity. These relationships were more pronounced in the WD region, which receives a fraction of the MAP and SAP of the CW region. Osburn et al. (2017) demonstrated that in the warmer, drier inland region of southwest Greenland, lakes had low hydrological connectivity with their catchments and high levels of degraded DOC.

With low mean annual precipitation, the lakes received limited soil-derived DOC.

Because hydrological connectivity is minimal, precipitation provides a crucial pathway for delivery of terrestrial organic matter and nutrients to these oligotrophic, chemically- dilute lakes. It is likely that terrestrial run-off from precipitation stimulates primary production and in-lake production and processing of DOC, as evidenced by the connection of Chl a to MAP in both the WD and CW regions. This study has illustrated the importance of regional climate differences, particularly MAP, as well as lake and landscape characteristics for C delivery to and processing within lakes. Saros et al. (2016) highlighted the importance of CDOM metrics for controlling light attenuation in Arctic lakes, and this study connects these CDOM metrics and 1% PAR to climate factors like

MAP. Climate controls on C cycling and light attenuation in lakes may also influence water temperature and inputs to the microbial loop, as the concentration and composition of DOC is altered with precipitation and temperature. In future scenarios, some regions of the Arctic may become drier (Bring et al. 2016). With reduced hydrological connectivity between these lakes and their catchments, drivers of light attenuation and C cycling may parallel those identified in this study, namely MAP.

By identifying the high levels of within- and between-region temporal coherence of lake variables in west Greenland Arctic lakes, and associating these lake responses with specific climate drivers, we have demonstrated how climate forcing affects Arctic lake features that contribute to light attenuation and C cycling. Notably, MAP was associated with increases in soil-derived DOC metrics and Chl a, and decreases in 1%

PAR. These effects were more prominent in the WD region. Interannual coherence of

DOC and 1% PAR was also higher in the WD region, possibly due to greater CEI and

C:L ratios in this region. With ongoing rapid climate change in the Arctic, climate-

mediated influences on terrestrial-aquatic linkages will likely elicit widespread impacts on light attenuation and C cycling in Arctic lakes.

CHAPTER 4

PALEOLIMNOLOGICAL COMPARISON OF ALGAL CHANGES IN A CLEAR

LAKE VERSUS A BROWN LAKE OVER THE LAST TWO CENTURIES IN

THE NORTHEASTERN U.S.A.

Abstract

We reconstructed fossil diatom and algal pigment records from a clear versus a brown lake in the northeastern U.S. to compare ecological responses to reduced acid deposition in recent decades in the context of a 140-year record, during which time multiple external drivers influenced both lakes. In the clear lake, diatom community structure changed continually from the beginning of the record, while it was more static in the brown lake over time, with the period of greatest change occurring after 1990.

Concentrations of algal pigments were low in the clear lake until 1940, then increased during the 1940 to 1990 period, after which they slightly declined. The opposite pattern occurred in the brown lake—algal pigments were high until 1940, decreased in the period from 1940 to 1990, and increased after 1990. The clear lake was more responsive to long- term climate warming beginning at the end of the Little Ice Age. During the period of higher acid deposition, light availability was more important for controlling algal responses in the clear lake, while nutrient subsidies from allochthonous dissolved organic matter (DOM) were likely the primary control in the brown lake. With reductions in atmospheric sulfate deposition, both lakes showed signs of recovery toward pre- acidification conditions, but these changes were dampened in the clear lake, suggesting greater sensitivity to acid deposition and other external drivers such as effects of climate change. Our study highlights strong and divergent algal responses in a clear vs. a brown

lake during a period of higher acid deposition, and considers these changes in the context of multiple drivers of environmental change over the past 140 years.

Introduction

The Paleoecological Investigation of Recent Lake Acidification (PIRLA) project of the 1990s provided conclusive evidence that anthropogenic sources of acidic atmospheric deposition in the twentieth century caused acidification of lakes across eastern North America, with signals beginning circa 1940 (Charles et al. 1990;

Whitehead et al. 1990). Paleolimnological reconstructions indicated that sensitivity to acid deposition varied across regions, and that reduced sensitivity or delayed effects of acidification in some lakes may have resulted from buffering by the lakes and their watersheds (Davis et al. 1983; Davis 1987). In PIRLA study lakes affected by acid deposition, dissolved organic carbon (DOC) generally declined with pH (Kingston and

Birks 1990). Since the successful implementation of the 1990 Clean Air Act

Amendments, increasing surface water DOC concentration is a major change that has been occurring in many lakes of eastern North America in response to reductions in acid deposition (Driscoll et al. 2003; Jeffries et al. 2003). Increasing soil pH can reduce ionic strength of the soil, increasing solubility of organic matter and facilitating its transport from watersheds to lakes (Monteith et al. 2007). Since the 1990s, a major research objective of lake ecology has been to understand how recovery from acidification (Keller et al. 1999; Stoddard et al. 1999) and subsequent DOC change (Donahue et al. 1998;

Jeffries et al. 2003) affect aquatic systems.

DOC is a primary control of water transparency in oligotrophic lakes, with important implications for lake ecology (Morris et al. 1995; Williamson et al. 1996; Gunn et al. 2001). Algal community structure and primary productivity are controlled in part by 69

light level (Berger et al. 2010), as are abundance (Williamson et al. 2015) and distribution (Rose et al. 2012) of zooplankton grazers. Reductions in water transparency can increase the amount of solar radiation absorbed in lake surface waters, which can alter lake thermal stratification timing and depth (Fee et al. 1996; Keller et al. 2006).

These changes in thermal structure may also reshape algal community structure, as key algal taxa respond to shifts in epilimnion depth and corresponding changes in temperature and/or nutrients, for example (Saros et al. 2016).

Patterns of recovery from acid deposition appear dependent, in part, on lakewater clarity. Clearer lakes have been hypothesized to be more sensitive to environmental change than darker lakes, as minor changes in water chemistry, precipitation, or energy could have amplified effects on transparency, physical structure, and other ecological variables (Snucins & Gunn 2000; Rose et al. 2009). Using data from the U.S.

Environmental Protection Agency’s National Lakes Assessment (NLA), Leech et al.

(2018) found that oligotrophic “blue” lakes decreased by 18% from 2007 to 2012 due to simultaneous “greening” and “browning” from eutrophication and run-off of chromophoric dissolved organic matter (CDOM) from the watershed. Williamson et al.

(2015) compared ecological consequences of long-term browning in two small lakes in

Pennsylvania—one a dystrophic, brown-water lake, and the other an oligotrophic, clear- water lake. Long-term trends were more apparent in the clear-water lake— reduced transparency in this lake had stronger effects on both epilimnetic and hypolimnetic water temperature (Williamson et al. 2015). This region of the U.S. has experienced increased precipitation and reduced acid deposition in recent decades.

Climate change can also be a major driver of lakewater clarity. Rising temperature may enhance production and release of DOC (Weyhenmeyer and Karlsson

2009; Clair et al. 2011), while shifts in precipitation intensity and frequency can boost flux of organic matter (Pace and Cole 2002; Weyhenmeyer et al. 2004; Zhang et al.

2010). For example, increased hydrological flows through the upper soil horizon flush organic matter to streams (McDowell and Likens 1988). The magnitude of this flux may increase with organic matter solubility. More than any other region of the U.S., extreme precipitation events in the northeastern U.S. have increased 60-70% since 1950 (Spierre et al. 2010), and Strock et al. (2016) found that in extreme wet years, DOC increased broadly in lakes across the northeastern U.S.

To better understand the ecological responses of clear vs. brown lakes exposed to multiple environmental drivers, including recovery from acidification and the simultaneous effects of rapid climate change, we performed a paleolimnological comparison of lakes in Acadia National Park (ANP) in Maine. ANP has experienced reductions in wet sulfate deposition and increases in temperature (Strock et al. 2017) and precipitation for several decades. Previous work has described recent physical and chemical trends in ANP lakes, but less is known about changes in algal ecology. Within

ANP, we collected sediment cores from a clear lake and a brown lake and used fossil diatom and algal pigment reconstructions to characterize algal responses to multiple drivers of recent environmental change in the context of long-term records. With the trend toward brownification of lakes across the U.S., our results provide a timely comparison of algal responses of a clear vs. a brown lake to recovery from acidification and climate change.

Methods

Site history and description of study lakes

Freshwater lakes cover 2,600 acres of the 35,000-acre landscape that makes up

ANP (Mount Desert Island, Maine, USA; 44.3°N, 68.2°W; Figure 4.1). ANP is mostly underlain by granite, with shallow, well-drained soils at high elevations and silts and clays at lower elevations (Gilman et al. 1988). By 1820, farming, lumbering, fishing, and shipbuilding were prevalent on Mount Desert Island. The area was granted national park status in 1919.

Jordan Pond

44.3 N Seal Cove Pond

10 km

68.2 W

Figure 4.1. Map highlighting Jordan Pond and Seal Cove Pond in Acadia National Park.

In the northeastern U.S., the onset of acidic deposition began in the late 1800s with the combustion of fossil fuels. Data from the Hubbard Brook Experimental Forest in

New Hampshire comprise the longest record of precipitation chemistry in North America

2- -1 and show that precipitation-weighted SO4 in wet deposition was 3.1 mg L in the mid-

1960s (Likens 1989). In ANP from 1981 to 2017, the concentration of precipitation-

2- -1 -1 weighted SO4 in wet deposition declined 0.04 mg L yr at the ME 98 weather station

(44.4°N, 68.3°W; NADP/NTN; Figure 4.2). The concentration of precipitation-weighted

- -1 -1 NO3 in wet deposition declined 0.01 mg L yr during this same period (NADP/NTN).

From 1940 to 2013, average minimum temperature in ANP increased by 0.015C yr-1, which accelerated to 0.15C yr-1 beginning in the early 1990s, and annual precipitation increased by 3.81 mm yr-1 (NOAA NCDC; Figure 4.3).

2.0

1.5

)

- mg L mg

( 1.0

4 S0

0.5

0.0 1981 1986 1991 1996 2001 2006 2011 2016

2- Figure 4.2. SO4 in wet deposition at ME-98 station from 1981-2017, data acquired from NADP/NTN.

14 2000 Minimum temperature mean 1800 12 Annual precipitation 1600

10 1400

1200 8

1000

6 800

600 Mean temperature ( 4

400 annual precipitation (mm) Total 2 200

0 0 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 4.3. Mean minimum temperature and total annual precipitation at ME-98 station from 1940-2013, data acquired from NCDC.

We selected two lakes in ANP with differing water transparency for this study.

Jordan Pond, an oligotrophic lake, is 0.8 km2 with a maximum depth of 46 m (Table 4.1).

Mean DOC concentration from 1995 to 2008 was 1.9 mg L-1 and mean Secchi depth was

13.8 m. Seal Cove Pond, a mesotrophic lake, is 1.0 km2 with a maximum depth of 13 m.

From 1995 to 2008, mean DOC concentration was 4.7 mg L-1 and mean Secchi depth was

6.9 m (U.S. National Park Service Monitoring Program, https://irma.nps.gov/Portal).

Table 4.1. Select characteristics of Jordan Pond and Seal Cove Pond. DOC, Secchi depth, and Chl a data are 1995-2008 averages from the U.S. National Park Service Monitoring Program, WA:LA is the ratio of watershed area to lake area, WRT is water residence time. Lake Mean Maximum Surface WA:LA WRT Fetch Elevation DOC Secchi Chl a depth (m) depth (m) area (km2) (yr) (km) (m a.s.l.) (mg L-1) depth (m) (µg L-1) Jordan Pond 26 46 0.8 6.0 5.9 2.0 83 1.9 13.8 1.0 Seal Cove Pond 5 13 1.0 7.4 0.5 2.9 12 4.7 6.9 2.8

Long-term monitoring data in ANP lakes document changes in DOC and water transparency since the 1890s. Water transparency in a suite of ANP lakes increased from the early-1980s to the mid-1990s, followed by significant declines from the mid-1990s to

2008. These declines in water transparency were synchronous with increases in DOC

(Strock et al. 2017; Figure 4.4). From 1995 to 2008, DOC increased from 1.6 to 2.2 mg

L-1 in Jordan Pond and 3.9 to 5.3 mg L-1 in Seal Cove Pond, while Secchi depth decreased from 13.9 to 12.1 m in Jordan Pond and 7.7 to 6.1 m in Seal Cove Pond (U.S. National

Park Service Monitoring Program, https://irma.nps.gov/Portal). The lakes are both circumneutral, and mean Chl a from 1995 to 2008 was 1.0 µg L-1 in Jordan Pond and 2.8

µg L-1 in Seal Cove Pond, with little change over that period (U.S. National Park Service

Monitoring Program, https://irma.nps.gov/Portal).

Figure 4.4. Changes in DOC and Secchi depth in Jordan Pond (a, c) and Seal Cove Pond (b, d) from 1995-2008. Note different y-axis scales.

The entire Jordan Pond watershed is included in ANP; land in the southern and eastern parts of the Seal Cove Pond watershed were acquired by ANP from 1932 to 1956, whereas the western side of the Seal Cove Pond watershed remains outside the ANP boundary. The northern half of the Jordan Pond watershed burned in the 1947 fire that engulfed almost 10,000 acres of ANP. The Jordan Pond watershed is currently made up of 35% evergreen forest and 3% wetland, and the Seal Cove Pond watershed contains

59% evergreen forest and 11% wetland by area.

Core collection and analysis

A sediment core was collected from the deepest part of Jordan Pond in 2008 and

Seal Cove Pond in 2016 using a gravity corer deployed from a boat. The Jordan Pond

core was 15 cm in length and the Seal Cove Pond core was 32 cm. Both cores were extruded in the field in 0.25-cm increments from 0 to 10 cm, and 0.50-cm increments from 10 cm to the bottom of the core. Sediment was stored in WhirlPak® bags at 4°C.

Sediments were dated at the University of Maine, using 210Pb activity counted with a high-purity germanium well detector with a low-background graded lead shield.

Chronologies were based on the constant rate of supply model (Appleby and Oldfield

1978). Percent loss on ignition (LOI) at 550°C was used as a proxy for percent sediment organic matter. It is important to note that sample resolution is lower in Jordan Pond than

Seal Cove Pond, and that the Jordan Pond record terminates at 2008, compared with 2016 for Seal Cove Pond, as the Jordan Pond core was collected for an earlier project and archived.

Historical diatom assemblages were identified and enumerated following standard methods (Battarbee 2001). Ten percent HCl and 30% H2O2 were used to digest carbonate and organic matter in each sample. The processed samples were mounted onto slides with

Naphrax®, and a minimum of 300 valves per slide were counted and identified on an

Olympus BX51 microscope at 1000x magnification. Taxonomic identification was based on Krammer & Lange-Bertalot (1986-1991) and Camburn & Charles (2000).

High performance liquid chromatography (HPLC) was used to identify and quantify algal pigments present in the core samples using the methods of Chen et al.

(2001). Sediment subsamples were stored at -80°C, then freeze-dried and extracted overnight in a mixture of acetone:methanol:water (80:15:5) at -20°C. Extracts were filtered through 0.22 µm PTFE syringe filters and dried under N2 gas, then re-dissolved in a 70:25:5 mixture of acetone, ion pairing reagent (0.75 g tetrabutyl ammonium acetate

and 7.7 g ammonium acetate in 100 ml water), and methanol prior to injection into the

Agilent HPLC unit. Commercial standards were used to calibrate peak areas (DHI

Denmark). Concentrations are reported in nanomoles of pigment per gram of organic sediment.

Data analysis

Principal components analysis (PCA) was conducted on pigment concentrations to identify variations over time (Legendre and Birks 2001). To estimate total algal biomass, we summed two chemically-stable indicators (β-carotene and pheophytin a;

Leavitt and Carpenter 1990). Detrended correspondence analysis (DCA) with down- weighting of rare taxa was applied to the percent relative abundance of all diatom species to evaluate diatom community turnover down core (Hill and Gauch 1990). Analyses were performed using the vegan package (Oksanen et al. 2013) in R version 3.4.2 (R

Development Core Team 2017). The resulting species scores from DCA axis 1, plotted over time, indicate diatom community turnover between samples in units of standard deviation (Hobbs et al. 2011). A square-root transformation to stabilize between-taxa variances did not change the DCA axis 1 scores, so relative abundance data were left untransformed prior to DCA. To infer past lake mixing depth, the mixing depth index

(MDI) was calculated for both lakes using a model of known diatom autecology, similar to those used by Wang et al. (2008), Boeff et al. (2016), Brown et al. (2016), and Stone et al. (2016). We calculated MDI as the ratio of the relative abundance of a diatom species indicating deeper mixing depth to a species indicating shallower mixing depth. In Jordan

Pond and Seal Cove Pond, we used Aulacoseira spp.:Discostella stelligera (Cleve &

Grunow) Houk & Klee relative abundance. For example, MDI is lower when D.

stelligera is more abundant, indicating shallower mixing depth. To reconstruct past water transparency, we calculated the ratio of planktic:benthic (P:B) diatom relative abundance.

Broadly, lower P:B indicates that more light can reach benthic habitats, implying greater water transparency. However, P:B may also reflect other lake conditions such as water balance, and can be affected by lake morphometry (Stone and Fritz 2004). For comparison purposes between Jordan Pond and Seal Cove Pond, sediment core dates were binned into three ranges: 1880-1940, 1940-1990, and 1990 to the end of the record

(2008 for Jordan Pond and 2016 for Seal Cove Pond). These bins broadly correspond to levels of atmospheric sulfate deposition in the U.S., i.e. relatively low from 1880-1940, reaching a maximum from 1940-1990, and declining in recent decades as a result of amendments to the Clean Air Act in 1990. The 2011 National Land Cover Database

(USGS 2014) was used to characterize land cover in the Jordan Pond and Seal Cove Pond watersheds.

Results

Core chronologies

Both the Jordan Pond and Seal Cove Pond cores had exponential declines in unsupported 210Pb activity with core depth (Figure 4.5). Unsupported 210Pb activity was contained within the top 5.5 cm for Jordan Pond and the top 25 cm for Seal Cove Pond.

Average error in estimated dates was ± 3 years for Jordan Pond and ± 6 years for Seal

Cove Pond. The Jordan Pond core extends from ~1874 to 2008 and the Seal Cove Pond core from ~1883-2016.

Jordan Pond 210Pb Activity (Bq g-1) Year 0.0 1.0 2.0 2010 1965 1920 1875 1830 0 a. b. 1

Depth (cm) Depth 4

7 Seal Cove Pond

0.0 0.1 0.2 2020 1990 1960 1930 1900 1870 1840 0 c. d. 5

Depth (cm) Depth 20

Figure 4.5. Stratigraphic plots of 210Pb activities and sediment chronologies in Jordan Pond (a, b) and Seal Cove Pond (c, d). Chronology is based on constant rate of supply (CRS) model and error bars represent error in 210Pb-estimated dates.

Diatom profiles

Changes in diatom communities were more apparent from the beginning of the record in 1880 to 1940 in the clear lake, Jordan Pond. Longer-term trends were marked by shifts occurring from 1940-1990 (Figure 4.6a). D. stelligera and Lindavia radiosa

(Grunow) De Toni & Forti dominated Jordan Pond communities at 14-36% relative abundance until 1990, at which time D. stelligera relative abundance remained high and

L. radiosa relative abundance dropped to 1-10%. The Lindavia comensis/tripartita/rossii complex (Cremer et al. 2001) was present throughout the core at 9-17% relative abundance, with percentages increasing in the late 1990s, around the same time that

Lindavia bodanica (Eulenstein ex Grunow) Nakov, Guillory, Julius, Theriot & Alverson began to increase in relative abundance (to a maximum of 12%). Eunotia spp. assemblages also increased during this period, while Aulacoseira spp., dominated by A. distans (Ehrenberg) Simonson, A. lacustris (Grunow) Krammer, A. lirata (Ehrenberg)

Ross, A. perglabra (Oestrup) Haworth, and A. perglabra var. floriniae (Camburn)

Haworth, had low relative abundance. Relative abundances of D. stelligera and

Aulacoseira spp. had an inverse relationship. The mean relative abundance of D. stelligera was 24% from 1880-1940, 28% from 1940-1990, and 29% from 1990-2008.

From 1880-1940, the mean relative abundance for Aulacoseira spp. was 3.7%, which dropped to 3.1% from 1940-1990, and declined again to 1.8% from 1990-2008. Diatom richness was lower in Jordan Pond than Seal Cove Pond; Jordan Pond had 60 taxa with relative abundance > 1% out of 100 taxa identified, while Seal Cove Pond had 76 taxa with relative abundance > 1% out of 248 taxa identified.

a. Jordan Pond

Relative abundance (%) b. Seal Cove Pond

Relative abundance (%)

Figure 4.6. Relative abundances of selected diatom taxa, mixing depth index (MDI), planktic:benthic diatom ratio (P:B), and DCA Axis 1 in SD units for Jordan Pond (a) and Seal Cove Pond (b) sediment cores.

Indices of diatom community structure also illustrate early and continuous change in the Jordan Pond record. The mean P:B ratio in Jordan Pond decreased over time

(Figure 4.6a). The 1880-1940 mean P:B was 3.5, the 1940-1990 mean P:B was 3.0, and the 1990-2008 mean P:B was 2.1. The P:B trend increased from 1990-2008, however.

The 1940-1990 P:B was marked by a peak P:B of 3.9 in 1960, which declined to a low value of 1.6 during the decade from 1987-1997. The P:B peak of 1960 and the subsequent decline were primarily driven by shifting assemblages of planktic D. stelligera and Lindavia spp., with influences from benthic Achnanthes and Eunotia spp.

The general decline in the MDI over time indicates shallowing mixing depth from 1880 to 1940, followed by a deeper mixing depth in 1960, and then a subsequent shallowing until 2008. Mean MDI was shallowest from 1990-2008. Shallowing MDI values were driven by increases in D. stelligera and decreases in Aulacoseira spp. The deep MDI in

1960 represents a dip in D. stelligera and a sharp increase in relative abundance of

Aulacoseira spp. Diatom community turnover, represented by DCA axis 1 scores, increased steadily through time.

In the brown lake, Seal Cove Pond, trends in diatom communities and related indices were generally more static through time, with the largest changes in community turnover occurring after 1990 (Figure 4.6b). Seal Cove Pond diatom assemblages were dominated by D. stelligera (13-33% relative abundance) and Aulacoseira spp. (7-29% relative abundance). Lindavia bodanica occurred at low relative abundance, from 0.3-7% of the assemblage. Discostella stelligera increased from 1880-1990, at which point it stayed high, with the highest relative abundance occurring in 2016, at 33%. The average

D. stelligera relative abundance was 16% from 1880-1940, 21% from 1940-1990, and

23% from 1990-2016. In Seal Cove Pond, Aulacoseira spp. were comprised of A. alpigena (Grunnow) Krammer, A. ambigua (Grunow) Simonsen, A. distans, A. lacustris,

A. lirata, A. nygaardii (Camburn) Camburn & Charles, A. perglabra, A. perglabra var. floriniae, A. subarctica (O. Müller) Haworth, and A. tenella (Nygaard) Simonsen.

Relative abundance of Aulacoseira spp. was highest near the bottom of the core, with mean relative abundance of 21% from 1880-1940. Mean relative abundance dropped to

12% from 1940-1990, and increased to 14% from 1990-2016.

Similarly, the average annual P:B ratio in the brown lake showed little directional change over time, and when changes in P:B occurred, they were smaller in magnitude than those in Jordan Pond (Figure 4.6b). The mean P:B was 1.5 from 1880-1940, 1.4 from 1940-1990, and 1.5 from 1990-2016. This trend was influenced by declines in

Aulacoseira spp. from 1940-1990. While the P:B increase from 1990-2016 was largely driven by increases in relative abundance of both D. stelligera and Aulacoseira spp., several primarily benthic taxa belonging to the aggregated genera of small Fragilaria,

Achnanthes, Cymbella, Brachysira, Gomphonema, and Eunotia spp. also increased in relative abundance during this period. MDI was deepest in 1880 and shallowed until

1940, then stayed shallow, but variable until 2016. The MDI trend was driven by declines in Aulacoseira spp. relative to increases in D. stelligera until 1990. Diatom community turnover was stable until 1990, when it rapidly increased until 2016.

Algal pigments

Jordan Pond had fewer types and lower concentrations of algal pigments than Seal

Cove Pond (Figure 4.7a). Alloxanthin (cryptophytes), canthaxanthin (cyanobacteria), lutein-zeaxanthin (chlorophytes-cyanobacteria), and Chl a, pheophytin a, and β-carotene

(all algal groups) were present in the core, and concentrations of all types of pigments generally changed synchronously over time. PCA scores were strong predictors of patterns of algal change—PCA axis 1 explained 84% of the variance. Pigment loadings on PCA Axis 1 ranged from -0.14 (β-carotene) to -0.80 (pheophytin a; Table 4.2). From

1880-1940, average pigment concentrations were low, with a synchronous peak among pigments occurring during the 1940-1990 period, followed by a subsequent decline in most pigments from 1990-2008. Estimated mean total algal biomass was 25 nmol g-1 from 1880-1940, 42 nmol g-1 from 1940-1990, and 38 nmol g-1 from 1990-2008. These temporally coherent shifts in all pigments suggest that algal taxa responded similarly to external drivers. Mean LOI was 7.7% from 1880-1940, 9.1% from 1940-1990, and 7.1% from 1990-2008.

Table 4.2. PCA Axis 1 loadings for Jordan Pond and Seal Cove Pond pigments.

Pigment Jordan Pond Seal Cove Pond PCA Axis 1 PCA Axis 1 loading loading Fucoxanthin - 0.04 Diatoxanthin - 0.48 Alloxanthin -0.34 -0.11 Chl b - -0.03 Canthaxanthin -0.23 0.00 Lutein-zeaxanthin -0.21 0.15 Echinenone - 0.04 Pheophorbide a - 0.76 Chl a -0.35 -0.25 Pheophytin a -0.80 0.02 β-carotene -0.14 0.19

Figure 4.7. Extracted pigment concentrations, percent LOI, and PCA Axis 1 in Jordan Pond (a) and Seal Cove Pond (b) sediment cores.

By comparison, Seal Cove Pond had more types of pigments, at 2-6 times the concentration of those in Jordan Pond (Figure 4.7b). In addition to the Jordan Pond pigments, fucoxanthin and diatoxanthin (siliceous algae), Chl b (chlorophytes), echinenone (cyanobacteria), and pheophorbide a (all algae) were present in the Seal Cove

Pond core. PCA axis 1 explained 51% of the variance. Pigment loadings on PCA Axis 1 ranged from -0.25 (chlorophyll a) to 0.76 (pheophorbide a; Table 4.2). The trends in pigment concentration and LOI were generally inverse to those of Jordan Pond. For example, pigment concentrations were relatively high from 1880-1940, followed by a decline from 1940-1990, and a return to higher levels from 1990-2016. Estimated mean total algal biomass was 50 nmol g-1 from 1880-1940, 39 nmol g-1 from 1940-1990, and returned to 51 nmol g-1 from 1990-2016. While many pigments followed this trend, not all pigments behaved coherently, as in Jordan Pond. For example, concentrations of biomarkers from siliceous algae such as diatoms (diatoxanthin), and an indicator of photodegradation and grazing pressure (pheophorbide a), remained low after dropping around 1940. Cryptophytes (alloxanthin) increased after 1990, while a replacement and/or reduction in cyanobacterial pigments occurred. Colonial cyanobacteria

(canthaxanthin) increased while indicators of total cyanobacteria (echinenone, β-carotene, and lutein-zeaxanthin, which changed similarly to other cyanobacterial pigments and may be predominantly zeaxanthin) declined. Chlorophyll a concentration decreased slightly from 1940-1990, and rapidly increased starting in 2000. However, Chl a can degrade rapidly, so we interpret the higher concentration at the top of the core with caution. LOI was approximately 3 times greater in Seal Cove Pond than Jordan Pond. Average annual

LOI was 26% from 1880-1940, 24% from 1940-1990, and 25% from 1990-2016.

Discussion The PIRLA project identified widespread effects of anthropogenic acid deposition on lakes of northeast North America from the 1940s-1990s. However, not all lakes in

New England responded similarly to acid deposition. Many lakes in Maine, where most of the New England study lakes were located, became slightly more alkaline due to buffering of lakes and their catchments against atmospheric inputs (Davis et al. 1983;

Davis 1997; Stauffer 1990). For example, diatom indicators of acidification in several

Maine lakes lagged decades to a century behind emissions of early atmospheric pollutants

(Davis et al. 1983). Additionally, clear and brown lakes exhibit different responses to recovery from acidification and other external pressures (Williamson et al. 2015). Our study highlights strong algal responses by a clear and a brown lake in Maine and contrasts responses of the two lakes to changes in acid deposition in recent decades in the context of multiple drivers of environmental change over the last 140 years.

Clear Jordan Pond and brown Seal Cove Pond exhibited varying responses to long-term climate change and atmospheric deposition. As the Little Ice Age ended in the latter half of the 19th century, diatom records indicate that Jordan Pond may have been more sensitive to climate warming than Seal Cove Pond. Diatom community turnover increased and mixing depth steadily shallowed in Jordan Pond, while mixing depth shallowed in Seal Cove Pond, but community turnover was relatively low. Clear Jordan

Pond may have been more sensitive to warming because the thermal structure of small lakes is largely controlled by water clarity (Snucins and Gunn 2000) and small changes in energy inputs are thought to cause amplified responses in thermal structure and ecosystem linkages (Rose et al. 2009). Additionally, increases in DOC concentration have a stronger effect on light attenuation in a clear lake than in a brown lake

(Williamson et al. 2015). Read and Rose (2013) found that the relationship between climate sensitivity and lake color is nonlinear, explaining why diatom community turnover in Jordan Pond may have been more responsive to warming than in Seal Cove

Pond.

Jordan Pond and Seal Cove Pond experienced pronounced, but opposite, ecological responses to acid deposition from approximately 1940-1990. In Jordan Pond, mixing depth and the relative abundance of planktic diatoms experienced a peak in 1960 followed by a sharp decline, and diatom community turnover rapidly increased. All detectable algal pigments from the Jordan Pond record increased to peak levels from

1940-1990, and algal biomass was highest during this time. The coherent increase in algal pigments in Jordan Pond may have been due to deepening of the epilimnion. It is important to note, however, that our inference of deeper MDI during this time is based on one sample. As DOC inputs to Jordan Pond decreased and transparency increased with acid deposition, the mixing depth likely deepened as less photosynthetically active radiation (PAR) was attenuated by DOC. This change would have allowed light to penetrate deeper into the water column, where more nutrients were also available. In recent years, Jordan Pond has had a deep chlorophyll maximum (DCM) below 15 m, with water column TP and TN at higher levels below the epilimnion (Saros, unpublished data).

Saros et al. (2005) found that algal biomass in the DCM was correlated with greater nutrient availability in the hypolimnion instead of UV avoidance in oligotrophic lakes.

Deeper light penetration coupled with increased nutrient availability and a deeper mixing depth may have increased habitat availability for planktic algae in Jordan Pond from

1940-1990, which is supported by the increase in algal pigments during this time.

While concentrations of algal pigments increased synchronously during the period of acid deposition in Jordan Pond, algal pigments in Seal Cove Pond were at their lowest concentrations during this time. Light availability may have been a less likely mechanism of shifting algal responses in Seal Cove Pond than in Jordan Pond. Seal Cove Pond is surrounded by four times more wetland area than Jordan Pond. The wetland likely provides a continuous supply of DOM to Seal Cove Pond, reducing temporal variability of light availability and providing a buffer to long-term environmental changes. In recent years, the mixing depth of Seal Cove Pond has been approximately 6 m (Saros, unpublished data), while average Secchi depth is almost 7 m and mean depth is 5 m. The relative proportion of Secchi depth to mean lake depth is much greater in Seal Cove Pond than in Jordan Pond. In many areas of the lake, light can penetrate to the bottom of the water column. Similarly, the mixing depth of Seal Cove Pond, even at its most shallow, was likely deep enough so that after 1940, the entire water column could have been mixed in many areas of the lake. Fluctuations in P:B ratio were small throughout Seal

Cove Pond record, indicating that changes in light availability were minor and/or exerted minimal influence on the composition of diatom functional groups.

Loss of nutrient subsidies resulting from acid deposition was likely a more important driver than light of algal responses in Seal Cove Pond during the period of higher acid deposition. In a study of boreal lakes, Daggett et al. (2015) found that regardless of the nutrient limitation pattern of a lake, algal biomass increased with dissolved organic matter (DOM) enrichment, and that experimental DOM additions simultaneously stimulated many taxa. DOM is increasingly viewed as a nutrient subsidy

(Granelí et al. 1999; Klug 2002; Kissman et al. 2013), and Klug (2002) discovered that

doubling DOM concentration from 10-20 mg L-1 in a moderately colored lake had positive effects on total phytoplankton growth. In ANP, Seal Cove Pond is a higher-DOC lake, with wetland and evergreen forest covering a large proportion of its watershed.

Compared to Jordan Pond, Seal Cove Pond has a greater watershed:lake area (WA:LA) with shallower watershed slope and a much shorter water residence time. Thus, terrestrial subsides of DOM may comparatively be more important sources of nutrients in Seal

Cove Pond than in Jordan Pond. While we do not have long-term nutrient limitation data for the lakes in this study, it is reasonable to infer that reduced inputs of landscape DOC to Seal Cove Pond during the period of high acid deposition diminished nutrient availability to algae and resulted in widespread reductions in biomass across algal taxa.

Since 1990, atmospheric sulfate deposition has decreased in ANP, while annual precipitation and average minimum temperature have risen. Diatom and algal pigment records indicate that both Jordan Pond and Seal Cove Pond have experienced substantial, and in some cases opposite, ecological changes during this acid recovery period. In both lakes, diatom community turnover continued to increase rapidly and MDI became shallower, suggesting strong responses of both lakes to multiple environmental drivers.

However, the P:B ratio decreased abruptly in Jordan Pond, while it increased slightly in

Seal Cove Pond. This shift supports the assertion that changes in light environment elicit stronger responses in Jordan Pond than Seal Cove Pond. As DOC increased and transparency decreased in recent decades, diatom assemblages quickly responded. Post- acidification shifts in algal taxa highlight rapid changes in both lakes. Algal pigments generally declined in Jordan Pond, while concentration of most algal pigments in Seal

Cove Pond increased to pre-acidification levels. With documented increases in DOC and

reductions in transparency during this time, greater shading and attenuation of PAR by

DOC may have negatively affected primary production in Jordan Pond. In Seal Cove

Pond, increased DOM may have increased nutrient availability and boosted algal production, with reduction in light availability eliciting limited algal response.

In this study, clear Jordan Pond was more sensitive to long-term climatic changes than brown Seal Cove Pond, which displayed more subtle responses from the end of the

LIA to 1940. Both lakes responded strongly to acid deposition occurring in ANP from

1940-1990, but the ecological responses of the two lakes were often in contrast to each other. When atmospheric sulfate deposition declined after 1990 and precipitation and temperature continued to increase, both Jordan Pond and Seal Cove Pond experienced notable changes. Some of the ecological changes in recent decades were the greatest of the sediment records—for example, diatom community turnover was at its highest for both lakes, while mixing depth was at its shallowest. Discostella stelligera reached the highest relative abundance of the record in the most recent core samples. Notably, shifts in algal pigments might indicate some form of recovery in both lakes, as concentrations of pigments of most taxa return to pre-acidification levels after 1990. This change is most prominent in Seal Cove Pond, where bioindicators of most algal taxa increase to pre- acidification levels after 1990. In Jordan Pond, many algal taxa decline after 1990, but pigments do not reach the low concentrations that existed prior to 1940. In fact, total algal biomass remains much higher than it was during the years prior to acidification.

However, we are careful to note that the higher pigment concentrations in these samples may be due to lower rates of degradation than in earlier samples. In nutrient limitation and DOM addition experiments, Daggett et al. (2015) found that Jordan Pond was co-

limited by N and P in 2009, and that additions of 1 mg L-1 DOM doubled algal biomass.

Reductions in atmospheric sulfate and nitrate deposition in ANP since 1990 paired with increasing DOC in Jordan Pond may have had interacting effects that resulted in the relatively small reductions in algal biomass during the 1990-2016 period that remained greater than concentrations prior to 1940. Altogether, these changes in diatom metrics and algal pigments in both lakes likely illustrate some recovery from the effects of acid deposition and a return toward pre-acidification conditions, especially in Seal Cove Pond.

While the differing effects of acidification and recovery are apparent in these records from a clear vs. a brown lake, it is important to note that recent effects of climate change have also influenced the ecological responses of both lakes. For example, increased precipitation and temperature likely amplify inputs of allochthonous DOC to both lakes.

Lake bathymetry and watershed characteristics also influence ecological responses. Even so, clear Jordan Pond responded more quickly to past climate warming than brown Seal

Cove Pond, and has been slower to recover from the effects of atmospheric sulfate deposition as precipitation and temperature continue to increase in the northeastern

U.S.A.

Conclusions

This study has investigated the ecological responses of a clear vs. a brown lake exposed to multiple drivers of environmental change over the last 140 years, including longer-term climate change marked by a period of acid deposition. We found that clear

Jordan Pond was more sensitive to long-term changes in climate, while brown Seal Cove

Pond exhibited more subtle responses. The lakes had varying responses to acid deposition. In Jordan Pond, mixing depth deepened while bioindicators of algal taxa

increased synchronously, likely due to increased water transparency, as evidenced by the

P:B ratio. In Seal Cove Pond, changes in light had little effect on algal responses, but losses of nutrient subsidies due to reduced allochthonous DOC may have been responsible for declines in algal biomass across taxa. As the rate of acid deposition declined after 1990, both lakes exhibited some return toward pre-acidification conditions, although recovery in Jordan Pond has been slower. These recent ecological changes reflect responses to multiple external pressures, including abrupt shifts in temperature and precipitation, in addition to reduced atmospheric sulfate deposition. In addition to lake color, it is necessary to consider morphological features of lakes as well as their watershed characteristics when interpreting responses to multiple environmental drivers of ecological change.

CHAPTER 5

DECIPHERING DIVERGENT HARMFUL ALGAL BLOOM OCCURRENCE IN

TWO SOUTHEAST AUSTRALIAN DEEP LAKES: IMPLICATIONS FOR LAKE

CONSERVATION IN THE 21ST CENTURY

Abstract

Record high temperatures and drought conditions paired with increases in harmful algal blooms (HABs) are threatening water resources prized for their ecological, recreational, and scientific value in southeast Australia. We used paleolimnological and contemporary techniques to assess HAB incidence over the last century in two deep crater lakes and better understand why HABs currently develop in only one of the lakes.

Diatoms in both lakes were responsive to increased salinity resulting from drought, but shifts in diatom assemblages were more pronounced in mesosaline Lake Bullen Merri than fresh Lake Purrumbete. After 1913, Cyclotella choctawhatcheeana dominated the diatom assemblage in Bullen Merri. Algal pigments from Bullen Merri sediments were generally higher during periods of drought. A period of low algal biomass from the

1940s-1990s may have resulted from higher precipitation and reduced light availability during this time, and/or the introduction of fish stocking in Bullen Merri. Most pigments, including those of cyanobacterial taxa, reached peak concentrations during the severe drought from 1997-2010. Pigment concentrations were lower in Purrumbete than Bullen

Merri, but remained relatively high until 1947, the end of a drought and a period of high fish stocking in Purrumbete. We demonstrated that throughout the last century, cyanobacterial taxa were present in both lakes, and that algae in the lakes were sensitive to environmental changes. However, different hydrological and physicochemical

parameters may cause Lake Bullen Merri to experience regular nuisance blooms while

Lake Purrumbete does not. HABs in Bullen Merri are likely caused by drought and resulting increases in salinity, as well as long water residence time within the lake, interactive effects of nutrients, and strong thermal stratification. With predicted increases in temperature and drought conditions, we expect that future HAB formation will be less likely in Purrumbete, but ongoing in Bullen Merri.

Introduction

The development and persistence of harmful algal blooms (HABs) in southeast

(SE) Australia lakes have intensified in recent years. HABs are deleterious to human health and can disrupt economies that depend on affected water resources (Hallegraeff

1993). Nodularia spumigena, a nitrogen (N)-fixing cyanobacteria, and many of the other common HAB-forming genera produce hepatotoxins and neurotoxins (Baker & Humpage

1994), and are suspected to cause carcinogenic effects in humans (Falconer et al. 2001).

While the health care costs associated with HABs are staggering (Falconer et al. 2001),

HABs also strain Australia’s economy by reducing recreational value of waterways and damaging aquaculture, agriculture, fishing, and tourism industries, costing an estimated

$180-240 million AUD per year (Atech 2000).

Some of the blooms have been attributed to nutrient run-off from human activities and resulting eutrophication of water bodies (Cook et al. 2010). Paerl and Huisman

(2008) note that HABs are also caused by changes in water bodies due to climate change effects, such as increased water temperature, enhanced stratification, shifts in salinity, and longer water residence time. Higher water temperature contributes to HABs by allowing toxic cyanobacterial species a competitive advantage over non-toxic algal

species. For example, HAB-forming Microcystis aeruginosa experiences peak growth rates at approximately 30°C, while species of non-toxic dinoflagellates, diatoms, and chlorophytes experience peak growth rates between 20-25°C and exhibit rapid die-off at higher temperatures (Paerl & Paul 2012). Nodularia spumigena in particular is dominant in temperate Australian lakes, likely due to long water residence times, intermediate salinity (5-20 ppt), and elevated phosphorus (P) supply from catchments and sediments of anoxic bottom waters (Cook et al. 2016). Nodularia spumigena is a buoyant and slow- growing species that grows optimally with enhanced water column stratification and residence times (Paerl 2014).

Climate in SE Australia is largely controlled by the El Niño Southern Oscillation

(ENSO), the Southern Annual Mode (SAM), and the Indian Ocean Dipole (IOD), all of which can impact seasonal rainfall patterns and undergo natural patterns of variability

(CSIRO 2010). Natural variations in ENSO, SAM, and the IOD are likely amplified by impacts of human-driven global warming on Hadley Cell circulation, which is correlated with decreasing autumn rainfall (Murphy & Timbal 2008, CSIRO 2010). SE Australia underwent a record-breaking drought from 1997-2010 that lasted longer and was more severe than any drought in recorded history (Timbal 2009; CSIRO 2010). Temperatures in this region increased by approximately 0.5-1°C over the past 100 years (BOM &

CSIRO 2014), and rose above the 1961-1990 mean every year between 1996-2010

(CSIRO 2010). These rapid shifts in precipitation and temperature affect the precipitation/evaporation (P/E) ratio of lakes as well as hydrological inputs (Jones et al.

2001; Yihdego et al. 2016), and can alter salinity, nutrient cycling, and thermal structure of lakes that contribute to HAB development (Paerl 2014).

Previous paleolimnological work in SE Australia provided detailed diatom-based reconstructions of past environmental change, particularly drought and its effects on lake level (Gell et al. 2012). Since the 1850s, decreases in effective rainfall have contributed to declining lake level in large crater lakes, and landscape modifications since European settlement ca. 1840 have reduced the resilience of wetlands in this area to recent drought conditions (Gell et al. 2012). Using historical records, survey data, and a lake balance model, Jones et al. (2001) found that lake level dropped more than 20 m in some deep crater lakes since 1859. After this time, wetlands in the region shifted to more saline conditions, and diatom bioindicators of disturbance and/or pollution became more prevalent in lake records (Gell et al. 2012). While lake level modelling and fossil diatom records indicate that climate in SE Australia has changed since the 1850s, comparatively few algal pigment records are available to illustrate how changing climatic conditions may have affected ecology of lakes in this region. Cook et al. (2016) used algal pigment biomarkers to show an increase in cyanobacterial pigments in a temperate SE Australian lagoon system during recent blooms of N. spumigena. In a 100-year sediment record from a lake within the same lagoon system, chlorophyll a (Chl a) concentration peaked after the 1980s, coincident with the rise of HABs in this area (Saunders et al. 2008).

As lake level in SE Australia is projected to continue declining (Jones 2010), a few deep (>5 m) crater lakes may be the only lakes in this region to persist throughout the

21st century (Leahy et al. 2010), as lakes of these depths are uncommon in mainland

Australia (Timms 1976). These deep lakes are essential water resources in SE Australia, and are of national and international significance, as some support migratory bird species and endangered plants (Leahy et al. 2010). They are valued recreational fisheries, tourist

100

areas, and sites of scientific research. With ongoing effects of drought, Yihdego et al.

(2016) predict that frequent seasonal drying of shallower lakes in the region will cause a deep crater lake, Lake Purrumbete, to become an important refugium for aquatic organisms. Maintaining the quality of deep crater lakes in SE Australia is therefore a top priority of local and state water management authorities as climate and other anthropogenic pressures escalate (Kirono et al. 2009; Victoria Department of

Environment, Land, Water and Planning 2017).

Two of the deep crater lakes of greatest ecological, recreational, and scientific significance in this region are Lake Bullen Merri and Lake Purrumbete. Both lakes are located in the western district lakes region of Victoria and were created by past volcanic activity. Bullen Merri is a brackish lake that has for decades been an important fishing, boating, and camping destination. Much of its catchment is dotted by cattle pastures of dairy and beef farms. Water quality in Lake Bullen Merri has declined in recent decades, with persistent N. spumigena blooms often causing closures during the busy summer season. As of 2019, the Department of Environment, Land, Water and Planning

(DELWP) restricted public usage of Bullen Merri for several summer months during each of the past five years (Thompson 2018; “Blue-green disgrace” 2019). In contrast, HABs are largely absent from fresh Lake Purrumbete, a designated nature conservation reserve and another popular recreational fishery. Our objectives were to 1) use paleolimnological proxies to identify the temporal pattern of HABs in Lake Bullen Merri in the context of the longer-term lake record, and 2) to compare and contrast ecological indicators in Lakes

Bullen Merri and Purrumbete to better understand the disparity in HAB occurrence between the two deep crater lakes. It is our intent that the results of this study may

101

provide insights useful to managers of Australian lakes as they develop applications for controlling the deleterious impacts of HABs on water resources.

Methods

Site descriptions

Lakes Bullen Merri (38.25°S, 143.10°E) and Purrumbete (38.25°S, 143.22°E) are both volcanic maar lakes in the Western District of Victoria, a region characterized by a large plain scattered with volcanic features and widespread red volcanic soils (Leahy et al. 2010; Figure 5.1). Bullen Merri is P-limited (Herpich 2002) and eutrophic, with low water clarity (Table 5.1; Timms 1976). Silica concentration ranged from 0.1-0.4 mg L-1 in 2007-2008 (Leahy et al. 2010). Purrumbete is N-limited (Herpich 2002) and mesotrophic, with higher water clarity (Timms 1976). Silica concentration in Purrumbete ranged from 1.7-3.0 mg L-1 in 2007-2008 (Leahy et al. 2010). Major ions in the lakes are similar due to comparable catchment geology, and Purrumbete is more exposed to wind because the altitude of its crater rim is lower than that of Bullen Merri (Timms 1976).

Because Purrumbete is fresher and receives more wind, the thermocline is deeper in

Purrumbete than in Bullen Merri (average 25 m vs. 22 m January-June 1970-1972;

Timms 1976). Stratification of the lakes begins in October-November and continues through late June (Purrumbete) or early July (Bullen Merri; Timms 1976). The surface waters of both lakes are well-oxygenated, and reach an oxygen minimum during peak stratification from May-July, when the percent saturation of oxygen below the thermocline is undetectable (Timms 1976). Groundwater input to Bullen Merri is low

(Jones 2001) and there is no stream in- or outflow, so lake level is largely controlled by the balance of precipitation and evaporation (Leahy et al. 2010). Purrumbete receives

102

some surface water inflow and groundwater input, and has an outflow channel to the

Curdies River (Yihdego 2010). Lake level is often below that of the outflow channel, and the level of Lake Purrumbete is also primarily controlled by precipitation and evaporation

(Yihdego 2010).

Figure 5.1. Map of deep crater lake study sites in SE Australia.

103

Table 5.1. Select characteristics of Lakes Bullen Merri and Purrumbete. Morphometric data from Timms (1976).

Lake Lat (°) Long (°) Mean Max Surface Volume Fetch Elevation depth (m) depth (m) area (km2) (x 10-6 m3) (km) (m a.s.l.) Bullen Merri 38.25 S 143.10 E 39 66 4.9 190 2.6 146 Purrumbete 38.25 S 143.22 E 29 45 5.5 160 3.2 139

104

Lake sampling

We sampled Lakes Bullen Merri and Purrumbete for basic physicochemical data in February 2017, anchored by raft at the point of maximum depth for each lake. We used a submersible Aquaread water quality sonde (Aquaread Ltd., Kent, England) to measure water temperature, pH, dissolved oxygen, salinity, and the location of the thermocline along a vertical water column profile of 0-10 m depth. To estimate phycocyanin concentration, which is generally associated with harmful algal blooms, we took readings with a FluoroSense handheld fluorometer (Turner Designs, Sunnyvale, CA) at the surface of each lake. Water clarity was measured using a Secchi disc. We collected water from the epilimnion of each lake for quantification of DOC, Chl a, and nutrients using a horizontal van Dorn bottle. Water samples were filtered through pre-combusted, pre- rinsed 0.7 µm GF/F filters, except samples for total nutrient analysis, which were unfiltered. The filtered water samples for DOC and dissolved nutrient analysis were chilled and Chl a filters were frozen for return to and analysis at the University of Maine.

Permits were not required for this study, as Lakes Bullen Merri and Purrumbete are open to public access with no special permission required. We did not sample any protected species, and only performed lake water and sediment analyses.

Laboratory methods

DOC concentration was analyzed using a Shimadzu TOC-5000 analyzer

(Shimadzu Corporation, Kyoto, Japan) by high-temperature catalytic oxidation. Chl a analyses were performed according to standard methods (APHA 2000) within three weeks of filtration. Briefly, Chl a was extracted with 90% acetone and clarified via centrifugation, then analyzed using a Varian Cary-300 Ultraviolet-Visible

105

spectrophotometer (Agilent Technologies, Santa Clara, California). Total nitrogen (TN) and total phosphorus (TP) samples were analyzed on a Lachat QuickChem 8500 analyzer

+ via persulfate digestion (Hach Company, Loveland, CO), and ammonium (NH4 ), nitrate

- (NO3 ), and soluble reactive phosphorus (SRP) were analyzed with the phenate, cadmium reduction, and ascorbic acid methods, respectively (APHA 2000). Quantification limits

-1 -1 -1 + -1 were 1 µg L for TP and SRP, and 10 µg L for TN, 5 µg L for NH4 , and 2.5 µg L for

- + - NO3 . Dissolved inorganic nitrogen (DIN) was calculated as the sum of NH4 and NO3 .

- -1 When NO3 concentration was below detection, we used 1 µg L in calculation of DIN, as in Burpee et al. (2016).

Core collection and analysis

A sediment core was collected from the deepest part of each lake using a gravity corer deployed from a raft. The Bullen Merri core was 39 cm and the Purrumbete core was 9 cm. Both cores were extruded in the field in 0.25-cm increments. After 10 cm, the

Bullen Merri core was extruded in 0.5-cm increments to the bottom of the core. Sediment was stored in WhirlPak® bags and chilled, and returned to the University of Maine for processing, at which time sediments were subsampled for diatom, algal pigment, and dating analysis. Sediments were dated at the University of Maine using 210Pb activity counted with a high-purity germanium well detector with a low-background graded lead shield. Chronologies were based on the constant rate of supply model (Appleby and

Oldfield 1978). Percent loss on ignition (LOI) at 550°C was used as a proxy for percent sediment organic matter.

Historical diatom assemblages were identified and enumerated following standard methods (Battarbee 2001). Ten percent HCl and 30% H2O2 were used to digest carbonate

106

and organic matter in each sample. The processed samples were mounted onto slides with

Naphrax®, and a minimum of 300 valves per slide were counted and identified on an

Olympus BX51 microscope at 1000x magnification. Taxonomic identification was primarily based on Krammer & Lange-Bertalot (1986-1991), Sonneman et al. (2000), and

Gell et al. (1999).

High performance liquid chromatography (HPLC) was used to identify and quantify algal pigments present in the core samples using the methods of Chen (2001).

Sediment subsamples were stored at -80°C, then freeze-dried and extracted overnight in a mixture of acetone:methanol:water (80:15:5) at -20°C. Extracts were filtered through 0.22

µm PTFE syringe filters and dried under N2 gas, then re-dissolved in a 70:25:5 mixture of acetone, ion pairing reagent (0.75 g tetrabutyl ammonium acetate and 7.7 g ammonium acetate in 100 ml water), and methanol prior to injection into the Agilent HPLC unit.

Commercial standards were used to calibrate peak areas (DHI Denmark). Concentrations are reported in nanomoles of pigment per gram of organic sediment. A power outage occurred during analysis of a subset (6-19 cm) of pigment samples from Lake Bullen

Merri, and these samples were left unrefrigerated overnight. Analysis of a portion of these samples (6-8 cm) was completed before the power outage occurred. After re- analyzing all the samples from 6-19 cm, we compared the concentrations of pigments in the earlier run of undegraded 6-8 cm samples with those that had been unrefrigerated overnight to estimate the magnitude of pigment degradation that may have occurred in the subset of samples that were not analyzed prior to the power outage (8-19 cm).

107

Data analysis

Principal components analysis (PCA) was conducted on pigment concentrations to identify variations over time (Legendre and Birks 2012). Canthaxanthin, echinenone, and zeaxanthin were identified as cyanobacterial pigments (Cook et al. 2016). To estimate total algal biomass, we summed two chemically-stable indicators (β-carotene and pheophytin a; Leavitt and Carpenter 1990). Detrended correspondence analysis

(DCA) with down-weighting of rare taxa was applied to the percent relative abundance of all diatom species to evaluate diatom community turnover down core (Hill and Gauch

1990). A square-root transformation to stabilize between-taxa variances was used prior to

DCA. Analyses were performed using the vegan package (Oksanen et al. 2013) in R version 3.4.2 (R Development Core Team 2017). The resulting species scores from DCA

Axis 1, plotted over time, indicate diatom community turnover between samples in units of standard deviation (Hobbs et al. 2011). We estimated diatom-inferred salinity (DI-S) using salinity optima from Fritz et al. (1991), Gasse et al. (1997), Gell (1997), and Gell et al. (2002). To reconstruct past lake level of Lake Purrumbete, we calculated the ratio of planktic:benthic (P:B) diatom relative abundance. Lower P:B in a lake with simple basin morphometry suggests a shallower lake level, as more light can reach benthic habitats

(Stone & Fritz 2004). Lower P:B may also reflect greater water clarity. We did not calculate P:B for Lake Bullen Merri because diatom assemblages after 1900 were dominated (~80%) by one taxon.

108

Results

Pelagic variables

Lake Bullen Merri was mesosaline with high DOC and low water clarity (Table

5.2). Lake Purrumbete was fresh with lower DOC and higher water clarity. The lakes were both alkaline, with similar epilimnion temperature, dissolved oxygen (DO), and Chl a. Based on the ratio of TN:TP, Bullen Merri was P-limited and Purrumbete was N- limited, although TN was very high and inorganic nitrogen was low in both lakes.

Surface levels of phycocyanin were greater in Bullen Merri than Purrumbete, and reactive silica was an order of magnitude lower in Bullen Merri than in Purrumbete.

Table 5.2. Pelagic characteristics of Lakes Bullen Merri and Purrumbete. Water samples collected from lake epilimnia during February 2017; bd = below detection. Silica data from Leahy et al. (2010).

Lake Bullen Merri Purrumbete Temperature (°C) 20 20 Secchi depth (m) 2 6 pH 9.5 9.3 DO (mg L-1) 9 9 Salinity (ppt) 10 0.4 DOC (mg L-1) 43 9 Chl a (µg L-1) 5 4 Ammonium (µg L-1) 8 9 Nitrate (µg L-1) bd bd Total nitrogen (µg L-1) 2,910 1,250 Soluble reactive phosphorus (µg L-1) 3 120 Total phosphorus (µg L-1) 21 140 Dissolved inorganic nitrogen (DIN):TP 0.4 0.1 TN:TP 140 9 Phycocyanin (RFU) 11 3 Reactive silica (mg L-1) 0.1-0.4 1.7-3.0

109

Core chronologies Both Lake Bullen Merri and Lake Purrumbete cores exhibited exponential declines in unsupported 210Pb activity with core depth (Figure 5.2). Unsupported 210Pb activity was contained within the top 28 cm for Bullen Merri and the top 7.75 cm for

Purrumbete. Average error in estimated dates was ± 14 years for Bullen Merri and ± 8 years for Purrumbete. The 210Pb dates extend from ~1883 to 2017 for the Bullen Merri core and ~1916 to 2017 for the Purrumbete core.

110

Figure 5.2. Stratigraphic plots of 210Pb activities and sediment chronologies in Lakes Bullen Merri (a, b) and Purrumbete (c, d). Error bars represent error in 210Pb-estimated dates.

111

Diatom profiles

After the turn of the 20th century, diatom species richness in Lake Bullen Merri was extremely low. Cyclotella choctawhatcheeana (Prasad) dominated the record, ranging from 89-100% relative abundance after 1913 (Figure 5.3). Prior to this shift, some benthic species were present in the assemblage. Relative abundance ranged from 7-

46% Stauroforma exiguiformis (Lange-Bertalot) Flower, Jones, and Round, 13-29%

Encyonema neogracile (Krammer), and 12-25% Amphora spp. Diploneis ovalis (Hilse)

Cleve, Staurosira construens var. venter (Ehrenberg) Hamilton, and Cocconeis placentula (Ehrenberg) were also present during this time at lower relative abundance.

DCA Axis 1 scores declined after the shift to C. choctawhatcheeana dominance in the early 1900s. Of the 33 taxa identified in the Bullen Merri record, 13 taxa had relative abundances ≥ 1% in at least one sample. DI-S reconstruction showed a substantial increase in salinity occurred starting ~1883, when C. choctawhatcheeana became dominant and C. placentula, S. construens var. venter, and S. exiguiformis declined in relative abundance. While absolute values of DI-S were higher than the current salinity in

Bullen Merri, the direction and magnitude of change in the DI-S trend are informative of general salinity patterns. There was little indication of frustule degradation throughout the core.

112

Figure 5.3. Relative abundances of selected diatom taxa, DCA Axis 1 in SD units, and diatom-inferred salinity (DI-S) for Lake Bullen Merri sediment core.

Diatom species richness was much greater in Lake Purrumbete. While 210Pb dating for this lake does not extend past the early 1900s when the diatom assemblage dramatically shifted in Bullen Merri, 26 of the 64 taxa present in the Purrumbete core had relative abundances ≥ 1% in at least one sample. Relative abundance of Aulacoseira granulata (Ehrenberg) Simonson was high (12-28%) throughout the core, and increased to 36-56% after 2008 (Figure 5.4). During the same period, Discostella stelligera (Cleve and Grunow) Houk and Klee declined to ~2% relative abundance after maintaining ~10% relative abundance from the beginning of the record. Centric diatoms Cyclotella meneghiniana (Kützing) and Lindavia ocellata (Pantocsek) increased to ~5% relative abundance during this period after appearing infrequently earlier in the record.

113

Simultaneously, Fragilariaceae declined in Purrumbete after 2008 and C. placentula increased to a maximum relative abundance of 8%. Pseudostaurosira brevistriata

(Grunow) Williams and Round occurred at 23-40% relative abundance earlier in the record, then dropped to 14% relative abundance after 2008. Similarly, S. construens var. venter and Staurosirella pinnata (Ehrenberg) Williams and Round declined in relative abundance at this time.

Contrary to the trend in the Lake Bullen Merri core, the DCA Axis 1 scores in the

Purrumbete record increased over time, with the greatest rate of increase occurring from the early 2000s to 2017. The P:B ratio reached its lowest values around 1915, 1945, and

2005, corresponding to the three most severe SE Australia droughts of the last century

(BOM 2018). The highest P:B ratio of the record occurred in 2015. While absolute DI-S values were higher than current salinity in Purrumbete, we note changes in magnitude and direction of the DI-S trend over time. The magnitude of DI-S change was lower in

Purrumbete than in Bullen Merri, however, diatoms in Purrumbete also responded to changes in salinity. DI-S peaked in 2010 and was also high throughout the 1940s and

1950s. The lowest DI-S values occurred in the 1930s and 1990s. There was limited valve degradation throughout the core, although some C. meneghiniana valves were slightly dissolved. Selective dissolution of C. meneghiniana valves was also detected in Lake

Purrumbete by Gasse et al. (1997), with implications for overrepresentation of the freshwater A. granulata in paleoenvironmental reconstructions.

114

Figure 5.4. Relative abundances of selected diatom taxa, planktic:benthic diatom ratio (P:B), DCA Axis 1 in SD units, and diatom-inferred salinity (DI-S) for Lake Purrumbete sediment core.

Algal pigments

Sediment cores from Lakes Bullen Merri and Purrumbete contained the same set of algal pigments, but pigment concentrations were generally higher in Bullen Merri.

PCA scores were strong predictors of algal change in Bullen Merri, as PCA Axis 1 scores explained 83% of the variance. Pigment loadings on PCA Axis 1 ranged from -0.87

(lutein-zeaxanthin) to -0.04 (alloxanthin; Table 5.3). The trend in PCA Axis 1 scores over time followed that of LOI. We were unable to separate the lutein-zeaxanthin peak, but this biomarker changed similarly over time to the cyanobacterial pigments canthaxanthin and echinenone, so it may be useful as a cyanobacterial proxy. Even with the likely

115

degradation that occurred in samples from 8-19 cm (approximately 1940-1997) during

HPLC analysis, the widespread decline in pigment concentration among algal pigments from the 1940s to 1990s matches the decline in LOI at this time (Figure 5.5). Because

LOI was not affected by HPLC analysis, this coherent decline in pigment concentration likely represents actual reductions in algal biomass. Most pigment concentrations were relatively high until the 1940s before declining to minimum levels between 1940 and

1990. Concentrations of all pigments reached a coherent peak from the late 1990s to the early 2000s. After the early 2000s, pigment concentrations, including cyanobacterial biomarkers, declined slightly, but remained high relative to pre-1990s concentrations.

During this time, PCA Axis 1 scores were greater than any other period in the record.

Mean organic content of the sediment, as estimated by LOI, was 25% in the period before

1900, 19% from 1900 to 1997, and 22% after 1997.

Table 5.3. PCA Axis 1 loadings for Lake Bullen Merri and Lake Purrumbete pigments.

Pigment Bullen Merri Purrumbete PCA Axis 1 PCA Axis 1 loading loading Diatoxanthin -0.09 -0.10 Alloxanthin -0.04 -0.16 Chl b -0.21 -0.25 Canthaxanthin -0.17 -0.05 Lutein-zeaxanthin -0.87 -0.20 Echinenone -0.24 0.35 Chl a -0.19 -0.78 Pheophytin a -0.19 -0.21 β-carotene -0.18 -0.29

116

Figure 5.5. Extracted pigment concentrations, percent LOI, and PCA Axis 1 scores in SD units in Lake Bullen Merri sediment core. Red bars indicate estimated algal pigment concentrations prior to possible degradation during analysis.

The concentration of most algal pigments in Lake Purrumbete was higher earlier in the record (~1890 to 1940), before declining after the 1940s (Figure 5.6). However, sample resolution for this sediment core was much coarser than for the Bullen Merri core, so we were unable to detect changes in algal pigments from 1997-2017, the period coinciding with coherent maxima in Bullen Merri pigment concentrations. Despite low resolution, we can infer that echinenone, which is a marker pigment for N. spumigena

(Engström-Öst et al. 2002), remained high after the 1940s, while other algal pigments and total algal biomass declined. Concentrations of the cyanobacterial pigments canthaxanthin and echinenone were approximately half of those measured in Bullen

Merri. The high concentrations of pigments in the sample at the top of the Purrumbete core may indicate unequal levels of pigment preservation between this sample and those earlier in the record, and should thus be interpreted with caution. Average sediment organic matter was 39% throughout much of the record, dropping to 31% in the most recent sample. PCA Axis 1 explained 60% of the variance in the record and closely

117

followed the LOI trend. Pigment loadings on PCA Axis 1 ranged from -0.78 (chlorophyll a) to 0.35 (echinenone; Table 5.3).

Figure 5.6. Extracted pigment concentrations, percent LOI, and PCA Axis 1 scores in SD units in Lake Purrumbete sediment core.

Discussion

We investigated two deep crater lakes in SE Australia—one with regular HABs

(Bullen Merri) and one without (Purrumbete). Our aims were to identify HAB occurrences in the lakes over the period covered by the sediment record and to use paleolimnological indicators in conjunction with contemporary lake sampling to better understand why HABs are prevalent in Lake Bullen Merri and not in Lake Purrumbete.

Our key findings were that diatoms in both Bullen Merri and Purrumbete were responsive to increased salinity resulting from drought, but that shifts in diatom assemblages were more pronounced in mesosaline Bullen Merri than fresh Purrumbete. Algal pigments from Bullen Merri sediments were generally higher during periods of drought. A period of low algal biomass from the 1940s-1990s may have resulted from higher precipitation and reduced light availability during this time, and/or the introduction of fish stocking in

Bullen Merri, with implications for algal biomass resulting from food web alterations.

118

Most pigments, including those of cyanobacterial taxa, reached peak concentrations during the period of severe drought from 1997-2010. Pigment concentrations were much lower in Purrumbete than Bullen Merri, but remained relatively high until 1947, the end of a drought and a period of high fish stocking in Purrumbete. Echinenone, a cyanobacterial marker, remained high after this time, while concentrations of other pigments decreased. We have provided evidence that within the past century, cyanobacterial taxa were present in both lakes, and that algae were responsive to environmental changes. However, differences in hydrological inputs and physicochemical variables between the lakes may be responsible for the disparity in recent HAB developments.

In Bullen Merri, drought and the resulting increases in salinity likely precipitated the low-diversity, planktonic C. choctawhatcheeana-dominated diatom assemblage after the turn of the 20th century. Changes in lake salinity can have outsized impacts on diatom diversity and community composition, and at high salinities, diatom diversity is usually low (Fritz 2013). In saline lakes, even small changes in the hydrological budget can be rapidly reflected by physicochemical and biological lake responses (Timms 2005). Since

1841, Bullen Merri lake level has declined 28 m (Leahy et al. 2010), and a severe drought occurred in SE Australia in 1914-1915 (Fluin et al. 2007). The low ratio of planktonic to benthic diatoms during this period may have partially resulted from slumping of the core.

After this time, benthic and tychoplanktic taxa including Amphora spp., Encyonema neogracile, Diploneis ovalis, and small Fragilarioids, were largely absent from the sediment record. Cyclotella choctawhatcheeana has a broader salinity tolerance (5-80 ppt) than the taxa it replaced, and a salinity optimum around 20 ppt (Prasad & Nienow

119

2006). This taxon is also stimulated by nutrient loading (Prasad et al. 1990), and often co- occurs in deep, hyposaline, alkaline lakes with N. spumigena (Oliva et al. 2008). By

1900, many of the native overstory trees surrounding Bullen Merri had been cleared for grazing (Joyce & Evans 1964), and now the lake is surrounded by exotic grasses with some tree belts. The swampland surrounding the lake has been drained, and it is farmed year-round (Jones et al. 2001). These land use changes occurring around 1900 may have enhanced nutrient loading to Bullen Merri, favoring growth of C. choctawhatcheeana.

The concentration of reactive silica in Bullen Merri is also relatively low and potentially limiting for some diatoms, which have an absolute requirement for silicon to build their cell walls (Lewin 1962). Si-limitation may have also led to lower diatom diversity, and

Si-driven competition may have allowed other phytoplankton groups to flourish. By

2100, the salinity of Bullen Merri, currently 10 ppt, is projected to increase to 20 ppt

(Kirono et al. 2009). High salinity paired with elevated nutrient inputs and low reactive silica could allow C. choctawhatcheeana, with its broad salinity and nutrient tolerances, to dominate the diatom assemblage in this lake.

The most prominent change in the diatom assemblage of Lake Purrumbete occurred around the turn of the 21st century, near the end of a drought that lasted from

1997-2010. Due to its hydrological features, namely inputs from groundwater and surface water, short-term fluctuations in precipitation and temperature have less impact on salinity and lake level in Lake Purrumbete than Bullen Merri, and the 1997-2010 drought resulted in a relatively small decrease in lake level and increase in salinity in Purrumbete

(Yihdego et al. 2016). The shift in the diatom assemblage in Purrumbete was marked by increases in A. granulata, C. meneghiniana, L. ocellata, and C. placentula relative

120

abundances concurrent with a decrease in D. stelligera. At the same time, relative abundance of Fragilarioids also decreased. Aulacoseira granulata is a freshwater species, while C. meneghiniana and C. placentula have higher salinity optima (Fritz et al. 1991;

Gell et al. 2002). The small Fragilarioids in Lake Purrumbete are typically found in fresh to oligosaline conditions (Gell 1997; Gell et al. 2002). In SE Australia, S. pinnata and S. construens var. venter are found in lower ranges of mesotrophic conditions, and the TP optimum for D. stelligera may be <20 µg L-1 (Tibby 2004). Optimum TP for Aulacoseira granulata, however, is >25 µg L-1, and C. placentula and C. meneginiana are also found in eutrophic conditions (Tibby 2004; Gell et al. 2005). At the time of lake water sampling in 2017, surface water TP was 140 µg L-1, therefore high TP in Purrumbete may partially explain recent changes in diatom assemblages. Purrumbete is surrounded by farmland, may be exposed to sewer effluent, and is connected to some natural and manmade drainage channels (EPA Victoria 2010; Yihdego 2010), but overall, nutrient fluxes in this lake are thought to be dominated by internal cycling (Herpich 2002). Macrophytes are abundant in Lake Purrumbete, and the sediment has a high organic content with abundant plant debris (Cadwallader & Eden 1981; Timms 1976). During the period of stratification, dissolved oxygen in the hypolimnion often falls below detectable limits

(Timms 1976), so it is possible that release of P from anoxic sediments may have contributed to high levels of water column P available for diatom uptake.

Cyclotella taxa are controlled by multiple factors in addition to nutrient availability, dependent on several lake variables (Saros & Anderson 2015). For example,

D. stelligera was more abundant at shallow (<5 m) mixing depths in lakes with moderate and high nutrient levels (Saros et al. 2012). Strong thermal stratification in lakes often

121

favors small planktonic taxa, like D. stelligera (Saros et al. 2016), while heavily silicified taxa are more common in well-mixed lakes with wind-induced turbulence, likely because they are better able to maintain their position within the photic zone (Rühland et al.

2003). Stone et al. (2016) used a diatom-inferred stratification index in which

Aulacoseira spp. indicated deeper mixing than smaller Cyclotelloids. Because fresh water is more easily mixed than saline water, and Purrumbete has a longer fetch and is more wind-exposed than Bullen Merri, Purrumbete has a deeper mixing depth (Timms 1976).

Purrumbete is slightly larger than 5 km2, and so mixing depth may primarily be controlled by wind strength in this lake (Fee et al. 1996). However, since the late 2000s, temperature in SE Australia has been anomalously high, which may contribute to greater thermal stability. Even so, recent wildfire activity anecdotally attributed to hot, dry, windy conditions (Tennison et al. 2018) may provide evidence that wind contributed to deeper mixing in Purrumbete. However, we do not have local wind data to confirm that wind speed has changed in the last decade. Increased wildfires since 2003 were followed by major flooding in 2011 and 2012, and severe fires have occurred in the years since

(Nyman et al. 2011; Hurley et al. 2018). Entrapment of burned vegetative debris by flash floods produced large erosion events and increased run-off (Nyman et al. 2011), which may have elevated nutrient loading to lakes in this area. We suggest that nutrient loading in N-limited Lake Purrumbete paired with changes in lake thermal structure contributed to recent changes in diatom community structure.

Trends in algal pigments in Lake Bullen Merri correspond to periods of drought as well as stocking of the Chinook salmon fishery in the lake. In SE Australia, the period from 1900-1950 was very dry, followed by wetter decades from the 1950s-1970s. The

122

early 1990s were also wet, followed by extreme drought and high temperatures from

1997-2010 (Leahy et al. 2010). Some of the driest years were 1915, 1945, and 2005

(BOM 2018). In Bullen Merri, algal pigment concentrations were generally higher during periods of drought. While precipitation transports nutrients to lakes, it can also reduce transparency (Rose et al. 2017). Additionally, internal fluxes dominate nutrient cycling in eutrophic Bullen Merri (Herpich 2002), so allochthonous nutrient input may not be a primary regulator of algal biomass. Instead of limiting algal biomass, droughts may have stimulated phytoplankton growth by improving light conditions, particularly for benthic taxa, via lake level reduction and increased photic zone habitat in this low-clarity system.

Concentrations of algal pigments reached peak levels during the 1997-2010 drought, suggesting that elevated salinity was not prohibitive to growth of various phytoplankton taxa.

The effect of zooplankton grazing is also an important control of algal biomass.

Prior to the 1940s, concentrations of many of the algal pigments in Bullen Merri were relatively high. The lake was stocked with Chinook salmon from 1936-1952, and regular stocking resumed in 1978 (Cadwallader & Eden 1981; Nguyen et al. 2012). After 2007, stocking was paused due to complications from the drought, but continued in 2012 and

2016 (“Chinook Salmon Fisheries Return” 2018). Chinook salmon feed primarily on galaxiids, which eat insects, molluscs, and crustaceans such as Daphnia (Hunt et al.

2014). Daphnia are generalist grazers of algae (Boon et al. 1994), and have been found in

Lake Bullen Merri (De Deckker 1982), so increased Daphnia populations resulting from fish stocking may have contributed to the reductions in algal pigments from the 1940s-

1980s. Salmon production and stocking suffered during the 1997-2010 drought

123

(“Chinook Salmon Fisheries Return” 2018), and so grazing pressure may have been partially released, resulting in increased concentrations of algal pigments. Reduced lake level and increased light availability paired with elevated nutrient loading from land use pressures (i.e. agriculture and tourism; Victoria DELWP 2017), may have improved conditions for phytoplankton growth as well.

In Lake Purrumbete, concentrations of most algal pigments were high until the

1940s, which may reflect high algal biomass during drought conditions, and/or the introduction of Chinook salmon as described for Lake Bullen Merri. After this time, algal pigments declined in the sediment record, with the exception of echinenone, a biomarker for N. spumigena (Engström-Öst et al. 2002), as well as some other cyanobacterial taxa.

After 1947, the end of a drought in SE Australia and a period of high levels of salmon stocking, the primary cyanobacterial pigment switched from canthaxanthin to echinenone, which remained elevated (although at much lower levels than in Bullen

Merri) until 1998, after which time sample resolution was low. Daphnia spp. has been described in Lake Purrumbete (Cadwallader & Eden 1981; De Deckker 1982), and as in

Bullen Merri, salmon stocking may have resulted in greater abundance of algal grazers like Daphnia in the lake. While Daphnia are generalists, they may preferentially consume cryptophytes (alloxanthin) and diatoms (diatoxanthin; Infante & Litt 1985) to N. spumigena, which lacks long-chain fatty acids and is a poorer-quality food source

(Holland et al. 2012). Zooplankton grazing also liberates dissolved organic matter

(DOM) to the water column (Møller 2005), which can stimulate N. spumigena growth

(Stolte et al. 2006). Furthermore, N. spumigena is an N-fixer, and can outcompete other phytoplankton taxa in N-limited systems (Holland et al. 2012). Zooplankton grazing

124

paired with N-fixation by N. spumigena may explain why echinenone remained high in the sediment record while concentrations of other pigments were low after 1947. Because sample resolution is low after 1998, we are unable to speculate about the effects of the

1997-2010 drought, fish stocking, and/or land use change on concentrations of algal pigments, including cyanobacterial biomarkers, after this time.

We found evidence of HAB-forming cyanobacteria in both Lake Bullen Merri and Lake Purrumbete, but concentrations of HAB biomarkers were higher in Bullen

Merri, especially in recent decades. McGregor et al. (2012) described the difficulty of assigning causative factors to HAB development, but suggested that drought, salinity, increased water retention time, and elevated nutrient status all contributed to blooms in

Australian lakes. In a temperate Australian lagoon system, high phosphorus supply combined with long water residence time and intermediate salinity (5-20 ppt) likely favored N. spumigena blooms (Cook & Holland 2012; Cook et al. 2016). While high TP and SRP in Purrumbete may promote HAB development, water residence time and salinity are much lower in Purrumbete than Bullen Merri. The differences in HAB occurrence between Bullen Merri and Purrumbete are likely due to contrasting hydrological and physicochemical parameters—despite the proximity of the lakes, drought effects on lake level and salinity fluctuations have been much greater in Bullen

Merri, while Purrumbete remains relatively buffered from short-term climate effects

(Yihdego 2016). However, our results showed diatom-inferred salinity and the diatom

P:B ratio in Purrumbete were responsive to environmental changes throughout the past century. Water retention time is longer in Bullen Merri, and less wind exposure and stronger thermal stratification likely encourage bloom formation, as cyanobacteria rely on

125

strong stratification for buoyancy control within the photic zone (Holland et al. 2012).

Droughts are predicted to have limited influence on lake level and salinity in Purrumbete compared to Bullen Merri, where lake level is dropping 10-20 cm yr-1 (Kirono et al.

2009; Yihdego et al. 2016; Victoria DELWP 2017). Therefore, we expect that N. spumigena blooms are less likely to develop in Purrumbete over the next century, even though this species may be present in the phytoplankton community, as evidenced by the presence of echinenone in the sediment record.

While it may be difficult to address many of the drivers of N. spumigena blooms in Lake Bullen Merri, such as drought, salinity, and water residence time, efforts to weaken thermal stratification may provide effective bloom prevention, thereby reducing threats to human health and minimizing the frequency and duration of lake closures.

Currently, elevated nutrient inputs are thought to be responsible for recent HABs

(Victoria DELWP 2017; “Blue-green disgrace” 2019), and methods for reducing nutrient loading, including native vegetation restoration, monitoring of nearby septic systems, and fuel reduction burns, are being considered (Victoria DELWP 2017). While trophic state may contribute to HABs in Bullen Merri, N. spumigena, the dominant bloom-forming species, is an N-fixer and often outcompetes other phytoplankton taxa even in low- nutrient conditions (Holland et al. 2012). Indeed, DIN (available for uptake by most phytoplankton) is already low in Bullen Merri. Consequently, in addition to efforts to reduce nutrient inputs, we suggest that an artificial aeration system could help weaken thermal stratification and limit HAB frequency and/or severity. A mechanical pumping system was installed in Bullen Merri in 1982, and water quality was thought to have improved as a result of weakened stratification (Stoessel 2009). However, vandalism,

126

expense, and inefficient operation reduced the system’s effectiveness, and episodic blooms resulted. Herpich (2002) suggested that positioning, size, and operation of the aeration system should be re-evaluated to improve its efficiency. We concur that a more efficient aeration system would reduce HAB formation in Lake Bullen Merri by weakening thermal stratification, and that the benefits of reduced lake closure time and health risks would outweigh associated costs of closures for this popular tourism destination and trophy fishery. Additionally, we suggest that conservation efforts should prioritize preservation of the ecological character of Lake Purrumbete, which may serve as a refugium for aquatic organisms in future decades.

127

CHAPTER 6

CONCLUSIONS

DOC dynamics in Arctic and boreal systems

The objectives of this dissertation were to uncover drivers of recent DOC quantity and quality change in a set of Arctic lakes and to interpret the effects of changing lakewater DOC on algal responses in a clear versus a brown lake recovering from atmospheric acid deposition in Acadia National Park. This work demonstrated that DOC quantity and quality in northern hemisphere (Arctic and Maine) lakes are under complex control of climate, landscape, and lake features. Climate drivers affected terrestrial- aquatic linkages in west Greenland lakes, and catchment and lake characteristics filtered these climate signals so that regional responses by lake variables displayed coherent temporal trends. In lakes of ANP experiencing increased DOC, algal responses to recovery from acidification and increased temperature and precipitation differed in a clear vs. a brown lake, showing that similar external pressures can result in different ecological responses in lakes of varying color. These findings highlight control of climate and atmospheric drivers on DOC and algal ecology as well as the role of catchment and in-lake characteristics for modification of these signals. With ongoing abrupt climate change (ACC) in the Arctic and widespread browning of lakes at mid-latitudes, my results provide a timely snapshot of climate effects on carbon cycling in northern hemisphere lakes with examples of algal responses in lakes experiencing changing DOC.

In Chapter 2, I found that even though recent declines in DOC concentration of west Greenland lakes were not explained by the tested mechanisms (dust inputs, bacterial activity, or light), several of these mechanisms were important for DOC composition, and

128

that controls of DOC quality may work in opposite directions and at different magnitudes depending on lake properties and the initial character of DOC in the lakes. Furthermore,

DOC pools in water samples from these lakes were highly photoreactive despite long water residence times of the lakes. The results of this study contribute to descriptions of variability in DOC concentration and composition across Arctic lakes and further our understanding of how future climate-mediated changes may influence DOC concentration and quality in these lakes.

In Chapter 3, I provide evidence for strong coherence of interannual lake responses (DOC concentration and quality, chlorophyll a, and light attenuation metrics) associated with climate trends (mean annual precipitation, spring average temperature, spring average precipitation, and the winter North Atlantic Oscillation (NAO) index).

Notably, mean annual precipitation was associated with increases in soil-derived DOC quality metrics and chlorophyll a, as well as decreases in depth of 1% photosynthetically active radiation (PAR). These associations were stronger in the warmer and dryer of the two lake regions, and interannual coherence of DOC and 1% PAR was higher in this region as well. Regional differences were likely caused by variations in climate exposure and catchment area : lake area ratios. With ongoing ACC in the Arctic, these results suggest that there will be widespread impacts on key lake responses that are important for light attenuation and C cycling.

To compare ecological responses to reduced acid deposition in a clear vs. a brown lake in ANP, I reconstructed fossil diatom and algal pigment records from sediment cores in both lakes. The results of Chapter 4 show that during the period of high atmospheric deposition, light availability was important for algal ecology in the clear lake, while

129

nutrient subsidies from allochthonous DOC were the primary control of algal biomass in the brown lake. While diatom community structure in the clear lake changed continually from the beginning of the record, coincident with the end of the Little Ice Age, the largest shift in diatom community structure in the brown lake occurred after the implementation of the 1990 Clean Air Act Amendments. These findings suggest that the clear lake was more responsive to long-term warming. After 1990, both lakes showed signs of recovery toward pre-acidification conditions, but these changes were dampened in the clear lake, suggesting greater sensitivity to acid deposition and other external drivers such as effects of climate change.

The Arctic is experiencing rapid, nonlinear climate change unlike any other region in the world and the Arctic and sub-arctic contain almost one-quarter of the world’s lake area. Because these lakes contribute to global carbon emissions and bury carbon in their sediments, our understanding of drivers of carbon cycling in these lakes is a key component of global carbon budgets. However, Arctic lakes are highly variable and understudied. My research demonstrated key connections between climate and lake variables, showing temporally coherent responses of Arctic lakes to climate drivers across regions. This work helps fill knowledge gaps about climate effects on carbon cycling and lake ecology in the Arctic. My dissertation research also contrasted ecological responses of a clear vs. a brown lake in Maine to multiple drivers, including increased DOC concentration resulting from acidification recovery and climate change.

With ongoing brownification of lakes in northeastern North America and Europe, the results of this work provide timely context for potential ecological consequences of browning in lakes, particularly those recovering from acidification.

130

The results of this research highlight the need for broader reach of long-term monitoring programs, particularly in the Arctic. While I demonstrated strong temporal coherence of regional lake responses to climate drivers in west Greenland lakes, this coherence broke down when catchment and lake characteristics displayed greater heterogeneity. DOC concentration and quality are highly variable in Arctic lakes, and this research underlined the complexity of DOC controls in these lakes. To better understand mechanisms of DOC change in west Greenland lakes, future work could include three- factor experiments that test the effects of light x dust inputs x bacterial activity.

Additionally, increased delivery of sulfur to lakes via atmospheric, marine, and volcanic inputs could provide a mechanism for DOC loss and should be investigated further. I also recommend continuation and further development of long-term monitoring programs for

Arctic lakes. Decades-long monitoring programs at mid-latitudes have provided the fields of limnology and ecology with valuable data allowing for high-quality statistical inferences of climate effects on physical, chemical, and biological features of lakes. As

Arctic lakes are highly variable in morphometry, hydrology, and chemistry, implementation of long-term monitoring across a wider array of Arctic sites would expand our understanding of climate controls on carbon cycling and algal ecology in

Arctic lakes.

Finally, my research contributed to the developing body of work showing that lake color is an important control of algal responses to multiple drivers like recovery from acidification and climate change. However, lake morphometry and watershed characteristics also contribute to the direction and scale of these responses. As with

Arctic lakes, it will be important to consider a broad range of lake types when evaluating

131

how brownification will affect lake ecology in future scenarios, and this field of study would benefit from extended monitoring. Jordan Pond, the clearest monitored lake in

ANP and in the state of Maine, has been equipped with a high-frequency water quality monitoring buoy for the duration of each open water season since 2013. Installation of a similar system in Seal Cove Pond, the brown lake in this study, would provide a real-time comparison of DOC, chlorophyll a, and water clarity responses to storm events, for example. These sorts of high-resolution monitoring programs paired with paleolimnological records are invaluable resources for understanding how the pressures of ACC and other drivers affect lake responses within the context of past environmental change.

IGERT Collaborative Immersion Project (CIP) Experience

Drought and temperatures in southeast Australia are projected to increase in coming decades, and it is likely that deep crater lakes in this region may be the only lakes to persist through the end of this century. As the effects of climate change intensify,

HABs across the region have become more severe in duration and frequency. The primary findings of Chapter 5 suggest that annual HABs in a deep crater lake of southeast

Australia prized for its Chinook salmon fishery are likely caused by drought conditions and subsequent increased salinity combined with long water residence time, high nutrients, and strong thermal stratification. These conditions will likely continue despite conservation efforts, but installation of an aeration system may weaken thermal stratification in this lake and reduce the frequency and duration of HABs. The deep crater lake free of HABs is fresh, with greater input from surface and groundwater flow and weaker thermal stratification. Despite high nutrients, the lake will likely remain HAB-

132

free, and as it could act as a refugium for aquatic organisms, conservation efforts should focus on preservation of hydrological connections and maintenance of biodiversity in this lake. Studying these Australian lakes for my CIP provided valuable experience working in a different lake and human stressor system, and comparison of this system with Arctic and boreal systems enhanced the academic breadth of my adaptation to abrupt climate change (ACC) work.

133

REFERENCES

American Public Health Association (APHA), 2000. Standard methods for the examination of water and wastewater, 20th ed. Washington, DC, USA.

Anderson, N. J., and Stedmon, C., 2007. The effect of evapoconcentration on dissolved organic carbon concentration and quality in lakes of SW Greenland. Freshwater Biology 52: 280-289, doi:10.1111/j.1365-2427.2006.01688.x

Anderson, N. J., D'Andrea, W., and Fritz, S. C., 2009. Holocene carbon burial by lakes in SW Greenland. Global Change Biology, 15(11): 2590-2598, doi:10.1111/j.1365- 2486.2009.01942.x

Anderson, N. J., Harriman, R., Ryves, D. B., and Patrick, S. T., 2001. Dominant factors controlling variability in the ionic composition of west Greenland lakes. Arctic, Antarctic, and Alpine Research, 33(4): 418–425, doi:10.1080/15230430.2001.12003450

Anderson, N. J., Saros, J. E., Bullard, J. E., Cahoon, S. M. P., McGowan, S., Bagshaw, E. A., Barry, C. D., Bindler, R., Burpee, B. T., Carrivick, J. L. Fowler, R. A., Fox, A. D., Fritz, S. C., Giles, M. E., Hamerlik, L., Ingeman-Nielsen, T., Law, A. C., Mernild, S. H., Northington, R. M., Osburn, C. L., Pla-Rabés, S., Post, E., Telling, J., Stroud, D. A., Whiteford, E. J., Yallop, M. L. and Yde, J. C., 2017. The Arctic in the twenty-first century: Changing biogeochemical linkages across a paraglacial landscape of Greenland. BioScience 67(2): 118-133, doi:10.1093/biosci/biw158

Appleby, P. G. and Oldfield, F., 1978. The calculation of lead-210 dates assuming a constant rate of supply of unsupported 210Pb to the sediment. Catena, 5: 1-8, doi:10.1016/S0341-8162(78)80002-2

Arrigo, K. R., van Dijken, G., and Pabi, S., 2008. Impact of a shrinking Arctic ice cover on marine primary production. Geophysical Research Letters, 35(19): L19603, doi:10.1029/2008GL035028

Atech, 2000. Cost of algal blooms. Land and Water Resources Research and Development Corporation (LWRRDC) Occasional Paper 26/99. Land and Water Resources Research and Development Corporation, Canberra, ACT, Australia, 42 pp.

Baddock, M. C., Mockford, T., Bullard, J. E., and Thorsteinsson, T., 2017. Pathways of high-latitude dust in the North Atlantic. Earth and Planetary Science Letters, 459: 170- 182, doi:10.1016/j.epsl.2016.11.034

Baines, S., Webster, K., Kratz, T., Carpenter, S., and Magnuson, J., 2000. Synchronous behavior of temperature, calcium, and chlorophyll in lakes of northern Wisconsin. Ecology, 81(3): 815-825, doi:10.1890/0012-9658(2000)081[0815:SBOTCA]2.0.CO;2

134

Baker, P.D. and Humpage, A.R., 1994. Toxicity associated with commonly occurring cyanobacteria in surface waters of the Murray-Darling Basin, Australia. Australian Journal of Marine and Freshwater Research, 45: 773-786, doi:10.1071/MF9940773

Ballantyne, A. P., Brahney, J., Fernandez, D., Lawrence, C. L., Saros, J. E., and Neff, J. C., 2011. Biogeochemical response of alpine lakes to a recent increase in dust deposition in the Southwestern, US. Biogeosciences, 8: 2689–2706, doi:10.5194/bg-8-2689-2011

Baron, J. and Caine, N., 2000. Temporal coherence of two alpine lake basins of the Colorado Front Range, U.S.A. Freshwater Biology, 43: 463-476, doi:10.1046/j.1365- 2427.2000.00517.x

Battarbee, R. W., Jones, V. J., Flower, B. P., Cameron, N. G., Bennion, H., Carvalho, L., and Juggins, S., 2001. Diatoms. In: Smol J, Birks HJB, Last WM (eds) Tracking environmental change using lake sediments, vol. 3, Terrestrial, Algal, and Siliceous Indicators. Kluwer, Dordrecht, pp 155-201.

Benson, B., Lenters, J., Magnuson, J., Stubbs, M., Kratz, T., Dillon, P., Hecky, R., and Lathrop, R., 2000. Regional coherence of climatic and lake thermal variables of four lake districts in the Upper Great Lakes Region of North America. Freshwater Biology, 43: 517-527, doi:10.1046/j.1365-2427.2000.00572.x

Berger, S. A., Diehl, S., Stibor, H., Trommer, G., and Ruhenstroth, M., 2010. Water temperature and stratification depth independently shift cardinal events during plankton spring succession. Global Change Biology, 16: 1954-1965. doi: 10.1111/j.1365- 2486.2009.02134.x

Blenckner, T., 2005. A conceptual model of climate-related effects on lake ecosystems. Hydrobiologia, 533: 1-14, doi:10.1007/s10750-004-1463-4

Blue-green disgrace. (2019, January 17). Camperdown Chronicle. Retrieved from http://www.camperdownchronicle.com.au/2019/01/blue-green-disgrace/

Boeff, K. A., Strock, K.E., and Saros, J. E., 2016. Evaluating planktonic diatom response to climate change across three lakes with differing morphometry. Journal of Paleolimnology 56: 33-47, doi:10.1007/s10933-016-9889-z

Boon, P. I., Bunn, S. E., Green, J. D., and Shiel, R. J., 1994. Consumption of cyanobacteria by freshwater zooplankton: Implications for the success of ‘top-down’ control of cyanobacterial blooms in Australia. Australian Journal of Marine Freshwater Research, 45: 875-887, doi:10.1071/MF9940875

Bosson, E., Lindborg, T., Berglund, S., Gustafsson, L. G., Selroos, J. O., Laudon, H., Claesson, L. L., and Destouni, G., 2013. Water balance and its intra-annual variability in a permafrost catchment: hydrological interactions between catchment, lake and talik. 135

Hydrology and Earth System Sciences Discussions, 10: 9271-9308, doi:10.5194/hessd- 10-9271-2013

Bring, A., Fedorova, I., Dibike, Y., Hinzman, L., Mård, J., Mernild, S. H., Prowse, T., Semenova, O., Stuefer, S. L., and Woo, M.-K., 2016. Arctic terrestrial hydrology: A synthesis of processes, regional effects, and research challenges. Journal of Geophysical Research: Biogeosciences, 121: 621-649, doi:10.1002/2015JG003131

Brodersen, K. and Anderson, N. J., 2000. Subfossil insect remains (Chironomidae) and lake-water temperature inference in the Sisimiut-Kangerlussuaq region, southern West Greenland. Geology of Greenland Survey Bulletin, 186, 78–82.

Brophy, J. E., and Carlson, D. J., 1989. Production of biologically refractory dissolved organic carbon by natural seawater microbial populations. Deep Sea Research Part A. Oceanographic Research Papers, 36(4): 497-507, doi:10.1016/0198-0149(89)90002-2

Brown, R. E., Nelson, S. J., and Saros, J. E., 2017. Paleolimnological evidence of the consequences of recent increased dissolved organic carbon (DOC) in lake of the northeastern USA. Journal of Paleolimnology, 57: 19-35, doi:10.1007/s10933-016-9913- 3

Bullard, J. E., 2013. Contemporary glacigenic inputs to the dust cycle. Earth Surface Processes and Landforms, 38(1): 71-89, doi:10.1002/esp.3315

Bullard, J.E., and Austin, M. J., 2011. Dust generation on a proglacial floodplain, West Greenland. Aeolian Research, 3(1): 43-54, doi:10.1126/science.1217447

Bureau of Meteorology (BOM) and Commonwealth Scientific and Industrial Research Organisation (CSIRO), 2014. State of the Climate 2014.

Bureau of Meteorology (BOM), 2018. Archive- Twelve-monthly rainfall deciles for Australia. Retrieved from http://www.bom.gov.au/jsp/awap/rain/archive.jsp?colour=colour&map=decile&year=201 3&month=12&period=12month&area=nat

Burpee, B., Saros, J. E., Northington, R. M., and Simon, K. S., 2016. Microbial nutrient limitation in Arctic lakes in a permafrost landscape of southwest Greenland. Biogeosciences, 13: 365-374, doi:10.5194/bg-13-365-2016

Cadwallader, P. L. and Eden, A. K., 1981. Food and growth of hatchery-produced chinook salmon, Oncorhynchus tshawytscha (Walbaum), in landlocked Lake Purrumbete, Victoria, Australian Journal of Fish Biology, 18: 321-330, doi:10.1111/j.1095- 8649.1981.tb03773.x

136

Camburn, K. E. and Charles, D. F., 2000. Diatoms of low-alkalinity lakes in the Northeastern United States. Academy of Natural Sciences Philadelphia, Special Publication 18, 152 pp.

Cappelen, J. (ed), In press. Weather observations from Greenland 1958-2018: Observation data with description. DMI Report 19-08.

Carlsson, P., and Caron, D. A., 2001. Seasonal variation of phosphorus limitation of bacterial growth in a small lake. Limnology and Oceanography 46(1): 108-120, doi:10.4319/lo.2001.46.1.0108

Charles, D. F., Binford, M. W., Furlong, E. T., Hites, R. A., Mitchell, M. J., Norton, S. A., Oldfield, F., Paterson, M. J., Smol, J. P., Uutala, A. J., White, J. R., Whitehead, D. R., and Wise, R. J., 1990. Paleoecological investigation of recent lake acidification in the Adirondack Mountains, N.Y. Journal of Paleolimnology, 3: 195-241, doi:10.1007/BF00219459

Chen, N., Bianchi, T. S., McKee, B. A., and Bland, J. M., 2001. Historical trends of hypoxia on the Louisiana shelf: application of pigments as biomarkers. Organic Geochemistry, 32: 543-561, doi:10.1016/S0146-6380(00)00194-7

Chinook Salmon Fisheries Return, 2018. Victoria Fisheries Authority. Retrieved from https://vfa.vic.gov.au/science-in-fisheries/featured/chinook-salmon-stocking-returns

Clair, T. A., Dennis, I. F., and Vet, R., 2011. Water chemistry and dissolved organic carbon trends in lakes from Canada’s Atlantic Provinces: no recovery from acidification measured after 25 years of lake monitoring. Canadian Journal of Fisheries and Aquatic Sciences, 68: 663-674. doi:10.1139/F11-013

Cole, J., Prairie, Y., Caraco, N., McDowell, W., Tranvik, L., Striegl, R., et al., 2007. Plumbing the global carbon cycle: Integrating inland waters into the terrestrial carbon budget. Ecosystems, 10, 171-184. doi:10.1007/s10021-006-9013-8.

Commonwealth Scientific and Industrial Research Organisation (CSIRO), 2010. Climate variability and change in south-eastern Australia: A synthesis of findings from Phase 1 of the South Eastern Australian Climate Initiative (SEACI).

Cook, P. L. M. and Holland, D. P., 2012. Long term nutrient loads and chlorophyll dynamics in a large temperate Australian lagoon system affected by recurring blooms of cyanobacteria. Biogeochemistry, 107: 261-274, doi:10.1007/sl0533-010-955 1-1

Cook, P. L. M., Holland, D. P., and Longmore, A. R., 2010. Effect of a flood event on the dynamics of phytoplankton and biogeochemistry in a large temperate Australian lagoon. Limnology and Oceanography, 55: 1123-1133, doi:10.4319/lo.2010.55.3.1123

137

Cook, P. L. M., Jennings, M., Holland, D. P., Beardall, J., Briles, C., Zawadzki, A., Doan, P., Mills, K., and Gell, P., 2016. Blooms of cyanobacteria in a temperate Australian lagoon system post and prior to European settlement. Biogeosciences, 13: 3677-3686, doi:10.5194/bg-13-3677

Cory, R. M., Ward, C. P., Crump, B. C., and Kling, G. W., 2014. Sunlight controls water column processing of carbon in arctic fresh waters. Science, 345(6199): 925-928, doi: 10.1126/science.1253119

Cremer, H., Wagner, B., Melles, M., and Hubberton, H.-W., 2001. The postglacial environmental development of Raffles Sø, East Greenland: inferences from a 10,000 year diatom record. Journal of Paleolimnology, 26: 67-87, doi: 10.1023/A:1011179321529

Curtis, P. and Adams, H., 1995. Dissolved organic matter quantity and quality from freshwater and saltwater lakes in east-central Alberta. Biogeochemistry, 30(1): 59-76, doi:10.1007/BF02181040

Daggett, C. T., Saros, J. E., Lafrancois, B. M., Simon, K. S., and Amirbahman, A., 2015. Effects of increased concentrations of inorganic nitrogen and dissolved organic matter on phytoplankton in boreal lakes with differing nutrient limitation patterns. Aquatic Sciences, 77: 511-521. doi: 10.1007/s00027-015-0396-5

Davis, R. B., 1987. Paleolimnological diatom studies of acidification of lakes by acid rain: an application of quaternary science. Quaternary Science Reviews, 6: 147-163. doi: 10.1016/0277-3791(87)90031-X

Davis, R. B., Anderson, D. S., Norton, S. A., and Whiting, M.C., 1994. Acidity of twelve northern New England (U.S.A.) lakes in recent centuries. Journal of Paleolimnology, 12: 103-154. doi: 10.1007/BF00678090

Davis, R. B., Norton, S. A., Hess, C. T., and Brakke, D. F., 1983. Paleolimnological reconstruction of the effects of atmospheric deposition of acids and heavy metals on the chemistry and biology of lakes in New England and Norway. Hydrobiologia, 103: 113- 123. doi: 10.1007/BF00028438

De Deckker, P., 1982. Holocene ostracods, other invertebrates and fish remains from cores of four maar lakes in southeastern Australia. Proceedings of the Royal Society of Victoria 94.4: 183-220.

Dillon, P. J. and Molot, L. A., 1997. Effect of landscape form on export of dissolved organic carbon, iron, and phosphorus from forested stream catchments. Water Resources Research, 33(11): 2591-2600, doi:10.1029/97WR01921

138

Donahue, W. F., Schindler, D. W., Page, S. J., and Stainton, M. P., 1998. Acid-induced changes in DOC quality in an experimental whole-lake manipulation. Environmental Science and Technology, 32: 2954-2960. doi:10.1021/es980306u

Driscoll, C. T., Driscoll, K. M., Roy, K. M., and Mitchell, M. J., 2003. Chemical responses of lakes in the Adirondack region of New York to declines in acid deposition. Environmental Science and Technology, 37: 2036-2042. doi:10.1021/es020924h

Ducklow, H. W., Purdie, D. A., Williams, P. J. L., and Davies, J. M., 1986. Bacterioplankton: a sink for carbon in a coastal marine plankton community. Science, 232(4752): 865-867, doi:10.1126/science.232.4752.865

Eastman, R. and Warren, S. G., 2013. A 39-yr survey of cloud changes from land stations worldwide 1971–2009: Long-term trends, relation to aerosols, and expansion of the Tropical Belt. Journal of Climate, 26: 1286–1303, doi:10.1175/JCLI-D-12-00280.1

Engels, S., 2003. Dust in the wind: modern and historical aeolian deposition rates in an uphill lake catchment in the Kangerlussuaq region, West Greenland. Internal Report, Faculty of Geographical Sciences, Utrecht University.

Engström-Öst, J., Koski, M., Schmidt, K., Viitasalo, M., Jónasdóttir, S. H., Kokkonen, M., Repka, S., and Sivonen, K., 2002. Effects of toxic cyanobacteria on a plankton assemblage: community development during decay of Nodularia spumigena. Marine Ecology Progress Series, 232: 1-14, doi:10.3354/meps232001

EPA, 2010. Environmental quality guidelines for Victorian Lakes. EPA Victoria (EPA Publication 1302).

Falconer, I. R., 2001. Toxic cyanobacterial bloom problems in Australian waters: risks and impacts on human health. Phycologia, 40(3): 228-233, doi:10.2216/i0031-8884-40-3- 228.1

Farjalla, V. F., Azevedo, D. A., Esteves, F. A., Bozelli, R. L., Roland, F., and Enrich- Prast, A., 2006. Influence of hydrological pulse on bacterial growth and DOC uptake in a clear-water Amazonian lake. Microbial Ecology, 52: 334-344, doi:10.1007/s00248-006- 9021-4

Fee, E. J., Hecky, R. E., Kasian, S. E. M., and Cruikshank, D. R., 1996. Effects of lake size, water clarity, and climatic variability on mixing depths in Canadian Shield lakes. Limnology and Oceanography, 41: 912-920, doi:10.4319/lo.1996.41.5.0912

Fichot, C. G., and Benner, R., 2012. The spectral slope coefficient of chromophoric dissolved organic matter (S275–295) as a tracer of terrigenous dissolved organic carbon in river-influenced ocean margins. Limnology and Oceanography, 57(5): 1453-1466, doi:10.4319/lo.2012.57.5.1453

139

Fluin, J., Gell, P., Haynes, D., Tibby, J., and Hancock, G., 2007. Palaeolimnological evidence for the independent evolution of neighbouring terminal lakes, the Murray Darling Basin, Australia. Hydrobiologia, 591: 117-134, doi: 10.1007/s10750-007-0799-y

Fox, L. E. and Wofsy, S. C., 1983. Kinetics of removal of iron colloids from estuaries. Geochimica et Cosmochimica Acta, 47: 211-216, doi:10.1016/0016-7037(83)90134-5

Francko, D. A. and Heath, R. T., 1982. UV-sensitive complex phosphorus: association with dissolved humic material and iron in a bog lake. Limnology and Oceanography, 27: 564-569, doi:10.4319/lo.1982.27.3.0564

Frey, K. E., and Smith, L. C., 2005. Amplified carbon release from vast west Siberian peatlands by 2100. Geophysical Research Letters, 32: L09401, doi:10.1029/2004GL022025.

Fritz, S. C., 2013. Salinity and climate reconstructions from continental lakes, p. 507-515. In: S. A. Elias and C. J. Mock, (eds), Encyclopedia of Quaternary Science, 2nd ed.

Fritz, S. C., Juggins, S., Battarbee, R. W., and Engstrom, D. R., 1991. Reconstruction of past changes in salinity and climate using a diatom-based transfer function. Nature, 352: 706-708, doi:10.1038/352706a0

Gasse, F., Barker, P., Gell, P. A., Fritz, S. C., and Chalié, F., 1997. Diatom-inferred salinity in palaeolakes: An indirect tracer of climate change. Quaternary Science Reviews, 16: 547-563, doi:10.1016/S0277-3791(96)00081-9

Gauthier, P., Sigmarsson, O., Gouhier, M., Haddadi, B., and Moune, S., 2016. Elevated gas flux and trace metal degassing from the 2014–2015 fissure eruption at the Bárðarbunga volcanic system, Iceland. Journal of Geophysical Research: Solid Earth, 121: 1610-1630, doi: 10.1002/2015JB012111

Gell, P. A., Sluiter, I. R., and Fluin, J., 2002. Seasonal and interannual variations in diatom assemblages in Murray River wetlands in north-west Victoria, Australia. Marine and Freshwater Research, 53: 981-992, doi:10.1071/MF01021

Gell, P., 1997. The development of a diatom database for inferring lake salinity, western Victoria, Australia: Towards a quantitative approach for reconstructing past climates. Australian Journal of Botany, 45: 389-423, doi:10.1071/BT96036

Gell, P., Mills, K., and Grundell, R., 2012. A legacy of climate and catchment change: the real challenge for wetland management. Hydrobiologia doi: 10.1007/s10750-012- 1163-4

140

Gell, P., Sonneman, J. A., Reid, M. A., Illman, M. A., and Sincock, A. J., 1999. An illustrated key to common diatom genera from southern Australia. Cooperative Research Centre for Freshwater Ecology Identification Guide No. 26, Thurgoona, NSW, Australia, 63 pp.

Gell, P., Tibby, J., Fluin, J., Leahy, P., Reid, M., Adamson, K., Bulpin, S., MacGregor, A., Wallbrink, P., Hancock, G., and Walsh, B., 2005. River Research and Applications, 21: 257-269, doi:10.1002/rra.845

George, D., Talling, J., and Rigg, E., 2000. Factors influencing the temporal coherence of five lakes in the English Lake District. Freshwater Biology, 43: 449-461, doi:10.1046/j.1365-2427.2000.00566.x

Gilman, R. A., Chapman, C. A., Lowell, T. V., and Borns, H. W., 1988. The Geology of Mount Desert Island: A Visitor’s Guide to the Geology of Acadia National Park, Maine Geological Survey, Department of Conservation, Augusta, Maine, 50 pp.

Granéli, W., Bertilsson, S., and Philibert, A., 2004. Phosphorus limitation of bacterial growth in high Arctic lakes and ponds. Aquatic Sciences, 66: 430-439, doi:10.1007/s00027-004-0732-7

Granéli, W., Lindell, M., and Tranvik, L., 1996. Photo-oxidative production of dissolved inorganic carbon in lakes of different humic content. Limnology and Oceanography, 41: 698-706. doi: 10.4319/lo.1996.41.4.0698

Guillemette, F., and del Giorgio, P. A., 2011. Reconstructing the various facets of dissolved organic carbon bioavailability in freshwater ecosystems. Limnology and Oceanography, 56(2): 734-748, doi:10.4319/lo.2011.56.2.0734

Guillemette, F., and del Giorgio, P. A., 2012. Simultaneous consumption and production of fluorescent dissolved organic matter by lake bacterioplankton. Environmental Microbiology, 14(6): 1432-1443, doi:10.1111/j.1462-2920.2012.02728.x

Gunn, J. M., Snucins, E., Yan, N. D., and Arts, M. T., 200. Use of water clarity to monitor the effects of climate change and other stressors on oligotrophic lakes. Environmental Monitoring and Assessment, 67: 69-88, doi:10.1023/A:1006435721636

Hallegraeff, G.M, 1993. A review of harmful algal blooms and their apparent global increase. Phycologia, 32(2): 79-99.

Helms, J. R., Mao, J., Schmidt-Rohr, K., Abdulla, H., and Mopper, K., 2013. Photochemical flocculation of terrestrial dissolved organic matter and iron. Geochimica et Cosmochimica Acta, 121: 398-413, doi:10.1016/j.gca.2013.07.025

141

Helms, J. R., Stubbins, A., Ritchie, J. D., Minor, E. C., Kieber, D. J., and Mopper, K., 2008. Absorption spectral slopes and slope ratios as indicators of molecular weight, source, and photobleaching of chromophoric dissolved organic matter. Limnology and Oceanography, 53(3): 955-969, doi:10.4319/lo.2008.53.3.0955

Herpich, M. A., 2002. Nutrient and phytoplankton status of Lakes Purrumbete and Bullen Merri, western Victoria. Deakin University.

Hill, M. O. and Gauch, H. G. Jr., 1980. Detrended correspondence analysis: an improved ordination technique. Vegetatio, 42: 47-58, doi:10.1007/978-94-009-9197-2_7

Hobbs, W. O., Vinebrooke, R. D., and Wolfe, A. P., 2011. Biogeochemical responses of two alpine lakes to climate change and atmospheric deposition, Jasper and Banff National Parks, Canadian Rocky Mountains. Canadian Journal of Fisheries and Aquatic Sciences, 68: 1480-1494. doi:10.1139/F2011-058

Holland, D. P., van Erp, I., Beardall, J., and Cook, P. L. M., 2012. Environmental controls on the nitrogen-fixing cyanobacterium Nodularia spumigena in a temperate lagoon system in SE Australia. Marine Ecology Progress Series, 461: 47-57, doi:10.3354/meps09843

Hunt, T. L., Khageswor, G., Brown, P., Ingram, B. A., Jones, P. L., Laurenson, L. J. B., and Wallis, A. M., 2014. Consequences of fish stocking density in a recreational fishery. Canadian Journal of Fisheries and Aquatic Sciences, 71: 1554-1560, doi:10.1139/cjfas- 2013-0517

Hurley, D., Tennison, R., and Gardiner, E. Massive fire storm devastates rural communities in Victoria’s west. (2018, March 19). Herald Sun. Retrieved from https://www.news.com.au/national/victoria/massive-fire-storm-devastates-rural- communities-in-victorias-west/news-story/ef5a3bca472a898c25996eae3052b5f3

Infante, A., and Litt, A. H., 1985. Differences between two species of Daphnia in the use of 10 species of algae in Lake Washington. Limnology and Oceanography, 30(5): 1053- 1059, doi:10.4319/lo.1985.30.5.1053

Järvinen, M., M. Rask, J. Ruuhijärvi, and L. Arvola. 2002. Temporal coherence in water temperature and chemistry under the ice of boreal lakes (Finland). Water Research, 36: 3949-3956, doi:10.1016/S0043-1354(02)00128-8

Jeffries, D. S., Clair, T. A., Couture, S., Dillon, P. J., Dupont, J., Keller, W., McNicol, D. K., Turner, M. A., Vet, R., and Weeber, R., 2003. Assessing the recovery of lakes in southeastern Canada from the effects of acid deposition. Ambio 32: 176-182. doi: 10.1579/0044-7447-32.3.176

142

Jickells, T. D., An, Z. S., Andersen, K. K., Baker, A. R., Bergametti, G., Brooks, N., Cao, J. J., Boyd, P.W., Duce, R. A., Hunter, K. A., Kawahata, H., Kubilay, N., laRoche, J., Liss, P. S., Mahowald, N., Prospero, J. M., Ridgwell, A. J., Tegen, and I., Torres, R., 2005. Global iron connections between desert dust, ocean biogeochemistry, and climate. Science, 308: 67–71, doi:10.1126/science.1105959

Johansson, E., Berglund, S., Lindborg, T., Petrone, J., van As, D., Gustafsson, L. G., Näslund, J. O., and Laudon, H., 2015. Hydrological and meteorological investigations in a periglacial lake catchment near Kangerlussuaq, west Greenland–presentation of a new multi-parameter data set. Earth System Science Data, 7(1): 93-108, doi:10.5194/essd-7- 93-2015

Johansson, E., Gustafsson, L., Berglund, S., Lindborg, T., Selroos, J., Liljedahl, L., and Destouni, G., 2015. Data evaluation and numerical modeling of hydrological interactions between active layer, lake and talik in a permafrost catchment, Western Greenland. Journal of Hydrology, 527: 688-703, doi:10.1016/j.jhydrol.2015.05.026

Jones, D. L., and Willett, V. B., 2006. Experimental evaluation of methods to quantify dissolved organic nitrogen (DON) and dissolved organic carbon (DOC) in soil. Soil Biology and Biochemistry, 38: 991-999, doi:10.1016/j.soilbio.2005.08.012

Jones, R. I., Shaw, P. J., and De Haan, H., 1993. Effects of dissolved humic substances on the speciation of iron and phosphate at different pH and ionic strength. Environmental Science and Technology, 27(6): 1052-1059, doi:10.1021/es00043a003

Jones, R. N., 2010. North central Victorian climate: Past, present, and future. Proceedings of the Royal Society of Victoria, 122(2): 147-160, doi:10.1071/RS10023

Jones, R. N., McMahon, T. A., and Bowler, J. M., 2001. Modelling historical lake levels and recent climate change at three closed lakes, Western Victoria, (c. 1840-1990). Journal of Hydrology, 246: 159-180, doi:10.1016/S0022-1694(01)00369-9

Joyce, E. B. and Evans, R. S., 1964. Areas of landslide activity in Victoria. Proceedings of the Royal Society of Victoria, 74: 99-108.

Kalbitz, K., Schmerwitz, J., Schwesig, D., and Matzner, E., 2003. Biodegradation of soil- derived dissolved organic matter as related to its properties. Geoderma, 113: 273-291, doi:10.1016/S0016-7061(02)00365-8

Kawasaki, N., and Benner, R., 2006. Bacterial release of dissolved organic matter during cell growth and decline: molecular origin and composition. Limnology and Oceanography, 51(5): 2170-2180, doi:10.4319/lo.2006.51.5.2170

Keller, W., Gunn, J. M., and Yan, N. D., 1999. Acid rain—perspectives on lake recovery. Journal of Aquatic Ecosystem Stress and Recovery, 6: 207-216. doi:10.1023/A:1009983318502

143

Keller, W., Heneberry, J., Leduc, J., Gunn, J., and Yan, N., 2006. Variations in epilimnion thickness in small boreal shield lakes: relationships with transparency, weather and acidification. Environmental Monitoring and Assessment, 115: 419-431, doi:10.1007/s10661-006-7237-x

Khan, S. A., Aschwanden, A., Bjørk, A. A., Wahr, J., Kjeldsen, K. K., and Kjær, K. H., 2015. Greenland ice sheet mass balance: a review. Reports on Progress in Physics, 78(4): 046801, doi:10.1088/0034-4885/78/4/046801

Kirono, D. G. C., Jones, R. N., and Kent, D. M., 2009. Modelling future lake levels and salinity at three lakes, western Victoria. National Research Flagships Climate Adaptation, CSIRO, 24 pp.

Kissman, C. E. H., Williamson, C. E., Rose, K. C., Saros, J. E., 2013. Response of phytoplankton in an alpine lake to inputs of dissolved organic matter through nutrient enrichment and trophic forcing. Limnology and Oceanography, 58: 867-880, doi:10.4319/lo.2013.58.3.0867

Kling, G., Kipphut, G., Miller, M., and O’Brien, J., 2000. Integration of lakes and streams in a landscape perspective: The importance of material processing on spatial patterns and temporal coherence. Freshwater Biology, 43: 477-497, doi:10.1046/j.1365- 2427.2000.00515.x

Klug, J., 2002. Positive and negative effects of allochthonous dissolved organic matter and inorganic nutrients on phytoplankton growth. Canadian Journal of Fisheries and Aquatic Sciences, 59: 85-95, doi:10.1139/F01-194

Krammer, K. and Lange-Bertalot, H., (1986–1991) Bacillariophyceae. In: Ettl H, Gärtner G, Gerloff J, Heynig H, Mollenhauer D (eds) Süßwasserflora von Mitteleuropa. Gustav Fischer Verlag, Stuttgart, vol 2(1–4).

Kratz, T., Soranno, P., Baines, S., Benson, B., Magnuson, J., Frost, T., and Lathrop, R., 1998. Inter-annual synchronous dynamics in north temperate lakes in Wisconsin, USA, p. 273-287. In: D. George, J. Jones, P. Puncochar, D. Reynolds, and D. Sutcliffe (eds), Management of lakes and reservoirs during global climate change. Kluwer Academic Publishers, The Netherlands.

Kratz, T., Webster, K., Bowser, C., Magnuson, J., and Benson, B., 1997. The influence of landscape position on lakes in northern Wisconsin. Freshwater Biology, 37: 209-217, doi:10.1046/j.1365-2427.1997.00149.x

Lawrence, C. R., Painter, T. H., Landry, C. C., and Neff, J. C., 2010. Contemporary geochemical composition and flux of aeolian dust to the San Juan Mountains, Colorado,

144

United States. Journal of Geophysical Research: Biogeosciences, 115: G03007, doi:10.1029/2009JG001077.

Leahy, P., Robinson, D., Patten, R., and Kramer, A., 2010. Lakes in the western district of Victoria and climate change. EPA Victoria (EPA Publication 1359).

Leavitt, P. R. and Carpenter, S. R, 1990. Regulation of pigment sedimentation by photooxidation and herbivore grazing. Canadian Journal of Fisheries and Aquatic Sciences, 47: 1166-1176, doi: 10.1139/f90-136

Leavitt, P. R., Fritz, S. C., Anderson, N. J., Baker, P. A., Blenckner, T., Bunting, L., Catalan, J., Conley, D. J., Hobbs, W. O., Jeppesen, E., Korhola, A., McGowan, S., Rühland, K., Rusak, J. A., Simpson, G. L., Solovieva, N., and Werne, J., 2009. Paleolimnological evidence of the effects on lakes of energy and mass transfer from climate and humans. Limnology and Oceanography, 54(6, part 2): 2330-2348, doi:10.4319/lo.2009.54.6_part_2.2330

Leech, D. M., Pollard, A. I., Labou, S. G., and Hampton, S. E., 2018. Fewer blue lakes and more murky lakes across the continental U.S.: Implications for planktonic food webs. Limnology and Oceanography, 63: 2661-2680, doi:10.1002/lno.10967

Legendre, P. and Birks, H. J. B., 2012. From Classical to Canonical Ordination. In: Birks H., Lotter A., Juggins S., Smol J. (eds) Tracking Environmental Change Using Lake Sediments. Developments in Paleoenvironmental Research, vol 5. Springer, Dordrecht.

Leng, M. J., and Anderson, N. J., 2003. Isotopic variation in modern lake waters from western Greenland. The Holocene, 13(4): 605–611, doi:10.1191/0959683603hl620rr

Lewin, J. C., 1962. Silicification. In: Lewin, R. A. (ed) Physiology and Biochemistry of Algae. Academic Press, New York, pp. 445–55.

Lindell, M. J., Granéli, W., and Tranvik, L. J., 1995. Enhanced bacterial growth in response to photochemical transformation of dissolved organic matter. Limnology and Oceanography, 40(1): 195-199, doi:10.4319/lo.1995.40.1.0195

Lu, Y., Bauer, J. E., Canuel, E. A., Yamashita, Y., Chambers, R. M., and Jaffé, R., 2013. Photochemical and bacterial alteration of dissolved organic matter in temperate headwater streams associated with different land use. Journal of Geophysical Research: Biogeosciences, 118: 1-15, doi:10.1002/jgrg.20048

Magnuson, J., Benson, B., and Kratz, T., 1990. Temporal coherence in the limnology of a suite of lakes in Wisconsin, U.S.A. Freshwater Biology, 23: 145-159, doi:10.1111/j.1365- 2427.1990.tb00259.x

145

Magnuson, J., Benson, B., and Kratz, T., 2004. Patterns of coherent dynamics within and between lake districts at local to intercontinental scales. Boreal Environment Research, 9: 359-369.

Magnuson, J., Benson, B., Jensen, O., Clark, T., Card, V., Futter, M., Soranno, P., and Stewart, K., 2005. Persistence of coherence of ice-off dates for inland lakes across the Laurentian Great Lakes region. Verhandlungen des Internationalen Verein Limnologie, 29: 521-527.

Magnuson, J., Kratz, T., Benson, B., and Webster, K., 2005. Coherent dynamics among lakes, p. 89-106. In: J. Magnuson, T. Kratz, and B. Benson (eds), Chapter 5: Long-term dynamics of lakes in the landscape. Oxford University Press.

Magnuson, J., Robertson, D. M., Benson, B. J., Wynne, R. H., Livingstone, D. M., Arai, T., Assel, R. A., Barry, R. G., Card, V., Kuusisto, E., Granin, N. G., Prowse, T. D., Stewart, K. M., and Vuglinksi, V. S., 2000. Historical trends in lake and river ice cover in the northern hemisphere. Science, 289: 1743-1746, doi:10.1126/science.289.5485.1743

Mahowald, N., Baker, A., Bergametti, G., Brooks, N., Duce, R., Jickells, T., Kubilay, N., Prospero, J., and Tegen, I., 2005. Atmospheric global dust cycle and iron inputs to the ocean. Global Biogeochemistry Cycles, 19(4): GB4025, doi:10.1029/2004GB002402

Malik, H. and Saros, J., 2016. Effects of temperature, light, and nutrients on five Cyclotella sensu lato taxa assessed with in situ experiments in arctic lakes. Journal of Plankton Research, 38(3): 431-442, doi:10.1093/plankt/fbw002

Maloney, K. O., Morris, D. P., Moses, C. O., and Osburn, C. L., 2005. The role of iron and dissolved organic carbon in the absorption of ultraviolet radiation in humic lake water. Biogeochemistry, 75: 393-407, doi:10.1007/s10533-005-1675-3

Markager, S. and Vincent, W., 2000. Spectral light attenuation and the absorption of UV and blue light in natural waters. Limnology and Oceanography, 45(3): 642-650, doi:10.4319/lo.2000.45.3.0642

Mayewski, P. A., Sneed, S. B., Birkel, S. D., Kurbatov, A. V., and Maasch, K. A., 2014. Holocene warming marked by abrupt onset of longer summers and reduced storm frequency around Greenland. Journal of Quaternary Science, 29(1): 99-104, doi:10.1002/jqs.2684

McDowell, W. H. and Likens, G. E., 1988. Origin, composition, and flux of dissolved organic carbon in the Hubbard Brook Valley. Ecological Monographs, 58: 177-195. doi: 10.2307/2937024

McGregor, G. B., Stewart, I., Sendall, B. C., Sadler, R., Reardon, K., Carter, S., Wruck, D., and Wickramasinghe, W., 2012. First Report of a Toxic Nodularia spumigena

146

(Nostocales/ Cyanobacteria) Bloom in Sub-Tropical Australia. I. Phycological and Public Health Investigations. International Journal of Environmental Research and Public Health, 9: 2396-2411, doi:10.3390/ijerph9072396

Mernild, S. H., Hasholt B., Tidwell A. C., and Kane D. L., 2008. Jökulhlaup observed at Greenland ice sheet. Eos, 89(165): 161-172.

Miller, W. L. and Moran, M. A., 1997. Interaction of photochemical and bacterial processes in the degradation of refractory dissolved organic matter from a coastal marine environment. Limnology and Oceanography, 42(6): 1317-1324, doi:10.4319/lo.1997.42.6.1317

Mladenov, N., Williams, M. W., Schmidt, S. K., and Cawley, K., 2012. Atmospheric deposition as a source of carbon and nutrients to an alpine catchment of the Colorado Rocky Mountains. Biogeosciences 9: 3337-3355, doi:10.5194/bg-9-3337-2012

Møller, E. F., 2005. Sloppy feeding in marine copepods: prey-size-dependent production of dissolved organic carbon. Journal of Plankton Research, 27(1): 27-35, doi:10.1093/plankt/fbh147

Monteith, D. T., Stoddard, J. L., Evans, C. D., de Wit, H. A., Forsius, M., Høgäsen, T., Wilander, A., Skelkvåle, B. L., Jeffries, D. S., Vuorenmaa, J., Keller, B., Kopácek, J., and Vesely, J., 2007. Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature, 450: 537-540, doi:10.1038/nature06316

Moran, M. A. and Hodson, R. E., 1990. Bacterial production on humic and nonhumic components of dissolved organic carbon. Limnology and Oceanography, 35(8): 1744- 1756, doi:10.4319/lo.1990.35.8.1744

Moran, M. A. and Zepp, R. G., 1997. Role of photoreactions in the formation of biologically labile compounds from dissolved organic matter. Limnology and Oceanography, 42(6): 1307-1316, doi:10.4319/lo.1997.42.6.1307

Moran, M. A., Wade, S. M., and Zepp, R. G., 2000. Carbon loss and optical property changes during long-term photochemical and biological degradation of estuarine dissolved organic matter. Limnology and Oceanography, 45(6): 1254-1264, doi:10.4319/lo.2000.45.6.1254

Morris, D. P. and Hargreaves, B. R., 1997. The role of photochemical degradation of dissolved organic carbon in regulating the UV transparency of three lakes on the Pocono Plateau. Limnology and Oceanography, 42(2): 239-249, doi:10.4319/lo.1997.42.2.0239

Morris, D. P., Zagarese, H., Williamson, C. E., Balseiro, E. G., Hargreaves, B. R., Modenutti, B., Moeller, R., and Queimalinos, C., 1995. The attenuation of solar UV

147

radiation in lakes and the role of dissolved organic carbon. Limnology and Oceanography, 40: 1381-1391. doi: 10.4319/lo.1995.40.8.1381

Mueller, S. H., Goldfarb, R. J., Farmer, G. L., Sanzolone, R., Adams, M., Theodorakos, P. M., Richmond, S. A., and McCleskey, R. B., 2002. Trace, minor and major element data for ground water near Fairbanks, Alaska, 1999-2000. US Department of the Interior, US Geological Survey.

Murphy, B. and Timbal, B., 2008. A review of recent climate variability and climate change in southeastern Australia. International Journal of Climatology, 28: 859-879, doi:10.1002/joc.1627

National Atmospheric Deposition Program (NRSP-3). 2019. NADP Program Office, Wisconsin State Laboratory of Hygiene, 465 Henry Mall, Madison, WI 53706.

Neff, J. C., Ballantyne, A. P., Farmer, G. L., Mahowald, N. M., Conroy, J. L., Landry, C. C., Overpeck, J. T., Painter, T. H. Lawrence, C. R., Reynolds, R. L., 2008. Increasing eolian dust deposition in the western United States linked to human activity. Nature Geoscience, 1: 189-195, doi:10.1038/ngeo133

Nguyen, T. T. T. and Ingram, B. A., 2012. Genetic status of Chinook salmon held at Snobs Creek and implications for improved broodstock management. Fisheries Victoria Research Report Series No. 55.

Nyman, P., Sheridan, G. J., Smith, H. G., and Lane, P. N. J., 2011. Evidence of debris flow occurrence after wildfire in upland catchments of south-east Australia. Geomorphology, 125: 383-401, doi:10.1016/j.geomorph.2010.10.016

Obernosterer, I. and Benner, R., 2004. Competition between biological and photochemical processes in the mineralization of dissolved organic carbon. Limnology and Oceanography, 49(1): 117-124, doi:10.4319/lo.2004.49.1.0117

Obernosterer, I., Reitner, B., and Herndl, G. J., 1999. Contrasting effects of solar radiation on dissolved organic matter and its bioavailability to marine bacterioplankton. Limnology and Oceanography, 44(7): 1645-1654, doi: 10.4319/lo.1999.44.7.1645

Ogawa, H., Amagai, Y., Koike, I., Kaiser, K., and Benner, R., 2001. Production of refractory dissolved organic matter by bacteria. Science, 292(5518): 917-920, doi:10.1126/science.1057627

Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, R. R., O’Hara, R. B., Simpson, G. L., Solymos, P., Stevens, H. H., and Wagner, H., 2013. vegan: community ecology package. R package version 2.0–10. http://CRAN.R-project.org/package=vegan

148

Oliva, M. G., Lugo, A., Alcocer, J., and Cantoral-Urize, E. A., 2008. Morphological study of Cyclotella choctawhatcheeana Prasad (Stephanodiscaceae) from a saline Mexican lake. Saline Systems, 4: 17, doi:10.1186/1746-1448-4-17

Osburn, C. L., Morris, D. P., Thorn, K. A., and Moeller, R. E., 2001. Chemical and optical changes in freshwater dissolved organic matter exposed to solar radiation. Biogeochemistry, 54: 251-278, doi:10.1023/A:1010657428418

Osburn, C. L., Zagarese, H. E., Morris, D. P., Hargreaves, B. R., and Cravero, W. E., 2001. Calculation of spectral weighting functions for the solar photobleaching of chromophoric dissolved organic matter in temperate lakes. Limnology and Oceanography, 46(6): 1455-1467, doi: 10.4319/lo.2001.46.6.1455

Osburn, C., Anderson, N. J., Stedmon, C., Giles, M., Whiteford, E., McGenity, T., Dumbrell, A., and Underwood, G., 2017. Shifts in the source and composition of dissolved organic matter in southwest Greenland lakes along a regional hydro-climatic gradient. Journal of Geophysical Research: Biogeosciences, 122(12): 3431-3445, doi:10.1002/2017JG003999

Osburn, C., Wigdahl, C., Fritz, S., and Saros, J., 2011. Dissolved organic matter composition and photoreactivity in prairie lakes of the U.S. Great Plains. Limnology and Oceanography, 56(6): 2371-2390, doi:10.4319/lo.2011.56.6.2371

Pace, M. L., Cole, J. J., 2002. Synchronous variation of dissolved organic carbon and color in lakes. Limnology and Oceanography, 47: 333-342. doi:10.4319/lo.2002.47.2.0333

Paerl, H. W. and Paul, V. J., 2012. Climate change: links to global expansion of harmful cyanobacteria. Water Research 46(5): 1349-1363, doi:10.1016/j.watres.2011.08.002

Paerl, H. W., 2014. Mitigating harmful cyanobacterial blooms in a human- and climatically-impacted world. Life, 4: 988-1012, doi:10.3390/life4040988

Paerl, H.W. and Huisman, J., 2008. Blooms Like It Hot. Science, 320: 57-58, doi:10.1126/science.1155398

Patoine, A. and Leavitt, P., 2006. Century-long synchrony of fossil algae in a chain of Canadian prairie lakes. Ecology 87(7): 1710-1721, doi:10.1890/0012- 9658(2006)87[1710:CSOFAI]2.0.CO;2

Pham, S., Leavitt, P., McGowan, S., and Peres-Neto, P., 2008. Spatial variability of climate and land-use effects on lakes of the northern Great Plains. Limnology and Oceanography, 53(2): 728-742, doi:10.4319/lo.2008.53.2.0728

149

Pham, S., Leavitt, P., McGowan, S., Wissel, B., and Wassenaar, L., 2009. Spatial and temporal variability of prairie lake hydrology as revealed using stable isotopes of hydrogen and oxygen. Limnology and Oceanography, 54(1): 101-118, doi:10.4319/lo.2009.54.1.0101

Prasad, A. K. S. K. and Nieow, J. A. 2006. The centric diatom genus Cyclotella, (Stephanodiscaceae: Bacillariophyta) from Florida Bay, USA, with special reference to Cyclotella choctawhatcheeana and Cyclotella desikacharyi, a new marine species related to the Cyclotella striata complex. Phycologia, 45(2): 127-140, doi:10.2216/05-13.1

Prasad, A. K. S. K., Nienow, J. A., Livingston, R. J., 1990. The genus Cyclotella (Bacillariophyta) in Choctawhatchee Bay, Florida, with special reference to C. striata and C. choctawhatcheeana sp. nov. Phycologia, 29: 418-436.

Prospero, J. M., Bullard, J. E., and Hodgkins, R., 2012. High-latitude dust over the North Atlantic: inputs from Icelandic proglacial dust storms. Science, 335(6072): 1078-1082, doi:10.1126/science.1217447

R Core Development Team, 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R- project.org/

Rose, K. C., Greb, S. R., Diebel, M., and Turner, M. G., 2017. Annual precipitation regulates spatial and temporal drivers of lake water clarity. Ecological Applications, 27(2): 632-643, doi:10.1002/eap.1471

Rose, K. C., Williamson, C. E., Fischer, J. M., Connelly, S. J., Olson, M., Tucker, A. J., Noe, D. A., 2012. The role of ultraviolet radiation and fish in regulating the vertical distribution of Daphnia. Limnology and Oceanography, 57: 1867-1876, doi:10.4319/lo.2012.57.6.1867

Rose, K. C., Williamson, C. E., Saros, J. E., Sommaruga, R., and Fischer, J. M., 2009. Differences in UV transparency and thermal structure between alpine and subalpine lakes: implications for organisms. Photochemical and Photobiological Sciences, 8: 1244- 1256, doi: 10.1039/b905616e

Rühland, K. M., Smol, J. P., and Pienitz, R., 2003. Ecology and spatial distributions of surface-sediment diatoms from 77 lakes in the subarctic Canadian treeline region. Canadian Journal of Botany, 81: 57-73, doi:10.1139/b03-005

SanClements, M., G. Oelsner, D. McKnight, J. Stoddard, and S. Nelson. 2012. New insights into the source of decadal increases of dissolved organic matter in acid-sensitive lakes of the northeastern United States. Environmental Science and Technology, 46: 3212-3219, doi:10.1021/es204321x

150

Saros, J. E. and Anderson, N. J., 2015. The ecology of the planktonic diatom Cyclotella and its implications for global environmental change studies. Biological Reviews, 90: 522-541, doi:10.1111/brv.12120

Saros, J. E., Interlandi, S. J., Doyle, S., Michel, T. J., and Williamson, C. E., 2005. Are the deep chlorophyll maxima in alpine lakes primarily induced by nutrient availability, not UV avoidance? Arctic, Antarctic, and Alpine Research, 37: 557-563, doi: 10.1657/1523-0430(2005)037[0557:ATDCMI]2.0.CO;2

Saros, J. E., Northington, R. M., Anderson, D. S., Anderson, N. J., 2016. A whole-lake experiment confirms a small centric diatom species as an indicator of changing lake thermal structure. Limnology and Oceanography Letters, 1: 27-35. doi:10.1002/lol2.10024

Saros, J. E., Osburn, C. L., Northington, R. M., Birkel, S. D., Auger, J. D., Stedmon, C. A., and Anderson, N. J., 2015. Recent decrease in DOC concentrations in Arctic lakes of southwest Greenland. Geophysical Research Letters, 42(16): 6703-6709, doi:10.1002/2015GL065075

Saros, J. E., Stone, J. R., Pederson, G. T., Slemmons, K. E. H., Spanbauer, T., Schliep, A., Cahl, D., Williamson, C. E., and Engstrom, D. R., 2012. Climate-induced changes in lake ecosystem structure inferred from coupled neo- and paleoecological approaches. Ecology, 93(10): 2155-2164, doi:10.1890/11-2218.1

Saros, J., Northington, R., Osburn, C., Burpee, B., and Anderson, N. J., 2016. Thermal stratification in small arctic lakes of southwest Greenland affected by water transparency and epilimnetic temperatures. Limnology and Oceanography, 61: 1530-1542, doi:10.1002/lno.10314

Saunders, K. M., Hodgson, D. A., Harrison, J., and McMinn, A., 2008. Palaeoecological tools for improving the management of coastal ecosystems: a case study from Lake King (Gippsland Lakes) Australia. Journal of Paleolimnology, 40: 33-47, doi:10.1007/s10933- 007-9132-z

Shapiro, J., 1966. On the measurement of ferrous iron in natural waters. Limnology and Oceanography, 11(2): 293-298. Smol, J., Wolfe, A., Birks, H., Douglas, M., Jones, V., Korhola, A. Pienitz, R., Rühland, K., Sorvari, S., Antoniades, D., Brooks, S., Fallu, M.-A., Hughes, M., Keatley, B., Laing, T., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitonen, S., Solvieva, N., and Weckström, J., 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the Natural Academy of Sciences of the United States of America, 102(12), 4397-4402, doi:10.1073/pnas.0500245102

151

Snucins, E. and Gunn, J., 2000. Interannual variation in the thermal structure of clear and colored lakes. Limnology and Oceanography, 45: 1639-1646, doi:10.4319/lo.2000.45.7.1639

Sobek, S., Tranvik, L., Prairie, Y., Kortelainen, P., and Cole, J., 2007. Patterns and regulation of dissolved organic carbon: An analysis of 7,500 widely distributed lakes. Limnology and Oceanography, 52(3): 1208-1219, doi:10.4319/lo.2007.52.3.1208

Sonneman, J. A., Sincock, A., Fluin, J., Reid, M., Newall, P., Tibby, J., and Gell, P., 2002. An Illustrated Guide to Common Stream Diatom Species from Temperate Australia. Cooperative Research Centre for Freshwater Ecology Identification Guide No. 33, Thurgoona, NSW, Australia, 166 pp.

Soranno, P., Webster, K. E., Riera, J. L., Kratz, T. K., Baron, J. S., Bukaveckas, P. A., Kling, G. W., White, D. S., Caine, N., Lathrop, R. C., and Leavitt, P. R., 1999. Spatial variation among lakes within landscapes: Ecological organization along lake chains. Ecosystems, 2: 395-410, doi:10.1007/s100219900089

Spencer, R. G. M., Aiken, G. R., Butler, K. D., Dornblaser, M. M., Striegl, R. G., and Hernes, P. J., 2009. Utilizing chromophoric dissolved organic matter measurements to derive export and reactivity of dissolved organic carbon exported to the Arctic Ocean: a case study of the Yukon River, Alaska. Geophysical Research Letters, 36(6): L06401, doi: 10.1029/2008GL036831.

Spierre, S. G., Wake, C., Gittell, R., Hurtt, G., Kelly, T., Markham, A., Schaefer, D., and VanDeveer, S., 2010. Trends in extreme precipitation events for the northeastern United States. Carbon Solutions New England, University of New Hampshire, Durham, NH, 14 pp.

Stauffer, R., 1990. Alkalinities of Maine lakes: Are they really changing? Limnology and Oceanography, 35: 1238-1257, doi:10.4319/lo.1990.35.6.1238

Stoddard, J. L., Jeffries, D. S., Lükewille, A., Clair, T. A., Dillon, P. J., Driscoll, C. T., Forsius, M., Johannessen, M., Kahl, J. S., Kellogg, J. H., Kemp, A., Mannio, J., Monteith, D. T., Murdoch, P. S., Patrick, S., Rebsdorf, A., Skelkvåle, B. L., Stainton, M. P., Traaen, T., van Dam, H., Webster, K. E., Wieting, J., and Wilander, A., 1999. Regional trends in aquatic recovery from acidification in North America and Europe. Nature, 401: 575-578, doi:10.1038/44114

Stoessel, D., 2009. Artificial aeration of Lake Bullen Merri—Review of Operations. Victorian Fisheries Authority. Retrieved from https://vfa.vic.gov.au/about/publications- and-resources/fisheries-reports/your-licence-fees-at-work-reports/2009-2010/artificial- aeration-of-lake-bullen-merri-review-of-operations

152

Stolte, W., Balode, M., Carlsson, P., Grzebyk, D., Janson, S., Lips, I., Panosso, R., Ward, C. J., and Granéli, E., 2006. Stimulation of nitrogen-fixing cyanobacteria in a Baltic Sea plankton community by land-derived organic matter or iron addition. Marine Ecology Progress Series, 327: 71-82, doi:10.3354/meps327071

Stone, J. R. and Fritz, S. C., 2004. Three-dimensional modeling of lacustrine diatom habitat areas: Improving paleolimnological interpretation of planktic : benthic ratios. Limnology and Oceanography, 49(5): 1540-1548, doi:10.4319/lo.2004.49.5.1540

Stone, J. R., Saros, J. E., and Pederson, G. T., 2016. Coherent late-Holocene climate- driven shifts in the structure of three Rocky Mountain lakes. The Holocene, 26(7): 1103- 1111, doi:10.1177/0959683616632886

Striegl, R. G., Aiken, G. R., Dornblaser, M. M., Raymond, P. A., and Wickland, K. P., 2005. A decrease in discharge-normalized DOC export by the Yukon River during summer through autumn, Geophysical Research Letters, 32: L21413, doi:10.1029/2005GL024413

Strock, K. E., Saros, J. E., Nelson, S. J., Birkel, S. D., Kahl, J. S., and McDowell, W. H., 2016. Extreme weather years drive episodic changes in lake chemistry: implications for recovery from sulfate deposition and long-term trends in dissolved organic carbon. Biogeochemistry, 127(2): 353-365, doi:10.1007/s10533-016-0185-9

Strock, K. E., Theodore, N., Gawley, W. G., Ellsworth, A. C., and Saros, J. E., 2017. Increasing dissolved organic carbon concentrations in northern boreal lakes: Implications for lake water transparency and thermal structure. Journal of Geophysical Research: Biogeosciences, 122: 1022-1035, doi:10.1002/2017JG003767

Tennison, R., Simonis, A., Hamblin, A., Hurley, D., and Fogarty, D. Victorian homes lost as wind change puts towns in fire danger. (2018, March 18). Herald Sun. Retrieved from https://www.heraldsun.com.au/news/victoria/victorians-warned-to-brace-for-severe-fire- danger/news-story/a43cea90a20f2c018355eb90a16441b3

Thingstad, T. F., Hagström, A., and Rassoulzadegan, F, 1997. Accumulation of degradable DOC in surface waters: Is it caused by a malfunctioning microbial loop? Limnology and Oceanography, 42(2): 398-404, doi:10.4319/lo.1997.42.2.0398

Thomas, J. D., 1997. The role of dissolved organic matter, particularly free amino acids and humic substances, in freshwater ecosystems. Freshwater Biology, 38: 1-36, doi:10.1046/j.1365-2427.1997.00206.x

Thompson, A. Blue-green algae: Camperdown lake expected to worsen. (2018, January 4). The Weekly Times. Retrieved from https://www.weeklytimesnow.com.au/news/bluegreen-alae-camperdown-lake-expected- to-worsen/news-story/502e2d3b74e033ace277a8b2a18bdffa

153

Tibby, J., 2004. Development of a diatom-based model for inferring total phosphorus in southeastern Australian water storages. Journal of Paleolimnology, 31: 23-36, doi:10.1023/B:JOPL.0000013272.25122.2a

Timbal, B., 2009. The continuing decline in south-east Australian rainfall: update to May 2009. Centre for Australian Weather and Climate Research (CAWCR) Research Letters, 2: 4-10.

Timms, B. V., 1976. A comparative study of the limnology of three maar lakes in western Victoria I. Physiography and physicochemical features. Australian Journal of Marine and Freshwater Research, 27: 35-60, doi:10.1071/MF9760035

Timms, B. V., 2005. Salt lakes in Australia: present problems and prognosis for the future. Hydrobiologia, 552: 1-15, doi:10.1007/s10750-005-1501-x

Tipping, E., 1981. The adsorption of aquatic humic substances by iron oxides. Geochimica et Cosmochimica Acta, 45: 191-199, doi:10.1016/0016-7037(81)90162-9

Tranvik, L. J. and Bertilsson, S., 2001. Contrasting effects of solar UV radiation on dissolved organic sources for bacterial growth. Ecology Letters, 4: 458-463, doi:10.1046/j.1461-0248.2001.00245.x

Tranvik, L. J., 1993. Microbial transformation of labile dissolved organic matter into humic-like matter in seawater. FEMS Microbiology Ecology, 12: 177-183, doi:10.1111/j.1574-6941.1993.tb00030.x

Tranvik, L. J., and Höfle, M. G., 1987. Bacterial growth in mixed cultures on dissolved organic carbon from humic and clear waters. Applied and Environmental Microbiology, 53(3): 482-488.

U.S. Geological Survey (2014) NLCD 2011 Land Cover (2011 Edition, amended 2014) - National Geospatial Data Asset (NGDA) Land Use Land Cover: U.S. Geological Survey.

Van As, D., Andersen, M. L., Petersen, D., Fettweis, X., Van Angelen, J. H., Lenaerts, J., Van Den Broeke, M. R., Lea, J. M., Bøggild, C. E., Ahlstrøm, A. P., and Steffen, K., 2014. Increasing meltwater discharge from the Nuuk region of the Greenland ice sheet and implications for mass balance (1960–2012). Journal of Glaciology, 60(220): 314- 322, doi:10.3189/2014jog13j065

Victoria Department of Environmental, Land, Water and Planning (DELWP), 2017. Lake Bullen Merri Management Plan.

Wang, L., Lu, H., Liu, J., Gu, Z., Mingram, J., Chu, G., Li, J., Rioual, P., Negendank, J. F. W., Han, J., and Liu, T., 2008. Diatom-based inference of variations in the strength of

154

Asian winter monsoon winds between 17,500 and 6000 calendar years B.P. Journal of Geophysical Research, 113: D21101, doi:10.1029/2008JD010145

Webster, K. E., Soranno, P. A., Baines, S. B., Kratz, T. K., Bowser, C. L., Dillon, P. J., Campbell, P., Fee, E. J., and Hecky, R. E., 2000. Structuring features of lake districts: Landscape controls on lake chemical responses to drought. Freshwater Biology, 43: 499- 515, doi:10.1046/j.1365-2427.2000.00571.x

Weishaar, J. L., Aiken, G. R., Bergamaschi, B. A., Fram, M. S., Fujii, R., and Mopper, K., 2003. Evaluation of specific ultraviolet absorbance as an indicator of the chemical composition and reactivity of dissolved organic carbon. Environmental Science and Technology, 37(20): 4702-4708, doi:10.1021/es030360x, doi:10.1021/es030360x

Wetzel, R. G., Hatcher, P. G., and Bianchi, T. S., 1995. Natural photolysis by ultraviolet irradiance of recalcitrant dissolved organic matter to simple substrates for rapid bacterial metabolism. Limnology and Oceanography, 40(8): 1369-1380, doi:10.4319/lo.1995.40.8.1369

Weyhenmeyer, G. A. and Karlsson, J., 2009. Nonlinear response of dissolved organic carbon concentrations in boreal lakes to increasing temperatures. Limnology and Oceanography, 54: 2513-2519, doi:10.4319/lo.2009.54.6_part_2.2513

Weyhenmeyer, G. A., Willén, E., and Sonesten, L., 2004. Effects of an extreme precipitation event on water chemistry and phytoplankton in the Swedish Lake Mälaren. Boreal Environment Research, 9: 409-420.

Whitehead, D. R., Charles, D. F., and Goldstein, R. A., 1990. The PIRLA project (Paleoecological Investigation of Recent Lake Acidification): an introduction to the synthesis of the project. Journal of Paleolimnology, 3: 187-194. doi: 10.1007/BF00219458

Williams, C. J., Yamashita, Y., Wilson, H. F., Jaffé, R., and Xenopoulos, M. A., 2010. Unraveling the role of land use and bacterial activity in shaping dissolved organic matter characteristics in stream ecosystems. Limnology and Oceanography, 55(3): 1159-1171, doi: 10.4319/lo.2010.55.3.1159

Williamson, C. E., Morris, D. P., Pace, M. L., and Olson, O. G., 1999. Dissolved organic carbon and nutrients as regulators of lake ecosystems: resurrection of a more integrated paradigm. Limnology and Oceanography, 44(3, Part 2): 795-803, doi:10.4319/lo.1999.44.3_part_2.0795

Williamson, C. E., Overholt, E. P., Pilla, R. M., Leach, T. H., Brentrup, J. A., Knoll, L. B., Mette, E. M., and Moeller, R. E., 2015. Ecological consequences of long-term browning in lakes. Scientific Reports, 5: 18666, doi:10.1038/srep18666

155

Williamson, C. E., Saros, J. E., Vincent, W. F., and Smol, J. P., 2009. Lakes and reservoirs as sentinels, integrators, and regulators of climate change. Limnology and Oceanography, 54(6, Part 2): 2273-2282, doi:10.4319/lo.2009.54.6_part_2.2273

Williamson, C. E., Stemberger, R. S., Morris, D. P., Frost, T. M., and Paulsen, S. G., 1996. Ultraviolet radiation in North American lakes: Attenuation estimates from DOC measurements and implications for plankton communities. Limnology and Oceanography, 41: 1024-1034, doi:10.4319/lo.1996.41.5.1024

Yihdego, Y., 2010. Modelling of lake level and salinity for Lake Purrumbete in western Victoria. EPA Victoria Technical Report, 37 pp.

Yihdego, Y., Webb, J., and Leahy, P., 2016. Modelling water and salt balances in a deep, groundwater-throughflow lake—Lake Purrumbete, southeastern Australia. Hydrological Sciences Journal, 61(1); 186-199, doi:10.1080/02626667.2014.975132

Zhang, J., Hudson, J., Neal, R., Sereda, J., Clair, T., Turner, M., Jeffries, D., Dillon, P., Molot, L., Somers, K., and Hesslein, R., 2010. Long-term patterns of dissolved organic carbon in lakes across eastern Canada: Evidence of a pronounced climate effect. Limnology and Oceanography, 55: 30-42, doi:10.4319/lo.2010.55.1.0030

156

APPENDIX: RESULTS OF CHAPTER TWO DUST + LIGHT AND BACTERIA +

LIGHT EXPERIMENTS

Table A.1. Dust + light and bacteria + light experimental results from Chapter 2.

[DOC] SUVA S 254 275-295 (mg L-1) (L mg C-1 m-1) (nm-1) Expt. Lake Treatment Day 0 Day 7 Day 0 Day 7 Day 0 Day 7 Dark + 24.1 ± 24.1 ± 3.26 ± 2.80 ± -0.029 ± -0.035 ± no dust 0.0 0.1 0.04 0.03 0.001 0.000 Dark + 24.1 ± 24.4 ± 3.24 ± 3.22 ± -0.029 ± -0.030 ±

dust 0.2 0.2 0.06 0.02 0.000 0.000

SS2 Light + 24.5 ± 26.0 ± 3.20 ± 2.56 ± -0.029 -0.035 ± no dust 0.1 0.2 0.03 0.01 ±0.000 0.000 Light + 24.6 ± 25.9 ± 3.20 ± 3.00 ± -0.029 ± -0.030 ± dust 0.0 0.2 0.04 0.03 0.001 0.000 Dark + 26.7 ± 26.4 ± 2.22 ± 2.22 ± -0.025 ± -0.025 ± no dust 0.2 0.1 0.08 0.01 0.001 0.000

Dark + 26.4 ± 26.5 ± 2.19 ± 2.19 ± -0.026 ± -0.026 ±

dust 0.1 0.0 0.02 0.02 0.000 0.000

Light + 26.7 ± 29.6 ± 2.13 ± 1.80 ± -0.026 ± -0.028 ± SS1381 no dust 0.2 0.1 0.06 0.01 0.000 0.001 Light + 26.5 ± 29.7 ± 2.21 ± 1.78 ± -0.026 ± -0.027 ±

dust 0.1 0.3 0.01 0.01 0.001 0.001 Dust + Light + Dust Dark + 7.4 ± 7.8 ± 3.89 ± 3.67 ± -0.025 ± -0.026 ± no dust 0.0 0.3 0.01 0.12 0.000 0.000

Dark + 7.3 ± 8.2 ± 3.91 ± 3.43 ± -0.025 ± -0.026 ±

dust 0.1 0.7 0.03 0.28 0.000 0.000

Light + 7.4 ± 9.9 ± 4.00 ± 2.54 ± -0.025 ± -0.031 ± SS901 no dust 0.0 0.4 0.10 0.07 0.000 0.001 Light + 7.5 ± 10.5 ± 3.88 ± 2.42 ± -0.025 ± -0.032 ± dust 0.0 0.5 0.05 0.04 0.001 0.000

69.5 ± 51.5 ± 8.37 ± 11.40 ± -0.013 ± -0.012 ± No dust 1.0 0.7 0.20 0.12 0.000 0.000 66.2 ± 51.8 ± 8.67 ± 11.23 ± -0.012 ± -0.012 ± Dust Soilwater 2.4 1.0 0.30 0.15 0.000 0.000

157

Table A.1 continued.

[DOC] SUVA S 254 275-295 (mg L-1) (L mg C-1 m-1) (nm-1) Expt. Lake Treatment Day 0 Day 7 Day 0 Day 7 Day 0 Day 7 Dark + 28.2 ± 28.8 ± 2.85 ± 2.81 ± -0.030 ± -0.031 ± filtered 0.1 0.3 0.01 0.02 0.000 0.000 Dark + 27.4 ± 27.8 ± 2.96 ± 2.90 ± -0.030 ± -0.031 ±

inoculated 0.1 0.0 0.03 0.02 0.000 0.000

SS2 Light + 28.2 ± 28.3 ± 2.83 ± 2.52 ± -0.031 ± -0.034 ± filtered 0.1 0.1 0.01 0.02 0.000 0.000 Light + 27.3 ± 27.3 ± 2.93 ± 2.55 ± -0.031 ± -0.035 ± inoculated 0.1 0.2 0.02 0.01 0.000 0.000 Dark + 32.8 ± 33.2 ± 2.06 ± 2.00 ± -0.034 ± -0.035 ±

filtered 0.1 0.1 0.02 0.01 0.001 0.001

Dark + 33.0 ± 33.2 ± 2.00 ± 2.00 ± -0.035 ± -0.035 ± inoculated 0.1 0.0 0.01 0.01 0.000 0.001

Light + 32.6 ± 33.4 2.06 ± 1.83 ± -0.034 ± -0.039 ± SS1381 filtered 0.1 ±0.2 0.02 0.01 0.000 0.001

Bacteria Light + Bacteria Light + 33.0 ± 33.6 ± 2.00 ± 1.81 ± -0.035 ± -0.038 ± inoculated 0.2 0.1 0.03 0.01 0.001 0.000 Dark + 7.6 ± 8.0 ± 3.66 ± 3.33 ± -0.025 ± -0.027 ± filtered 0.0 0.2 0.04 0.26 0.000 0.001

Dark + 7.7 ± 8.5 ± 3.59 ± 2.88 ± -0.026 ± -0.027 ±

inoculated 0.0 0.8 0.02 0.88 0.000 0.001

Light + 7.6 ± 8.3 ± 3.72 ± 2.78 ± -0.025 ± -0.030 ± SS901 filtered 0.0 0.5 0.05 0.46 0.000 0.000 Light + 7.6 ± 8.0 ± 3.68 ± 3.07 ± -0.026 ± -0.029 ± inoculated 0.0 0.0 0.04 0.13 0.000 0.002

158

BIOGRAPHY OF THE AUTHOR

Rachel Anne (Dicker) Fowler was born in Presque Isle, Maine in December 1987.

She was raised in Westfield, Maine and graduated from Presque Isle High School in

2005. She attended Bowdoin College and graduated in 2009 with a Bachelor of Arts degree in Biology and a minor in Anthropology. She went on to earn her Master of

Science in Environmental, Earth, and Ocean Science from the University of

Massachusetts Boston in 2011, completing a thesis under the guidance of Dr. Juanita

Urban-Rich entitled “Release of extracellular polymeric substances (EPS) by Aurelia aurita and mucus blob flux.” After graduation, Rachel worked as a certified arborist in

Worcester, Massachusetts before moving back to Maine in 2012 and performing quality control in a microbiological facility. Rachel began her doctoral work at the University of

Maine in June 2014. She is a candidate for the Doctor of Philosophy degree in Ecology and Environmental Science from the University of Maine in May 2019.

159