THE EFFECTS OF THERMAL HABITAT AND MACROINVERTEBRATE

PREDATION ON THE CRUSTACEAN ZOOPLANKTON COMMUNITY OF A SMALL

BOREAL SHIELD LAKE

by

SHANNON ANNE MACPHEE

A thesis submitted to the Department of Biology

in conformity with the requirements for

the degree of Master of Science

Queen’s University

Kingston, Ontario, Canada

March 2009

Copyright © Shannon A. MacPhee, 2009

i 

    ISBN:978-0-494-48187-5    

 ABSTRACT

Climate change will affect all freshwater ecosystems via both direct physiological and indirect, biologically-mediated effects. Small lakes (< 10 ha) numerically dominate the Boreal Shield and represent an important habitat for aquatic biota. Small, shallow lakes are particularly responsive to climate-induced changes in thermal structure.

Furthermore, biological interactions may be particularly important in small lakes where space, habitat heterogeneity, and thermal refugia are limited. Therefore, it is critical to understand and predict the consequences of climate change for community dynamics in small Boreal Shield lakes.

Using 10 years of monitoring data and a field experiment I tested for differences in crustacean zooplankton community structure between warm and cool lake habitats. I classified years from a small, shallow Boreal Shield lake as ‘warm’ or ‘cool’ based on several characteristics of lake thermal structure. Since macroinvertebrates are often the main predators in small, shallow lakes, I further tested for potential interactions between lake thermal structure and spatially-dependent macroinvertebrate predation using in situ mesocosms.

Body sizes of two ubiquitous crustacean zooplankton taxa, Leptodiaptomus minutus and Bosmina spp., were reduced in warm years, but no differences in abundance or diversity were detected at the annual scale. In contrast, in 15d enclosure experiments, crustacean zooplankton abundance and calanoid copepodid body size were reduced by the vertically-migrating predator Chaoborus punctipennis, but only in warm isothermal conditions. Zooplankton lowered their daytime depth distribution to avoid the surface- orienting notonectid predator, Buenoa macrotibialis, regardless of thermal habitat. No

ii predation effect was detected in a hot (25ºC) isothermal habitat where both Chaoborus and notonectids were likely heat-stressed.

Differences in abundance effects between the enclosure and monitoring data are likely due to the scales at which the analyses were conducted. Over short timescales predator-prey dynamics depended on lake thermal structure. However, over annual timescales zooplankton response was averaged across periods of seasonal change in thermal structure and biological processes, which may dampen the short-term effects associated with strong predation in isothermal conditions. Therefore, the importance of macroinvertebrate predators in regulating crustacean zooplankton community structure may increase if small lakes become progressively more isothermal with future climate change.

iii CO-AUTHORSHIP

This thesis conforms to the manuscript format as outlined by the School of

Graduate Studies and Research. The Acknowledgements and References sections apply to the entire thesis. The manuscripts that are a direct result of this thesis and its coauthors are:

MacPhee, S.A., Keller, W., and Arnott, S.E. In preparation. Differences in crustacean zooplankton community structure between warm and cool years in a small Boreal Shield lake.

MacPhee, S.A., Arnott, S.E., and Keller, W. In preparation. The effects of lake thermal habitat and spatially-dependent macroinvertebrate predation on crustacean zooplankton community structure.

iv ACKNOWLEDGEMENTS Thanks to my two supervisors, Shelley Arnott and Bill Keller. I consider you both as role models who manage to balance being successful scientists with your active personal lives. Thank you both for supporting me by sharing your homes during the field season and during the final stages of writing. You are both optimistic individuals who inspire me to stay in the field of aquatic ecology and I hope to interact with you in the future.

To the Arnott lab – Thank you to Angela Strecker and Alison Derry for being leaders during the early stages of my thesis. Liz Hatton, you are a true friend. Anneli

Jokela, I appreciate all of your helpful comments, constructive criticism, advice, and support - you are a natural leader and a great friend. Leah James, it was wonderful to share discussions with you, both academic and personal. Thanks also to Justin Shead,

Mike Pedruski, Colleen Inglis, Derek Gray, Johanna Pokorny, Theresa Patenaude, Shirley

French, Amy Tropea and Adam Jeziorski for being supportive colleagues and friends.

Thanks to Leah James, Hannah Kent, and Colin MacLeod for help with the experimental set-up.

Corrine Daly – I could not have done this without you. You are a truly amazing, motivated individual and incredible friend. Thanks for your support, hard work and friendship. Your honours project (2008) provided much insight into the interpretation of my experimental results.

Thanks to the MacPhee’s, especially to John for help with data management, to

Mike for help with programming, to Kathy for help editing and to Matt for letting me convert your room into an office.

v Leon Boegman kindly provided example MatLab® code. Thanks to Tom and

Karen Johnston for use of your personal canoe on Clearwater Lake. Thanks to Chantal

Sarazin-Delay, Kim Fram, Dan Dechaine and Donna Strang from the Bughouse for lending equipment and for help collecting notonectids, and to Lynn Witty for patient taxonomic training – Say “No”-to- nectids!

To the Waterhouse – Amanda Valois, you were a wonderful peer and were exactly far enough ahead of me to provide ‘fresh’ advice. I consider you a friend and colleague.

Nat Webster, thanks for your emotional support, especially during the field season and writing process, and for your friendship and zest for life. Andrea Ford, thank you for your help with data sharing, data management suggestions, and bug collection. Jocelynne

Heneberry, you are a true mentor, colleague and friend, and I can’t thank you enough for the positive experiences we’ve shared at the Coop Unit. You are diligent with sample collection and management and have deep knowledge of your long-term monitoring dataset.

Thanks to the entire Cooperative Freshwater Ecology Unit, including those aforementioned plus Karen Oman for your friendship. Thanks to the Dorset

Environmental Science Center for my first working exposure to limnology. I am grateful for all previous opportunities made available to me along the way, including the U of G

Coop program, the DESC, CWS, and the ELA. Committee members John Smol and Paul

Grogan provided helpful comments to improve the thesis. Funding was provided by an

Ontario Ministry of the Environment ‘Best in Science’ grant to Shelley Arnott and Bill

Keller and a Queen’s Graduate Award to Shannon MacPhee.

vi TABLE OF CONTENTS

ABSTRACT...... ii CO-AUTHORSHIP ...... iv ACKNOWLEDGEMENTS ...... v TABLE OF CONTENTS ...... vii LIST OF ABBREVIATIONS ...... ix LIST OF TABLES ...... x LIST OF FIGURES ...... xi

CHAPTER 1: GENERAL INTRODUCTION AND LITERATURE REVIEW ...... 1 1.1 Global Climate Change ...... 1 1.2. Implications of Climate Change in the Boreal Shield Ecozone...... 1 1.3. Climate Effects on Lake Thermal Structure...... 2 1. 4. Freshwater Crustacean Zooplankton ...... 5 1.5 Implications of Climate Warming for Freshwater Zooplankton Communities ...... 7 1.6 Macroinvertebrate Predators ...... 9 1.7 Thesis Objectives ...... 13

CHAPTER 2 THE INFLUENCE OF LAKE THERMAL STRUCTURE ON MEAN ANNUAL CRUSTACEAN ZOOPLANKTON COMMUNITY STRUCTURE...... 17 Abstract...... 18 Introduction...... 19 Methods...... 22 Study site ...... 22 Field collections...... 23 Chemical analyses ...... 24 Zooplankton Enumeration ...... 24 Statistical analyses...... 25 Measures of zooplankton community structure ...... 26 Results ...... 28 Lake thermal habitat in warm and cool years ...... 28 Water chemistry ...... 29 Zooplankton community structure ...... 29 Discussion...... 30

CHAPTER 3 THE EFFECTS OF LAKE THERMAL HABITAT AND SPATIALLY- DEPENDENT MACROINVERTEBRATE PREDATION ON CRUSTACEAN ...... 43 ZOOPLANKTON COMMUNITY STRUCTURE ...... 43 Abstract...... 44 Introduction...... 46 Methods...... 51 Experimental design...... 51 Enclosure deployment and stocking...... 52 Enclosure sampling...... 54 Chemical analyses ...... 55

vii Zooplankton enumeration ...... 55 Statistical analyses...... 56 Results ...... 57 Enclosure conditions...... 57 Treatment effects on mean abundance and body size...... 59 Discussion...... 60

CHAPTER 4: GENERAL DISCUSSION AND FUTURE DIRECTIONS ...... 81

SUMMARY ...... 89 REFERENCES...... 90

APPENDIX I – SELECTED SWAN LAKE MONITORING DATA ...... 105 APPENDIX II– PREPARATION OF FORMALIN PRESERVATIVE...... 111 Procedure for making formalin – Sudbury method (Chapter 2)...... 111 Procedure for making formalin – Dorset method (Chapter 3)...... 112 APPENDIX III –DEVIATION FROM LONG-TERM MEAN AIR TEMPERATURE...... 113 APPENDIX IV – BODY SIZE MEASUREMENT ...... 122 APPENDIX V - PREY SELECTION ANALYSIS ...... 123 APPENDIX VI – SUPPLEMENTARY EXPERIMENTAL DATA ...... 129

viii

LIST OF ABBREVIATIONS

ANCOVA – analysis of covariance

ANOVA – analysis of variance

DOC – dissolved organic carbon

DVM – diel vertical migration

PCA – principle components analysis

PCA1 – the primary axis from a principle components analysis

RDA – redundancy analysis

RM – ANOVA – repeated-measures analysis of variance

SE – standard error of the mean

SRSi – soluble reactive silicon

UVR – ultraviolet radiation

VWT – volume-weighted water temperature

ix LIST OF TABLES

CHAPTER 2:

TABLE 2.1. Mean warm- and cool-year abundances (no. per m-3) of crustacean zooplankton functional groups...... 38

TABLE 2.2. Mean warm- and cool-year relative abundances (pi) of crustacean zooplankton functional groups. Significant values are in bold (P<0.05)...... 39

TABLE 2.3. Mean warm- and cool-year body size (mm) of commonly occurring crustacean zooplankton species in Swan Lake. Significant values are in bold (P<0.05)...... 40

CHAPTER 3:

TABLE 3.1. The effects of thermal habitat and the presence of Chaoborus and notonectids on mean total crustacean zooplankton abundance (no.per m3) pooled over depth using a three-way ANCOVA with initial density as a covariate. Significant values are in bold (P<0.05)...... 71

TABLE 3.2. The interactive effects of thermal habitat, Chaoborus, and notonectids on mean body size (mm) for Bosmina spp., calanoid copepodids, and adult Leptodiaptomus minutus using a three-way ANCOVA pooling over depth with initial body size (mm) as a covariate. Significant values are in bold (P<0.05)...... 72

TABLE 3.3. The effects of thermal habitat and the presence of Chaoborus and notonectids on abundance-weighted mean total crustacean zooplankton depth (m) using a three-way ANOVA. Significant values are in bold (P<0.05)...... 73

TABLE 3.4. Results of a 3 x 2 x 2 repeated measures ANOVA with depth as the repeated measure and thermal habitat and the presence of Chaoborus and notonectids as factors affecting total crustacean zooplankton community abundance (no. per m3). Univariate Greenhouse-Geiser (ε) F-values are reported. Significant values are in bold (P<0.05)...... 74

x LIST OF FIGURES

CHAPTER 1:

FIGURE 1.1 Conceptual framework of how climate warming will mediate processes that structure crustacean zooplankton community composition (adapted from Keller 2007)...... 14

FIGURE 1.2. Size class distribution of (a) Ontario lakes (data from Cox 1978); lakes greater than 10 000 ha not shown (n = 64), and (b) Global lakes (data from Downing et al. 2006); lakes greater than 1 000 ha not shown (n = 17 357). Note the differing size classes and frequency scales...... 15

FIGURE 1.3. Number of peer-reviewed primary and review articles published worldwide (in English) since 1975 related to (□) climate effects on lentic systems and to (●) implications of climate change for freshwater biological interactions. (□) I performed a search on ISI Web of Science using climate-related keywords (climate, temperature, weather, season) and common phrases for freshwater lentic systems (lake, loch, pond, reservoir); (●) I re-ran the analysis searching specifically for studies addressing questions related to climate effects on freshwater biological interactions (trophic, food-web, food chain, predator-prey, predation, competition)...... 16

CHAPTER 2:

FIGURE 2.1. (a) Map of Canada highlighting the Sudbury, Ontario region and (b) bathymetry of Swan Lake (46° 22' 81° 04'). Circled numbers represent sampling stations along a lake transect. Thick black line indicates where mesocosm experiments have been conducted (Chapter 3). Map courtesy of the Cooperative Freshwater Ecology Unit...... 41

FIGURE 2.2. PCA of selected meteorological and lake thermal variables from 1993 – 2002. Scaling is focussed on inter-sample (year) distances. Blue squares indicate cool years and orange circles indicate warm years. SPRING = mean monthly total degree days for May; ICE = mean air temperature for February, March, and April (°C) as a proxy for the timing of ice-out (Gao and Stefan 1999); VWT = mean annual volume-corrected whole-lake water temperature (°C); EPI = mean mid- summer (July) epilimnion temperature (°C); TURN = timing of fall turnover (Julian day); BTM = mean late-summer (August) bottom (8m) water temperature. .. 42

CHAPTER 3:

FIGURE 3.1. 3 x 2 x 2 factorial experimental design: three levels of a thermal habitat treatment (cool, warm, and hot), two levels of a Chaoborus treatment (present, absent) and two levels of a notonectid treatment (present, absent) with four replicates each...... 75

xi LIST OF FIGURES CONTINUED....

FIGURE 3.2. Lake temperature progression (°C) from 1 – 6 m over the 15 day duration of the (a) cool, (b) warm, and (c) hot thermal trials...... 76

FIGURE 3.3. Boxplots of mean crustacean zooplankton abundance (no. individuals per m3) by predator treatment within the (a) cool, (b) warm, and (c) hot thermal trials. Line represents median, box represents upper and lower (25 % and 75 %) quartiles, bars represent variance...... 77

FIGURE 3.4. Mean abundance-weighted depth by predator treatment for the (a) cool, (b) warm, and (c) hot thermal trials. Line represents median, box represents upper and lower (25 % and 75 %) quartiles, bars represent variance...... 78

FIGURE 3.5. Mean crustacean zooplankton abundance (no. individuals ·m-3) by thermal and predator treatments and by depth. Bars represent + SE...... 79

FIGURE 3.6. Mean total chlorophyll a concentration (µg·L-1) from 1 – 6 m on day 15 in the (a) cool, (b) warm, and (c) hot thermal trials. Bars represent + SE...... 80

xii CHAPTER 1: GENERAL INTRODUCTION AND LITERATURE REVIEW 1.1 Global Climate Change

Global climate change has occurred as a result of anthropogenic emissions of greenhouse gases that have altered the composition of the atmosphere (Hansen et al.

2007, IPCC 2007). Emissions produced from the combustion of fossil fuels are the primary source of carbon dioxide that has led to recent climate change (Hansen and Sato

2004, IPCC 2007). As a result of these emissions, mean global air temperatures have increased and will continue to rise throughout the 21st Century (IPCC 2007). Climate change will have profound effects on the global provision of ecological services through declines in biodiversity and altered structure and function of both terrestrial and aquatic ecosystems (Sala et al. 2000, Loreau et al. 2001, Brönmark and Hansson 2002, Walther et al. 2002, Millenium Ecosystem Assessment 2005).

1.2. Implications of Climate Change in the Boreal Shield Ecozone

The Boreal Shield Ecozone spans the northern hemisphere covering 20% of

Canada’s land mass and containing 22% of the country’s freshwater surface area. It is expected to be greatly impacted by climate warming - second in sensitivity only to Arctic regions (Urquizo et al. 2000, IPCC 2007). This region is economically important to resource- and recreation-based industry. In 2000, the Boreal Shield Ecozone generated a gross domestic product of approximately $50 billion CDN and accounted for 15% of

Canada’s resource-based employment (Urquizo et al. 2000). Future climate warming threatens the socio-economic status and the cultural and intrinsic value of the Boreal

Shield Ecozone and its supporting ecosystem services (Millennium Ecosystem

Assessment 2005) both directly and through interactions with multiple anthropogenic

1 stressors (Schindler 2001, Brönmark and Hanssen 2002, Keller 2007, Yan et al. 2008), for which climate change is an integrating factor (Fig. 1.1).

Climate change will alter not only the temperature, but also the thermal structure of Boreal Shield lakes (Fee et al. 1996, Magnuson et al. 1997, Snucins and Gunn 2000,

Keller et al. 2005), and these changes will occur both in time and space. Most temperate lakes thermally stratify (Wetzel 2001), which creates a habitat gradient in resource quality and quantity due to differences in temperature, oxygen, nutrients, mixing, light conditions

(Sommer 1985, Findlay et al. 2001) and predation refuges (Lampert 1989, Wissel and

Ramcharan 2003, Wissel et al. 2003a,b) for aquatic organisms. This spatially and temporally dynamic habitat allows the coexistence of a diverse assemblage of species compared to what otherwise might exist in a more homogeneous environment

(Hutchinson 1961). Changes in both the seasonal progression and spatial structure of lake thermal stratification may lead to thermal niche differentiation in time (e.g. Hampton

2005) or space (e.g. Helland et al. 2007), and these changes may alter predatory and competitive interactions (Schindler 2001). Climate-induced changes in lake thermal regimes will therefore have important implications for species assemblages in Boreal

Shield lakes.

1.3. Climate Effects on Lake Thermal Structure

Climate change will alter lake thermal structure through direct changes in meteorological factors such as air temperature, wind speed, and insolation (De Stasio et al. 1996, Gao and Stefan 1999, Cahill et al. 2005, Keller et al. 2005, IPCC 2007,

Tanentzap et al. 2008). However, in lakes <500 ha in size, transparency is the single most important determinant of lake thermal structure (Fee et al. 1996), where clear lakes warm deeply and coloured lakes strongly stratify with a warm epilimnion and cool hypolimnion

2 (Perez-Fuentetaja et al. 1999, Snucins and Gunn 2000). Transparency in oligotrophic

Shield lakes is principally determined by the coloured fraction of light-attenuating dissolved organic carbon (DOC) concentrations (Fee et al. 1996, Williamson et al. 1996) mainly derived from the catchment (Rasmussen et al. 1989). Climate regulates the production, decomposition and transport of DOC from a watershed (Schindler et al. 1997,

Pace and Cole 2002, Dillon and Molot 2005, Keller et al. 2008). Therefore, climate change will alter lake thermal habitat not only directly through meteorological forcing, but through complex processes that affect transparency and are mediated by land-water interactions (Fig. 1.1) (Keller et al. 2005, Keller 2007).

Models predict increases in the surface temperatures of stratified lakes of the

Boreal Shield and the Laurentian Great Lakes regions of 1–7 °C with a doubling of atmospheric CO2 (De Stasio et al. 1996, Magnuson et al. 1997). Later ice freeze-up and earlier break-up have been recorded across the Northern Hemisphere over a 150 year period (Magnuson et al. 2000) and recent observations include increased duration of the ice-free season in the Boreal Shield (Schindler et al. 1990, Futter 2003, Keller 2007). In moderately-sized, deep lakes the duration of the stratified season may also increase if stratification strengthens (De Stasio et al. 1996, Keller 2007). This increase in stratification strength may actually lead to increases in coldwater habitat and decreases in volume-weighted whole lake water temperature with climate change (De Stasio et al.

1996, Keller et al. 2005, Tanentzap et al. 2008).

In contrast, shallow lakes will experience increased mixed layer heating and, depending on transparency, weaker stratification stability and increased whole lake volume-weighted water temperature with warming (De Stasio et al. 1996, Schindler

1997). This may result in the complete loss of a thermocline (De Stasio et al. 1996) and

3 bottom coldwater habitat in shallow lakes (Moore et al. 1996, Cahill et al. 2005). Lakes of small volume and shallow depth will be especially sensitive to changes in the duration of ice cover and whole-lake water temperature because their limited thermal mass provides a minimal buffer against increases in air temperature and radiative forcing (Shuter et al.

1983, Gao and Stefan 1999, Gerten and Adrian 2001). Small lakes (< 10 ha) are abundant, numerically accounting for approximately 50% of the lakes in Ontario (Cox

1978) (Fig. 1.2a) with a similar distribution throughout the rest of the Boreal Shield.

Similarly, of the ~18 000 lakes >1 ha in the northeast USA, 70% are <10 ha in size

(Paulsen et al. 1991). In fact, using a combination of survey data and GIS, Downing et al.

(2006) determined small systems to be the most abundant lakes on the planet (Fig. 1.2b) and found that globally, small lakes cover more surface area than do larger systems. They further suggest that, since small lakes dominate the planet both in abundance and surface area, processes operating in small lakes and ponds may be globally significant.

Despite their vast abundance and cumulative surface area, small lakes have been largely ignored in major sampling programs, perhaps because they do not appear on many printed maps or GIS grids, because their surface area is too small to land a plane

(although they are accessible via helicopter), because they are generally too small to support cottage development, or because of their general lack of sport fish populations.

However, small lakes represent an important habitat for many minnow species, wildlife, and particularly breeding waterfowl (McNicol and Wayland 1992). They contain a diverse assemblage of macroinvertebrates (Bendell and McNicol 1987a,b) and both littoral and pelagic crustacean zooplankton species (Keller and Conlon 1994, Heneberry et al. 2009, Malkin et al. 2006). Also, their abundance on the landscape makes small lakes especially important as a source for the dispersal of colonists to new habitats (e.g. Cohen

4 and Shurin 2003). This is especially important in the context of biological recovery from acidification in Boreal Shield lakes, where dispersal may be a limiting factor in recovery(Yan et al. 2003).

1. 4. Freshwater Crustacean Zooplankton

Crustacean zooplankton (Phylum Arthropoda, Class Crustacea) are a widespread and diverse group of small-bodied (<5.0 mm) aquatic invertebrates (Covich and Thorp

2001). They have short generation times (on the order of weeks to months) and produce several generations per year (Allan 1976, Williamson and Reid 2001). They also inhabit variable environments which can change on the order of hours (during vertical migration) to days and weeks (seasonal succession). It is their relatively short generation times and widespread distribution in variable environments that make crustacean zooplankton excellent indicators of environmental change. Furthermore, zooplankton represent a critical link between primary producers and higher consumers in aquatic food webs

(Covich and Thorp 2001), and they respond to both bottom-up and top-down community structuring effects (Stemberger and Lazorchak 1994).

Crustacean zooplankton consist of two distinct Orders: the Copepoda (suborders

Calanoida and Cyclopoida) and Cladocera (Thorp and Covich 2001). Cladocerans can reproduce sexually or parthenogenically depending on resource availability and environmental stress (Dodson and Frey 2001). Copepods are obligate-sexual and demonstrate several morphologically distinct life history stages as nauplii and copepodids before adult maturation (Allan 1976, Williamson and Reid 2001). Both cladocerans and copepods can produce viable resting eggs that remain dormant and resistant to desiccation and can alter community structure via dispersal through time (Hairston 1996) as well as in space (e.g. Shurin 2001). In addition to dispersal abilities, crustacean zooplankton display

5 other unique adaptations including inducible morphological (e.g. Laforsch et al. 2004) and life history (Riessen 1999) adaptations to reduce predation risk. They can also display predator avoidance behaviour through a diel vertical migration (DVM) pattern, moving into dark bottom water to avoid visual predators and ultraviolet radiation (UVR) during the day (e.g. Boeing et al. 2004) and migrating up to feed in warmer waters at night. A reverse DVM pattern can occur in response to invertebrate predation (e.g. Ohman et al.

1983, Neill 1990). However, this behavioural avoidance of predators and UVR comes with a metabolic cost (Dawidowicz and Loose 1992) and as a trade-off with food availability and optimal temperature for growth and reproduction in stratified lakes

(reviewed in Lampert 1989).

Crustacean zooplankton must tolerate a range of temperatures to survive within their broad geographic distribution and seasonal environments (Carter et al. 1980, Patalas

1990, Gillooly and Dodson 2000). Climate, often varying along a latitudinal scale, is the single most important factor in determining regional crustacean zooplankton species richness (Patalas 1990, Gillooly and Dodson 2000). Patalas (1990) found that richness is maximized at mean July air temperatures of approximately 15°C but declines at higher temperatures. However, there is considerable variation in optimal water temperatures for the growth and reproduction of crustacean zooplankton in temperate lake systems

(reviewed in Moore et al. 1996) and some clonal species have been shown to genetically adapt to various thermal optima (Carvalho 1987, Mitchell and Lampert 2000).

Direct, physiological thermal stress has been observed in zooplankton at temperatures exceeding 25°C where fecundity, reproductive success and survival rates of many Daphnia species decrease (Moore et al. 1996). Thermal tolerances of copepods are less well known (Moore et al. 1996). Zooplankton are sensitive to climate warming

6 because optimal temperatures lie only slightly below lethal temperatures (Mitchell and

Lampert 2000). Temperature also mediates sensitivity to other stressors such as toxic pollutants (Moore and Folt 1993), calcium depletion (Ashforth and Yan 2008), and UVR

(Williamson et al. 2002), and can have implications for the colonization success and range expansion of both native and exotic zooplankton species (Lennon et al. 2001,

Holzapfel and Vinebrooke 2005). Temperature controls on physiological processes in crustacean zooplankton include, but are not limited to, controls on locomotion (swimming speed), filtering/feeding efficiency, body size at maturation, rates of growth and reproduction, the timing of the switch from hatching to diapausing eggs, and, ultimately, survival (Allan 1976, Hairston et al. 1990, Atkinson 1994, Moore and Folt 1993, Moore et al. 1996, Gillooly 2000). While the thermal biology is well-described for some crustacean zooplankton species, ecological complexity makes predicting community-level changes associated with climate-induced shifts in lake thermal regime a challenge.

1.5 Implications of Climate Warming for Freshwater Zooplankton Communities

In addition to the direct effects of temperature on individuals and populations, climate warming will impact complex community-level processes. Trophic dynamics regulate many lake processes (Brooks and Dodson 1965, Carpenter et al. 1987, Vanni and

Layne 1997) but differential responses between trophic levels and complex inter- and intra-specific interactions make predicting the community-level effects of climate warming difficult (reviewed in Stenseth et al. 2002, Walther et al. 2002). Also, the effects of climate change on food web processes are difficult to predict because they occur not only over broad temporal scales (seasonally, inter-annually) but also in space due to thermal habitat heterogeneity and temporal (e.g. Hampton 2005) and spatial (e.g. De

Stasio et al. 1996, Helland et al. 2007) niche differentiation facilitated by lake thermal

7 stratification. Our current ability to forecast freshwater ecosystem responses to future climate change is limited by our understanding of food-web responses (Schindler 2001,

Keller 2007), but this critical field of research is slowly gaining momentum (Fig. 1.3).

To date, much of the research on climate-related changes in freshwater zooplankton communities has focussed on temperature-mediated differences in population demographic responses (reviewed in Moore et al. 1996; Chen and Folt 1996,

Chen and Folt 2002) or on phytoplankton-zooplankton dynamics (Beisner et al. 1996,

1997; Strecker et al. 2004). Recent work determined that large-scale climate patterns can drive the synchronous response of Boreal Shield zooplankton populations to climate across a regional scale (Rusak et al. 1999, Arnott et al. 2003, Rusak et al. 2008). Also, species-specific phenological responses to warming (e.g. Chen and Folt 2002, Gerten and

Adrian 2002, Hampton 2005) often associated with wide variations in large-scale climate indices (ecological responses reviewed in Stenseth et al. 2002; climate indices reviewed in Stenseth et al. 2003) can lead to the temporal uncoupling of trophic interactions. This uncoupling can result in a ‘mismatch’ (sensu Cushing 1974) between peak zooplankton population densities with the spring phytoplankton bloom (Winder and Schindler 2004a, b) or an increased overlap of zooplankton with planktivorous fish (Wagner and Benndorf

2007), both which have led to reductions in zooplankton abundance in large (>500 ha) lakes.

Climate-related changes in lake thermal habitat structure may also alter spatially- dependent predator-prey interactions. The few previous studies examining the potential spatial uncoupling of trophic interactions with climate change include model simulations

(De Stasio et al. 1996, Jansen and Hesslein 2004) and predictions based on seasonal thermal niche selection in planktivorous fish (Helland et al. 2007) in deep lakes with

8 stable hypolimnia. Thermal regime changes could be especially important in shallow lakes that may lose stratification, coldwater habitat, and thermal refuges with future climate warming, but these effects have not been investigated. Previous investigations into climate-related changes in small lake zooplankton communities are limited to studies in shallow eutrophic ponds (McKee et al. 2002, Carvalho and Kirika 2003, Moss et al.

2003, Gylleström et al. 2005, Jackson et al. 2007) that are not typical of the Boreal

Shield.

1.6 Macroinvertebrate Predators

Macroinvertebrate predators have become extremely abundant throughout parts of the Boreal Shield Ecozone following the extirpation of fish from thousands of historically acidified small lakes (Bendell and McNicol 1987a, Arnott et al. 2006). They are also widespread in naturally fishless lakes (Malkin et al. 2006), ponds, streams, wetlands, and systems containing fish (Merritt and Cummins 1996). Macroinvertebrates represent the top predators in fishless systems (Arnott et al. 2006) and although invertebrate predators have a lower overall feeding rate than planktivorous fish, they are found at much higher densities and can therefore exert a significant impact on zooplankton community structure

(Riessen et al. 1988). Macroinvertebrate predators consist of a diverse group of orders with many predatory species capable of regulating crustacean zooplankton assemblages

(e.g. Murdoch and Scott 1984, Notonectidae; Yan et al. 1991, Chaoboridae; Arnott et al.

2006, Dytiscidae).

Perhaps the most voracious and arguably the most abundant and widespread

(Carter et al. 1980) macroinvertebrate predator of crustacean zooplankton on the Boreal

Shield is the planktonic phantom midge larvae, Chaoborus (Family: Chaoboridae).

Chaoborus has a broad zoogeographic distribution throughout North America (Carter et

9 al. 1980). Third and fourth instar Chaoborus can over-winter in lake sediments at approximately 4°C with first emergence occurring in early June and peak abundance in early July (von Ende 1982). Many species vertically migrate into lake sediments to avoid fish predation (Wissel et al. 2003a) and UVR (Persaud et al. 2003) during the day. In fact, species that do not exhibit diel migration, such as C. americanus, are only abundant in fishless lakes (von Ende 1979, Wissel et al. 2003a, Garcia and Mittelbach 2008).

Chaoborus punctipennis Say is a small species which often coexists with fish and is commonly found in shallow lakes of the Boreal Shield (Yan et al. 1991, Wissel et al.

2003a).

Chaoborus are visual and tactile ambush predators with a Type II (i.e. density dependent) functional response (Holling 1966, Pastorek 1980, Spitze 1985, Moore

1988a). They are able to feed with equal efficiency in both the light and dark (Swift and

Forward 1981). Chaoborus can significantly impact both the structure and function of zooplankton communities, consuming a significant portion of zooplankton production

(Yan et al. 1991) and altering the population size structure by size-selective predation

(Riessen et al. 1988, Arnott and Vanni 1993, Wissel et al. 2003b). They have four instars that exhibit different feeding preferences, where early instars prefer to feed on rotifer prey and later instars select for crustacean zooplankton (Fedorenko 1975a, Moore and Gilbert

1987, Moore 1988b). Chaoborus can induce behavioural, morphological, and life history responses in crusctacean zooplankton prey (Nesbitt et al. 1996, reviewed in Riessen

1999).

Notonectidae (backswimmers) are a surface-orienting predaceous family of the order Heteroptera (“true bugs”) distributed globally in the nekton of small freshwater lakes and ponds (Merrit and Cummins 1996). Within North America the family has been

10 reported to occur in all provinces and territories in Canada except Nunavut, and all

American states except Alaska (Chordas et al. 2005). Notonectids are abundant in many small lakes and ponds across the Boreal Shield Ecozone, particularly in naturally fishless lakes and the thousands of lakes and ponds where fish were locally extirpated due to acid deposition (Bendell 1986, Bendell and McNicol 1987a). The two most common species of Notonectidae occurring in the Boreal Shield are Notonecta borealis Bueno and Hussey, and Buenoa macrotibialis Hungerford (Bendell and McNicol 1987b).

Notonectids typically experience six life stages and over-winter as eggs (Rice

1954, Hampton et al. 2000). Hatching is temperature-dependent, but in general hatching occurs in mid-May, adults begin appearing by late June (Hampton et al. 2000, S.

MacPhee, personal observation), and first mating occurs in early July (Rice 1954).

Notonectids are surface-orienting ambush planktivores that feed on other aquatic insects and zooplankton by piercing and sucking with a beak-like rostrum (Gittelman 1974).

They are extremely active near the surface in the pelagic zone (Bendell and McNicol

1987b, Hampton 2004) but are also found patchily distributed amongst macrophytes in the littoral zone (Hampton 2004) of small lakes. Notonectids represent an important food source for fish and particularly for juvenile waterfowl (McNicol and Wayland 1992), but are also voracious predators capable of reducing crustacean zooplankton species richness, abundance, and biomass (Hampton et al. 2000, Shurin 2001). They are visual and mechanosensing predators (Cooper et al. 1985) with a Type II functional response

(Holling 1966, Streams 1994) that can forage during both day and night, but the ability to detect smaller prey is reduced in the dark (Diéguez and Gilbert 2003, Hampton 2004).

Direct lethal effects can create significant changes in zooplankton prey community structure, particularly because notonectids typically select for the largest available prey

11 items and often reduce the abundance of large Daphnia and copepods (Shurin 2001,

Hampton et al. 2000, Diéguez and Gilbert 2003). Selection for larger zooplankton prey items can create trophic cascades that release smaller zooplankton, such as rotifers, from competition and predation (Hampton et al. 2000, Gilbert and Hampton 2001). Other indirect predation effects of notonectids on zooplankton prey include inducing behavioural (Herwig and Schindler 1996, Nesbitt et al. 1996, Hampton 2004) morphological (Riessen 1999), and life history (Dodson and Havel 1988) responses.

Higher water temperatures associated with climate warming may increase the metabolic demand of predators (e.g. Buns and Ratte 1991) and perhaps increase recruitment by adding additional generations within a single year (Gillooly 2000).

Predator body size may also decrease with increased temperature, which is a well- described pattern for most ectotherms (Hanazato et al. 1989, Atkinson 1994). Warmer falls and winters associated with a warming climate may increase the over-wintering recruitment success or size-structure of macroinvertebrate predators (Moore et al. 1996) with potential consequences for the crustacean zooplankton prey community. Conversely, over-wintering success may actually be reduced in warmer winters when metabolic demands are high but food is limiting (Neill 1988).

In addition to the potential direct effects of temperature on macroinvertebrate foraging rate, population size-structure, and recruitment (Moore et al. 1996), changes in the strength of lake stratification may alter the spatial coupling of predator-prey interactions. Potential interactions between thermal stratification and predation could be especially important for biological interactions in small, shallow lakes that are expected to weakly stratify or become warm isothermal with future climate warming. Small lakes are highly sensitive to shifts in thermal regime (Shuter et al. 1983, Gao and Stefan 1999,

12 Gerten and Adrian 2001) and often represent habitats where macroinvertebrate predators are the main top-down force structuring crustacean zooplankton communities (Bendell and McNicol 1987a, Arnott et al. 2006). Changes in lake thermal structure may have implications for the food acquisition/surface-predator avoidance trade-off experienced by zooplankton (Lampert 1989, Herwig and Schindler 1996) if predation risk is altered through changes in the spatial overlap and encounter rate of zooplankton and macroinvertebrate predators (Fedorenko 1975b, Williamson et al. 1989, Williamson

1993). The potential interaction between lake thermal structure and macroinvertebrate predation has not been investigated and may have important implications for crustacean zooplankton community structure with future climate change.

1.7 Thesis Objectives

The objectives of this thesis are to (1) determine characteristics of lake thermal habitat structure in warm conditions in a typical small, shallow Boreal Shield lake; (2) determine how climate-induced changes in the thermal habitat of small Boreal Shield lakes affect crustacean zooplankton community structure at the annual scale; and (3) determine if predation interacts with lake stratification to mediate these effects. I accomplished this using 10 years of monitoring data from a small lake to determine if there are differences in crustacean zooplankton body size, abundance and diversity measures between warm and cool lake years. I also used a field experiment to determine if spatially-dependent macroinvertebrate predation on crustacean zooplankton depends on lake thermal habitat structure. Investigating how changes in lake thermal structure will affect crustacean zooplankton community structure and predator-prey interactions in small, shallow lakes will help predict the effects of climate change for thousands of lakes across the Boreal Shield.

13 Changing Climate

Effects on Direct Effects Watersheds On Lakes (DOC, Nutrients, (Sunshine, Wind, Hydrology) Air Temperature) Altered Thermal Recovery from Habitat Altered Acidification (Spatially, Temporally) pathway of introductions

Other stressors Species Range (Acid, Metals, UV, Expansions Base Cations) Food Resources (Native & Exotic) Quality Quantity

Predators Competitors Recruitment Nutrient requirements Foraging rate Grazing ability Size-structure Over-winter strategy Overlap Reproductive success Predator avoidance Niche overlap (space, time)

Niche overlap (space, time)

Zooplankton Population Zooplankton Individual Demographic Responses Physiological Responses

Crustacean zooplankton community

FIGURE 1.1. Conceptual framework of how climate warming will mediate processes that structure crustacean zooplankton community composition (adapted from Keller 2007).

14

(a) 0.50

0.40 0.30

0.20

n = 9 847 n = 88 836 n = 81 426 0.10 n = 842

0.00

999 9999 100 - 1000 - 1 - 9.9 10 - 99

1.0

Proportion of lakes (%) (b)

0.8

0.6

0.4

0.2 n = 2 097 000 n = 182 300 n = 241 200 000

n = 277 4000 000 0.0

-

10 - 99 0.1-1 999 100

1 - 9.9

Size Class (ha)

FIGURE 1.2. Size class distribution of (a) Ontario lakes (data from Cox 1978); lakes greater than 10 000 ha not shown (n = 64), and (b) Global lakes (data from Downing et al. 2006); lakes greater than 1 000 ha not shown (n = 17 357). Note the differing size classes and frequency scales.

15

Number of publications

Publication year

FIGURE 1.3. Number of peer-reviewed primary and review articles published worldwide (in English) since 1979 related to (□) climate effects on lentic systems and to (●) implications of climate change for freshwater biological interactions. (□) I performed a search on ISI Web of Science using climate-related keywords (climate, temperature, weather, season) and common phrases for freshwater lentic systems (lake, loch, pond, reservoir); (●) I re-ran the analysis searching specifically for studies addressing questions related to climate effects on freshwater biological interactions (trophic, food-web, food chain, predator-prey, predation, competition).

16

CHAPTER 2

THE INFLUENCE OF LAKE THERMAL STRUCTURE ON MEAN ANNUAL CRUSTACEAN

ZOOPLANKTON COMMUNITY STRUCTURE

17 Abstract

Climate change will alter the thermal structure of Boreal Shield lakes. Small lakes

(< 10 ha) numerically comprise almost 50% of the freshwater resources within the Boreal

Shield Ecozone and are particularly responsive to climate forcing on thermal structure.

This sensitivity in thermal habitat has biological implications. However, long-term monitoring data from which to assess potential climate change impacts on small lakes are sparse. In the current study, Swan Lake (5. 8 ha, max depth 8.8 m, mean depth 2.2 m) near Sudbury, ON, was sampled approximately twice-monthly for > 10 years and was used as a model small lake system. A multivariate approach was used to identify years from 1993 – 2002 as ‘warm’ or ‘cool’ based on several key characteristics related to the seasonal structure of lake thermal habitat experienced by aquatic biota, such as the timing of ice-out and late summer bottom water temperature. In general, warm years had earlier ice-out, higher mid-summer surface and late-summer bottom water temperatures, and later fall turnover than cool years. Several univariate indices of crustacean zooplankton community structure (abundance, richness, evenness, diversity, body size) were tested for broad overall differences at the annual scale between warm and cool years. Mean body sizes of the two most dominant taxa, Leptodiaptomus minutus and Bosmina spp., were significantly reduced by 3% and 11%, respectively, in warm years. Also, the relative abundance of small cladocerans increased in warm years, concurrent with declines in the proportion of calanoid copepods. Zooplankton abundance appears resilient to impact from stressors such as climate change, at least at the annual scale, and this is likely due to species compensation over seasonally variable conditions. However, a community shift towards smaller-bodied individuals and species with future climate change has

18 implications for ecosystem functioning and the transfer of nutrients and energy to higher trophic levels.

Introduction

Climate warming will affect both the physicochemical and biological nature of freshwater ecosystems of the Boreal Shield (Schindler 2001, Keller 2007). These effects may be different in large, deep versus small, shallow lakes because morphometry is an important factor in determining lake thermal structure (Fee et al. 1996). Small lakes numerically dominate the Boreal Shield. For example, small lakes (< 10 ha) account for almost 50% of the lakes in Ontario alone (Cox 1978) with a similar pattern in Quebec

(Downing et al. 2006). Understanding how climate warming will affect the physical structure and biota of small lakes is important because many processes related to lake metabolism and primary productivity, such as internal nutrient recycling, are related to the epilimnetic area-to-volume ratio (Fee 1979, Kelly et al. 2001), which, in general, scales to lake size (McKee et al. 2002). Also, biotic factors such as predation and competition may be more important in small lakes where space, habitat heterogeneity, and predation refugia are limited (Roff et al. 1981, Keller and Conlon 1994, Moore et al.

1996, Post et al. 2000). Small lakes represent an important habitat for wildlife and particularly for breeding waterfowl (McNicol and Wayland 1992) and represent a source of colonisers for dispersal to new habitats (Cohen and Shurin 2003).

The thermal habitat of small lakes may be particularly vulnerable to climate change. In comparison to larger systems, small, shallow lakes have limited thermal mass and therefore respond more quickly to changes in climate and local weather conditions

(Shuter et al. 1983, Gao and Stefan 1999, Arnott et al. 2001, Gerten and Adrian 2001,

McKee et al. 2002). The allochthonous production and transport of dissolved organic

19 carbon (DOC), the main determinant of lake transparency and thermal structure in Boreal

Shield lakes (Fee et al. 1996, Perez-Fuentetaja et al. 1999, Snucins and Gunn 2000, Keller et al. 2005), varies with temperature and precipitation (Schindler et al. 1997, Pace and

Cole 2002, Dillon and Molot 2005, Keller et al. 2008). Small lakes typically have relatively high DOC concentrations due to small surface area – to – catchment ratios and short residence times leading to reduced in-lake DOC removal (Fee and Hecky 1992,

Keller and Conlon 1994, Schindler et al. 1997). Short residence time also leads to greater variation in inter-annual concentrations of DOC in small lakes (Schindler et al. 1990,

Curtis and Schindler 1997, Keller et al. 2008).

These factors such as residence time, DOC concentration, and surface area – to – catchment ratio can cause small or shallow lakes to respond differently to climate than larger systems within a region (e.g. Gerten and Adrian 2000, Arnott et al. 2003). For example, while moderate-sized lakes may actually experience decreased volume- corrected whole-lake and bottom water temperature related to increased thermal stability with climate change (Hondzo and Stefan 1991, De Stasio et al. 1996, Keller et al. 2005,

Tanentzap et al. 2008), small, shallow lakes may become isothermal with bottom water temperatures exceeding 20°C by late-summer in warm years (Schindler et al. 1990,

Moore et al. 1996, Schindler 1997). This loss of a thermal refuge may have important implications for aquatic biota.

Changes in lake thermal habitat have profound effects on freshwater plankton communities (Schindler 2001, Keller 2007). Climate-related changes in crustacean zooplankton communities are of much ecological interest because zooplankton represent a critical trophic link between primary producers and higher level consumers, and therefore are able to respond to both top-down and bottom up structuring forces

20 (Stemberger and Lazorchak 1994). Effects of warming conditions on zooplankton community structure include physiological and metabolic responses to temperature that typically result in intra-specific reductions in body size, a community shift to dominance by smaller-bodied species, and reductions in richness and biomass, but not abundance

(Moore and Folt 1993, Atkinson 1994, Moore et al. 1996, Stemberger et al. 1996,

Gillooly and Dodson 2000, McKee et al. 2002, Strecker et al. 2004, Yan et al. 2008).

Additionally, warming leads to increases in reproductive rates and decreases in generation time (Allan 1976, 1977; Gillooly 2000) that are species-specific and temperature- dependent (Allan 1976, 1977; Chen and Folt 2002). These differential demographic responses will structure zooplankton communities under future climate warming scenarios.

To date, most studies of the effects of climate change on plankton communities in small lakes have been limited to examinations in macrophyte-dominated eutrophic systems (McKee et al. 2002, Carvalho and Kirika 2003, Moss et al. 2003, Gylleström et al. 2005, Jackson et al. 2007). There have been few investigations into the effects of climate warming on the crustacean zooplankton communities in the more dilute, small, shallow lakes of the Boreal Shield, although some paleolimnological data exist (e.g.

Rühland et al. 2008). Previous studies have examined lake physical (Snucins and Gunn

2000) or biological differences (Chen and Folt 1996, 2002) between single warm and cool years or consecutive warm/cool periods (Gerten and Adrian 2000, Gerten and Adrian

2002) in single, or multiple (Stemberger et al. 1996) lakes. In contrast, the present study compares differences in zooplankton community structure between multiple, non- consecutive warm and cool years in a single small, shallow lake. Since aspects of lake morphometry are important in determining thermal structure (Fee et al. 1996), the

21 advantage of using years as replicates is that morphometry is held constant and there is reduced variability in local, within-lake factors such as nutrient dynamics, food-web dynamics, and other stressors - unless they interact with climate (e.g. DOC concentration, predator recruitment). Therefore, climate signals should represent an important source of variation within the system.

The objectives of this study were to (i) characterize years as ‘warm’ or ‘cool’ based on several key lake thermal habitat parameters and (ii) to examine broad-scale differences in crustacean zooplankton community structure (mean annual abundance, body size, and diversity measures) between warm and cool years.

Methods

Study site

Swan Lake (46°22' 81°04') is small (5.8 ha), shallow (mean depth 2.8 m, maximum depth 8.5 m) lake located approximately 13 km from the Copper Cliff smelter in Sudbury, Ontario, Canada (Fig. 2.1). It is recovering from the effects of acid deposition

(Keller et al. 1992a, Arnott et al. 2001), and therefore is representative of many thousands of Boreal Shield lakes that are expected to recover across eastern Canada (Jeffries et al.

2003). Also, the lake has undergone periods of drought-induced re-acidification (Keller et al. 1992a, Yan et al. 1996, Arnott et al. 2001, Appendix I Fig. I.2). Swan Lake responds quickly to changes in weather because of its short residence time (~ 0.6 yr) and limited thermal mass. Water chemistry records from the Ontario Ministry of the Environment (W.

Keller, Cooperative Freshwater Ecology Unit, Sudbury, Ontario, Canada, unpublished data) indicate an overall recovery of average annual pH from < 4.0 in 1977 to ~ 5.7 in

2008.

22 Swan Lake is fishless, although paleolimnological investigations of Chaoborus head capsule remains indicate that it contained fish prior to industrialization (Uutala and

Smol 1996). The lake contains no Daphnia and the zooplankton community is dominated

(> 90% of total crustacean biomass) by the acid-tolerant calanoid copepod

Leptodiaptomus minutus Lilljeborg, with lower densities of the small cladocerans

Bosmina (Bosmina) spp. and Diaphanosoma birgei Kořínik (Appendix I Table I.2, Fig.

I.1), which are three of the most ubiquitous crustacean zooplankton species across eastern

North America (Carter et al. 1980).

Field collections

Zooplankton were sampled approximately every two weeks during the ice-free season from the late 1970s to the present using a 30 L, 30 µm mesh Schindler / Patalas sampler. Zooplankton samples were collected from three stations along a lake transect

(Fig. 2.1) and preserved in 13% sugared, buffered formalin (Appendix II). Three zooplankton samples at 1 m and 4 m, and one sample at 7 m were collected to account for patchiness in zooplankton distribution. Volume-weighted water samples for chemical parameters (metals, ions, pH / conductivity) were collected from the lake surface to the lake bottom, and nutrients (total phosphorus, DOC, total Kjeldahl nitrogen, etc., see

Chemical Analyses below) were collected from the euphotic zone (defined as twice the

Secchi depth). Total chlorophyll a was sampled discretely at 0.5 m, 4.5 m, and 7.5 m

-1 using a van Dorn sampler; samples were chilled and preserved with 1.5 mg ·L of CaCO3 prior to filtration and extraction. Dissolved oxygen and temperature profiles were measured using a YSI model 54/55-type probe (Yellow Springs Institute, Yellow Springs,

Ohio, USA), and Secchi depth and water level were recorded. Weather data (no. bright sunshine hours, i.e. the hours of sunshine intense enough to burn a paper card in a

23 Campbell-Stokes recorder; air temperature (°C), wind velocity (km ·hr-1) and precipitation (mm)) were measured hourly from 1955 – 2008 at the Greater Sudbury

Airport Environment Canada weather station located approximately 30 km from Swan

Lake and were obtained from Environment Canada’s National Climate Data and

Information Archive (Environment Canada 2008).

Chemical analyses

Conductivity and pH were measured using a CDM2 and PHM64 Research probe, respectively (Bach Simpson Ltd., London, ON, Canada). Water chemistry samples were analyzed for nutrients (Total Kjeldahl nitrogen, total phosphorus, total chlorophyll a,

DOC), soluble reactive silica (SRSi) and calcium following Ontario Ministry of the

Environment standard protocols (Ontario Ministry of the Environment 2009a,b,c,d,e, respectively).

Zooplankton Enumeration

Zooplankton were enumerated with a protocol designed to reasonably represent the occurrence of rare species (Girard and Reid 1990). Sub-samples were generated with a Folsom plankton splitter and a minimum of 250 individuals were enumerated, with a maximum of 30 nauplii and 40 adults and copepodids of each species counted per sample.

For all samples, adult zooplankton were identified to species and juvenile copepods as

Calanoida or Cyclopoida. Body length was measured using the semi-automated sample processing system ZEBRA2M (Allen et al. 1994, Pawson and Yan 1993), with an average of 6930 individuals measured per year. A Leica MZ 12.5 dissecting microscope (Leica

Microsystems, Richmond Hill, Ontario, Canada) and the same taxonomists were used to process all samples. Bosmina (Bosmina) freyi and Bosmina (Bosmina) liederi were pooled

24 due to uncertainty in taxonomic identification (Taylor et al. 2002). See Appendix I (Table

I.1) for a complete list of taxonomic revisions.

Statistical analyses

The entire data record for Swan Lake covers a wide range in both acidity and local weather conditions (Appendix I Table I.3, Fig. I.2, Fig. I.3). To limit confounding results in effects on zooplankton richness and abundance related to recovery from acidification, the analyses were restricted to only include data after 1992. This is after the period of increase in pH and DOC, and stabilization of phytoplankton community structure

(Graham et al. 2007) following the 1988 drought-induced re-acidification event (Yan et al. 1996, Arnott et al. 2001, Appendix I Fig. I.2). At the time of the current study, zooplankton data were only available through 2002.

A multivariate analysis was used to classify Swan Lake years from 1993 – 2002 as

‘warm’ or ‘cool’. The seasonal progression of lake thermal structure was expected to be biologically important, therefore warm and cool lake years were quantified based on parameters related to the timing and strength of thermal stratification. Variables related to the seasonal progression of lake thermal structure were obtained from annual season depth-time isotherm diagrams (Appendix III Fig. III.2) produced in the program

MATLAB® (The MathworksTM Inc, Natick, MA) using water temperature data obtained from HOBOware® Tidbit temperature loggers (www.onsetcomp.com/products /data- loggers) deployed at 1m intervals throughout the ice-free season (years > 2000) and from approximately twice-monthly collected thermal profile data.

To quantify the extent of lake warming, mean annual ice-free season volume- corrected whole lake water temperature (°C) and mean mid-summer (July) epilimnion temperature (°C) were included in the analysis. The epilimnion was defined as the

25 surface layer where the gradient of temperature change was < 0.5°C ·m-1 (Wetzel 2001).

Lake thermal data for early spring were not available in some years and were not included in the analysis because multivariate ordination techniques do not handle missing data

(Lepš and Šmilauer 2003). Instead, mean air temperature (°C) for February-April was included as a proxy for the timing of ice-out (Gao and Stefan 1999). The ice-out proxy together with total degree days for May (no. days) was a combined measure of the time to thermal stratification (adapted from Cahill et al. 2005). Time to fall turnover (Julian day) was readily determined from thermal isopleth figures (Appendix III Fig. III.2). Mean

August bottom water temperature (°C) was also included in the analysis because Swan

Lake is shallow and stratification breaks down by late-summer in some years, and this may affect zooplankton abundance (e.g. Chen and Folt 1996).

The selected ice-free season lake thermal habitat parameters described above were input into a preliminary detrended correspondence analysis (DCA) using the statistical package CANOCO 4.5 (ter Braak and Šmilauer 2002). Based on the DCA it was determined that linear ordination methods were appropriate because the gradient length of the first axis was short (< 3) (Lepš and Šmilauer 2003). The multivariate linear ordination technique principle components analysis (PCA) was then used to visually separate Swan Lake years as ‘warm’ or ‘cool’ by approximating dissimilarity between sample points (lake years) based on their Euclidean distance (Lepš and Šmilauer 2003).

Graphical scaling focussed on inter-sample (i.e. inter-annual) distances.

Measures of zooplankton community structure

Several univariate metrics were compared between warm and cool lake years using two-tailed Student’s t-tests. Response variables included mean annual adult crustacean zooplankton species richness, evenness, and diversity, mean annual total

26 crustacean zooplankton abundance (copepodids and nauplii included), and mean annual body size by taxa. Total body length for Cladocerans was measured along the longest body axis from the top of the head to the base of the spine, excluding the tail spine.

Copepod total length was measured along the longest body axis from the top of the head to the base of the caudal rami, excluding caudal setae (details in Appendix IV). Mean annual zooplankton species richness was calculated as the total number of taxa (excluding juveniles) across all depths (1 m, 4 m, 7 m) per sampling date, divided by the number of sampling dates per year. The corrected Shannon-Wiener diversity index (Jost 2006) was calculated as:

S (1) ' −= ∑ ln pipiEXPH j=1 where S is the species richness of the sample and pi is the proportion of S composed of the ith taxa.

The formula for Evar was used because it is recommended as the best evenness measure for general use (Smith and Wilson 1996):

2 ⎡ S ⎛ S ⎞ ⎤ (2) E Π−= arctan/21 ⎢ ⎜ pi − ⎟ //)ln()ln( SSpi ⎥ var ∑∑⎜ ⎟ ⎣⎢ i==11⎝ j ⎠ ⎦⎥ where pi is the number of individuals of taxa i, jn is the number of individuals of taxa j, and S is the total number of taxa in a sample. Evenness ranges from 0 (minimum evenness) to 1 (maximum evenness). Parametric assumptions were verified using histograms and boxplots. Abundance data were log10-transformed to improve homogeneity of variance.

27 Results Lake thermal habitat in warm and cool years

The multivariate analysis captured seasonal aspects of lake thermal structure that differed between lake years. Three ‘warm’ and seven ‘cool’ lake years were identified from 1993 – 2002 (Fig. 2.2). The timing of ice-out, as inferred from late winter air temperature (Gao and Stefan 1999), was earlier in warm compared to cool years and loaded strongly on the primary axis, while late-summer bottom water temperature and the timing of fall turnover were associated with the second axis (Fig. 2.2). In general, warm years stratified earlier, reached higher mid-summer epilimnion temperatures, and had later fall turnover than cool years. However, one warm year (2002) had a relatively cool spring but had high mid-summer epilimnion temperatures, high late-summer bottom water temperatures, and late fall turnover. Cool years typically had a slow progression of warming (see isopleth figures in Appendix III Fig. III.2), cooler epilimnia, and earlier fall turnover than warm years. Warm years were isothermal for periods during late summer, but this also occurred in some cool years (i.e. 1997) that had a very slow, constant progression of warming. However, surface temperatures were always higher in warm compared to cool years. In order to produce a balanced experimental design, the three warm (1998, 1999, 2002) and three coolest (1993, 1994, 1997) years for which corresponding zooplankton data existed were selected as replicates for analyses of zooplankton community structure.

Warm years had higher mean annual air temperature (Student’s t-test, t=4.37,

P=0.04) and number of bright sunshine hours (Student’s t-test, t=2.17, P=0.04) than cool years. Mean annual precipitation (Student’s t-test, t=0.89, P=0.45) and wind velocity

28 (Student’s t-test, t=0.01, P=0.96) did not differ between warm and cool Swan Lake years

(Appendix I Fig. I.3).

Water chemistry

There were no significant differences in mean annual pH (two-tailed Student’s t- test; t=-0.96, P=0.42) or concentrations of total phosphorus (two-tailed Student’s t-test; t=-0.45, P=0.68), total Kjeldahl nitrogen (two-tailed Student’s t-test; t=-1.68, P=0.17), total chlorophyll a (two-tailed Student’s t-test; t=1.75, P=0.20), calcium (two-tailed

Student’s t-test; t=0.36, P=0.74) or SRSi (two-tailed Student’s t-test; t=0.17, P=0.87) between warm and cool years from 1993 – 2002 (Appendix I Table. I.1, Fig. I.2). Mean annual concentrations of DOC were significantly greater in warm than cool years (two- tailed Student’s t-test; t=2.32, P=0.04) but the warm and cool year means only differed by

0.3 mg ·L-1.

Zooplankton community structure

There were no depth-related differences in mean annual zooplankton abundance

(RM-ANOVA; depth x year-type: F (1.5, 12.0) =38.40, P=0.35) or richness (RM-ANOVA; depth x year-type: F (1.3, 10.4) =0.52, P=0.53) between warm and cool years; therefore, discrete depth samples were pooled for further analysis by averaging across the three depths (1 m, 4 m, 7 m). Total crustacean zooplankton richness (two-tailed Student’s t- test; t =1.06, P=0.38), evenness (two-tailed Student’s t-test, t=-0.21, P=0.85), and diversity (two-tailed Student’s t-test, t=1.62, P=0.25) were not significantly different between warm and cool years in Swan Lake. Most species were rare, often with only a single occurrence (Appendix I Fig. I.2). Crustacean zooplankton community biomass was

> 99% dominated by L. minutus and Bosmina spp. in both warm and cool years. There were no significant differences in mean annual crustacean zooplankton abundance (two-

29 tailed Student’s t-test, t=-0.11, P =0.92) or the abundance of any functional group (Table

2.1) between warm and cool years. However, the mean relative abundance of small cladocerans increased from 8% to 13% in warm years, concurrent with declines in the mean relative abundance of calanoid copepods from 90% - 84% of community composition (Table 2.2). Both L. minutus and Bosmina spp. were significantly larger in cool versus warm years (Table 2.3). No significant differences in body size were detected for any other crustacean zooplankton species; however, these were too rare to yield any statistical power for detecting effects.

Discussion

There was large inter-annual variation in the thermal structure of a small, shallow

Boreal Shield lake. However, years could be effectively separated as ‘warm’ or ‘cool’ based on differences in characteristics related to the phenology of lake warming, such as the timing of ice-out, the timing and strength of spring stratification, and the timing of fall turnover. Including measures of the magnitude of lake warming, such as mid-summer epilimnion temperature and late-summer bottom water temperature, captured aspects of the thermal habitat actually experienced by aquatic organisms.

In general, warm years had longer periods of isothermal conditions, higher bottom water temperature in late-summer, and later fall turnover compared to cool years. This exposed crustacean zooplankton to periods of higher water temperatures without a thermal refuge. Also, warm years typically had higher mid-summer surface water and epilimnion temperatures that occasionally reached 25°C, the upper lethal thermal limit for many temperate-region zooplankton species (Hanazato and Yasuno 1989, Moore et al.

1996, Folt et al. 1999, Chen and Folt 2002). However, while Swan Lake has undergone dramatic weather-related changes in thermal habitat between warm and cool years, the

30 current study revealed only differences in crustacean zooplankton body size and relative abundance, but not abundance or measures of diversity, on an overall annual basis.

Elevated temperatures typically shift crustacean zooplankton communities towards dominance by small-bodied species and individuals (Moore and Folt 1993,

Atkinson 1994, Moore et al. 1996). The mechanisms for a reduction in body size at elevated temperatures are poorly understood, but a net energy deficit may occur as warmer temperatures increase the metabolic demands of invertebrates (e.g. Buns and

Ratte 1991) but reduce assimilation efficiency (Giguère 1981) resulting in less energy diverted to growth. This pattern is so well-described for zooplankton (Moore and Folt

1993, Moore et al. 1996, Gillooly and Dodson 2000) and ectotherms in general (reviewed in Atkinson 1994), that Moore et al. (1996) suggested mean zooplankton body size be used as a ‘sensitive index of exposure’ to elevated temperatures in aquatic systems. The results of the current study support this suggestion. Consistent with expectations, warm years produced smaller-bodied individuals of the two most dominant zooplankton species in Swan Lake, L. minutus and Bosmina spp.

Factors other than temperature that can affect body size are likely not responsible for the observed reductions of body size in warm years. Total phosphorus and total chlorophyll a did not differ between warm and cool years, although these parameters tend to affect community size structure instead of individual body size (Hall et al. 1976, Yan et al. 2008). Also, while calcium concentrations are declining across eastern North America

(Keller et al. 2001, Jeziorski et al. 2008), calcium concentrations were always > 1.5 mg

·L-1, a critical threshold even for Ca-rich daphniids very sensitive to calcium reductions

(Jeziorski and Yan 2006, Ashforth and Yan 2008) and calcium not differ between warm and cool years. While DOC was greater in warm years, the mean difference was only 0.3

31 mg ·L-1. In any case, this change in DOC should not affect predation vulnerability because the main predators in Swan Lake (i.e. Chaoborus) are not visual feeders. Finally, regional reductions in acidity and some chemical recovery in the Sudbury area (Jeffries et al. 2003) likely do not explain changes in body size because pH did not differ between warm and cool years and the Swan Lake zooplankton community was acid-tolerant.

Some implications of a shift to smaller zooplankton community size structure include reductions in reproductive success, increased invertebrate predation risk, and changes in nutrient recycling and grazing that favour smaller-celled phytoplankton

(reviewed in Moore and Folt 1993). In small, fishless lakes, a temperature-induced shift to a smaller community size class may increase zooplankton vulnerability to predation from gape-limited invertebrate predators such as Chaoborus that typically prefer the smallest available prey (Swift and Fedorenko 1975, Pastorok 1981, Riessen et al. 1984,

Riessen et al. 1988). Furthermore, this potential increase in zooplankton predation risk may act synergistically with possible increases in predator foraging rate at elevated temperatures (Atkinson 1994, Moore et al. 1996, Chapter 3) to greatly reduce zooplankton abundance with future climate warming. For example, in a field enclosure experiment, Chaoborus reduced crustacean zooplankton abundance when exposed to constant warm temperatures, likely due to increased foraging associated with greater metabolic demand (Chapter 3). However, the small reduction in body size of ~ 11% for

Bosmina spp. and only 3% for L. minutus in warm Swan Lake years is not likely to alter the predation risk of the already small-bodied (< 1.0 mm) Swan Lake zooplankton community. This is corroborated by a lack of effect on mean annual zooplankton abundance in warm years. Typical reductions in body size of single zooplankton populations associated with elevated temperatures (> 25°C) range from 7 – 26% (Moore

32 et al. 1996). For predation risk to increase in lakes small such as Swan Lake due to changes in body size, a greater shift in the size frequency distribution of the entire zooplankton community would need to occur with future warming.

In contrast to the general pattern of reductions in zooplankton body size, zooplankton abundance is typically not directly affected by elevated temperatures (Moore et al. 1996), although climate can synchronize the fluctuations in annual abundance of several species of zooplankton between lakes within regions of the Boreal Shield (Rusak et al. 2008). Although smaller-bodied individuals, in general, produce fewer eggs per clutch, this effect is apparently compensated by increases in development rate and the number of clutches produced annually at elevated temperatures (e.g. Allen 1976, Orcutt and Porter 1983, Moore et al. 1996, Gillooly 2000). The lack of detectable differences in crustacean zooplankton abundance in Swan Lake between warm and cool years is consistent with previous investigations of zooplankton abundance between warm and cool conditions in eutrophic European systems (e.g. McKee et al. 2002).

It is possible that the temperatures experienced by zooplankton during warm years in Swan Lake were not thermally stressful, and this may explain why no differences in abundance were detected. However, this is unlikely because surface water temperatures occasionally reached 25°C, which is a lethal tolerance threshold for some temperate zooplankton species (e.g. Moore et al. 1996, Chen and Folt 2002). Also, even though mid-summer heat waves are typically of short duration (1 – 2 weeks) this period is long enough to affect short-lived zooplankton communities (Moore and Folt 1993). Moreover,

25°C surface water temperatures have previously reduced zooplankton survival in enclosure studies conducted in Swan Lake (Hasek 2008). Therefore, given the above, and the warm-year reductions in body size, it is likely that Swan Lake zooplankton were

33 thermally stressed in warm years, but that this stress did not translate into detectable differences in total crustacean zooplankton abundance at the annual scale.

Climate-related changes in the seasonal abundance of crustacean zooplankton abundance may be buffered at the annual scale by shifts in community composition related to species compensation (Schindler 1995). The relative abundance of small cladocerans in Swan Lake increased, while the proportion of calanoid copepods correspondingly decreased during warm years. Cooler conditions may favour copepod populations. L. minutus is typically considered to be a cool-water, oligotrophic, boreal species with a northerly range (Torke 2001) and although it is tolerant to a broad range of temperatures within its geographic distribution (Carter et al. 1980), the abundance of L. minutus is strongly associated with a longer duration of ice cover and later timing of ice- out (Rusak et al. 2008). McKee et al. (2002) found that copepods were more abundant in control than in 3°C warming treatments during their two-year mesocosm experiment. In contrast, warm conditions favour rapid cladoceran population growth (Allan 1976, 1977;

McCauley and Murdoch 1987, Moore et al. 1996, Chen and Folt 2002). A higher thermal tolerance and rapid development during periods of warming may be advantageous in unstratified lakes lacking a coldwater refuge (Chen and Folt 1996, 2002) and likely explains the increased relative proportion of small cladocerans (mainly Bosmina spp.) in warm versus cool years in Swan Lake. This may have consequences for ecosystem function in warm conditions. Although Bosmina are small-bodied, clearance rate (i.e. the volume of water cleared of suspended material per unit time) is much more dependent on animal density than size, and Bosmina grazing activity can exert greater control on phytoplankton biovolume than do larger cladoceran species, particularly at low food concentrations such as in oligotrophic lakes (e.g. DeMott 1982). Unfortunately, few

34 studies have examined grazing or clearance rates for L. minutus (but see Chow-Fraser

1986) and so it is unclear how a shift from community dominance by L. minutus to

Bosmina spp. will affect overall zooplankton community grazing rates.

Despite slight increases in the relative proportion of small cladocerans in warm conditions, there were no significant differences in crustacean zooplankton species richness, evenness, or diversity between warm and cool years in Swan Lake. Similarly,

MacLennan (2007) did not detect a difference in crustacean zooplankton species richness or diversity between a warm and cool year in Killarney Park near Sudbury, Ontario.

McKee et al. (2002) also found no difference in cladoceran species richness between a control and 3°C warming treatment in a two-year shallow, eutrophic pond mesocosm experiment. In contrast, in a comparative study of zooplankton richness between a warm and cool year across several hundred lakes, Stemberger et al. (1996) found that the richness of small cladoceran species was greater in the cool year. However, they detected no differences in the richness of calanoid copepods or small cyclopoids which are both abundant in Swan Lake. It is therefore not surprising that measures of overall diversity

(diversity, richness, evenness) did not change between warm and cool years in Swan

Lake.

Perhaps warming conditions would not reduce species richness in small, shallow lakes if exposure history pre-conditions species to survive periods of time without a cool thermal refuge. For example, Folt et al. (1999) found that Daphnia isolated from a pond that typically experienced elevated (> 25°C) surface water temperatures were more tolerant of 30°C conditions than those collected from a lake with a coldwater refuge.

Also, Carvalho (1987) demonstrated that cladoceran responses to thermal stress may vary between populations with different environmental histories. While small Boreal Shield

35 lakes can support a diverse assemblage of littoral and pelagic zooplankton species (Keller and Conlon 1994, Malkin et al. 2006, Heneberry et al. 2009), they tend to lack hypolimnetic species and consequently support lower species richness than larger, deeper lakes (Patalas and Salki 1993, Keller and Conlon 1994, Post et al. 2000). Swan Lake does not contain any hypolimnetic or glacial relict species (e.g. Limnocalanus macrurus,

Senecella calanoide, Cyclops scutifer) so it may be that richness would not be reduced in warming conditions because the thermal tolerances of the species present allow them to survive periods of warming. For example, the thermal tolerance of Bosmina, a ubiquitous species common to many small Shield lakes (Keller and Conlon 1994), is among the highest of all Cladocera (Drenner et al. 1981). However, as previously mentioned, L. minutus likely has a cooler thermal optima. The importance of exposure history in mediating thermal stress effects remains to be tested for temperate zooplankton species, and the thermal tolerances of many temperate zone zooplankton species are unknown

(Moore et al. 1996).

Finally, given the low pH of Swan Lake, perhaps it is not surprising that diversity did not differ between warm and cool lake years. Although Swan Lake was anthropogenically acidified (Keller et al. 1992a, Arnott et al. 2001), paleolimnological investigations indicate historic pH was slightly acidic and ranged from 5.6 – 6.1 (Dixit et al. 1989, Tropea 2008). The low pH of Swan Lake may have reduced the establishment success of potential colonisers, particularly Daphnia species (Binks et al. 2005), in both warm and cool years. Also, the Swan Lake zooplankton community is dominated by a few acid-tolerant taxa (Marmorek and Korman 1993) and biological resistance to potential colonisers through local biotic factors such as competition for algal food resources or macroinvertebrate predation (Binks et al. 2005), may have also been

36 important in limiting changes in species richness, evenness, or diversity between warm and cool lake years.

In summary, climate change will greatly alter the thermal structure of Boreal

Shield lakes. Small lakes numerically dominate the valuable freshwater resources across the Boreal Shield Ecozone (Fig. 1.2)(Cox 1978, Downing et al. 2006) but are also particularly sensitive to climate-induced changes in thermal structure (Schindler et al.

1990, Moore et al. 1996, Schindler 1997, Arnott et al. 2001). The results of the current study demonstrate wide inter-annual variability in key thermal habitat characteristics in a typical small Boreal Shield lake. Despite these changes, many characteristics of community structure (abundance, diversity measures) were resistant to changes in climate and lake thermal habitat in a broad, annual-scale comparison. However, an increase in the relative abundance of small cladocerans, declines in the proportion of calanoid copepods, and reductions in mean annual body size of dominant crustacean zooplankton taxa occurred in warm years. These changes in body size and the relative abundance of small grazers may have important implications for the recycling of nutrients and transfer of energy in freshwater ecosystems with future climate change.

37 TABLE 2.1. Mean warm- and cool-year abundances (no. per m3) of crustacean zooplankton functional groups.

Functional Group Warm Mean Cool Mean t-ratio P-value (no. per m-3) (no. per m-3) Calanoid copepods 38 142.50 41 322.50 -0.46 0.67

Cyclopoid copepods 663.98 470.86 0.76 0.50

Large cladocerans 170.34 217.50 -0.09 0.94

Small cladocerans 5 726.41 3 763.30 1.49 0.23

Littoral Species 195.60 297.46 -1.11 0.35

38 TABLE 2.2. Mean warm- and cool-year relative abundances of crustacean zooplankton functional groups. Significant values are in bold (P<0.05).

Functional Group Warm Mean Cool Mean t-ratio P-value

Calanoid copepods 0.84 0.90 -2.29 0.04

Cyclopoid copepods 0.02 0.01 0.97 0.39

Large cladocerans 0.004 0.004 -0.29 0.40

Small cladocerans 0.13 0.08 2.22 0.04

Littoral Species 0.004 0.006 -1.78 0.08

39 TABLE 2.3. Mean warm- and cool-year body size (mm) of commonly occurring crustacean zooplankton species in Swan Lake. Significant values are in bold (P<0.05).

Taxa Warm Mean (mm) Cool Mean (mm) t-ratio P-value

Bosmina spp. 0.32 0.36 -4.38 < 0.0001

Calanoid copepodids 0.65 0.65 -0.50 0.63

Cyclopoid copepodids 0.56 0.62 -1.02 0.33

D. birgei 0.62 0.63 -0.88 0.40

L. minutus 0.90 0.93 -2.39 0.03

40 (a)

• Sudbury

(b)

3

2

1

FIGURE 2.1. (a) Map of Canada highlighting the Sudbury, Ontario region and (b) bathymetry of Swan Lake (46° 22' 81° 04'). Circled numbers represent sampling stations along a lake transect. Thick black line indicates where mesocosm experiments have been conducted (Chapter 3). Map courtesy of the Cooperative Freshwater Ecology Unit.

41

1.0

1994

0.5 2001 SPRING 1995 1998 1993 2000 ICE 0.0 1999 PCA2 = 24% VWT 1996 -0.5 1997 EPI

2002 BTM TURN -1.0 -0.5 0.0 0.5 1.0 1.5

PCA1 = 44%

FIGURE 2.2. PCA of selected meteorological and lake thermal variables from 1993 – 2002. Scaling is focussed on inter-sample (year) distances. Blue squares indicate cool years and orange circles indicate warm years. SPRING = mean monthly total degree days for May; ICE = mean air temperature for February, March, and April (°C) as a proxy for the timing of ice-out (Gao and Stefan 1999); VWT = mean annual volume-corrected whole-lake water temperature (°C); EPI = mean mid-summer (July) epilimnion temperature (°C); TURN = timing of fall turnover (Julian day); BTM = mean late- summer (August) bottom (8m) water temperature.

42

CHAPTER 3

THE EFFECTS OF LAKE THERMAL HABITAT AND SPATIALLY-DEPENDENT

MACROINVERTEBRATE PREDATION ON CRUSTACEAN

ZOOPLANKTON COMMUNITY STRUCTURE

43 Abstract

Climate change will impact lake thermal structure with implications for biological interactions. The effects of climate change on spatially-dependent predator-prey interactions are poorly understood. Small, shallow Boreal Shield lakes are particularly responsive to climate-induced changes in thermal structure. In these systems, macroinvertebrates are often the top predators regulating crustacean zooplankton community structure. Potential interactions between macroinvertebrate predation and lake thermal structure may alter the outcomes of predator-prey interactions if the spatial distributions of prey in relation to predators or predator foraging rates differ across various thermal gradients. I performed a factorial in situ enclosure experiment using predators that exhibit different spatially-dependent predation styles (surface-orienting notonectid Buenoa macrotibialis and vertically-migrating phantom midge Chaoborus punctipennis) and three different thermal habitats representing potential future climate change scenarios for small, shallow lakes. These include a typical ‘cool’ stratified habitat, a ‘warm’ isothermal habitat, and a ‘hot’ isothermal habitat. I expected predation from surface-orienting predators to be greatest in stratified conditions when algal food resources are concentrated near the surface in contact with predators. I also expected that zooplankton would adopt intermediate depth distributions when surface-orienting and vertically-migrating predators co-occurred. Surprisingly, surface predators had no effect on zooplankton abundance or body size, likely because zooplankton avoided surface predators regardless of thermal habitat structure. In contrast, Chaoborus had an unexpected strong predation impact in a warm isothermal habitat. I hypothesize that this effect is related to increased predator metabolism, foraging, and ingestion rates when migrating through a thermally homogenous warm water column without access to a cool

44 hypolimnion. Predators were apparently thermally stressed in the ‘hot’ thermal trial as neither Chaoborus nor notonectids affected zooplankton abundance or body size under these conditions. These results demonstrate that the effects of climate change may be indirect, and biologically-mediated effects may result in strong, unexpected consequences for predator-prey dynamics. Predation impact may increase with lake water temperature and reductions in stratification strength, but there may be a threshold temperature where predators become physiologically stressed and direct temperature effects regulate community structure. Therefore, the importance of macroinvertebrate predators in regulating zooplankton community structure in small shallow lakes will depend on the extent to which climate change alters lake thermal habitat.

45 Introduction

Global climate change will alter the structure and function of terrestrial and aquatic ecosystems (IPCC 2007). These changes will occur through both direct thermal effects on physiological processes affecting survival, growth and reproduction at the level of the individual organism or population, and through complex indirect effects acting on community-level processes. Ignoring ecological complexity will limit the ability to predict community-level responses to environmental change. In particular, the biologically-mediated indirect effects of climate are poorly understood (Walther et al.

2002, Schmitz et al. 2003) and often create unexpected responses at the community-level that cannot be predicted based solely on intra-specific tolerances to warming conditions.

For example, strong predation may dampen the effects of climate on density-dependent prey populations that track climate fluctuations in the absence of predators (Wilmers et al.

2007). Alternatively, climate may increase the temporal synchrony in predator recruitment, perhaps increasing the regional coherence in prey population fluxes (Rusak et al. 2008). Temperature can influence the outcome of predator-prey dynamics in terrestrial (Davies et al. 1998, Post et al. 1999) and aquatic (Beisner et al. 1996, Sandford

1999, Jiang and Morin 2004) ecosystems, or temporally uncouple trophic interactions because of differential, species-specific phenological responses (reviewed in Stenseth et al. 2002).

In temperate freshwater systems, climate change will impact trophic interactions not only in time (e.g. seasonally, inter-annually) but also in space (vertically and horizontally due to lake thermal stratification). The temporal and spatial environmental heterogeneity of lake ecosystems creates many niches allowing a diverse assemblage of species to coexist (Hutchinson 1961). Changes in lake thermal habitat may lead to

46 species-specific differences in temporal (e.g. Hampton 2005) or spatial (Helland et al.

2007) niche differentiation that may alter predatory and competitive interactions

(Schindler 2001). Although many processes in lakes are regulated by food web dynamics

(Brooks and Dodson 1965, Carpenter et al. 1987, Vanni and Layne 1997) the ability to forecast freshwater ecosystem responses to future climate change is limited by an understanding of how such changes in lake thermal habitat may alter biological interactions (Schindler 2001, Keller 2007).

The thermal stratification of Boreal Shield lakes will be altered by climate change

(Schindler et al. 1990, Fee et al. 1996, Snucins and Gunn 2000, Keller et al. 2005). Food- web interactions in lakes depend on both the seasonal and spatial overlap of potential prey with predator species (Williamson et al. 1989). Recent studies demonstrate that seasonal changes in lake thermal habitat can alter trophic interactions due to species-specific phenological responses (Gerten and Adrian 2002) that affect the temporal overlap of zooplankton with their algal food resources (Winder and Schindler 2004a,b) or with predators (Wagner and Benndorf 2007). Analogously, the strength of lake stratification may affect spatially-dependent predator-prey interactions such as those between crustacean zooplankton and surface-orienting or vertically-migrating macroinvertebrate predators. Crustacean zooplankton are of much ecological interest because they represent a critical link in the energy transfer between primary producers and higher trophic levels.

To date, few studies have investigated how climate-related changes in lake thermal habitat may alter the outcomes of spatially-dependent biotic interactions. These include modelling scenarios of fish thermal habitat availability (De Stasio et al. 1996, Jansen and

Hesslein 2004) and predictions based on the seasonality (Helland et al. 2007) of vertical distribution and habitat overlap of planktivorous fish and zooplankton.

47 Crustacean zooplankton can behaviourally avoid visually-sensing predators such as fish (Zaret and Suffern 1976, Stich and Lampert 1981) and surface-orienting macroinvertebrate predators (e.g. Arts et al. 1981, Herwig and Schindler 1996) by diel vertical migration (DVM), whereby plankton migrate downwards into deeper, darker waters during the day. In contrast, mechano-sensing Chaoborus (Neill 1990) and other invertebrate predators (Ohman et al. 1983) that exhibit a ‘normal’ DVM to avoid fish predation can induce a ‘reverse’ migration in prey, such that prey remain higher in the water column during the day and migrate downwards at night. However, there is an energetic cost associated with migration itself (Swift 1976, Giguère and Dill 1980) and a fitness trade-off in growth and development between predator avoidance behaviour and habitat choice for both metabolically optimal temperature and efficient food acquisition

(reviewed in Lampert 1989).

Potential interactions between lake thermal habitat and spatially-dependent predation may alter the cost of this trade-off by altering predation risk (sensu Williamson et al. 1989) created by the overlap of predators and prey. The spatial overlap of predators and prey may change if thermal niches differentially shift (e.g. De Stasio et al. 1996).

Phytoplankton composition and biovolume may also change with levels of thermal stratification (Sommer 1985, Findlay et al. 2001, Winder and Hunter 2008) that affect light, temperature, nutrient and mixing regimes. Changes in the vertical gradient of algal resource availability may affect zooplankton depth distribution, and, subsequently, predation risk.

Spatial differences in predator-prey dynamics due to lake thermal structure may be particularly important in small lakes (< 10 ha) where space, habitat heterogeneity, and thermal refugia are limited (Keller and Conlon 1994, Moore et al. 1996, Post et al. 2000).

48 While climate warming is expected to increase thermal stability and coldwater habitat in moderate-sized lakes (De Stasio et al. 1996, Keller et al. 2005, Tanentzap et al. 2008), in small, shallow lakes, increased mixed layer warming may, depending on transparency, result in weaker stratification and decreased bottom coldwater habitat (Fee et al. 1996,

Moore et al. 1996, Schindler 1997, Snucins and Gunn 2000), potentially leading to isothermal conditions (Schindler et al. 1990, Moore et al. 1996, Schindler 1997).

Many small lakes are naturally fishless due to their headwater position or tendency for winterkill (Malkin et al. 2006) or have become so due to fish extirpations associated with cultural acidification (Bendell and McNicol 1987a, Uutala and Smol

1996). In such systems, macroinvertebrates represent the top predators on zooplankton prey (Wissel et al. 2003b, Arnott et al. 2006).

The phantom midge Chaoborus is arguably the most abundant, widespread

(Carter et al. 1980), and voracious macroinvertebrate predator on zooplankton prey.

Chaoborus induces behavioural, morphological and life history responses in crustacean zooplankton (e.g. Havel and Dodson 1984, Dodson 1988, Lünning 1992) and can regulate crustacean zooplankton community structure and function (e.g. Elser et al. 1987, Yan et al. 1991, Arnott and Vanni 1993). Chaoborus punctipennis is a vertically migrating species that is abundant in shallow lakes (Wissel et al. 2003a). C. punctipennis remains in the sediments during the day and migrates to feed at the surface at night while avoiding

UVR (Persaud et al. 2003) and predation from visually-sensing fish (von Ende 1979,

Wissel et al. 2003a). Changes in lake thermal structure may alter the mean exposure temperature of vertically-migrating predators such as Chaoborus, with potential implications for its control on the zooplankton prey community.

49 In addition, the importance of surface-orienting predators in structuring the crustacean zooplankton community may change if small lakes weakly stratify or become warm isothermal. The backswimmer Buenoa macrotibialis (Notonectidae) is ubiquitous in eastern North America (Chordas et al. 2005). Like Chaoborus, notonectids can also induce morphological, behavioural, and life history predator defences in crustacean zooplankton (e.g. Dodson 1988, Dodson and Havel 1988, Lünning 1992). In contrast to

Chaoborus, which ingest prey whole and typically prefer the smallest available prey items (Swift and Fedorenko 1975, Pastorok 1981, Riessen et al. 1984, Riessen et al.

1988), notonectids alter the size structure of zooplankton communities by selecting for larger individuals (Cooper 1983, Murdoch et al. 1984, Hampton et al. 2000, Gilbert and

Hampton 2001, Shurin 2001). If a lake is well-mixed (isothermal), then zooplankton may lower their depth distribution to obtain algal food resources at greater depth while avoiding time spent near the surface in contact with predators such as dytiscids (Arts et al.

1981) or notonectids (Herwig and Schindler 1996, Nesbitt et al. 1996, Gilbert and

Hampton 2001).

I used in situ enclosures to address the following questions: (i) Are there differences in the vertical distribution of crustacean zooplankton associated with differences in lake thermal structure?; (ii) do crustacean zooplankton demonstrate predator avoidance behaviour dependent on lake thermal structure?; and (iii) does macroinvertebrate predation interact with lake thermal structure to regulate crustacean zooplankton community abundance and body size?

Specifically, I hypothesized that (i) notonectids will have a greater predation impact on crustacean zooplankton abundance and body size in thermally stratified conditions when zooplankton algal food resources are concentrated near the surface

50 compared to isothermal conditions when zooplankton may adjust their depth distribution to obtain food at depth and spend less time in contact with surface predators; and (ii) crustacean zooplankton will maintain an intermediate depth distribution in the presence of both Chaoborus and notonectids (Nesbitt et al. 1996). I expected these depth distribution patterns to reduce predation impact on crustacean zooplankton abundance and body size.

Methods

Experimental design

I deployed enclosures in Swan Lake near Sudbury, ON (see Arnott et al. 2001), within a 3 x 2 x 2 factorial experimental design that included three levels of lake thermal habitat (cool-stratified, warm-isothermal, hot-isothermal), two levels of a vertically- migrating predator treatment with Chaoboridae: Chaoborus punctipennis Say (present, absent), and two levels of a surface-orienting predator treatment with Notonectidae:

Buenoa macrotibialis Hungerford (present, absent). Predator treatments were replicated four times within each level of thermal habitat (Fig. 1). To alter the thermal habitat of the enclosures I took advantage of natural seasonal changes in the thermal structure of Swan

Lake (Chapter 2, Appendix III Fig. III.2) to incubate the enclosures by performing three separate thermal trials during summer 2007.All thermal trials were deployed for 15 d, a period long enough to detect macroinvertebrate predation effects (e.g. Hasek 2008) without thermal structure changing dramatically over the course of the experiment.

The first thermal trial (June 4 – 18) simulated thermal conditions of a typical

‘cool’ year in which a lake is stratified with a warm epilimnion and cooler water below a thermocline. The second thermal trial (August 1 – 15) simulated ‘hot’ conditions where a lake is isothermal at ~ 25°C. This represents a high temperature scenario for most Boreal

Shield lakes (W. Keller, Cooperative Freshwater Ecology Unit, Sudbury, Ontario,

51 unpublished data). Finally, the third thermal trial (August 23 – September 6) simulated conditions typical of a ‘warm’ year where a lake is isothermal at a temperature of ~ 21°C.

These latter two scenarios correspond to potential thermal conditions with future climate change for small, shallow, moderately clear lakes, where the whole lake is weakly stratified or is warm isothermal, with no thermocline or bottom coldwater habitat

(Schindler et al. 1990, Fee et al. 1996, Moore et al. 1996, Schindler 1997, Snucins and

Gunn 2000).

Enclosure deployment and stocking

Enclosures (1 m diameter, 6.5 m length, ~ 5100 L) were constructed from 5-mm clear cylindrical polyethylene bags [Filmtech Plastic, Brampton, Ontario, Canada] suspended from wooden frames floating at the lake surface on Styrofoam® blocks.

Enclosures were filled with filtered Swan Lake water using a Honda gas-powered 2-inch centrifugal water pump with a 50 µm mesh filter on the output hose to exclude zooplankton. They were covered with a 0.25 cm mesh screen to prevent invertebrate colonisation and emigration of notonectids. The enclosures were stocked with zooplankton collected from Swan Lake using a 36 cm diameter, 80 µm mesh conical plankton net. Stocking densities were aimed to be representative of the long-term mean zooplankton density for June - August from 1993 – 2002 (~ 14 calanoid copepod adults and copepodids · L-1). However, actual stocking densities were 14, 9, and 4 individuals ·

L-1 in the cool, warm, and hot thermal trials, respectively. Zooplankton were allowed to acclimate in the enclosures overnight prior to the start of each thermal trial.

I assumed that the thermal habitat treatments would not be confounded by the seasonal succession of species because the relative proportion of zooplankton species in

Swan Lake did not greatly vary seasonally or annually from 1993 – 2002 (Appendix I

52 Fig. I.1). However, during the August 23rd warm thermal trial the dominant zooplankton species in Swan Lake switched from Leptodiaptomus minutus to Diaphanosoma birgei and the relative proportion of calanoid copepods in Swan Lake was significantly reduced compared to the first two (cool and hot) thermal trials. Therefore, for the third experiment

(warm isothermal) zooplankton were collected from Clearwater Lake (~ 1 km from Swan

Lake) and filtered through a 400 µm mesh to exclude Holopedium glacialis and large D. birgei present in the samples. Swan Lake and Clearwater Lake have similar acidification histories, although Clearwater has experienced significant chemical recovery as of 2007 to pH 6.31 in comparison to Swan Lake with a 2007 pH of 5.45 (W. Keller, Cooperative

Freshwater Ecology Unit, Sudbury, Ontario, unpublished data). In a reciprocal transplant experiment, Derry and Arnott (2007) demonstrated that crustacean zooplankton colonists can survive in treatment water of similar chemistry to the source lake. Also, the dominant species in the Clearwater Lake zooplankton community (L. minutus, Bosmina spp., D. birgei) were acid-tolerant (Marmorek and Korman 1993) and a transfer to slightly more acidic conditions should not affect survival.

On June 3rd, 7th, and 11th a mixture of III and mostly IV instar C. punctipennis larvae was collected 45 min after dark from Swan Lake using a 30 cm diameter, 153 µm mesh conical net towed from 6 m to the surface. I stored the Chaoborus at 4°C

(conditions typically experienced by over-wintering III and IV instars; von Ende 1982) for the study duration to slow development and ensure consistent predator assemblages

(i.e. density and age-structure) for each thermal trial. Mites and other conspicuous aquatic macroinvertebrates were excluded from the inoculum. On Day 0 of each experiment,

Chaoborus were added live to treatment enclosures at the ambient June lake density of

2.75 individuals ·L-1. The notonectid predator B. macrotibialis was collected on Day 0

53 from the pelagic zone near the enclosure frames and from the shoreline using kitchen sieves and D-frame nets and stocked at a visually-determined ambient density of 16 per enclosure (~5 individuals·m-2).

Enclosure sampling

Zooplankton, chlorophyll a, water chemistry (dissolved organic carbon [DOC], pH, and conductivity), Secchi depth, light and temperature/dissolved oxygen (DO) profiles were sampled at mid-day on Day 1 and Day 15 of each two week thermal trial.

Light was measured at 0.5 m intervals using a LI-COR LI-1400 data logger and light meter (LI-COR Biosciences, Lincoln, Nebraska, USA). Temperature/DO profiles were measured at 0.5 m intervals using a YSI model 55 probe (Yellow Springs Institute,

Yellow Springs, Ohio, USA). During each two week experiment, HOBOware® Tidbit temperature loggers (www.onsetcomp.com/products/data-loggers) were installed in a control bag and temperature was measured every hour at 1 m intervals. Chemical parameters (pH, conductivity, dissolved organic carbon) were sampled with a 1.8 cm diameter, 5 m integrated tube sampler. In addition, Day 1 total chlorophyll a samples were collected with the tube sampler from 0 - 5 m. Total chlorophyll a was sampled on

Day 15 from 0 - 6 m at 1 m intervals using a Van Dorn sampler.

Day 1 (initial) zooplankton were sampled from each enclosure at 0 - 6 m with a

7.5 cm diameter, 80 µm mesh conical tow net. Day 1 zooplankton samples were collected to test if stocking densities were similar among and within thermal trials; however, predation was always assessed based on comparisons relative to control treatments on

Day 15. Zooplankton were sampled on Day 15 (final) with a 12 L, 63 µm mesh

Schindler-Patalas trap from 0 - 6 m at 1 m intervals starting near the surface and working towards the bottom of the enclosures. Zooplankton sample collection alternated with

54 discrete total chlorophyll a collection at each depth interval to minimize mixing of the water column upon sampling the next strata.

Chemical analyses

Conductivity and pH were measured using a PHM64 Research probe (Bach

Simpson Ltd., London, ON, Canada). For total chlorophyll a analysis, 500 mL of lake water was filtered through 47 mm diameter Whatman GF/C pore filters and frozen prior to a 24 hr extraction in methanol and fluorometric analysis using a TD 700 Fluorometer

(Turner Designs, Sunnyvale, California, USA). DOC samples were processed at the

Dorset Environmental Science Centre, Dorset, Ontario, Canada, following standard

Ontario Ministry of the Environment protocols (Ontario Ministry of the Environment

2009d).

Zooplankton enumeration

Zooplankton were preserved in 5.5% sugared buffered formalin (Haney and Hall

1974, Appendix II) and enumerated using a Leica MZ12.5 stereomicroscope (Leica

Microsystems, Richmond Hill, Ontario, Canada). A total of 100 - 150 individuals from at least two equal fractions were counted per initial (Day 1) sample; sub -samples were generated using a Folsom plankton splitter (after Girard and Reid 1990). Final (Day 15) samples were counted in their entirety. For all samples, adult zooplankton were identified to species and body length was measured using the semi-automated sample processing system ZEBRA2M (Pawson and Yan 1993, Allen et al. 1994). Juvenile copepods and nauplii were identified to order (Calanoid or Cyclopoid). Immatures, i.e. nauplii and juvenile Bosmina (< 0.20 mm), were not considered in initial enclosure stocking calculations because they were likely to develop over the 15 d duration of the experiment and also were not vulnerable to notonectid predation (Cooper 1983). While nauplii are

55 vulnerable to Chaoborus predation (Swift and Forward 1981), they are typically only consumed by first and second instar larvae (Fedorenko 1975a) and Moore (1988b) reports nauplii are rarely ingested by IV C. punctipennis. Subsequently, immatures were excluded from all analyses. Bosmina (Bosmina) freyi and Bosmina (Bosmina) liederi were pooled due to uncertainty in taxonomic identification (Taylor et al. 2002). Taxonomic keys used included Smith and Fernando (1978), Thorp and Covich (2001), and Witty

(2004).

Statistical analyses

I tested for the interactive effect of thermal habitat and predators on mean zooplankton abundance and mean body size with a 3 x 2 x 2 factorial analysis of co- variance (ANCOVA) with initial abundance and initial body size as covariates. I also performed a 3 x 2 x 2 factorial analysis of variance (ANOVA) on mean abundance- weighted zooplankton depth calculated as:

= ∑ (nd ·d) / ∑ d

-3 where nd is the number of individuals at a given depth (no. individuals ·m ) and d is depth (m). All ANOVAs were performed using the statistical package R (R Development

Core Team 2007). I used Levene’s test and visually examined histograms, boxplots, and residual plots to determine that ANOVA assumptions were met. Abundance data were log10-transformed to improve homogeneity of variance and normality. During the cool thermal trial one “both predator” replicate bag developed a hole and I excluded this replicate from the analyses, resulting in an unbalanced design. Since a preliminary one- way ANOVA indicated that initial stocking densities differed between thermal trials (F (2,

45) =37.74, P<0.001) I considered only predator or thermal x predator interaction effects to reflect treatment differences. Although the covariates used in the analysis of mean

56 abundance and body size could not be separated from thermal treatment effects, including covariates reduced the residual error and improved the explanatory power of the

ANOVAs.

Finally, I used a 3 x 2 x 2 repeated-measures analysis of variance (RM-ANOVA) with depth as the repeated measure to test for treatment effects on total crustacean zooplankton depth distribution using the statistical package JMP 7.0 (SAS Institute Inc.

2007). I report the univariate Greenhouse-Geiser adjusted F-ratio to avoid problems with satisfying the assumption of sphericity (Quinn and Keough 2002). I did not include a covariate in the RM-ANOVA because I did not have discrete depth information for initial zooplankton abundance.

Results

Enclosure conditions

Mean epilimnion temperature was 22.0 °C (+ 0.14°C) in the cool-stratified trial and the thermocline remained at 4 m and the enclosures became more strongly stratified over the duration of the experiment. Thermal conditions were constant throughout the water column at 21.2 °C (+ 0.41 °C) and 25.1 °C (+ 0.06 °C) in the warm and hot isothermal trials, respectively (Fig. 3.2). Temperatures in Swan Lake had rarely exceeded 25°C (W. Keller, Cooperative Freshwater Ecology Unit, Sudbury, Ontario,

Canada, unpublished data) and this represents an extremely high temperature for lakes within the Boreal Shield Ecozone.

The initial pH (ANOVA: F(2, 45)= 6.90, P=0.002) and conductivity (ANOVA:

F(2,45)=1330.8, P<0.001) of the enclosures were statistically different across thermal treatments, but these slight differences are not ecologically relevant for the acid-tolerant zooplankton community (Marmorek and Korman 1993) and are within the natural range

57 in pH (Max - Min 1993 – 2007: 5.84 - 5.42) and conductivity (Max - Min 1993 – 2007:

47.50 – 37.91) for Swan Lake (Appendix I Table I.3). Furthermore, paleolimnological chrysophyte-inferred pH indicates that Swan Lake was naturally acidic with a pre- industrial pH of 5.6 – 6.1 (Dixit et al. 1989). Initial DOC concentrations differed between

-1 - the thermal trials (ANOVA, F (2, 45) =8.13, P =0.001) with means of 3.0 mg·L , 2.9 mg·L

1, and 3.2 mg·L-1 in the cool, warm, and hot thermal trials, respectively, but the actual range in DOC was only 0.8 mg ·L-1. Mean initial total chlorophyll a concentrations were similar at 0.30 and 0.35 µg·L-1 in the warm and hot isothermal trials, respectively, but

-1 were significantly greater at 0.57 µg·L in the stratified cool thermal trial (ANOVA: F (2,

45) =107.49, P<0.001, Tukey’s hsd: P<0.05). Secchi depth did not vary between replicate enclosures within each thermal trial and was 2.8 m, 5.9 m, and 5.4 m in the cool, warm, and hot thermal trials, respectively. Chemical conditions at the end of each experiment are described in Appendix VI.

Enclosure communities were 90% dominated by calanoids (L. minutus) except in the warm treatment where calanoids comprised 45% and cladocerans (Bosmina spp.) comprised 50% of total crustacean zooplankton community composition. The relative abundance of calanoids did not differ among predator treatments within thermal trials (3 x

2 x 2 ANOVA; Thermal x Chaoborus: F (2, 2) =0.70, P=0.50; Thermal x Notonectid: F (2, 2)

=0.05, P=0.95). There were differences in initial densities of crustacean zooplankton between thermal trials (F (2, 45) =37.74, P<0.001).

Notonectid survival was 80 – 82% by the end of the cool and warm thermal trials, but only 70% at the end of the hot trial. While some Chaoborus emergence did occur during each trial, adults were captured in the mesh enclosure covers and there were no apparent differences in adult density between thermal trials based on visual inspection.

58 Chaoborus survival was not enumerated but live individuals were captured in some 6 m samples.

Treatment effects on mean abundance and body size

Chaoborus only reduced zooplankton abundance in warm isothermal conditions

(Table 3.1). Mean crustacean zooplankton abundance was 20.6 individuals ·L-1 (+ 3.2) and 12.5 individuals ·L-1 (+ 1.8) across all predator treatments in the cool-stratified and hot-isothermal trials, respectively (Fig. 3.3). However, in the warm isothermal habitat, the presence of Chaoborus reduced crustacean zooplankton abundance by 73% (Fig. 3.3), from a mean of 33.7 individuals ·L-1 (+ 3.4) in treatments where Chaoborus was absent to a mean of 9.0 individuals ·L-1 (+ 1.1) in the treatments where Chaoborus was present.

This Chaoborus effect did not change in the presence of notonectids (Table 3.1).

Notonectids did not significantly affect mean crustacean zooplankton abundance in any thermal habitat (Table 3.1, Fig. 3.3).

Neither Chaoborus nor notonectids, when present either alone or combined, caused a shift in the size distribution of Bosmina spp. or adult L. minutus in any thermal habitat conditions (Table 3.2). However, Chaoborus significantly reduced the mean body size of calanoid copepodids (Table 3.2), but only in the warm thermal habitat (Table 3.2).

Mean final body size for Bosmina spp. (0.28 mm + 0.004 SE), calanoid copepodids (0.55 mm + 0.011 SE) and adult L. minutus (0.80 mm + 0.004 SE) across all enclosures is comparable to the 22-year annual mean body size for Bosmina (0.32 mm + 0.002 SE) calanoid copepods (0.62 mm + 0.003 SE) and L. minutus (0.89 mm + 0.003 SE) in Swan

Lake (W. Keller, Cooperative Freshwater Ecology Unit, Sudbury, Ontario, Canada, unpublished data).

59 Treatment effects on zooplankton depth distribution

Mean abundance-weighted zooplankton depth was the same across all treatment combinations (Table 3.3, Fig. 3.4), but the range in depth distribution was greater in stratified conditions (Fig. 3.4). Notonectids had a marginal effect (Table 3.3) on shifting mean abundance-weighted zooplankton depth deeper in the water column. In general, zooplankton were evenly distributed throughout the water column but were less abundant near the surface 1 m (Fig. 3.5).

Thermal habitat significantly affected daytime zooplankton depth distribution

(P<0.01; Table 3.4) such that zooplankton abundance was more variable between depths in the cool-stratified thermal trial than in the warm- and hot- isothermal trials (Fig. 3.5).

There was no interaction between the Chaoborus predation effect in the warm isothermal treatment and daytime zooplankton depth distribution (P=0.83; Table 3.4, Fig. 3.5).

Notonectids had a marginal effect on lowering zooplankton depth distribution, but this effect was independent of thermal habitat structure (P=0.07; Tables 3.3 and 3.4).

Discussion

Climate-induced changes in lake thermal structure may impact predator-prey dynamics. To date, there are no published field data regarding the interaction between lake stratification and macroinvertebrate predation on crustacean zooplankton. In a controlled field experiment, the impact on crustacean zooplankton of a ubiquitous vertically migrating aquatic invertebrate predator depended on the thermal characteristics of the water column. Chaoborus punctipennis reduced crustacean zooplankton abundance and body size in a warm isothermal habitat, but had no impact in hot isothermal or cool stratified conditions. Contrary to expectations, crustacean zooplankton did not adjust their daytime depth distribution based on an interaction between lake thermal structure and

60 surface-orienting predators. Instead, zooplankton were located deeper in the water column in response to the surface-orienting notonectid predator Buenoa macrotibialis, and this occurred in both stratified and isothermal conditions. Consequently, notonectids had no significant effect on overall crustacean zooplankton abundance or body size despite the occurrence of high predation rates (up to 42 prey ·L-1 ·d-1) in 1 L aquaria laboratory feeding trials (Daly 2008).

Logistical constraints associated with artificially creating and maintaining an in situ thermal gradient amidst a mass of lake water have restricted experimental studies on the effects of temperature on invertebrate community dynamics to small-scale, short duration laboratory aquaria (e.g. Murdoch et al. 1984, Spitze 1985, Beisner et al. 1997), to narrow vertical columns in the laboratory (e.g. Bass and Sweet 1984) or in situ (e.g.

Boeing et al. 2004), to small pond enclosures too shallow to stratify or allow vertical migration (e.g. Moss et al. 2003, Strecker et al. 2004, Baulch et al. 2005), or to greenhouse canopy enclosures that only heat the surface of the water column (e.g.

Christensen et al. 2006), although Baulch et al. (2003) employed an electric heating mechanism in the near-shore region that may be adjusted for use in the pelagic zone.

In the current experiment, the use of seasonal changes in lake thermal structure to simulate various climate change scenarios under which to examine the outcomes of predator-prey interactions was a novel approach conducted at a realistic scale. While mesocosms have been criticized for their lack of realism (Carpenter 1996) I contend that my enclosures realistically simulated natural conditions for several reasons. Firstly, the enclosures were deep enough to allow vertical migration and to develop heterogeneity in thermal and resource gradients at a scale comparable to in-lake conditions. Secondly, my focus was on invertebrate communities that had enough space for natural movement and

61 interactions in the enclosures. Finally, the study duration was long enough to detect predation effects at natural predator-prey densities without artificially raising densities to increase encounter rate (discussed in Sutor et al. 2001).

Two possible explanations to account for the unexpected strong Chaoborus predation in warm isothermal conditions include potential increased frequency of predator-prey encounters in warm conditions, and increased predator metabolic demand resulting in increased consumption at constant warm temperatures.

Chaoborus are ambush predators. As such, predation efficiency is related to the frequency of encounter with prey items, which depends on both prey swimming speed and density (Swift and Fedorenko 1975, Gerritsen and Strickler 1977, Pastorok 1980

1981, Riessen et al. 1984). Swimming speeds of Daphnia (Pastorok 1980, Spitze 1985) and Diaptomus (includes Leptodiaptomus) (Pastorok 1980) have been shown to increase with temperature, ultimately resulting in higher Chaoborus attack frequency. However,

Swift and Fedorenko (1975) only observed a temperature effect on swimming speed for

Polyphemus, but not for Bosmina or Diaptomus, the dominant species in my experiment.

Chaoborus may have had higher encounter rates with zooplankton prey items during the warm thermal trial, but this seems unlikely given the work of Swift and Fedorenko

(1975).

Although initial zooplankton prey densities differed between the three thermal trials, it is unlikely that this accounts for the 73% reduction in zooplankton abundance by

Chaoborus in warm isothermal conditions. Prey density does affect the encounter frequency with ambush predators such as Chaoborus (Fedorenko 1975b, Pastorok 1980,

Vinyard and Menger 1980, Spitze 1985, Moore 1988b). In my experiment, Chaoborus were never saturated with prey in any thermal trial (measured saturation densities:

62 Fedorenko 1975b, IV C. americanus and IV C. trivittatus at > 20 copepods ·L-1; Vinyard and Menger, C. americanus at 90 (<1.0 mm size) prey ·L-1; Spitze 1985, C. americanus >

150 Daphnia (0.8 mm size) prey ·L-1; Daly 2008, C. punctipennis, > 150 (< 0.8 mm size) prey ·L-1) and so were never limited by handling time (Holling 1965). Also, prey densities

(4 – 14 prey ·L-1) were not so low that predators were encounter-limited. For example,

Fedorenko (1975b) measured a 50% consumption effect of C. americanus at 4 copepod prey ·L-1. Therefore, a Chaoborus predation effect would be detectable had it occurred in any of the thermal treatments. Also, to account for stochastic processes in the prey community over the duration of the experiment, predation effect was always measured relative to a control treatment and not to initial prey densities. Hence, I conclude that differences in initial prey density most likely did not affect the ability to detect a

Chaoborus predation effect.

Likewise, slight differences in prey community composition do not likely account for the strong Chaoborus predation effect in the warm isothermal trial. The warm isothermal trial contained a greater relative proportion of Bosmina spp. than the cool and hot trials, but Chaoborus forage opportunistically based on encounter frequency with prey when they are not prey-saturated and therefore not limited by prey handling time

(Fedorenko 1975b). This is supported by an analysis of prey selection using Ivlev’s

Electivity Index (Ivlev 1961) on the enclosure data (Appendix V) and on 24 hr laboratory feeding trials of C. punctipennis foraging on a mixed assemblage of Swan Lake crustacean zooplankton (Daly 2008). In each of these analyses Chaoborus displayed no prey selection and Chaoborus diet composition was proportional to the relative abundance of prey in the environment. A regression of PCA1-transformed zooplankton prey community composition in the Chaoborus diet versus PCA1 prey composition in the

63 environment indicated that Chaoborus were foraging opportunistically in the enclosures

(R2 =0.93, P < 0.001, Appendix V) and in 1 L aquaria (R2 = 0.99, P < 0.01, Daly 2008).

Therefore, the change in prey composition between thermal trials did not affect

Chaoborus predation rates.

The water temperatures that Chaoborus experience throughout their diel migration may influence their predation impact on zooplankton. Chaoborus respiration (Swift

1976, Sigmon et al. 1978) and metabolism (Kajak and Rybak 1979, Giguère 1980, Buns and Ratte 1991) increase with temperature. Therefore, attack rate and ingestion must concurrently increase to satisfy metabolic demands when Chaoborus larvae are exposed to elevated temperatures (e.g. Fedorenko 1975b, C. trivitattus; Kajak and Rybak 1979, C. flavicans; Croteau et al. 2002, C. punctipennis). Digestion rate, the limiting factor in handling time, also increases with temperature (Giguère 1981, Spitze 1985). Therefore, as the average temperature experienced by Chaoborus increases, the rates of processes associated with metabolism, foraging, and ingestion also increase.

Migration through stratified water results in a decreased mean exposure temperature for Chaoborus because of the time spent in the cool bottom water. In my experiment, the cool-stratified trial had a mean water column temperature over the duration of the 15 d experiment of 17.5°C compared to 21.3°C in the warm isothermal trial. Exposure to cool water may lower basal respiration and metabolism, reducing overall foraging and ingestion rate. In contrast, with exposure to warm, isothermal conditions, assimilation efficiency decreases, metabolic demand increases, and predators spend more time actively foraging, likely resulting in greater ingestion and predation impact. Sutor et al. (2001) recommend that the predation impact of migrating and non- migrating Chaoborus populations be considered separately due to differences in

64 consumption associated with a temporal and spatial prey refuge. My experiments further suggest that among migrating Chaoborus populations, those migrating through homogeneous, warm water temperatures may have a greater predation impact than those migrating through a thermal gradient with access to a cool water refuge that lowers overall metabolic demand.

The lack of a Chaoborus predation effect in hot isothermal conditions may be due to direct thermal stress on Chaoborus. Although I did not measure Chaoborus survival in the enclosures, there is evidence to suggest that it may be reduced under high temperatures. Fourth instar C. punctipennis (Sigmon et al. 1978) and C. flavicans

(Hanazato and Yasuno 1989) experience mortality at temperatures greater than 25°C.

Halat and Lehman (1996) developed a Chaoborus bioenergetics model predicting starvation of first instar C. crystallinus in < 1 day without food at 25°C, but the model predicted survival for several days without food at lower temperatures. Also, Daly (2008) found that IV C. punctipennis from Swan Lake had low survival and required longer acclimation periods when incubated at temperatures > 24°C than at lower temperatures

(18 – 22°C).

Similarly, B. macrotibialis collected from Swan Lake had lower survival in the laboratory at temperatures exceeding 22°C (Daly 2008) and were likely also thermally stressed in enclosures at 25°C. In my enclosures, notonectids experienced a 12% reduction in survival in the hot isothermal trial compared to in cooler conditions. Perhaps the small-bodied zooplankton prey were not sufficient to meet notonectid metabolic demands, which increase with temperature (Murdoch et al. 1984), particularly since prey were able to alter their depth distribution and reduce time spent in contact with predators.

65 There was only a slight shift in zooplankton depth distribution in the enclosures associated with avoidance of notonectids, and, in contrast to expectations, and although algal food resources (as measured by total chlorophyll a concentration) differed with depth in stratified conditions (RM - ANOVA: depth x thermal, P<0.001, Appendix VI

Table VI.1), this effect was independent of thermal habitat treatment. In isothermal conditions (warm and hot trials), zooplankton avoided the surface, especially in the presence of notonectids, but were otherwise evenly distributed throughout the water column. This is likely because temperature, food availability and, except in the notonectid treatments, predation risk, were similar across depths (Fig. 3.6). There was greater variability in zooplankton depth distribution in stratified conditions where food resources, as determined by total chlorophyll a concentrations, were more heterogeneously distributed (Fig. 3.6). Nevertheless, the weak avoidance behaviour may have been sufficient to offset notonectid predation since there was no detectable notonectid effect on zooplankton body size or abundance.

My experimental results are in agreement with previous studies demonstrating that crustacean zooplankton migrate to limit time spent in contact with surface-orienting aquatic insect predators. Gilbert and Hampton (2001) also observed downward migration in zooplankton (Tropcyclops extensus) in response to B. macrotibialis. In shallow pond enclosures, predatory, surface-orienting dytiscid larvae reduced the abundance and deepened the depth distribution of Daphnia (Arts et al. 1981). When Herwig and

Schindler (1996) removed all surface-orienting aquatic insect predators from a shallow treatment pond, mean Daphnia body size increased and the migration strategy was reversed to spend more time higher in the water column during the day in comparison to pre-manipulation patterns and to a reference system. Likewise, in my study, migration

66 away from surface predators resulted in reduced predation impact on crustacean zooplankton abundance and body size.

In my experiment, crustacean zooplankton avoided notonectids near the surface in all thermal treatments. Zooplankton did not appear to avoid Chaoborus at the bottom during the daytime. In a similar study, Nesbitt et al. (1996) reported, in contrast, that

Daphnia adjusted their depth distribution to minimize overlap with both Chaoborus americanus and Notonecta undulata when the predators co-occurred. However, in their study, Chaoborus americanus were present throughout much of the mid-water column during the day. In contrast, C. punctipennis remained at the bottom of my enclosures during the day and were never present in the water column, except in a few 6 m samples, which likely explains why zooplankton did not adopt an intermediate depth distribution to avoid both Chaoborus and notonectids.

Although Buenoa can consume zooplankton < 0.5 mm in size (Hampton et al.

2000, Gilbert and Hampton 2001, Hampton and Gilbert 2001) it may be that notonectids are not significant predators in the pelagic zone of lakes dominated by small-bodied zooplankton. With the exception of Gilbert and Hampton (2001), previous investigations of interactions between surface predators and crustacean zooplankton have focussed on a single species, Daphnia pulex (Arts et al. 1981, Herwig and Schindler 1996, Nesbitt et al.

1996), that is likely more vulnerable to predation by notonectids because of its large size

(Cooper 1983, Murdoch and Scott 1984) relative to the small-bodied zooplankton (< 1.0 mm) in my enclosures. Although Buenoa inhabit the open water (Gittelman and

Bergstrom 1977, Hampton et al. 2000, SAM personal observation), they are also present in the near shore area (Hampton et al. 2000, SAM personal observation) and may preferably forage on larger littoral cladocerans present in Swan Lake, such as Chydorus

67 spp. In small-bodied crustacean zooplankton communities, Chaoborus are likely the dominant predator over notonectids because their whole-prey ingestion predation style allows them to preferably forage on small-bodied prey items whereas the piercing and sucking feeding style of notonectids requires more prey handling time (from several minutes to hours, Giller and McNeil 1981), and zooplankton can migrate to avoid notonectids in the pelagic zone of shallow lakes. This idea is supported by the weak predator avoidance behaviour and associated lack of a notonectid predation effect, and the surprisingly strong Chaoborus effect, under warm isothermal conditions in my enclosures.

Invertebrate predation appears to be much more important in structuring zooplankton communities in small, shallow lakes than in larger systems (Roff et al. 1981,

Keller and Conlon 1994, Malkin et al. 2006). Furthermore, small, shallow lakes are particularly responsive to climate-induced changes in lake thermal structure and may become isothermal for parts of the ice-free season (Moore et al. 1996, Schindler 1997,

Chapter 2). Despite the sensitivity to changes in thermal regime and importance of invertebrate predation in small lakes, to date, no previous study has investigated the potential interaction between spatially-dependent invertebrate predation and lake thermal habitat in regulating crustacean zooplankton community structure in these systems. This research is important because small lakes (< 10 ha) are numerically the most abundant aquatic systems across the Boreal Shield (Fig. 1.2)(e.g. Cox 1978, Downing et al. 2006) and globally (Downing et al. 2006).

My results suggest that the presence of surface-orienting predators may induce a downwards shift in crustacean zooplankton depth distribution that is independent of lake thermal structure and the presence of Chaoborus, and this apparently reduces predation.

68 In contrast, the importance of vertically-migrating predators in reducing crustacean zooplankton abundance and body size may increase in warm isothermal habitats expected to occur in small shallow lakes with future climate change (Schindler et al. 1990, Moore et al. 1996, Schindler 1997, Chapter 2). However, there appears to be an upper threshold temperature where macroinvertebrate predators become thermally stressed and fail to significantly impact crustacean zooplankton community structure. Therefore, the future importance of macroinvertebrate predators in regulating crustacean zooplankton community structure will depend on the extent to which climate change alters the thermal habitat of small, shallow Boreal Shield lakes.

My results demonstrate possible macroinvertebrate predation effects on crustacean zooplankton under various climate change scenarios. Although this study was phenomenological in nature and specific mechanisms for the observed Chaoborus predation effect remain unclear, the results allow the construction of hypotheses concerning temperature effects on spatially-dependent predator-prey interactions, particularly under fluctuating thermal regimes. I hypothesize that homogenous warm water temperatures increase the mean exposure temperature for vertically-migrating predators such as Chaoborus resulting in increased foraging rate and impact on prey community abundance and size structure. Further investigation is required to elucidate the specific mechanisms regulating the Chaoborus predation effect and of spatially- dependent interactions between crustacean zooplankton and macroinvertebrate predators.

Lake water temperature affects the behaviour and physiology of individuals, population demographics, and community-level dynamics of aquatic organisms. Climate change will temporally and spatially alter the lake thermal habitat upon which complex biological interactions are superimposed. Predicting the outcomes of biological

69 interactions under various thermal regimes is a major challenge in ecology. My study demonstrates that the effects of climate warming may have strong unexpected consequences for predator-prey dynamics.

70 TABLE 3.1. The effects of thermal habitat and the presence of Chaoborus and notonectids on mean total crustacean zooplankton abundance (no. per m3) pooled over depth using a three-way ANCOVA with initial density as a covariate. Significant values are in bold (P<0.05).

Factor df SS F P-value

Thermal 2 0.05 0.9929 0.38

Chaoborus 1 0.27 10.5419 <0.01

Notonectid 1 2.00 x 10-5 0.0009 0.98

Thermal x Chaoborus 2 0.89 17.25 <0.0001

Thermal x Notonectid 2 5.83 x 10-4 0.01 0.99

Chaoborus x Notonectid 1 0.03 1.19 0.28

Thermal x Chaoborus x Notonectid 2 0.01 0.18 0.84

71 TABLE 3.2. The interactive effects of thermal habitat, Chaoborus, and notonectids on mean body size (mm) for Bosmina spp., calanoid copepodids, and adult Leptodiaptomus minutus using a three-way ANCOVA pooling over depth with initial body size (mm) as a covariate. Significant values are in bold (P<0.05).

SS Taxa Factor df F P-value (x 10-3) Thermal 2 10.61 20.95 <0.0001

Chaoborus 1 0.02 0.10 0.76

Notonectid 1 0.18 0.72 0.40

Thermal x Chaoborus 2 0.27 0.54 0.59 Bosmina spp. Thermal x Notonectid 2 0.99 1.96 0.16

Chaoborus x Notonectid 1 0.08 0.32 0.58

Thermal x Chaoborus x Notonectid 2 0.34 0.66 0.52

Thermal 2 179.46 113.3 <0.0001 7 Chaoborus 1 0.10 0.12 0.73

Notonectid 1 0.06 0.07 0.79

Calanoid Thermal x Chaoborus 2 5.88 3.72 0.03 copepodids Thermal x Notonectid 2 1.25 0.79 0.46

Chaoborus x Notonectid 1 2.97 3.75 0.06

Thermal x Chaoborus x Notonectid 2 1.83 1.16 0.33

Thermal 2 3.41 2.06 0.14

Chaoborus 1 0.15 0.19 0.67

Notonectid 1 0.45 0.54 0.47

L. minutus Thermal x Chaoborus 2 0.47 0.28 0.75

Thermal x Notonectid 2 1.01 0.61 0.55

Chaoborus x Notonectid 1 1.11 1.33 0.26

Thermal x Chaoborus x Notonectid 2 1.08 0.65 0.53

72 TABLE 3.3. The effects of thermal habitat and the presence of Chaoborus and notonectids on abundance-weighted mean total crustacean zooplankton depth (m) using a three-way ANOVA. Significant values are in bold (P<0.05).

Factor df SS F P-value

Thermal 2 0.68 1.57 0.22

Chaoborus 1 0.24 1.09 0.30

Notonectid 1 0.79 3.63 0.07

Thermal x Chaoborus 2 0.01 0.03 0.97

Thermal x Notonectid 2 0.49 1.12 0.34

Chaoborus x Notonectid 1 0.02 0.09 0.76

Thermal x Chaoborus x Notonectid 2 0.06 0.13 0.88

73 TABLE 3.4. Results of a 3 x 2 x 2 repeated measures ANOVA with depth as the repeated measure and thermal habitat and the presence of Chaoborus and notonectids as factors affecting total crustacean zooplankton community abundance (no. per m3). Univariate Greenhouse-Geiser (ε) F-values are reported. Significant values are in bold (P<0.05).

Num. Den. Effect Factor ε F P-value df df Between- Thermal 2 36 0.24 4.38 0.02

Subjects Chaoborus 1 36 0.24 8.47 <0.01

Notonectid 1 36 0.03 1.07 0.31

Thermal x Chaoborus 2 36 0.58 10.45 <0.001

Thermal x Notonectid 2 36 0.001 0.03 0.97

Chaoborus x Notonectid 1 36 0.11 3.82 0.06 Thermal x Chaoborus x 2 36 0.08 1.47 0.24 Notonectid Within- Depth 2.53 91.17 0.51 23.19 <0.001

Subjects Depth x Thermal 5.07 91.17 0.51 3.61 <0.01

Depth x Chaoborus 2.53 91.17 0.51 1.53 0.22

Depth x Notonectid 2.53 91.17 0.51 2.46 0.07 Depth x Thermal x 5.07 91.17 0.51 0.43 0.83 Chaoborus Depth x Thermal x 5.07 91.17 0.51 1.26 0.29 Notonectid Depth x Chaoborus x 2.53 91.17 0.51 0.50 0.65 Notonectid Depth x Thermal x 5.07 91.17 0.51 1.08 0.38 Chaoborus x Notonectid

74 1 m

1m 2m x 4 Temperature 3m reps loggers in 4m 6.5 m control bag 5m 6m

t = 15 d t = 15 d t = 15 d

“cool” scenario “warm” scenario “hot” scenario [June trial] [Late-August trial] [Early August trial]

FIGURE 3.1. 3 x 2 x 2 factorial experimental design: three levels of a thermal habitat treatment (cool, warm, and hot), two levels of a Chaoborus treatment (present, absent) and two levels of a notonectid treatment (present, absent) with four replicates each.

75

(a) 0 - - 26 June 4th – 18th 1 - - 24

2 - - 22

3 - Cool Stratified - 20

4 - - 18

5 - - 16 20 – 23°C 6 - - 14

(b) 0 - - 26 Aug. 23rd – Sept. 06th

1 - - 24

→

2 - - 22 )

°C ( 3 - Warm Isothermal - 20

Depth (m)

4 - - 18 erature p ←

5 - - 16 Tem 20 – 21°C 6 - - 14

(c) 0 - - 26 Aug. 1st – 15th

1 - - 24

2 - - 22

3 - Hot Isothermal - 20

4 - - 18

5 - - 16 25 – 26°C 6 - ' ' ' ' ' ' ' ' - 14 1 3 5 7 9 11 13 15

Day of Experiment

FIGURE 3.2. Lake temperature progression (°C) from 1 – 6 m over the 15 day duration of the (a) cool, (b) warm, and (c) hot thermal trials.

76 X 10 3 (a) 50-

40-

30- )

-3 20-

10-

3 X 10C tlNt tid Bth 50- (b) Chaoborus P = 0.002 40-

·m (no. individuals abundance 30-

B B 20-

10- A A

3 X 10 50- (c) Mean crustacean zooplankton 40-

30-

20-

10-

Control Notonectid Notonectid Chaoborus Notonectid

Chaoborus +

FIGURE 3.3. Boxplots of mean crustacean zooplankton abundance (no. individuals per m3) by predator treatment within the (a) cool, (b) warm, and (c) hot thermal trials. Line represents median, box represents upper and lower (25 % and 75 %) quartiles, bars represent variance.

77

0

(a) . 2 5 . 2.52

0 . 3

5 .

3.53

0 . 4

5 . 4.54

0 . 5 (b)

2.5

3.5

4.5

(c)

2.5 Mean abundance-weighted zooplankton depth (m)

3.5

4.5

Control

Notonectid Notonectid Chaoborus Notonectid Chaoborus +

FIGURE 3.4. Mean abundance-weighted depth by predator treatment for the (a) cool, (b) warm, and (c) hot thermal trials. Line represents median, box represents upper and lower (25 % and 75 %) quartiles, bars represent variance.

78

Control Notonectid Chaoborus Chaoborus + Notonectid

1 1 1-- 1 d 2 2-- 2 2 3 3 3-- 3 4 4 4-- 4 5 5 - 5 5- Warm Isothermal - Control Warm Isothermal - Notonectid 6 Warm Isothermal - Chaoborus 6 Warm Isothermal - Both Cool Stratifie - 6 6- 0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 600 1-- 1 1 1 2-- 2 2 2 - 3 Depth (m) 3- 3 3 - 4- 4 4 4

← - 5 5 5 5- Hot Isothermal - Control Hot Isothermal - Chaoborus Hot Isothermal - Both 79 Hot Isothermal -

- 6 6 6 Warm Isothermal 6- 0 20000 40000 60000 0 20000Notonectid 40000 60000 0 20000 40000 60000 0 20000 40000 600

1-- 1 1 1 2-- 2 2 2 3-- 3 3 3

- 4 4 4 4- - 5 5- 5 5

- 6 6 6 Hot Isothermal 6- ' ' ' ' ' ' ' 3' ' ' ' ' 0 2000 0 40000 60000 0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 60000

Mean crustacean zooplankton abundance (no. individuals ·m-3) → FIGURE 3.5. Mean crustacean zooplankton abundance (no. individuals ·m-3) by thermal and predator treatments and by depth. Bars represent + SE.

79

Mean total chlorophyll a concentration (µg·m-3) 0.0 0.2 0.4 0.6 0.8 (a) 1 2

3 4

5

6

(b) 1 2

3

4

Depth (m) 5

6 (c) 1

2 3

4 5

6

FIGURE 3.6. Mean total chlorophyll a concentration (µg·L-1) from 1 – 6 m on day 15 in the (a) cool, (b) warm, and (c) hot thermal trials. Bars represent + SE.

80

CHAPTER 4: GENERAL DISCUSSION AND FUTURE DIRECTIONS

Global mean annual air temperatures are expected to rise during the next century

(IPCC 2007). Climate change will affect all freshwater ecosystems and has implications for individuals, populations, and communities inhabiting Boreal Shield lakes. Small lakes

(< 10 ha) are numerically the most abundant lakes within the Boreal Shield Ecozone (Fig.

1.2)(Cox 1978, Downing et al. 2006) and therefore represent an important habitat for aquatic biota. Small, shallow Boreal Shield lakes are typically understudied and are particularly responsive to weather-induced changes in thermal structure (e.g. Gao and

Stefan 1999, Arnott et al. 2001, Chapter 2, Appendix III). Furthermore, biological interactions are likely more intense in small lakes and ponds than in larger systems (Roff et al. 1981, Keller and Conlon 1994) and knowledge of climate change effects on biological interactions is a limiting factor in predicting the responses of freshwater ecosystems to future climate change (Schindler 2001, Keller 2007). Therefore, it is critical to understand and predict the consequences of climate change for community dynamics in small lakes.

Monitoring data coupled with mechanistic experiments provide a good way of predicting the effects of future climate change on aquatic ecosystems. Using a combination of monitoring data (Chapter 2) and a field experiment (Chapter 3) I tested for differences in crustacean zooplankton community structure between warm and cool lake habitats with differing thermal structure. I also tested for interactions between macroinvertebrate predators and lake thermal structure (Chapter 3) because small lakes are often fishless and dominated by macroinvertebrate predators.

Using monitoring data from Swan Lake, I detected reductions in mean annual body size, a sensitive indicator of thermal stress (Moore and Folt 1993), of the two most

81

dominant taxa in warm lake years, but I found no difference in mean crustacean zooplankton community abundance between warm and cool years at the annual scale

(Chapter 2). However, at the scale of two week field enclosures, the predation effect of

Chaoborus, a voracious, ubiquitous macroinvertebrate predator, greatly reduced crustacean zooplankton abundance, and this effect was dependent on warm isothermal conditions (Chapter 3).

Differences in the abundance effect between monitoring (Chapter 2) and experimental data (Chapter 3) are likely due to differences in the temporal scale at which the analyses were conducted. The lake monitoring data covered broader seasonal effects than the short-duration enclosure experiment. Warm lake years were only isothermal for short periods during late summer, and predator effects on zooplankton abundance over short timescales are likely ‘averaged out’ by seasonal, stochastic processes and by species compensation over the duration of the ice-free season. Therefore, if the duration of the late-summer warm isothermal period increases as predicted with future climate change

(Schindler et al. 1990, Moore et al. 1996, Schindler 1997, Chapter 3, Appendix III Fig.

III.2, Fig. III.3), then the importance of macroinvertebrate predators such as Chaoborus in structuring the crustacean zooplankton community may become increasingly apparent at the annual scale.

Climate-related effects on zooplankton abundance are likely strongly seasonal and averaged out at the annual scale by complex interactions between biotic and abiotic factors. Yan et al. (2008) suggest that for mean annual zooplankton standing stock to be altered with future climate warming, ‘extreme’ circumstance must occur to destabilize community dynamics. They suggest that possible factors may include drastic changes in nutrient supply or impact from generalized predators with high consumption rates. For

82

example, if secondary zooplankton production does not exceed Chaoborus consumption, zooplankton abundance may be reduced (Neill and Peacock 1980, Neill 1981, Elser et al.

1987, Yan et al. 1991). If climate warming increases the duration of warm isothermal periods in late summer, as was the case for many ‘warm’ years in Swan Lake (see

Appendix III Fig. III.2 for data from 1993 – 2008), predators such as Chaoborus may be exposed to constant high temperature with no access to cool bottom waters. This may increase predator activity and predation impact on crustacean zooplankton (Chapter 3).

Although I did not detect differences in crustacean zooplankton abundance between warm and cool lake years, this does not preclude abundance effects with future climate change. It may be that more extreme, perhaps season-specific temperature differences are required to affect abundance (see Appendix III Table III.1, Fig. III.2, Fig.

III.3). For example, differential responses to temperature of phytoplankton and zooplankton can results in the temporal uncoupling of the peak in zooplankton abundance from the spring phytoplankton bloom (Winder and Schindler 2004a,b). Rusak et al.

(2008) predict a decline in L. minutus and Bosmina spp. abundance in years with earlier ice-out, a contemporary trend expected to continue with future climate warming

(Magnuson et al. 2000, Futter 2003) and observed in warm lake years (Chapter 2). L. minutus and Bosmina spp. are two of the most ubiquitous species in eastern North

America (Carter et al. 1980). If their abundance is reduced in depauperate communities such as in Swan Lake, total zooplankton community abundance may be reduced.

Clearly, small lakes show large inter-annual variation in thermal structure

(Chapter 2, Appendix III Fig. III.2 and Fig. III.3). However, zooplankton community structure in this small lake responded similarly to other larger systems, i.e. reductions in body size but not abundance in warm conditions. No change in diversity was detected

83

between warm and cool years. Climate effects on crustacean zooplankton species diversity are typically variable and may decrease (Stemberger et al. 1996, rotifers and small cladocereans; Strecker et al. 2004, cladocerans) or show no change (Stemberger et al. 1996, large cladocerans and copopods; McKee et al. 2002, MacLennan 2007 and

Chapter 2, all functional groups) with warming. The effect of climate change on zooplankton diversity may depend on the background diversity of the system (e.g. acid- tolerant, species-poor community due to low pH in Swan Lake) and represent a fundamental difference between oligotrophic, species-poor Boreal Shield lakes, and more diverse, productive systems.

Future Directions

My intent was to examine broad-scale climate effects on zooplankton community structure by pooling twice-monthly collected zooplankton samples at the annual scale.

However, this may have overlooked some potentially interesting seasonal effects. For example, zooplankton community structure in other systems was related to seasonal factors such as the timing of ice-out (Rusak et al. 2008) and the timing of the spring phytoplankton bloom (e.g. Winder and Schindler 2004a,b). Future work may include a more extensive analysis using the multivariate approach employed in the current study with an expanded dataset including more recent lake years in a larger analysis (Appendix

III Fig. III.3). These additional data would provide enough warm/cool year replicates to more rigorously examine the role of differences in seasonal changes in lake thermal habitat on zooplankton community structure. With more lake years available (1993 –

2008), a direct gradient multivariate approach (e.g. RDA) could also be used to determine which thermal drivers are particularly important in structuring the crustacean zooplankton

84

community because ordination techniques require a greater number of samples (i.e. lake years) than environmental predictor variables (Lepš and Šmilauer 2003).

It is interesting to note that two consecutive years, 1998 and 1999, were the two warmest years during the Swan Lake monitoring record (Appendix III). Zooplankton abundances were higher in 1998 than in any other year. In contrast, 1999, the second warmest year on record, had abundances an order of magnitude lower than in any other year. Similarly, Chen and Folt (1996) observed reductions in the abundance of D. catawba after a mid-summer heat wave in the previous year. They suggest that increases in fall temperatures, such as what occurred in warm Swan Lake years (Chapter 2,

Appendix III Table III.1,Fig. III.2, Fig. III.3), are likely to have important consequences for the over-wintering strategies and seasonal phenologies of zooplankton, and may even determine community structure due to species-specific ecological and life history characteristics (Chen and Folt 1996). This may be important because zooplankton resting stages and emergence of individuals from historically-deposited eggs support ecosystem resilience to stressors such as climate change. Further investigation of the potential carry- over effects of the previous years’ thermal regime on zooplankton community structure may provide greater insight into thermal drivers of community dynamics, particularly regarding the timing and strategy of over-wintering and spring emergence, and perhaps predator recruitment success.

Additionally, future investigation of calanoid grazing rates is necessary in order to predict the ecosystem-level consequences of increases in the relative proportion of small cladocerans with associated reductions in copepod grazers. In their review paper, Moore et al. (1996) suggest that climate warming may alter freshwater predatory interactions by changing the habitat overlap of predators and prey and/ or by altering predator foraging

85

rates, abundance, and population size structure. In a field experiment I found no interaction between the spatial overlap of surface-orienting predators and their zooplankton prey (i.e. daytime zooplankton depth distribution) with lake thermal structure. However, the foraging rate of a vertically-migrating predator was dependent on lake thermal structure (Chapter 3). Also, differences in predator recruitment (abundance) and the size (age) structure of the predator community may have implications for the temporal overlap of predators and prey. Effects on zooplankton community abundance and diversity due to increases in predator recruitment associated with warmer conditions have been demonstrated in biogeographical studies of fish communities in small, shallow lakes (Gylleström et al. 2005, Jackson et al. 2007) but never for macroinvertebrates.

However, Wissel et al. (2003a) suggest that the high abundance of C. punctipennis in shallow fishless lakes is due to enhanced growth rates in warmer waters, and the abundance of Chaoborus can track large-scale climate signals (Keller et al. 1992b).

One limitation of this study was that it was performed in a single, species-poor, low pH lake, where it may not be surprising that no changes in zooplankton diversity were detected. Ideally, the type of warm/cool year analysis conducted in the current study

(Chapter 2) could be extended to include other small lake systems. Unfortunately, long- term monitoring data are limited for small lakes, and those samples that do exist are often collected only once annually (e.g. Heneberry et al. 2009) although some paleolimnological data are available (e.g. Rühland et al. 2008).

Finally, few previous studies examine the effects of invertebrate predation at temperatures exceeding 25°C (Moore et al. 1996; but see Chen and Folt 1996, 2002; Daly

2008). It is difficult to predict climate effects on Boreal Shield zooplankton communities or to interpret experimental results, for example the potential thermal stress in the hot

86

(25ºC) isothermal field enclosures (Chapter 3), because the thermal tolerances of many temperate and northern zooplankton species, particularly copepods, are not well known.

More work exploring the thermal tolerances, effects on life history traits, and predator- prey dynamics of aquatic biota across a range of elevated temperatures that surpass upper lethal limits is required in order to better predict the effects of climate change on freshwater communities.

Conclusions

Complex biological interactions occur within the framework of lake thermal habitat. Climate change will alter lake thermal structure, potentially altering crustacean zooplankton community structure and biological interactions. Freshwater communities must be viewed in the context of changing background conditions and multiple anthropogenic stressors (Yan et al. 2008), for which climate change is an over-arching factor. Further research is necessary to understand and define how changes in weather actually translate into thermal habitat changes in small lakes, and to determine how biota will respond, both directly and indirectly, to such habitat changes. Predicting the consequences of future climate change on freshwater ecosystems requires information on how outcomes of predator prey interactions may differ under changing thermal regimes and is currently poorly understood (Schindler 2001, Keller 2007).

The research presented in this thesis advances the current understanding of potential climate change effects on freshwater communities by examining the importance of indirect, biologically-mediated temperature effects in regulating crustacean zooplankton community structure. My results support the idea that climate change will indirectly affect zooplankton abundance via altered biological interactions. This thesis also demonstrates that small Boreal Shield lakes show wide inter-annual variation in

87

thermal structure. Warm conditions may alter community composition by causing a shift towards smaller-bodied species and individuals. Furthermore, increased periods of warm, isothermal conditions may substantially increase predation risk for crustacean zooplankton communities. Taken together, these changes in predation risk and body size may have important implications for the function of freshwater ecosystems with future climate change.

88

SUMMARY

1. Warm years in a small lake had earlier ice-out and spring stratification, higher mid-

summer surface, epilimnion, and late-summer bottom water temperatures, and later

fall turnover than cool years.

2. Mean crustacean zooplankton body size was greater in cool versus warm years for

the dominant taxa in a small lake, indicating zooplankton were thermally stressed in

warm conditions.

3. There were no detectable differences in crustacean zooplankton community

abundance or measures of diversity between warm and cool lake years, but the

relative abundance of small cladocerans increased in warm years concurrent with

the decline in relative of abundance of calanoid copepods.

4. Crustacean zooplankton were not restricted to spending more time in contact with

surface predators in stratified than in isothermal habitats. Instead, zooplankton

lowered their daytime depth distribution to avoid notonectids, regardless of lake

thermal structure. This reduced notonectid predation to levels beyond detection.

5. Vertically-migrating Chaoborus had a strong predation effect on crustacean

zooplankton, but only in warm isothermal conditions. This may be related to

increased rates of metabolism, foraging, and ingestion associated with a higher

mean exposure temperature experienced during a diel migration through warm

isothermal conditions compared to at a lower mean exposure temperature.

6. The importance of macroinvertebrate predators in regulating crustacean

zooplankton community structure may increase with future climate warming, but

there appears to be a threshold temperature where predators become physiologically

stressed and direct thermal effects regulate the crustacean zooplankton community.

89

REFERENCES Allan, D.J. 1976. Life history patterns in zooplankton. American Naturalist 110: 165– 180. Allan, D.J. 1977. An analysis of seasonal dynamics of a mixed population of Daphnia, and the associated cladoceran community. Freshwater Biology 7: 505-512. Allen, G., Yan, N.D., and Geiling, W.T. 1994. ZEBRA-2 - Zooplankton enumeration and biomass routines for APIOS: a semi-automated sample processing system for zooplankton ecologists. Ontario Ministry of Environment and Energy, Dorset, Ontario, Canada. Arnott, S.E., and Vanni, M.J. 1993. Zooplankton assemblages in fishless bog lakes: influence of biotic and abiotic factors. Ecology 74: 2361–2380. Arnott, S.E., Yan, N.D., Keller, W., and Nichols, K. 2001. The influence of drought- induced acidification on the recovery of plankton from Swan Lake (Canada). Ecological Applications 11: 747–763. Arnott, S.E., Keller, B., Dillon, P.J., Yan, N., Paterson, M., and Findlay, D. 2003. Using temporal coherence to determine the response to climate change in Boreal Shield lakes. Environmental Monitoring and Assessment 88: 365-388. Arnott, S.E., Jackson, A.B., and Alarie, Y. 2006. Distribution and potential effects of water beetles in lakes recovering from acidification. Journal of the North American Benthological Society 25: 811–824. Arts, M.T., Maly, E.J., and Pasitschniak. M. 1981. The influence of Acilius (Dytiscidae) predation on Daphnia in a small pond. Limnology and Oceanography 26: 1172– 1175. Ashforth, D. and Yan, N.D. 2008. The interactive effects of Ca concentration and temperature on the survival and reproduction of Daphnia pulex at high and low food concentrations. Limnology and Oceanography 53: 420–432. Atkinson, D. 1994. Temperature and organism size: a biological law for ectotherms? Advances in Ecological Research 25: 1–58. Bass, D., and Sweet, M.H. 1984. Do Chaoborus larvae migrate in temporary pools? Hydrobiologia 108: 181–185. Baulch, H.M., Nord, T.W., Ackerman, A.Y., Dale, J.D., Hazewinkel. R.R.O., Schindler, D.W., and Vinebrooke, R.D. 2003. Climate warming experiments: Design of a mesocosm heating system. Limnology and Oceanography Methods 1: 10–15. Baulch, H.M., Schindler, D.W., Turner, M.A., Findlay, D.L., Paterson, M.J., and Vinebrooke, R.D. 2005. Effects of warming on benthic communities in a boreal lake: Implications of climate change. Limnology and Oceanography 50: 437–452. Beisner, B.E., McCauley, E., and Wrona, F.J. 1996. Temperature-mediated dynamics of planktonic food chains: the effect of an invertebrate carnivore. Freshwater Biology 35: 219–232. Beisner, B.E., McCauley, E., and Wrona, F.J. 1997. The influence of temperature and food chain length on plankton predator-prey dynamics. Canadian Journal of Fisheries and Aquatic Sciences 54: 586–595. Bendell, B.E. 1986. The effects of fish and pH on the distribution and abundance of backswimmers (Hemiptera: Notonectidae). Canadian Journal of Zoology 64: 2696– 2699. Bendell, B.E., and McNicol, D.K. 1987a. Fish predation, lake acidity and the composition of aquatic insect assemblages. Hydrobiologia 150: 193–202.

90

Bendell, B.E., and McNicol, D.K. 1987b. Estimation of nektonic insect populations. Freshwater Biology 18: 105–108 Binks, J.A., S.E. Arnott, and Sprules, W.G. 2005. Local factors and colonist dispersal influence crustacean zooplankton recovery from cultural acidification. Ecological Applications 15: 2025–2036. Boeing, W.J., Leech, D.M., Williamson, C.E., Cooke, S.E., and Torres, L. 2004. Damaging UV radiation and invertebrate predation: conflicting selective pressures for zooplankton vertical distribution in the water column of low DOC lakes. Oecologia 138: 603–312. Brönmark, C., and Hansson, L.-A. 2002. Environmental issues in lakes and ponds: current state and perspectives. Environmental Conservation 29: 290–306. Brooks, J.L., and Dodson, S.I. 1965. Predation, body size, and composition of plankton. Science 150: 28–35. Büns, M., and Ratte, H.T. 1991. The combined effects of temperature and food consumption on body weight, egg production and development time in Chaoborus crystallinus De Geer (Diptera: Chaoboridae). Some new evidence for the adaptive value of vertical migration. Oecologia 88: 470–476. Cahill, K.L., Gunn, J.M., and Futter, M.N. 2005. Modelling ice cover, timing of spring stratification, and end-of-season mixing depth in small Precambrian Shield lakes. Can. J. Fish. Aquat. Sci. 62: 2134–2142. Carpenter, S.R. 1996. Microcosm experiments have limited relevance for community and ecosystem ecology. Ecology 77: 677–680. Carpenter, S.R., Kitchell, J.F., Hodgson, J.R., Cochran, P.A., Elser, J.J., Elser, M.M., Lodge, D.M., Kretchmer, D., and He, X. 1987. Regulation of primary productivity by food web structure. Ecology 68: 1863–1876. Carter, J.C.H., Dadswell, M.J., Roff, J.C., and Sprules, W.G. 1980. Distribution and zoogeography of planktonic crustaceans and dipterans in glaciated eastern North America. Canadian Journal of Zoology 58: 1355–1387. Carvalho, G.R. 1987. The clonal ecology of Daphnia magna (Crustacea: Cladocera): II. Thermal differentiation among seasonal clones. Journal of Animal Ecology 56: 469–478. Carvalho, L., and Kirika, A. 2003. Changes in shallow lake functioning: response to climate change and nutrient reduction. Hydrobiologia 506/509: 789–796. Chen, C.Y. and Folt, C.L. 1996. Consequences of fall warming for zooplankton over- wintering success. Limnology and Oceanography 41: 1077–1086. Chen, C.Y. and Folt, C.L. 2002. Ecophysiological responses to warming events by two sympatric zooplankton species. Journal of Plankton Research 24: 579–589. Chordas III, S.W., Ferreira, R.N., and Stewart, R.L. Jr. 2005. Synopsis of the backswimmers (Hemiptera: Notonectidae) of New Hampshire. Northeastern Naturalist 12: 187–194. Chow-Frazer, P. 1986. An empirical model to predict in situ grazing rates of Diaptomus minutus Lilljeborg on small algal particles. Canadian Journal of Fisheries and Aquatic Sciences 43: 1065 –1070. Christensen, M. R., Graham, M.D., Vinebrooke, R.D., Findlay, D.L., Paterson, M.J., and Turner, M.A. 2006. Multiple anthropogenic stressor cause ecological surprises in boreal lakes. Global Change Biology 12: 2316–2322.

91

Cohen, G.M., and Shurin, J.B. 2003. Scale-dependence and mechanisms of dispersal in freshwater zooplankton. Oikos 103: 603–617. Cooper, S.D. 1983. Selective predation on cladocerans by common pond insects Canadian Journal of Zoology 61: 879–886. Cooper, S.D., Smith, D.W., and Bence, J.R. 1985. Prey selection by freshwater predators with different foraging strategies. Canadian Journal of Fisheries and Aquatic Sciences 42: 1720–1732. Covich, A.P., and Thorp, J.H. 2001. Introduction to the Subphylum Crustacea. pp. 777– 809. IN: Thorp, J.H. and Covich, A.P. (Eds.). Ecology and classification of North American freshwater invertebrates. Academic Press, San Diego, California, USA. Cox, E.T. 1978. Counts and measures of Ontario lakes. Ontario Ministry of Natural Resources Report, Toronto, Ontario, Canada. Croteau, M.-N., Hare, L., and Tessier, A. 2002. Influence of temperature on Cd accumulation by species of the biomonitor Chaoborus. Limnology and Oceanography 47: 505–514. Curtis, P.J., and Schindler, D.W. 1997. Hydrologic control of dissolved organic matter in low-order Precambrian Shield lakes. Biogeochemistry 36: 125–138. Cushing, D. H. 1974. Sea fisheries research. John Wiley and Sons, New York, New York, USA. Daly, C. 2008. The influence of temperature on the predation rate of macroinvertebrate predators, Chaoborus punctipennis and Buenoa macrotibialis, on a natural zooplankton community in Swan Lake, Sudbury, Ontario. BSc. Thesis, Queen’s University, Kingston, Ontario, Canada. Davies, A.J., Lawton, J.H., Shorrocks, B., and Jenkinson, L.S. 1998. Individualistic species responses invalidate simple physiological models of community dynamics under global environmental change. Journal of Animal Ecology 67: 600–312. Dawidowicz, P., and Loose, C.J. 1992. Metabolic costs during predator-induced diel vertical migration of Daphnia. Limnology and Oceanography 37: 1589–1595. DeMott, W.R. 1982. Feeding selectivities and relative ingestion rates of Daphnia and Bosmina. Limnology and Oceanography 27: 518–527. Derry, A.M., and Arnott, S.E. 2007. Adaptive reversals in acid tolerance in copepods from lakes recovering from historical stress. Ecological Applications 17: 1116– 1126. De Stasio, B.T., Hill, J.M., Leinhans, N.P., Nibbelink, N.P., and Magnuson, J.J. 1996. Potential effects of global climate change on small north temperate lakes: physics, fish and plankton. Limnology and Oceanography 41: 1136–1149. Diéguez, M.C, and Gilbert, J.J. 2003. Predation by Buenoa macrotibialis (Insecta, Hemiptera) on zooplankton: effect of light on selection and consumption of prey. Journal of Plankton Research 25: 759–769. Dillon, P.J., and Molot, L.A. 2005. Long-term trends in catchment export and lake retention of dissolved organic carbon, dissolved organic nitrogen, total iron, and total phosphorus: The Dorset, Ontario, study, 1978–1998. Journal of Geophysical Research 110, G01002, doi:10.11029/2004JG000003. Dixit, S.S., Dixit, A.S., and Smol, J.P. 1989. Lake acidification recovery can be monitored using Chrysophycean microfossils. Canadian Journal of Fisheries and Aquatic Sciences. 46: 1309–1312.

92

Dodson, S. I. 1988. The ecological role of chemical stimuli for the zooplankton: Predator- avoidance behavior in Daphnia. Limnology and Oceanography 33: 1431–1439. Dodson, S.I. and Havel, J.E. 1988. Indirect prey effects – some morphological and life- history responses to Daphnia pulex exposed to Notonecta undulata. Limnology and Oceanography 33: 1274–1285. Dodson, S.I. and Frey, D.G. 2001. Cladocera and other Branchiopoda. pp. 723–786 IN: Thorp, J.H., and Covich, A.P. (Eds.). Ecology and classification of North American freshwater invertebrates. Academic Press, San Diego, California, USA. Downing, J.A., Prairie, Y.T., Cole, J.J., Duarte, C.M., Tranvik, L.J., Striegl, R.G., McDowell, W.H., Kortelainen, P., Caraco, N.F., Melack, J.M., and Middelburg, J.J. 2006. The global abundance and size distribution of lakes, ponds, and impoundments. Limnology and Oceanography 51: 2388–2397. Drenner, R.W., Vinyard, G.L., O’Brien, W.J., Triplett, J.R., and Wagner, J. 1981. The zooplankton community of Lacygne Lake: a cooling pond in Kansas. Southwestern Association of Naturalists. 26: 243–249. Elser, M.M., von Ende, C.N., Sorrano, P., and Carpenter, S.R. 1987. Chaoborus populations: Response to food web manipulation and potential effects on zooplankton communities. Canadian Journal of Zoology 65: 2846–2852. Environment Canada. 2008. National Climate Data and Information Archive. Online. Available: http://www.msc- smc.ec.gc.ca/education/elnino/comparing/enso1950_2002_e.html Fedorenko, A.Y. 1975a. Instar and species-specific diets in two species of Chaoborus. Limnology and Oceanography 20: 238–249. Fedorenko, A.Y. 1975b. Feeding characteristics and predation impact of Chaoborus (Diptera, Chaoboridae) larvae in a small lake. Limnology and Oceanography 20: 250–258. Fee, E.J. 1979. A relation between lake morphometry and primary productivity and its use in interpreting whole-lake eutrophication experiments. Limnology and Oceanography 24: 401–416. Fee, E.J., and Hecky, R.E. 1992. Introduction to the northwest Ontario lake size series (NOLSS). Canadian Journal of Fisheries and Aquatic Sciences 49: 2434–44. 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. Findlay D. L., Kasian, S.E.M., Stainton, M.P, Beaty, K., and Lyng, M. 2001. Climatic influences on algal populations of boreal forest lakes in the Experimental Lakes Area. Limnology and Oceanography 46: 1784–1793. Folt, C.L., Chen, C.Y., Moore, M.V., and Burnaford, J. 1999. Synergism and antagonism among multiple stressors. Limnology and Oceanography 44: 864–877. Futter, M. 2003. Patterns and trends in Southern Ontario lake ice phenology. Environmental Monitoring and Assessment 88: 431–444. Gao, S. and Stefan, H.G. 1999. Multiple linear regression for lake ice and lake temperature characteristics. Journal of Cold Regions Engineering 13: 59–77. Garcia, E.A., and Mittelbach, G.G. 2008. Regional coexistence and local dominance in Chaoborus: species sorting along a predation gradient. Ecology 89: 1703–1713.

93

Gerritson and Strickler. 1977. Encounter probabilities and community structure in zooplankton: a mathematical model. Journal of the Fisheries Research Board of Canada. 34: 73–82. Gerten, D., and Adrian, R. 2000. Climate-driven changes in spring plankton dynamics and the sensitivity of shallow polymictic lakes to the North Atlantic Oscillation. Limnology and Oceanography 45: 1058–1066. Gerten, D., and Adrian, R. 2001. Differences in persistency of the North Atlantic Oscillation signal among lakes. Limnology and Oceanography 46: 448–455. Gerten, D. and Adrian, R. 2002. Species-specific changes in phenology and peak abundance in freshwater copepods in response to warm summers. Freshwater Biology 47: 2163–2173. Giguère, L.A. 1980. Metabolic expenditures in Chaoborus larvae. Limnology and Oceanography 25: 922–928. Giguère, L.A. 1981. Food assimilation efficiency as a function of temperature and meal size in larvae of Chaoborus trivittatus (Diptera: Chaoboridae). Journal of Animal Ecology 50: 103–109. Giguère, L.A. and L.M. Dill. 1980. Seasonal patterns of vertical migration: a model for Chaoborus trivittatus. pp. 112–128. IN: Kerfoot, W.C. (Ed.). Evolution and Ecology of Zooplankton Communities. University Press of New England, Hanover, New Hampshire, USA. Gilbert, J.J., and Hampton, S.E. 2001. Diel vertical migrations of zooplankton in a shallow, fishless pond: a possible avoidance-response cascade induced by notonectids. Freshwater Biology 46: 611–621. Giller, P.S., and McNeill, S. 1981. Predation strategies, resource partitioning and habitat selection in Notonecta (Hemiptera/Heteroptera). Journal of Animal Ecology 50: 798–808. Gillooly, J.F. 2000. Effect of body size and temperature on generation time in zooplankton. Journal of Plankton Research 22: 241–251. Gillooly, J.F. and Dodson, S.I. 2000. Latitudinal patterns in the size distribution and seasonal dynamics of New World, freshwater cladocerans. Limnology and Oceanography 45: 2–30. Girard, R.E., and Reid, R.A. 1990. Dorset Research Centre Study lakes: sampling methodology (1986–1989) and lake morphometry. Ontario Ministry of the Environment Technical Report, Dorset, Ontario, Canada. Gittelman, S.H. 1974. Locomotion and predatory strategy in backswimmers (Hemiptera: Notonectidae). American Midland Naturalist 92: 496–500. Gittelman and Bergstrom. 1977. Depth selection in two species of Buenoa. Annals of the Entomological Society of America 70: 469–476. Graham, M.D., Vinebrooke, R.D., Keller, B., Heneberry, J., Nicholls, K.H., and Findlay, D.L. 2007. Comparative responses of phytoplankton during chemical recovery in atmospherically and experimentally acidified lakes. Journal of Phycology 43: 908– 923. Gyllström, M., Hansson, L.-A., Jeppesen, E., Garcia-Criado, F., Gross, E., Irvine, K., Kairesalo, T., Kornijow, R., Miracle, M.R., Nykänen, M., Nõges, T., Romo, S., Stephen, D., Van Donk, E., and Moss, B. 2005. The role of climate in shaping zooplankton communities of shallow lakes. Limnology and Oceanography 50: 2008–2021.

94

Hairston Jr., N.G. 1996. Zooplankton egg banks as biotic reservoirs in changing environments. Limnology and Oceanography 41: 1087–1092. Hairston, Jr., N.G. Dillon, T.A., and De Stasio Jr., B.T. 1990. Field test for the cues of diapause in a freshwater copepod. Ecology 71: 2218–2223. Halat, K.M., and Lehman, J.T. 1996. Temperature-dependent energetics of Chaoborus populations: hypothesis for anomalous distributions in the great lakes of East Africa. Hydrobiologia 330: 31–36. Hall, D.J., Threlkeld, S.T., and Burns, C.W. 1976. The size-efficiency hypothesis and the size structure of zooplankton communities. Annual Review of Ecology and Systematics 7: 177 – 208. Hampton, S.E. 2004. Habitat overlap of enemies: temporal patterns and the role of spatial complexity. Oecologia 138: 475–484. Hampton, S.E. 2005. Increased niche differentiation between two Conochilus species over 33 years of climate change and food web alteration. Limnology and Oceanography 50: 421–426. Hampton, S.E., Gilbert, J.J., and Burns, C.E. 2000. Direct and indirect effects of juvenile Buenoa macrotibialis (Hempitera: Notonectidae) on the zooplankton of a shallow pond. Limnology and Oceanography 45: 1006–1012. Hampton, S.E., and Gilbert, J.J. 2001. Observations of insect predation on rotifers. Hydrobiologia 446/447: 115–121. Hanazato, T., and Yasuno, M. 1989. Effect of temperature in laboratory studies on growth of Chaoborus flavicans (Diptera, Chaoboridae). Archiv für Hydrobiologie 114: 497–504. Haney, J.F., and Hall, D.J. 1974. Sugar coated Daphnia: a preservation technique for Cladocerans. Limnology and Oceanography 18: 331–333. Hansen J., and Sato, M. 2004. Greenhouse gas growth rates. Proceedings of the National Academy of Sciences 101: 16109–16114. Hansen, J., Sato, M., Ruedy, R., Kharecha, P., Lacis, A., Miller, R., Nazarenko, L., Lo, K., Schmidt, G. A., Russell, G., Aleinov, I., Bauer, S., Baum, E., Cairns, B., Canuto, V., Chandler, M., Cheng, Y., Cohen, A., Del Genio, A., Faluvegi, G., Fleming, E., Friend, A., Hall, T., Jackman, C., Jonas, J., Kelley, M., Kiang, N. Y., Koch, D., Labow, G., Lerner, J., Menon, S., Novakov, T., Oinas, V., Perlwitz, Ja., Perlwitz, Ju., Rind, D., Romanou, A., Schmunk, R., Shindell, D., Stone, P., Sun, S., Streets, D., Tausnev, N., Thresher, D., Unger, N., Yao, M., and Zhang, S. 2007. Dangerous human-made interference with climate: a GISS modelE study. Atmospheric Chemistry and Physics 7: 2287–2312. Hasek, D. 2008. The role of biological resistance on zooplankton recovery from acidification stress. BSc. Thesis, Queen’s University, Kingston, Ontario, Canada. Havel, J.E., and Dodson, S.I. 1984. Chaoborus predation on typical and spined morphs of Daphnia pulex: behavioural observations. Limnology and Oceanography 29: 487– 494. Helland, I.P., Freyhof, J., Kasprzak, P., and Mehner, T. 2007. Temperature sensitivity of vertical distributions of zooplankton and planktivorous fish in a stratified lake. Oecologia 151: 322–330. Heneberry, J., Keller, W., McNicol, D., and Weeber, R. 2009. Characterization of crustacean zooplankton communities in 33 small lakes near Sudbury, Ontario. Canadian Wildlife Service Report, Environment Canada, Ontario Region.

95

Herwig, B.R., and Schindler, D.E. 1996 Effects of aquatic insect predators on zooplankton in fishless ponds. Hydrobiologia 324: 141–147. Holling, C.S. 1966. The functional response of invertebrate predators to prey density. Memoirs of the Entomological Society of Canada 48: 1–86. Holzapfel, A.M., and Vinebrooke, R.D. 2005. Environmental warming increases invasion potential of alpine lake communities by imported species. Global Change Biology 11: 2009–2015. Hondzo, M., and Stefan, H.G. 1991. Three case studies of lake temperature and stratification response to warmer climate. Water Resources Research 27: 1837– 1846. Hutchinson, G.E. 1961. The paradox of the plankton. American Naturalist 95: 137–145. IPCC. 2007. Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Pachauri, R.K and Reisinger, A. (Eds.)]. IPCC, Geneva, Switzerland. Ivlev, V. S. 1961. pg. 101 IN: Krebs, C. J. 1989. Ecological Methodology. Harper Collins Publishers, New York, USA. Jackson, L.J., Torben, L.L., Søndergaard, M., and Jeppesen, E. 2007. A comparison of shallow Danish and Canadian lakes and implications of climate change. Freshwater Biology 52: 1782–1792. Jansen, W., and Hesslein, R.H. 2004. Potential effects of climate warming on fish habitats in temperate zone lakes with special reference to Lake 239 of the experimental lakes area (ELA), north-western Ontario. Environmental Biology of Fishes 70: 1–22. 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 acidic deposition. Ambio 32: 176–182. Jeziorski, A., and Yan, N.D. 2006. Species identity and aqueous calcium concentrations as determinants of calcium concentrations of freshwater crustacean zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 63: 1007–1013. Jeziorski, A., Yan, N.D., Paterson, A.M., DeSellas, A.M., Turner, M.A., Jeffries, D.S., Keller, B., Weeber, R.C., McNicol, D.K., Palmer, M.E., McIver, K., Arseneau, K., Ginn, B.K., Cumming, B.F., and Smol, J.P. 2008. The widespread threat of calcium decline in fresh waters. Science 322: 1374–1377. Jiang, L., and Morin, P. 2004. Temperature-dependent interactions explain unexpected responses to environmental warming in communities of competitors. Journal of Animal Ecology 73: 569–576. Jost, L. 2006. Entropy and diversity. Oikos 113: 363–375. Kajak, Z., and Rybak, J. 1979. The feeding of Chaoborus flavicans Meigen (Diptera, Chaoboridae) and its predation on lake zooplankton. Archiv für Hydrobiologie 64: 361–378. Keller, W. 2007. Implications of climate warming in Boreal Shield lakes: a review and synthesis. Environmental Reviews 15: 99–112. Keller, W., Pitblado, J.R., and Carbone, J. 1992a. Chemical responses of acidic lakes in the Sudbury, Ontario, area to reduced smelter emissions, 1981 – 89. Canadian Journal of Fisheries and Aquatic Sciences Supplement 1: 25–32.

96

Keller, W., Yan, N.D., Howell, T., Molot, L.A., and Taylor, W.D. 1992b. Changes in zooplankton during the experimental neutralization and early reacidification of Bowland Lake near Sudbury, Ontario. Canadian Journal of Fisheries and Aquatic Sciences Supplement 1: 52–62. Keller, W., and Conlon, M. 1994. Crustacean zooplankton communities and lake morphometry in Precambrian Shield lakes. Canadian Journal of Fisheries and Aquatic Sciences 51: 2424–2434. Keller, W., Dixit, S.S., and Heneberry, J. 2001. Calcium declines in northeastern Ontario lakes. Canadian Journal of Fisheries and Aquatic Sciences 58: 2011–2020. Keller, W., Heneberry, J., and Leduc, J. 2005. Linkages between weather, dissolved organic carbon, and cold-water habitat in a Boreal Shield lake recovering from acidification. Canadian Journal of Fisheries and Aquatic Sciences 62: 341–347. Keller, W., Paterson, A.M., Somers, K.M., Dillon, P.J., Heneberry, J., and Ford, A. 2008. Relationships between dissolved organic carbon concentrations, weather, and acidification in small Boreal Shield lakes. Canadian Journal of Fisheries and Aquatic Sciences 65: 786–795. Kelly, C.A., Fee, E., Ramlal, P.S., Rudd, J.W.M., Hesslein, R.H., Anema, C., and Schindler, E.U. 2001. Natural variability of carbon dioxide and net epilimnetic production in the surface waters of Boreal lakes of different sizes. Limnology and Oceanography 46: 1054–1064. Laforsch, C., Ngwa, W., Grill, W. and Tollrian, R. 2004. An acoustic microscopy technique reveals hidden morphological defenses in Daphnia. Proceeding of the National Academy of Sciences USA 101: 15911–15914. Lampert, W. 1989. The adaptive significance of diel vertical migration of zooplankton. Functional Ecology 3: 21–27. Legendre, P., and Gallagher, E.D. 2001. Ecologically meaningful transformations for ordinations of species data. Oecologia 129: 271–280. Lennon, J.T., Smith, V.H., and Williams, K. 2001. Influence of temperature on exotic Daphnia lumholtzi and implications for invasion success. Journal of Plankton Research 23: 425–434. Lepš, J., and Šmilauer, P. 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge, UK. Loreau, M., Naeem, S., Inchausti, P., Bengtsson, J., Grime, J.P., Hector, A., Hooper, D.U., Huston, M.A., Raffaelli, D., Schmid, B., Tilman, D. and Wardle, D.A. 2001. Biodiversity and ecosystem functioning: current knowledge and future challenges. Science 294: 804–808. Lünning, J. 1992. Phenotypic plasticity of Daphnia pulex in the presence of invertebrate predators: morphological and life history responses. Oecologia 92: 383–390. MacLennan, M. 2007. Changes in thermal structure and zooplankton assemblages in response to elevated summer temperatures in clear and coloured lakes of Killarney Park, Ontario. BSc. Honours Thesis, Queen’s University, Kingston, Ontario, Canada. Magnuson, J.J., Webster, K.E. Assel, R.A., Bowser, C.J., Dillon, P.J., Eaton, J.G., Evans, H.E., Fee, D.J., Hall, R.I., Mortsch, L.R., Schindler, D.W., and Quinn, F.H. 1997. Potential effects of climate change on aquatic systems: Laurentian Great Lakes and Precambrian Shield Region. Journal of Hydrological Processes 11: 825–871.

97

Magnuson, J.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 Vuglinski, V.S. 2000. Historical trends in lake and river ice cover in the Northern Hemisphere. Science 289: 1743–1746. Malkin, S.Y., Johannssen, O.E., and Taylor, B. 2006. Small-bodied zooplankton communities yet strong top-down effects on phytoplankton in the absence of fish. Archiv für Hydrobiologie 165: 313–338. Marmorek, D.R., and Korman, J. 1993. The use of zooplankton in a biomonitoring program to detect lake acidification and recovery. Water, Air, and Soil Pollution 69: 223–241. MathWorks, The Inc.TM. MatLab Version 7.0. Natick, Massachusetts. McCauley, E., and Murdoch, W.W. 1987. Cyclic and stable populations: Plankton as paradigm. American Naturalist 129: 97–121. McKee, D., Atkinson, D., Collings, S., Eaton, J., Harvey, I., Heyes, I., Hatton, K., Wilson, D., and Moss, B. 2002. Macro-zooplankter responses to simulated climate warming in experimental freshwater microcosms. Freshwater Biology 47: 1557– 1570. McNicol, D.K, and Wayland, M. 1992. Distribution of waterfowl broods in Sudbury area lakes in relation to fish, macroinvertebrates, and water chemistry. Canadian Journal of Fisheries and Aquatic Sciences 49: 122–133. Merritt, R.W., and Cummins, K.W. 1996. An introduction to the aquatic insects of North America. Kendall Hunt Publishing Company, Dubuque, Iowa, USA, 862pp. Millenium Ecosystem Assessment 2005. Ecosystems and human well-being: summary for decision makers. World Resources Institute, Washington, DC. Mitchell, S.W., and Lampert, W. 2000. Temperature adaptation in a geographically widespread zooplankter, Daphnia magna. Journal of Evolutionary Biology 13: 371– 382. Moore, M.V. 1988a. Density-dependent predation of early instar Chaoborus feeding on multi-species prey assemblages. Limnology and Oceanography 33: 258–270. Moore, M.V. 1988b. Differential use of food resources by the instars of Chaoborus punctipennis. Freshwater Biology 19: 249–268. Moore, M.V., and Gilbert, J.J. 1987. Age-specific Chaoborus predation on rotifer prey. Freshwater Biology. 17: 223–236. Moore, M,. and Folt, C. 1993. Zooplankton body size and community structure: effects of thermal and toxicant stress. Trends in Ecology and Evolution 8: 178–183. Moore, M.V., Folt, C.L., and Stemberger, R.S. 1996. Consequences of elevated temperatures for zooplankton assemblages in temperate lakes. Archiv für Hydrobiologie 135: 289–319. Moss, B., McKee, D., Atkinson, D., Collings, S.E., Eaton, J.W., Gill, A.B., Harvey, I., Hatton, K., Heyes, T., and Wilson, D. 2003. How important is climate? Effects of warming, nutrient addition and fish on phytoplankton in shallow lake microcosms. Journal of Applied Ecology 40: 782–792. Murdoch, W.W., and Scott, M.A. 1984. Stability and extinction of laboratory populations of zooplankton preyed on by the backswimmer Notonecta. Ecology 65: 1231–1248. Murdoch, W.W., Scott, M.A., and Ebsworth, P. 1984. Effects of the general predator Notonecta (Hemiptera) upon a freshwater community. Journal of Animal Ecology 53: 791–808.

98

Neill, W.E. 1981. Impact of Chaoborus predation upon the structure and dynamics of a crustacean zooplankton community. Oecologia 48: 164–177. Neill, W.E. 1988. Community responses to experimental nutrient perturbations in oligotrophic lakes: the importance of bottlenecks in size-structured populations. pp. 236–254. IN: Ebenman, B., and Persson, L. (Eds.). Size-structured populations. Springer-Verlag, Berlin, Germany. Neill, W.E. 1990. Induced vertical migration in copepods as a defence against invertebrate predation. Nature 345: 524–526. Neill, W.E., and Peacock, A. 1980. Breaking the bottleneck: Interactions of invertebrate predators and nutrients in oligotrophic lakes. pp. 715–724 IN: Kerfoot, W.C. (Ed.). Evolution and ecology of zooplankton communities. University Press of New England, Hanover, New Hampshire, USA. Nesbitt, L.M., Riessen, H.P., and Ramcharan, C.W. 1996. Opposing predation pressures and induced vertical migration responses in Daphnia. Limnology and Oceanography 41: 1306–1311. Ohman, M.D., Frost, B.W., and Cohen, E.B. 1983. Reverse diel vertical migration: an escape from invertebrate predators. Science 220: 1404–1407. Ontario Ministry of the Environment, Laboratory Services Branch. 2009a. The determination of total Kjeldahl nitrogen in surface water and precipitation by colourimetry. Ontario Ministry of the Environment, Rexdale, Ontario, Canada. Ontario Ministry of the Environment, Laboratory Services Branch. 2009b. The determination of total phosphorus in water by colourimetry. Ontario Ministry of the Environment, Rexdale, Ontario, Canada. Ontario Ministry of the Environment, Laboratory Services Branch. 2009c. The determination of chlorophyll in river and lake samples by spectrophotometry. Ontario Ministry of the Environment, Rexdale, Ontario, Canada. Ontario Ministry of the Environment, Laboratory Services Branch. 2009d. The determination of molybdate reactive silicates and dissolved organic carbon in water and precipitation by colourimetry. Ontario Ministry of the Environment, Rexdale, Ontario, Canada. Ontario Ministry of the Environment, Laboratory Services Branch. 2009e. The determination of cations in Precambrian Shield waters by atomic absorption spectrophotometry (AAS). Ontario Ministry of the Environment, Rexdale, Ontario, Canada. Orcutt, J.D., and Porter, K.G. 1983. Diel vertical migration by zooplankton: Constant and fluctuating temperature effects on life history parameters of Daphnia. Limnology and Oceanography 28: 720–730. Pace, M.L., and Cole, J.J. 2002. Synchronous variation of dissolved organic carbon and color in lakes. Limnology and Oceanography 47: 333–342. Pastorek, R.A. 1980. The effects of predator hunger and food abundance on prey selection by Chaoborus larvae. Limnology and Oceanography 25: 910–921. Pastorok, R.A. 1981. Prey vulnerability and size selection by Chaoborus larvae. Ecology 62: 1311–1324. Patalas, K. 1990. Diversity of the zooplankton communities in Canadian lakes as a function of climate. Archiv für Hydrobiologie 24: 360–368. Patalas, K., and Salki, A. 1993. Spatial variation of crustacean zooplankton in lakes of different size. Canadian Journal of Fisheries and Aquatic Sciences 50: 2626–2640.

99

Paulsen, S.G., Larsen, D.P., Kaufman, P., Whittier, T.R., Baker, J.R., Peck, D.V., McGue, J., Hughes, R.M., McMullen, D., Stevens, D., Stoddard, J.L., Lazorchak, J.M., Kinney, W., Selle, A.R., and Hjort, R. 1991. EMAP-surface waters monitoring and research strategy fiscal year 1991. US Environmental Protection Agency, Corvallis, Oregon, EPA/600/3-91/022, 184pp. Pawson, T.W., and Yan, N.D. 1993. Zooplankton data and sample archive database design for the Dorset Research Centre. Ontario Ministry of the Environment Technical Report, Dorset, Ontario, Canada. Perez-Fuentetaja, A., Dillon, P.J., Yan, N.D., and McQueen, D.J. 1999. Significance of dissolved organic carbon in the prediction of thermocline depth in small Canadian Shield lakes. Aquatic Ecology 33: 127–133. Persaud, A.D., Arts, M.T., and Yan, N.D. 2003. Photoresponses of late instar Chaoborus punctipennis larvae to UVR. Aquatic Ecology 37: 257–265. Post, E., Peterson, R.O., Stenseth, N.C., and McLaren, B.E. 1999. Ecosystem consequences of wolf behavioural responses to climate. Nature 401: 905–907. Post, D.M., Pace, M.L., and Hairston, N.G. Jr. 2000. Ecosystem size determines food- chain length in lakes. Nature 405: 1047–1049. Quinn, G.P., and M.J. Keough. 2002. Experimental design and data analysis for biologists. Cambridge University Press, Cambridge, UK. R Development Core Team. 2007. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3- 900051-07-0, URL http://www.R-project.org. Rasmussen, J.B., Godbout, L., and Schallenberg, M. 1989. The humic content of lake water and its relationship to watershed and lake morphometry. Limnology and Oceanography 34: 1336–1343. Rice, L.A. 1954. Observations on the biology of ten Notonectoid species found in the Douglas Lake, Michigan region. American Midland Naturalist 51: 105–132. Riessen, H.P., O’Brien, W.J., and Loveless, B. 1984. An analysis of the components of Chaoborus predation on zooplankton and the calculation of relative prey vulnerabilities. Ecology 65: 514–522. Riessen, H.P., Sommerville, J.W., Chiappari, C., and Gustafson, D. 1988. Chaoborus predation, prey vulnerability, and their effect n zooplankton communities. Canadian Journal of Fisheries and Aquatic Sciences 45: 1912–1920. Riessen, H.P. 1999. Predator-induced life history shifts in Daphnia: a synthesis of studies using meta-analysis. Canadian Journal of Fisheries and Aquatic Sciences 56: 2487– 2494. Roff, J. C., Sprules, W. G., Carter, J. C. H., and Dadswell, M. J. 1981. The structure of crustacean zooplankton communities in glaciated eastern North America. Canadian Journal of Fisheries and Aquatic Sciences 38: 1428–1437. Rühland, K., Paterson, A.M., and Smol, J.P. 2008. Hemispheric-scale patterns of climate- related shifts in planktonic diatoms from North American and European lakes. Global Change Biology 14: 2740 – 2754. Rusak, J.A., Yan, N.D., Somers, K.M., and McQueen, D.J. 1999. The temporal coherence of zooplankton population abundances in neighbouring north-temperate lakes. The American Naturalist 153: 46–58.

100

Rusak, J.A., Yan, N.D., and Somers, K.M. 2008. Regional climatic drivers of synchronous zooplankton dynamics in north-temperate lakes. Canadian Journal of Fisheries and Aquatic Sciences 65: 878–889. Sala, O.E., Chapin III, F.S., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R., Huber- Sanwald, E., Huenneke, L.F., Jackson, R.B., Kinzig, A., Leemans, R., Lodge, D.M., Mooney, H.A., Oesterheld, M., Poff, N.L., Sykes, M.T., Walker, B.H.,Walker, M., and Wall, D.H. 2000. Global biodiversity scenarios for the year 2100. Science 287: 1770–1774. Sandford, E. 1999. Regulation of keystone predation by small changes in ocean temperature. Science 283: 2095–2097. SAS Institute Inc. 2007. JMP Version 7.0 for Windows. SAS Institute, Cary, NC, USA. Schindler, D.W. 1995. Linking species and communities to ecosystem management: A perspective from the Experimental Lakes experience, pp. 313–325. IN: Jones, C.G., and Lawton, J.H. (Eds.). Linking species and ecosystems. Chapman and Hall, New York, USA. Schindler, D.W. 1997. Widespread effects of climatic warming on freshwater ecosystems in North America. Hydrological Processes 11: 1043–1067. Schindler, D.W. 2001. The cumulative effects of climate warming and other human stresses on Canadian freshwaters in the new millennium. Canadian Journal of Fisheries and Aquatic Sciences 58: 18–29. Schindler, D.W., Beaty, K.G., Fee, E.J., Cruikshank, D.R., DeBruyn, E.R., Findlay, D.L., Linsey, G.A., Shearer, J.A., Stainton, M.P, and Turner, M.A. 1990. Effects of climate warming on lakes of the central boreal forest. Science 250: 967–970. Schindler, D.W., Curtics, P.J., Bayley, S.E., Parker, B.R., Beaty, K.G., and Stainton, M.P. 1997. Climate-induced changes in the dissolved organic carbon budgets of boreal lakes. Biogeochemistry 36: 9–28. Schmitz, O.J., Post, E., Burns, C.E., and Johnston, K.M. 2003. Ecosystem responses to global climate change: moving beyond color mapping. Bioscience 53: 1199–1205. Shurin, J.B. 2001. Interactive effects of predation and dispersal on zooplankton communities. Ecology 82: 3404–3416. Shuter, B.J., Schlesinger, D.A., and Zimmerman, A.P. 1983. Empirical predictors of annual surface water temperature cycles in North American lakes. Canadian Journal of Fisheries and Aquatic Sciences 40: 1838–1845. Sigmon, C.F., Tombes, A.S., and Tilly, L. 1978. Diel oxygen uptake in Chaoborus punctipennis (Diptera: Culicidae) 61: 69–73. Smith, K., and Fernando, C.H. 1978. A Guide to the Freshwater Calanoid and Cyclopoid Copepod Crustacea of Ontario. Department of Biology, University of Waterloo, Waterloo, Ontario, Canada. Smith, B. and Wilson, J.B. 1996. A consumer’s guide to evenness indices. OIKOS 76: 70–82. Snucins, E., and Gunn, J. 2000. Interannual variation in the thermal structure of clear and coloured lakes. Limnology and Oceanography 45: 1639–1646. Sommer, U. 1985. Seasonal succession of phytoplankton in Lake Constance. Bioscience 35: 351–357. Spitze, K. 1985. Functional response of an ambush predator: Chaoborus americanus predation on Daphnia pulex. Ecology 66: 938–949.

101

Stemberger, R.S., and Lazorchak, J.M. 1994. Zooplankton assemblage responses to disturbance gradients. Canadian Journal of Fisheries and Aquatic Sciences 51: 2435–2447. Stemberger, R.S., Herlihy, A.T., Kugler, D.L., and Paulsen, S.G. 1996. Climatic forcing on zooplankton richness in lakes of the northeastern United States. Limnology and Oceanography 41: 1093–1101. Stenseth, N.C., Mysterud, A., Ottersen, G., Hurrell, J.W., Chan, K.-S., and Lima, M. 2002. Ecological effects of climate fluctuations. Science 297: 1292–1296. Stenseth, N.C., Ottersen, G., Hurrel, J.W. Mysterud, A., Lima, M., Chan, K.-S., Yoccoz, N.G., and Adlansvik, B. 2003. Studying climate effects on ecology through the use of climate indices: the North Atlantic Oscillation, El Niño Southern Oscillation and beyond. Proceedings of the Royal Society of London 270: 2087–2096. Stich, H.-B., and Lampert, W. 1981. Predator evasion as an explanation of diurnal vertical migration by zooplankton. Nature 293: 396–398. Streams, F.A. 1994. Effect of prey size on attack components of the functional response by Notonecta undulata. Oecologia 98: 57–63. Strecker, A.L., Cobb, T.P., and Vinebrooke, R.D. 2004. Effects of experimental greenhouse warming on phytoplankton and zooplankton communities in fishless alpine ponds. Limnology and Oceanography 49: 1182–1190. Sutor, M., Ramcharan, C., and Downer, R.G. 2001. Predation effects of two densities of fourth-instar Chaoborus trivittatus on a freshwater zooplankton assemblage. Hydrobiologia 464: 121–131. Swift, M.C. 1976. Energetics of vertical migration in Chaoborus trivittatus larvae. Ecology 57: 900–914. Swift, M.C., and Fedorenko, A.Y. 1975. Some aspects of prey capture by Chaoborus larvae. Limnology and Oceanography 20: 418–425. Swift, M.C., and Forward, R.B. Jr. 1981. Chaoborus prey capture efficiency in the light and dark. Limnology and Oceanography 26: 461–466. Tanentzap, A.J., Yan, N.D., Keller, B., Girard, R., Heneberry, J., Gunn, J.M., Hamilton, D.P., and Taylor, P.A. 2008. Cooling lakes while the world warms: Effects of forest regrowth and increased dissolved organic matter on the thermal regime of a temperate, urban lake. Limnology and Oceanography 53: 404–410. Taylor, D.J., Ishikane, C.R., and Haney, R.A. 2002. The systematics of Holarctic bosminids and a revision that reconciles molecular and morphological evolution. Limnology and Oceanography 47: 1486–1495. ter Braak, C.J.F, and Šmilauer, P. 2002. CANOCO Version 4.5. Biometris – Plant Research International, Wageningen, The Netherlands. Thorp, J.H., and Covich, A.P. 2001. Ecology and classification of North American freshwater invertebrates, 2nd ed. Academic Press, San Diego, California, USA. Torke, B. 2001. The distribution of calanoid copepods in the plankton of Wisconsin Lakes. Hydrobiologia 453/454: 351–365. Tropea, A.E. 2008. Assessing biological recovery from acidification and metal contamination in urban lakes from Sudbury, Canada: a paleolimnological approach. MSc. Thesis, Queen’s University, Kingston, Ontario, Canada. 243pp. Urquizo, N., Bastedo, J., Brydges, T., and Shear, H. 2000. Ecological assessment of the Boreal Shield ecozone. Environment Canada Report. ISBN-662-28679-0. Available: http://www.ec.gc.ca/soer-ree/English/SOER/CRAengfin.cfm.

102

Uutala, A.J. and Smol, J.P. 1996. Paleolimnological reconstructions of long-term changes in fisheries status in Sudbury area lakes. Canadian Journal of Fisheries and Aquatic Sciences 53: 174–180. Vanni, M.J., and Layne, C. D. 1997. Nutrient recycling and herbivory as mechanisms in the top-down effect of fish on algae in lakes. Ecology 78: 21– 40. Vinyard, G.L., and Menger, R.A. 1980. Chaoborus americanus predation on various zooplankters; functional response and behavioural observations. Oecologia 45: 90– 93. von Ende, C.N. 1979. Fish predation, interspecific predation, and the distribution of two Chaoborus [americanus, Chaoborus punctipennis] species. Aquatic Ecology 60: 119–128. von Ende, C.N. 1982. Phenology of four Chaoborus species. Environmental Entomology 11: 9–16. Wagner, A., and Benndorf, J. 2007. Climate-driven warming during spring destabilizes a Daphnia population: a mechanistic food web approach. Oecologia 151: 351–364. Walther, G.-R., Post, E., Convey, P., Menzel, A., Parmensank, C., Beebee, T.J.C., Fromentin, J.-M., Hoegh-Guldber, O., and Bairlein, F. 2002. Ecological responses to recent climate change. Nature 416: 389–395. Wetzel, R.G. 2001. Limnology: Lake and River Ecosystems. Third Ed. Academic Press, San Diego, California, USA. Williamson, C.E., Stoeckel, M.E., and Schoeneck, L.J. 1989. Predation risk and the structure of freshwater zooplankton communities. Oecologia 79: 76–82. Williamson, C.W. 1993. Linking predation risk models with behavioural mechanisms: identifying population bottlenecks. Ecology 74: 320–331. Williamson, C.W., 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. Williamson, C.W. and Reid, J.W. 2001. Copepoda. pp. 787–822 IN: Thorp, J.H., and Covich, A.P. (Eds.). Ecology and classification of North American freshwater invertebrates. Academic Press, San Diego, California, USA. Willimason, C.E., Grad, G., De Lange, H.J., and Gilroy, S. 2002. Temperature-dependent ultraviolet responses in zooplankton: Implications of climate change. Limnology and Oceanography 47: 1844–1848. Wilmers, C.C., Post, E., and Hastings, A. 2007. The anatomy of predatory-prey dynamics in a changing climate. Journal of Animal Ecology 76: 1037–1044. Winder, M, and Schindler, D.E. 2004a. Climate change uncouples trophic interactions in an aquatic ecosystem. Ecology 85: 2100–2106. Winder, M, and Schindler, D.E. 2004b. Climatic effects on the phenology of lake processes. Global Change Biology 10: 1844–1856. Winder, M., and Hunter, D.A. 2008. Temporal organization of phytoplankton communities linked to physical forcing. Oecologia 156: 179–192. Wissel, B., and Ramcharan, C.W. 2003. Plasticity of vertical distribution of crustacean zooplankton in lakes with varying levels of water colour. Journal of Plankton Research 25: 1047–1057.

103

Wissel, B., Yan, N.D., and Ramcharan, C.W. 2003a. Predation and refugia: implications for Chaoborus abundance and species composition. Freshwater Biology 48: 1421– 1431. Wissel, B., Boeing, W.J., and Ramcharan, C.W. 2003b. Effects of water color on predation regimes and zooplankton assemblages and in freshwater lakes. Limnology and Oceanography 48: 1965–1976. Witty, L.M. 2004. Practical guide to identifying freshwater crustacean zooplankton, 2nd ed. Cooperative Freshwater Ecology Unit, Sudbury, Ontario, Canada. Yan, N.D., Keller, W., MacIsaac, H.J., and McEachern, L.J. 1991. Regulation of zooplankton community structure of an acidified lake by Chaoborus. Ecological Applications 1: 2–65. Yan, N.D., Keller, W., Scully, N.M., Lean, D.R.S., and Dillon, P.J. 1996. Increased UV- B penetration in a lake owing to drought-induced acidification. Nature 381: 141– 143. Yan, N.D., Leung, B., Keller, W., Arnott, S.E., Gunn, J.M., and Raddum, G. 2003. Developing conceptual ecological frameworks for the recovery of aquatic biota from acidification. Ambio 32: 165–169. Yan, N.D., Somers, K.M., Girard, R.E., Paterson, A.M., Keller, W., Ramcharan, C.W., Rusak, J.A., Ingram, R., Morgan, G.E., and Gunn, J.M. 2008. Long-term trends in zooplankton of Dorset, Ontario, lakes: The probably interactive effects of changes in pH, total phosphorus, dissolved organic carbon, and predators. Canadian Journal of Fisheries and Aquatic Sciences 65: 862–877. Zaret, T.M., and Suffern, J.M. 1976. Vertical migration of zooplankton as a predator avoidance mechanism. Limnology and Oceanography 21: 804–813.

104

APPENDIX I – SELECTED SWAN LAKE MONITORING DATA

TABLE. I.1. List of zooplankton species sampled in Swan Lake, 1993-2002. Taxonomic authorities in parentheses indicate genus name has changed since original description.

Species Taxonomic Authority Functional Acantholeberis curvirostris Müller Large cladoceran Acroperus harpae (Baird) Littoral Alona quadrangularis Müller Littoral Bosmina spp. b De Melo & Hebert Small cladoceran Chydorus sphaericus Müller/Frey Littoral Daphnia ambigua Scourfield Large cladoceran Daphnia mendotae Birge Large cladoceran Diacyclops bicuspidatus (Forbes) Cyclopoid Diaphanasoma birgei Kořínik Large cladoceran Eubosmina coregoni d (Baird) Small cladoceran Eucyclops agilis (Koch) Cyclopoid Holopedium glacialis Zaddach Large cladoceran Latona setifera (Müller) Littoral Leptodiaptomus ashlandi (Marsh) Calanoid Leptodiaptomus minutus (Lilljeborg) Calanoid Limnocalanus macrurus Sars Calanoid Macrothrix spp. Baird Large cladoceran Mesocyclops edax (Forbes) Cyclopoid Orthocyclops modestus (Herrick) Cyclopoid Polyphemus pediculus (Linnaeus) Predatory Sida crystallina (Müller) Littoral Simocephalus serrulatus Koch Large cladoceran Skistodiaptomus (Lilljeborg) Calanoid a Formerly Acanthocyclops vernalis complex b Bosmina spp. = Bosmina (Sinobosmina) spp. + Bosmina longirostris + Bosmina freyi + Bosmina liederi c Daphnia pulex complex = Daphnia pulex + Daphnia pulicaria + Daphnia catawba + Daphnia minnehaha d Formerly Bosmina (Eubosmina) coregoni

105

TABLE I.2. Crustacean zooplankton species presence/absence in Swan Lake during the 1993–2002 study period. x = present, * = single occurrence

Species 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Acantholeberis curvirostris * x * x Acroperus harpae x x x x x x x x x x Alona quadrangularis * * Bosmina spp. x x x x x x x x x x Chydorus sphaericus * x * * x * Daphnia ambigua * * Daphnia mendotae * x * x Diacyclops bicuspidatus thomasi * * * * Diaphanasoma birgei x x x x x x x x Eubosmina coregoni * * Eucyclops agilis * * Holopedium glacialis x x Latona setifera * * Leptodiaptomus ashlandi * * Leptodiaptomus minutus x x x x x x x x x x Limnocalanus macrurus * * Macrothrix spp. x * x * Mesocyclops edax * x * x Orthocyclops modestus * * x * * * x * Polyphemus pediculus x x x x Sida crystallina x x x x x x Simocephalus serrulatus * *

Skistodiaptomus oregonensis * *

106

TABLE I.3. Mean annual ice-free season chemical data for selected parameters for Swan Lake from 1993 - 2007.

COND DOC Ca SiO CHL a TKN TP Year pH 2 (µS·cm-1) (mg·L-1) (mg·L-1) (mg·L-1) (µg·L-1)* (mg·L-1) (µg·L-1)

1993 5.53 46.95 2.04 3.91 0.21 2.11 0.29 7.50 1994 5.42 44.53 2.24 3.62 0.10 2.20 0.32 8.22 1995 5.60 43.52 2.85 3.23 0.07 1.32 0.39 10.40 1996 5.64 40.80 2.80 3.14 0.15 6.40 0.33 8.00 1997 5.50 37.91 2.31 3.13 0.15 3.92 0.35 10.28 1998 5.45 45.19 2.31 3.81 0.10 6.17 0.31 10.00 1999 5.47 47.06 2.71 3.82 0.13 9.13 0.27 6.92 2000 5.49 46.92 3.03 3.83 0.12 3.63 0.30 9.83 2001 5.57 45.20 2.85 3.42 0.31 4.96 0.30 9.00 2002 5.43 43.05 2.56 3.33 0.26 2.88 0.27 7.38 2003 5.70 47.50 2.74 3.43 0.34 4.57 0.31 11.16 2004 5.54 42.15 2.59 3.15 0.18 2.49 0.22 5.96 2005 5.84 43.48 2.46 3.26 0.20 2.56 0.31 8.82 2006 5.54 45.35 3.02 3.25 0.24 4.47 0.28 7.05 2007 5.45 46.49 2.91 3.23 0.14 2.35 0.28 7.38

COND = specific conductivity; DOC = dissolved organic carbon; CHL a = total chlorophyll a; TKN = total Kjeldahl nitrogen; TP = total phosphorus * Total chlorophyll a was pooled over depth (0.5 m, 4.5 m, 7.5 m) on each sampling date Note: Mean annual pH was calculated by converting pH to hydrogen ion concentration, averaging, and then re-converting to pH by the relationship: pH = -LOG10[H+]

107

(a)

) 1.0 -3 0.8

0.6

0.4

ity (no. ·individuals m 0.2

0.0 May June July Aug

(b) 1.0

tal zooplankton commun 0.8

0.6

0.4

0.2

Relative proportion of to 0.0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

FIGURE I.1. Relative abundance of crustacean zooplankton species in Swan Lake at the (a) mean monthly scale, and (b) mean annual scale, both from 1993 – 2002. Legend is the same for both figures. ( ) Calanoid copepods (i.e. Leptodiaptomus minutes), ( ) Cyclopoid copepods, ( ) Diaphanosoma birgei, and ( ) Bosmina spp.

108

8 6 ) -1 · L 6 5.5

4 5 pH ) and Ca (mg -1

2 4.5

DOC (mg ·L 0 4 1981 1986 1991 1996 2001 2006

FIGURE I.2. Mean annual Ca (○, mg ·L-1), dissolved organic carbon (, mg ·L-1), and pH (▲) in Swan Lake, 1982 – 2005. Vertical arrow highlights the 1988 drought-induced re-acidification event. To avoid having acid recovery as a confounding factor, years prior to 1993 were omitted from the study period (grey-shaded section).

109

6 1200

1000 5 800

4 600

400 3 (mm) precipitation 200 Total annual rainfall and Mean annual precipitation (mm) Mean annual air temperature (°C) 2 0

1980 1985 1990 1995 2000 2005

FIGURE I.3. Mean annual air temperature (▲, °C), total annual rainfall (, mm), and total annual precipitation (○, mm) in the Sudbury-area, 1982 – 2005. Dashed horizontal line represents long-term mean annual air temperature (°C).

110

APPENDIX II– PREPARATION OF FORMALIN PRESERVATIVE

Procedure for making formalin – Sudbury method (Chapter 2): The following recipe is for 32.5% sugared buffered formalin. The sugar serves to keep the Cladocera from ballooning and dropping any eggs they may hold, which is important if the samples are to be used for production estimates. The formalin is buffered because it is normally acidic and can be tough on zooplankton preserved for long time periods, and it can also destroy fine-bodied protozoa and phytoplankton. The final preservative concentration for sample storage is 13.5% sugared buffered formalin. It is recommended to mix a 32.5% solution for field preservation because, in many cases, all the lake water is not drained from the plankton sample and this water will add to the dilution of the formalin. When collecting field samples, ensure that the entire samples has been collected and properly preserved by rinsing the zooplankton sampling bucket with

32.5% formalin.

1. Recipe for 32.5% sweetened buffered formalin stock (Yield = 4 L): a. Add 1300 mL of 37% formaldehyde to a 4 litre plastic jug (properly labelled) b. Dissolve 240 grams of sugar in the 37% formaldehyde (purchased from dealer) c. Dilute the solution with 2700 mL distilled water d. Strongly agitate the solution to ensure thorough mixing e. Buffer with NaHCO3 to a pH of 7.2

2. Recipe for 13.5% formalin used for sample storage (Yield = 500 mL):

Add together in a 500 mL squeeze bottle 208 mL of the sweetened buffered formalin (see recipe in step 1 above) and 292 mL distilled water. Shake to mix well.

111

Procedure for making formalin – Dorset method (Chapter 3): The sugared formalin solution uses the following ingredients resulting approximately 6 L of an 11% formalin solution. This solution is mixed 50:50 in the field with sample volume in order to achieve a 5.5% final storage solution.

1. Take 5 L distilled water to which 359.92 grams of white granular sugar are added.

2. Using an eye dropper add 43 drops of 0.5N NaOH to buffer.

3. Add 620 mL Formaldehyde (distributed from MOE Stores at 37% strength). Stir this solution and store it within a fume hood prior to dispensing it into field use containers.

4. Mix solution 50:50 in the field with sample volume. The product should now be a

5.5% formalin solution with a pH approximating 7.

112

APPENDIX III –DEVIATION FROM LONG-TERM MEAN AIR TEMPERATURE We performed a preliminary analysis to classify years as warm or cool using air temperature data. We followed the methods of Somers (2000) to classify years as warm or cool by expressing air temperature as the percentage deviation from the long-term mean of total degree days (TDD) at both the monthly and annual scales. Daily air temperature data were obtained from the Sudbury Airport located ~30 km from Swan

Lake. To calculate the percentage deviation from long-term mean air temperature we first shifted all values into positives by adding + 40°C to each mean daily temperature. We summed the daily mean values by month and year to obtain an estimate TDD on a monthly and annual basis. Next we calculated the long-term mean annual and monthly air temperature for the period from 1955 – 2008. To standardize the data we subtracted the long-term mean from the observed monthly (or yearly) value, divided by the long-term mean, and multiplied by 100% to obtain the percent deviation from the long-term mean value. Values were considered significant deviations at the monthly scale if they differed from the long-term mean by 5% or more. Deviations of 1% or greater were considered significant at the annual scale.

At the annual level, there were three cool years and seven warm years deviating from the long-term mean by greater than 1% during the study period from 1993 – 2008

(Fig. III.1). Warm years were greater anomalies showing a stronger deviation from the long-term mean than cool years (Fig. III.1). Two of the warmest years (1998, 1999) coincided with the strongest recorded El Ninõ event on record (Environment Canada

2008). The coolest two years occurred in 1996 and 1997. Years from 1993 – 1997 were

113

all cooler than the long-term mean due to high concentrations of volcanic ash causing atmospheric shading following the 1991 eruption of Mount Pinatubo, Philippines.

At the monthly scale, the summer months of June – August rarely exhibited deviations from the long-term mean (Table III.1). During the 16 year study period, extreme deviations from the 50 yr mean are most common in November through March, suggesting that winter months may be very important in determining various physical, chemical, and biological lake characteristics in a given year. Perhaps this is not surprising since the timing of ice-off is dependent on lake size but also mainly on mean air temperature from February – April (Gao and Stefan 1999).

In 1995, extreme deviations vary in direction from January - April resulting in no classification as either warm or cool at the annual scale. Also, some years with the greatest number of deviations in a single direction did not show a large deviation at the annual scale (i.e. 1994, 2000, 2003, 2004, 2008). For this reason, and because air temperature alone is not the best predictor of lake water temperature, we further examined classifying years using a combination of both meteorological and lake thermal data in a multivariate analysis (Chapter 2, Fig 2.2). Depth-time thermal isopleth diagrams were used to determine various lake thermal parameters from 1993 – 2008 (Fig. III.2).

REFERENCES, APPENDIX III Environment Canada. 2008. National Climate Data and Information Archive. Online. Available: http://www.msc- smc.ec.gc.ca/education/elnino/comparing/enso1950_2002_e.html Gao, S. and Stefan, H.G. 1999. Multiple linear regression for lake ice and lake temperature characteristics. Journal of Cold Regions Engineering 13: 59–77. Somers, K.M. 2000. Patterns in climate parameters: understanding the role of climate signals. In Proceedings of the 2000 OPG/NSERC workshop. Trent University, Peterborough, Ontario.

114

TABLE III.1. Monthly total-degree days (TDD) during the 1993 – 2008 study period; expressed as a percent deviation from the long- term mean air temperature (1955–2008). Values warmer than 5% are highlighted in orange and values cooler than 5% are highlighted in blue. Years highlighted in bold red represent El Ninõ events based on the Southern Oscillation Index (SOI).

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

1993 4.99 -11.68 1.35 -0.65 -0.79 -1.60 1.70 3.08 -4.55 -5.96 -6.22 2.33 1994 -32.06 -11.55 1.84 -1.57 -1.59 1.72 -0.69 -3.19 1.71 3.32 4.78 15.22 1995 14.79 -8.09 6.65 -6.19 -0.48 4.34 0.45 2.93 -3.24 2.61 -12.69 -9.42 1996 -4.74 0.57 -6.41 -9.83 -2.57 1.95 -1.92 0.96 3.66 -0.95 -7.82 7.98 1997 -10.22 -0.43 -5.17 -0.82 -6.62 4.23 -0.59 -2.79 -0.34 0.13 -2.47 6.07 1998 6.29 23.17 3.85 6.72 7.70 0.60 0.28 2.47 1.49 0.86 5.20 10.02 1999 0.23 4.93 3.04 2.36 -1.16 4.35 -2.30 115 14.58 5.62 6.53 6.21 9.55 2000 -2.12 17.03 14.57 0.96 1.33 -2.42 -2.75 -1.02 -2.75 2.24 0.81 -14.56 2001 6.95 -4.10 -0.30 4.20 3.95 2.15 -0.72 4.03 0.31 1.21 9.97 18.56 2002 20.08 7.06 -5.58 -0.52 -4.97 0.40 2.36 2.17 6.68 -5.71 -5.17 9.05 2003 -5.60 -11.77 -0.44 -6.29 -0.04 0.77 -0.17 2.07 2.53 -0.98 3.79 13.18 2004 -19.17 13.51 6.37 -1.22 -2.20 -2.65 -1.11 -3.14 5.65 2.86 -1.98 -5.01 2005 -4.83 10.39 -0.26 5.61 0.46 5.99 2.76 2.89 6.04 5.41 1.34 0.95 2006 22.67 1.00 8.08 6.08 4.49 1.89 2.40 -0.17 -1.02 -2.34 5.71 18.17 2007 10.78 -9.70 3.90 1.23 3.19 2.52 -0.93 1.39 3.52 6.49 -0.89 2.46 2008 17.12 3.27 -4.41 5.89 -2.93 1.60 -0.51 0.17 1.05 -0.97 0.66 -4.10

115

5 4 3 2

annual air temperature (%) 1 0

-1 2000 2002 1994 2004 1992 1996 2006 2008 -2 1998 -3 -4 -5

Deviation from long-term mean FIGURE III.1. Percentage deviation from long-term mean annual total degree days from 1993 – 2008. Values deviating by greater than -1% are considered cool years (blue); values deviating by greater than 1% are warm years (orange).

116

1993 1994

)

°C

) ( m

(

th p

erature De 1995 1996 p

Tem 117

Julian Day of Year FIGURE III.2a. Thermal isopleths from May through October for 1993 - 1996. Figures show the temperature from the lake surface to the bottom from May 1st to October 31st of each year. Interpolation is between actual thermal profile measurements and white space indicates where no data were available.

117

1997 1998

) °C (

)

m ( th erature p

p De 1999 2000 Tem 118

Julian Day of Year FIGURE III.2b. Thermal isopleths from May through October for 1997 - 2000. Figures show the temperature from the lake surface to the bottom from May 1st to October 31st of each year. Interpolation is between actual thermal profile measurements and white space indicates where no data were available.

118

2001 2002

) °C

(

)

m ( th erature p p

De

2003 2004 Tem 119

Julian Day of Year FIGURE III.2c. Thermal isopleths from May through October for 2001 - 2004. Figures show the temperature from the lake surface to the bottom from May 1st to October 31st of each year. Interpolation is between actual thermal profile measurements and white space indicates where no data were available.

119

(2005) (2006)

Depth (m) (2007) (2008)

120 Temperature (°C)

Julian Day of Year FIGURE III.2d. Thermal isopleths from May through October for 2005 - 2008. Figures show the temperature from the lake surface to the bottom from May 1st to October 31st of each year. Interpolation is between actual thermal profile measurements and white space indicates where no data were available.

120

1.5 BTM 1.0 2007 1997 2006

0.5 2004 VWT

ICE 2005 2002 EPI 0.0 SPRING 1999 2000 PCA2 = 21% TURN 1994 2003 1996 1998 1995 -0.5 2001 1993 2008

-1.0

-1.5 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 PCA1 = 40%

FIGURE III.3 PCA of selected meteorological and lake thermal variables from 1993 – 2008. Scaling is focussed on inter-sample (year) distances. Blue squares indicate cool years, black triangles indicate intermediate years, and orange circles indicate warm years. SPRING = mean monthly total degree days for May; ICE = mean air temperature for February, March, and April (°C) as a proxy for the timing of ice-out (Gao and Stefan 1999); VWT = mean annual volume-corrected whole-lake water temperature (°C); EPI = mean mid-summer (July) epilimnion temperature (°C); TURN = timing of fall turnover (Julian day); BTM = mean late-summer (August) bottom (8m) water temperature.

121

APPENDIX IV – BODY SIZE MEASUREMENT Crustacean zooplankton body size was measured using the semi-automated sample processing program ZEBRA2M (Allen et al. 1994). Organisms were identified using a Leica MZ 12.5 dissecting microscope (Leica Microsystems, Richmond Hill,

Ontario, Canada). The microscope field view was output to a video screen and measurements of total body length were made with Ultra-Cal Mark III 6″ digital calipers

(Fred V. Fowler Co. Inc., Newton, Massachusetts, USA). The magnification factor was specified and measurements per individual organism were recorded in ZEBRA2M.

Calipers were calibrated each day against a stage micrometer.

Total length measurements for Cladocerans were from the top of the head to the anterior base of the spine, not including the tail spine (Fig. IV.1a). Copepods were measured from the top of the head to the base of the caudal rami, not including the caudal setae (Fig. IV.1b).

(a) (b)

FIGURE IV.1. Measurement points for two common zooplankton groups; (a) example of Cladoceran total length measurement for Bosmina spp.; (b) example of Copepod total length for Leptodiaptomus minutus.

REFERENCES – APPENDIX IV Allen, G., Yan, N.D., and Geiling, W.T. 1994. ZEBRA-2 - Zooplankton enumeration and biomass routines for APIOS: a semi-automated sample processing system for zooplankton ecologists. Ontario Ministry of Environment and Energy, Dorset, Ontario, Canada.

122

APPENDIX V - PREY SELECTION ANALYSIS Prey Selection - Methods

I used Ivlev’s Electivity Index (Ivlev 1961) to test for diet selection in each thermal treatment for Chaoborus, notonectids, and both predators combined. Ivlev’s

Electivity measure (Ei) for species i is:

Ei = ri – ni , ri + ni were ri is the percentage of species i in the diet, and ni is the percentage of species i in the environment. This yields values between +1 and -1 indicating the selection of a particular prey item by a given predator. Values between 0 and +1 indicate preference and values between 0 and -1 indicate avoidance of particular prey items. We used the control replicates (n = 4) within each thermal trial to represent the prey environment, ni, and all possible combinations of control-predator differences (n = 16) within each thermal trial to estimate the diet, ri, for each thermal x predator treatment combination.

I performed a one-way ANOVA using JMP 7.0 (SAS Institute 2007) to determine if electivity values differed between thermal trials for each predator. Tukey’s post-hoc comparison was performed where significant differences were detected. Crustacean zooplankton species abundance data were Hellinger-transformed to down-weight the importance of rare species (Legendre and Gallagher 2001) and analyzed across all thermal treatments with the linear ordination technique principle components analysis

(PCA) using the program CANOCO 4.5 (ter Braak and Šmilauer 2002). To test if there was a relationship between zooplankton community composition in the diet versus that of the environment, PCA axis 1 scores of diet were linearly regressed against PCA axis 1 scores of the prey environment for each predator treatment.

123

Prey Selection - Results

Neither Chaoborus nor notonectids demonstrated strong positive or negative selection for any particular crustacean zooplankton prey group, as indicated by low Ei values (-0.5 < Ei < 0.5, Table V.1). In general, Chaoborus displayed limited selection for calanoids and cyclopoids and against cladocerans. Selection for cladocerans by

Chaoborus was significantly influenced by thermal habitat such it was greatest in isothermal habitats when compared to stratified conditions (P < 0.001,Tukey’s hsd, Table

V.1, Fig.V.1). Selection for cyclopoids was greatest at intermediate temperatures in the warm isothermal habitat (P =0.003, Tukey’s hsd, Table V.1, Fig. V.1). These results for selection by Chaoborus coincide with prey abundance in the environment (Fig. V.3).

Notonectids positively selected for all major crustacean zooplankton taxa, with the greatest selection occurring for cyclopoids in hot isothermal conditions (Tukey’s hsd,

Table V.1, Fig. V.1). Cyclopoids were mainly represented by the large-bodied

Mesocyclops edax that comprised < 5% of the zooplankton community. This selection for M. edax, the largest zooplankton prey item available in the community, may be due to high metabolic demands by notonectids in hot isothermal conditions. Notonectids were extremely active and had reduced survival during the hot isothermal trial in comparison to the cooler thermal trials.

Finally, there was no strong prey selection in the presence of both predators. All major zooplankton taxa were positively selected, and cyclopoids were particularly selected in warm isothermal conditions (P < 0.0001, Tukey’s hsd). The diet of both predators combined was similar to that of both predators individually and this relationship did not change with thermal habitat.

124

One-to-one plots and a regression of PCA axis 1 scores for the zooplankton prey community available in the environment (i.e. control bags) versus the PCA axis 1 scores for the diet of Chaoborus and notonectids are shown in Figure V.2. There was a strong positive linear relationship between available zooplankton prey items in the environment and in the diet of Chaoborus (P < 0.001, R2 =0.93) and notonectids (P < 0.001, R2 =

0.95).

Prey Selection - Conclusions

Chaoborus and notonectids demonstrated limited selection for any major crustacean zooplankton taxonomic group under any thermal habitat conditions. Electivity did not change in the presence of both predators, suggesting that neither predator switched prey items based on inter-specific competition. All three predator treatments had the same diet composition, which was proportional to environmental prey availability.

These results suggest that predators are feeding opportunistically based on encounter with prey that are most abundant in the environment. Daly (2008) performed a similar analysis involving 24 hr laboratory experiments of Chaoborus and notonectid predators feeding on known densities of zooplankton prey items across an 18–30 °C temperature range. Our results support the findings of Daly (2008), in which prey selection was minimal and was not a factor in overall predation

125

TABLE V.1. Ivlev’s electivity index (Ei) for the predation of C. punctipennis, B. macrotibialis (notonectids), and both predators’ on crustacean zooplankton prey items. Predation was determined from 15 d field enclosure experiments, using all combinations of control replicates (n=4) versus predator treatment replicates (n=16). *Significant where P < 0.05.

Predator Prey Group Mean Ei SE df MS F P Tukey HSD

A) C. punctipennis Calanoids 0.002 0.018 2 0.034 1.23 0.300 -

Cylcopoids 0.204 0.021 2 0.213 6.81 0.003* Cool A, Warm B, Hot A

Cladocerans -0.125 0.045 2 0.490 12.19 < 0.001* Cool A, Warm B, Hot B

B) B. macrotibialis Calanoids 0.013 0.034 2 0.076 2.91 0.065 - 126

Cylcopoids 0.003 0.037 2 0.203 6.43 0.004* Cool A, Warm A, Hot B

Cladocerans 0.029 0.019 2 0.084 2.45 0.098 -

C) Both predators Calanoids -0.049 0.021 2 0.043 1.54 0.255 -

Cylcopoids 0.17 0.040 2 0.448 12.18 < 0.001* Cool A, Warm B, Hot A

Cladocerans 0.014 0.010 2 0.003 0.06 0.942 -

126

Chaoborus Notonectid

0.6 (b) (a) 0.6

0.2 0.2

Calanoids

-0.2 0.2

-0.6 0.6

(c) (d) 0.6 0.6 ) i A AB

2

0.2 0. oids p clo

y -0.2 0.2 A B A C

Mean Electivity (E

-0.6 0.6

Cool Warm Hot (e) (f) 0.6 0.6 A B B

0.2 0.2

-0.2 0.2 Cladocerans Cladocerans

-0.6 0.6

Cool Warm Hot Cool Warm Hot

FIGURE V.1. Boxplots of Ivlev’s prey electivity (Ei) for (a,b) Calanoids, (c,d) Cyclopoids, and (e,f) Cladocerans by Chaoborus and notonectid predators. Line represents mean, box represents upper (25 %) and lower (75 %) quartiles, bars represent variance. Electivity values where -0.5 > Ei > 0.5 are considered high negative/positive selection. Letters indicate where treatment differences were significant (P < 0.05, Tukey’s hsd).

127

2

1

diet PCA1 (96%) -2 -1 012

-1 Chaoborus y = 0.9624x + 0.0047 R2 = 0.9259

-2

2

1

-2 -1 0 12 Notonectid diet PCA1 (97%) -1 y = 0.9746x + 0.0054

R2 = 0.9497

-2

Prey environment PCA (96%)

FIGURE V.2. PCA1 of zooplankton community composition in the diet of (a) Chaoborus punctipennis and (b) Buenoa macrotibialis versus the prey environment. Black line represents regression, equation shown in figure inset.

128

APPENDIX VI – SUPPLEMENTARY EXPERIMENTAL DATA

TABLE. VI.1. Final (day 15) total crustacean zooplankton abundance data (no.per m3) for each replicate thermal x predator treatment x depth combination. Nau plii and juvenile Bosmina not included. Cool Bag P was excluded due to a suspected hole.

Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance

Cool Absent Present A 1 3166.67

Cool Absent Present A 2 11250.00

Cool Absent Present A 3 17833.33

Cool Absent Present A 4 21833.33

Cool Absent Present A 5 23416.67

Cool Absent Present A 6 21416.67

Cool Present Absent B 1 7416.67

Cool Present Absent B 2 14166.67

Cool Present Absent B 3 66000.00

Cool Present Absent B 4 32750.00

Cool Present Absent B 5 28416.67

Cool Present Absent B 6 39000.00

Cool Absent Absent C 1 9750.00

Cool Absent Absent C 2 8750.00

Cool Absent Absent C 3 10583.33

Cool Absent Absent C 4 11833.33

Cool Absent Absent C 5 25000.00

Cool Absent Absent C 6 21666.67

Cool Present Present D 1 5416.67

Cool Present Present D 2 7166.67

Cool Present Present D 3 20500.00

Cool Present Present D 4 21083.33

Cool Present Present D 5 29916.67

129

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Cool Present Present D 6 34583.33

Cool Present Absent E 1 4583.33

Cool Present Absent E 2 3833.33

Cool Present Absent E 3 9416.67

Cool Present Absent E 4 17166.67

Cool Present Absent E 5 14083.33

Cool Present Absent E 6 19166.67

Cool Absent Present F 1 2266.67

Cool Absent Present F 2 7666.67

Cool Absent Present F 3 16833.33

Cool Absent Present F 4 16750.00

Cool Absent Present F 5 18333.33

Cool Absent Present F 6 39000.00

Cool Present Present G 1 33333.33

Cool Present Present G 2 43666.67

Cool Present Present G 3 39250.00

Cool Present Present G 4 31833.33

Cool Present Present G 5 22250.00

Cool Present Present G 6 21166.67

Cool Absent Absent H 1 28333.33

Cool Absent Absent H 2 16666.67

Cool Absent Absent H 3 19166.67

Cool Absent Absent H 4 19250.00

Cool Absent Absent H 5 17416.67

Cool Absent Absent H 6 13916.67

Cool Present Absent I 1 50000.00

Cool Present Absent I 2 20166.67

130

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Cool Present Absent I 3 9500.00

Cool Present Absent I 4 2833.33

Cool Present Absent I 5 9833.33

Cool Present Absent I 6 12000.00

Cool Absent Present J 1 4500.00

Cool Absent Present J 2 5833.33

Cool Absent Present J 3 12500.00

Cool Absent Present J 4 20500.00

Cool Absent Present J 5 14916.67

Cool Absent Present J 6 19416.67

Cool Present Present K 1 10444.44

Cool Present Present K 2 12083.33

Cool Present Present K 3 25500.00

Cool Present Present K 4 34083.33

Cool Present Present K 5 17083.33

Cool Present Present K 6 26750.00

Cool Absent Absent L 1 43833.33

Cool Absent Absent L 2 12666.67

Cool Absent Absent L 3 13583.33

Cool Absent Absent L 4 20083.33

Cool Absent Absent L 5 10000.00

Cool Absent Absent L 6 16666.67

Cool Absent Present M 1 22333.33

Cool Absent Present M 2 13500.00

Cool Absent Present M 3 16166.67

Cool Absent Present M 4 34333.33

Cool Absent Present M 5 20916.67

131

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Cool Absent Present M 6 18250.00

Cool Present Absent N 1 23416.67

Cool Present Absent N 2 30000.00

Cool Present Absent N 3 17416.67

Cool Present Absent N 4 18250.00

Cool Present Absent N 5 10916.67

Cool Present Absent N 6 24916.67

Cool Absent Absent O 1 20000.00

Cool Absent Absent O 2 25750.00

Cool Absent Absent O 3 22250.00

Cool Absent Absent O 4 35166.67

Cool Absent Absent O 5 14833.33

Cool Absent Absent O 6 46333.33

Hot Absent Present A 1 13583.33

Hot Absent Present A 2 11250.00

Hot Absent Present A 3 16083.33

Hot Absent Present A 4 14000.00

Hot Absent Present A 5 14833.33

Hot Absent Present A 6 14333.33

Hot Present Absent B 1 9250.00

Hot Present Absent B 2 17000.00

Hot Present Absent B 3 15916.67

Hot Present Absent B 4 12083.33

Hot Present Absent B 5 14000.00

Hot Present Absent B 6 14666.67

Hot Absent Absent C 1 8833.33

Hot Absent Absent C 2 10166.67

132

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Hot Absent Absent C 3 10083.33

Hot Absent Absent C 4 10916.67

Hot Absent Absent C 5 9500.00

Hot Absent Absent C 6 11833.33

Hot Present Present D 1 9833.33

Hot Present Present D 2 20000.00

Hot Present Present D 3 25416.67

Hot Present Present D 4 17250.00

Hot Present Present D 5 12833.33

Hot Present Present D 6 11500.00

Hot Present Absent E 1 11166.67

Hot Present Absent E 2 10500.00

Hot Present Absent E 3 12666.67

Hot Present Absent E 4 13000.00

Hot Present Absent E 5 8333.33

Hot Present Absent E 6 7250.00

Hot Absent Present F 1 6833.33

Hot Absent Present F 2 13250.00

Hot Absent Present F 3 18916.67

Hot Absent Present F 4 18500.00

Hot Absent Present F 5 10833.33

Hot Absent Present F 6 10333.33

Hot Present Present G 1 13083.33

Hot Present Present G 2 16083.33

Hot Present Present G 3 20416.67

Hot Present Present G 4 19500.00

Hot Present Present G 5 12250.00

133

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Hot Present Present G 6 10333.33

Hot Absent Absent H 1 7250.00

Hot Absent Absent H 2 10833.33

Hot Absent Absent H 3 20000.00

Hot Absent Absent H 4 15250.00

Hot Absent Absent H 5 15833.33

Hot Absent Absent H 6 12083.33

Hot Present Absent I 1 6250.00

Hot Present Absent I 2 14333.33

Hot Present Absent I 3 10916.67

Hot Present Absent I 4 6833.33

Hot Present Absent I 5 8666.67

Hot Present Absent I 6 7916.67

Hot Absent Present J 1 3500.00

Hot Absent Present J 2 6166.67

Hot Absent Present J 3 9916.67

Hot Absent Present J 4 13333.33

Hot Absent Present J 5 15833.33

Hot Absent Present J 6 10416.67

Hot Present Present K 1 5500.00

Hot Present Present K 2 7666.67

Hot Present Present K 3 11666.67

Hot Present Present K 4 8333.33

Hot Present Present K 5 17750.00

Hot Present Present K 6 14583.33

Hot Absent Absent L 1 5333.33

Hot Absent Absent L 2 8416.67

134

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Hot Absent Absent L 3 14666.67

Hot Absent Absent L 4 10916.67

Hot Absent Absent L 5 20500.00

Hot Absent Absent L 6 13000.00

Hot Absent Present M 1 17333.33

Hot Absent Present M 2 16916.67

Hot Absent Present M 3 20750.00

Hot Absent Present M 4 19750.00

Hot Absent Present M 5 12516.67

Hot Absent Present M 6 16833.33

Hot Present Absent N 1 3166.67

Hot Present Absent N 2 9500.00

Hot Present Absent N 3 12166.67

Hot Present Absent N 4 15333.33

Hot Present Absent N 5 10962.96

Hot Present Absent N 6 7111.11

Hot Absent Absent O 1 2333.33

Hot Absent Absent O 2 15500.00

Hot Absent Absent O 3 10145.83

Hot Absent Absent O 4 13916.67

Hot Absent Absent O 5 8833.33

Hot Absent Absent P 1 10145.83

Hot Present Present P 2 2250.00

Hot Present Present P 3 4083.33

Hot Present Present P 4 13083.33

Hot Present Present P 5 8750.00

Hot Present Present P 6 6833.33

135

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Warm Absent Present A 1 9583.33

Warm Absent Present A 2 1500.00

Warm Absent Present A 3 8166.67

Warm Absent Present A 4 12000.00

Warm Absent Present A 5 20166.67

Warm Absent Present A 6 20833.33

Warm Present Absent B 1 26800.00

Warm Present Absent B 2 4666.67

Warm Present Absent B 3 2500.00

Warm Present Absent B 4 4500.00

Warm Present Absent B 5 8500.00

Warm Present Absent B 6 6500.00

Warm Absent Absent C 1 13250.00

Warm Absent Absent C 2 3083.33

Warm Absent Absent C 3 11000.00

Warm Absent Absent C 4 15333.33

Warm Absent Absent C 5 19555.56

Warm Absent Absent C 6 20000.00

Warm Present Present D 1 31000.00

Warm Present Present D 2 2666.67

Warm Present Present D 3 3333.33

Warm Present Present D 4 6333.33

Warm Present Present D 5 12666.67

Warm Present Present D 6 8583.33

Warm Present Absent E 1 10000.00

Warm Present Absent E 2 2000.00

Warm Present Absent E 3 3416.67

136

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Warm Present Absent E 4 19750.00

Warm Present Absent E 5 15066.67

Warm Present Absent E 6 12444.44

Warm Absent Present F 1 16000.00

Warm Absent Present F 2 13555.56

Warm Absent Present F 3 29666.67

Warm Absent Present F 4 40000.00

Warm Absent Present F 5 41777.78

Warm Absent Present F 6 44444.44

Warm Present Present G 1 27333.33

Warm Present Present G 2 5083.33

Warm Present Present G 3 8833.33

Warm Present Present G 4 23833.33

Warm Present Present G 5 22333.33

Warm Present Present G 6 23333.33

Warm Absent Absent H 1 13916.67

Warm Absent Absent H 2 7666.67

Warm Absent Absent H 3 36666.67

Warm Absent Absent H 4 32083.33

Warm Absent Absent H 5 88666.67

Warm Absent Absent H 6 53333.33

Warm Present Absent I 1 59333.33

Warm Present Absent I 2 3750.00

Warm Present Absent I 3 4333.33

Warm Present Absent I 4 8888.89

Warm Present Absent I 5 5583.33

Warm Present Absent I 6 7166.67

137

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Warm Absent Present J 1 6666.67

Warm Absent Present J 2 14000.00

Warm Absent Present J 3 12666.67

Warm Absent Present J 4 62666.67

Warm Absent Present J 5 38666.67

Warm Absent Present J 6 66666.67

Warm Present Present K 1 53333.33

Warm Present Present K 2 1916.67

Warm Present Present K 3 4250.00

Warm Present Present K 4 3666.67

Warm Present Present K 5 3166.67

Warm Present Present K 6 6666.67

Warm Absent Absent L 1 4583.33

Warm Absent Absent L 2 21538.46

Warm Absent Absent L 3 57666.67

Warm Absent Absent L 4 70666.67

Warm Absent Absent L 5 31666.67

Warm Absent Absent L 6 59333.33

Warm Absent Present M 1 34000.00

Warm Absent Present M 2 6166.67

Warm Absent Present M 3 42666.67

Warm Absent Present M 4 24666.67

Warm Absent Present M 5 61333.33

Warm Absent Present M 6 35666.67

Warm Present Absent N 1 31025.64

Warm Present Absent N 2 5416.67

Warm Present Absent N 3 5666.67

138

TABLE. VI.1. Cont’d... Thermal Chaoborus Notonectid Bag Depth (m) Total Abundance Warm Present Absent N 4 5333.33

Warm Present Absent N 5 6833.33

Warm Present Absent N 6 2666.67

Warm Absent Absent O 1 2500.00

Warm Absent Absent O 2 8400.00

Warm Absent Absent O 3 50379.33

Warm Absent Absent O 5 50666.67

Warm Absent Absent O 6 36422.53

Warm Present Present P 1 38333.33

Warm Present Present P 2 34333.33

Warm Present Present P 3 7250.00

Warm Present Present P 4 19833.33

Warm Present Present P 5 23416.67

Warm Present Present P 6 12500.00

139

TABLE VI.2. ANOVA results for final (Day 15) chemical conditions in the enclosures in the cool-stratified, warm- and hot-isothermal trials. Significant values are in bold. ‡Note 2 that SiO2 was analysed with a Kruskal-Wallis non-parametric ANOVA (χ test statistic).

Tukey’s Parameter df SS MS F P-value post hoc Cool A, Warm A , pH 2 0.09 0.04 15.34 <0.001 Hot B Conductivity Cool A , Hot B, 2 813.09 406.54 1330.80 <0.001 (µS·cm-1) Warm C Warm A, Hot B, DOC (mg·L-1) 2 2.75 1.37 36.30 <0.001 Cool C Warm A, Hot AB, ‡ Silicon (mg·L-1) 2 - - χ2=6.86 0.03 Cool B

DOC= dissolved organic carbon ‡ Soluble reactive silica was analysed with a non-parametric Kruskal-Wallis ANOVA (χ2 test statistic)

140

TABLE VI.3. Results of a 3 x 2 x 2 repeated-measures analysis of variance of the interactive effects of thermal habitat and the presence of Chaoborus and notonectids on total chlorophyll a concentration (µg·m-3) at the end of the 15 d enclosure experiments, with depth (m) as the repeated measure. Univariate Greenhouse-Geiser (ε) F-values are reported. Significant values are in bold (P<0.05).

Num. Den. Effect Factor ε F P-value df df Between- Thermal 2 36 12.97 233.38 <0.0001

Subjects Chaoborus 1 36 0.24 8.71 <0.01

Notonectid 1 36 0.01 0.44 0.51

Thermal x Chaoborus 2 36 0.38 6.80 <0.01

Thermal x Notonectid 2 36 0.08 1.46 0.25 Chaoborus x 1 36 0.03 0.99 0.33 Notonectid Thermal x Chaoborus 2 36 0.06 1.07 0.36 x Notonectid Within- Depth 3.50 126.15 0.70 8.59 <0.0001

Subjects Depth x Thermal 7.01 126.15 0.70 4.40 <0.001

Depth x Chaoborus 3.50 126.15 0.70 1.45 0.23

Depth x Notonectid 3.50 126.15 0.70 0.25 0.89 Depth x Thermal x 7.01 126.15 0.70 0.77 0.59 Chaoborus Depth x Thermal x 7.01 126.15 0.70 0.29 0.96 Notonectid Depth x Chaoborus x 3.50 126.15 0.70 0.54 0.69 Notonectid Depth x Thermal x Chaoborus x 7.01 126.15 0.70 2.48 0.02 Notonectid

141