Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
ACIDIFICATION AND CLIMATE WARMING: UNDERSTANDING THE IMPACT OF MULTIPLE ANTHROPOGENIC STRESSORS ON ADIRONDACK (NY, USA) LAKES by Kristina Marie Allport Arseneau A thesis submitted to the Department of Biology in conformity with the requirements for the Degree of Doctor of Philosophy Queen’s University Kingston, Ontario, Canada May, 2014 Copyright © Kristina M. A. Arseneau, 2014 ABSTRACT Lakes in the Adirondack Park (NY, USA) are undergoing chemical recovery from acidification. There is now a pressing research need to define recovery targets for acid-impacted sites. Researchers attempting to designate such targets are hampered by two issues: 1) a lack of long-term monitoring data, and 2) the influence of multiple stressors on recovering lakes. This thesis addresses both difficulties by applying paleolimnological techniques within a regional reference lake framework. Using a set of stringent selection criteria, 31 lakes protected from acidification, eutrophication, road salt seepage, and piscivore introductions were identified from 1,469 Adirondack lakes. Ordination techniques showed that the lakes are representative of 2436% of the chemical/morphological variation of Adirondack lakes. Qualitative and quantitative historic analyses found that many of the lakes experienced early watershed and/or fisheries disturbance, highlighting the danger of assuming that a lake’s condition remains static over time. A top-bottom paleolimnological study revealed that the reference lakes have undergone a ‘shifting baseline’ in species assemblages, with increases in colonial and/or warm-water chrysophyte taxa from pre-1900 to present, changes most likely due to regional warming and/or oligotrophication. A subset of three reference lakes were then paired with two Adirondack lakes that acidified and are undergoing chemical recovery from acidification. The acidified lakes underwent a significant shift in species composition since the 1995 implementation of the US Acid Rain Program, indicating biological recovery from acidification. However, both reference and acidified lakes showed increases in colonial chrysophytes since ca. 1970-1980, a trend correlated with mean annual air temperature and ice-cover measures in the two reference lakes. Long-term species changes in acidified/reference lakes suggest that the recovering lakes will not return to their pre-disturbance state but will instead move to a state characterized by an increased abundance of colonial taxa/warm-water species. Overall, this thesis demonstrates the utility of ii pairing paleolimnological techniques with a regional reference site dataset for tracking shifting baselines and defining recovery targets, a method that could be applied to examine other stressors in other regions, thereby addressing a critical management need. iii CO-AUTHORSHIP Chapter 3 was co-authored by Dr. Charles T. Driscoll, Graham Pope, Cassandra Cummings, and Dr. Brian F. Cumming, and represents original work completed as part of my Ph.D. thesis. I was responsible for project design, led all field work, performed the qualitative historical analyses and ordination analyses, and am the principal author on the paper. Chapter 4 was co-authored by Dr. Charles T. Driscoll and Dr. Brian F. Cumming, and represents original work completed as part of my Ph.D. thesis. I was responsible for project design, led all field work, collected and analyzed all chrysophyte data, performed the ordination, univariate, and ANOSIM analyses, and am the principal author on the paper. Chapter 5 was co-authored by Dr. Charles T. Driscoll, Cassandra M. Cummings, Ayla Fenton, and Dr. Brian F. Cumming, and represents original work completed as part of my Ph.D. thesis. I was responsible for project design, led all field work, collected and analyzed chrysophyte data from Moss Lake, South Lake, and Queer Lake, performed the ANOSIM, ordination, and correlation analyses, and am the principal author on the paper. iv ACKNOWLEDGEMENTS If the old adage that it takes a village to raise a child is true, then it certainly must take a lifetime of relationships to create a Ph.D. thesis. If I neglect to mention one of the seemingly endless number of people who have supported me in this, that oversight is entirely mine and not a reflection of the value of their contribution. First, I must thank my supervisor and mentor Dr. Brian F. Cumming, who taught me that above all else science is about constantly asking and refining questions. His guidance in this project has been absolutely invaluable and his constant support has been deeply appreciated. I must also thank my committee: Dr. John P. Smol, Dr. Neal Scott, and Dr. Andrew Paterson, whose support and suggestions have greatly improved these manuscripts. I also want to thank the collective members of the Paleoecological Environmental Assessment and Research Laboratory both past and present, for their support, friendship, and their willingness to discuss any obscure matter of paleo-science with gusto at any time of day or night. This dissertation required a great deal of field work and would have been impossible without the hard work and dedication of my numerous field crews. I would like to sincerely thank: Cassandra Cummings, Ayla Fenton, Iain MacKenzie, Jerome Velasco, Colin Robinson, Mark Kelly, and Brendan Wiltse. I also want to acknowledge the support I received from individuals at the New York State Department of Environmental Conservation when navigating the permitting process for this project, particularly Mr. Scott Healy. The support of my colleagues and friends in New York State was absolutely invaluable in this project. I would like to thank Dr. Curt Stager for his many informative discussions of Adirondack ‘Heritage Lakes’ which were inspirational in this study design. I want to acknowledge the researchers and staff at the Adirondack Ecological Centre who hosted my field v crew and I at the Huntington Wildlife Forest, particularly Dr. Colin Beier who also provided icecover records for two reference lakes in this project. This project would not have been possible without the long-term monitoring data collected by the Adirondack Lakes Survey Corporation. I want to particularly acknowledge the help I have received from Mr. Jed Dukett, who was always happy to answer any questions I had about their data. Lastly, I must acknowledge the collective members of the Adirondack Research Council, whose yearly conference introduced me to the new datasets like the Adirondack GIS portal and greatly deepened my understanding of the intricacies of the private-public partnerships at the heart of the Adirondack Park. A variety of funding agencies have provided support to the work undertaken in this dissertation. The majority of the funding for this research was supported by an NSERC Discovery Grant awarded to my supervisor, Dr. Brian Cumming. I have received funding support from NSERC, Ontario Graduate Scholarships, and Queen’s University and my co-author Dr. Charles T. Driscoll received funding from the New York State Energy Research and Development Authority (NYSERDA). Funding to support water chemistry monitoring by the Adirondack Lakes Survey Corporation has been provided by the NYSERDA and the US Environmental Protection Agency. Lastly, but certainly not least, I must thank my family and friends for their support: my father Réjean Arseneau who was my first true science mentor, my mother Deborah Allport who taught me so many of the things not covered in university, my brother Vincent Arseneau, my dear friends Katie Griffiths, Katy Marsh, Matt Marsh, and Heather Mawby, and, of course, my partner Laura Schaefli, who was always my first reviewer and editor-in-chief, my toughest critic and biggest cheerleader. Danke, mein Herz. vi TABLE OF CONTENTS Section Page Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Co-authorship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Table of contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii List of tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi List of figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii List of abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Chapter 1 – General introduction General introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Literature cited. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Chapter 2 - Literature review Impact of acid deposition on lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Sources, forms, and the terrestrial-aquatic link . . . . . . . . . . . . . . . . . . . . . . . 9 Influence on the chemical & biological characteristics of lakes . . . . . . . . . . 11 Emission controls and chemical/biological recovery . . . . . . . . . . . . . . . . . . 12 Impact of climate change on lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Physical/chemical effects of climate change on lakes . . . . . . . . . . . . . . . . . . 17 Biological effects of climate change on lakes . . . . . . . . . . . . . . . . . . . . . . . . 20 Understanding the effects of multiple stressors in lake environments . . . . . . . . 26 Reference sites & reference conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii 27 Chrysophytes – general ecology & use as paleo-indicators . . . . . . . . . . . . . . . 30 The Adirondack Park . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Literature cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Chapter 3 – Tracking shifting baseline conditions due to climate change in lakes: a novel use of regional reference sites and paleolimnology Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Literature cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Tables & table captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Figures & figure captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Chapter 4 – Adirondack (NY, USA) reference lakes show a pronounced shift in chrysophyte species composition since ca. 1900 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Literature cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 viii Tables & table captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Figures & figure captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Chapter 5 – Understanding biological recovery from acidification: applying a reference-site framework in the Adirondacks (NY, USA) Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Literature cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Tables & table captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Figures & figure captions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Chapter 6 – General discussion and conclusions General discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Literature cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Appendices Appendix A. Histogram plots of Adirondack Lakes Survey (ALS) variables . . . 186 Appendix B. List of analytes measured in 2010 and 2011 water chemistry samples from reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Appendix C. Location information for 31 Adirondack reference lakes . . . . . . . 188 Appendix D. Summary diagram of species assemblages in 31 Adirondack reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix 189 Appendix E. Relationship between colonial chrysophyte taxa and DOC in 31 reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 Appendix F. Stocking history, recent netting survey results, land-use characteristics, and watershed disturbance histories for South Lake, Queer Lake, Moss Lake, Arbutus Lake, and Wolf Lake . . . . 191 Appendix G. Seasonal Kendall τ results for long-term monthly chemistry data. 193 Appendix H. Stratigraphies of chrysophyte relative abundance . . . . . . . . . . . . 194 Appendix I. List of taxon codes used in nMDS ordination diagram . . . . . . . . . 196 Appendix J. Principal Components Analysis (PCA) results for acidified and reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 197 LIST OF TABLES Table Page Chapter 3 Table 3.1. Summaries of lake and watershed characteristics, lake morphometrics, and water chemistry variables for 31 Adirondack reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Table 3.2. Summaries of early watershed disturbance and fisheries disturbance in 31 Adirondack reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Chapter 4 Table 4.1 SIMPER test results showing taxa that contributed to the significant difference in species composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Chapter 5 Table 5.1. ANOSIM and SIMPER test results assessing whether chrysophyte assemblages have changed between two a priori defined time periods. 166 Table 5.2. Kendall τ correlation test results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Appendices Table B1. List of analytes measured in 2010 and 2011 water chemistry samples from reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Table C1. Location information for 31 Adirondack reference lakes . . . . . . . . . . . 188 Table F1. Stocking history, recent netting survey results, land-use characteristics, and watershed disturbance histories for South Lake, Queer Lake, Moss Lake, Arbutus Lake, and Wolf Lake . . . . . . . . . . . . . xi 191 Table G1. Seasonal Kendall τ results for long-term monthly chemistry data . . . . 193 Table I1. List of taxon codes used in nMDS ordination diagram . . . . . . . . . . . . . 196 xii LIST OF FIGURES Figure Page Chapter 3 Figure 3.1. Conceptual diagram outlining the biological response of two lakes (Lake A and Lake B) to acidification and climate change since preindustrial times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Figure 3.2. Conceptual spatial diagram of a paleolimnological reference site study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Figure 3.3. Schematic outlining the reference lake selection and characterization process followed in this investigation . . . . . . . . . . . . . . . . . . . . . . . . . 99 Figure 3.4. Map of the Adirondack Park (NY, USA) showing the locations of a set 31 Adirondack reference lakes minimally impacted by anthropogenic stressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Figure 3.5. PCA ordinations of ALS survey data . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Chapter 4 Figure 4.1. Schematic showing the paleolimnological study design . . . . . . . . . . . 131 Figure 4.2. PCA of environmental variables and Redundancy analysis (RDA) using PCA-axes scores as input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Figure 4.3. PCA of chrysophyte assemblages in top (present-day) and bottom (pre-1900) samples in 26 reference lakes . . . . . . . . . . . . . . . . . . . . . . 133 Figure 4.4. Total relative abundance (%) of colonial chrysophyte taxa in top (present-day) and bottom (pre-1900) samples in 26 reference lake . . xiii 134 Chapter 5 Figure 5.1. Long-term chemistry trends in acidified and reference lakes . . . . . . . 168 Figure 5.2. 210Pb and 214Bi activities in sediment cores from acidified and reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure 5.3. Total relative abundance of colonial chrysophytes (% colonial) and chrysophyte-inferred pH (CI-pH) values for two acidified lakes (South Lake and Queer Lake) and three reference lakes (Arbutus Lake, Wolf Lake, Moss Lake) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Figure 5.4. Non-metric Multidimensional Scaling (nMDS) plots for acidified and reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Figure 5.5. Non-metric Multidimensional Scaling (nMDS) plots relating to South Lake species assemblages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Figure 5.6. Non-metric Multidimensional Scaling (nMDS) plots relating to Queer Lake species assemblages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Figure A1. Histogram plots of Adirondack Lakes Survey (ALS) variables . . . . . . 186 Appendices Figure D1. Summary diagram of species assemblages in 31 Adirondack reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Figure EI. Relationship between colonial taxa and DOC in 31 reference lakes . . 190 Figure H1. Stratigraphies of chrysophyte relative in acidified lakes . . . . . . . . . . . 194 Figure H2. Stratigraphies of chrysophyte relative abundance in reference lakes . 195 Figure I1. Principal Components Analysis (PCA) results for acidified and reference lakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv 197 LIST OF ABBREVIATIONS Symbol Meaning ABC-top AEZ ALS ALSC ANC ANOSIM APA BC BDL CAAA CFCs CI-pH DEM DCA DIC DI-pH DOC ELS EMAP FAM FDM GIS NAAQS NAPAP NYSDEC PAR PCA PIRLA RDA RMSEP NY US USA US ARP US EPA SA SIMPER SUNY-ESF SWAP TP TN WS-SA Zmax Average Bray-Curtis similarity of replicate top samples Adirondack Ecological Zone Adirondack Lakes Survey Adirondack Lakes Survey Corporation Acid neutralizing capacity Analysis of Similarities Adirondack Park Agency Bray-Curtis similarity Below detection limit Clean Air Act Amendments (1970 & 1990) Chlorofluorohydrocarbons Chrysophyte-inferred pH Digital elevation model Detrended Correspondence Analysis Dissolved inorganic carbon Diatom-inferred pH Dissolved organic carbon Eastern Lakes Survey Environmental Monitoring and Assessment Program Flow accumulation model Flow direction model Geographic information system National Ambient Air Quality Standards National Acid Precipitation Program New York State Department of Environmental Conservation Photosynthetically active radiation Principal Components Analysis Paleoecological Investigation of Recent Lake Acidification Redundancy Analysis Root-mean squared error of prediction New York United States United States of America United States Acid Rain Program United States Environmental Protection Agency Surface area Similarities Percentages test State University of New York School of Environmental Science and Forestry Surface Water Acidification Programme Total phosphorus Total nitrogen Watershed to surface area ratio Maximum depth xv CHAPTER 1 GENERAL INTRODUCTION In 2000, the Nobel-prize winning chemist Paul Crutzen and the renowned diatomist Eugene Stoermer argued that the past three centuries should be renamed the Anthropocene, a new geological epoch reflecting the profound influence of human activity on ecosystems around the globe (Crutzen and Stoermer 2000). Though controversial, the term serves to highlight how humans have affected all levels of ecosystem functioning in marine, terrestrial, and freshwater environments. Indeed, even remote alpine and arctic sites have recorded the effects of anthropogenic stressors (Catalan et al. 2013), suggesting that ‘pristine’ sites protected from all human activities are unlikely to exist (Bennion et al. 2011). As anthropogenic influence has increased around the world, a growing number of environmental stressors have begun to act in tandem on freshwater ecosystems. Since the mid1990s, there has been a shift in the ecological literature away from a focus on single-stressor impacts (e.g. acidification or eutrophication) to an emphasis on multiple stressor systems (e.g. acidification and/or, land-use change, invasive species, climate change, etc.). For example, in a highly cited manuscript, Schindler et al. (1996) noted that the interactive effects of acidification and climate change increased UV-B penetration in an experimentally acidified lake more than in an unmanipulated reference lake in the Experimental Lakes Area (ON, Canada). Similarly, in a seminal case-study review of both terrestrial and aquatic environments, Paine et al. (1998) argued that multiple stressors can lead to ‘ecological surprises’, or novel species communities, which may be irreversible. Systems impacted by multiple stressors pose a particular management challenge. Fundamentally, most environmental stressors have legislative controls which are largely separate from other, possibly interacting, stressors. To take the example of Schindler et al. (1996) above, 1 in the United States acid deposition is regulated under Title IV of the US Clean Air Act, ozonedepleting compounds like chlorofluorocarbons (CFCs) are regulated under Title VI of the Clean Air Act, and the US Environmental Protection Agency is currently engaged in developing regulatory strategies for CO2 emissions under the Clean Air Act mandate. While each stressor may be subject to different legislative controls under the Clean Air Act, acidification, UV-B radiation, and climate change can all influence the same waterbody. Moreover, while legislative and management strategies dealing with single environmental stressors like acidification often have either an explicit or implicit recovery goal of a return to pre-disturbance state (Baker et al. 1990), the interactive effects of multiple stressors may prevent recovering systems from returning to pre-disturbance conditions. Such ‘shifting baselines’ make defining recovery targets difficult – in the face of changing climate conditions, what should a ‘recovered’ site look like? Researchers and managers who are attempting to designate recovery targets for impacted systems must address two issues: 1) the confounding influence of multiple stressors and, more fundamentally, 2) a lack of long-term monitoring data. In a recent meta-analysis of both terrestrial and aquatic recovery studies, Jones and Schmitz (2009) found that only 20% of the studies had access to pre-disturbance data. In aquatic ecosystems, long-term monitoring records, where they exist, are typically short (< 50 years) and are insufficient to track the effects of stressors like climate change which work on decadal-to-centennial timescales. Fortunately, paleolimnological techniques can be used to reconstruct the environmental histories of lakes from yearly-to-centennial (or indeed, millennial) timescales (Smol 2008). Paleolimnological techniques have been used to understand the effects of acidification and eutrophication in aquatic ecosystems, as well as define biological recovery from those stressors (Battarbee 1999; Battarbee 2 et al. 2005) and study the effects of multiple co-occurring stressors in lake ecosystems (Smol 2010). When it comes to understanding shifting baselines, paleolimnological studies have a rich history of using remote, minimally-disturbed alpine or arctic sites to track the effects of climate change (Battarbee et al. 2002; Catalan et al. 2013; Gregory-Eaves et al. 1999). However, such sites are typically removed from areas heavily influenced by anthropogenic stressors like acidification, eutrophication, and/or land-use change and may be subject to different climate controls (i.e. high elevation/arctic vs. temperate lowlands), making them less than ideal for defining recovery targets. Conversely, there is an expansive limnological bioassessment literature that uses minimally-disturbed reference sites within anthropogenically-impacted regions to define recovery targets for impacted sites, often by using modelling techniques (Bailey et al. 2004; Downes et al. 2002; Stoddard et al. 2006). However, these studies usually lack the long-term perspective offered by paleolimnological studies and so are unable to define predisturbance conditions for impacted sites and identify whether or not shifting baselines are occurring in the reference/impacted systems due to climate change. This project aims to bridge the paleolimnological and bioassessment literature by applying paleolimnological techniques in a study of a regional set of minimally-disturbed reference lakes in the Adirondack Park (NY, USA). The main goal of this project is to use the regional reference sites to track shifting baselines due to regional stressors like climate change and then use that information to define reasonable recovery targets for lakes recovering from acidification in that region. The Adirondack region is an ideal study location for this project because the area has been heavily influenced by acid deposition (Driscoll et al. 1991), the acidification history of the region has been documented in previous paleolimnological studies 3 (Cumming et al. 1992; Cumming et al. 1994), and large synoptic surveys have been undertaken in the region (Krester et al. 1989). Additionally, long-term chemical monitoring data exist for many Adirondack lakes (Roy et al. 2011) and recent studies have found that some lakes are undergoing chemical recovery from acidification (Driscoll et al. 2007; Waller et al. 2012). However, while chemical recovery may be progressing, the Adirondacks are also undergoing regional warming, a trend that has been most pronounced since the 1970s (Jenkins 2010). This dissertation contains six chapters, including the general introduction (Chapter 1) and literature review (Chapter 2). In Chapter 3, a conceptual framework is presented for pairing paleolimnological techniques with a regional reference site approach. Key concepts are discussed such as: what ‘good’ reference sites are and how they can be identified; the useful role reference sites play in studies of multiple stressors; why reference site studies should be regional in nature, rather than relying on a small number of sites; and how reference sites can be used to help define recovery targets. This framework is then applied in the Adirondack Park. A set of 31 reference sites is identified out of the 1,469 lakes surveyed as part of the Adirondack Lakes Survey using a set of stringent selection criteria. The reference sites are then evaluated critically using both qualitative and quantitative historical methods and ordination techniques. In Chapter 4, a paleolimnological ‘top-bottom’ study of the reference sites is carried out using scaled chrysophytes as a proxy group (Chrysophyceae and Synurophyceae). The relationship between the present-day distribution of scaled chrysophyte taxa and physical/chemical variables in the reference lakes is examined using direct ordination techniques. The change in chrysophyte assemblages from pre-1900 times to present is then examined in the reference lakes using indirect ordination methods and quantified using both univariate and multivariate analyses. Because of the reference lake study design, it is possible to conclude that 4 changes in chrysophyte assemblages from pre-1900 to present are not due to acidification, eutrophication, road salt seepage, or piscivores introductions. Significant shifts in the composition of chrysophyte species assemblages in the reference lakes are discussed in the context of regional environmental change and shifting baselines for biological recovery. In Chapter 5, a subset of the three reference lakes is paired with two Adirondack lakes which acidified and are undergoing chemical recovery from acidification. Long-term changes in the chrysophyte assemblages of the five lakes are reconstructed and compared. In the acidified lakes, multivariate analyses are used to identify if the species assemblages have undergone a significant shift in species composition since the implementation of the 1995 US Acid Rain Program and are experiencing biological recovery from acidification. Then, indirect ordination methods are used to assess if the post-1995 species assemblages of each acid-impacted lake are returning to a species assemblage characteristic of pre-disturbance times. In a subsequent analysis, the pre-1900 assemblages of the acid-impacted lakes are compared with those of the entire regional reference lake set to identify which reference lakes were most similar to the acidlakes in species composition pre-1900. A subset of reference lakes are then used to define probable recovery endpoints for the acidified lakes. Lastly, changes in chrysophyte species composition in two reference lakes are compared with long-term temperature and ice-cover records. Finally, Chapter 6 presents a general discussion of the results of this investigation and suggests possible future research directions. This dissertation makes several important contributions to the ecological literature. First, it improves our understanding of biological recovery from acidification, a subject that is currently under-documented (Verdonschot et al. 2013). Second, it contributes to our understanding of how climate change, possibly interacting with the long-term effects of acid 5 deposition, influences algal assemblages, a topic that is of keen research interest as the impact of climate change on lakes continues to grow (Winder and Sommer 2012). Lastly, this dissertation provides a critical link between the paleolimnological and bioassessment literature and provides a framework for examining the effects of multiple stressors on biological recovery. Though the focus of this dissertation is on acidification, the framework outlined could be easily adapted and applied to look at other pressing issues, including eutrophication, invasive species introductions, or land-use change. Researchers, policy makers, and lake managers no longer have the luxury of considering stressors in isolation. Frameworks like the one used in this dissertation that combine the long-term perspective of paleolimnology with the rigorous study design used in bioassessment work will become increasingly useful and necessary as lakes both within and outside protected areas are increasingly influenced by a growing number of anthropogenic stressors. Literature Cited Bailey RC, Norris RH, Reynoldson TB. (2004) Bioassessment of freshwater ecosystems: using the reference condition approach. Kluwer Academic Publishers, New York, 170 p. Baker JP, Bernard DP, Christensen SW, Sale MJ. (1990) Biological effects of changes in surface water acid-base chemistry. Report 13. Acidic deposition: state of science and technology. National Acid Precipitation Assessment Program (NAPAP), Washington, DC, pp. 13-1-13-381. Battarbee RW. (1999) The importance of palaeolimnology to lake restoration. Hydrobiologia 395/396:149-159. Battarbee RW, Anderson NJ, Jeppesen E, Leavitt PR. (2005) Combining palaeolimnological and limnological approaches in assessing lake ecosystem response to nutrient reduction. Freshwater Biology 50:1772-1780. Battarbee RW, Thompson R, Catalan J, Grytnes J-A, Birks HJB. (2002) Climate variability and ecosystem dynamics of remote alpine and arctic lakes: the MOLAR project. Journal of Paleolimnology 28:1-6. 6 Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. Catalan J, Pla-Rabés S, Wolfe AP, Smol JP, Rühland KM, Anderson NJ, Kopáček J, Stuchlík E, Schmidt R, Koinig KA, Camarero L, Flower RJ, Heiri O, Kamenik C, Korhola A, Leavitt PR, Psenner R, Renberg I. (2013) Global change revealed by palaeolimnological records from remote lakes: a review. Journal of Paleolimnology 49:513-535. Crutzen PJ, Stoermer EF. (2000) The "Anthropocene". Global Change Newsletter 41:17-18. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Cumming BF, Smol JP, Kingston JC, Charles DF, Birks HJB, Camburn KE, Dixit SS, Uutala AJ, Selle AR. (1992) How much acidification has occurred in Adirondack region lakes (New York, USA) since preindustrial times?. Canadian Journal of Fisheries and Aquatic Sciences 49:128-141. Downes BJ, Barmuta LA, Fairweather PG, Faith DP, Keough MJ, Lake PS, Mapstone BD, Quinn GP. (2002) Monitoring ecological impacts: concept and practice in flowing waters. Cambridge University Press, Cambridge, 452 p. Driscoll CT, Driscoll KM, Roy KM, Dukett J. (2007) Changes in the chemistry of lakes in the Adirondack region of New York following declines in acidic deposition. Applied Geochemistry 22:1181-1188. Driscoll CT, Newton RM, Gubala CP, Baker JP, Christensen S. (1991) Adirondack mountains. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer-Verlag, New York, pp. 133-202. Gregory-Eaves I, Smol JP, Finney BP, Edwards ME. (1999) Diatom-based transfer functions for inferring past climatic and environmental changes in Alaska, USA. Arctic, Antarctic, and Alpine Research 31:353-365. Jenkins J. (2010) Climate change in the Adirondacks: the path to sustainability. Cornell University Press, Ithaca, New York, 183 p. Jones HP, Schmitz OJ. (2009) Rapid recovery of damaged ecosystems. PLoS ONE 4:e5653-1- e5653-6. Krester W, Gallagher J, Nicolette J. (1989) Adirondack lakes survey 1984-1987: an evaluation of fish communities and water chemistry. Adirondack Lakes Survey Corporation, Ray Brook, NY, 437 p. 7 Paine RT, Tegner MJ, Johnson EA. (1998) Compounded perturbations yield ecological surprises. Ecosystems 1:535-545. Roy K, Houck N, Hyde P, Cantwell M, Brown J. (2011) The Adirondack long-term monitoring lakes: a compendium of site descriptions, recent chemistry and selected research information. New York State Energy Research and Development Authority, Albany, 298 p. Schindler DW, Curtis PJ, Parker BR, Stainton MP. (1996) Consequences of climate warming and lake acidification for UV-B penetration in North American boreal lakes. Nature 379:705-708. Smol JP. (2008) Polluion of lakes and rivers: a paleoenvironmental perspective. Blackwell Publishing Ltd, Malden. Smol JP. (2010) The power of the past: using sediments to track the effects of multiple stressors on lake ecosystems. Freshwater Biology 55 (Suppl. 1):43-59. Stoddard JL, Larsen DP, Hawkins CP, Johnson RK, Norris RH. (2006) Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16:1267-1276. Verdonschot PFM, Spears BM, Feld CK, Brucet S, Keizer-Vlek H, Borja A, Elliott M, Kernan M, Johnson RK. (2013) A comparative review of recovery processes in rivers, lakes, estuarine and coastal waters. Hydrobiologia 704:453-474. Waller K, Driscoll C, Lynch J, Newcomb D, Roy K. (2012) Long-term recovery of lakes in the Adirondack region of New York to decreases in acidic deposition. Atmospheric Environment 46:56-64. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. 8 CHAPTER 2 LITERATUTE REVIEW Outline The goal of this literature review is to provide the necessary background information for a greater appreciation of chapters 3-5. Several topics are discussed including: the effects of acid deposition on lakes, the impact of climate change on lakes, tools for understanding the influence of multiple stressors on lakes (with a focus on using reference sites), the ecology of scaled chrysophytes and their use as paleo-indicators, and the influence of acid deposition and climate change in the Adirondack Park. The impact of acid deposition on lakes Sources, forms, and the terrestrial-aquatic link Acid deposition is a globally-occurring phenomenon that affects North America (primarily eastern North America and the western US), Europe, and parts of Asia (Driscoll 2012). Acid deposition is primarily a by-product of fossil fuel combustion. The combustion of fossil fuels from sources like electric utilities and motor vehicles leads to the production of SO2 and NOx emissions. SO2 and NOx react with water in the atmosphere to produce sulphuric and nitric acids (H2SO4 and HNO3, respectively), which then return from the atmosphere in the form of wet and dry deposition1. Wet deposition includes acid rain, snow, and sleet whereas dry deposition includes acids bound to particles, gases, or a combination of both (aerosols) (Munson and Gherini 1991). 1 In agricultural areas, NH3 emissions may also be an important source of acid deposition (Munson and Gherini 1991). 9 In the absence of acid deposition, the acid-base status of waterbodies is primarily controlled by bicarbonate buffering and organic acids (Munson and Gherini 1991). Organic acids can act as buffers but their presence tends to decrease lake-water pH (Kullberg et al. 1993). In areas influenced by acid deposition, the acid-base status of waterbodies is primarily determined by mineral acids and is intimately linked to the terrestrial environment. In general, the primary determinant of the acid-base status of a lake is the hydrologic flow path (Goldstein et al. 1984). In the soil matrix, H+ associated with incoming acids will undergo cation exchange. Cations with a higher valence (3+, 2+) will be preferentially exchanged for the H+ and sulphate will adsorb to iron and aluminum oxides (Jenkins et al. 2007). Soil pH tends to increase with soil depth, and so the deeper precipitation percolates through the soil matrix, the higher the likelihood that H+ will be replaced by a cation, neutralizing the acids prior to groundwater flow entering a nearby waterbody (Munson and Gherini 1991). For this reason, lakes surrounded by thicker watershed soils tend to have more alkaline surface waters (Newton and Driscoll 1990). The buffering function of watershed soils can be exceeded by high levels of acid deposition. If the leaching of base cations occurs at a rate greater than they are supplied to the soil matrix by weathering, the soils will acidify. Sulphate adsorption is concentration dependent, and so increasing soil sulphate concentration leads to a greater release of sulphate to nearby waterbodies (Jenkins et al. 2007). The solubility of Al also increases as soil pH drops below 5.5, leading to the export of toxic inorganic Al to nearby waterbodies (Munson and Gherini 1991). In general, when soil base-cation concentration is less than 20%, the input of strong acids will result in an incomplete neutralization of H+ and a mobilization of toxic inorganic aluminum (Cronan and Schofield 1990). 10 While sulphate deposition may make the largest contribution to acid anion concentration in surface waters (Baker et al. 1990b), nitrate is often the main constituent of episodic acidification (Wigington et al. 1996). Terrestrial environments are typically N-limited and so NO3- is actively taken up during the growing season. During the winter, NO3- builds up in the snowpack and the spring snowmelt results in an acid pulse to receiving waterbodies (Munson and Gherini 1991). In areas heavily influenced by acid deposition, episodic acidification can be sufficient to result in spring fish kills (Baker et al. 1996). In general, waterbodies with an acid neutralizing capacity (ANC) greater than 50 μeq L-1 are considered protected from episodic acidification whereas lakes with an ANC < 0 μeq L-1 are considered chronically acidic, and any additional input of acid will result in a decrease in pH (Driscoll et al. 2001). Influence of acid deposition on the chemical & biological characteristics of lakes Acid deposition causes a suite of chemical changes in freshwaters. Most importantly, acid deposition leads to acidification (i.e. decreases in lake-water pH and ANC) when the buffering capacity of surrounding watershed soils and in-lake processes are surpassed. High levels of acid deposition also lead to an increased concentration of strong acid anions and toxic forms of inorganic aluminum in surface waters (Munson and Gherini 1991). An increased flux of cations increases the total solute concentration (and conductivity) of receiving waters and can lead to calcium depletion in watershed soils (Likens et al. 1996). Acid deposition is also associated with reductions in lake-water DOC concentrations, allowing for increased UV-B penetration (Schindler et al. 1996b) and changes in lake thermostructure (Snucins and Gunn 2000). The influence of acid deposition on TP concentrations is less well-understood. An increased concentration of SO42- in sediments may reduce the sequestration of TP (Caraco et al. 1991). In contrast, soil acidification may lead to a decreased export of TP from watershed soils (Eimers et 11 al. 2009). Overall, reviews of acidification research have not revealed a consistent link between acidification and eutrophication/oligotrophication (Olsson and Pettersson 1993; Schindler 1988). In general, the acidification of freshwaters leads to a simplification of aquatic foodwebs. The toxic effects of acidification on aquatic taxa is primarily related to an interruption of normal osmoregulatory function, as well as impairment of respiratory processes by inorganic aluminum (Baker et al. 1990a). Acidification may also affect aquatic organisms indirectly by changing food availability, foodweb structures, and/or predation levels (Baker and Christensen 1991). Most acid-sensitive aquatic organisms can persist when pH is greater than 6.0 (Baker et al. 1990a). Numerous studies have shown that decreases in pH lead to a progressive loss of taxa in phytoplankton (Siegfried et al. 1989), zooplankton (Siegfried and Sutherland 1992), and fish assemblages (Krester et al. 1989). Coarse metrics, like biomass and production, within taxonomic groups often do not show consistent patterns with acidification (Baker and Christensen 1991). The fact that coarse metrics are not typically reliable indicators in acidification studies is likely due to functional complementarity: as one taxon is lost with acidification, another more acid-tolerant taxon is released from competition and increases in abundance until carrying capacity is reached (Havens and Carlson 1998). Despite these species replacements, acidification is a key management concern, as the acidification of freshwaters leads to the loss of desirable sportfish taxa like lake trout (Salvelinus namaycush) and increases in nuisance filamentous algae in the littoral zone (Baker and Christensen 1991). Emission controls and chemical/biological recovery Growing concerns about the human health effects of SO2 and NOx emissions, as well as the effects of acid deposition on terrestrial and aquatic environments, led to the creation of legislative control programs for SO2 and NOx in North America and Europe. Driscoll et al. 12 (2010) provides a useful summary of the main research/policy steps taken in Europe and the United States to address acid deposition. Briefly, in the US, legislation was implemented as part of the 1970 and 1990 Clean Air Act Amendments (CAAA). In 1970, National Ambient Air Quality Standards (NAAQS) were introduced that focused primarily on limiting concentrations of individual pollutants in urban areas, rather than their long-distance transport. In 1980, the National Acid Precipitation Assessment Program (NAPAP) was created, a multi-agency research program tasked with examining all issues associated with acid deposition. In 1990, NAPAP released a series of reports that provided the scientific foundation for the 1990 CAAA, which included the creation of the US Acid Rain Program (ARP). The ARP initiated an innovative ‘cap and trade’ program for SO2 emissions with the goal of a 50% reduction below 1980 levels by 2010, as well as intensity-based NOx emissions targets for electricity generating units (Burns et al. 2011). S-emissions for the continental United States peaked in 1970 and have since declined (Husar et al. 1991). In 2009, SO2 emissions from sources governed by the ARP were 67% lower than 1980 levels (Burns et al. 2011). In the same year, NOx emissions from sources governed by the ARP were 67% lower than 1995 levels. Overall, emissions of SO2 from all sources (including those not governed by the ARP) were 59% lower than 1990 levels and NOx emissions were 40% lower than 1990 levels. Chestnut and Mills (2005) estimated that the human health and environmental benefits of the US ARP are over $100 billion, compared to an implementation cost of $3 billion, arguably making the 1990 CAAA one of the most successful pollution-control measures ever implemented in the United States. Clean air legislation in the US, Canada, and Europe has resulted in chemical recovery from acidification in some regions. Declines in lake sulphate concentration have been almost 13 universally documented in acid-impacted areas, while some regions have also experienced increases in ANC and decreases in H+ (Garmo et al. 2014; Skjelkvåle et al. 2005). Driscoll et al. (2001) lists three chemical recovery targets for freshwaters: 1) pH > 6.0 (to protect acid-sensitive organisms), 2) ANC > 50 μeqL-1 (to protect against episodic acidification), and 3) inorganic aluminum levels < 2 μmol L-1 (to protect fish from the toxic effects of aluminum). Modelling studies have suggested that chemical recovery will be a protracted process (Chen and Driscoll 2005; Sullivan et al. 2011). Several factors are likely to slow chemical recovery in acid-impacted regions, including the release of stored sulphate from watershed soils (Driscoll et al. 1998), base cation depletion in soils (Likens et al. 1996), and increases in DOC concentrations. While DOC increases should be expected with decreased acid deposition (Monteith et al. 2007), in the shortterm increased DOC concentrations may decrease pH and ANC in recovering surface waters (Lawrence et al. 2013; Waller et al. 2012). Increases in DOC may, however, lead to a decrease in the concentration of toxic inorganic monomeric aluminum, improving conditions for fish (Lawrence et al. 2013). It is expected that chemical recovery will promote biological recovery from acidification after a lag period that will likely differ between taxonomic groups (Driscoll et al. 2001). Recovery is expected to be characterized by a decreased abundance of acid-tolerant taxa and an increased abundance of acid-sensitive taxa, as well as the re-establishment of extirpated species (Baker et al. 1990b). Biological recovery can be measured using a variety of metrics including species relative abundances, the presence/abundance of indicator species, and metrics like species richness. Case studies in the United States and elsewhere have identified biological recovery from acidification (Arseneau et al. 2011; Greenaway et al. 2012a; Havas et al. 1995; Hynynen and Meriläinen 2005; Kopáček et al. 2002), though in many cases species responses are 14 modest. Beyond insufficient chemical recovery, there are several factors that may delay biological recovery from acidification, including the limited dispersal ability for some taxa (Keller and Yan 1998), damage to sedimentary eggs banks (Binks et al. 2005), increased predation (Nilssen and Wærvågen 2002), and biological resistance from established acid-tolerant species communities (Derry and Arnott 2007). There is growing recognition that recovering species assemblages will likely not return to their pre-disturbance state, even with a return to predisturbance chemical conditions (if possible), due to the influence of additional environmental stressors. In particular, paleolimnological studies of biological recovery from acidification have suggested that climate change may prevent recovering species assemblages from returning to their pre-disturbance state (Arseneau et al. 2011; Battarbee et al. 2014). Shifting baselines caused by regional stressors like climate change pose a particular problem to lake managers and policy makers when it comes to designing suitable recovery targets for acid-impacted systems. The next section explores some of the physical, chemical, and biological effects of climate change on lakes, followed by a discussion of tools ecologists can use to understand the influence of multiple anthropogenic stressors on aquatic ecosystems. The impact of climate change on lakes Globally, from 1880 to 2012, mean surface temperature increased by an estimated 0.85ºC (Hartmann et al. 2013) and models predict an additional increase between 0.3-4.8 ºC by 20812100 relative to the 1986-2005 time period (Collins et al. 2013). Paleoclimatic studies suggest that temperatures experienced in the Northern Hemisphere during the last thirty years were likely the warmest over the last 1,400 years (Masson-Delmotte et al. 2013).The effects of global warming will be regionally-specific and may be difficult to predict. In North America, the warming trend has been most pronounced over the past 50 years (Field et al. 2007) but this 15 warming trend is not evenly distributed. By the late 21st century, temperatures in the southern, western, and eastern parts of North America will likely have warmed by an estimated 2-3 ºC while warming in the north may exceed 5ºC (Christensen et al. 2007). Changes in precipitation will also vary regionally. Increases in drought frequency in the US Southwest are likely to be accompanied by increased rainfall in the US Northeast (Field et al. 2007). Such changes in temperature and precipitation regimes are currently having, and will continue to have, wideranging and complex effects on both terrestrial and aquatic ecosystems, influencing chemical and nutrient cycling (George 2010), seasonal match-or-mismatch between predator and prey species (Durant et al. 2007), parasitism and disease (Marcogliese 2001), species distributions (Walther 2010), and more. Indeed, climate warming is frequently cited as one of the main conservation concerns of the 21st century (Baron et al. 2009; Strayer and Dudgeon 2010). Lakes can be considered ‘sentinels’ of climate change as they integrate the effects of regional warming from the atmosphere, the catchment, and within-lake processes (Adrian et al. 2009). To understand the effects climate change is having and will have on lakes, researchers draw on data from a variety of sources. In some cases, high-quality long-term monitoring records of lake physical, chemical, or biological data exist (Winder and Schindler 2004), while in other cases inferences are drawn from unusual or extreme weather events (e.g. unusually warm years, Forsström et al. 2005; Jankowski et al. 2006). Modelling studies can also be useful when trying to project the impacts of regional warming (e.g. Fang and Stefan 2009). While some processes have been well-characterized (e.g. reduction in the ice-covered period, Duguay et al. 2006; Magnuson et al. 2000), others are more poorly understood (e.g. the interactions of climate change with multiple stressors like eutrophication, Moss 2012). Complications also arise when comparing different lake-types (i.e. deep, dimictic lakes vs. shallow, polymitic lakes), as they 16 differ in their thermal/chemical structures, as well as their response to environmental change (e.g. nutrient loading, Taranu et al. 2010). Certainly, an exhaustive review of the effects of climate warming across a range of lake-types would be beyond the scope of this dissertation. As such, this section will focus primarily on the likely physical, chemical, and biological impacts of climate change on North-temperate dimictic lakes. Physical/chemical effects of climate change on lakes One of the most visible effects of regional warming on lakes is the loss of ice-cover. Regional surveys have shown a loss of seasonal ice-cover in lakes and rivers in the northern hemisphere due to earlier ice-off dates (break-up dates in the spring) and/or later ice-on dates (freeze-up dates in the fall) (Duguay et al. 2006; Futter 2003; Magnuson et al. 2000)2. Using a process-oriented, water-quality model (MINLAKE96), Fang and Stefan (2009) suggested that warming in the United States will be associated with a reduced ice-covered period of up to 90 days at 2xCO2 concentrations and a reduced ice thickness of up to 0.44 m. Changes in ice-cover dynamics may differ regionally. For example, Beier et al. (2012) found that a reduced icecovered period in five Adirondack lakes was due primarily to a delay in ice-on rather than earlier ice-off dates and hypothesized that heavy snowfall in the Adirondack region may minimize changes in ice-off dates. The study by Beier et al. (2012) represents an interesting contrast to work in other parts of North America which showed greater changes in ice-off rather than ice-on dates (Duguay et al. 2006). Changes in the length of the ice-covered period have an important influence on lake thermostructure. A reduction in the ice-covered period can lead to a longer period of thermal stratification and warming in epilimnetic waters can lead to an enhancement of thermal gradients 2 For a review of the controls on ice phenology, see Kirillin et al. (2012) and Livingston et al. (2010). 17 (Vincent 2009). Fang and Stefan (2009) suggested that in a 2xCO2 scenario, the summer stratification period in US lakes may be extended by up to 67 days and surface waters may warm by up to 5.2ºC. An earlier ice-off date can result in an earlier onset of stratification (De Stasio et al. 1996), though some lakes may instead show an extended period of spring mixing (Weyhenmeyer et al. 1999). Changes in the onset of stratification and surface-water warming will likely influence thermocline depth by leading to either deeper and shallower thermoclines (De Stasio et al. 1996). For lakes less than 500 ha in surface area, water transparency is likely the primary determinant of lake thermostructure (Fee et al. 1996), with clear lakes having deeper thermoclines (Schindler et al. 1996a). The influence of water transparency on the thermal structure of small lakes partially explains why increases in surface-water temperatures are more commonly documented with regional warming than increases in hypolimnetic-water temperatures (Adrian et al. 2009). In coloured lakes, increases in air temperatures may have little effect on hypolimnetic waters due to reduced solar heating of profundal waters (Snucins and Gunn 2000). Longer and/or stronger periods of thermal stratification can have important consequences for chemical gradients in lakes. In oxic conditions, bacterial decomposition increases nutrient concentrations in the hypolimnion. If oxygen is depleted and anoxic conditions occur, redox reactions lead to the release of reduced iron and stored phosphate from sediments (see summaries in Moss 2012 and Wetzel 2001). The internal loading of phosphorus can lead to hypolimnetic nutrient concentrations greater than those of inflowing waters and can stimulate cyanobacterial blooms (Pettersson 1998). Thus, if longer/stronger periods of thermal stratification increase the length of summertime anoxic periods (Fang and Stefan 2009), the internal loading of phosphorus from sediments may increase (Jankowski et al. 2006). In contrast, a longer stratification period 18 may lead to nutrient depletion in the epilimnion due to prolonged isolation from the nutrient-rich hypolimnion (Winder and Sommer 2012). Changes in precipitation regimes with regional climate change will also influence lakewater chemistry. Regional droughts can release previously stored SO42- and metals from wetland/littoral sediments and lead to reacidification events (Arnott et al. 2001). In contrast, increased precipitation can lead to an increased leaching of dissolved organic and inorganic compounds from the watershed, as both stream and groundwater fluxes of anions, cations, and dissolved organic matter are closely tied to precipitation events (Deyton et al. 2009; Eimers et al. 2008; Inamdar et al. 2009). Recently, studies have noted increases in DOC concentrations in North American and European lakes (Evans et al. 2005; Monteith et al. 2007). The potential causes of increases in DOC are currently debated and may be due to reduced sulphate deposition, increases in temperature, and/or increases in precipitation/runoff (Erlandsson et al. 2008; Evans et al. 2006; Monteith et al. 2007; Weyhenmeyer and Karlsson 2009). In general, there is a need to improve understanding of the controls on DOC dynamics, because DOC trends can be sitespecific (Adrian et al. 2009) and differ between regions (Zhang et al. 2010). This mini-review has highlighted some of the complex effects climate change may have on physical/chemical characteristics of lakes. These complexities deepen when considering the biological effects that climate change can have on aquatic ecosystem, as individual taxa are influenced by the physical/chemical characteristics of a lake, as well as by their competitive interactions and predator-prey relationships with other individuals and species. The section below discusses some of the biological effects of climate change on lakes, focusing on phytoplankton, zooplankton, and fish. 19 Biological effects of climate change on lakes While lakes may appear superficially to be relatively simple ecosystems, a single temperate lake may be home to thousands of co-existing and competing species of phytoplankton, zooplankton, invertebrates, and vertebrates. A review of the impacts of climate warming across so many biological families is evidently beyond the purview of this dissertation. This section, therefore, focuses on the groups of most interest to this project: phytoplankton (particularly the Chrysophyceae and Synurophyceae), zooplankton (particularly cladoceran zooplankton) and, to a lesser extent, planktivorous and piscivorous fish. Phytoplankton responses to climate change: Climate change is already influencing the seasonality, productivity, and composition of algal assemblages. A number of studies have shown that lakes are experiencing an earlier spring phytoplankton bloom (Adrian et al. 1999; Weyhenmeyer et al. 1999; Winder and Schindler 2004), the timing of which is largely determined by light regime (Sommer et al. 2012). Modelling work by De Senerpont Domis et al. (2013) concluded that a reduced ice-covered period and an earlier onset of stratification with regional warming can lead to advancements in the spring phytoplankton bloom in temperate lakes. The effects of climate change on overall phytoplankton productivity may be more difficult to predict. In a study of 205 small, oligotrophic lakes in Sweden along a latitudinal gradient, Weyhenmeyer et al. (2013) found that, out of 7 physical/chemical variables, the duration of the open-water (ice-free) season best explained phytoplankton biomass, suggesting that an increasing ice-free period with regional warming may increase overall phytoplankton biomass. Similarly, an increased internal loading of phosphorus due to longer periods of stratification may stimulate phytoplankton production (Jeppesen et al. 2009). In contrast, increased water 20 temperatures with regional warming may increase grazing rates on phytoplankton, reducing phytoplankton biomass (Winder and Sommer 2012). Reynolds (1988) provides a useful framework for understanding phytoplankton ecology and, by extension, phytoplankton responses to climate warming. Phytoplankton maximal growth is limited by two factors: access to essential nutrients and access to light. Nutrient limitation can be considered a form of stress on phytoplankton, and involuntary vertical translocations in the water column by wind-induced mixing can be considered a form of disturbance. Phytoplankton can therefore be grouped together based on their adaptive strategies to stress and disturbance as: competitors (C-strategists; selected during periods of low stress & low disturbance), ruderals (Rstrategists; selected during periods of low stress & high disturbance), and stress-tolerant species (S-strategists; selected during periods of high stress & low disturbance)3. Small, fast-growing diatoms with a low sinking-rate can be considered C-strategists; large, heavily-silicified diatoms that depend on turbulent mixing to maintain their position in the euphotic zone can be considered R-strategists; and, mobile taxa adapted to low-nutrient conditions can be considered S-strategists. The C-R-S groups should not be considered mutually exclusive; rather, taxa may show a variety of adaptations that represent a gradation between the three strategies (see, for example, Olrik 1994). The emphasis Reynolds (1988) puts on nutrients and turbulence (as well as its corollary, stratification) can help us understand which algal groups may benefit from regional warming. For example, as detailed above, a reduced period of ice-cover may increase the length or strength 3 Reynolds (1988) argues that there is no adaptive strategy for plankton in high-stress, high-disturbance conditions (e.g. oligotrophic, polymictic sites). While this is likely an over generalization, such sites would most likely be dominated by benthic algal taxa. In another classification scheme, Reynolds et al. (2002) lists 31 trait-differentiated functional groups of algae. While useful for understanding why certain algal groups may be present in certain habitats seasonally, the schematic is arguably less intuitive to use for hypothesis testing of algal responses to climate warming than Reynolds (1988). 21 of thermal stratification (Vincent 2009). As such, we may expect that an increased period of thermal stratification (i.e. a reduction in disturbance from mixing) may favour C- and/or Sstrategists over R-strategists. Indeed, a recent paleolimnological meta-analysis demonstrated that recent pan-continental increases in small Cyclotella taxa (C-strategists) have been accompanied by declines in large, heavily-silicified diatoms (R-strategists) (Rühland et al. 2008). Similarly, increases in the abundance of colonial chrysophyte taxa (highly mobile taxa adapted to low nutrient conditions, S-strategists4) documented in paleolimnological studies across North America (Arseneau et al. 2011; Flear 2011; Ginn et al. 2010; Paterson et al. 2004) are likely related, at least in part, to warming. In general, recent reviews have suggested that longer periods of thermal stratification may result in an increased abundance of small taxa with a high surfaceto-volume ratio (and, therefore, lower sinking rates), buoyant taxa (such as cyanobacteria) and motile taxa such as flagellated species (Winder and Sommer 2012). Frameworks such as the one proposed by Reynolds (1988) are appealing because they are intuitive and provide researchers with the ability to make predictions about how phytoplankton species composition may change with climate warming. However, the difficulty remains that climate warming influences physical, chemical, and biological factors in tandem and so unexpected species responses may occur. For example, Shatwell et al. (2008) found that a shift in the timing of spring succession events in Lake Mügglesse associated with regional warming led to a an unexpected increase in cold-adapted diatom species. Typically, it is suggested that warmwater plankton species will increase with increases in air temperatures (Winder and Sommer 4 Olrik (1994) suggests that the scaled chrysophytes can be considered a transitional group between C- and Rstrategists and between C- and S-strategists. Similarly, Kristiansen (2005) suggests that most chrysophyte taxa cannot be exclusively grouped into any of the 3 strategies. Presumably, some of this ambiguity arises from the different seasonal distributions of chrysophyte taxa (Siver & Hamer 1992). However, as scaled chrysophytes are motile taxa most commonly occurring in low nutrient conditions, the ‘S-strategy’ described by Reynolds (1988) may be a suitable as a general description for the group. For additional information, see ‘Chrysophytes (the golden-brown algae): general ecology and use as paleo-indicators’ in this introduction. 22 2012), though the study by Shatwell et al. (2008) emphasizes the point that exceptions to general ecological assumptions occur. As an additional complicating factor, changes in DOC concentration will also likely play an important role in controlling phytoplankton seasonality and vertical structure, as DOC influences lake thermostructure (Fee et al. 1996), irradiance regimes (Snucins and Gunn 2000), and the concentration of certain micronutrients (i.e. chelated iron, Fuss et al. 2010). Part of the difficulty in predicting algal species responses to regional warming arises from the fact that phytoplankton are subject to both bottom-up and top-down control. For this reason, some of the effects of climate change on zooplankton and fish taxa are discussed below. Zooplankton responses to climate change: In North-temperate lakes, the spring phytoplankton bloom is typically followed by a peak in zooplankton abundance, sometimes resulting in a clearwater phase (Sommer et al. 2012). The timing and magnitude of the spring phytoplankton bloom determines the timing and size of the subsequent zooplankton bloom. For this reason, there has been a considerable amount of research interest in a possible ‘mismatch’ between phytoplankton and zooplankton with the advancement of the spring phytoplankton bloom (Durant et al. 2007; Winder and Schindler 2004). Such a decoupling of phytoplankton and zooplankton peak abundances has not been seen universally (Adrian et al. 2006; Preston and Rusak 2010). De Senerpont Domis et al. (2007) argued that daphniid overwintering strategy would likely determine whether or not a mismatch in spring phenology would occur, as taxa that overwinter are less likely to suffer a mismatch than taxa that must recruit from the sedimentary eggbank. Beyond the potential decoupling of phytoplankton and zooplankton phenology, changes in water temperature will likely have an influence on zooplankton species assemblages. Different zooplankton taxa have different physiological tolerances to temperature (Moore et al. 1996). 23 Taxa with a low thermal optimum and taxa with a thermal optimum just below their lethal thermal limit may be more likely to undergo local extinctions with increases in surface temperatures (De Senerpont Domis et al. 2007). Changes in temperature may also influence zooplankton size structure. In a recent meta-analysis, Daufresne et al. (2009) summarized a set of three biological rules that suggest that small-bodied taxa may benefit from regional warming5. In a recent top-bottom study of 44 south-central Ontario lakes, Korosi et al. (2010) found that Daphnia decreased in size from pre-industrial times to present and attributed that change to either an effect of regional warming and/or an acid deposition. However, studies of cladoceran zooplankton have not universally demonstrated the group’s sensitivity to climate change. For example, in a transect study of arctic and subarctic lakes, Sweetman et al. (2008) found no consistent response in cladoceran assemblages to regional warming from pre-industrial times to present. Because zooplankton are subject to top-down control, the effects of climate change on fish populations will likely influence the overall abundance of zooplankton, as well as the composition of the group. The following section examines some of the possible effects of climate change on populations of planktivorous/piscivorous fish. Fish responses to climate change: Different planktivorous and piscivorous fish taxa have different thermal preferences and are often described as cold-water, cool-water, or warm-water taxa. There is a concern that surface-water warming will result in a reduction in cold-water habitat for desirable taxa like lake trout (Salvelinus namaycush) and an expansion in warm-water habitat for taxa like yellow perch (Perca flavescens) (Ficke et al. 2007; Stager and Thill 2010). An increased hypoxic zone in lakes with longer periods of stratification would place cold-water 5 As cited in Daufresne et al. (2009), the three rules are: Bergmann’s rule – small-bodied species are more common in warmer environments James’ rule – within a species, small-sized organisms are generally found in warmer environments Temperature-size rule – for ectotherms, individual body size decreases with increasing temperature 24 taxa under a “temperature-oxygen squeeze”, forcing cold-water taxa to abandon cold-water refugia in the hypolimnion and move into warmer waters where they may experience an increased incidence of stress and disease (Ficke et al. 2007). While the loss of cold-water habitat is a key management concern, observational studies and modelling studies have not consistently shown a reduction in cold-water habitat with regional warming. For example, Snucins and Gunn (2000) showed an increase in cold-water habitat during a warm year, likely because an early establishment of strong thermal gradients inhibited mixing between surface and bottom waters. Similarly, a modelling study by De Stasio et al. (1996) showed an increase in habitat available to warm, cool, and cold-water fish taxa with a 2xCO2 scenario, though some modelling simulations did predict a reduction in habitat for coldwater taxa. Modelling studies by Fang and Stefan (2009) did show an increased potential for fishkills during summer time anoxic events with a 2xCO2, but also showed a decreased potential for winterkill due to a shortened ice-cover period. Overall, changes in the composition/abundance of freshwater fishes will likely have an important top-down influence on zooplankton/phytoplankton assemblages through trophic cascades that may be difficult to predict (Sommer et al. 2012). This mini-review has highlighted how climate change can affect the physical/chemical/biological processes of lakes. These interactions can be complex, difficult to predict, and are likely to increase in intensity over time. As lakes are increasingly influenced by multiple anthropogenic stressors, these stressors may interact with climate change to produce unexpected ‘ecological surprises’ (Paine et al. 1998). The following section discusses some of the strategies ecologists can use to understand the effects of multiple stressors on lakes. 25 Understanding the effects of multiple stressors in lake environments The introduction sections above discuss the effects of acid deposition and climate change on lakes in isolation from other stressors. This is something of an oversimplification as a growing number of lakes are increasingly influenced by multiple anthropogenic stressors (Keller 2009). Acidification, eutrophication, land-use change, invasive species, and climate change can all have complex and interactive effects on aquatic ecosystems. Regional droughts can lead to reacidification events and alter species assemblages (Arnott et al. 2001; Faulkenham et al. 2003). Changes in land-use can change the quantity of DOC entering aquatic environments (Mattsson et al. 2009), influencing lake thermal structure and chemical gradients, a trend that may be exacerbated by regional warming. The fact that all of these stressors and more may co-occur in lakes makes both understanding the environmental effects of multiple stressors and the remediation of impacted sites difficult. Ecologists have employed a variety of approaches to understand the influence of multiple stressors on individual taxa and whole species assemblages. Microcosms can be used to test for the influence of toxicants in isolation and in combination (e.g. Locke 1991), though such experiments may be criticized for being an oversimplification of aquatic ecosystems. Mesocosm experiments have also been used to document the synergistic and antagonistic effects of multiple stressors (Christensen et al. 2006; Folt et al. 1999), though again, mesocosms are unable to mimic the full complexity of freshwater ecosystems. More rarely, whole lake manipulations have been used to elucidate the effects of multiple stressors on waterbodies (e.g. acidification and climate change, Schindler et al. 1996b). Such large-scale manipulations are uncommon but provide valuable insights into changes in the structure and function of aquatic ecosystems. 26 In general, a lack of long-term monitoring data for aquatic ecosystems can make understanding the effect of multiple stressors on species assemblages difficult. Here, paleolimnological studies can play an important role by providing the long-term perspective needed for management and environmental stewardship (Wolfe et al. 2012). Often, paleolimnologists employ a multi-proxy approach when examining the influence of multiple stressors on aquatic ecosystems (Michelutti and Smol 2013), an effective but time-consuming approach. In such studies, multiple paleolimnological indicator groups are reconstructed and the different ecological sensitivities of the organism groups are used to reconstruct the influence of different stressors in the waterbody (Quinlan et al. 2008). Alternatively, mathematical models can be applied in an attempt to help distinguish between the effects of individual stressors on species assemblages (Simpson and Anderson 2009). However, such approaches are complex, require a deep understanding of statistics, and are constrained by the available data, the correlational structure of underlying data, and many assumptions. Arguably, the most intuitive approach to understand the effects of multiple stressors on aquatic ecosystems is the reference site (or reference condition) approach. When impacted sites are paired with reference sites, changes in species assemblages in the reference sites can be used to understand regional environmental changes (for example, in climate; Yan et al. 1996). The reference sites can be used to determine recovery targets for the impacted sites that account for regional stressors (Yan et al. 2004, USEPA 2012). The reference lake concept has been applied in a variety of ways in ecological studies, and so is examined in greater detail below. Reference sites & reference conditions The concept of a ‘reference site’ or ‘reference condition’ is frequently applied in limnological (and paleolimnological) work but different studies apply the reference term in 27 different ways, resulting in confusion about its meaning. First, a reference site is not normally a control site in the traditional use of the word. The word ‘control’ is normally used to designate sites (or, more accurately, sampling units; Bailey et al. 2004) that are identical to impacted sites in physical, chemical and/or biological characteristics, except for the presence of a stressor (or stressors) under study (Downes et al. 2002). True controls are rare in most non-manipulative limnological studies however (though, for an example see Keatley et al. 2008), and so limnologists often rely on ‘reference’ frameworks for comparison between impacted and nonimpacted sites. In its simplest application, a reference site is a site defined by the absence of a stressor of interest (e.g. logging, Paterson et al. 1998; acidification, Jeziorski et al. 2013; eutrophication, Hawryshyn et al. 2012; metal contamination, Ilyashuk et al. 2003). The main appeal of this approach is that it is relatively easy to implement (Chessman et al. 2008) and can be effective at detecting the influence of stressors on species assemblages, especially if executed in a balanced study design (Persson 2008). The main weakness of this approach, however, is that these reference sites may be influenced by environmental stressors other than the main stressor of interest. The uncontrolled-for stressors may have a significant influence on the species assemblages independent of the stressor of interest, affecting the study results. An alternative, more rigorous reference approach is to designate sites as being in ‘reference’ condition based on the absence of multiple anthropogenic stressors (e.g., acidification, eutrophication, introduced species, road salt seepage, etc.). In this framework, the reference sites represent the best condition that could be expected for impacted sites in the region (Reynoldson et al. 1997). Minimally-disturbed reference sites possess a high degree of biotic integrity (or ‘naturalness’) because they have been protected from multiple anthropogenic 28 stressors (Hamilton et al. 2010; Stoddard et al. 2006). A variety of methods can be used to identify minimally-disturbed reference sites, including selection criteria and best professional judgement, multivariate or multimetric approaches, and/or modelling or ordination techniques (Hawkins et al. 2010; Soranno et al. 2011). The minimally-disturbed reference site approach is commonly used in bioassessment studies. In such studies, the present-day ecological condition of the reference sites is used to define recovery targets for impacted systems, often by using modelling techniques (e.g., which taxa would be present at this impacted site if it were in reference condition?, sensu Bailey et al. 2004). Whatever approach is used, the criteria used to designate reference sites must be clearly defined; otherwise, the sites have little interpretative or comparative value (Brucet et al. 2013). Arguably, the ideal reference site would be a hypothetical ‘pristine’ site that has never experienced any anthropogenic impact (Bennion et al. 2011). Unfortunately, the pervasive influence of atmospheric deposition and climate change make it highly unlikely that such a site exists. Alternatively, in paleolimnological studies a pre-disturbance time period, typically one prior to the industrial revolution (e.g. ca. 1850, Battarbee et al. 2011), can be designated to serve as a recovery target for lakes impacted by anthropogenic stressors. While this historical perspective is critically important, especially for identifying if the present-day species composition of impacted sites is significantly different from their historical norm, a historical condition may not always be an appropriate recovery target. There is growing recognition that the influence of multiple anthropogenic stressors on lakes can prevent impacted sites from returning to their pre-disturbance, or historical, condition, even with management intervention (Bennion et al. 2011; Hobbs et al. 2009). This phenomenon is often referred to as a “shifting baseline”; specifically, the influence of an additional, non-remediated stressor (such as invasive 29 species, land-use change, or climate warming) can prevent recovering sites from returning to their pre-disturbance state (or baseline condition). The best-attainable condition for that impacted site may then become a state that is still ‘degraded’ when compared to its historical (or baseline) condition. A growing number of recovery case studies have suggested that shifting baselines due to climate change may already be occurring (Arseneau et al. 2011; Battarbee et al. 2012; Greenaway et al. 2012b; Helliwell and Simpson 2010). Indeed, in a recent report the USEPA called for long-term monitoring at regional reference sites to specifically monitor the effect of climate warming on the sites and track shifting baseline conditions (USEPA 2012). In a shifting baseline scenario, reference sites can be used to define recovery targets for impacted sites in the same geographic area (Yan et al. 1996; Yan et al. 2004). Regardless of the approach ecologists use to understand the influence of multiple stressors on aquatic ecosystems and designate recovery targets, they must employ taxa in their studies that are sensitive to the stressors being investigated. This dissertation focuses on the scaled chrysophytes, an algal group that has frequently been employed in paleolimnological studies of lake acidification. The following section discusses the ecology of chrysophytes and their use as paleolimnological indicators. Chrysophytes (the golden-brown algae) - general ecology & use as paleo-indicators In general terms, chrysophyte algae have two flagella unequal in length, and possess chlorophyll a and c, as well as accessory pigments (β-carotene and phytoxanthylls) which give the algae their characteristic golden-brown colour (Kristiansen 2005). Morphologically, chrysophytes are a diverse group, including motile and non-motile taxa, unicellular and colonial taxa, and taxa with branched structures (Kristiansen 2005). The cells of chrysophytes can be naked (unadorned), protected in loricas, or covered in siliceous scales and bristles (Sandgren 30 1995). This last group of chrysophytes, the scaled chrysophytes, are members of the classes Synurophyceae and Chrysophyceae. The scaled chrysophytes are of particular interest to paleolimnological studies because their siliceous scales/bristles preserve well in lake sediments and possess morphologically distinct features that can be used for identification (Smol 1995). Scaled chrysophytes are used as an indicator group in this investigation and so when the term ‘chrysophytes’ is used in this dissertation, it is primarily in reference to the scaled chrysophytes. Members of the Synurophyceae and Chrysophyceae differ in morphology, size, and life strategies. Scaled chrysophytes are euplanktonic and both unicellular (Mallomonas, Paraphysomonas, Spiniferomonas) and colonial taxa (Synura, Chrysosphaerella, Chrysodidymus) exist. For colonial taxa, the colonies range in size from 2 cells (Chrysodidymus synuroides, Wujek and Wee 1983) to colonies > 200 µm in diameter (Synura uvella, Siver 2003). Due to their size, large chrysophytes such as M. caudata and large synuran colonies are only susceptible to predation by large herbivores (i.e. large Daphnia taxa) whereas smaller taxa can be predated upon by small zooplankton (i.e. Bosmina/Eubosmina, copepods, and small daphniids) (Sandgren and William 1995). Some scaled chrysophyte taxa such as Chrysosphaerella have been reported to be phagotrophic (Holen and Boraas 1995). Scaled chrysophytes can reproduce both asexually and sexually and all chrysophytes form siliceous resting stages known as stomatocysts or statospores (or more simply as cysts) (Kristiansen 2005). Like their siliceous scales, chrysophyte cysts preserve well in sediments and can be used in paleolimnological studies, though often it is not known which cyst is derived from which species (Duff et al. 1995; Wilkinson et al. 2002). Scaled chrysophytes are known to occur on all of the continents except for Antarctica and have been identified in both temperate and tropical locales (Kristiansen 2008). In general, 31 chrysophytes are considered characteristic of cool, oligotrophic, soft-water lakes, with low conductivity and neutral-to-slightly acidic pH (Sandgren 1988). However, numerous investigations have demonstrated that this is an oversimplification. For example, chrysophyte taxa have been identified in warm, tropical locales (e.g. Northeast India, Saha and Wujek 1990) and eutrophic waters (Kristiansen 1988). The fact that chrysophytes tend to be a dominant flora at low-TP conditions does not appear to be because they are indifferent to TP concentrations. In a survey of 91 lakes, Watson et al. (1997) demonstrated that the total abundance of chrysophytes increased with TP concentrations when values were < 10 µg L-1, after which their summertime biomass levelled out. The authors argued that chrysophytes may have been excluded from higher-TP lakes due to increased herbivory by large Daphnia or because the high-TP lakes in the dataset were primarily shallow, polymictic ponds that were unsuitable for large colonial chrysophytes. The seasonal occurrence of chrysophyte taxa is controlled by several factors including water column stability, light conditions, temperature, nutrient availability, and predation (Kristiansen 2005). Typically, chrysophytes are considered a major component of the spring flora (Siver 1995). While true in many lakes, different unicellular and colonial chrysophyte taxa have different seasonal occurrences in the water column; some taxa are known to be tightly tied to the summer months (e.g. Mallomonas punctifera, Siver and Hamer 1992) or to form blooms under-ice (e.g. Synura petersenii, Watson et al. 2001). During periods of thermal stratification, large stable metalimnetic populations of both unicellular (M. acaroides, M. caudata, M. crassisquama) and colonial chrysophyte taxa (Synura, Chrysosphaerella) can occur near or below the thermocline (Nicholls 1995; Sandgren 1988; Siver 2003). Such blooms have been associated with taste and odour issues in lakes and reservoirs (e.g. S. petersenii blooms, Nicholls 32 and Gerrath 1985; Watson et al. 2001) and are disrupted with the deterioration of thermal stratification at the fall overturn (Fee 1976). The causes of metalimnetic chrysophyte blooms are not well understood. While chrysophytes can persist in low-light, low-temperature conditions, such conditions may not be optimal for their growth (Healey 1983). Nutrient conditions may be more favourable at or below the thermocline due to increased nutrient concentration from decomposition (Sandgren 1988). Additionally, the metalimnetic position of the blooms may provide a refuge from predation by large Daphnia (Nicholls 1995) or protection from harmful UV-B radiation (Vinebrooke et al. 2002). As such, during extended periods of thermal stratification, flagellated algae like chrysophytes may have a competitive advantage over nonflagellated taxa like diatoms because of their ability to seek out optimal conditions in the water column (Winder and Sommer 2012). Furthermore, at an assemblage level, large colonial chrysophytes may have an advantage over unicellular taxa because large flagellates are often superior swimmers (i.e. have faster swimming speeds and show larger diurnal migrations, Sommer 1988). As proxies of past environmental change, the occurrence of scaled-chrysophytes along environmental gradients has been well-characterized. pH is found consistently to an important chemical variable structuring chrysophyte assemblages in different regions (e.g. New York: Cumming et al. 1992a, Connecticut: Siver and Hamer 1989, the US Northeast: Hyatt et al. 2010, Ontario: Dixit et al. 1988; Paterson et al. 2001, maritime Canada: Ginn et al. 2010). Chrysophytes are also sensitive to specific conductance (Siver and Hamer 1989), temperature (Siver and Hamer 1989; Siver and Lott 2012), and lake trophic status (Siver and Marsicano 1996). As indicators, they have primarily been used in paleolimnological studies of acidification (Cumming et al. 1992b; Dixit et al. 1988), though other applications include salinization (Zeeb 33 and Smol 1991), eutrophication (Siver and Marsicano 1996), and seasonality reconstructions (Pla-Rabes and Catalan 2011). Scaled-chrysophytes may respond earlier than diatoms to acidification (Cumming et al. 1992b; Cumming et al. 1994; Dixit et al. 2002) and chemical recovery (Arseneau et al. 2011; Greenaway et al. 2012b) because they are euplanktonic and because some taxa are common in the spring and so may be more sensitive to episodic acidification (e.g. Mallomonas hindonii, Cumming et al. 1992a). Scaled-chrysophytes were chosen as a paleo-indicator in this investigation primarily because of their sensitivity to pH and because they have previously been shown to be abundant in the surface waters of the Adirondack Park, the study location of this dissertation. The following section provides information on the Adirondack Park, particularly the influence of acid deposition and climate change on waterbodies in the region. The Adirondack Park The Adirondack Park is the largest protected area in the contiguous United States at ~6 million acres6. Unlike many protected areas in North America, the Adirondack Park follows a mixed-use model, including both publically and privately-owned lands. Approximately 51% of the land in the Adirondack Park is privately owned, 43% is state land, and 6% is water (APA 2011a). Based on 2010 census data, the Adirondack Park Agency (APA) estimated that approximately 130,000 people live within the Adirondack Park (Barge 2011), though millions of seasonal tourists visit the region each year. Areas within the Adirondack Park have different land-use classifications and are subject to different levels of protection. The classifications range from intensive-use areas (e.g. the Whiteface ski area near Lake Placid) to wild forest and wilderness areas (the most restrictive land-use designation). While recreational activities like 6 The Adirondack Ecological Zone (AEZ) extends beyond the boundaries of the Adirondack Park and is defined as the areas greater than 305 m (1000 ft) in elevation in the Adirondack Mountains (Krester et al. 1989). 34 hiking and canoeing are permitted in wild forest and wilderness areas, logging and development are not (APA 2011b). The climate of the Adirondacks can be described as cool continental. Average monthly temperatures range from approximately -7ºC in the winter to 21-27ºC in the summer (Jenkins and Keal 2004). Precipitation is evenly distributed throughout the year, with precipitation amounts of ~3-4 inches (~76-100 mm) per month (Jenkins and Keal 2004). Precipitation falls as snow or as snow mixed with rain from November until April and the snowpack accumulates from December until spring (Driscoll et al. 1991). The growing season is short, approximately 100-120 days per year (Jenkins and Keal 2004). Northern hardwoods comprise approximately 50% of Adirondack forests and are dominated by yellow birch (Betula alieghaniensis), beech (Fagus grandifolia), and sugar maple (Acer saccharum). Another 25% of Adirondack forests are mixed forests of hardwoods and softwoods, including red spruce (Picea rubens), balsam fir (Abies balsamea), eastern hemlock (Tsuga canadensis), and white pine (Pinus strobus). Ten percent of Adirondack forests are coniferous stands, and approximately 5% of all Adirondack vegetation are wetlands (Driscoll et al. 1991). The Adirondack region receives high levels of acid deposition. Major wind patterns in the United States move from west to east, and so SO2 and NOx emissions from the industrial Midwest are transported to the Adirondacks (Driscoll et al. 1991). A precipitation gradient exists in the Adirondack Mountains, with the highest amounts of the most acidic precipitation falling in the Southwest (Ito et al. 2002). The sensitivity of Adirondack lakes to acid precipitation is caused by the fact that the region is primarily underlain by Ca-poor granitic bedrock, that soils in the region are primarily shallow acidic Spodosols, and that the steep slopes in the region contribute to waterbodies having a relatively short water residence time (< 1 year), limiting the 35 potential for in-lake processes to neutralize incoming acids (Driscoll et al. 1991). A 1980s survey of 1,469 Adirondack lakes found that 27% were chronically acidic and 24% had pH values less than 5 (Krester et al. 1989). The majority of lakes with a pH < 5 were located in the Southwestern Adirondacks whereas lakes with a pH > 6 were more evenly dispersed throughout the region (Gallagher and Baker 1990). Lakes insensitive to acid deposition (> 50 μeq L-1) were either associated with thick glacial till deposits or were carbonate-influenced (i.e. associated with CaCO3 bearing rock in till or exposed bedrock) (Jenkins et al. 2007). Since the implementation of the 1990 CAAA, Adirondack lakes have begun to undergo chemical recovery from acidification. Long-term monitoring data have documented a decrease in SO42-, NO3-, and inorganic monomeric aluminum in surface waters since the 1990s, as well as increases in pH and ANC (Driscoll et al. 2007; Waller et al. 2012). A recent analysis of longterm monitoring data also found that many Adirondack lakes are experiencing increases in DOC, but that the trend is not universally documented (Lawrence et al. 2013). Modelling studies suggest that chemical recovery will be a protracted process in the Adirondacks, as many Adirondack lakes will still have ANC values less than 50 μeq L-1 by 2050 (Chen and Driscoll 2005). Without further reductions in sulphate deposition, many Adirondack watersheds will experience deposition levels greater than target loads for maintaining lake-water ANC > 50 μeq L-1 by 2100 (Sullivan et al. 2012). While lakes in the Adirondacks may be undergoing chemical recovery from acidification, the region is also experiencing the effects of climate change. Annual air temperatures in the Adirondacks have increased at a rate of 1.3ºC per 100 years since 1900 (Jenkins 2010). The rate of warming in the Adirondacks is somewhat higher than the rate documented for the Northeast as a whole (0.8ºC per 100 years) (Hayhoe et al. 2007). In both the Adirondacks and the Northeast as 36 a whole, the rate of temperature change has increased since approximately 1970, with winter temperatures increasing at a greater rate than summer ones (Hayhoe et al. 2007; Jenkins 2010). However, in an analysis of monthly temperature data from 1975-2005, Stager et al. (2009) found the June and September temperatures have increased significantly over the time period, though many other months showed non-significant increases. Stager et al. (2009) also found no statistically significant trends in total monthly precipitation trends in Adirondack records from 1975-2005, though precipitation levels in the Northeast are expected to increase by 7-14% by 2070-2099 (Hayhoe et al. 2007). Total snowfall (as snow-water-equivalent) in the Northeast has decreased at a rate of ~3.5 mm per decade from 1970-2000 (Hayhoe et al. 2007) and monitoring records in the Adirondacks either showed a significant decrease in snowfall from 1948-2005, or no significant change (Jenkins 2010). Changes in temperature/precipitation regimes in the Northeast have already been associated with an earlier timing of peak spring stream flows, earlier ice-out dates, and an earlier timing of first leafing and flower blooming (Hayhoe et al. 2007; Hodgkins et al. 2002) and in the Adirondacks, regional warming has been associated with a decreased ice-covered period on lakes (Beier et al. 2012; Stager et al. 2009). The effects of climate change are likely to have a profound influence on Adirondack flora and fauna. For example, Momen et al. (2006) noted an increase in productivity in seven Adirondack lakes (measured as chlorophyll a) from 1994 to 2003. Though the authors attributed this increase in productivity to chemical recovery from acidification, paleolimnological studies have attributed recent increases in chlorophyll a concentration to regional warming (Michelutti et al. 2005), a phenomenon not examined by Momen et al. (2006). Importantly, the Adirondack region represents the southernmost range limit of many boreal species, many of which may be extirpated from the region with significant warming (Jenkins 37 2010). Regional warming may result in the loss of cold-water fish taxa, rare boreal habitats (high-elevation Krummholtz and alpine-tundra areas), and winter recreation opportunities in the region (skiing, snowmobiling) (Frumhoff et al. 2007). Climate change will also affect how lake ecosystems are undergoing biological recovery from acidification and may lead to novel (and potentially undesirable) aquatic species communities in recovered systems (Arseneau et al. 2011). Summary This literature review has highlighted the complex effects acid deposition and climate change have on aquatic ecosystems. Ecologists have a variety of methods available to them to understand the influence of multiple stressors on species assemblages, one of the most intuitive being the use of reference systems. In a recent review, Gray and Arnott (2009) argued that the ideal recovery study would include the use of reference sites, historical data, and a temporal analysis. This is precisely the method followed in this investigation. First, a regional set of minimally-disturbed reference sites were identified for the Adirondack Park and critically evaluated (Chapter 3). Second, the reference sites are used in a ‘top-bottom’ paleolimnological study of scaled chrysophyte assemblages to identify if regional stressors have led to a significant shift in species assemblages since pre-1900, creating a ‘shifting baseline’ (Chapter 4). Lastly, chrysophyte assemblages are compared between a subset of reference lakes and two Adirondack lakes recovering from acidification. Paleolimnological techniques are used to identify if biological recovery from acidification is occurring and reference sites are used to define reasonable recovery endpoints for the acid-impacted systems (Chapter 5). Overall, this project will improve our understanding of biological recovery from acidification, the influence of multiple stressors on aquatic ecosystems and, perhaps most importantly, highlight the utility of 38 pairing a rigorous regional reference site approach commonly used in bioassessment studies with the valuable long-term perspective of paleolimnology. In Chapter 6, the general implications of this project are discussed, along with some possible directions for future research. Literature Cited Adrian R, O'Reilly CM, Zagarese H, Baines SB, Hessen DO, Keller W, Livingstone DM, Sommaruga R, Straile D, Van Donk E, Weyhenmeyer GA, Winder M. (2009) Lakes as sentinels of climate change. Limnology and Oceanography 54:2283-2297. Adrian R, Walz N, Hintze T, Hoeg S, Rusche R. (1999) Effects of ice duration on plankton succession during spring in a shallow polymictic lake. Freshwater Biology 41:621-634. Adrian R, Wilhelm S, Gerten D. (2006) Life-history traits of lake plankton species may govern their phenological response to climate warming. Global Change Biology 12:652-661. Adirondack Park Agency (APA). (2011a) Adirondack Park land use classificiation statistics Adirondack Park Agency. Ray Brook, New York. http://apa.ny.gov/gis/stats/colc1108.htm. Adirondack Park Agency (APA). (2011b) State of New York Adirondack Park state land master plan. Adirondack Park Agency Ray Brook, NY, 125 p. Arnott SE, Yan N, Keller W, Nicholls K. (2001) The influence of drought-induced acidification on the recovery of plankton in Swan Lake (Canada). Ecological Applications 11:747-763. Arseneau KMA, Driscoll CT, Brager LM, Ross KA, Cumming BF. (2011) Recent evidence of biological recovery from acidification in the Adirondacks (New York, USA): a multiproxy paleolimnological investigation of Big Moose Lake. Canadian Journal of Fisheries and Aquatic Sciences 68:575-592. Bailey RC, Norris RH, Reynoldson TB. (2004) Bioassessment of freshwater ecosystems: using the reference condition approach. Kluwer Academic Publishers, New York, 170 p. Baker JP, Van Sickle J, Gagen CJ, DeWalle DR, Sharpe WE, Carline RF, Baldigo BP, Murdoch PS, Bath DW, Krester WA. (1996) Episodic acidification of small streams in the northeastern United States: effects on fish populations. Ecological Applications 6:422-437. 39 Baker JP, Bernard DP, Christensen SW, Sale MJ. (1990a) Biological effects of changes in surface water acid-base chemistry. Report 13. Acidic deposition: State of science and technology. National Acid Precipitation Assessment Program (NAPAP), Washington, DC, pp. 13-1-13-381. Baker JP, Christensen SW. (1991) Effects of acidification on biological communities in aquatic ecosystems In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer-Verlag, New York, New York, pp. 83-106. Baker LA, Kaufmann PR, Brakke DF, Herlihy AT, Eilers JM. (1990b) Current status of surface-water acid-base chemistry. Report 9. Acidic deposition: state of science and technology. National Acid Precipitation Assessment Program (NAPAP), Washington, DC, pp. 9-1-9-367. Barge J. (2011) Adirondack Park population trends - 2010. Adirondack Park Agency, Ray Brook, New York. http://apa.ny.gov/gis/_assets/AdirondackParkPopulationTrends2010.pdf. Baron JS, Gunderson L, Allen CD, Fleishman E, McKenzie D, Meyerson LA, Oropeza J, Stephenson N. (2009) Options for national parks and reserves for adapting to climate change. Environmental Management 44:1033-1042. Battarbee RW, Anderson NJ, Bennion H, Simpson GL. (2012) Combining limnological and palaeolimnological data to disentangle the effects of nutrient pollution and climate change on lake ecosystems: problems and potential. Freshwater Biology 57:2091-2106. Battarbee RW, Morley D, Bennion H, Simpson GL, Hughes M, Bauere V. (2011) A palaeolimnological metadatabase for assessing the ecological status of lakes. Journal of Paleolimnology 45:405-414. Battarbee RW, Simpson GL, Shilland EM, Flower RJ, Kreiser A, Yang H, Clarke G. (2014) Recovery of UK lakes from acidification: an assessment using combined palaeoecological and contemporary diatom assemblage data. Ecological Indicators 37, Part B:365-380. Beier CM, Stella JC, Dovčiak M, McNulty SA. (2012) Local climatic drivers of changes in phenology at a borealtemperate ecotone in eastern North America. Climatic Change 115:399-417. Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. 40 Binks JA, Arnott SE, Sprules WG. (2005) Local factors and colonist dispersal influence crustacean zooplankton recovery from cultural acidification. Ecological Applications 15:2025-2036. Brucet S, Poikane S, Lyche‐Solheim A, Birk S. (2013) Biological assessment of European lakes: ecological rationale and human impacts. Freshwater Biology 58:1106-1115. Burns DA, Lynch JA, Cosby BJ, Fenn ME, Baron JS. (2011) National Acid Precipitation Assessment Program report to Congress 2011: an integrated assessment. National Science and Technology Council, Washington, DC, 114 p. Caraco N, Cole JJ, Likens GE. (1991) A cross-system study of phosphorus release from lake sediments. In: Cole J, Lovett G, Findlay S (eds), Comparative analyses of ecosystems: patterns, mechanisms, and theories. Springer New York, pp. 241-258. Chen LM, Driscoll CT. (2005) Regional assessment of the response of the acid-base status of lake watersheds in the adirondack region of New York to changes in atmospheric deposition using PnET-BGC. Environmental Science and Technology 39:787-794. Chessman BC, Muschal M, Royal MJ. (2008) Comparing apples with apples: use of limiting environmental differences to match reference and stressor-exposure sites for bioassessment of streams. River Research and Applications 24:103-117. Chestnut LG, Mills DM. (2005) A fresh look at the benefits and costs of the US acid rain program. Journal of Environmental Management 77:252-266. Christensen JH, Hewitson B, Busuioc A, Chen A, Gao X, Held I, Jones R, Kolli RK, Kwon W-T, Laprise R, Magaña Rueda V, Mearns L, Menéndez CG, Räisänen J, Rinke A, Sarr A, Whetton P. (2007) Regional climate projections. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds), Climate change 2007: The physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, pp. 848-940. Christensen MR, Graham MD, Vinebrooke RD, Findlay DL, Paterson MJ, Turner MA. (2006) Multiple anthropogenic stressors cause ecological surprises in boreal lakes. Global Change Biology 12:2316-2322. Collins M, Knutti R, Arblaster J, Dufresne J-L, Fichefet T, Friedlingstein P, Gao X, Gutowski WJ, Johns T, Krinner G, Shongwe M, Tebaldi C, Weaver AJ, Wehner M. (2013) Long-term climate change: projections, 41 commitments and irreversibility. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley GF (eds), Climate change 2013: The physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge Univerity Press, Cambridge, pp. 1029-1136. Cronan CS, Schofield CL. (1990) Relationships between aqueous aluminum and acidic deposition in forested watersheds of North America and northern Europe. Environmental Science and Technology 24:1100-1105. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Cumming BF, Smol JP, Birks HJB. (1992a) Scaled chrysophytes (Chrysophyceae and Synurophyceae) from Adirondack drainage lakes and their relationship to environmental variables. Journal of Phycology 28:162178. Cumming BF, Smol JP, Kingston JC, Charles DF, Birks HJB, Camburn KE, Dixit SS, Uutala AJ, Selle AR. (1992b) How much acidification has occurred in Adirondack region lakes (New York, USA) since preindustrial times?. Canadian Journal of Fisheries and Aquatic Sciences 49:128-141. Daufresne M, Lengfellner K, Sommer U. (2009) Global warming benefits the small in aquatic ecosystems. Proceedings of the National Academy of Sciences 106:12788-12793. De Senerpont Domis LN, Elser JJ, Gsell AS, Huszar VLM, Ibelings BW, Jeppesen E, Kosten S, Mooij WM, Roland F, Sommer U, Van Donk E, Winder M, Lürling M. (2013) Plankton dynamics under different climatic conditions in space and time. Freshwater Biology 58:463-482. De Senerpont Domis LN, Mooij WM, Hülsmann S, van Nes EH, Scheffer M. (2007) Can overwintering versus diapausing strategy in Daphnia determine match–mismatch events in zooplankton–algae interactions? Oecologia 150:682-698. De Stasio BT, Hill DK, Kleinhans JM, Nibbelink NP, Magnuson JJ. (1996) Potential effects of global climate change on small north-temperate lakes: Physics, fish, and plankton. Limnology and Oceanography 41:1136-1149. Derry AM, Arnott SE. (2007) Zooplankton community response to experimental acidification in boreal shield lakes with different ecological histories. Canadian Journal of Fisheries and Aquatic Sciences 64:887-898. 42 Deyton EB, Schwartz JS, Robinson RB, Neff KJ, Moore SE, Kulp MA. (2009) Characterizing episodic stream acidity during stormflows in the Great Smoky Mountains National Park. Water, Air, and Soil Pollution 196:3-18. Dixit SS, Dixit AS, Evans RD. (1988) Scaled chrysophytes (Chrysophyceae) as indicators of pH in Sudbury, Ontario, lakes. Canadian Journal of Fisheries and Aquatic Sciences 45:1411-1421. Dixit SS, Dixit AS, Smol JP. (1992) Assessment of changes in lake water chemistry in Sudbury area lakes since preindustrial times. Canadian Journal of Fisheries and Aquatic Sciences 49:8-16. Dixit SS, Dixit AS, Smol JP. (2002) Diatom and chrysophyte functions and inferences of post-industrial acidification and recent recovery trends in Killarney lakes (Ontario, Canada). Journal of Paleolimnology 27:79-96. Downes BJ, Barmuta LA, Fairweather PG, Faith DP, Keough MJ, Lake PS, Mapstone BD, Quinn GP. (2002) Monitoring ecological impacts: concept and practice in flowing waters. Cambridge University Press, Cambridge, 452 p. Driscoll CT. (2012) Ecological effects of acidic deposition. In: Jørgensen SE, Fath BD (eds), Encyclopedia of ecology. Elsevier, Oxford, pp. 1012-1019. Driscoll CT, Cowling EB, Grennfelt P, Galloway JM, Dennis RL. (2010) Integrated assessment of ecosystem effects of atmospheric deposition. Air and Waste Management Associations Magazine for Environmental Managers 11:6-13. Driscoll CT, Driscoll KM, Roy KM, Dukett J. (2007) Changes in the chemistry of lakes in the Adirondack region of New York following declines in acidic deposition. Applied Geochemistry 22:1181-1188. Driscoll CT, Lawrence GB, Bulger AJ, Butler TJ, Cronan CS, Eagar C, Lambert KF, Likens GE, Stoddard JL, Weathers KC. (2001) Acidic deposition in the northeastern United States: sources and inputs, ecosystem effects, and management strategies. BioScience 51:180-198. Driscoll CT, Likens GE, Church MR. (1998) Recovery of surface waters in the northeasten U.S. from decreases in atmospheric deposition of sulfur. Water, Air, and Soil Pollution 105:319-329. Driscoll CT, Newton RM, Gubala CP, Baker JP, Christensen S. (1991) Adirondack mountains. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer-Verlag, New York, pp. 133-202. 43 Duff KE, Zeeb BA, Smol J. (1995) Atlas of chrysophycean cysts. Kluwer Academic Publishers, Dordrecth, The Netherlands, 189 p. Duguay CR, Prowse TD, Bonsal BR, Brown RD, Lacroix MP, Ménard P. (2006) Recent trends in Canadian lake ice cover. Hydrological Processes 20:781-801. Durant JM, Hjermann DO, Ottersen G, Stenseth NC. (2007) Climate and the match or mismatch between predator requirements and resource availability. Climate Research 33:271-283. Eimers MC, Buttle J, Watmough SA. (2008) Influence of seasonal changes in runoff and extreme events on dissolved organic carbon trends in wetland- and upland-draining streams. Canadian Journal of Fisheries and Aquatic Sciences 65:796-808. Eimers MC, Watmough SA, Paterson AM, Dillon PJ, Yao HX. (2009) Long-term declines in phosphorus export from forested catchments in south-central Ontario. Canadian Journal of Fisheries and Aquatic Sciences 66:1682-1692. Erlandsson M, Buffam I, Fölster J, Laudon H, Temnerud J, Weyhenmeyer GA, Bishop K. (2008) Thirty-five years of synchrony in the organic matter concentrations of Swedish rivers explained by variation in flow and sulphate. Global Change Biology 14:1191-1198. Evans CD, Chapman PJ, Clark JM, Monteith DT, Cresser MS. (2006) Alternative explanations for rising dissolved organic carbon export from organic soils. Global Change Biology 12:2044-2053. Evans CD, Monteith DT, Cooper DM. (2005) Long-term increases in surface water dissolved organic carbon: Observations, possible causes and environmental impacts. Environmental Pollution 137:55-71. Fang X, Stefan HG. (2009) Simulations of climate effects on water temperature, dissolved oxygen, and ice and snow covers in lakes of the contiguous United States under past and future climate scenarios. Limnology and Oceanography 54:2359-2370. Faulkenham SE, Hall RI, Dillon PJ, Karst-Riddoch T. (2003) Effect of drought-induced acidification on diatom communities in acid-sensitive Ontario lakes. Limnology and Oceanography 48:1662-1673. Fee EJ. (1976) Vertical and seasonal distribution of chlorophyll in lakes of Experimental Lakes Area, Northwestern Ontario - implications for primary production estimates. Limnology and Oceanography 21:767-783. Fee EJ, Hecky RE, Kasian SEM, Cruikshank DR. (1996) Effects of lake size, water clarity, and climatic variability on mixing depths in Canadian Shield lakes. Limnology and Oceanography 41:912-920. 44 Ficke AD, Myrick CA, Hansen LJ. (2007) Potential impacts of global climate change on freshwater fisheries. Reviews in Fish Biology and Fisheries 17:581-613. Field CB, Mortsch LD, Brklacich M, Forbes DL, Kovacs P, Patz JA, Running SW, Scott MJ. (2007) North America. In: Parry ML, Canziani OF, Palutikof JP, van der Linden PJ, Hanson CE (eds), Climate change 2007: Impacts, adaptation and vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambrige, UK, pp. 617-652. Flear K. (2011) Changes in scaled-chrysophyte assemblages in response to recent climate change in Northwestern Ontario. Department of Biology. Queen's University Kingston, Ontario, 121 p. Folt CL, Chen CY, Moore MV, Burnaford J. (1999) Synergism and antagonism among multiple stressors. Limnology and Oceanography 44:864-877. Forsström L, Sorvari S, Korhola A, Rautio M. (2005) Seasonality of phytoplankton in subarctic Lake Saanajärvi in NW Finnish Lapland. Polar Biology 28:846-861. Frumhoff PC, McCarthy JJ, Melillo JM, Moser SC, Wuebbles DJ. (2007) Confronting climate change in the U.S. Northeast: science, impacts, and solutions. Synthesis report of the Norhteast Climate Impacts Assessment (NECIA). Union of Concerned Scientists, Cambridge, MA, 146 p. Fuss CB, Driscoll CT, Johnson CE, Petras RJ, Fahey TJ. (2010) Dynamics of oxidized and reduced iron in a northern hardwood forest. Biogeochemistry 104:103-119. Futter M. (2003) Patterns and Trends in Southern Ontario Lake ice Phenology. Environmental Monitoring and Assessment 88:431-444. Gallagher J, Baker J. (1990) Current status of fish communities in Adirondack Lakes. Adirondack Lakes Survey: An interpretive analysis of fish communities and water chemistry, 1984-1987. Adirondack Lakes Survey Corporation, Ray Brook, New York, pp. 3-1-13-44. Garmo ØA, Skjelkvåle BL, de Wit HA, Colombo L, Curtis C, Fölster J, Hoffmann A, Hruška J, Høgåsen T, Jeffries DS, Keller WB, Krám P, Majer V, Monteith DT, Paterson AM, Rogora M, Rzychon D, Steingruber S, Stoddard JL, Vuorenmaa J, Worsztynowicz A. (2014) Trends in surface water chemistry in acidified areas in Europe and North America from 1990 to 2008. Water, Air, and Soil Pollution 225:1880-1-1880-14. George DG. (2010) The impact of climate change on European lakes. Springer, Dordrecht, The Netherlands, 507 p. 45 Ginn BK, Rate M, Cumming BF, Smol JP. (2010) Ecological distribution of scaled-chrysophyte assemblages from the sediments of 54 lakes in Nova Scotia and southern New Brunswick, Canada. Journal of Paleolimnology 43:293-308. Goldsten RA, Gherini SA, Chen CW, Mok L, Hudson RJM, Kramer JR, Brown DJA, Chester DF. (1984) Integrated acidification study (ILWAS): A mechanistic ecosystem analysis. Philosophical Transactions of the Royal Society of London B, Biological Sciences 305:409-425. Gray DK, Arnott SE. (2009) Recovery of acid damaged zooplankton communities: measurement, extent, and limiting factors. Environmental Reviews 17:81-99. Greenaway CM, Paterson AM, Keller W, Smol JP. (2012a) Dramatic diatom species assemblage responses in lakes recovering from acidification and metal contamination near Wawa, Ontario, Canada: a paleolimnological perspective. Canadian Journal of Fisheries and Aquatic Sciences 69:656-669. Greenaway CM, Paterson AM, Keller WB, Smol JP. (2012b) Scaled-chrysophyte assemblage changes in the sediment records of lakes recovering from marked acidification and metal contamination near Wawa, Ontario, Canada. Journal of Limnology 71:267-278. Hamilton AT, Barbour MT, Bierwagen BG. (2010) Implications of global change for the maintenance of water quality and ecological integrity in the context of current water laws and environmental policies. Hydrobiologia 657:263-278. Hartmann DL, Klein Tank AMG, Rusticucci M, Alexander LV, Brönnimann S, Charabi Y, Dentener FJ, Dlugokencky EJ, Easterling DR, Kaplan A, Soden BJ, Thorne PW, Wild M, Zhai PM. (2013) Observations: atmosphere and surface. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds), Climate change 2013: The physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climat Change Cambridge University Press, Cambridge, pp. 159-254. Havas M, Woodfine DG, Lutz P, Yung K, MacIsaac HJ, Hutchinson TC. (1995) Biological recovery of two previously acidified, metal-contaminated lakes near Sudbury Ontario, Canada. Water, Air, and Soil Pollution 85:791-796. Havens KE, Carlson RE. (1998) Functional complementarity in plankton communities along a gradient of acid stress. Environmental Pollution 101:427-436. 46 Hawkins CP, Olson JR, Hill RA. (2010) The reference condition: predicting benchmarks for ecological and waterquality assessments. Journal of the North American Benthological Society 29:312-343. Hawryshyn J, Rühland KM, Quinlan R, Smol JP. (2012) Long-term water quality changes in a multiple-stressor system: a diatom-based paleolimnological study of Lake Simcoe (Ontario, Canada). Canadian Journal of Fisheries and Aquatic Sciences 69:24-40. Hayhoe K, Wake CP, Huntington TG, Luo L, Schwartz MD, Sheffield J, Wood E, Anderson B, Bradbury J, DeGaetano A, Troy TJ, Wolfe D. (2007) Past and future changes in climate and hydrological indicators in the US Northeast. Climate Dynamics 28:381-407. Healey FP. (1983) Effect of temperature and light intensity on the growth rate of Synura sphagnicola. Journal of Plankton Research 5:767-774. Helliwell RC, Simpson GL. (2010) The present is the key to the past, but what does the future hold for the recovery of surface waters from acidification? Water Research 44:3166-3180. Hobbs RJ, Higgs E, Harris JA. (2009) Novel ecosystems: implications for conservation and restoration. Trends in Ecology & Evolution 24:599-605. Hodgkins GA, James IC, Huntington TG. (2002) Historical changes in lake ice-out dates as indicators of climate change in New England, 1850–2000. International Journal of Climatology 22:1819-1827. Holen DA, Boraas ME. (1995) Mixotrophy in chrysophytes. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: ecology, phylogeny and development. Cambridge University Press, Cambridge, United Kingdom, pp. 119-140. Husar RB, Sullivan TJ, Charles DF. (1991) Historical trends in atmospheric sulfur deposition and methods for assessing long-term trends in surface water chemistry. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer New York, pp. 65-82. Hyatt CV, Paterson AM, Cumming BF, Smol JP. (2010) Factors related to regional and temporal variation in the distribution of scaled chrysophytes in north-eastern North America: Evidence from lake sediments. Nova Hedwigia, Beiheft 136:87–102. Hynynen J, Meriläinen JJ. (2005) Recovery from acidification in boreal lakes inferred from macroinvertebrates and subfossil chironomids. Hydrobiologia 541:155-173. 47 Ilyashuk B, Ilyashuk E, Dauvalter V. (2003) Chironomid responses to long-term metal contamination: a paleolimnological study in two bays of Lake Imandra, Kola Peninsula, northern Russia. Journal of Paleolimnology 30:217-230. Inamdar S, Rupp J, Mitchell M. (2009) Groundwater flushing of solutes at wetland and hillslope positions during storm events in a small glaciated catchment in western New York, USA. Hydrological Processes 23:19121926. Ito M, Mitchell MJ, Driscoll CT. (2002) Spatial patterns of precipitation quantity and chemistry and air temperature in the Adirondack region of New York. Atmospheric Environment 36:1051-1062. Jankowski T, Livingstone DM, Bührer H, Forster R, Niederhauser P. (2006) Consequences of the 2003 European heat wave for lake temperature profiles, thermal stability, and hypolimnetic oxygen depletion: Implications for a warmer world. Limnology and Oceanography 51:815-819. Jenkins J. (2010) Climate change in the Adirondacks: the path to sustainability. Cornell University Press, Ithaca, New York, 183 p. Jenkins J, Keal A. (2004) The Adirondack atlas: a geographic portrait of the Adirondack Park. Syracuse University Press, Syracuse, New York, 296 p. Jenkins JC, Roy K, Driscoll C, Buerkett C. (2007) Acid rain in the Adirondacks: an environmental history. Cornell University Press, Ithaca, New York, 246 p. Jeppesen E, Kronvang B, Meerhoff M, Søndergaard M, Hansen KM, Andersen HE, Lauridsen TL, Liboriussen L, Beklioglu M, Özen A, Olesen JE. (2009) Climate change effects on runoff, catchment phosphorus loading and lake ecological state, and potential adaptations. Journal of Environmental Quality 38:1930-1941. Jeziorski A, Keller B, Paterson A, Greenaway C, Smol J. (2013) Aquatic ecosystem responses to rapid recovery from extreme acidification and metal contamination in lakes near Wawa, Ontario. Ecosystems 16:209-223. Keatley BE, Douglas MSV, Smol JP. (2008) Prolonged ice cover dampens diatom community responses to recent climatic change in high Arctic lakes. Arctic, Antarctic, and Alpine Research 40:364-372. Keller W. (2009) Limnology in northeastern Ontario: from acidification to multiple stressors. Canadian Journal of Fisheries and Aquatic Sciences 66:1189-1198. Keller W, Yan ND. (1998) Biological recovery from lake acidification: zooplankton communities as a model of patterns and processes. Restoration Ecology 6:364-375. 48 Kirillin G, Leppäranta M, Terzhevik A, Granin N, Bernhardt J, Engelhardt C, Efremova T, Golosov S, Palshin N, Sherstyankin P, Zdorovennova G, Zdorovennov R. (2012) Physics of seasonally ice-covered lakes: a review. Aquatic Sciences 74:659-682. Kopáček J, Stuchlĺk E, Veselý J, Schaumburg J, Anderson IA, Fott J, Hejzlar J, Vrba J. (2002) Hysteresis in reversal of central European mountain lakes from atmospheric acidification. Water, Air, & Soil Pollution: Focus 2:91-114. Korosi JB, Paterson AM, DeSellas AM, Smol JP. (2010) A comparison of pre-industrial and present-day changes in Bosmina and Daphnia size structure from soft-water Ontario lakes. Canadian Journal of Fisheries and Aquatic Sciences 67:754-762. Krester W, Gallagher J, Nicolette J. (1989) Adirondack lakes survey 1984-1987: an evaluation of fish communities and water chemistry. Adirondack Lakes Survey Corporation, Ray Brook, NY, 437 p. Kristiansen J. (1988) Seasonal occurrence of silica-scaled chrysophytes under eutrophic conditions. Hydrobiologia 161:171-184. Kristiansen J. (2005) Golden algae: a biology of chrysophytes. A.R.G. Gantner Verlag, Königstein, Germany, 167 p. Kristiansen J. (2008) Dispersal and biogeography of silica-scaled chrysophytes. Biodiversity and Conservation 17:419-426. Kullberg A, Bishop KH, Hargeby A, Jansson M. (1993) The ecological significance of dissolved organic carbon in acidified waters. AMBIO 22:331-337. Lawrence GB, Dukett JE, Houck N, Snyder P, Capone S. (2013) Increases in dissolved organic carbon accelerate loss of toxic Al in Adirondack lakes recovering from acidification. Environmental Science and Technology 47:7095-7100. Likens GE, Driscoll CT, Buso DC. (1996) Long-term effects of acid rain: response and recovery of a forest ecosystem. Science 272:244-245. Livingstone DM, Adrian R, Blenckner T, George G, Weyhenmeyer GA. (2010) Lake ice phenology. In: George DG (ed), The impact of climate change on European lakes, Springer, Dordrecht, The Netherlands, pp. 51-61. Locke A. (1991) Zooplankton responses to acidification: A review of laboratory bioassays. Water, Air, and Soil Pollution 60:135-148. 49 Magnuson JJ, Robertson DM, Benson BJ, Wynne RH, Livingstone DM, Arai T, Assel RA, Barry RG, Card V, Kuusisto E, Granin NG, Prowse TD, Stewart KM, Vuglinski VS. (2000) Historical trends in lake and river ice cover in the Northern Hemisphere. Science 289:1743-1746. Marcogliese DJ. (2001) Implications of climate change for parasitism of animals in the aquatic environment. Canadian Journal of Zoology 79:1331-1352. Masson-Delmotte V, Schulz M, Abe-Ouchi A, Beer J, Ganopolski A, González Rouco JF, Jansen E, Lambeck K, Luterbacher J, Naish T, Osborn T, Otto-Bliesner B, Quinn T, Ramesh R, Rojas M, Shao X, Timmermann A. (2013) Information from paleoclimate archives. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds), Climate change 2013: The physcial science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, pp. 383-464. Mattsson T, Kortelainen P, Laubel A, Evans D, Pujo-Pay M, Räike A, Conan P. (2009) Export of dissolved organic matter in relation to land use along a European climatic gradient. Science of The Total Environment 407:1967-1976. Michelutti N, Smol JP. (2013) Multiproxy approaches. In: Elias SA (ed), The Encyclopedia of Quarternary Science. Elsevier, Amsterdam, The Netherlands, pp. 339-348. Michelutti N, Wolfe AP, Vinebrooke RD, Rivard B, Briner JP. (2005) Recent primary production increases in arctic lakes. Geophysical Research Letters 32: L19715.1-L19715.4. Momen B, Lawrence GB, Nierzwicki-Bauer SA, Sutherland JW, Eichler LW, Harrison JP, Boylen CW. (2006) Trends in summer chemistry linked to productivity in lakes recovering from acid deposition in the Adirondack region of New York. Ecosystems 9:1306-1317. Monteith DT, Stoddard JL, Evans CD, de Wit HA, Forsius M, Høgasen T, Wilander A, Skjelkvåle BL, Jeffries DS, Vuorenmaa J, Keller B, Kopácek J, Vesely J. (2007) Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature 450:537-540. Moore MV, Folt CL, Stemberger RS. (1996) Consequences of elevated temperatures for zooplankton assemblages in temperate lakes. Archiv für Hydrobiologie 135:289-319. Moss B. (2012) Cogs in the endless machine: lakes, climate change and nutrient cycles: a review. Science of The Total Environment 434:130-142. 50 Munson RK, Gherini SA. (1991) Processes influencing the acid-base chemistry of surface waters. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer New York, pp. 9-34. Newton RM, Driscoll CT. (1990) Classificiation of ALSC lakes. In: Adirondack Lakes Survey: An interpretive analysis of fish communities and water chemistry, 1984-1987. Adirondack Lakes Survey Corporation, Ray Brook, New York, pp. 2-70-2-91. Nicholls KH. (1995) Chrysophyte blooms in the plankton and neuston of marine and freshwater systems. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: ecology, phylogeny and development. Cambridge University Press, Cambridge, UK, pp. 181-213. Nicholls KH, Gerrath JF. (1985) The taxonomy of Synura (Chrysophyceae) in Ontario with special reference to taste and odor in water-supplies. Canadian Journal of Botany 63:1482-1493. Nilssen JP, Wærvågen SB. (2002) Intensive fish predation: an obstacle to biological recovery following liming of acidified lakes? Journal of Aquatic Ecosystem Stress and Recovery 9:73-84. Olrik K. (1994) Phytoplankton - ecology: determining factors for the distribution of phytoplankton in freshwater and the sea. Miljøbiologisk Laboratorium ApS, Copenhagen, 183 p. Olsson HP, Pettersson A. (1993) Oligotrophication of acidified lakes - a review of hypotheses. AMBIO 22:312-317. Paine RT, Tegner MJ, Johnson EA. (1998) Compounded perturbations yield ecological surprises. Ecosystems 1:535-545. Paterson AM, Cumming BF, Smol JP, Blais JM, France RL. (1998) Assessment of the effects of logging, forest fires and drought on lakes in northwestern Ontario: a 30-year paleolimnological perspective. Canadian Journal of Forest Research 28:1546-1556. Paterson AM, Cumming BF, Smol JP, Hall RI. (2001) Scaled chrysophytes as indicators of water quality changes since preindustrial times in the Muskoka-Haliburton region, Ontario, Canada. Canadian Journal of Fisheries and Aquatic Sciences 58:2468-2481. Paterson AM, Cumming BF, Smol JP, Hall RI. (2004) Marked recent increases of colonial scaled chrysophytes in boreal lakes: implications for the management of taste and odour events. Freshwater Biology 49:199-207. Persson G. (2008) Zooplankton response to long-term liming: comparison of 15 limed and 15 reference lakes in Sweden. Limnologica 38:1-13. Pettersson K. (1998) Mechanisms for internal loading of phosphorus in lakes. Hydrobiologia 373/374:21-25. 51 Pla-Rabes S, Catalan J. (2011) Deciphering chrysophyte responses to climate seasonality. Journal of Paleolimnology 46:139-150. Preston ND, Rusak JA. (2010) Homage to Hutchinson: Does inter-annual climate variability affect zooplankton density and diversity? Hydrobiologia 653:165-177. Quinlan R, Hall RI, Paterson AM, Cumming BF, Smol JP. (2008) Long-term assessments of ecological effects of anthropogenic stressors on aquatic ecosystems from paleoecological analyses: challenges to perspectives of lake management. Canadian Journal of Fisheries and Aquatic Sciences 65:933-944. Reynolds CS. (1988) Functional morphology and the adapive strategies of freshwater plankton. In: Sandgren CD (ed), Growth and reproductive strateiges of freshwater plankton. Cambridge University Press, Cambridge, pp. 388-433. Reynolds CS, Huszar V, Kruk C, Naselli-Flores L, Melo S. (2002) Towards a functional classification of the freshwater phytoplankton. Journal of Plankton Research 24:417-428. Reynoldson TB, Norris RH, Resh VH, Day KE, Rosenberg DM. (1997) The reference condition: a comparison of multimetric and multivariate approaches to assess water-quality impairment using benthic macroinvertebrates. Journal of the North American Benthological Society 16:833-852. Rühland K, Paterson AM, Smol JP. (2008) Hemispheric-scale patterns of climate-related shifts in planktonic diatoms from North American and European lakes. Global Change Biology 14:2740-2754. Saha LC, Wujek DE. (1990) Scale-bearing chrysophytes from tropical Northeast India. Nordic Journal of Botany 10:343-355. Sandgren CD. (1988) The ecology of chrysophyte flagellates: their growth and perennation strategies as freshwater phytoplankton. In: Sandgren CD (ed), Growth and reproductive strategies of freshwater phytoplankton. Cambridge Univeristy Press, Cambridge, United Kingdom, pp. 9-104. Sandgren CDW, William E. (1995) The influence of zooplankton herbvory on the biogeography of chrysophyte algae. In: Sandgren CDS, John P. , Kristiansen J (eds), Chrysophyte algae: ecology, phylogeny and development. Cambridge University Press Cambridge, UK, pp. 269-302. Schindler DW. (1988) Effects of acid rain on freshwater ecosystems. Science 239:149-157. 52 Schindler DW, Bayley SE, Parker BR, Beaty K, Cruikshank D, Fee E, Schindler E, Stainton M. (1996a) The effects of climatic warming on the properties of boreal lakes and streams at the Experimental Lakes Area, northwestern Ontario. Limnology and Oceanography 41:1004-1017. Schindler DW, Curtis PJ, Parker BR, Stainton MP. (1996b) Consequences of climate warming and lake acidification for UV-B penetration in North American boreal lakes. Nature 379:705-708. Shatwell T, Köhler J, Nicklisch A. (2008) Warming promotes cold-adapted phytoplankton in temperate lakes and opens a loophole for Oscillatoriales in spring. Global Change Biology 14:2194-2200. Siegfried CA, Bloomfield JA, Sutherland JW. (1989) Acidity status and phytoplankton species richness, standing crop, and community composition in Adirondack, New York, U.S.A. lakes. Hydrobiologia 175:13-32. Siegfried CA, Sutherland JW. (1992) Zooplankton communities of Adirondack lakes: changes in community structure associated with acidification. Journal of Freshwater Ecology 7:97-112. Simpson GL, Anderson NJ. (2009) Deciphering the effect of climate change and separating the influence of confounding factors in sediment core records using additive models. Limnology and Oceanography 54:2529-2541. Siver P, Marsicano L. (1996) Inferring lake trophic status using scaled chrysophytes. Nova Hedwigia Beiheft 114:233-246. Siver PA. (1995) The distribution of chrysophytes along environmental gradients: their use as biological indicators. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: Ecology, phylogeny and development. Cambridge University Press, Cambridge, UK, pp. 232-268. Siver PA. (2003) Synurophyte algae. In: Wehr JD (ed), Freshwater algae of North America: ecology and classification. Academic Press, San Diego, California, pp. 523-558. Siver PA, Hamer JS. (1989) Multivariate statistical analysis of the factors controlling the distribution of scaled chrysophytes. Limnology and Oceanography 34:368-381. Siver PA, Hamer JS. (1992) Seasonal periodicity of chrysophyceae and synurophyceae in a small New England lake: implications for paleolimnological research. Journal of Phycology 28:186-198. Siver PA, Lott AM. (2012) Biogeographic patterns in scaled chrysophytes from the east coast of North America. Freshwater Biology 57:451-466. 53 Skjelkvåle BL, Stoddard JL, Jeffries DS, Tørseth K, Høgasen T, Bowman J, Mannio J, Monteith DT, Mosello R, Rogora M, Rzychon D, Vesely J, Wieting J, Wilander A, Worsztynowicz A. (2005) Regional scale evidence for improvements in surface water chemistry 1990–2001. Environmental Pollution 137:165-176. Smol JP. (1995) Application of chrysophytes to problems of paleoecology. In: Sandgren CDS, John P., Kristiansen J (eds), Chrysophyte algae: Ecology, phylogeny and development. Cambridge University Press Cambridge, UK, pp. 303-329. Snucins E, Gunn J. (2000) Interannual variation in the thermal structure of clear and colored lakes. Limnology and Oceanography 45:1639-1646. Sommer U. (1988) Some size relationships in phytoflagellate motility. Hydrobiologia 161:125-131. Sommer U, Adrian R, De Senerpont Domis L, Elser JJ, Gaedke U, Ibelings B, Jeppesen E, Lürling M, Molinero JC, Mooij WM. (2012) Beyond the Plankton Ecology Group (PEG) model: mechanisms driving plankton succession. Annual Review of Ecology, Evolution, and Systematics 43:429-448. Soranno PA, Wagner T, Martin SL, McLean C, Novitski LN, Provence CD, Rober AR. (2011) Quantifying regional reference conditions for freshwater ecosystem management: a comparison of approaches and future research needs. Lake and Reservoir Management 27:138-148. Stager JC, McNulty S, Beier C, Chiarenzelli J. (2009) Historical patterns and effects of changes in Adirondack climates since the early 20th century. Adirondack Journal of Environmental Studies 15:14-24. Stager JC, Thill M. (2010) Climate change in the Champlain basin: what natural resource managemers can expect and do. The Nature Conservancy, 38 p. Stoddard JL, Larsen DP, Hawkins CP, Johnson RK, Norris RH. (2006) Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16:1267-1276. Strayer DL, Dudgeon D. (2010) Freshwater biodiversity conservation: recent progress and future challenges. Journal of the North American Benthological Society 29:344-358. Sullivan TJ, Cosby BJ, Driscoll CT, McDonnell TC, Herlihy AT, Burns DA. (2012) Target loads of atmospheric sulfur and nitrogen deposition for protection of acid sensitive aquatic resources in the Adirondack Mountains, New York. Water Resources Research 48:W01547-1-W01547-16. 54 Sullivan TJ, Cosby BJ, Jackson WA, Snyder KU, Herlihy AT. (2011) Acidification and prognosis for future recovery of acid-sensitive streams in the southern blue ridge province. Water, Air, & Soil Pollution 219:1126. Sweetman JN, LaFace E, Rühland KM, Smol JP. (2008) Evaluating the response of cladocera to recent environmental changes in lakes from the Central Canadian Arctic treeline region. Arctic, Antarctic, and Alpine Research 40:584-591. Taranu ZE, Koster D, Hall RI, Charette T, Forrest F, Cwynar LC, Gregory-Eaves I. (2010) Contrasting responses of dimictic and polymictic lakes to environmental change: a spatial and temporal study. Aquatic Sciences 72:97-115. United States Environmental Protection Agency (USEPA). (2012) Implications of climate change for state bioassessment programs and approaches to account for effects (final report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-11/036F. Vincent WF. (2009) Effects of climate change on lakes. In: Likens GE (ed), Encyclopedia of inland waters. Elsevier, Oxford, United Kingdom, pp. 55-60. Vinebrooke RD, Dixit SS, Graham MD, Gunn JM, Chen Y-W, Belzile N. (2002) Whole-lake algal responses to a century of acidic industrial deposition on the Canadian Shield. Canadian Journal of Fisheries and Aquatic Sciences 59:483-493. Waller K, Driscoll C, Lynch J, Newcomb D, Roy K. (2012) Long-term recovery of lakes in the Adirondack region of New York to decreases in acidic deposition. Atmospheric Environment 46:56-64. Walther G-R. (2010) Community and ecosystem responses to recent climate change. Philosophical Transactions of the Royal Society B, Biological Sciences 365:2019-2024. Watson SB, McCauley E, Downing JA. (1997) Patterns in phytoplankton taxonomic composition across temperate lakes of differing nutrient status. Limnology and Oceanography 42:487-495. Watson SB, Satchwill T, Dixon E, McCauley E. (2001) Under-ice blooms and source-water odour in a nutrient-poor reservoir: biological, ecological and applied perspectives. Freshwater Biology 46:1553-1567. Wetzel RG. (2001) Limnology: Lake and river ecosystems Academic Press, New York, 1006 p. Weyhenmeyer GA, Blenckner T, Petterson K. (1999) Changes of the plankton spring outburst related to the North Atlantic Oscillation. Limnology and Oceanography 44:1788-1792. 55 Weyhenmeyer GA, Karlsson J. (2009) Nonlinear response of dissolved organic carbon concentrations in boreal lakes to increasing temperatures. Limnology and Oceanography 54:2513-2519. Weyhenmeyer GA, Hanner P, Willén E. (2013) Shifts in phytoplankton species richness and biomass along a latitudinal gradient – consequences for relationships between biodiversity and ecosystem functioning. Freshwater Biology 58:612-623. Wigington PJ, DeWalle DR, Murdoch PS, Kretser WA, Simonin HA, Van Sickle J, Baker JP. (1996) Episodic acidification of small streams in the northeastern United States: ionic controls of episodes. Ecological Applications 6:389-407. Wilkinson AN, Zeeb BA, Smol JP. (2002) Atlas of chrysophycean cysts. Kluwer Academic Publishers, Dordrecht, The Netherlands, 169 p. Winder M, Schindler DE. (2004) Climate change uncouples trophic interactions in an aquatic ecosystem. Ecology 85:2100-2106. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. Wolfe BB, Hall RI, Edwards TWD, Johnston JW. (2012) Developing temporal hydroecological perspectives to inform stewardship of a northern floodplain landscape subject to multiple stressors: paleolimnological investigation of the Peace-Athabasca Delta. Environmental Reviews 20:191-210. Wujek DE, Wee JL. (1983) Chrysodidymus in the United States. Transactions of the American Microscopical Society 102:77-80. Yan ND, Girard R, Heneberry JH, Keller WB, Gunn JM, Dillon PJ. (2004) Recovery of copepod, but not cladoceran, zooplankton from severe and chronic effects of multiple stressors. Ecology Letters 7:452-460. Yan ND, Keller W, Somers KM, Pawson TW, Girard RE. (1996) Recovery of crustacean zooplankton communities from acid and metal contamination: comparing manipulated and reference lakes. Canadian Journal of Fisheries and Aquatic Sciences 53:1301-1327. Zeeb BA, Smol JP. (1991) Paleolimnological investigation of the effects of road salt seepage on scaled chrysophytes in Fonda Lake, Michigan. Journal of Paleolimnology 5:263-266. Zhang J, Hudson J, Neal R, Sereda J, Clair T, Turner M, Jeffries D, Dillon P, Molot L, Somers K, Hesslein R. (2010) Long-term patterns of dissolved organic carbon in lakes across eastern Canada: evidence of a pronounced climate effect. Limnology and Oceanography 55:30-42. 56 CHAPTER 3 TRACKING SHIFTING BASELINE CONDITIONS DUE TO CLIMATE CHANGE IN LAKES: A NOVEL USE OF REGIONAL REFERENCE SITES AND PALEOLIMNOLOGY Abstract Climate warming may prevent lakes recovering from acidification or eutrophication from returning to their pre-disturbance state, creating the ‘shifting baseline’ phenomenon. To define recovery targets in a multiple stressor environment, researchers must: 1) have knowledge of predisturbance conditions; and 2) understand how stressors/recovery processes influence species assemblages, a difficult task as long-term monitoring data are rare. Paleolimnological studies of minimally-disturbed reference sites can address both problems. Paleolimnological studies use the physical, chemical, and biological properties of sediments to infer how aquatic environments have changed over time. By reconstructing past biotic assemblages of reference lakes, we can infer if/how climate warming is influencing species assemblages, which can help characterize the shifting baseline and define recovery targets. In this manuscript, we present a framework for using paleolimnological techniques in a regional reference site approach and apply it in the Adirondack region (NY, USA). Simple selection criteria were used to identify reference lakes that were relatively unimpacted by acidification, eutrophication, road salt seepage, and nonnative piscivore introductions (31 out of 1469 lakes). Ordination methods found that the reference lakes were representative of 24-36% of the chemical/physical characteristics of Adirondack lakes. GIS methods, historic maps, and historic fisheries information were used to quantify historic disturbance and separate the reference lakes into classes. Only 9 lakes experienced little or no historic watershed disturbance, highlighting the danger of assuming a site’s condition has been static over time. As the effects of climate change become more 57 prevalent, the historical perspective offered by paleolimnological studies will become increasingly important for management. Introduction Limits on sulphur dioxide emissions implemented via ‘clean air’ legislation in North America and Europe have led to significant declines in acid deposition. In some regions, waterbodies have begun to undergo chemical recovery from acidification, with long-term increases acid neutralizing capacity (ANC) and decreases in H+ (Garmo et al. 2014; Skjelkvåle et al. 2005). Given this chemical recovery, there is now an acute research need to characterize biological recovery from acidification. However, researchers tasked with identifying biological recovery are faced with two major problems: 1) a lack of long-term monitoring data; and 2) the difficulty of assessing recovery in systems influenced by multiple stressors. Paleolimnological methods can be used to identify biological recovery when there is a lack of long-term monitoring data. Paleolimnological studies use the chemical, physical, and biological properties of lake sediments to infer the environmental histories of lakes (Smol 2008). Commonly, researchers will identify the remains of aquatic organisms in radiometrically-dated sediments and reconstruct how the species assemblages have responded to environmental change over time. This approach has been used widely to identify the onset of acidification (Battarbee 1990; Cumming et al. 1994) and, more recently, biological recovery from acidification (Arseneau et al. 2011; Hesthagen et al. 2011). Interpretations of both paleolimnological studies and studies that rely solely on modern limnological sampling are difficult when systems are impacted by multiple anthropogenic stressors (e.g. acidification, eutrophication, landscape alteration, climate warming), an increasingly common problem (Keller 2009). Researchers working on lakes that are influenced 58 by multiple stressors may face challenges in attributing the response of species or assemblages to a single stressor. Climate change is of particular concern in this regard, given its regional nature. Climate warming may result in particular “syndromes” of species change in lakes, such as an increase in the abundance of warm-water taxa, small planktonic algae, or flagellated/buoyant algae as water temperature increases and/or longer stratification periods occur with regional warming (Winder and Sommer 2012). In such a scenario, how are researchers to distinguish between a species’ response to a novel stressor such as climate change and, for example, the recovery of algal or invertebrate assemblages from acid deposition? Furthermore, climate change acting either alone or in conjunction with other stressors may prevent recovering systems from returning to their pre-disturbance state, even with remediative action, creating a phenomenon known as the “shifting baseline” (Bennion et al. 2011). When the baseline of an undisturbed lake changes, it becomes increasingly difficult to define a reasonable recovery target for an impacted site. The goal of this manuscript is to provide a framework for using paleolimnological techniques in regional studies of reference lakes (sites) to track shifting baseline conditions and define recovery targets for impacted sites. In a recent report on environmental assessment practices in the United States, the USEPA highlighted the problem that climate change poses to bioassessment programs and called for long-term monitoring of regional reference sites (i.e. sites minimally disturbed by anthropogenic stressors) so that they could serve as ‘sentinels’ of climate warming and track shifting baseline conditions (USEPA 2012). However, long-term monitoring data of aquatic ecosystems, where they exist, typically cover short time periods (< 50 years) and may not be suited for studies of climate change. Here, paleolimnological studies can play an important role. Using minimally-disturbed reference sites in paleolimnological studies can allow 59 researchers to identify a species’ response to climate change independent of other stressors and identify if a ‘shifting baseline’ effect has occurred. An understanding of the shifting baseline can, in turn, provide the necessary information to define recovery targets for impacted systems. Unfortunately, most environmental assessment studies lack the historical perspective offered by paleolimnological work (Stoddard et al. 2006). Similarly, while paleolimnological studies of environmental stressors often include a small number of reference sites, the regionally-based reference site approach that is often used in bioassessment studies (e.g. Bailey et al. 2004) has been less commonly applied in paleolimnological investigations (though for an early example, see Battarbee et al. 2002). As such, in this manuscript we will: 1) provide a framework for pairing paleolimnological techniques with a regional reference site approach and; 2) describe a case study in the Adirondack Park (NY, USA) where a set of 31 minimallydisturbed reference sites were identified and assessed for use in a regional paleolimnological study of recent climate warming and biological recovery from acidification. The methods used in this investigation could be easily modified and applied in other protected areas to fulfill a critical research and management objective (Baron et al. 2009). To that end, we will elaborate on the rationale behind the proposed framework and the methods used, while future manuscripts will focus on the results of our paleolimnological investigations of changes in algal and invertebrate assemblages in these reference systems. Methods A conceptual framework for using paleolimnological techniques in regional reference site studies 60 In the following section we outline some key concepts that should be considered when designing a paleolimnological study of biological recovery that utilizes regional reference sites. These concepts include: what are ‘good’ reference sites and how are they identified; the useful role of reference sites in studies of multiple stressors and in identifying reasonable recovery targets for impacted systems; and why paleo-studies should be regional in nature, rather than relying on a small number of reference sites. Subsequently, we present a case study that identified and assessed a group of reference lakes in the Adirondack region (NY, USA). Reference sites in limnological vs. paleolimnological studies – a distinction: Before elaborating on our framework, it is useful to distinguish between the use of reference sites in studies that rely solely on modern limnological sampling and those that use paleolimnological techniques. In limnological studies, the regional reference site approach (or reference condition approach, sensu Bailey et al. 2004) involves the identification of sites minimally-disturbed by anthropogenic stressors (reference sites) whose present-day ecological condition is used to define recovery targets for impacted sites, often by using modelling techniques (i.e., what species would be present at this site if it were in reference condition?). The modelling approach may be necessary because the pre-disturbance condition of an impacted site is not usually known. In contrast, paleolimnological studies can reconstruct the pre-disturbance or historic species assemblages (e.g., of diatoms, chrysophytes, invertebrates) of an impacted site. The purpose of including reference sites in paleolimnological studies is to track regional changes in species composition over time which are occurring independent of a stressor of interest (e.g., acidification or eutrophication); information which, in turn, can be used to define recovery targets for impacted sites. 61 What makes a ‘good’ reference site in paleolimnological studies?: Given the pervasive influence of environmental stressors such as acid deposition or climate change, it is unlikely that researchers will be able to identify ‘pristine’ reference sites (i.e., sites that have never experienced anthropogenic influence, Bennion et al. 2011). Nevertheless, the reference sites should possess a high degree of biotic integrity (or, ‘naturalness’), in that they have been largely protected from human impact over the time period being investigated (Hamilton et al. 2010; Stoddard et al. 2006). To be useful in research on ecosystem recovery, the reference sites must have been protected from the main stressor of interest (e.g., acidification or eutrophication), and, ideally, other stressors common in the study region (e.g., watershed development, introduced species, road salt application, Bailey et al. 2004). How can reference sites be identified?: The methods used to identify reference sites will depend in part on the research question being addressed. Regardless of the question, however, the selection criteria used to designate reference sites must be clearly defined. These criteria may differ among regions, depending on levels of human impact or natural variability in variables of interest (e.g. total phosphorus, Cl- concentration; Herlihy et al. 2008). Methods used to identify reference sites include the use of selection criteria or ‘best professional judgement’, multivariate or multimetric techniques, or a hybrid approach using both (see recent reviews by Hawkins et al. 2010; Soranno et al. 2011). Ultimately, the selection process used will depend on the availability of information on the candidate sites. Paleolimnological studies that use data unsuited to quantification (e.g., descriptive regional histories) may need to use methods that include a reliance on selection criteria and/or best professional judgement. How can reference sites be used to understand the effects of multiple stressors?: The implicit null hypothesis of paleolimnological studies is one of ‘no change over time’ or, more 62 specifically, no change exceeding the natural variability of the system at a decadal/sub-decadal temporal resolution (Smol 2008). When paleolimnologists identify a species change that appears to be outside natural variability, they must attribute that change to a mechanism based on their knowledge of the study site/region and the ecology of the taxa they have identified in the paleorecord. Attributing species response to a particular mechanism can be difficult in ecosystems affected by multiple anthropogenic stressors (Quinlan et al. 2008). The growing influence of climate change on the physical, chemical, and biological properties of lakes adds complexities to a system’s response to other environmental stressors (Smol 2010). Studies of climate change typically involve a two-step process: a change must be first detected in a particular system and then that change must be attributed to a causal mechanism (Hegerl et al. 2007). By using reference sites, the paleolimnologist is following one of the basic principles of good scientific design; namely that researchers should attempt to simplify their systems so that they can examine the effect(s) of stressors in isolation (Jager and Loonman 1995). The careful use of reference systems may allow researchers to identify shifts in species composition due to regional climate change independent of other stressors, an inference that can be strengthened by comparing paleolimnological data with long-term climate monitoring data (Battarbee et al. 2012). Thus, in a multiple stressor environment, using reference sites to track long-term species response to regional climate change can help identify whether or not a ‘shifting baseline’ effect is likely occurring in a region. How can reference sites be used to define recovery targets?: In paleolimnological studies of environmental stressors, the recovery target for a lake is usually defined as the pre-disturbance condition of that lake (e.g. ca. 1850, Battarbee et al. 2011). However, a lake responding to multiple stressors over the recovery period is unlikely to return to its pre-disturbance state, even 63 with management intervention (Bennion et al. 2011; Hobbs et al. 2009). In this case, a reasonable recovery target may be defined by comparing impacted species assemblages to those of suitable reference sites (i.e. lakes with higher pH/ANC values that have not experienced chronic acidification, See Figure 3.1). For example, if the reference sites document a recent increase in warm-water taxa or small planktonic algae, changes predicted to occur with regional climate warming (Winder and Sommer 2012), a recovery target for an impacted site should reflect the regional shifts in species composition. Effectively, the recovery target would be recovery to a “novel” state that is similar to suitable reference systems (Hobbs et al. 2006). The ‘recovered’ species assemblages of the impacted site would contain elements of the predisturbance assemblage such as previously abundant acid-sensitive taxa and ‘novel’ additions such as an increased abundance of warm-water taxa. Note that in this framework, the reference sites do not serve as ‘controls’ for the impacted sites in the traditional sense (i.e. sites with identical species composition, morphology; Downes et al. 2002) but should, nevertheless, be similar to the impacted site(s) in general characteristics (i.e. surface area, depth, watershed characterstics) so that meaningful comparisons among the sites can be made. Why is it important that paleolimnological studies of reference sites be ‘regional’ in nature?: Though paleolimnological studies of environmental stressors may use reference sites (e.g. Alahuhta et al. 2009; Charles et al. 1990; Jeziorski et al. 2013), such studies often have fewer reference sites than impacted sites, sometimes relying on only a single reference site. Such a study design is understandable given the time-intensive nature of paleolimnological sampling. However, a limitation in this approach is the possibility that the chosen reference site(s) is/are not characteristic of the region (Downes 2010). Thus, rather than identifying a common regional response to climate warming, the reference site may be responding to a local phenomenon such 64 as an unknown introduced species or watershed disturbance. Similarly, paleolimnological studies have a robust history of using remote, minimally-disturbed sites in arctic or alpine areas to track the influence of environmental stressors like climate change (Catalan et al. 2013). While important, these sites are by design remote from populated areas and so are not ideal for defining recovery targets (e.g. alpine areas are under different climate-controls than more populated, temperate lowland areas). An alternative approach to using a small number of reference sites in paleolimnological studies or to using a number of remote sites would be to combine two different types of paleolimnological approaches: i) a ‘top-bottom’ approach (Cumming et al. 1992) and ii) a traditional ‘down-core’ paleolimnological study embedded within an anthropogenicallyimpacted area (See Figure 3.2). In ‘top-bottom’ studies, researchers only analyze a present-day sediment sample (the ‘top’ of a sediment core, normally between 0-1 cm) and a sample from a set depth in the core that is taken to represent pre-industrial conditions (e.g., 20 cm to represent ~1850). The top-bottom approach is similar to a BACI design, in that it compares a time period minimally influenced by regional stressors like climate change (the ‘bottom’ sample) with a recent time where such stressors may be more important (the ‘top’ sample). However, preindustrial conditions may not represent a “pristine” time-period and so researchers should attempt to characterize the early disturbance history of the sites (e.g., land-clearance/logging history), if possible. The top-bottom approach gives a ‘snap-shot’ of change throughout the region. Because the temporal resolution at each study site is low (i.e., 2 samples), researchers can analyze data from a much larger number of sites than is possible in a traditional ‘down-core’ paleolimnological study. With top-bottom data, researchers can test for a significant difference in species composition from pre-industrial times to present and, depending on the study design, 65 attribute changes to stressors such as regional warming (Enache et al. 2011). A variety of statistical frameworks can be used to analyze top-bottom paleo-data, including univariate methods (Wilcoxon Sign-Rank test, e.g., Korosi and Smol 2012) or multivariate methods (e.g. ordination techniques, e.g., Hyatt et al. 2010; or Analysis of Similarities, e.g., DeSellas et al. 2011). Analyses can be performed on the entire paleo-species assemblage, on key individual taxa, or, if transfer functions are applied to the paleo-species data, on reconstructed inferred-pH (or TP, etc.) values (Juggins and Birks 2012). However, interpretations of such mathematical reconstructions must be rooted in an understanding of the biological assemblage as a whole, as inferred-water chemistry values may ‘mask’ important changes in individual taxa of management concern (e.g., increases in taxa that cause taste and odour issues, Paterson et al. 2004). While the top-bottom approach is effective at building a regional picture of species change from pre-industrial times to present, the method provides limited information for each individual site. If one of the goals of the reference site approach is to use the reference sites to define recovery targets for impacted sites, more detailed temporal information may be required. Thus, a complementary approach to the top-bottom study is to compare a subset of the reference lakes with a set of impacted lakes in a down-core study. In a down-core study, many sediment sub-samples (e.g., n > 20) are analyzed per core, providing a more detailed temporal understanding of changes in species assemblages over time at a site. The down-core study would allow researchers to compare the timing of species changes between the reference and impacted sites, as well as compare paleolimnological data to long-term monitoring data, if available. Again, this comparison can be done using univariate methods (e.g., correlations between key taxa and long-term chemical/climatic monitoring data, Battarbee et al. 2012), multivariate 66 methods (e.g., plotting both impacted and reference sites in ordination space, Yan et al. 2004) or a combination of both. Ultimately, researchers must be conscious of scale in their study design (Allen and Hoekstra 1992). The effects of climate change are regional in nature but those effects are filtered at the local scale by a given watershed and its waterbody and are reflected in the physical, chemical, and/or biological characteristics of the system (Adrian et al. 2009). A reference lake study design should therefore allow for an understanding of both broad regional trends and lake-specific responses, and combining the top-bottom and down-core approaches can provide both. While it is difficult to determine the minimum number of sites required in a regional study, given the variation in lake physical/chemical characteristics, a minimum of at least 25 reference sites is likely needed (Bailey et al. 2004). Implementing a regional paleolimnological study of reference sites in the Adirondacks (NY, USA) The study region: The Adirondack Park presents an excellent opportunity to demonstrate how paleolimnological techniques can be used to investigate biological recovery from acidification within a regional reference-site framework. The Adirondack region is a mountainous area that is a largely forested mosaic of public and private lands (Liroff 1981). Adirondack lakes have been characterized extensively in regional surveys (e.g., Krester et al. 1989) and long-term chemical monitoring data are available for many sites (Roy et al. 2011). The region has been heavily impacted by acid deposition (Driscoll et al. 1991) and the acidification history of the region has been well-documented in previous paleolimnological studies (Cumming et al. 1992b; Cumming et al. 1994). Following declines in acid deposition, many lakes in the region are now undergoing chemical recovery from acidification (Driscoll et al. 2007; Waller et al. 2012). However, while chemical recovery may be progressing, the Adirondacks are also experiencing the influence of 67 regional warming. Mean annual temperature has increased by a rate of 1.3°C per 100 years, a warming trend that has been most pronounced since the 1970s (Jenkins 2010). Project goals: Following the principles outlined above, the goal of this investigation was to identify a set of reference lakes to be used in a two-part paleolimnological analysis that included a top-bottom study and a down-core study. The purpose of the top-bottom study was to identify if changes in species assemblages were occurring in Adirondack lakes independent of several common stressors. Subsequently, a subset of the reference lakes were compared with a set of Adirondack lakes that were known to have acidified by acid deposition and are now experiencing chemical recovery from acidification (based on long-term monitoring data) in a down-core paleolimnological study. By comparing long-term trends in reference sites and impacted sites, we sought to identify if a shifting baseline in background conditions was occurring and, if so, use the reference sites to define recovery targets for the impacted sites. Reference site selection criteria: This study builds on the work of Stager and Sanger (2003) who introduced the concept of “Heritage Lakes”, or minimally-disturbed lakes in the Adirondacks. Our primary interest was to identify a set of Adirondack lakes that had been largely protected from several stressors common in the Adirondack Park, namely: acidification, eutrophication from watershed/shoreline development, road salt seepage, and the introduction of non-native piscivores. The reference lakes were identified from an online database of 1,469 Adirondack lakes sampled from 1984-1987 as part of the Adirondack Lakes Survey (ALS) (Krester et al. 1989) (database available at: www.adirondacklakessurvey.org). The ALS used a modified random sampling design. In 1984-1986, the survey focused on three of the five large watersheds in the Adirondack Ecological Zone (AEZ), a region that extends beyond the boundary of the Adirondack Park and covers the area that falls within the 305 m (1000 ft.) contour line of the 68 Adirondack Uplands. The three watersheds contained 1,862 lakes greater than 0.2 ha in surface area, of which 1,200 lakes were sampled. Sixty percent of the lakes were chosen randomly and then additional waterbodies within 2 km of the randomly selected sites were sampled. In 1987, a pure randomized design was used to sample a minimum of 250 waters in the remaining two watersheds. In total, the 1,469 lakes sampled in the ALS survey are considered a representative sample of the 2,759 lakes greater than 0.2 ha that fall within the AEZ. The six selection criteria used to identify reference lakes in the ALS are listed below. The criteria are based on the findings of previous limnological or paleolimnological studies of the Adirondack Park and elsewhere. 1. Public access: The search was restricted to lakes with public access. Lakes that were not accessible via hiking trails and lakes with excessively long or difficult hikes were also excluded. 2. 1980s pH > 6.5; 1980s ANC > 50 μeq L-1: These paired criteria were used to exclude lakes that were likely to have acidified. Cumming et al. (1992) demonstrated that few Adirondack lakes with a pH > 6.5 in the late 1970s and early 1980s underwent significant chronic acidification. Additionally, an acid neutralizing capacity (ANC) > 50 μeq L-1 largely protects acid-sensitive lakes from episodic acidification (Driscoll et al. 2001).7 3. Watershed & shoreline development < 5%: These paired criteria were used to exclude lakes that were likely to have experienced eutrophication due to development along their shorelines or in their watersheds. A 5% development cut-off was used so that Adirondack lakes with small camps would not be excluded. 4. Cl- < 2 mg L-1: In the ALS, lakes were considered ‘salt impacted’ based on a cut-off criterion of 20 µmol L-1 (~0.71 mg L-1) (Newton and Driscoll 1990). We applied a less restrictive selection criterion, allowing lakes with a modest concentration of salt to be included in the survey. In general, it was found that lakes excluded from the reference set 7 Middle Branch Lake was included in the survey, despite having an ANC < 50 μeq L-1 in the ALS survey (pH 6.9; ANC 18.4 μeq L-1). Inclusion was based on the fact that the lake had a pH > 6.5 at the time of sampling and longterm monitoring data from Middle Branch Lake suggests that the lake has generally maintained a pH > 6.0 since 1993, with summertime pH values frequently reaching > 6.5 (Roy et al. 2011). 69 based on high Cl- concentration were also excluded based on other criteria, primarily shoreline or watershed development. 5. No introduced piscivores: One aim of this paleolimnological study is to reconstruct cladoceran zooplankton assemblages, a group that includes Daphnia species that are sensitive to changes in fish predation (Korhola and Rautio 2001). As such, lakes that had non-indigenous piscivores (e.g., largemouth bass, Micropterus salmoides) in the ALS survey were excluded from the study. Lakes with rainbow smelt (Osmerus mordax), a species introduced in the Adirondack uplands that can compete with lake trout fry, were also excluded. While it may have been desirable to exclude lakes with any nonindigenous fish species, some taxa such as yellow perch (Perca flavescens) and the golden shiner (Notemigonus crysoleucas) are so widespread in the Adirondacks that excluding lakes based on the presence of these taxa would have reduced the number of sites to an insufficient number for study. Non-native piscivores were identified using George (1981), Kraft et al. (2006), and USGS (2013). 6. Maximum depth > 5 m; surface area > 4 ha: These selection criteria were included because some of the paleo-indicators of interest (scaled-chrysophytes, the Chrysophyceae and Synurophyceae; and Daphnia taxa) are planktonic taxa that are rare or absent in small, shallow waterbodies (Dixit et al. 1999; Ginn et al. 2010; Jeziorski et al. 2012). Recognizing that many Adirondack lakes are shallow and < 4 ha in size (Krester et al. 1989), the size selection criteria used in this investigation were chosen to maximize the total number of possible reference sites while minimizing the inclusion of sites unlikely to contain the paleo-indicators of interest. Identifying historic landscape disturbance & fisheries disturbance: Once the reference sites were identified in the ALS, we sought to distinguish between sites that had experienced historic watershed and fisheries disturbance from those that had not. In paleolimnological studies, the process of attributing a species response to a particular stressor can be facilitated with knowledge of the disturbance history of a particular lake. Such historical information is often incomplete or 70 qualitative. However, even incomplete historical data can help paleolimnologists distinguish between species responses to a local stressor versus a regional stressor such as climate change. This investigation used both quantitative and qualitative methods to define the landscape disturbance history of the reference sites, identify alterations to historic fish assemblages, and group the lakes into ‘classes’ based on their disturbance history (see Figure 3.3). Two historic maps were used to characterize early watershed disturbance in the reference lakes. The first historic map is a digital version of an 1890 map made available online by the New York State Archives and the Adirondack Museum Library. The map shows the condition of Adirondack forests prior to the creation of the Adirondack Park in 1892 (Original map scale-bar shown in digital version, geographic coordinates not provided; NYS Forest Commission 1890). The second data source is a 1916 “fire protection map” (hereafter referred to as the 1916 fire map) digitized by the Adirondack Park Agency (APA) for use as a tool to determine historic disturbance in the Adirondack Park (Scale 1:126720; APA 2000). 1890 and 1916 are two important time periods in the development history of the Adirondack Park. Prior to 1890, logging practices in what is now the Adirondack Park were fairly benign, involving the selective logging of merchantable timber (primarily large softwoods), a process that did not alter the fundamental composition of the forest (McMartin 1994). The original 1892 boundary of the Adirondack Park was set to encompass lands that were primarily virgin forest or areas logged for softwoods pre-1890, while areas outside the boundary included regions that had been burned for charcoal production (denuded) or cleared for agriculture/grazing. Thus, 1890 is a critical time-window that represents a period where most of the lakes within the original 1892 boundary can be considered minimally disturbed. With the creation of the 1894 “forever wild” provision in the New York State constitution, logging was 71 prohibited on State lands in the park. However, logging increased on private lands from the 1890s until ~1910, after which harvest generally declined (McMartin 1994). Additionally, parts of the Adirondack region (primarily private land) experienced major fires in 1903, 1908, and 1913 caused by coal sparks along now largely abandoned railways, fires that were exacerbated by logging and land-clearance practices (McMartin 1994). From 1910 onwards, there was an expansion of the area covered by the Adirondack Park and an enlargement of State-held lands protected from logging (Jenkins and Keal 2004). Thus, 1916 represents a time-window that captures much of the historic watershed disturbance experienced by Adirondack lakes on private or public lands. Focusing on two early time periods (1890 & 1916) is a sufficient level of historical investigation for a top-bottom study because of the method’s coarse temporal resolution (i.e., a two-sample comparison between the present-day ‘top’ sample and a pre-industrial ‘bottom’ sample). Using these two early maps provides us with information on historic local disturbance, a disturbance that may be reflected in the species assemblages of the ‘bottom’ sample. It is difficult to assign an exact date to the ‘bottom’ sample because of differences in sedimentation rate (the rate at which sediment accumulates in a lake basin) among lakes (Cumming et al. 1992b). For example, a sediment slice from 20 cm in one lake may represent sediment from 1850, while in another it may be ca. 1900 or later. Thus, while we may have been justified to restrict our analysis to the pre-1900s, to be conservative in our approach we chose to examine disturbance in both 1890 and 1916. In the subsequent down-core paleolimnological study, additional historical resources were used to reconstruct the disturbance histories of a subset of the reference lakes over a longer time period (not discussed here). 72 The two historic maps provided us information on the spatial extent of logging, fire disturbance, and land clearance in 1890 and 1916 in what is now the Adirondack Park. From a paleolimnological perspective, the importance of a watershed disturbance is related to its influence on the chemical composition of lake-water (e.g., increased export of cations, anions, nutrients, dissolved and suspended organic matter, trace metals) which, in turn, can influence the composition of aquatic species assemblages. Selective logging can influence runoff concentrations of cations and anions but has a short-lived impact on stream chemistry (Scott et al. 2001; Wang et al. 2006). Fire can increase runoff and erosion rates, increase nutrient export and, in cases where large portions of the watershed are denuded, cause vegetative succession (Binkley 1999; Committee on Hydrologic Impacts of Forest Management 2008). Lastly, persistent land clearance has many potentially long-lasting influences on aquatic ecosystems, including increased organic matter export to nearby lakes, changes in stream hydrology, and eutrophication from increased phosphorus loading (Foster et al. 2003). Thus, when the reference lakes in this investigation were ranked based on historic disturbance, logging, fire, and land clearance, in that order, were viewed as increasingly detrimental to water quality and increasingly influential on aquatic species. The 1890 map was analyzed qualitatively. The map lacks geographic co-ordinates and contains discrepancies in the placement of some smaller waterbodies and so is not suitable for quantitative analysis. The reference lakes were subdivided into “A” lakes and “B” lakes based on their land-cover classification in the 1890 map. “A” lakes were classed as virgin forest or logged for softwoods in 1890 and are considered minimally disturbed whereas “B” lakes were in denuded or cleared areas. In cases where a reference lake was not clearly labelled in the 1890 73 map (typically smaller or more remote sites), the placement of the lake was inferred based on the proximity to other geographic features (e.g., large waterbodies, mountains). The 1916 fire map was analyzed quantitatively using GIS methods (see Jenson and Domingue 1988). First, watershed maps were generated based on elevational gradients around the lake using ARCGIS 10.0. The 10 m US National Elevation Dataset was used as the base Digital Elevation Model (DEM) (USGS 1999). The DEM was corrected to fill pits or sinks that, while either real-world phenomena or digitizing errors, would corrupt the hydrological models. After this process, a flow direction model (FDM) was created. The FDM examined the corrected DEM on a cell-by-cell basis and, for each cell, evaluated which direction water would flow based on the elevation of the surrounding 8 cells. The FDM was then used to create a flow accumulation model (FAM) which mirrored the drainage pattern of the landscape. Using the FAM, the point of highest accumulation for each lake was manually digitized as a point. The point was then used to calculate the watershed for that lake using the FDM, which stepped backwards up slopes from the starting point until a watershed divide was reached (i.e., flow direction = null and/or accumulation = 0). We defined watershed disturbance by calculating the percentage of each watershed that had been affected by logging, fire, or land clearance on the 1916 map. Many GIS studies quantify not only the overall extent of watershed disturbance (i.e. % disturbed area) but also the spatial structure of the disturbance (e.g., amount of riparian vegetation cleared, slope of cleared areas, Rogan and Miller 2007; Uuemaa et al. 2009). However, we used a non-spatial metric because the 1916 digitized fire map showed discrepancies between its ‘water’ classification and the placement of lakes in Adirondack watersheds. These discrepancies are not uniform or systematic and likely originate from the digitization process of the original 1916 fire map (a 74 hard-copy map including warped areas, digitized into 12 coverages edge-matched and appended together, APA 2000). Given these discrepancies, interpreting fine-scale differences in watershed disturbance was not appropriate. Regardless, the majority of the disturbed areas in the 1916 map cover large areas and so this approach is sufficient to provide a general understanding of historic watershed disturbance in the reference lakes. GIS studies of digitized historic maps often quantify disturbance using non-spatial metrics (e.g., change in ‘% forest area’ or ‘% arable land’, Bender et al. 2005; Skaloš et al. 2012), and % forest cover is often found to be the most important determinant of water quality (Hunsaker and Hughes 2002). To calculate disturbed areas in the watersheds of the reference lakes, the watershed maps were converted to polygon (from raster) datasets. The area of the watersheds’ polygons were calculated, and intersected with the 1916 fire map. Intersecting areas retained the original size field and so a new size calculation of the intersected polygon revealed the percentage of a land cover layer that fell within a watershed. The 6 terrestrial land-use/land-cover classifications in the 1916 fire map were simplified into 4 groups and scored as progressively detrimental: green timber (0); selective logging, including areas logged for softwoods, and areas logged for softwoods/hardwoods (-1); fire damage, including burned, with slash, and burned, denuded (-2); and open land (-3). Each lake was assigned a score based on the percentage of the watershed that had been impacted by a particular disturbance type. For example, a lake that had experienced selective logging in 50% of its terrestrial watershed, with the remaining area classed as green timber would receive a score of -0.5, and a lake that had been logged in 50% of its watershed, burned in 25% of its watershed, and had green timber in the remaining areas would receive a score of -1. The reference lakes were then separated into 3 disturbance classes based on these scores: Class 1 (least disturbed), 100% of the watershed listed as green timber (score = 0); Class 75 2, at least 1 form of disturbance in less than 50% of the terrestrial watershed (0 > score > -1); and Class 3, at least 1 form of disturbance in more than 50% of the watershed (score < -1). Lastly, reference lakes that experienced reclamation and/or had non-native piscivores in a recent (post-ALS) survey were distinguished from those without known changes. Reclamation is a fisheries management strategy employed by the New York State Department of Environmental Conservation (NYSDEC) in the Adirondack Park. Reclamation involves the addition of rotenone to a lake to eliminate all fish taxa and allow for subsequent restocking of desirable taxa, a process that can influence the size and structure of zooplankton species assemblages (Harig and Bain 1998). As cladoceran zooplankton taxa are sensitive to alterations in fish assemblages and are a paleo-indicator of interest, reclamation is considered a form of disturbance in this investigation. A list of lakes reclaimed prior to 2011 was provided by R. Preall (NYSDEC). Recent (1990-2011) fisheries surveys (NYSDEC 2012; Roy et al. 2011; R. Preall pers. comm.) were available for 24 of the reference lakes and were examined to identify the presence of nonnative piscivores in the most recent fisheries survey. For the 7 remaining reference sites, the most recent fisheries survey is the ALS data. When identifying fisheries disturbance, the most recent survey was used (rather than all surveys combined), because not all of the reference lakes included in the NYSDEC data had been surveyed multiple times (minimum 1 survey; maximum 10). In the disturbance classes defined using the historic watershed maps, lakes that have experienced reclamation and/or the introduction of non-native piscivores are demarked by a negative (-) sign (i.e., a lake that was virgin forest in 1890 and 100% green timber in 1916 but reclaimed in the 1950s is listed as Class 1A-, See Figure 3.3). Comparing the reference lakes to the ALS population: The reference lake selection criteria were chosen to identify lakes that are minimally influenced by anthropogenic stressors but the criteria 76 also excluded certain kinds of Adirondack lakes (e.g., naturally acidic ponds). If reference lakes are chosen based on a priori defined selection criteria, it is important to identify how ‘representative’ the reference lakes are of the different lake-types in a region (e.g., high-DOC sites vs. low-DOC sites, deep lakes vs. shallow ponds). Thus, we plotted the 1,469 ALS lakes in a Principal Components Analysis (PCA) of environmental variables (Lepš and Šmilauer 2003) to compare the range of chemical/morphological characteristics of the reference lakes with that of the ALS population; a population that is itself representative of 2,759 lakes in the AEZ. Prior to inclusion in PCA, missing samples in the environmental matrix were replaced with the variable’s mean value (Lepš and Šmilauer 2003). Variables with more than 4% of their values missing and strongly correlated environmental variables (R > 0.9) were excluded from analysis. A Shapiro-Wilks normality test was applied to test the normality of all environmental variables (and sqrt/log transformed datasets). However, none of the environmental variables were found to have a normal distribution, although many were minimally skewed. Rather than arbitrarily exclude outliers (which would affect the representativeness of the ALS population), we opted to visually inspect histogram plots of the environmental variables. As PCA is fairly robust to minor violations of the assumption of normality, we selected 16 environmental variables (or their sqrt/log transformed equivalents) that were roughly normally distributed for inclusion (See Appendix A). Three ordinations were performed: one on chemical/morphological data, one on chemical data only, and one on morphological data only. PCA ordinations were performed using CANOCO 5.0 (ter Braak and Šmilauer 2012). In a PCA of environmental variables, samples that are similar in chemical/morphological characteristics plot close together in the ordination diagram. This aspect of PCA allows us to identify two interesting lake populations in the ALS. First, using the placement of the reference 77 sites in the PCA to define the outer bounds of a polygon in PCA space, we can identify a population of ALS lakes that, although they failed to meet one or more of the selection criteria, are similar in physical/chemical characteristics to the reference lakes (hereafter referred to as polygon-sites). Secondly, assuming that the primary gradient in the PCA is a pH/ANC gradient, ALS sites that fall within the same range as the reference-lake polygon on the secondary axis, but are outside the polygon on the low end of the pH/ANC gradient, form a population of lowpH/ANC lakes that the reference sites can be used to define recovery targets for. This comparison is possible because the low pH/ANC sites have physical/chemical characteristics (size, colour, nutrient levels, etc.) to the reference lakes. Because the ALS is considered representative of 2,759 lakes in the AEZ, we can extrapolate the percentage of sites in the ALS that meet our selection criteria to lakes in the AEZ. The extrapolated population is the (relatively small) number of Adirondack lakes that the reference sites are most representative of (or, most similar to). However, the polygon-sites are also similar in physical/chemical characteristics to the reference sites, though the polygon-sites failed to meet one or more selection criteria. The polygon-sites form a larger ‘representative’ population that is of interest as a study population for additional research questions (e.g. comparing species responses in reference sites to an equal number of polygon sites that have experienced watershed development). Field sampling & additional GIS methods: To characterize the present-day water chemistry of the reference sites, water samples were collected by hand in sterile glass and polypropylene jars from ~0.5 m beneath the surface. All chemical analyses were performed by the Center for Environmental Systems Engineering (Syracuse University, NY) using standard US EPA methods (See Appendix B). Additionally, current land-use in the watersheds of the reference was defined 78 using the 2006 National Land Cover Database (Homer et al. 2012). Percent area of land-use categories was identified using the methods outlined above. Results Site descriptions: Using the selection criteria described earlier, we identified a total of 30 reference sites in the ALS but one lake had a Cl- concentration exceeding 2 mg L-1 in 2010 and so was excluded post-hoc from the dataset, leaving 29 lakes (~2.0% of ALS lakes) (See Figure 3.4 & Appendix C). If we extrapolate this percentage (~2.0%) to the 2759 AEZ lakes, we find that there may be up to 54 AEZ lakes that meet the selection criteria. Two additional lakes, Arbutus Lake and Wolf Lake, which were not part of the ALS, were added post-hoc to the reference lake dataset. Both lakes are located in the Huntington Wildlife Forest, a private research property owned and protected by the State University of New York, School of Environmental Science and Forestry (SUNY-ESF) since the 1930s. The lakes have been protected from acidification, watershed and shoreline development, lack non-native piscivores, meet the depth and surface area criteria (Roy et al. 2011; Stager and Sanger 2003), and currently have Cl- concentrations < 2 mg L-1. Most of the reference lakes are protected in either ‘wild forest’ or ‘wilderness’ preserves in the Adirondack Park. According to 2006 land-cover data, the two most dominant land-cover types in the watersheds of the reference lakes, excluding open water, are deciduous forest and coniferous forest. The least common land-cover type is developed land (See Table 3.1). In general, the reference lakes are small (mean SA: 20.5 ha), moderately deep (mean Zmax: 10.9 m), and are located at elevations greater than 300 m. The mean pH of the reference lakes was 6.4 when sampled in 2010 and 2011 and 28 lakes (90%) had ANC values greater than 50 μeq L-1. All 79 of the lakes had Cl- concentrations less than 0.5 mg L-1. The lakes are oligotrophic (mean TP = 1.3 µg L-1), calcium-poor (mean Ca2+ = 2.3 mg L-1) , with a DOC range of ~2 mg L-1 to 8.8 mg L-1. Disturbance Rankings: A summary of the watershed disturbance experienced by each reference lake in 1890 and 1916 is provided in Table 3.2, as well as a summary of fisheries alterations. In 1890, 24 (77%) of the reference were located in minimally-disturbed forests (virgin forest or selectively logged for softwoods; “A” lakes). The lakes located in denuded or cleared regions (“B” lakes) are primarily located in what is now the eastern Adirondack Park (See Figure 3.4). In 1916, 13 lakes (42%) had the terrestrial areas of their watershed listed as 100% green timber (Class 1), of which 9 were considered minimally disturbed in 1890 (Class 1A). An additional 6 lakes (19%) had experienced at least 1 form of disturbance in less than 50% of their watershed (Class 2), and 12 lakes (39%) had experienced a least 1 form of disturbance in more than 50% of their watershed (Class 3). Forest fires were the most common form of anthropogenic disturbance experienced by the reference lakes in 1916, followed by selective logging, and agricultural clearance. In terms of fisheries disturbance, as of 2011, 20 of the reference lakes (65%) had been reclaimed and 6 lakes (19%) had non-native piscivores identified in their most recent post-ALS fisheries survey (See Table 3.2). In total, 5 Class-1 reference lakes (16%) (and 4 Class 1A lakes), have no record of reclamation and lacked non-native piscivores in their most recent fisheries survey. Ordination Results: The first two axes of the PCA plots based on chemical/morphological data, chemical data only, and morphological data only captured approximately 50%, 57%, and 73% of the variation in environmental data of samples in the ALS dataset respectively (See Figure 3.5). 80 In general, when compared to the total ALS population, the reference lakes are comparatively low in DOC/TP, are generally deeper and have larger surface areas/watershed areas, and are intermediate in pH, conductivity, and concentration of major ions. Polygons created to bound the reference lake dataset in the PCA contained 356 (24.2% of the population), 534 (36.4%), 451 (30.7%) sites in the chemical/morphological, chemical only, and morphological only PCAs respectively. There was significant overlap in the polygon-sites between the 3 PCAs (48.8%91.9%). In the chemical/morphological PCA (the most restrictive test), the most common reason why the polygon-sites were excluded from the reference lake set was the presence of non-native piscivores. Discussion Understanding reference sites: The myth of the pristine & the dangers of presentism The ideal reference site would be a hypothetical “pristine” site that has never experienced any anthropogenic disturbance. Given the pervasive influence of stressors such as atmospheric deposition and climate change, such a site is unlikely to exist (Stager and Sanger 2003). However, this scarcity does not imply that the presence of present-day or historical anthropogenic influence on a lake makes it unsuitable as a reference site. The process of identifying reference sites is inherently subjective, even if ordination methods are used to ‘organically’ identify sites (Yates and Bailey 2010), because it depends on the research goals of the study and the input data used. A site deemed to be in “reference” condition in one study could be inadmissible in another. Some studies employ simple criteria (e.g., absence of stressor of interest, Chessman et al. 2008) while others use more holistic measures of anthropogenic disturbance (Wang et al. 2008). Researchers must therefore be able to answer the question 81 “reference for what?”. The selection criteria and methods used to identify the reference sites must be clearly defined and correspond to the goals of the study. While the selection criteria used in this investigation identified lakes that have been largely protected from the detrimental effects of acid deposition, watershed/shoreline development, and road salt seepage, historical analysis shows that many of the sites have experienced early watershed and/or fisheries disturbance. Like many protected areas, the Adirondack Park has a long history of human impact, including land clearance, logging, forest fires, and non-native piscivores introductions (George 1981; McMartin 1994) and the 31 reference lakes in this investigation reflect that history. This study thus serves as a reminder of the danger of presentism in ecological research; namely, the assumption that the present-day physical, chemical, or biological characteristics of a system are representative of the long-term historical condition of the site (Leira et al. 2006). Landscape disturbance may have long-lasting effects on waterbodies in certain circumstances (Martin et al. 2011; Renberg et al. 2009), and so limnological studies should ideally include a thorough historical analysis. The historical perspective offered by paleolimnology can play an essential role in developing a more complete understanding of the present-day ecology of a region and the response of species assemblages to environmental stressors. Unfortunately, most studies that rely solely on modern limnological sampling lack this historical perspective. For example, in a recent meta-analysis of 240 recovery studies that included both terrestrial and aquatic studies, Jones and Schmitz (2009) found that only 20% used pre-disturbance data (and, interestingly, only 58% used undisturbed reference sites). This lack of pre-disturbance data is problematic because, while many recovery studies have an implied recovery goal of a return to pre-disturbance state, without the historical perspective it is difficult to define such a recovery target, let alone assess whether 82 or not the recovering sites are experiencing a shifting baseline. Indeed, the fact that the 2012 US National Lakes Assessment includes a paleo-component (USEPA 2011), and that the EU Water Framework Directive supports the use of paleolimnological data to determine ‘reference conditions’ (Bennion et al. 2011), highlights the growing recognition of the value the paleoapproach provides in environmental assessment. Using paleolimnological techniques within a reference site framework allows us to answer the question “what condition should impacted sites recover to?”. While the ideal situation would be a return to pre-disturbance condition (e.g. conditions ca. 1850), this may not be a feasible recovery goal in a period of global environmental change. Defining recovery targets: making meaningful comparisons between reference & impacted sites Ideally, reference sites in paleolimnological studies should be minimally-disturbed by multiple environmental stressors. However, such minimally-disturbed sites may not necessarily be representative of all lake-types in a region. If reference sites are to be used to define recovery targets for an impacted site, the site must be similar in general characteristics to the reference systems to allow for meaningful comparisons. For example, our ordination of the ALS survey found that the reference lakes are, in general, mid-sized, mid-elevation, low DOC/TP lakes with circumneutral pH. As such, none of our reference sites would be appropriate to define a recovery target for a high DOC acidic Adirondack pond. To address this data gap, we could increase the pool of reference sites by relaxing some of the selection criteria (in this example, the pH / depth criteria) but this could result in the inclusion of a number of acidified sites. The tradeoff between maximizing the number of reference sites and minimizing the possibility of classifying ‘disturbed’ sites as reference sites should be carefully considered when implementing regional paleolimnological studies in a reference site framework. 83 Testing hypotheses about climate change Using reference lakes provides us with the opportunity to test hypotheses about climate change on species assemblages. Given the rigor of the selection criteria used, if similar species changes are documented in the reference lakes, it will be possible to conclude that: 1) the species changes are not the result of acidification, eutrophication, salinization, or piscivores introductions; 2) that a regional stressor is responsible for the species changes; and 3) that climate change could be the cause of the species changes noted. This last inference can be made through the use of statistical techniques such as ordination methods (Hyatt et al. 2010), a strong understanding on the ecology of the paleo-indicators under study, and by comparing ‘down-core’ paleorecords with long-term climate monitoring data (Battarbee et al. 2012). The historical analysis done in this study also allows us to conduct ‘natural experiments’ to test the influence of local versus regional stressors on species assemblages. For example, it will be possible to identify the influence of fisheries disturbance on the assemblages by comparing the species responses in an equal number of reclaimed lakes and non-reclaimed lakes using either univariate or multivariate frameworks. Similarly, it will be possible to test the influence of early watershed disturbance by comparing lakes in different disturbance classes. Presumably, if local stressors are influencing species assemblages, there should be a significant difference in species composition between disturbed and undisturbed sites at a centennial scale. If not, a regional stressor is likely the cause of the change in species assemblage. Taken together, these analyses can help develop a holistic view of how aquatic species assemblages are likely to respond to climate change both independent of and synergistically with other stressors. 84 Applications in other regions The successful implementation of this project in the Adirondack Park relied on three factors: the availability of data from a large synoptic survey; previous paleolimnological studies that provided the basis of appropriate selection criteria; and historical GIS data and fisheries information for the study sites. During the acid-rain crisis, many regions in North America and Europe were surveyed as part of large, synoptic surveys (e.g., ELS, Landers et al. 1988; EMAP, Hughes et al. 2000; PIRLA-I and PIRLA-II, Charles and Smol 1990; Charles and Whitehead 1986; SWAP, Battarbee and Renberg 1990), many of which included paleolimnological studies. There is also a growing availability of digital historic maps and online archival data for use in the reconstruction of land-use change over time. Thus, the approach followed in this investigation could be modified and applied in other regions where similar data exist. Of particular interest would be the development of reference lakes in protected areas (Baron et al. 2009), which could serve two roles: first, identifying the influence of climate warming on lakes in the region, and second, serving as reference sites for lakes in developed areas. Landscape development threatens to degrade the condition of reference sites across the United States (USEPA 2012). As such, the importance of reference sites in protected areas will continue to increase as the number of reference sites in or near urban, suburban, and agricultural areas decreases. Conclusions Lakes in the Adirondack Park and elsewhere in North America are increasingly influenced by multiple stressors, with anthropogenic climate change posing a growing threat to management and conservation efforts. Paleolimnological studies of minimally-disturbed reference lakes can be used to identify the influence of climate change on aquatic species assemblages independent of other stressors and define recovery targets for impacted sites. We 85 described and applied a conceptual framework for pairing paleolimnological techniques with a regional reference site framework in a study of the Adirondack Park. Though not “pristine”, the 31 reference lakes have been largely protected from common environmental stressors and can be used to define reasonable recovery targets for lakes recovering from acidification and test hypotheses about climate change. The methods used in this investigation could be easily modified and applied in other regions. As land-use change continues to threaten reference sites outside of protected areas and the impact of climate warming on freshwater systems continues to grow, reference sites in protected regions will become an increasingly important tool for biodiversity conservation and policy development. Literature Cited Adrian R, O'Reilly CM, Zagarese H, Baines SB, Hessen DO, Keller W, Livingstone DM, Sommaruga R, Straile D, Van Donk E, Weyhenmeyer GA, Winder M. (2009) Lakes as sentinels of climate change. Limnology and Oceanography 54:2283-2297. Adirondack Park Agency (APA). (2000) 1916 Fire protection areas of the Adirondack Park. Adirondack Park Agency, Ray Brook, NY. http://www.apa.ny.gov/gis/shared/htmlpages/metadata/1916fire.html. Alahuhta J, Vuori K-M, Hellsten S, Järvinen M, Olin M, Rask M, Palomäki A. (2009) Defining the ecological status of small forest lakes using multiple biological quality elements and palaeolimnological analysis. Fundamental and Applied Limnology 175:203-216. Allen TFH, Hoekstra TW. (1992) Toward a unified ecology. Columbia University Press, New York, 384 p. Arseneau KMA, Driscoll CT, Brager LM, Ross KA, Cumming BF. (2011) Recent evidence of biological recovery from acidification in the Adirondacks (New York, USA): a multiproxy paleolimnological investigation of Big Moose Lake. Canadian Journal of Fisheries and Aquatic Sciences 68:575-592. Bailey RC, Norris RH, Reynoldson TB. (2004) Bioassessment of freshwater ecosystems: using the reference condition approach. Kluwer Academic Publishers, New York, 170 p. 86 Baron JS, Gunderson L, Allen CD, Fleishman E, McKenzie D, Meyerson LA, Oropeza J, Stephenson N. (2009) Options for national parks and reserves for adapting to climate change. Environmental Management 44:1033-1042. Battarbee RW. (1990) The causes of lake acidification, with special reference to the role of acid deposition. Philosophical Transactions of the Royal Society B, Biological Sciences 327:339-347. Battarbee RW, Anderson NJ, Bennion H, Simpson GL. (2012) Combining limnological and palaeolimnological data to disentangle the effects of nutrient pollution and climate change on lake ecosystems: problems and potential. Freshwater Biology 57:2091-2106. Battarbee RW, Morley D, Bennion H, Simpson GL, Hughes M, Bauere V. (2011) A palaeolimnological metadatabase for assessing the ecological status of lakes. Journal of Paleolimnology 45:405-414. Battarbee RW, Renberg I. (1990) The surface water acidification project (SWAP) palaeolimnology programme. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 327:227-232. Battarbee RW, Thompson R, Catalan J, Grytnes J-A, Birks HJB. (2002) Climate variability and ecosystem dynamics of remote alpine and arctic lakes: the MOLAR project. Journal of Paleolimnology 28:1-6. Bender O, Boehmer HJ, Jens D, Schumacher KP. (2005) Analysis of land-use change in a sector of Upper Franconia (Bavaria, Germany) since 1850 using land register records. Landscape Ecology 20:149-163. Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. Binkley D. (1999) Disturbance in temperate forests of the Northern hemisphere. In: Walker LR (ed), Ecosystems of disturbed ground. Elsevier Science Amsterdam, The Netherlands, pp. 453-466. Catalan J, Pla-Rabés S, Wolfe AP, Smol JP, Rühland KM, Anderson NJ, Kopáček J, Stuchlík E, Schmidt R, Koinig KA, Camarero L, Flower RJ, Heiri O, Kamenik C, Korhola A, Leavitt PR, Psenner R, Renberg I. (2013) Global change revealed by palaeolimnological records from remote lakes: a review. Journal of Paleolimnology 49:513-535. Charles D, Smol J. (1990) The PIRLA II project: regional assessment of lake acidification trends. Internationale Vereinigung, für Theoretische und Angewardte Limnologie, Verhandlungen 24:474-480. 87 Charles D, Whitehead D. (1986) The PIRLA project: Paleoecological investigations of recent lake acidification. Hydrobiologia 143:13-20. Charles DF, Binford MW, Furlong ET, Hites RA, Mitchell MJ, Norton SA, Oldfield F, Paterson MJ, Smol JP, Uutala AJ, White JR, Whitehead DR, Wise RJ. (1990) Paleoecological investigation of recent lake acidification in the Adirondack Mountains, N.Y. Journal of Paleolimnology 3:195-241. Chessman BC, Muschal M, Royal MJ. (2008) Comparing apples with apples: use of limiting environmental differences to match reference and stressor-exposure sites for bioassessment of streams. River Research and Applications 24:103-117. Committee on Hydrologic Impacts of Forest Management NRC. (2008) Hydrological effects of a changing forest. The National Academic Press, Washington, DC, 168 p. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Cumming BF, Smol JP, Kingston JC, Charles DF, Birks HJB, Camburn KE, Dixit SS, Uutala AJ, Selle AR. (1992) How much acidification has occurred in Adirondack region lakes (New York, USA) since preindustrial times?. Canadian Journal of Fisheries and Aquatic Sciences 49:128-141. DeSellas AM, Paterson AM, Sweetman JN, Smol JP. (2011) Assessing the effect of multiple environmental stressors on zooplankton assemblages from Boreal Shiled lakes since pre-industrial times. Journal of Limnology 70:41-56. Dixit SS, Smol JP, Charles DF, Hughes RM, Paulsen SG, Collins GB. (1999) Assessing water quality changes in the lakes of the northeastern United States using sediment diatoms. Canadian Journal of Fisheries and Aquatic Sciences 56:131-152. Downes BJ. (2010) Back to the future: little-used tools and principles of scientific inference can help disentangle effects of multiple stressors on freshwater ecosystems. Freshwater Biology 55:60-79. Downes BJ, Barmuta LA, Fairweather PG, Faith DP, Keough MJ, Lake PS, Mapstone BD, Quinn GP. (2002) Monitoring ecological impacts: concept and practice in flowing waters. Cambridge University Press, Cambridge, 452 p. 88 Driscoll CT, Driscoll KM, Roy KM, Dukett J. (2007) Changes in the chemistry of lakes in the Adirondack region of New York following declines in acidic deposition. Applied Geochemistry 22:1181-1188. Driscoll CT, Lawrence GB, Bulger AJ, Butler TJ, Cronan CS, Eagar C, Lambert KF, Likens GE, Stoddard JL, Weathers KC. (2001) Acidic deposition in the northeastern United States: sources and inputs, ecosystem effects, and management strategies. BioScience 51:180-198. Driscoll CT, Newton RM, Gubala CP, Baker JP, Christensen S. (1991) Adirondack mountains. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer-Verlag, New York, pp. 133-202. Enache MD, Paterson AM, Cumming BF. (2011) Changes in diatom assemblages since pre-industrial times in 40 reference lakes from the Experimental Lakes Area (northwestern Ontario, Canada). Journal of Paleolimnology 46:1-15. Foster D, Swanson F, Aber J, Burke I, Brokaw N, Tilman D, Knapp A. (2003) The Importance of Land-Use Legacies to Ecology and Conservation. BioScience 53:77-88. Garmo ØA, Skjelkvåle BL, de Wit HA, Colombo L, Curtis C, Fölster J, Hoffmann A, Hruška J, Høgåsen T, Jeffries DS, Keller WB, Krám P, Majer V, Monteith DT, Paterson AM, Rogora M, Rzychon D, Steingruber S, Stoddard JL, Vuorenmaa J, Worsztynowicz A. (2014) Trends in surface water chemistry in acidified areas in Europe and North America from 1990 to 2008. Water, Air, and Soil Pollution 225:1880-1-1880-14. George CJ. (1981) The fishes of the Adirondacks. New York State Department of Environmental Conservation, 93 p. Ginn BK, Rate M, Cumming BF, Smol JP. (2010) Ecological distribution of scaled-chrysophyte assemblages from the sediments of 54 lakes in Nova Scotia and southern New Brunswick, Canada. Journal of Paleolimnology 43:293-308. Hamilton AT, Barbour MT, Bierwagen BG. (2010) Implications of global change for the maintenance of water quality and ecological integrity in the context of current water laws and environmental policies. Hydrobiologia 657:263-278. Harig AL, Bain MB. (1998) Defining and restoring biological integrity in wilderness lakes. Ecological Applications 8:71-87. Hawkins CP, Olson JR, Hill RA. (2010) The reference condition: predicting benchmarks for ecological and waterquality assessments. Journal of the North American Benthological Society 29:312-343. 89 Hegerl GC, Zwiers FW, Braconnot P, Gillet NP, Luo Y, Marengo Orsini JA, Nicholls N, Penner JE, Stott PA. (2007) Understanding and attributing climate change. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds), Climate change 2007: the physical science basis Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingston, pp. 664-745. Herlihy AT, Paulsen SG, Van Sickle J, Stoddard JL, Hawkins CP, Yuan LL. (2008) Striving for consistency in a national assessment: the challenges of applying a reference-condition approach at a continental scale. Journal of the North American Benthological Society 27:860-877. Hesthagen T, Fjellheim A, Schartau AK, Wright RF, Saksgård R, Rosseland BO. (2011) Chemical and biological recovery of Lake Saudlandsvatn, a formerly highly acidified lake in southernmost Norway, in response to decreased acid deposition. Science of The Total Environment 409:2908-2916. Hobbs RJ, Arico S, Aronson J, Baron JS, Bridgewater P, Cramer VA, Epstein PR, Ewel JJ, Klink CA, Lugo AE, Norton D, Ojima D, Richardson DM, Sanderson EW, Valladares F, Vilà M, Zamora R, Zobel M. (2006) Novel ecosystems: theoretical and management aspects of the new ecological world order. Global Ecology and Biogeography 15:1-7. Hobbs RJ, Higgs E, Harris JA. (2009) Novel ecosystems: implications for conservation and restoration. Trends in Ecology & Evolution 24:599-605. Homer CH, Fry JA, Barnes CA. (2012) The National Land Cover Database, U.S. Geological Survey fact sheet. Earth Resources Observation and Science (EROS) Center, Sioux Falls, SD, 4 p. Hughes RM, Paulsen SG, Stoddard JL. (2000) EMAP-Surface Waters: a multiassemblage, probability survey of ecological integrity in the U.S.A. Hydrobiologia 422/423:429-443. Hunsaker CT, Hughes RM. (2002) Effects of landscape change on the physical and chemical components of aquatic ecosystems. In: Gutzwiller KJ (ed), Applying landscape ecology in biological conservation. SpringerVerlag, New York, pp. 286-308. Hyatt CV, Paterson AM, Cumming BF, Smol JP. (2010) Factors related to regional and temporal variation in the distribution of scaled chrysophytes in north-eastern North America: Evidence from lake sediments. Nova Hedwigia, Beiheft 136:87–102. 90 Jager JC, Loonman CWN. (1995) Data collection. Data analysis in community and landscape ecology. Cambridge University press, Cambridge, pp. 10-27. Jenkins J. (2010) Climate change in the Adirondacks: the path to sustainability. Cornell University Press, Ithaca, New York, 183 p. Jenkins J, Keal A. (2004) The Adirondack atlas: a geographic portrait of the Adirondack Park. Syracuse University Press, Syracuse, New Yor, 296 p. Jenson SK, Domingue JO. (1988) Extracting topographic structure from digital elevation data for Geographic Information System analysis. Photogrammetric Engineering and Remote Sensing 54:1593-1600. Jeziorski A, Keller B, Paterson A, Greenaway C, Smol J. (2013) Aquatic ecosystem responses to rapid recovery from extreme acidification and metal contamination in lakes near Wawa, Ontario. Ecosystems 16:209-223. Jeziorski A, Paterson AM, Smol JP. (2012) Crustacean zooplankton sedimentary remains from calcium-poor lakes: complex responses to threshold concentrations. Aquatic Sciences 74:121-131. Jones HP, Schmitz OJ. (2009) Rapid recovery of damaged ecosystems. PLoS ONE 4:e5653-1- e5653-6. Juggins S, Birks HJB. (2012) Quantitative environmental reconstructions from biological data. In: Birks HJB, Lotter AF, Juggins S, Smol JP (eds), Tracking environmental change using lake sediments: data handling and numerical techniques, Springer, Dordrecht, pp. 431-494. Keller W. (2009) Limnology in northeastern Ontario: from acidification to multiple stressors. Canadian Journal of Fisheries and Aquatic Sciences 66:1189-1198. Korhola A, Rautio M. (2001) Cladocera and other branchiopod crustaceans. In: Smol JP, Birks HJB, Last WM (eds), Tracking environmental change using lake sediments, Volume 4: Zoological indicators. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 5-41. Korosi JB, Smol JP. (2012) A comparison of present-day and pre-industrial cladoceran assemblages from softwater Nova Scotia (Canada) lakes with different regional acidification histories. Journal of Paleolimnology 47:4354. Kraft CE, Carlson DM, Carlson M. (2006) Inland fishes of New York (online). Version 4.0. Department of Natural Resources, Cornell University, and the New York State Department of Environmental Conservation. http:// fish.dnr.cornell.edu/nyfish/fish.html. 91 Krester W, Gallagher J, Nicolette J. (1989) Adirondack lakes survey 1984-1987: an evaluation of fish communities and water chemistry. Adirondack Lakes Survey Corporation, Ray Brook, NY, 437 p. Landers DH, Overton WS, Linthurst RA, Brakke DF. (1988) Eastern Lake Survey, regional estimates of lake chemistry. Environmental Science and Technology 22:128-135. Leira M, Jordan P, Taylor D, Dalton C, Bennion H, Rose N, Irvine K. (2006) Assessing the ecological status of candidate reference lakes in Ireland using palaeolimnology. Journal of Applied Ecology 43:816-827. Lepš J, Šmilauer P. (2003) Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge, United Kingdom, 284 pp. Liroff RA, Gordon GG. (1981) Protecting open space: land use control in the Adirondack Park. Ballinger Publishing Company, Cambridge, Massachusetts, 302 p. Martin SL, Hayes DB, Rutledge DT, Hyndman DW. (2011) The land-use legacy effect: adding temporal context to lake chemistry. Limnology and Oceanography 56:2362-2370. McMartin B. (1994) The great forest of the Adirondacks. North Country Books, Utica, New York, 254 p. Newton RM, Driscoll CT. (1990) Classificiation of ALSC lakes. Adirondack Lakes Survey: An interpretive analysis of fish communities and water chemistry, 1984-1987. Adirondack Lakes Survey Corporation, Ray Brook, New York, pp. 2-70-2-91. New York State Forest Commission (NYS Forest Commission) (1890) Map of the great forest of Northern New York showing boundaries (in red) of the forest area and boundaries (in blue) of the proposed Adirondack Park. New York State Archives Adirondack Museum Library, New York State. http://iarchives.nysed.gov/PubImageWeb/viewImageData.jsp?id=125178 New York State Department of Environmental Conservation (NYSDEC). (2012) Fish surveys (Adirondack Regional Geographic Information System). New York State Department of Environmental Conservation, Albany, New York. http://aprgis.org Paterson AM, Cumming BF, Smol JP, Hall RI. (2004) Marked recent increases of colonial scaled chrysophytes in boreal lakes: implications for the management of taste and odour events. Freshwater Biology 49:199-207. Quinlan R, Hall RI, Paterson AM, Cumming BF, Smol JP. (2008) Long-term assessments of ecological effects of anthropogenic stressors on aquatic ecosystems from paleoecological analyses: challenges to perspectives of lake management. Canadian Journal of Fisheries and Aquatic Sciences 65:933-944. 92 Renberg I, Bigler C, Bindler R, Norberg M, Rydberg J, Segerström U. (2009) Environmental history: a piece in the puzzle for establishing plans for environmental management. Journal of Environmental Management 90:2794-2800. Rogan J, Miller J. (2007) Integrating GIS and remotely senses data fro mapping forest disturbance and change. In: Wulder MA, Franklin SE (eds), Understanding forest disturbance and spatial pattern: remote sensing and GIS approaches. Taylor and Francis Group, Boca Raton, Florida, pp. 133-171. Roy K, Houck N, Hyde P, Cantwell M, Brown J. (2011) The Adirondack long-term monitoring lakes: a compendium of site descriptions, recent chemistry and selected research information. New York State Energy Research and Development Authority, Albany, 298 p. Scott NA, Likens GE, Eaton JS, Siccama TG. (2001) Trace metal loss following whole-tree harvest of a northeastern deciduous forest, USA. Biogeochemistry 54:197-217. Skaloš J, Engstová B, Trpáková I, Šantrůčková M, Podrázský V. (2012) Long-term changes in forest cover 1780– 2007 in central Bohemia, Czech Republic. European Journal of Forest Research 131:871-884. Skjelkvåle BL, Stoddard JL, Jeffries DS, Tørseth K, Høgasen T, Bowman J, Mannio J, Monteith DT, Mosello R, Rogora M, Rzychon D, Vesely J, Wieting J, Wilander A, Worsztynowicz A. (2005) Regional scale evidence for improvements in surface water chemistry 1990–2001. Environmental Pollution 137:165-176. Smol JP. (2008) Polluion of lakes and rivers: a paleoenvironmental perspective. Blackwell Publishing Ltd, Malden, 396 p. Smol JP. (2010) The power of the past: using sediments to track the effects of multiple stressors on lake ecosystems. Freshwater Biology 55 (Suppl. 1):43-59. Soranno PA, Wagner T, Martin SL, McLean C, Novitski LN, Provence CD, Rober AR. (2011) Quantifying regional reference conditions for freshwater ecosystem management: A comparison of approaches and future research needs. Lake and Reservoir Management 27:138-148. Stager JC, Sanger T. (2003) An Adirondack "heritage lake". Adirondack Journal of Environmental Studies 10:6-10. Stoddard JL, Larsen DP, Hawkins CP, Johnson RK, Norris RH. (2006) Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16:1267-1276. ter Braak CJF, Šmilauer P. (2012) CANOCO reference manual and user's guide: software for ordination (version 5.0). Microcomputer Power, Ithaca, New York, 496 p. 93 United States Environmental Protection Agency (USEPA). (2011) National Lakes Assessment 2012: a fact sheet for communities. USEPA Office of Water, Monitoring Branch, Washington, DC, EPA 841-F-11-007. United States Environmental Protection Agency (USEPA). (2012) Implications of climate change for state bioassessment programs and approaches to account for effects (final report) U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-11/036F. United States Geological Survey (USGS). (1999) National Elevation Dataset. United States Geological Survey, Sioux Falls, South Dakota. http://ned.usgs.gov/Ned/ned.html. United States Geological Survey (USGS). (2013) Nonindigenous aquatic species database. Gainesville, Florida. http:// nas.er.usgs.gov/queries/SpeciesList.aspx?group=Fishes&state=NY&Sortby#3 Uuemaa E, Antrop M, Roosaare J, Marja R, Mander Ü. (2009) Landscape metrics and indices: an overview of their use in landscape research. Living Reviews in Landscape Research 3:1. http://www.livingreviews.org/lrlr2009-1. Vinebrooke RD, Cottingham KL, Norberg M, Scheffer J, Dodson SI, Maberly SC, Sommer U. (2004) Impacts of multiple stressors on biodiversity and ecosystem functioning: The role of species co-tolerance. Oikos 104:451-457. Waller K, Driscoll C, Lynch J, Newcomb D, Roy K. (2012) Long-term recovery of lakes in the Adirondack region of New York to decreases in acidic deposition. Atmospheric Environment 46:56-64. Wang L, Brenden T, Seelbach P, Cooper A, Allan D, Clark R, Wiley M. (2008) Landscape based identification of human disturbance gradients and reference conditions for Michigan streams. Environmental Monitoring and Assessment 141:1-17. Wang X, Burns DA, Yanai RD, Briggs RD, Germain RH. (2006) Changes in stream chemistry and nutrient export following a partial harvest in the Catskill Mountains, New York, USA. Forest Ecology and Management 223:103–112. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. Yan ND, Girard R, Heneberry JH, Keller WB, Gunn JM, Dillon PJ. (2004) Recovery of copepod, but not cladoceran, zooplankton from severe and chronic effects of multiple stressors. Ecology Letters 7:452-460. Yates AG, Bailey RC. (2010) Selecting objectively defined reference sites for stream bioassessment programs. Environmental Monitoring and Assessment 170:129-140. 94 Tables & Table Captions Table 3.1. Summaries of lake and watershed characteristics, lake morphometrics, and water chemistry variables for 31 Adirondack reference lakes. Minimum (min.), maximum (max.), mean, and median of variables presented. Lake watershed land-cover & land-use characteristics are calculated as percent area (%) of the terrestrial watershed. Land cover & land use types are: deciduous forest (% deciduous), evergreen (coniferous) forest (% evergreen), woody wetland (% woody wetland), mixed forest (% mixed), shrubs/scrub (% shrub), emergent herbaceous wetland (% emergent), and developed land (open space) (% developed). Lake morphometric parameters are: elevation (m), surface area (ha), and maximum depth (Zmax, m). Chemistry variables are: pH, acid neutralizing capacity (ANC, µeq L-1), chloride (Cl-, mg L-1), total phosphorus (TP, µg L-1), dissolved organic carbon (DOC, mg C L-1), and calcium (Ca, mg L-1). BDL = below detection limit. Note TP values are from sites sampled in 2011 only (n = 20). Category Watershed Land Cover & Land Use Lake Morphometrics Water Chemistry Variable % deciduous % evergreen % woody wetland % mixed % shrub % emergent wetland % developed Elevation Surface Area Zmax pH ANC ClTP DOC Ca Min 6.3 0 0 0 0 0 0 349 4.7 5.2 5.9 35.3 0.11 0 (BDL) 2.1 0.5 95 Max 88.8 73.2 48.8 44.3 5.4 7.5 1.9 731 57 23.8 6.8 376.8 0.37 4 8.8 5.2 Mean 50.2 25.7 13.3 9.5 0.5 0.7 0.1 490 20.5 10.9 6.4 120.7 0.20 1.3 4.3 2.3 Median 51.5 15.9 9.2 4.8 0 0 0 486 13.9 10.1 6.3 85.8 0.17 1.0 3.9 2.0 Table 3.2. Summaries of early watershed disturbance and fisheries disturbance in 31 Adirondack reference lakes. Lakes are listed by their lake code given in Appendix C. A digital version of an 1890 forest commission map was consulted to identify the watersheds as virgin forest (VF), logged for softwoods (LS), denuded (DE), or cleared for agriculture (CL). GIS data of a digitized 1916 fire protection map were used to calculate the percent of each lake’s terrestrial watershed characterized as green timber, logged, burned, or cleared land (for agriculture or grazing) and assign the lake a score based on the type and amount of disturbance. The classes were sub-divided into ‘A’ lakes (lakes that were either virgin forest or logged for softwoods as of 1890) and ‘B’ lakes (lakes that were cleared or denuded as of 1890) and grouped into a disturbance class based on the 1916 GIS data, with Class 1 lakes being the least disturbed. Fisheries disturbance data documented whether or not the lake is known to have been reclaimed (Recl.; Y if reclaimed), and whether non-native piscivores (Non-nat.) were identified in the most recent NYSDEC fisheries surveys (Y if non-natives identified). Lakes that were reclaimed or had non-native piscivores in their more recent fisheries survey are denoted by a negative sign (-) in the lake class. Lake Code WolfL SampP StreL CoppP CascL MossL ArbL ENestL LongP IslaP USpecP GullP ChallP WilcL RockP EPineP FishP LFishP BassL CascP BootP RounP LSargP DeerP ClamP BessP NellP ClearP LydiP MidBL GrizO 1890 Forest Map VF VF VF LS VF VF VF VF VF CL CL CL DE VF VF LS LS LS DE LS VF VF VF VF LS LS LS LS LS LS & DE CL Green Timber 100 100 100 100 100 100 100 100 100 100 100 100 100 92.5 91.4 58.8 52.4 51.1 67.7 0 0 0 0 0 28.1 13.8 11.8 0.2 0 0 18.2 1916 Fire Map Logged Burned Cleared Land 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7.5 0 0 8.6 0 41.2 0 0.3 46.9 0.3 0.3 48.2 0.3 32.3 0 0 100 0 0 3.5 84.0 12.4 0 100 0 99.8 0.2 0 94.1 5.9 0 0 71.9 0 0 86.2 0 0 88.2 0 0 99.8 0 0 81.4 18.6 4.2 95.8 0 0 1.3 80.5 96 Score 0 0 0 0 0 0 0 0 0 0 0 0 0 -0.22 -0.26 -0.82 -0.95 -0.98 -0.32 -1.00 -2.09 -2.00 -1.00 -1.06 -1.44 -1.72 -1.76 -2.00 -2.19 -1.96 -2.44 Fisheries NonRecl. Nat. Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y - Lake Class 1A 1A 1A 1A 1A1A1A1A1A1B 1B1B1B2A 2A2A2A2A2B3A 3A 3A3A3A3A3A3A3A3A3B 3B- Figures & Figure Captions Figure 3.1. Conceptual diagram outlining the biological response of two lakes (Lake A and Lake B) to acidification and climate change since pre-industrial times. The lake response is shown across two gradients: one of increasing acidity and one of increasing air temperature. Lake A is a higher-pH/ANC ‘reference lake’ that is resistant to acidification. Lake B is a lake susceptible to acidification. The purpose of including a reference lake (Lake A) is to identify trends in species composition caused by regional warming (i.e., to characterize the ‘shifting baseline’). Historically, the influences of regional warming and acid deposition on the lakes are minimal. The lakes are similar in species composition (A hist, Bhist), though the taxa of Lake A are characteristic of higher pH/ANC conditions. With the onset of acid deposition, Lake B acidifies (solid arrow). Acid-sensitive taxa are lost and replaced with acid-tolerant ones (Bacid). Lake A does not acidify; its species assemblages do respond, however, to regional warming (solid arrow). There is an increase in warmer-water taxa, taxa that benefit from shorter-ice covered seasons, etc. (Apres). When the stress of acid deposition is reduced, there are four possible recovery trajectories for Lake B (dashed arrows): 1) Return to pre-disturbance state: this is only possible if Lake B does not respond to warming, which is unlikely; 2) No Recovery: Lake B responds to warming but shows no decline in acid-tolerant taxa; 3) Partial recovery: Lake B responds to warming and there is some decline in acid-tolerant taxa and/or increase in acid-sensitive taxa; 4) Recovery to a novel state: Lake B responds to warming and the abundance of acid-tolerant/acid-sensitive taxa returns to historic levels. However, regional warming has prevented the lake from returning to its pre-acidification state. Lake A and B are again similar in species composition, though the taxa of lake A remain characteristic of higher pH/ANC conditions. Conceptual diagram based ideas presented in Vinebrooke et al. (2004), Hobbs et al. (2009), Battarbee et al. (2012). 97 Figure 3.2. Conceptual spatial diagram of a paleolimnological reference site study. Reference sites (white circles), impacted sites (black circles), and sites that are neither reference nor impacted sites (grey circles) are distributed in a region. The reference sites are sampled as part of a top-bottom study (TB) and then a selection of reference sites are compared with a set of impacted sites in a down-core study (DC). 98 Figure 3.3. Schematic outlining the reference lake selection and characterization process followed in this investigation. Reference lakes were identified using 6 selection criteria from 1,469 lakes included in the Adirondack Lakes Survey. The selection criteria were: lakes had public access; pH > 6.5, ANC > 50 µeq L-1; % shoreline/watershed development < 5%; Cl < 2 mg L-1; no introduced piscivores (at time of sampling in ALS); maximum depth > 5 m, surface areas > 4 ha. The historic watershed disturbance experienced by each reference lake was qualitatively characterized using a digital version of an 1890 map and quantitatively characterized using GIS methods applied to a digitized 1916 map. Using the 1890 map, lakes were described as being part of a minimally disturbed forest in 1890 (‘A’ lakes) or a more disturbed area (‘B’ lakes). In the quantitative assessment, lakes were scored based on the amount and type of early watershed disturbance they experienced and sorted into disturbance classes (with Class 1 lakes being the least disturbed). Lastly, in the disturbance classes defined using the historic watershed maps, lakes that experienced reclamation and/or the introduction of non-native piscivores in a post-ALS fisheries survey were demarked by a negative (-) sign. The result was a series of lake classes that represent a gradation from least (Class 1A) to most (Class 3B-) disturbed reference lakes. 99 Figure 3.4. Map of the Adirondack Park (NY, USA) showing the locations of a set 31 Adirondack reference lakes minimally impacted by anthropogenic stressors. Map shows the position of the Adirondack Park in northeastern United States (enlarged area); inset shows lakes in the St. Regis Canoe area in detail – all other reference lakes are shown as points. 100 101 Figure 3.5. PCA ordinations of ALS survey data: A) PCA of morphological & chemical variables; B) PCA of chemical variables only; and C) PCA of morphological variables only. Morphological variables are elevation (elv, m) maximum depth (Zmax, m, log transformed), surface area (SArea, ha, log transformed), watershed area (WArea, ha, log transformed). Chemical variables are SO42-, F-, K+ (mg L-1, sqrt transformed), DOC (dissolved organic carbon, mg C L-1, sqrt transformed), pH (field pH measurement, sqrt transformed), Cl-, Na+, Ca2+, Fe2+, (mg L-1, log transformed), DIC (dissolved inorganic carbon, mg L-1, log transformed), TP (total phosphorus, mg L-1, log transformed) and specific conductance (SCOND, µmhos cm-1, log transformed). Reference sites (large circles) were used to demarcate a polygon in PCA space that contains sites that are similar in physical and/or chemical composition to the reference sites but failed to meet one or more selection criteria. These ‘polygon-sites’ are shown as grey crosses. Non-reference & non-polygon sites in the ALS are shown as points. Eigenvalues (λ) for PCA axis 1 and 2 are given. CHAPTER 4 ADIRONDACK (NY, USA) REFERENCE LAKES SHOW A PRONOUNCED SHIFT IN CHRYSOPHYTE SPECIES COMPOSITION SINCE CA. 1900 Abstract Lakes in both North America and Europe are undergoing chemical recovery from acidification and there is a pressing research need to define recovery targets for acid-impacted species assemblages. Researchers attempting to define such recovery targets are often hampered by two issues: 1) a lack of monitoring data; and 2) the confounding influence of multiple stressors, especially regional stressors like climate warming. Paleolimnological studies of minimally-disturbed reference sites can address both of these issues. By reconstructing reference lake histories, paleolimnologists can infer if/how regional stressors like climate change influence species assemblages independent of stressors such as acidification, eutrophication, and land-use change. This manuscript reconstructs changes in the scaled chrysophyte (Synurophyceae and Chrysophyceae) assemblages of 31 minimally-disturbed Adirondack (NY, USA) reference lakes from pre-ca. 1900 to present. The reference lakes have been minimally influenced by acidification, eutrophication, road salt seepage, and piscivore introductions, and so represent a unique opportunity to examine the effects of regional stressors in the Adirondack Park. First, the present-day distribution of chrysophyte taxa in the reference lakes was compared to measured limnological variables using Redundancy Analysis (RDA). RDA found that physical/chemical variables (summarized by PCA axis-scores) explained ~19% of the variation in chrysophytes species assemblage. S. petersenii, a taxon known to cause taste and odour issues in freshwaters, was common in deeper lakes with higher pH/ANC and DOC < ~5.2 mg C L-1. A top-bottom paleolimnological analysis examined changes in chrysophyte assemblages from pre-1900 to 102 present in the reference lakes. A significant shift in species composition from pre-1900 to present was found using both univariate (Wilcoxon signed-rank test) and multivariate (ANOSIM) analyses. Present-day assemblages were characterized by an increased abundance of colonial chrysophyte taxa, a trend that has been documented in other regions in North America. This study suggests that regional warming and/or oligotrophication are influencing the species assemblages of minimally-disturbed reference lakes, and so lakes recovering from acidification are unlikely to return to their pre-disturbance assemblages. Introduction Lakes are increasingly influenced by multiple anthropogenic stressors (Keller 2009), including introduced species, land-use change, acidification, eutrophication, and climate change. Multiple stressors can have both synergistic and antagonistic effects on aquatic species assemblages (Coors and De Meester 2008; Folt et al. 1999) and may lead to ‘ecological surprises’ (sensu Paine et al. 1998) in impacted systems. As such, systems affected by multiple stressors pose unique challenges to researchers, managers, and policy makers. Remediative strategies for environmental issues typically focus on a single stressor, often with an either explicitly stated or implied recovery target of a return to pre-disturbance state (e.g. clean air legislation targeting sulphur dioxide emissions, leading to declines in acid deposition and chemical recovery from acidification in some lakes in some regions, Garmo et al. 2014). However, there is a growing recognition that multiple stressors may prevent systems from returning to a pre-disturbance state, even with remediative action, a phenomenon referred to as the ‘shifting baseline’ (Bennion et al. 2011). In this regard, climate change represents a stressor of particular concern, given its regional nature. In a multiple stressor environment, climate change introduces additional complexities in the physical, chemical, and biological processes of 103 lakes (Smol 2010). While there are a growing number of case studies suggesting that regional warming may prevent lakes recovering from acidification or eutrophication from returning to their pre-disturbance state (e.g., Arseneau et al. 2011; Battarbee et al. 2012), the effect of climate change on recovery processes is surprisingly understudied (Verdonschot et al. 2013). There is a pressing research need to define reasonable recovery targets for impacted sites that account for changing climatic conditions. To address this issue, the United States Environmental Protection Agency (USEPA) recently called for long-term monitoring at minimally-disturbed reference sites across the United States (USEPA 2012). Reference sites (i.e., sites minimally impacted by multiple anthropogenic stressors, Bailey et al. 2004) are commonly used in bioassessment surveys to define recovery targets for impacted ecosystems. Long-term monitoring at reference sites can be used to monitor changes in baseline conditions due to regional stressors such as climate change which are occurring independent of more local stressors (e.g., land-use change, eutrophication, acidification, etc.). While laudable, the approach advocated by the USEPA is limited by the fact that longterm monitoring data of aquatic ecosystems are rare and, when they exist, typically cover relatively short periods (< 50 years). Here, paleolimnological studies can play an informative role. Paleolimnological studies of minimally-disturbed reference sites can identify regional changes in species composition that are occurring independent of many common stressors (for an early example, see Battarbee et al. 2002), information that can, in turn, be used to define recovery targets for impacted lakes in a region. For example, if a paleolimnological study of a set of regional reference sites were to identify increases in warm-water taxa or small planktonic algae over the past ~100-150 years (changes predicted to occur with regional warming; Winder and Sommer 2012), a recovery target for acid-impacted species assemblages in the region should 104 reflect those changes (Chapter 3). A variety of possible methods for comparing impacted and reference sites exist (Chapter 3), including both univariate methods (i.e. correlations between long-term monitoring data and key taxa in reference/impacted sites Battarbee et al. 2012) and multivariate methods (i.e. ordination techniques, Yan et al. 2004). The goal of this manuscript is to use a top-bottom paleolimnological approach to examine long-term changes in scaled chrysophyte assemblages (Synurophyceae and Chrysophyceae) in a set of 31 reference lakes in the Adirondack Park (NY, USA). The reference lakes were identified from an online database of 1,469 Adirondack lakes surveyed as part of the Adirondack Lakes Survey (ALS) using a set of stringent selection criteria (Krester et al. 1989). The reference lakes have been largely protected from several common stressors in the Adirondack Park, including: acidification (i.e. decreases in lake pH and ANC), eutrophication from watershed/shoreline development, leaching of road salt, and piscivore introductions (see Chapter 3). The reference lakes are not, however, ‘pristine’ (i.e. lakes protected from any anthropogenic influence, Bennion et al. 2011). Many of the sites have experienced historic watershed such as logging or fisheries disturbance such as reclamation. These forms of historic disturbance are common in the Adirondack Park and in other currently protected areas (Stager & Sanger 2003) and may have long-lasting chemical/biological legacies (Martin et al. 2011). While the reference sites may not be pristine, they have been protected from several common anthropogenic stressors and so can be used to identify what influence, if any, regional stressors such as climate change are having on Adirondack lakes; information that, in turn, can be used to define recovery targets for Adirondack lakes recovering from stressors such as acidification. Which that goal in mind, the following questions are addressed in this investigation: 105 1) What physical/chemical variables influence the present-day distribution of chrysophyte taxa in the reference lakes? Additionally, do historic watershed or fisheries disturbances influence the present-day distribution of chrysophyte taxa in the reference lakes? and, 2) Has there been a significant shift in chrysophyte species composition in the reference lakes since ca. 1900, both at the level of key individual taxa and the whole species assemblage? This manuscript is part of a series of papers whose goal is to highlight the important role paleolimnological studies of minimally-disturbed reference sites can play in defining reasonable recovery targets for lakes recovering from stressors such as acidification. This top-bottom study is aimed at identifying whether or not regional changes in chrysophyte species composition are occurring in minimally-disturbed Adirondack lakes and examine plausible mechanisms for any species changes noted. Subsequent articles will provide detailed down-core analyses of species assemblages in both reference and acidified lakes. Methods Study Sites: The lakes in this investigation are a set of 31 minimally-disturbed reference lakes in the Adirondack Park. The lakes were selected from an online database of 1,469 Adirondack lakes surveyed from 1984-1987 as part of the ALS. The lakes were chosen using a set of selection criteria designed to identify lakes that have been minimally impacted by common anthropogenic stressors (acidification, watershed development, road salt application, piscivore introductions). PCA-ordination showed that the reference lakes are representative of between 24 to 36% of the physical and/or chemical characteristics of the ALS lakes, a survey that is itself representative of 2,759 lakes in the Adirondack Ecological Zone (AEZ) (Chapter 3). Quantitative and qualitative analyses of historic maps and fisheries data found that many of the reference sites had experienced some form of historic disturbance since the late 1800s and 106 early 1900s (Chapter 3). An understanding of historic disturbance was used to group the reference lakes into classes. Qualitative analysis of an 1890 map showed that 77% of the lakes were located in minimally-disturbed forest (named ‘A’ lakes) whereas the remainder were located in cleared or denuded watersheds (‘B’ lakes). Quantitative (GIS) analysis of a 1916 map found that 42% of the reference lakes were listed in watersheds classed as 100% green timber (Class 1 lakes), of which 9 were considered minimally disturbed in 1890 (Class 1A). An additional 6 lakes (19%) had experienced at least one form of disturbance (logging, burns, or land-clearance) in less than 50% of their watershed (Class 2), and 12 lakes (39%) had experienced a least 1 form of historical disturbance in more than 50% of their watershed (Class 3). Forest fires were the most common form of anthropogenic disturbance experienced by the reference lakes in 1916, followed by selective logging and agricultural clearance. In terms of fisheries disturbance, as of 2011, 20 of the reference lakes (65%) had been reclaimed and 6 lakes (19%) had non-native piscivores identified in their most recent post-ALS fisheries survey. The reference sites are not ‘pristine’ (i.e., sites devoid of any anthropogenic influence, Bennion et al. 2011); indeed, given the pervasive influence of stressors such as atmospheric deposition and climate warming, the likelihood of identifying ‘pristine’ reference lakes is small. Rather, the Adirondack reference lakes reflect the Adirondack Park’s history of landscape disturbance and fisheries alterations (George 1981; McMartin 1994), disturbances that are common in many protected areas. Despite these early disturbances, the reference sites provide us with the opportunity to examine regional changes in species composition independent of many common contemporary stressors. Furthermore, the fact that some (but not all) of the reference sites have experienced either historic watershed and/or fisheries disturbance allows us to test if these early disturbances influence the present-day distribution of species assemblages. 107 Field Collection: Sediment cores were collected from the deepest basin of the study lakes in either summer 2010 or 2011 using a 7.6 cm (ID) Glew gravity corer, and were sectioned on-site into 0.25 cm using a vertical Glew extruder (Glew et al. 2001). Water samples were collected by hand at ~0.5 m below the water’s surface in glass and polypropylene jars. All chemistry analyses were performed by the Center for Environmental Systems Engineering (Syracuse University, NY) using standard US EPA methods (Chapter 3). Sample preparation: In a top-bottom paleolimnological investigation, a surface-sediment sample is taken to represent the present-day species assemblage and compared to a sediment sample taken from further down in the sediment core to represent pre-industrial times (Cumming et al. 1992b). The technique is an effective way to build a regional ‘snap-shot’ of species changes over time. In this investigation, the ‘top’ sample is an integrated sample from 0-1 cm in the sediment core and is estimated to represent the past 2-3 years. The ‘bottom’ sample is a 0.25 cm sediment subsection from ~20 cm (in 4 cases where the sediment core length was < 20 cm, the bottommost sediment interval was used; 3 cases ~17-18 cm, 1 case ~15 cm). A single 0.25 cm subsection was analyzed at ~20 cm, rather than a 1 cm subsection, as a decrease in water content in the sediment occurs downcore which increases the amount of ‘time’ represented per sediment subsection at this depth. A depth of 20 cm was chosen because a survey of recently 210Pb-dated cores from the Adirondacks suggests that supported 210Pb is reached at ~20 cm in the majority of Adirondack region lakes (Arseneau et al. 2011, Arseneau et al. unpublished; see also Cumming et al. 1992b; Cumming et al. 1994). Conservatively, we refer to the time period at ~20 cm in the sediment core as ‘pre-ca. 1900’, rather than pre-industrial times (ca. 1850), as we recognize that differences in sedimentation rate may alter the depth at which unsupported 210Pb is no longer 108 present. An additional set of samples from a depth of 30 cm was analyzed in the 10 reference lakes that had a sufficiently high concentration of chrysophyte scales to be counted to assess the variability of sedimentary assemblages from pre-1900 (see Reproducibility study below). Sediment samples were prepared for enumeration of chrysophyte scales using standard methods (Battarbee et al. 2001). Briefly, sediment samples were digested in a 1:1 molar mixture of strong H2SO4:HNO3. Samples were heated to ~70°C for 6-7 hours and were then rinsed repeatedly until samples reached the pH of distilled water (litmus test). Samples were plated as a series of four 100x dilutions on microscope slides using Naphrax®. Chrysophyte scales were identified using DIC optics under oil-immersion using a 100X Fluotar objective with a numerical aperture of 1.3. In cases where scales were sparse, a minimum of 200 scales were counted (Ginn et al. 2010). In cases were scales were abundant, a minimum of 400 scales were counted. A small number of pre-1900 samples were considered uncountable due to low scale concentration (i.e. < 10 scales encountered per transect). The principle taxonomic references referred to in this investigation are: Siver (1991), Nicholls and Gerrath (1985), and Kling and Kristiansen (1983), as well as unpublished photo reference materials. Small Mallomonas spp. scales that could not be identified using light microscopy were group into a Mallomonas ‘small’ category (Cumming et al. 1992a). Statistical Analyses Understanding factors related to modern chrysophyte distribution: A Redundancy Analysis (RDA) was used to examine the influence of water chemistry and lake morphometric variables on the present-day distribution of chrysophyte taxa in the Adirondack reference lakes. RDA, a linear direct ordination method, was selected because an initial Detrended Correspondence Analysis (DCA) of the present-day samples for the 31 reference lakes indicated that the length of 109 the longest species turnover gradient was short (< 3 s.d.) and so linear ordination techniques were deemed appropriate (Lepš and Šmilauer 2003). Because of the relatively small number of surface samples in this investigation (31 samples), the environmental data matrix was summarized as PCA-scores of the first four PCA axes (Lepš and Šmilauer 2003). The normality of environmental variables was assessed using the Shapiro-Wilks test prior to inclusion of the environmental variables in the PCA. Non-normal environmental variables were transformed using either square-root or log transformations, and were not included if transformation did not result in a normal distribution. Correlations between normal input variables were all less than 0.8. All environmental variables (PCA axis scores) were used in an initial RDA to test if the combined set could explain a significant proportion of the species variance along all canonical RDA axes (ter Braak and Šmilauer 2012). If so, then forward selection using Monte Carlo permutation tests was used to identify a minimum set of environmental variables that could explain a significant proportion of the species assemblage data. Shapiro-Wilks tests and correlation coefficients were determined using the computer program SYSTAT v. 11. PCA, RDA (on square-root transformed species data) and forward selection procedure were performed using the computer program CANOCO v. 5.0 (ter Braak and Šmilauer 2012). The species data were square-root transformed to reduce the importance of the dominant taxon Mallomonas duerrschmidtiae, a common Adirondack taxon with a wide tolerance to many environmental variables (Cumming et al. 1992a). Lastly, an Analysis of Similarities (ANOSIM) was used to test for a significant difference in species composition in the present-day species assemblages of the reference lakes grouped into different disturbance classes. ANOSIM is a non-parametric test of rank similarities that can be used to test for a significant difference in species composition between a priori defined 110 groups (Clarke 1993). The Similarities Percentages test, SIMPER, is then used to identify which species make an important contribution to the difference seen between groups (Clarke and Warwick 2001). In this investigation, if no significant difference is seen between different lake classes, it can be concluded that historic watershed/fisheries disturbances have minimal influence on the present-day chrysophyte assemblages of the reference lakes. ANOSIM was used to compare present-day chrysophyte taxa in: Class 1 lakes (minimally disturbed, n = 13) vs. Class 3 lakes (at least 1 form of disturbance in >50% of watershed, n = 12) and non-reclaimed lakes (n = 11) vs. a randomly-selected subset of reclaimed lakes (n = 11). A randomly selected subset of reclaimed lakes was used to ensure a balanced study design. ANOSIM (using Bray-Curtis similarities) was performed on square-root transformed species data using the computer program Primer v. 6.1.11. In all of the analyses listed above (and subsequently), taxa that reached > 2% relative abundance in more than two sediment samples were included in the analyses. This cut-off criterion was used primarily to exclude rare taxa, as they can have a large influence in ordinations with a small number of samples. Because the interest in this study are the general controls on chrysophyte distribution and general trends in the reference lakes over time, we opted to set a selection criterion that would eliminate rare taxa. Summarizing long-term species changes: A chrysophyte-inferred pH (CI-pH) was calculated for present-day and pre-1900 sediment intervals using an Adirondack calibration set based on 71 lakes that spanned a pH gradient from just over 4 to ~8 (Cumming et al. 1992a). CI-pH was calculated using a ln+1 species transformation, with tolerance down-weighing and inverse deshrinking (Cumming et al. 1994) using the program C2 v. 1.72. 111 For ease of interpretation, sedimentary assemblage data were summarized using ordination techniques. An initial DCA of present-day and pre-1900 samples for the 26 reference lakes with sufficient scales in the pre-1900 sample found that the length of the longest gradient was short (< 3 s.d.) and so linear ordination techniques were appropriate (Lepš and Šmilauer 2003). The sedimentary assemblages were summarized in a PCA ordination diagram using square-root transformed species data using CANOCO v. 5.0 (ter Braak and Šmilauer 2012). A summary diagram was composed listing the relative abundance of all the taxa in the samples, with lakes listed by PCA-1 axis scores of a PCA with all 31 surface sediments (Appendix D). Testing for a significant change in species composition from pre-1900 to present: Changes in species composition from pre-1900 to present were analyzed in the reference lakes using both univariate and multivariate techniques. Previous paleolimnological investigations have identified recent increases in colonial chrysophyte taxa in Ontario (Flear 2011; Paterson et al. 2004; Paterson et al. 2008) and Maritime Canada (Ginn et al. 2010), increases which the authors suggest may be due, at least in part, to regional warming. As such, in this investigation a Wilcoxon signed-rank test (a non-parametric analog of a paired t-test) was used to identify if there has been a significant increase in the relative abundance of colonial taxa in the reference lakes from pre-1900 to present. Subsequently, an ANOSIM test was used to test for a significant difference in species composition in the overall assemblages from pre-1900 to present. Lakes with chrysophyte concentrations too low to count in the pre-1900 interval were excluded from the analyses. Reproducibility study (counting variability and stability analysis): An implicit assumption of the top-bottom analysis is that differences in species assemblages noted between present-day and 112 pre-industrial (or, pre-1900) times are greater than those arising from counting variability and, as such, can likely be attributed to some form of environmental change. To test this assumption, in 2012 quadruplicate sediment cores from the deepest basin of 4 reference lakes were collected. Sediment samples from the core tops (0-1 cm), 20 cm, and, when possible, 30 cm were prepared and analyzed for each core using the standard methods listed above. Prior to counting, the sediment samples were given a non-descript identifier (sample I, sample II, etc.) to reduce possible counter bias. Taxa that reached a relative abundance > 2% in more than 2 samples were used to calculate Bray-Curtis (BC) similarities between groups of samples (see below). To test the assumption that the differences between top-and-bottom reference site samples were greater than expected counting variability, quadruplicate top samples were compared within each lake group in a triangle similarity matrix of Bray-Curtis similarities (see Figure 4.1). Then, the average BC similarity of all top sample comparisons was calculated (referred to as ABC-top). This average value represents the expected similarity between samples when variation is due primarily to counting variability (and core collection/sample preparation). Therefore, in the top-bottom analysis of the regional reference lakes, a reference lake with a topbottom BC-similarity that is less than ABC-top is unlikely to have a difference in species composition due to counting variability alone. A second implicit assumption of the top-bottom approach is that the sedimentary assemblages were relatively stable pre-industrially (or, here, pre-1900). To test this assumption, BC-similarities were calculated for all top, 20 cm, and 30 cm samples from all replicate cores within each lake group. Triplicate cores of sufficient length (> 30 cm) were collected from three of the reference lakes. A Wilcoxon signed-rank test was performed to examine if a significant difference existed between similarities calculated between top vs. 20 cm samples, and 20 vs. 30 113 cm samples. Presumably, if some form of environmental change has acted on the species assemblages over the past 100-150 years, the species assemblages should be less similar between the top vs. 20 cm samples than the 20 vs. 30 cm samples. To further examine this assumption, samples from a core depth of 30 cm were analyzed in all reference lakes from the regional topbottom study where sediment cores > 30 cm in length were collected. Ten lakes had a sufficiently high concentration of scales at 30 cm to be counted. ANOSIM was used to test for a significant difference in species composition samples from 20 and 30 cm in the 10 lakes. If no significant difference in species composition is found between the two groups, it can be concluded that the assemblages were relatively stable prior to 1900; or, at least, that the variation between the time periods is less than the regional variation in species assemblages captured by the 10 lakes. Results Controls on chrysophyte species distribution: PCA plots based on chemical/morphological data captured approximately 72% of the environmental variation when summarized in 4 PCA axes (See Figure 4.2A & 4.2B). The first PCA axis is related to ion concentration and ANC, PCA axis-2 is related to Zmax and DOC/TN (with pH loading positively on both axis 1 and axis 2), PCA axis-3 is related primarily to SO42-, and PCA axis-4 is related to elevation and K+. Unfortunately, filtered TP values were only available for lakes sampled in 2011 and so TP could not be included in the PCA. However, in the 2011 sites, TP was significantly positively correlated with TN (r = 0.5), and so higher TN sites were also likely associated with higher TP conditions. The initial RDA using all environmental variables explained a significant proportion of the species variance along all canonical axes (p = 0.002, variance explained = 24.7%). The 114 forward selection procedure using Monte Carlo permutation tests found that 2 environmental variables (PCA-axis scores) explained a significant proportion of the variation in the chrysophyte assemblages in the present-day reference lake assemblages (PCA-2: F = 3.6, p = 0.002; PCA-1: F = 2.8, p = 0.007). The first and second RDA axes explained approximately 19.1% of the variance in the species assemblage data, when constrained to the PCA axis 1 and 2 scores (See Figure 4.2C). The main separation in the redundancy analysis fell along a pH gradient and was between lakes characterized by a higher relative abundance of S. petersenii and lakes characterized by a higher relative abundance of S. sphagnicola. Lakes with a higher relative abundance of S. petersenii were associated with higher PCA-1 & PCA-2 scores (i.e. deeper lakes with higher pH values and lower DOC values). In general, lakes characterized with a higher relative abundance of S. sphagnicola were associated with lower pH values. While some lakes with a high relative abundance of S. sphagnicola were associated with low PCA-2 scores (i.e. shallower, high DOC lakes; e.g. Streeter Lake), others were associated with high PCA-2 scores (e.g. Deer Pond). ANOSIM did not detect a significant difference in the present-day chrysophytes assemblage between lakes that have been reclaimed (R = 0.02, p = 0.32) or that had experienced early watershed disturbance (Class 1 vs. Class 3: R = 0.001, p = 0.41). The ANOSIM tests had a large enough sample size to compute a p < 0.05 if a significant difference existed between groups (# permutation >> 999, Clarke and Warwick 2001). However, the similarity within each group was low for each comparison (reclaimed: 66.2%, non-reclaimed: 51.2%; Class 1: 59.6%, Class 3: 56.5%), due to the regional variation in chrysophyte assemblages. It may be that chrysophyte assemblages are influenced by reclamation/early disturbance but that the sample sizes used in the ANOSIM tests (n = 11-13) are too low to detect that influence, given the 115 variability between sites. Thus, while it is possible that reclamation/early disturbance have no influence on present-day chrysophyte assemblages, a more conservative conclusion is that the influence of the stressors is small and less than the regional variability captured by approximately one third of the reference sites. Long-term changes in species composition: The first two axes of the PCA ordination explained approximately 46% of the species variation in the present-day and pre-1900 assemblages of 26 Adirondack reference lakes (See Figure 4.3). In general, the relative abundance of colonial chrysophyte taxa has increased in a majority of the reference lakes from pre-1900 to present (See Figure 4.3 & Appendix D). The only colonial chrysophyte that generally had a higher relative abundance in the pre-1900 samples was Synura curtispina. Unicellular taxa such as M. duerrschmidtiae, Mallomonas pseudocoronota, Mallomonas crassisquama, and Mallomonas lychenensis have generally declined in relative abundance from pre-1900 to present. In contrast, Mallomonas punctifera appears to have increased since the 1900s, as has Mallomonas elongata. Mallomonas hamata, an acid-tolerant chrysophyte species (Cumming et al. 1992a), was not abundant in the reference lake dataset (i.e. relative abundance < 5%), although it has shown a small increase in relative abundance in some lakes. Lastly, CI-pH increased, decreased, or remained unchanged between the pre-1900 and present-day samples (See Appendix D). Only two lakes (Middle Branch Lake and Bass Lake) showed a decline in CI-pH greater than the RMSEP (0.51), suggesting that the majority of the reference lakes are unlikely to have acidified. Wilcoxon signed-rank test found that the relative abundance of colonial chrysophytes was significantly higher in the present-day than pre-1900 samples (Z = 3.9; p < 0.0001) (See Figure 4.4). ANOSIM also showed a significant shift in species composition from pre-1900 to present (R = 0.12; p = 0.002). The low R-coefficient is due to the low similarity between lakes 116 within each time period (average similarity present-day: 62.3; average similarity pre-1900: 62.1). However, the fact that there is a significant difference between time periods suggests that the change in species composition across time periods is greater than the variability in species assemblages seen spatially in the Adirondack Park in pre-1900 or present-day samples. The SIMPER test found that recent increases in the relative abundances S. petersenii and S. sphagnicola in the present-day samples make the largest contribution to the difference in species composition between the two time periods (11.2% and 10.5% contribution, respectively) (See Table 4.1). Reproducibility & stability analysis: Quadruplicate cores from 4 reference lakes found that the average BC similarity of replicate top samples (ABC-top) was 89.0% (range: 82.5% to 94.7%). In comparison, all BC-similarities calculated between the 26 top and bottom reference lake samples were less than the ABC-top value (average: 57.7%; range: 17.2% to 83.5%) and only one top-bottom comparison had a similarity greater than the minimum value calculated between any replicate top samples (82.5%). Thus, the species changes documented between top and bottom samples in the reference lake study are generally greater than expected differences due to counting variability alone. Wilcoxon signed-rank test showed that the BC similarities were lower between replicate top and 20 cm than replicate 20 and 30 cm samples (Z = 3.4, p < 0.0001). Similarly, ANOSIM did not find a significant difference between the 20 cm and 30 cm samples of 10 top-bottom reference lakes (R = -0.09, p = 0.95), suggesting that the assemblages were relatively stable between 20 and 30 cm or, at least, that the variability between the two time periods was less than the regional variability in species assemblages captured by the 10 reference sites. 117 Discussion Controls on present-day chrysophyte distribution in reference lakes: Even though the pH gradient in the reference lakes was relatively narrow (~5.9 to 6.8), the pH/ANC gradient captured by PCA axes 1 and 2 was tightly correlated with RDA axis-1. Numerous studies have shown the importance of pH for determining chrysophyte species distribution, both in the Adirondacks and elsewhere (Cumming et al. 1992a; Hyatt et al. 2010; Paterson et al. 2001; Siver and Hamer 1989). In this investigation, lakes with higher pH values and deeper maximum depths (higher PCA-1 & PCA-2 scores) were characterized by a higher relative abundance of S. petersenii. Interestingly, Flear (2011) found a similar separation in a study of 40 minimally-disturbed lakes in the Experimental Lakes Area. Deep, high-pH lakes were characterized by a higher relative abundance of S. petersenii and shallower, low-pH lakes characterized by a higher relative abundance of S. sphagnicola, M. acaroides, and M. punctifera. In the Adirondack reference lakes, S. sphagnicola was abundant both in shallower, high DOC sites (low PCA-2 scores) and deeper, low DOC sites, suggesting that pH rather than DOC/depth per say was more important in determining the abundance of S. sphagnicola. In contrast, a close examination of the DOC data reveals that S. petersenii was less abundant (< ~10% relative abundance) in lakes with a DOC concentration greater than 5.2 mg C L-1, which corresponds to a photic zone depth of ~5.2 m (Bukaveckas and Robbins-Forbes 2000) (See Appendix E). Therefore, high DOC levels may exclude S. petersenii from some reference lakes. The fact that the RDA only explained approximately 19.1% of the variation in the present-day chrysophyte assemblages suggests that one or more important variables were not included in the analysis. Lake thermostructure, for example, is known to influence chrysophyte species composition. Some chrysophyte taxa are considered warm-water species (M. punctifera) 118 and during long periods of thermal stratification, large metalimnetic populations of both colonial (Synura, Chrysosphaerella) and unicellular (Mallomonas caudata, M. acaroides, M. crassisquama) taxa can occur (Nicholls 1995; Sandgren 1988; Siver 1995; Siver 2003). The fact that maximum depth was important for structuring chrysophyte assemblages in the reference lakes suggests that the thermostructure of lakes is likely an important control on chrysophyte distribution, as shallower lakes are often associated with higher surface-water temperatures and may be more prone to summertime mixing (Scheffer 2004). Biological factors may also be an important control on chrysophyte species distribution. Large colonial chrysophytes likely have a size refuge from small herbivores but may be susceptible to predation by large daphniids (Sandgren and William 1995). A future paleolimnological investigation of the reference lakes will include cladoceran zooplankton as paleo-indicators and so it will be possible to determine what influence, if any, the presence/abundance of Daphnia has on chrysophyte species distributions in the reference lakes. The Adirondack reference lakes have experienced a significant shift in species composition from pre-1900 to present, with the present-day samples characterized by an increased abundance of colonial taxa (detailed discussion below). The RDA results suggest that both physical and chemical characteristics may control the distribution of chrysophyte taxa. Of particular interest are the controls on S. petersenii, a taxon that has been shown to cause taste and odour issues in lakes and reservoirs (Nicholls and Gerrath 1985; Watson et al. 2001). The results of this investigation suggest that S. petersenii is more common in deeper lakes with higher pH values and DOC concentrations < ~5.2 mg C L-1. Nuisance S. petersenii blooms are thus unlikely to be a concern in the many Adirondack lakes which are small (SA < 4 ha, Zmax < 5 m), relatively acidic (pH < 6), and/or higher in DOC concentrations (DOC > 5 mg C L-1) (Krester et al. 1989). 119 However, lakes which support sportfish like lake trout are often fairly large, deep, and characterized by higher ANC and hence may provide suitable habitat for S. petersenii. Increases in colonial chrysophyte taxa (and S. petersenii in particular) in these sites may therefore pose a management concern (Paterson et al. 2004), and so it is important to discuss the possible causal mechanisms behind the increases in colonial chrysophyte taxa noted in this investigation. Future top-bottom and down-core paleolimnological studies of the reference lakes will build on the discussion this manuscript provides by employing additional paleo-indicators and comparing long-term changes in species composition with long-term monitoring data. Changes in chrysophyte composition from pre-1900 to present: The reproducibility analysis suggests that changes between the present-day and pre-1900 assemblages of the reference lakes are unlikely to be due to counting variability. Similarly, the stability analysis suggested that the change in chrysophyte species assemblages between the present day and 20 cm sample was greater than changes earlier in the record (between 20 and 30 cm). Thus, we believe that the shift in chrysophyte composition between 20 cm and the present-day samples in the top-bottom study is interpretable and merits investigation. Both univariate and multivariate analyses found that there has been a significant increase in the relative abundance of colonial taxa from pre-1900 to present, along with increases in some unicellular taxa in some lakes (M. punctifera, M. elongata). Because of the regional distribution of the reference lakes, we can conclude that a regional rather than local stressor is likely responsible for the changes in chrysophyte species composition. Furthermore, because of the reference lake study design, we can conclude that these increases are not due to: acidification, eutrophication (from shoreline or watershed development), salinization, or introduced piscivores. 120 ANOSIM results also suggest that both historic fisheries disturbance and watershed disturbance have minimal influence on the present-day chrysophyte taxa in the reference lakes. Given the large impact atmospheric deposition has had in the Adirondack region, it is important to consider whether or not the changes in chrysophyte species composition noted in this investigation could be due to long-term effects of acid deposition. CI-pH reconstructions suggest that the majority of the reference lakes are unlikely to have acidified (i.e. experienced decreases in pH/ANC) but atmospheric deposition may have effects outside acidification. Increased leaching of cations with acid deposition is unlikely to be an important driver of the species changes noted from pre-1900 to present as chrysophyte taxa are broadly tolerant of a range of cations, though increased long-distance transport and deposition of certain trace metals may be important (i.e. Se, a micronutrient shown to stimulate chrysophyte blooms, Sandgren 1988). Long-term exposure to acid deposition can also increase the ionic strength of soil solutions, decreasing DOC flux to nearby waterbodies (see Monteith et al. 2007 and citations therein). Decreases in DOC flux, in turn, would increase light penetration in the water column and reduce suboptimal low-light conditions for metalimnetic colonial chrysophyte populations (Healy 1983). However, while S. petersenii may have benefitted in lakes where DOC decreased below 5.2 mg C L-1, S. sphagnicola and S. echinulata appear to be DOC-indifferent and so changes in DOC concentrations cannot explain increases in those taxa. Finally, long-term exposure to acid deposition may decrease the flux of phosphorus to lakes by increasing the adsorption of TP in watershed soils (Eimers et al. 2009). Furthermore, logging and subsequent forest regrowth may also decrease TP export to nearby lakes, and many of the reference sites experienced early logging disturbances (Hall and Smol 1996). The fact that M. lychenensis and S. curtispina decreased in abundance from pre-1900 to present in some lakes suggests that total 121 phosphorus (TP) concentrations may have decreased in those sites, as both taxa are associated with higher lake trophic status (Siver and Marsicano 1996; Siver 1995). Furthermore, in low-TP conditions, the metalimnetic position of colonial chrysophytes may provide them a competitive advantage by allowing access to nutrient-rich hypolimnetic waters (Nicholls 1995). Interestingly, though unicellular taxa generally declined in relative abundance from pre-1900 to present, a few lakes showed increases in the relative abundance of M. caudata and M. elongata, both of which can form metalimnetic blooms (Fee et al. 1978; Siver 2003). Thus, oligotrophication may have contributed to the changes in chrysophyte assemblages noted in the reference lakes. Regional warming may have also contributed to the changes in species composition noted in the reference lakes from pre-1900 to present. The Adirondack region has warmed are a rate of 1.3ºC per 100 years (Jenkins 2010), a warming trend associated with an increased ice-free period in Adirondack lakes (Beier et al. 2012). The increases in M. punctifera (a warm-water taxon, Siver 1991) noted in some reference lakes may therefore be due to warming surface waters. Similarly, an extended ice-covered season may benefit taxa like S. sphagnicola which persists in the water column through summer and fall and is lost at ice-on (Siver and Hamer 1992). Longer ice-free periods can also lead to a longer period of summer stratification (Vincent 2009), which may provide motile algae like chrysophytes a competitive advantage over non-motile algae like diatoms because they can seek out optimal temperature/nutrient/light conditions in the water column (Winder and Sommer 2012). At the assemblage level, longer periods of stratification may provide a competitive advantage to colonial chrysophyte taxa over unicellular ones because large flagellates tend to be superior swimmers, with faster swim velocities and larger migration amplitudes (Sommer 1988). Longer periods of stratification can also result in increased nutrient limitation in the epilimnion (Vincent 2009), providing chrysophytes that form metalimnetic 122 blooms with a competitive advantage because of their access to nutrient-rich hypolimnetic waters. Thus, the increases in colonial taxa noted in this investigation, along with increases in certain unicellular taxa like M. punctifera, are likely related at least in part to regional warming. It is possible that the long-term effects of regional warming and/or acid deposition may have also resulted in changes in biotic conditions that could have contributed to the increased abundance of colonial chrysophytes. Large colonial chrysophytes likely have a size refuge from small zooplankton (Bosmina, calanoid and cyclopoid copepods) but can be predated upon by large Daphnia taxa (i.e. > 1 mm in size) (Sandgren and William1995). It has been hypothesized that regional warming may benefit small zooplankton (Daufresne et al. 2009). Furthermore, long-term exposure to acid deposition may lead to calcium depletion in watershed soils (Likens et al. 1996), a phenomenon that can negatively affect Ca-sensitive Daphnia (Jeziorski et al. 2008). A subsequent paleolimnological study of cladoceran assemblages in the reference lakes will determine whether or not cladoceran assemblages have undergone a significant shift in species composition from pre-1900 to present and what link, if any, those changes may have to changes in colonial chrysophyte abundance in Adirondack lakes. Importance of reference lake results for recovery studies: Arguably, the most important finding of this top-bottom study is simply that minimally-disturbed reference lakes in the Adirondack Park show a pronounced shift in species composition from pre-1900 to present, likely due to climate warming and/or oligotrophication. Thus, it should be expected that lakes in the area recovering from acidification will not return to their pre-disturbance state. Rather, the recovery endpoints for acidified lakes will be one that is different from their pre-disturbance condition – a state characterized by an increased abundance of colonial and/or warm-water chrysophyte taxa. By repeating the top-bottom analysis of the reference sites with other paleo-indicators (diatoms, 123 cladocera, chironomids, etc.), we will be able to determine whether or not such shifting baselines are occurring in a suite of aquatic organisms. Moreover, by pairing both acidified and reference lakes in subsequent paleolimnological studies, we will be able to determine whether or not the recovery trajectories of impacted sites suggest that they are moving towards a novel ‘recovered’ condition similar to the reference sites (Chapter 3). While a growing number of studies have shown chemical recovery from acidification, lakes are increasingly influenced by multiple anthropogenic stressors, making the return of biological assemblages to a pre-disturbance state unlikely. Paleolimnology can and should play an important role in recovery studies because paleolimnological techniques can both identify species changes in impacted sites and, as we have illustrated, be applied in reference lakes to identify shifting baselines and define reasonable bounds for recovery targets. Literature Cited Arseneau KMA, Driscoll CT, Brager LM, Ross KA, Cumming BF. (2011) Recent evidence of biological recovery from acidification in the Adirondacks (New York, USA): a multiproxy paleolimnological investigation of Big Moose Lake. Canadian Journal of Fisheries and Aquatic Sciences 68:575-592. Bailey RC, Norris RH, Reynoldson TB. (2004) Bioassessment of freshwater ecosystems: using the reference condition approach. Kluwer Academic Publishers, New York, 170 p. Battarbee RW, Anderson NJ, Bennion H, Simpson GL. (2012) Combining limnological and palaeolimnological data to disentangle the effects of nutrient pollution and climate change on lake ecosystems: problems and potential. Freshwater Biology 57:2091-2106. Battarbee RW, Jones VJ, Flower RJ, Cameron NG, Bennion H, Carvalho L, Juggins S. (2001) Diatoms. In: Smol JP, Birks HJB, Last WM (eds), Tracking Environmental Change Using Lake Sediments. Volume 3:Terrestrial, Algal, and Siliceous Indicators. KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. 155-202. 124 Battarbee RW, Thompson R, Catalan J, Grytnes J-A, Birks HJB. (2002) Climate variability and ecosystem dynamics of remote alpine and arctic lakes: the MOLAR project. Journal of Paleolimnology 28:1-6. Beier CM, Stella JC, Dovčiak M, McNulty SA. (2012) Local climatic drivers of changes in phenology at a borealtemperate ecotone in eastern North America. Climatic Change 115:399-417. Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. Bukaveckas PA, Robbins-Forbes M. (2000) Role of dissolved organic carbon in the attenuation of photosynthetically active and ultraviolet radiation in Adirondack lakes. Freshwater Biology 43:339-354. Clarke KR. (1993) Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18:117-143. Clarke KR, Warwick RM. (2001) Change in marine communities: an approach to statistical analysis and interpretation, 2nd edition. PRIMER-E Ltd. , Plymouth, United Kingdom, 172 pp. Coors A, De Meester L. (2008) Synergistic, antagonistic and additive effects of multiple stressors: predation threat, parasitism and pesticide exposure in Daphnia magna. Journal of Applied Ecology 45:1820-1828. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Cumming BF, Smol JP, Birks HJB. (1992a) Scaled chrysophytes (Chrysophyceae and Synurophyceae) from Adirondack drainage lakes and their relationship to environmental variables. Journal of Phycology 28:162178. Cumming BF, Smol JP, Kingston JC, Charles DF, Birks HJB, Camburn KE, Dixit SS, Uutala AJ, Selle AR. (1992b) How much acidification has occurred in Adirondack region lakes (New York, USA) since preindustrial times?. Canadian Journal of Fisheries and Aquatic Sciences 49:128-141. Daufresne M, Lengfellner K, Sommer U. (2009) Global warming benefits the small in aquatic ecosystems. Proceedings of the National Academy of Sciences 106:12788-12793. 125 Eimers MC, Watmough SA, Paterson AM, Dillon PJ, Yao HX. (2009) Long-term declines in phosphorus export from forested catchments in south-central Ontario. Canadian Journal of Fisheries and Aquatic Sciences 66:1682-1692. Fee EJ, Shearer JA, DeClercq DR. (1978) In vivo chlorophyll profiles from lakes in the Experimental Lakes Area, Northwestern Ontario - 1976 Data. Fisheries & Marine Serivce. Department of Fisheries and the Environment, Winnipeg, Manitoba, 104 p. Flear K. (2011) Changes in scaled-chrysophyte assemblages in response to recent climate change in Northwestern Ontario. Department of Biology. Queen's University Kingston, Ontario, 121 p. Folt CL, Chen CY, Moore MV, Burnaford J. (1999) Synergism and antagonism among multiple stressors. Limnology and Oceanography 44:864-877. Garmo ØA, Skjelkvåle BL, de Wit HA, Colombo L, Curtis C, Fölster J, Hoffmann A, Hruška J, Høgåsen T, Jeffries DS, Keller WB, Krám P, Majer V, Monteith DT, Paterson AM, Rogora M, Rzychon D, Steingruber S, Stoddard JL, Vuorenmaa J, Worsztynowicz A. (2014) Trends in surface water chemistry in acidified areas in Europe and North America from 1990 to 2008. Water, Air, and Soil Pollution 225:1880-1-1880-14. George CJ. (1981) The fishes of the Adirondacks. New York State Department of Environmental Conservation, 93 p. Ginn BK, Rate M, Cumming BF, Smol JP. (2010) Ecological distribution of scaled-chrysophyte assemblages from the sediments of 54 lakes in Nova Scotia and southern New Brunswick, Canada. Journal of Paleolimnology 43:293-308. Glew JR, Smol J, Last WM. (2001) Sediment core collection and extrusion. In: Last WM, Smol J (eds), Tracking environmental change using lake sediments. Volume 1: basin analysis, coring, and chronological techniques Kluwer Academic Publishers, Dordrecht., pp. 73-105. Healy FP. (1983) Effect of temperature and light intensity on the growth rate of Synura sphagnicola. Journal of Plankton Research 5:767-774. Hyatt CV, Paterson AM, Cumming BF, Smol JP. (2010) Factors related to regional and temporal variation in the distribution of scaled chrysophytes in north-eastern North America: Evidence from lake sediments. Nova Hedwigia, Beiheft 136:87–102. 126 Jenkins J. (2010) Climate change in the Adirondacks: the path to sustainability. Cornell University Press, Ithaca, New York, 183 p. Jeziorski A, Yan ND, Paterson AM, DeSellas AM, Turner MA, Jeffries DS, Keller B, Weeber RC, McNicol DK, Palmer ME, McIver K, Arseneau K, Ginn BK, Cumming BF, Smol JP. (2008) The widespread threat of calcium decline in fresh waters. Science 322:1374-1377. Keller W. (2009) Limnology in northeastern Ontario: from acidification to multiple stressors. Canadian Journal of Fisheries and Aquatic Sciences 66:1189-1198. Kling HJ, Kristiansen J. (1983) Scale–bearing Chrysophyceae (Mallomonadaceae) from Central and Northern Canada. Nordic Journal of Botany 3:269-290. Krester W, Gallagher J, Nicolette J. (1989) Adirondack lakes survey 1984-1987: an evaluation of fish communities and water chemistry. Adirondack Lakes Survey Corporation, Ray Brook, NY, 437 p. Lepš J, Šmilauer P. (2003) Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge, United Kingdom, 284 p. Likens GE, Driscoll CT, Buso DC. (1996) Long-term effects of acid rain: response and recovery of a forest ecosystem. Science 272:244-245. Martin SL, Hayes DB, Rutledge DT, Hyndman DW. (2011) The land-use legacy effect: adding temporal context to lake chemistry. Limnology and Oceanography 56:2362-2370. McMartin B. (1994) The great forest of the Adirondacks. North Country Books, Utica, New York, 254 p. Monteith DT, Stoddard JL, Evans CD, de Wit HA, Forsius M, Høgasen T, Wilander A, Skjelkvåle BL, Jeffries DS, Vuorenmaa J, Keller B, Kopácek J, Vesely J. (2007) Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature 450:537-540. Nicholls KH. (1995) Chrysophyte blooms in the plankton and neuston of marine and freshwater systems. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: Ecology, phylogeny and development. Cambridge University Press, Cambridge, UK, pp. 181-213. Nicholls KH, Gerrath JF. (1985) The taxonomy of Synura (Chrysophyceae) in Ontario with special reference to taste and odor in water-supplies. Canadian Journal of Botany 63:1482-1493. Paine RT, Tegner MJ, Johnson EA. (1998) Compounded perturbations yield ecological surprises. Ecosystems 1:535-545. 127 Paterson AM, Cumming BF, Smol JP, Hall RI. (2001) Scaled chrysophytes as indicators of water quality changes since preindustrial times in the Muskoka-Haliburton region, Ontario, Canada. Canadian Journal of Fisheries and Aquatic Sciences 58:2468-2481. Paterson AM, Cumming BF, Smol JP, Hall RI. (2004) Marked recent increases of colonial scaled chrysophytes in boreal lakes: implications for the management of taste and odour events. Freshwater Biology 49:199-207. Paterson AM, Winter JG, Nicholls KH, Clark BJ, Ramcharan CW, Yan ND, Somers KM. (2008) Long-term changes in phytoplankton composition in seven Canadian Shield lakes in response to multiple anthropogenic stressors. Canadian Journal of Fisheries and Aquatic Sciences 65:846-861. Reynoldson TB, Norris RH, Resh VH, Day KE, Rosenberg DM. (1997) The reference condition: a comparison of multimetric and multivariate approaches to assess water-quality impairment using benthic macroinvertebrates. Journal of the North American Benthological Society 16:833-852. Sandgren CD. (1988) The ecology of chrysophyte flagellates: their growth and perennation strategies as freshwater phytoplankton. In: Sandgren CD (ed), Growth and reproductive strategies of freshwater phytoplankton. Cambridge Univeristy Press, Cambridge, United Kingdom, pp. 9-104. Sandgren CDW, William E. (1995) The influence of zooplankton herbvory on the biogeography of chrysophyte algae. In: Sandgren CDS, John P. , Kristiansen J (eds), Chrysophyte algae: ecology, phylogeny and development. Cambridge University Press Cambridge, UK, pp. 269-302. Scheffer M. (2004) Ecology of Shallow Lakes. Kluwer Academic Publishers, Dordrecht, The Netherlands, 357 p. Siver P, Marsicano L. (1996) Inferring lake trophic status using scaled chrysophytes. Nova Hedwigia Beiheft 114:233-246. Siver PA. (1991) The biology of Mallomnas: morphology, taxonomy and ecology. Kluwer Academic Publishers, Dordrecht, The Netherlands, 248 p. Siver PA. (1995) The distribution of chrysophytes along environmental gradients: their use as biological indicators. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: Ecology, phylogeny and development. Cambridge University Press, Cambridge, UK, pp. 232-268. Siver PA. (2003) Synurophyte algae. In: Wehr JD (ed), Freshwater algae of North America: ecology and classification. Academic Press, San Diego, California, pp. 523-558. 128 Siver PA, Hamer JS. (1989) Multivariate statistical analysis of the factors controlling the distribution of scaled chrysophytes. Limnology and Oceanography 34:368-381. Siver PA, Hamer JS. (1992) Seasonal periodicity of chrysophyceae and synurophyceae in a small New England lake: implications for paleolimnological research. Journal of Phycology 28:186-198. Smol JP. (2010) The power of the past: using sediments to track the effects of multiple stressors on lake ecosystems. Freshwater Biology 55 (Suppl. 1):43-59. Sommer U. (1988) Some size relationships in phytoflagellate motility. Hydrobiologia 161:125-131. ter Braak CJF, Šmilauer P. (2012) CANOCO reference manual and user's guide: software for ordination (version 5.0). Microcomputer Power, Ithaca, New York, 496 p. United States Environmental Protection Agency (USEPA). (2012) Implications of climate change for state bioassessment programs and approaches to account for effects (final report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-11/036F. Verdonschot PFM, Spears BM, Feld CK, Brucet S, Keizer-Vlek H, Borja A, Elliott M, Kernan M, Johnson RK. (2013) A comparative review of recovery processes in rivers, lakes, estuarine and coastal waters. Hydrobiologia 704:453-474. Vincent WF. (2009) Effects of climate change on lakes. In: Likens GE (ed), Encyclopedia of inland waters. Elsevier, Oxford, United Kingdom, pp. 55-60. Watson SB, Satchwill T, Dixon E, McCauley E. (2001) Under-ice blooms and source-water odour in a nutrient-poor reservoir: biological, ecological and applied perspectives. Freshwater Biology 46:1553-1567. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. Yan ND, Girard R, Heneberry JH, Keller WB, Gunn JM, Dillon PJ. (2004) Recovery of copepod, but not cladoceran, zooplankton from severe and chronic effects of multiple stressors. Ecology Letters 7:452-460. 129 Tables & Table Captions Table 4.1. SIMPER test results showing taxa that contributed to the significant difference in species composition between top and bottoms samples. ANOSIM identified significant differences between top (present-day) and bottom (pre-1900) samples in the 26 reference lakes (ANOSIM R = 0.12, p = 0.002). Taxa that made a > 5% contribution (% Cont.) to the dissimilarity between groups are listed. Average square-root relative abundance (%) of the species in the top (Av √RA TOP) and bottom (AV √RA Bottom) groups are provided, with the non-square root transformed average relative abundance given in brackets. Species S. petersenii S. sphagnicola M. duerrschmidtiae M. ‘small’ M. pseudocoronata S. echinulata M. crassisquama Chrysosphaerella spp. M. acaroides Av √RA TOP (%) 3.4 (16.8) 3.2 (15.3) 3.7 (15.2) 2.9 (11.0) 1.2 (2.7) 2.6 (8.5) 2.4 (7.3) 1.5 (3.8) 1.2 (2.6) 130 Av √RA Bottom (%) 2.02 (7.0) 2.4 (8.9) 5.4 (31.5) 2.9 (11.6) 2.3 (8.9) 1.6 (4.4) 2.7 (8.5) 0.8 (1.3) 1.3 (2.6) % Cont. 11.2 10.5 9.6 8.4 8.1 7.3 6.1 5.5 5.3 Figure & Figure Captions Figure 4.1. Schematic showing the paleolimnological study design. The study consists of two parts, a top-bottom study (a) and a reproducibility study (b). The reproducibility study consists of two parts, an analysis of counting variability (i) and a stability analysis (ii). i) Counting variability. The Bray-Curtis (BC) similarities between top (present-day) samples from 4 replicate cores are computed separately for 4 reference lakes. An average BCsimilarity (ABC-top) is then computed using all comparisons from all lakes. ii) Stability analysis. BC-similarities are calculated for all comparisons between top vs. 20 cm samples and 20 vs. 30 cm samples from 3 replicate cores and are computed separately for 3 reference lakes. A Wilcoxon signed-rank test is then performed to compare the BC similarities from top vs. 20 cm samples and 20 vs. 30 cm samples from all lakes. 131 132 PCA λ4 = 0.10 PCA λ2 = 0.21 PCA λ3 = 0.14 RDA λ1 = 0.14 Figure 4.2. Principal Components Analysis (PCA) of environmental variables and Redundancy analysis (RDA) using PCA-axes scores as input. A) Biplot of PCA axes 1 & 2. B) Biplot of PCA axes 3 & 4. Input variables are: pH, acid neutralizing capacity (ANC, μeq L-1), K, Na, SO42- (μmol L-1), Cl-, Mg2+ (μmol L1 , log transformed), dissolved organic carbon (DOC, μmol C L-1, log transformed), total nitrogen (TN, μmol L-1, log transformed), Elevation (elv, m), and maximum depth (Zmax, m, log transformed). C) RDA of present-day samples from 31 reference lakes. PCA axes (environmental variables) are open arrows, species are closed arrows, and samples are shown as points. Eigenvalues (λ) are given. PCA λ1 = 0.27 RDA λ2 = 0.05 133 PCA λ2 = 0.22 PCA λ2 = 0.22 PCA λ1 = 0.24 PCA λ1 = 0.24 Figure 4.3. Principal Components Analysis (PCA) of chrysophyte assemblages in top (present-day) and bottom (pre-1900) samples in 26 reference lakes. The PCA is summarized in multiple panels to reduce crowding in biplots. A) Species vectors. B) Present-day (filled circle) and pre-1900 (empty square) samples. Dashed lines show trajectory of species change in ordination space from pre-1900 to present. C) Present-day samples only (labelled). Eigenvalues (λ) are given. PCA λ1 = 0.24 PCA λ2 = 0.22 Figure 4.4. Total relative abundance (%) of colonial chrysophyte taxa in top (present-day) and bottom (pre-1900) samples in 26 reference lakes. Dashed line represents a one-to-one line. All samples above the grey dashed line have shown an increase in the total relative abundance of colonial chrysophytes from pre-1900 to present. 134 CHAPTER 5 UNDERSTANDING BIOLOGICAL RECOVERY FROM ACIDIFICATION: APPLYING A REFERENCE SITE FRAMEWORK IN THE ADIRONDACKS (NY, USA) Abstract Since the implementation of the US Acid Rain Program (US ARP), lakes in some areas of the United States have begun to undergo chemical recovery from acidification (e.g. increases in pH/ANC, decreases in aluminium concentrations). Chemical recovery is expected to promote biological recovery from acidification. However, it is often difficult to assess biological recovery due to a lack of long-term monitoring data. Furthermore, climate warming may prevent recovering lakes from returning to their pre-acidification state, even with management intervention. In such a scenario, a reasonable recovery target can be defined by comparing impacted species assemblages to those of higher pH/ANC reference lakes that have experienced limited effects of acid deposition. These reference systems can be used to identify regional changes in species assemblages. In this study, we use paleolimnological techniques within a reference site framework to provide evidence that biological recovery from acidification is occurring in two Adirondack (NY, USA) lakes undergoing chemical recovery. A set of three acid-insensitive reference lakes were used to compare and contrast recent changes in species composition in the acidified lakes and to identify the possible influence of climate warming on Adirondack lakes. The two lakes recovering from acidification showed a significant shift in chrysophyte species composition after the 1995 implementation of the US ARP. Declines in acid-tolerant chrysophyte taxa indicate a modest biological recovery. However, both the acidified lakes and reference lakes show recent increases in the relative abundance of colonial chrysophytes similar to patterns documented elsewhere in North America. A shift toward novel 135 chrysophyte assemblages since ca. 1970-1980 is illustrated in non-metric multidimensional scaling (nMDS) ordinations. The relative abundance of colonial taxa was significantly correlated with mean annual air temperature and ice-cover measures in two reference lakes. While modest biological recovery has occurred in Adirondack lakes following a decrease in acid deposition, the chrysophyte assemblages do not appear to be returning to their pre-acidification state, likely due to recent climate warming and/or oligotrophication. Introduction Acid deposition has impacted aquatic ecosystems across North America. Declines in pH/ANC (acid neutralizing capacity), increases in toxic inorganic monomeric aluminium, and other chemical and biological changes associated with long-term exposure to acid deposition result in the progressive loss of acid-sensitive taxa and a general simplification of aquatic foodwebs (Lovett et al. 2009). With the implementation of Title IV of the 1990 Clean Air Act Amendments (CAAA) and the 1995 implementation of the US Acid Rain Program (ARP), sulphur dioxide (SO2) emissions in the United States from sources governed by the US ARP have decreased 64% below their 1990 levels as of 2009 (Burns et al. 2011). In addition, there has been an ~70% decrease in nitrogen oxide (NOx) emissions from 1990 to 2009 as a result of the 1990 CAAA and the U.S. Environmental Protection Agency Nitrogen Budget Program (Waller et al. 2012). Consequently, many lakes in the United States have begun to undergo chemical recovery from acidification (Garmo et al. 2014; Skjelkvåle et al. 2005). In acid-impacted regions, it is expected that chemical recovery will lead to biological recovery from acidification, characterized by a decline in the abundance of acid-tolerant species and an increase in the abundance of acid-sensitive species as lake pH and ANC increase over time (e.g. Driscoll et al. 2001). However, studies attempting to identify biological recovery are 136 often hampered by a lack of long-term biological monitoring data. Fortunately, this problem can be addressed using paleolimnological techniques. Paleolimnological studies infer the environmental histories of lakes using the physical, chemical, and biological characteristics in dated sediment cores (Smol 2008). Paleolimnological techniques can be used to both identify the onset of acidification and provide evidence of biological recovery (Battarbee 1999). In paleolimnological studies, a lake’s recovery target is usually defined as its predisturbance condition (e.g. ca. 1850, Battarbee et al. 2011). However, a lake responding to multiple stressors over the recovery period is unlikely to return to its pre-disturbance state, even with management intervention (the ‘shifting baseline’ phenomenon, Hobbs et al. 2009; Bennion et al. 2011). In such a scenario, reasonable recovery targets can be defined by comparing impacted species assemblages to those of minimally-disturbed reference sites protected from multiple anthropogenic stressors (Bailey et al. 2004). Pairing acidified lakes with reference systems allows researchers to distinguish between species responses to acidification and chemical recovery versus responses to regional stressors such as climate change (Chapter 3). In this framework, the reference sites do not serve as ‘controls’ for the impacted sites in the traditional sense (i.e. sites with identical species composition, morphology, etc., Downes et al. 2002) but are, however, similar to the impacted sites in general characteristics (e.g. deep oligotrophic lakes with predominently forested catchments). As minimally-disturbed sites, the reference lakes serve as ‘sentinel’ systems that allow the detection of species’ responses to recent warming (USEPA 2012). Despite these benefits, reference sites are under-utilized in recovery research (Jones and Schmitz 2009). The Adirondack region represents a unique opportunity to highlight the benefits of using paleolimnological techniques to study biological recovery within a reference-site framework. 137 The Adirondacks have been heavily impacted by acid deposition (Driscoll et al. 1991) and the acidification history of the area has been well-documented in previous paleolimnological studies (Cumming et al. 1992b; Cumming et al. 1994). Following declines in acid deposition, many lakes in the region are undergoing chemical recovery from acidification (Driscoll et al. 2007; Waller et al. 2012). However, while chemical recovery may be progressing, the region is also experiencing the effects of climate change. In the Adirondacks, mean annual temperature has increased at a rate of 1.3°C per 100 years, a warming trend that has been most pronounced since the 1970s (Jenkins 2010) and is coincident with a shortened ice-covered period on lakes (Beier et al. 2012). To examine the possible influence of regional stressors like climate change on Adirondack lakes, Chapter 3 identified a regional set of 31 minimally-disturbed reference lakes from an online database of 1,469 lakes. The sites were chosen using a set of stringent selection criteria and have been largely protected from: acidification, eutrophication from shoreline or watershed development, road salt seepage, and piscivores introductions. The reference lakes were studied in a top-bottom paleolimnological study, a comparison of species assemblages from the present day and pre-1900 times (Cumming et al. 1992b). The top-bottom analysis lakes revealed that scaled-chrysophyte assemblages (algae belonging to the Synurophyceae and Chrysophyceae) showed a significant shift in species composition from pre-1900 to present, including an increased abundance of colonial taxa, taxa that form metalimnetic blooms, and warm-water species (Chapter 4). Similar changes have been documented elsewhere in North America (Paterson et al. 2001; Ginn et al. 2010; Flear 2011) and have been linked, at least in part, to recent warming. Because of the regional reference lake study design, were able to conclude that the changes in species composition noted were not due to acidification (decreases 138 in pH/ANC), eutrophication, road salt seepage, historic watershed disturbance, and fisheries alterations. We concluded that the species changes were likely due to the long-term effects of regional warming or oligotrophication. Moreover, we suggested that since Adirondack reference lakes had undergone a significant shift in species composition over time, lakes recovering from acidification are unlikely to return to their pre-disturbance state. This manuscript builds on the work of Chapter 4 and examines long-term changes in species composition in a subset of the regional reference lakes, pairing them with a set of lakes recovering from acidification. We hypothesized that the lakes recovering from acidification would show some evidence of biological recovery (i.e. declines in the abundance of acid-tolerant taxa) but would also show novel changes in species composition similar to those documented in the reference lakes. The following research questions are addressed: 1) Do the acidified lakes show a significant shift in species assemblage after the 1995 implementation of the US ARP? If so, is the change consistent with biological recovery from acidification? 2) Are the recent (post-1995) species assemblages of the acidified lakes moving towards their pre-acidification state? How do the changes in species composition compare with those noted in the subset of reference lakes? Can reference sites be used to extrapolate probable endpoints in the recovery trajectories of the acid-impacted sites? and, 3) Is the relative abundance of colonial chrysophytes correlated with climate variables such as mean annual air temperature or the duration of ice cover? In addition to providing information on the ecological dynamics of biological recovery from acidification (a subject that is not well documented, Verdonschot et al. 2013), this study will illustrate the advantages of pairing paleolimnological techniques within a regional referencesite framework to untangle the effects of two important stressors on aquatic systems: anthropogenic acidification and climate change. 139 Methods Site selection rationale: Sediment cores from five lakes were collected and examined in this investigation: two from acidified lakes and three from reference lakes. The two acidified lakes (South Lake, N 43°30’35”, W 74°52’35”; and Queer Lake, N 43°48’49”, W 74°46’38”) were chosen because previous paleolimnological investigations demonstrated that the lakes acidified under high loads of sulphate and nitrate deposition (South Lake ca.1930-1950, Cumming et al. 1994; Queer Lake in the late 1970s, Charles et al. 1990). Long-term chemical data indicate that both lakes are undergoing chemical recovery from acidification (See ‘chemistry trends’ below and Figure 5.1A). Three reference lakes were chosen from a regional set of 31 reference lakes (Chapter 3). Moss Lake (N 43°46’52”, W 74°51’11”) and Arbutus Lake (N 43°58’58”, W 74°14’09”) were included because they have been subject to long-term monitoring since the early 1980s (See ‘chemistry trends’ below and Fig 4.1B). Wolf Lake (N 44°01’42”, W 74°13’16”) was included because it is an Adirondack ‘Heritage lake’ (Stager and Sanger 2003) and is considered one of the most ecologically intact lakes in the Adirondacks (Beier et al. 2012). Site descriptions: Detailed descriptions of Arbutus Lake, Moss Lake, South Lake, and Queer Lake are provided by Roy et al. (2011) and a detailed description of Wolf Lake is provided by Stager and Sanger (2003) (see also Chapter 3). All of the study sites are relatively large (> 40 ha), deep (max depth ~8-21 m), and oligotrophic (TP < 10 μg L-1) drainage lakes located at elevations greater than 500 m. The lakes’ watersheds are predominantly forested with either a mixed deciduous-coniferous forest or northern hardwood forest. The lakes all have fish communities including both piscivorous and planktivorous fish and all the lakes, except for Wolf 140 Lake, have been stocked with brook trout and/or lake trout in the past (Appendix F). The watersheds of Arbutus Lake and Wolf Lake are protected within the Huntington Wildlife Forest, a property owned by the State University of New York College of Environmental Science and Forestry (SUNY-ESF) that was established in 1932. The watersheds of Moss Lake, Queer Lake, and South Lake have been protected since the early 1970s in ‘wild forest’ or ‘wilderness’ areas in the Adirondack Park where logging and watershed development are prohibited. The watersheds of Moss Lake and Queer Lake lie entirely within protected areas while 56% of the South Lake watershed is protected (Appendix F). The lakes have, however, all experienced forms of early disturbance, primarily from selective logging prior to the 1920s and forest ‘blowdown’ from a windstorm in 1950. The watershed of South Lake is the most disturbed as it experienced heavy forest blowdown in 1950 and currently has cottages on its lakeshore (Appendix F). Chemistry trends: Though not the main focus of this investigation, we summarize long-term chemistry trends in both the acidified and reference lakes to contextualize any drivers of biological change in these lakes. Detailed field and analytical methods used in the collection and analysis of long-term chemistry data are available in Driscoll and van Dreason (1993). South Lake had a pre-industrial chrysophyte-inferred (CI) pH of ~5.8 (Cumming et al. 1994) and Queer Lake had a pre-industrial diatom-inferred (DI) pH of ~6.4 (Charles et al. 1990). The CI-pH of South Lake began to decline ca. 1930-1950 and was near or below 5.0 in the early 1980s. The DI-pH of Queer Lake began to decline in the late 1970s and decreased below 5.5 in the early 1980s. In 1984, the measured pH of South Lake was 5.2 and ANC was 2.9 μeq L-1 (Roy et al. 2011). In 1986, the measured pH of Queer Lake was 5.5 and ANC was 1.9 μeq L-1 (Roy et al. 2011). From 1992-2010, chemistry records show that both lakes experienced a 141 significant decline in the concentration of strong acid anions (sulphate and nitrate), inorganic monomeric aluminium, and specific conductance (Figure 5.1A & Appendix G). Likewise, both lakes have experienced significant increases in ANC, pH, and dissolved organic carbon (DOC) over the same period (Figure 5.1A & Appendix G). The reference lakes also show long-term changes in water chemistry. From 1982-2010 in Moss Lake and from 1983-2010 in Arbutus Lake, there was a significant decline in the concentration of strong acid anions and specific conductance (Figure 5.1B & Appendix G). Moss Lake has also shown a significant increase in pH, ANC, and DOC since 1982. The increase in pH and ANC in Moss Lake appears to be due to a decline in the strength of episodic acidification. Long-term water chemistry data are not available for Wolf Lake. However, sporadic pH measurements show that Wolf Lake had a pH between 6.5 and 7.0 from 1950-1997 (Stager and Sanger 2003). Sediment collection and preparation: Sediment cores were collected from the deepest basin of the study lakes in May 2008 (South Lake, Moss Lake), November 2009 (Queer Lake), and May 2010 (Wolf Lake, Arbutus Lake). Sediment cores (~20-30 cm in length) were collected using a Glew gravity corer with an internal diameter of 7.6 cm and were sectioned on site into 0.25-cm intervals using a modified Glew vertical extruder (Glew et al. 2001). Scaled chrysophytes (Synurophyceae and Chrysophyceae; hereafter referred to as ‘chrysophytes’) are used in this investigation because they are sensitive to pH/aluminum levels and because their characteristic siliceous scales are well-preserved in lake sediments (Cumming et al. 1992a). Slides were prepared for chrysophyte enumeration following the methods of Battarbee et al. (2001). Briefly, 0.2-0.3 g of wet sediment was transferred to a labelled glass vial and digested in approximately 20 mL of concentrated HNO3-H2SO4 to isolate the siliceous 142 chrysophyte scales. Samples were heated in a hot water bath to ~70°C for 6 hours and then rinsed 8-9 times until the samples were acid-free (litmus test). The samples were plated as a series of four 100% dilutions and mounted on microscope slides using Naphrax®. Activities of 210Pb, 137Cs, and 214Bi were measured following the methods of Schelske et al. (1994). Briefly, ~20 sediment intervals per core were freeze-dried using a VirTis Advantage freeze-drier (SP Industries, Inc.). Freeze-dried sediment was added to polypropylene tubes, capped with 2-Ton epoxy, and kept at room temperature for at least two weeks before 210Pb, 137 Cs and 214Bi activities were measured in a high-purity germanium coaxial well-detector with a 4-mm active well depth with an internal diameter of 15.5 mm (EG& G Ortec) attached to a DSPEC digital spectrometer. Sediment age was calculated using the constant rate of supply (CRS) model (Appleby and Oldfield 1978). In cases where sediment intervals were inferred to represent < 1 year, species counts were combined in adjacent intervals to provide a minimum yearly resolution throughout the profile. Chrysophyte scales were enumerated on a Leica DMBR microscope using differential interference contrast (DIC) optics. A minimum of 400-500 chrysophyte scales were counted for samples from South Lake, Queer Lake, Moss Lake, and Wolf Lake. A minimum of 200 scales were counted for samples from Arbutus Lake, as the abundance of scales was low prior to 1980. The principle taxonomic references used in this investigation were: Kling and Kristiansen (1983), Nicholls and Gerrath (1985), Siver (1991), and unpublished reference materials. A Mallomonas ‘small’ category was used to group small Mallomonas taxa that are not easily distinguished using light microscopy (Cumming et al. 1992a). Statistical analyses – understanding long-term changes in chrysophyte assemblages: Unless otherwise stated, chrysophyte taxa that reached a relative abundance > 2% at least twice in the 143 sedimentary profile were included in the following analyses. CI-pH values were generated for the acidified and reference lakes from a calibration set of 71 Adirondack lakes (Cumming et al. 1992a). CI-pH values were calculated based on a weighted-averaging model with tolerance down-weighting and inverse deshrinking with a ln (x+1) species transformation (Cumming et al. 1994). Stratigraphies of the relative abundance of chrysophyte taxa were created using the computer program C2 v. 1.7.2 (S. Juggins, unpub. program, 2011) and are included in an appendix (Appendix H). Ordination techniques were used to summarize long-term changes in species composition in the acidified and reference lakes. Non-metric multidimensional scaling (nMDS) plots were used to summarize long-term (pre-1900 to present) changes in species assemblages in the acidified and reference lakes. nMDS is based on a ranked matrix of Bray-Curtis (BC) distances (dissimilarities), the twin concept to BC similarity (BC similarity = 1 – BC dissimilarity). Because nMDS is based on a ranked distance matrix, the technique avoids the assumptions other ordination methods like PCA make about the underlying structure of the species data. nMDS simplifies multidimensional species data into two (or three) dimensions and sedimentary assemblages with a similar species composition will plot close together in the ordination diagram. Thus, if recent species assemblages are in close proximity to the pre-acidification assemblages in the ordination diagram, it can be inferred that the recent assemblages are moving towards their pre-acidification state. While both PCA and nMDS are appropriate for illustrating long-term trends in species assemblages, nMDS was used in this investigation because the ordination is based directly on an underlying dissimilarity matrix. This consideration was important for an ordination used to illustrate how similar the pre-1900 species assemblages of the acidified lakes were to the pre-1900 species assemblages of a set of regional reference lakes (See 144 Statistical analyses – identifying a return to pre-disturbance state and recovery endpoints). PCA ordinations of long-term species changes in both the acidified and reference lakes are provided in an Appendix for comparison (see Results below). When designating reference periods and recovery targets in paleolimnological studies, it is important to consider factors such as landscape disturbance (Renberg et al. 2009). Known watershed disturbances in the acidified lakes and reference lakes were minimal prior to ca. 1900 (Appendix F). As such, we defined four time-periods in the nMDS for each lake: pre-disturbance (pre-1900), early disturbance (1900 to pre-1950), acidification (1950 to 1995), and post-ARP (post-1995 to present). The sedimentary assemblages of the reference lakes were grouped in the same time periods to allow for comparison of species changes between the reference and acidified lakes. nMDS plots were created using BC distances of square-root transformed species data. Permutations tests (999 permutations) were used to identify the nMDS ordinations summarized in 2 axes with the lowest stress (stress equation 2; primary distance measures). nMDS ordinations were performed on square-root transformed species data using the computer program CANOCO v. 5.0 (ter Braak & Šmilauer 2012). Statistical analyses – quantifying biological recovery: An Analysis of Similarities (ANOSIM) (Clarke 1993) was used to determine if the species assemblages in the acidified lakes showed a significant shift in species composition after the 1995 implementation of the US ARP in comparison to a time period of equal length prior to 1995 (details below). ANOSIM is a nonparametric test of ranked similarities that tests for a significant difference in species compositions between a priori defined groups in multivariate data. Subsequently, the similarities percentage test (SIMPER) can be used to identify which taxa contribute to the differences in species composition seen between groups. 145 In each core, two ~ten-year time periods were identified prior to and after the 1995 implementation of the US ARP. An analysis of long-term monitoring data from Adirondack lakes showed no systematic increase in pH or ANC from 1982-1994 (Driscoll et al. 1995). Thus, in the absence of pre-1992 long-term chemical monitoring data in South Lake and Queer Lake, the ANOSIM framework in this investigation attempts to designate two time periods: one period when the lakes were still acidic (Period 1, pre-1995) and one period when they are undergoing chemical recovery (Period 2, post-1995). We interpreted a decline in the relative abundance of Mallomonas hamata and Mallomonas hindonii (acid-tolerant taxa associated with low pH and high concentrations of inorganic monomeric aluminium, Cumming et al. 1992a), and a corresponding increase in taxa with higher pH optima in Period 2 as evidence of biological recovery from acidification. The approach used here is conservative in that it recognizes that 210 Pb dating has errors that are difficult to quantify (Binford 1990). By delineating two ten-year time periods and treating sedimentary intervals within each time period as replicates, major trends in species composition may be identified. ANOSIM (using BC similarity coefficient) and SIMPER were performed on square-root transformed species data using the computer program PRIMER v. 6.1.11. Statistical analyses – identifying a return to pre-disturbance state & recovery endpoints: By plotting the species assemblages of the acidified sites in nMDS ordination space, we can identify how the assemblages have changed over time and identify if the recent post-1995 assemblages have moved back towards a pre-disturbance (pre-1900) condition. While informative, this approach is somewhat limited as it is only able to identify species changes that have occurred in South Lake and Queer Lake up until the present and does not identify probable recovery 146 endpoints (i.e., what species assemblages are likely to occur in the lake when chemical recovery is complete?) We can expand on the long-term nMDS-ordination analysis for the acidified lakes by comparing the species trends in the acidified lakes to those documented in the regional set of 31 reference sites. First, we actively plotted the pre-1900 sedimentary assemblages of the acidified lakes with all pre-1900 samples from the regional reference lakes (the ‘bottom’ samples from the top-bottom analysis; 26 of 31 reference lakes had a sample from pre-1900). This allows us to identify which reference lakes were most similar in species composition to the acidified lakes in pre-disturbance (pre-1900) times. In a subsequent ordination, we actively plot the pre-1900 and present-day samples of a subset of the reference lakes that were similar in species composition to the acid-impacted lakes with the long-term record of each acidified lake. Active plotting is necessary in these ordinations so that taxa present in either the acidified lakes but absent from the reference lakes (such as M. hindonii) or vice versa contribute to the ordination. The ordination space inhabited by the present-day assemblages of the reference lakes represents a hypothetical ‘endpoint’ for biological recovery in the acidified lakes – a space that may be both different from the pre-disturbance condition and the present-day species composition of the acidified. The assumption underlying the above approach is that the species assemblages of the acidified and reference lakes were similar pre-1900 because the physical/chemical/biological conditions of the lakes were similar. Thus, changes in the reference lakes over time can be used as a proxy of changes in the acid-impacted sites. Presumably, reference sites that were more similar to the acid-impacted sites pre-1900 provide better proxies for expected changes in the recovering acidified lakes (i.e. sites 70-80% similar are better analogues than sites 40-50% similar). Taxa that made a > 2% contribution to more than two samples in the combined dataset 147 of the regional reference lakes and South Lake/Queer Lake were included in the ordination. The two reference lakes that had the highest average BC similarity with the pre-1900 samples from each acid-impacted lake in the first nMDS ordination were included in the second nMDS ordination. nMDS plots were created in CANOCO v. 5. Statistical analyses - correlations with climate variables: A Mann-Kendal correlation test (Helsel and Hirsch 2002) was used to determine if there was a relationship between total relative abundance of colonial chrysophytes and selected climate variables. The Mann-Kendall test is a non-parametric test that identifies if a monotonic trend exists between two data series (i.e. generally increasing or generally decreasing). Note that a ‘strong’ Mann-Kendall correlation typically has a smaller correlation coefficient than an equivalently strong Pearson correlation coefficient (i.e. R = 0.9 ≈ τ = 0.7, Helsel and Hirsch 2002). The correlation between total relative abundance of colonial taxa and ice-on date, ice-off date, and ice-cover duration were examined for Arbutus Lake and Wolf Lake from ca. 1970 to present (the only sites with long-term icecover data). Only years coinciding with the 210Pb-inferred dates of the sedimentary intervals were included in the temperature and ice cover data. The length of the ice-covered season has been shown to influence the timing and composition of spring phytoplankton blooms (Gerten and Adrian 2002). Therefore, in this investigation we compared ice-duration data to sedimentary intervals whose 210Pb-inferred dates matched the year in which ice-off occurred. Mann-Kendall tests were performed using the computer program of Helsel et al. (2006). Ice cover data were provided by Dr. C. Beier (Adirondack Ecological Center, SUNY- ESF). Quality-controlled mean-annual air-temperature data were provided by the United States Historical Climatology Network (Menne et al. 2009) and were from the meteorological station located within the 148 Adirondack Park nearest to each lake: Station 304102 Arbutus Lake (~ 26 km) and Wolf Lake (~ 30 km). Results 210 Pb activities and dates: All lakes showed exponential decay of unsupported 210Pb with cumulative dry mass in the core (Figure 5.2). 210Pb activity in the Wolf Lake sediment core did not reach background activity and so unsupported 210Pb activity was estimated using background levels from a previously collected sediment core (Stager and Sanger 2003). 137Cs profiles did not show a pronounced peak in activity, which is common in the highly-organic sediment typical of Adirondack lakes (Heit and Miller 1987). Standard errors of inferred-dates were low (< 1 year) in recent (post-1950) sediment intervals. Age-depth models were calculated based on a quadratic equation for South Lake, Moss Lake, and Arbutus Lake. For Queer Lake and Wolf Lake, a cubic equation was used as quadratic equations under-estimated the age of sediment at depths < 6-7 cm. Long-term changes in species composition Acidified lakes: In general, the acidification histories of South Lake and Queer Lake are similar to previously published results (Cumming et al. 1994, Charles et al. 1990). The CI-pH of South Lake began to decline ca. 1940-1950 and reached its lowest values in the 1980s (~5.0). Since the mid-1990s, there has been a modest increase in CI-pH, though values remained low (<5.5) (See Figure 5.3). In Queer Lake, the CI-pH began to decline in the 1970s, decreased below 6.5 in the 1980s, and has shown little change since. Stress tests of nMDS biplots of long-term species assemblage data from South Lake and Queer show that the 2-axes of the nMDS provide a good approximation of the multivariate data 149 (stress = 0.10; stress = 0.07, respectively) (See Figure 5.4 – species codes listed in Appendix I). The nMDS plots showed a similar separation between time-periods as PCA plots made on square-root transformed species data (See Appendix J). In both lakes, the pre-disturbance (pre1900) and early disturbance (1900 to pre-1950) species assemblages were similar in composition, though some pre-1950 assemblages in South Lake had a high relative abundance of S. echinulata, a colonial chrysophyte (See Figure 5.4 & Appendix H, Figure H1). The pre-1950 assemblages in both lakes are characterized by a higher relative abundance of taxa with higher pH optima such as M. acaroides (pH optimum = 5.9, Cumming et al. 1992a) and M. pseudocoronata (pH optimum = 7.2, Cumming et al. 1992a) and a lower relative abundance of the acid-tolerant taxa M. hamata and M. hindonii. In South Lake, M. hamata and M. hindonii reached their highest relative abundance from the 1980s to the early 1990s and then showed a subsequent decline while S. echinulata increased again post-1970. In Queer Lake, M. hamata increased in the late 1970s, peaked in the late 1980s, and then showed a subsequent decline. In contrast, Mallomonas punctifera and the colonial chrysophytes Synura petersenii, S. echinulata and Chrysosphaerella spp. increased in relative abundance in the 1980s. Reference Lakes: The reference lakes all maintained a CI-pH greater than 6.5 over the sedimentary record, suggesting that the lakes have not acidified over time (See Figure 5.3). Stress tests of nMDS biplots of long-term species assemblage data from Moss Lake, Arbutus Lake, and Wolf Lake show that the 2-axes of the nMDS provide a good approximation of the multivariate data (stress = 0.08; stress = 0.15; stress = 0.17, respectively) (See Figure 5.4 – species codes listed in Appendix I). In general, the pre-disturbance and the early disturbance periods in each lake are similar in species composition, although, similar to South Lake, some years in the early disturbance period of Moss Lake are characterized by an increased relative 150 abundance of a colonial chrysophyte (Moss Lake: S. sphagnicola; South Lake: S. echinulata) (See Figure 5.4 & Appendix H, Figure H2). Similar to the acidified lakes, all three reference lakes show an increasing relative abundance of colonial chrysophyte taxa since the 1970s and 1980s. The post-1995 assemblages of Moss Lake are generally characterized by an increased relative abundance of S. echinulata, M. punctifera, and S. spinosa, though some intervals still have a high relative abundance of the unicellular taxon M. duerrschmidtiae. The post-1995 assemblages of Arbutus Lake are characterized by an increased relative abundance of the colonial taxa S. sphagnicola and S. petersenii, though the relative abundance of S. petersenii has decreased since the early 2000s. Lastly, the post-1995 species assemblages of Wolf Lake are characterized by a decline in the relative abundance of M. pseudocoronata and an increase in the relative abundance of S. petersenii, though some of the pre-1950 species intervals had a similar relative abundance of S. petersenii. Species changes in the acidified lakes post-1995 US ARP: The ANOSIM results showed a significant shift in species composition post-1995 in both South Lake (R = 0.83, p = 0.001) and Queer Lake (R = 0.68, p = 0.002) (See Table 5.1). The SIMPER post-hoc test revealed that recent declines in the relative abundance of acid-tolerant species contributed to the dissimilarity between Periods 1 and 2 in South Lake (M. hamata, 24.8% and M. hindonii, 12.7%) and Queer Lake (M. hamata, 7.6%). However, in Queer Lake the recent increase in the relative abundance of S. petersenii and M. punctifera contributed more to the dissimilarity between periods (25.0% and 9.8%, respectively) than M. hamata. Return to pre-disturbance state and recovery endpoints: The post-1995 assemblages of South Lake plot intermediately on nMDS axis-1 between the pre-1900/pre-1950 samples and the 151 majority of the samples in the acidification time period (1950 to 1995), suggesting that the lake may be moving towards a pre-acidification assemblage (Figure 5.4). The intermediate placement is due to the persistence of M. hamata and M. hindonii and the fact that S. echinulata showed a peak in relative abundance in the early 1900s. In contrast, the post-1995 species assemblages of Queer Lake appear to be moving towards a unique chrysophyte assemblage, characterized by an increased relative abundance of S. petersenii, M. punctifera, S. echinulata, and Chrysosphaerella spp. Interestingly, the species assemblages of the three reference lakes also show a separation in ordination space between their recent and pre-1950 species assemblages, suggesting that they are also moving towards a novel chrysophyte assemblage characterized by an increased abundance of colonial chrysophytes (S. echinulata, S. sphagnicola, and/or S. petersenii) and an increased abundance of M. punctifera. The nMDS plot of the pre-1900 samples from South Lake and 26 regional reference lakes found that Nellie Pond (NellP) and Round Pond (RounP) were most similar to South Lake species assemblage pre-1900 (BC similarity 73% and 70%, respectively) (See Figure 5.5A). Both lakes have shown an increase in the relative abundance of colonial chrysophytes over time (S. sphagnicola, NellP; Chrysosphaerella spp., RounP). The present-day assemblages of the reference lakes are located in a different part of the ordination diagram than the recent assemblages of South Lake (See Figure 5.5B). The nMDS of pre-1900 samples from Queer Lake and the regional reference lakes showed that Wolf Lake (WolfL) and Moss Lake (MossL) were most similar to the pre-1900 Queer Lake assemblages (78% and 75%, respectively) (See Figure 5.6A). Because both lakes had long-term profiles created in this investigation, long-term changes in both lakes are compared separately with long-term changes in Queer Lake. The long-term nMDS plots show 152 that Queer Lake, Moss Lake, and Wolf Lake are moving towards novel chrysophyte assemblages. Wolf Lake has shown an increase in S. petersenii over time along with a decrease in abundance of the unicellular taxon M. pseudocoronata, though the abundance of S. petersenii is less than its current abundance in Queer Lake (See Figure 5.6B). The abundance of S. petersenii is also lower in Moss Lake, but Moss Lake has also shown an increase in S. echinulata and M. punctifera over time (generally < 2% and < 5% relative abundance in Queer Lake) (See Figure 5.6C). Correlations with air temperature/ice cover: The total relative abundance of colonial chrysophyte taxa was significantly positively correlated with mean annual air temperature and measures of ice-cover duration in Arbutus Lake and Wolf Lake (Table 5.2). The total percent relative abundance of colonial chrysophytes was positively correlated with ice-on date in Arbutus Lake (i.e. years with a later freeze-up date were associated with a higher relative abundance of colonial chrysophytes) and negatively correlated with ice-cover duration in Wolf Lake (i.e. years with shorter ice-cover periods were associated with a higher relative abundance of colonial chrysophytes). Discussion Biological recovery from acidification and novel species assemblages Both South Lake and Queer Lake are undergoing some biological recovery from acidification. ANOSIM documented a significant shift in species composition in both lakes after the 1995 implementation of the US ARP associated with a decline in the relative abundance of acid-tolerant taxa (M. hamata and M. hindonii, South Lake; M. hamata, Queer Lake). However, neither South Lake nor Queer Lake showed recent increases in taxa with higher pH-optima such 153 as M. acaroides or M. pseudocoronata. The pH of South Lake decreases below 5.0 most years and the pH of Queer Lake has only begun to reach consistently above 6.0 since 2001. Multiple studies have suggested that the pH ‘benchmarks’ of 5.0 and 6.0 must be maintained for acidsensitive species to persist in recovering lakes (Lovett et al. 2009). As such, recovery is likely to be limited until further improvements in water chemistry. The nMDS ordination of the South Lake samples alone suggested that South Lake was moving towards a chrysophyte assemblage more characteristic of pre-acidification conditions. However, an nMDS plot comparing the long-term changes in chrysophyte species composition in South Lake to two reference sites that had a similar species composition pre-1900 suggested that South Lake may not return to its pre-disturbance condition. Instead, the lake may show an increased abundance of colonial chrysophyte taxa currently rare or absent from South Lake (S. sphagnicola, currently < 1% and/or Chrysosphaerella, currently absent). Queer Lake appears to be moving towards a novel assemblage characterized by an increased relative abundance of colonial chrysophytes relative to pre-disturbance times, a trend noted in all three reference lakes. The comparative nMDS plots with Moss Lake and Wolf Lake suggest that the relative abundance of S. petersenii may eventually decline in Queer Lake and that other colonial taxa and/or M. punctifera may increase in relative abundance in the lake as recovery progress. In the study of 31 Adirondack reference lakes, S. petersenii was most abundance in lakes with DOC concentrations less than 5.2 mg C L-1 (Chapter 4). DOC is currently increasing in Queer Lake and if its concentration surpasses 5.2 mg C L-1, this could lead to a decrease in S. petersenii, as suggested by the comparative nMDS plots with Moss Lake and Wolf Lake. As increases in colonial chrysophytes were documented in both the acidified lakes and the reference lakes which have maintained a CI-pH of ~6.5 or greater throughout the sedimentary 154 record, the increased abundance of colonial chrysophytes does not appear to be related to acidification and chemical recovery. Because both the reference and acidified sites have been subject to long-term chemical monitoring, we are also able to dismiss other changes in water chemistry variables as possible drivers of the changes in chrysophyte assemblages. Chrysophyte taxa are known to be sensitive to specific conductance (Siver 1993). However, the declines in specific conductance measured in the study lakes are within the tolerance range of most taxa and so are unlikely to explain the recent increases in colonial chrysophytes (Siver 1993). The total relative abundance of colonial chrysophyte taxa increased in all sites, regardless of whether the lakes have shown an increase in DOC (Moss Lake, South Lake, Queer Lake) or no directional change in DOC concentrations over time (Arbutus Lake). However, as mentioned above, DOC may have an important influence of the relative abundance of S. petersenii. Indeed, in Arbutus Lake there has been a decrease in the relative abundance of S. petersenii since the early 2000s, a time period typically characterized by DOC concentrations greater than 5 mg C L-1. In contrast, as both S. echinulata and S. sphagnicola are likely DOC-indifferent in the reference lakes (Chapter 4), changes in DOC concentration cannot explain recent increases in the total relative abundance of colonial chrysophytes noted in both the acidified and reference lakes. Chapter 4 suggested that increases in colonial chrysophytes in the regional set of 31 Adirondack reference lakes were due to regional warming and/or oligotrophication due to increased adsorption of TP in watershed soils with long-term exposure to acid deposition and/or logging followed by forest regrowth. Similar conclusions have been reached in other North American studies (Paterson et al. 2001, Paterson et al. 2008, Ginn et al. 2010, Hyatt et al. 2010, Flear et al. 2011). With longer ice-free periods and longer periods of thermal stratification, flagellated algae like chrysophytes may have a competitive advantage over non-flagellated taxa 155 like diatoms because they have an ability to seek out optimal light/nutrient conditions in the water column (Winder and Sommer 2012). At an assemblage level, colonial taxa may have a competitive advantage over unicellular ones with longer periods of stratification, as large flagellates tend to be superior swimmers (Sommer 1988). Of course, the effects of regional warming could be interactive with oligotrophication. In low-nutrient conditions, colonial chrysophytes may be favoured because they form blooms at or below the thermocline where nutrient conditions are more favourable (Fee 1976). Furthermore, phagotrophic colonial taxa like Chrysosphaerella may have an advantage because they can consume alternative sources of nutrients (Holen and Boraas 1995). Unfortunately, the TP records from the study sites in this investigation were too short to be analyzed (< 5 years) and so future studies should examine long-term trends in TP concentrations in the Adirondacks and their potential influence on the recovery of algal assemblages. This study does however suggest a link between regional warming and changes in chrysophyte species composition in Adirondack lakes. Both mean annual temperature and measures of ice cover duration were significantly correlated with the total relative abundance of colonial chrysophytes in Arbutus Lake and Wolf Lake. Colonial chrysophytes can be found in both cool and warm waters (Siver 2002) and so colonial taxa are likely responding to indirect factors associated with warming such as ice cover duration or thermal stratification rather than directly to air or water temperature. In contrast, the fact that M. punctifera increased coincidently with colonial taxa in Moss Lake and Queer Lake likely reflects warming water temperatures as it is a warm-water chrysophyte taxon (Siver 1991). In Arbutus Lake, the relative abundance of colonial chrysophytes was positively correlated with ice-on date. S. sphagnicola, the main colonial chrysophyte in Arbutus Lake, typically blooms in the water column from the summer to 156 early autumn and is lost with the appearance of ice cover (Siver and Hamer 1992). Thus, delaying the onset of the ice-covered period would allow this taxon to persist in the water column for longer periods. In contrast, S. petersenii, the principal colonial chrysophyte in Wolf Lake, is a dominant taxon just after ice off, though it can remain in the water column throughout the year (Siver and Hamer 1992). The relatively modest increase in S. petersenii noted in Wolf Lake may therefore be due to the fact that the ice-off date for Wolf Lake has changed little since the 1970s (Beier et al. 2012). This investigation thus suggests that the response of taxa to regional warming may be affected by their seasonal occurrence in the water column. Given the findings of this and other studies (e.g., Paterson et al. 2004, Flear 2011), it is unlikely that lakes recovering from acidification in the Adirondacks and elsewhere will return to their pre-disturbance state. There is, however, a general need for more research into the factors that control algal species composition and seasonal dynamics in freshwater ecosystems. Without more detailed information, it is difficult to project phytoplankton responses to warming beyond general trends (e.g. increases in flagellated algae, Winder and Sommer 2012). Such projections would be highly beneficial, as some taxa that are likely to gain a competitive advantage with warming conditions are also known to cause taste and odour issue in freshwaters, an important management concern (e.g. S. petersenii, Nicholls and Gerrath 1985). The nMDS plots created in this investigation were an attempt to predict what possible recovery endpoints could be expected for the acidified lakes. This relatively simple approach could certainly be expanded upon by modelling work of the reference/impacted sites that accounted for the physical/chemical/biological effects of climate change. 157 Building context: a long-term perspective of chrysophyte assemblage change Using paleolimnological techniques to define biological recovery enables us to understand recent increases in colonial chrysophyte taxa in a historical context. This study revealed two long-term patterns in colonial chrysophyte species abundance. In Queer Lake, Arbutus Lake, and Wolf Lake, there was little change in the relative abundance of colonial chrysophyte taxa until ca. 1970-1980. In contrast, in Moss Lake and South Lake there were pre1950 peaks in the abundance of colonial chrysophytes (South Lake: early 1900s; Moss Lake: late 1940s), followed by subsequent increases post-1970. Both trends have been noted in previous paleolimnological investigations (Cumming et al. 1994) and merit further discussion. As pre-1950 peaks in colonial taxa were noted both in acidified and reference lakes, it does not appear that the trend is the result of an early onset of acidification. Northern New York experienced a modest warming from the early 1900s to the mid-20th century, followed by a rapid warming from the 1970s to present (Jenkins 2010). Thus, the mechanisms related to warming discussed above may have promoted pre-1950 increases in colonial chrysophytes. Alternatively, Davis et al. (2006) suggested that colonial chrysophytes increased in two New Hampshire lakes due to watershed disturbance from logging and development, early stressors experienced by South Lake and Moss Lake. However, Paterson et al. (1998) found that logging had little impact on chrysophyte assemblages. It is therefore unclear what role, if any, disturbance had in promoting early peaks in colonial chrysophytes in Adirondack lakes. Future studies should examine possible relationships among climate warming, landscape disturbance, and changes in algal assemblages. However, it must be emphasized that watershed disturbance cannot explain recent increases in colonial chrysophytes as the lakes examined in this investigation have 158 experienced little watershed disturbance since the 1930s (Wolf Lake and Arbutus Lake) and the 1970s (Moss Lake, South Lake, Queer Lake). Recovery in a multiple stressor environment Modest biological recovery from acidification is occurring in Adirondack lakes, though further recovery may be limited until key pH benchmarks are met. The acidified and reference lakes examined in this investigation appear to be moving towards a novel chrysophyte assemblage, characterized by recent (post-ca. 1970-1980) increases in colonial chrysophyte taxa and warm-water taxa. The recent increases in colonial chrysophytes cannot be explained by chemical recovery from acid deposition alone as they were noted in both acidified and nonacidified lakes. Rather, the increases may be related to recent climate warming and/or oligotrophication. Additionally, the pre-1950 peaks in colonial chrysophyte abundance noted in some lakes may be related to early warming and/or watershed disturbance. However, disturbance cannot explain recent increases in colonial chrysophytes as the lakes in this investigation have been protected from watershed disturbance since the 1930s and 1970s. This study illustrates that applying paleolimnological techniques within a reference lake framework provides researchers with powerful tools to contextualize species responses to multiple stressors, a growing research need as lakes become increasingly impacted by multiple stressors in North America (Keller 2009) and elsewhere. The US EPA recently recognized that current SO2/NOx emissions standards do not adequately protect sensitive ecosystems from acid deposition (Burns et al. 2011). Modelling studies have suggested that many Adirondack lakes will have a pH less than 6.0 and ANC values less than 50 μeq L-1 by 2050 (Chen and Driscoll 2005) and that, without further reductions in sulphate deposition, many watersheds will experience deposition greater than target loads for ANC of 50 μeq L-1 by 2100 (Sullivan et al. 159 2012). As such, biological recovery will likely be a protracted process, during which aquatic communities will be influenced by multiple anthropogenic stressors. Shifts towards novel species assemblages such as the recent increases in colonial chrysophytes documented in this investigation are likely to be prevalent as recovering communities in North America may pose unexpected management concerns, such as taste and odour issues. Literature Cited Appleby PG, Oldfield F. (1978) The calculation of lead-210 dates assuming a constant rate of supply of unsupported 210Pb to the sediment. CATENA 5:1-8. Bailey RC, Norris RH, Reynoldson TB. (2004) Bioassessment of freshwater ecosystems: using the reference condition approach. Kluwer Academic Publishers, New York, 170 p. Battarbee RW. (1999) The importance of palaeolimnology to lake restoration. Hydrobiologia 395/396:149-159. Battarbee RW, Anderson NJ, Bennion H, Simpson GL. (2012) Combining limnological and palaeolimnological data to disentangle the effects of nutrient pollution and climate change on lake ecosystems: problems and potential. Freshwater Biology 57:2091-2106. Battarbee RW, Jones VJ, Flower RJ, Cameron NG, Bennion H, Carvalho L, Juggins S. (2001) Diatoms. In: Smol JP, Birks HJB, Last WM (eds), Tracking Environmental Change Using Lake Sediments. Volume 3:Terrestrial, Algal, and Siliceous Indicators. KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. 155-202. Battarbee RW, Morley D, Bennion H, Simpson GL, Hughes M, Bauere V. (2011) A palaeolimnological metadatabase for assessing the ecological status of lakes. Journal of Paleolimnology 45:405-414. Beier CM, Stella JC, Dovčiak M, McNulty SA. (2012) Local climatic drivers of changes in phenology at a borealtemperate ecotone in eastern North America. Climatic Change 115:399-417. Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. 160 Binford MW. (1990) Calculation and uncertainty analysis of 210Pb dates for PIRLA project lake sediment cores. Journal of Paleolimnology 3:253-267. Burns DA, Lynch JA, Cosby BJ, Fenn ME, Baron JS. (2011) National Acid Precipitation Assessment Program report to Congress 2011: an integrated assessment. National Science and Technology Council, Washington, DC, 114 p. Charles DF, Binford MW, Furlong ET, Hites RA, Mitchell MJ, Norton SA, Oldfield F, Paterson MJ, Smol JP, Uutala AJ, White JR, Whitehead DR, Wise RJ. (1990) Paleoecological investigation of recent lake acidification in the Adirondack Mountains, N.Y. Journal of Paleolimnology 3:195-241. Chen LM, Driscoll CT. (2005) Regional assessment of the response of the acid-base status of lake watersheds in the adirondack region of New York to changes in atmospheric deposition using PnET-BGC. Environmental Science and Technology 39:787-794. Clarke KR. (1993) Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18:117-143. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Cumming BF, Smol JP, Birks HJB. (1992a) Scaled chrysophytes (Chrysophyceae and Synurophyceae) from Adirondack drainage lakes and their relationship to environmental variables. Journal of Phycology 28:162178. Cumming BF, Smol JP, Kingston JC, Charles DF, Birks HJB, Camburn KE, Dixit SS, Uutala AJ, Selle AR. (1992b) How much acidification has occurred in Adirondack region lakes (New York, USA) since preindustrial times?. Canadian Journal of Fisheries and Aquatic Sciences 49:128-141. Davis RB, Anderson DS, Dixit SS, Appleby PG, Schauffler M. (2006) Responses of two New Hampshire (USA) lakes to human impacts in recent centuries. Journal of Paleolimnology 35:669-697. Downes BJ, Barmuta LA, Fairweather PG, Faith DP, Keough MJ, Lake PS, Mapstone BD, Quinn GP. (2002) Monitoring ecological impacts: concept and practice in flowing waters. Cambridge University Press, Cambridge, 452 p. 161 Driscoll CT, van Dreason R. (1993) Seasonal and long-term temporal patterns in the chemistry of Adirondack lakes. Water, Air, and Soil Pollution 67:319-344. Driscoll CT, Driscoll KM, Roy KM, Dukett J. (2007) Changes in the chemistry of lakes in the Adirondack region of New York following declines in acidic deposition. Applied Geochemistry 22:1181-1188. Driscoll CT, Lawrence GB, Bulger AJ, Butler TJ, Cronan CS, Eagar C, Lambert KF, Likens GE, Stoddard JL, Weathers KC. (2001) Acidic deposition in the northeastern United States: sources and inputs, ecosystem effects, and management strategies. BioScience 51:180-198. Driscoll CT, Newton RM, Gubala CP, Baker JP, Christensen S. (1991) Adirondack mountains. In: Charles DF (ed), Acidic deposition and aquatic ecosystems: regional case studies. Springer-Verlag, New York, pp. 133-202. Driscoll CT, Postek KM, Kretser W, Raynal DJ. (1995) Long-term trends in the chemistry of precipitation and lake water in the Adirondack region of New York, USA. Water, Air, and Soil Pollution 85:583-588. Fee EJ. (1976) Vertical and seasonal distribution of chlorophyll in lakes of Experimental Lakes Area, Northwestern Ontario - implications for primary production estimates. Limnology and Oceanography 21:767-783. Flear K. (2011) Changes in scaled-chrysophyte assemblages in response to recent climate change in Northwestern Ontario. Department of Biology. Queen's University Kingston, Ontario, 121 p. Garmo ØA, Skjelkvåle BL, de Wit HA, Colombo L, Curtis C, Fölster J, Hoffmann A, Hruška J, Høgåsen T, Jeffries DS, Keller WB, Krám P, Majer V, Monteith DT, Paterson AM, Rogora M, Rzychon D, Steingruber S, Stoddard JL, Vuorenmaa J, Worsztynowicz A. (2014) Trends in surface water chemistry in acidified areas in Europe and North America from 1990 to 2008. Water, Air, and Soil Pollution 225:1880-1-1880-14. Gerten D, Adrian R. (2002) Effects of climate warming, North Atlantic Oscillation, and El Niño-Southern Oscillation on thermal conditions and plankton dynmaics in northern hemispheric lakes. The Scientific World Journal 2:586-606. Ginn BK, Rate M, Cumming BF, Smol JP. (2010) Ecological distribution of scaled-chrysophyte assemblages from the sediments of 54 lakes in Nova Scotia and southern New Brunswick, Canada. Journal of Paleolimnology 43:293-308. Glew JR, Smol J, Last WM. (2001) Sediment core collection and extrusion. In: Last WM, Smol J (eds), Tracking environmental change using lake sediments. Volume 1: basin analysis, coring, and chronological techniques Kluwer Academic Publishers, Dordrecht., pp. 73-105. 162 Heit M, Miller KM. (1987) Cesium-137 sediment depth profiles and inventories in Adirondack lake sediments. Biogeochemistry 3:243-265. Helsel DR, Hirsch RM. (2002) Correlation. Statistical methods in water resources techniques of water resources investigations, book 4, chapter A3. U.S. Geological Survey, pp. 209-220. Helsel DR, Mueller DK, Slack JR. (2006) Computer program for the Kendall family of trend tests. U.S. Geological Survey, Reston, Virginia, 4 p. Hobbs RJ, Higgs E, Harris JA. (2009) Novel ecosystems: implications for conservation and restoration. Trends in Ecology & Evolution 24:599-605. Holen DA, Boraas ME. (1995) Mixotrophy in chrysophytes. In: Sandgren CD, Smol JP, Kristiansen J (eds), Chrysophyte algae: ecology, phylogeny and development. Cambridge University Press, Cambridge, United Kingdom, pp. 119-140. Hyatt CV, Paterson AM, Cumming BF, Smol JP. (2010) Factors related to regional and temporal variation in the distribution of scaled chrysophytes in north-eastern North America: Evidence from lake sediments. Nova Hedwigia, Beiheft 136:87–102. Jenkins J. (2010) Climate change in the Adirondacks: the path to sustainability. Cornell University Press, Ithaca, New York, 183 p. Jones HP, Schmitz OJ. (2009) Rapid recovery of damaged ecosystems. PLoS ONE 4:e5653-1- e5653-6. Keller W. (2009) Limnology in northeastern Ontario: from acidification to multiple stressors. Canadian Journal of Fisheries and Aquatic Sciences 66:1189-1198. Kling HJ, Kristiansen J. (1983) Scale–bearing Chrysophyceae (Mallomonadaceae) from Central and Northern Canada. Nordic Journal of Botany 3:269-290. Lovett GM, Tear TH, Evers DC, Findlay SEG, Cosby BJ, Dunscomb JK, Driscoll CT, Weathers KC. (2009) Effects of air pollution on ecosystems and biological diversity in the eastern United States. Annals of the New York Academy of Sciences 1162:99-135. Menne MJ, Williams CN, Vose RS. (2009) The U.S. Historical Climatology Network monthly temperature data, version 2. Bulletin of the American Meteorological Society 90:993–1007. Nicholls KH, Gerrath JF. (1985) The taxonomy of Synura (Chrysophyceae) in Ontario with special reference to taste and odor in water-supplies. Canadian Journal of Botany 63:1482-1493. 163 Paterson AM, Cumming BF, Smol JP, Blais JM, France RL. (1998) Assessment of the effects of logging, forest fires and drought on lakes in northwestern Ontario: a 30-year paleolimnological perspective. Canadian Journal of Forest Research 28:1546-1556. Paterson AM, Cumming BF, Smol JP, Hall RI. (2001) Scaled chrysophytes as indicators of water quality changes since preindustrial times in the Muskoka-Haliburton region, Ontario, Canada. Canadian Journal of Fisheries and Aquatic Sciences 58:2468-2481. Paterson AM, Cumming BF, Smol JP, Hall RI. (2004) Marked recent increases of colonial scaled chrysophytes in boreal lakes: implications for the management of taste and odour events. Freshwater Biology 49:199-207. Paterson AM, Winter JG, Nicholls KH, Clark BJ, Ramcharan CW, Yan ND, Somers KM. (2008) Long-term changes in phytoplankton composition in seven Canadian Shield lakes in response to multiple anthropogenic stressors. Canadian Journal of Fisheries and Aquatic Sciences 65:846-861. Renberg I, Bigler C, Bindler R, Norberg M, Rydberg J, Segerström U. (2009) Environmental history: a piece in the puzzle for establishing plans for environmental management. Journal of Environmental Management 90:2794-2800. Roy K, Houck N, Hyde P, Cantwell M, Brown J. (2011) The Adirondack long-term monitoring lakes: a compendium of site descriptions, recent chemistry and selected research information. New York State Energy Research and Development Authority, Albany, 298 p. Schelske CL, Peplow A, Brenner M, Spencer CN. (1994) Low-background gamma counting: applications for 210Pb dating of sediments. Journal of Paleolimnology 10:115-128. Siver PA. (1991) The biology of Mallomnas: morphology, taxonomy and ecology. Kluwer Academic Publishers, Dordrecht, The Netherlands, 248 p. Siver PA. (1993) Inferring the specific conductivity of lake water with scaled chrysophytes. Limnology and Oceanography 38:1480-1492. Siver PA. (2003) Synurophyte algae. In: Wehr JD (ed), Freshwater algae of North America: ecology and classification. Academic Press, San Diego, California, pp. 523-558. Siver PA, Hamer JS. (1992) Seasonal periodicity of chrysophyceae and synurophyceae in a small New England lake: implications for paleolimnological research. Journal of Phycology 28:186-198. 164 Skjelkvåle BL, Stoddard JL, Jeffries DS, Tørseth K, Høgasen T, Bowman J, Mannio J, Monteith DT, Mosello R, Rogora M, Rzychon D, Vesely J, Wieting J, Wilander A, Worsztynowicz A. (2005) Regional scale evidence for improvements in surface water chemistry 1990–2001. Environmental Pollution 137:165-176. Smol JP. (2008) Polluion of lakes and rivers: a paleoenvironmental perspective. Blackwell Publishing Ltd, Malden. Sommer U. (1988) Some size relationships in phytoflagellate motility. Hydrobiologia 161:125-131. Stager JC, Sanger T. (2003) An Adirondack "heritage lake". Adirondack Journal of Environmental Studies 10:6-10. Sullivan TJ, Cosby BJ, Driscoll CT, McDonnell TC, Herlihy AT, Burns DA. (2012) Target loads of atmospheric sulfur and nitrogen deposition for protection of acid sensitive aquatic resources in the Adirondack Mountains, New York. Water Resources Research 48:W01547-1- W01547-16. ter Braak CJF, Šmilauer P. (2012) CANOCO reference manual and user's guide: software for ordination (version 5.0). Microcomputer Power, Ithaca, New York, 496 p. United States Environmental Protection Agency (USEPA). (2012) Implications of climate change for state bioassessment programs and approaches to account for effects (final report). U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-11/036F. Verdonschot PFM, Spears BM, Feld CK, Brucet S, Keizer-Vlek H, Borja A, Elliott M, Kernan M, Johnson RK. (2013) A comparative review of recovery processes in rivers, lakes, estuarine and coastal waters. Hydrobiologia 704:453-474. Waller K, Driscoll C, Lynch J, Newcomb D, Roy K. (2012) Long-term recovery of lakes in the Adirondack region of New York to decreases in acidic deposition. Atmospheric Environment 46:56-64. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. 165 Tables & Table Captions Table 5.1. Analysis of Similarity (ANOSIM) and Similarities Percentage (SIMPER) test results assessing whether chrysophyte assemblages have changed between two a priori defined time periods (Period 1 and Period 2). Time periods pre-1995 (Period 1, P1) and post-1995 (Period 2, P2) are listed for each lake with the number of samples per period (n). ANOSIM results (test statistic, R, and p-value) are given. ANOSIM results significant at p < 0.05 are highlighted in bold. SIMPER post-hoc test results are provided. Taxa that made a > 5% contribution (% Cont.) to the dissimilarity between groups are listed. The average square-root relative abundance (%) of the species in Period 1 (Av √RA P1) and Period 2 (AV √RA P2) is provided, with the non square-root transformed Time Periods ANOSIM results R p P1 P2 South Lake 1993-1981 (n = 7) 2008-1998 (n = 10) 0.828 0.001 Queer Lake 1993-1981 (n = 8) 2009-1997 (n = 7) 0.682 0.002 SIMPER Results Species M. hamata M. hindonii M. duerrschmidtiae S. echinulata S. sphagnicola M. acaroides M. caudata M. ‘small’ C. synuroides S. petersenii M. caudata M. punctifera Chrysosphaerella M. duerrschmidtiae S. echinulata M. hamata M. ‘small’ M. pseudocoronata 166 Av √RA P1 (%) 4.6 (26.2) 2.0 (4.2) 6.8 (46.2) 4.2 (17.6) 0.8 (0.6) 1.3 (1.7) 2.3 (5.3) 1.0 (1.0) 1.1 (1.2) 2.6 (6.8) 3.5 (12.3) 1.6 (2.6) 1.0 (1.0) 8.1 (65.6) 0.9 (0.8) 2.1 (4.4) 1.2 (1.4) 0.8 (0.6) Av √RA P2 (%) 3.2 (10.2) 1.4 (2.0) 7.5 (56.3) 4.7 (22.1) 0.2 (0.04) 1.0 (1.0) 2.1 (4.4) 0.9 (0.8) 1.2 (1.4) 4.1 (16.8) 2.8 (7.8) 2.2 (4.8) 1.5 (2.3) 7.6 (57.8) 1.4 (2.0) 1.7 (2.9) 0.9 (0.8) 0.6 (0.4) % Cont. 24.8 12.7 12.3 10.4 8.2 7.2 7.0 6.8 5.8 25.0 12.3 9.8 9.5 9.2 8.6 7.6 7.5 7.2 Table 5.2. Kendall τ correlation test results. Correlations between the total relative abundance of colonial chrysophyte taxa (% colonial) and mean annual air temperature (TMEAN, in °C), ice cover parameters (ICE ON, ice-on date; ICE OFF, ice-off date; ICE DUR, ice cover duration) are provided for Arbutus Lake and Wolf Lake. Correlation coefficient (τ), p-value (p), and sample size (n) are provided. Correlations significant at p < 0.05 are highlighted in bold (positive correlation, positive τ; negative correlation, negative τ). % colonial vs. TMEAN (ca. 1970-present) % colonial vs. ICE ON (ca. 1970-2007) % colonial vs. ICE OFF (ca. 1970-2007) % colonial vs. ICE DUR (ca. 1970-2007) Arbutus Lake τ p n 0.28 0.03 32 τ 0.28 0.31 0.02 27 0.08 0.61 25 0.03 0.87 27 -0.25 0.07 26 0.06 0.71 25 -0.41 0.01 24 167 Wolf Lake p n 0.03 30 Figures & Figure Captions Figure 5.1. Long-term chemistry trends in acidified and reference lakes. Chemistry data for A: South Lake (●) and Queer Lake (●) are from 1992-2010 and, B: Moss Lake (■) and Arbutus Lake (■) are from 1982/1983-2010. Sulphate and nitrate (SO42- + NO3-, μeq L-1), pH, acid neutralizing capacity (ANC, μeq L-1), inorganic monomeric aluminium (Ali, μg L-1), dissolved organic carbon (DOC, mg L-1) and specific conductance (Sp. Cond., μS cm-1) are shown. 168 Figure 5.2. 210Pb and 214Bi activities in sediment cores from acidified and reference lakes. Unsupported 210Pb (●) and 214Bi (■) activities (Bq per gram dry sediment, Bq g-1) plotted against cumulative dry mass (g cm-2) in sediment cores from acidified lakes, A: South Lake (open symbols) and Queer Lake (black symbols) and reference lakes, B: Moss Lake (grey symbols), Arbutus Lake (open symbols), and Wolf Lake (black symbols). Dates are inferred using the constant rate of supply Model (CRS). CRS-Inferred dates are plotted against sediment interval mid-point (cm) in sediment cores from acidified lakes, C: South Lake (○) and Queer Lake (●) and reference lakes, D: Moss Lake (▲), Arbutus Lake (○), and Wolf Lake (●). Age-depth models for Moss Lake, South Lake, and Arbutus Lake were based on a simple quadratic function (solid lines). Queer Lake and Wolf Lake were fitted with a cubic functions (solid lines), as quadratic equations (dashed lines) appeared to underestimate sediment-age in the top 6-7 cm of the core. Error bars represent the standard error. 169 170 Figure 5.3. Total relative abundance of colonial chrysophytes (% colonial) and chrysophyte-inferred pH (CI-pH) values for two acidified lakes (South Lake and Queer Lake) and three reference lakes (Arbutus Lake, Wolf Lake, Moss Lake). Acidified and reference lakes are separated by a grey, dashed line. Samples are arranged by 210Pb-inferred date. % Colonial data CI-pH data are presented at an ~ 5 year resolution post-1970, and an ~10-20 year resolution is present pre-1970 for all lakes. CI-pH is based on a calibration set of 71 Adirondack lakes (Cumming et al. 1992a). CI-pH was calculated by a weighted-averaging function with tolerance down-weighting, inverse deshrinking and a ln (x + 1) species transformation (r2 boot = 0.75, RMSEP = 0.51). Figure 5.4. Non-metric Multidimensional Scaling (nMDS) plots for acidified (A, South Lake; B, Queer Lake) and reference lakes (C, Moss Lake; D, Arbutus Lake; E, Wolf Lake). Species assemblages are divided into 4 time periods: pre-disturbance (pre-1900, □), early disturbance (1900 to pre-1950, Δ), acidification (1950 to 1995, ♦), and post-1995 implementation of the US Acid Rain Program (post-1995, ●). Species codes are provided in Appendix I. Species vectors are plotted passively in the ordination space. Stress values are provided. 171 b nMDS-2 nMDS-2 a Stress = 0.09 Stress = 0.17 nMDS-1 nMDS-1 Figure 5.5. non-metric Multidimensional Scaling (nMDS) plots relating to South Lake species assemblages. A) pre-1900 South Lake assemblages (shown as black stars) in comparison to all pre-1900 samples from a set of 26 Adirondack reference lakes. The two reference sites with the highest average similarity in comparison with the pre-1900 assemblages of South Lake are shown as grey squares. B) Long-term changes in chrysophyte species assemblages in South Lake (black circles). Note that data are presented at a sub-decadal resolution post-1970 and an approximately 10-20 resolution pre-1970 to reduce crowding in ordination plot. Present-day (‘top’; T samples) and pre-1900 (‘bottom’; B samples) samples for two reference lakes (NellP and RounP; white circles) are shown. Species codes are provided in Appendix I. Samples are plotted actively in the ordination space and species vectors are plotted passively in the ordination space. Stress values are provided. 172 a b nMDS-2 nMDS-2 nMDS-2 c Stress = 0.17 nMDS-1 Stress = 0.07 nMDS-1 Stress = 0.07 nMDS-1 Figure 5.6. Non-metric Multidimensional Scaling (nMDS) plots relating to Queer Lake species assemblages. A) pre-1900 Queer Lake assemblages (shown as black stars) in comparison to all pre-1900 samples from a set of 26 Adirondack reference lakes. The two reference sites with the highest average similarity in comparison with the pre-1900 assemblages of Queer Lake are shown as grey squares. B) Long-term changes in chrysophyte species assemblages in Queer Lake (black circles) versus Wolf Lake and C) Moss Lake (white circles). Note that data are presented at a sub-decadal resolution post-1970 and an approximately 10-20 resolution pre1970 to reduce crowding in ordination plot. Species codes are provided in Appendix I. Samples are plotted actively in the ordination space and species vectors are plotted passively in the ordination space. Stress values for the nMDS ordinations are provided. 173 CHAPTER 6 – GENERAL DISCUSSION General Discussion The 1990 Clean Air Act Amendments are arguably some of the most successful pollution control measures ever implemented in the United States. With declines in sulphate emissions, some areas in the United States are undergoing chemical recovery from acidification (Garmo et al. 2014; Skjelkvåle et al. 2005). As such, there is a pressing need to define recovery targets for acid-impacted lakes. Researchers trying to define recovery targets from acidification must overcome two issues: 1) a lack of long-term monitoring data, and 2) the confounding influence of multiple stressors. The difficulty posed by a lack of long-term monitoring data can be addressed by applying paleolimnological techniques in recovery studies and regional reference sites can be used to clarify the problem of multiple stressors. This dissertation aimed to bridge the paleolimnological and bioassessment literature by applying paleolimnological techniques in a study of a regional set of minimally-disturbed reference lakes in the Adirondacks (NY, USA). The main goal of this dissertation was to use the regional reference sites to track shifting baselines due to regional stressors like climate change and use that information to define recovery targets for lakes recovering from acidification in the Adirondack Park. Chapter 3 introduced the conceptual framework underlying this project and critically evaluated the 31 reference sites using both qualitative and quantitative approaches; Chapter 4 provided the results from a top-bottom study examining shifting baselines in Adirondack reference lakes, and; Chapter 5 provided an application of the reference site approach to defining recovery targets by pairing a set of reference lakes and a set of acidimpacted lakes recovering from acidification. Chapter 3 outlined the conceptual framework applied in this investigation. The premise behind this project was that a set of minimally-disturbed reference sites embedded in a region 174 impacted by multiple stressors could be used to track shifting baselines due to regional stressors like climate change; information that could, in turn, be used to define reasonable recovery targets for lakes recovering from acidification. For example, if the regional reference sites exhibited increases in warm-water taxa or motile taxa, changes predicted to occur with regional warming (Winder and Sommer 2012), those changes should be included in recovery targets for acidified lakes in the region. In essence, acidified lakes would be expected to recover to a ‘novel’ state similar to that of the reference sites with chemical recovery, rather than return to a predisturbance state. The approach adopted in this investigation is powerful because it bases recovery targets on historical data, a long-term temporal perspective, and the use of reference sites, three components Gray and Arnott (2009) argued would be ideal for studying recovery and defining recovery targets. A set of stringent selection criteria were used to identify a set of Adirondack lakes that have been protected from: acidification, eutrophication (from shoreline and watershed development), road salt seepage, and piscivores introductions. However, a critical evaluation of the reference sites using both qualitative and quantitative methods revealed that most of the reference lakes had experienced some form of historic fisheries or watershed disturbance, highlighting the danger of assuming that a site’s present-day condition is characteristic of its earlier history (Leira et al. 2006). Ordination methods found that the reference sites were representative of between 24-36% of the chemical and/or physical characteristics of the ALS lakes. The need to make meaningful comparisons between reference lakes and impacted sites was discussed, as was the possibility of implementing similar frameworks in other regions. In Chapter 4, a top-bottom paleolimnological investigation of the reference lakes demonstrated that a shifting baseline effect was occurring in Adirondack reference lakes. Both 175 univariate and multivariate analyses showed that there has been a significant shift in species composition, with present-day sediments characterized by an increased relative abundance of colonial chrysophyte taxa. The changes in species composition noted in the reference lakes from pre-1900 to present were greater than expected from counting variability, and so merited further investigation. Because of the reference lake study design, it was possible to conclude that the changes in species composition were not due to acidification (decreases in pH/ANC), eutrophication (from watershed or shoreline development), road salt seepage, and piscivores introductions. Furthermore, ANOSIM results found that historic watershed disturbance and fisheries disturbance had minimal influence on chrysophyte species composition in the presentday samples. The increases in colonial taxa noted in the investigation were likely due to either regional warming and/or oligotrophicaion resulting from increased adsorption of TP in watershed soils with long-term exposure to acid deposition and/or logging followed by subsequent forest regrowth. Both regional warming and oligotrophication could favour colonial chrysophytes and indeed would likely have interactive effects (e.g. a longer period of thermal stratification leading to nutrient depletion in the epilimnion). Regardless of the causal mechanism(s) for the species changes noted, the results of the top-bottom investigation suggest that lakes recovering from acidification in the region are unlikely to return to a pre-disturbance state. Rather, the lakes are likely to show an increased abundance of colonial taxa and/or warmwater species as recovery progresses. While an increase in colonial taxa from pre-1900 to present was seen in a majority of the reference lakes, both physical and chemical factors influence the distribution of chrysophyte taxa in Adirondack lakes. An RDA of the reference sites found that S. petersenii occurs in deeper lakes with higher pH values and DOC values less than 5.2 mg C L-1. In contrast, S. sphagnicola 176 occurred in sites with lower pH values, both in low and high DOC sites. Because many Adirondack lakes are shallow sites with pH values less than 6 (Kreseter et al. 1989), S. petersenii may not be widespread in Adirondack surface waters. Taste and odour issues caused by S. petersenii may, however, pose a problem in recreationally-important large, deep, alkaline lakes that are home to important sportfish taxa like lake trout (Salvelinus namaycush). Chapter 5 provides an application of the reference site approach that pairs a set of three reference sites with two lakes (South Lake & Queer Lake) which acidified and are undergoing chemical recovery from acidification. ANOSIM found that both of the acidified lakes had undergone a significant shift in species composition since the 1995 implementation of the US ARP. SIMPER found that the change in species composition was due both to a decline in the relative abundance of acid-tolerant taxa (M. hamata, M. hindonii), as well as increases in colonial chrysophyte species. Plotting the long-term species trajectories of the acid-impacted sites with two reference sites that had similar species assemblages pre-1900 in an nMDS ordination revealed that the probable recovery endpoints for South Lake and Queer Lake include an increased abundance of colonial taxa and/or warm-water taxa, rather than a return to predisturbance state. The results of Chapter 5 support statements made in Chapters 3 and 4 that acid-impacted lakes in the Adirondacks are unlikely to return to a pre-disturbance state. Paleolimnological reconstructions from all five study lakes showed an increase in colonial chrysophyte taxa since ca. 1970-1980. The fact that increases in colonial taxa occurred both in lakes that acidified and reference sites again highlights that the changes are not due to acidification and recovery. Similarly, the increases in colonial taxa were unlikely due to declines in specific conductance and though DOC may influence the abundance of S. petersenii, increases in DOC cannot explain 177 increases in the total relative abundance of colonial chrysophytes. Increases in colonial taxa were correlated to ice-cover measures and mean annual air temperature in two reference lakes where long-term monitoring data were available, linking increases in colonial taxa to regional warming, though as noted in Chapter 4, these changes may also be attributable to oligotrophication. Pre1950 peaks in colonial chrysophyte abundance noted in some lakes may be related to early warming and/or watershed disturbance but disturbance cannot explain recent increases in colonial chrysophytes as the study lakes have been protected from watershed disturbance since the 1930s (Wolf Lake and Arbutus Lake) and 1970s (South Lake, Queer Lake, Moss Lake). This dissertation has made several important contributions to the ecological literature, including improving our understanding of biological recovery from acidification (a subject that is not well documented, Verdonschot et al. 2013), and our understanding of the influence of climate warming on algal assemblages, a topic of growing research interest (Winder and Sommer 2012). Perhaps most globally, this dissertation provides an effective framework for examining the effects of multiple stressors on biological recovery in lakes that can be applied to other study other stressors in other regions. The application of this approach in the Adirondacks relied on three factors: the availability of data from a large synoptic survey, previous limnological/paleolimnological investigations to provide a basis for the reference lake selection criteria used, and historical watershed and fisheries disturbance data. Datasets like the National Lakes Assessment in the United States (USEPA 2011) or the European Diatom Database (Battarbee et al. 2014) highlight the growing availability of regional survey data. In cases where monitoring programs have already been established, including a paleolimnological component in such studies will provide valuable information about how both impacted and reference lakes have changed over time, tracking the ‘shifting baseline’ phenomnenon in response to regional 178 warming (USEPA 2012). However, the paleolimnological approach does have certain limitations. First, certain taxa of interest (i.e. soft-bodied zooplankton like copepods) may not be preserved in the sedimentary record, and so ideally the paleolimnological perspective should be coupled with long-term biological monitoring data (Gray & Arnott 2009). Second, the paleolimnological approach does require specialized equipment and knowledge, which can make the implementation of paleolimnological studies difficult (i.e., depending on the approach: taxonomic knowledge of subfossil species assemblages, dating techniques, multivariate data analysis, etc.). While the use of paleolimnological data in the National Lakes Assessment in the United States (USEPA 2011) and studies related to the implementation of the European Water Framework Directive (Bennion et al. 2011) highlight that there is a growing use of paleolimnological techniques in limnological assessment studies, long-standing methods like the use of indices of biotic integrity are more widely used by lake managers (Soranno et al. 2011). While such indices may be simpler to interpret than the multivariate species assemblage data commonly used in paleolimnological studies, they may also mask important ecological information. Likewise, if transfer functions are used in paleolimnological work to reconstruct inferred pH (or TP, etc.) values, such changes should be interpreted within the context of overall species assemblage change so that important ecological information is not overlooked (Quinlan et al. 2008). Lastly, despite the utility of the reference site design, there can be complexities when interpreting long-term species changes. For example, in Chapter 5 both early watershed disturbance and regional warming could have contributed to pre-1950 increases in the relative abundance of colonial chrysophytes in some lakes. Such complexity illustrates the importance of building a long-term perspective of land-use and land-cover change (Renberg et al. 2009). However, if early watershed disturbances have a lasting chemical/biological legacy in lakes, this 179 complexity is arguably present in most aquatic bioassessment work, regardless of whether or not a paleolimnological approach is applied (Martin et al. 2011). Having long-term paleolimnological data can, at least, allow researchers and lake managers to test for the effect of early disturbances (vs. regional warming) on historic and present day species assemblages, if a suitable experimental design is implemented (see example below). Overall, by combining the long-term perspective of paleolimnological studies with a rigorous reference site study design, this framework was used to both identify shifting baseline conditions in reference lakes and define reasonable recovery targets for impacted lakes, a critical research and management need. The results of this project also suggest several interesting possibilities for expansions that could be made to this work, which are discussed below. Future Directions Expanding the reference-lake top-bottom study: The first obvious expansion point for this project would be to include additional paleolimnological proxies in top-bottom studies of the reference sites to examine if shifting baselines are occurring in other species groups in Adirondack lakes. A group of particular interest would be the cladoceran zooplankton. While zooplankton may be sensitive to climate warming (Moore et al. 1996; Daufresne et al. 2009), the response of cladoceran to climate change is likely to be complex in regions influenced by acid deposition. Long-term exposure to acid deposition can lead to calcium depletion in watershed soils (Likens 1996). Many large daphniids are Ca-sensitive organisms because calcium is an integral structural component of their carapaces (Jeziorski and Yan 2006). While calcium-depletion in the Adirondack region could lead to a reduced abundance of Daphnia in impacted lakes (Jeziorski et al. 2008), it can be difficult to disentangle the effect of low calcium levels from acidification. Low pH values frequently co-occur with low Ca levels and most daphniids are pH-sensitive 180 organisms (Jeziorski et al. 2012). Here, this project can make a valuable contribution because the lakes span a calcium gradient (< 1 mg L-1 to > 5 mg L-1) and the majority of lakes are unlikely to have acidified. Moreover, the selection criteria used to identify the reference lakes included a depth criterion to increase the probability that daphniids would occur in the sites (see Chapter 3). This project could offer an opportunity to disentangle a Ca-effect from a low-pH effect and clarify our understanding of cladoceran responses to regional warming. Identifying the influence of oligotrophication on Adirondack lakes: The results of Chapter 4 and Chapter 5 suggest that climate warming and/or oligotrophication as a long-term consequence of acid deposition could have promoted the increases in colonial chrysophyte relative abundance noted in the lakes in this investigation. While there is evidence that regional warming is having an influence on Adirondack lakes (for example, by reducing the length of the ice-covered period, Beier et al. 2012), there is limited information on long-term TP trends in the region. Diatominferred TP reconstructions in the reference lakes could be used to identify whether or not oligotrophication is occurring in Adirondack lakes. Dixit et al. (1999) identified that the proportion of Adirondack lakes classified as oligotrophic or mesotrophic did not change from pre-industrial times to present, though lakes within each category showed both increases and decreases in diatom-inferred TP. However, the diatom flora may not be sufficiently sensitive to track ecologically-important but subtle changes in TP. For example, while a reduction in TP levels from 4 μg L-1 to 2 μgL-1 would likely be ecologically significant (a 50% reduction in available TP), each state would be characterized by ultra-oligotrophic diatom taxa and the change in inferred-TP could be within the error of the inference method (e.g. ln TP RMSEboot = 0.79, Dixit et al. 1999). For this reason, modelling studies may be needed to identify what influence, if any, oligotrophication from long-term exposure to acid deposition is having on 181 Adirondack lakes. Mesocosm studies could also be undertaken to evaluate chrysophyte species responses to both oligotrophication and changes in stratification patterns (e.g. nutrient-reduction trials, vs. experimental mixing/stratification manipulations). Disentangling the effects of early logging vs. early warming: In Chapter 5, both a reference site (Moss Lake) and acid-impacted lake (South Lake) show a pre-1950 peak in colonial chrysophyte abundance, a trend noted in previous paleolimnological studies of the region (Cumming et al. 1994). Such early increases in colonial chrysophyte may have been due to early warming pre1950 and/or early watershed disturbance. The effect of these two stressors could be disentangled using the reference site approach and a balanced study design. A detailed paleolimnological study of the four Class 1A lakes could be compared with four suitable Class 2 or Class 3 reference lakes that are known to have experienced early watershed disturbance (see Chapter 3). Ideally sampling with an equal temporal resolution, a comparison of the long-term changes in species assemblages in these lakes types would help identify what influence, if any, early warming and/or watershed disturbance had on chrysophytes assemblages. Chrysophyte species change at a Holocene-level temporal resolution: The results of this investigation suggest that Adirondack lakes have undergone a significant shift in species composition over the past 100-150 years, likely due to regional warming and/or oligotrophication. However, the reference lake analysis could be expanded to include a Holocene-level chrysophyte assemblage reconstruction. Such analysis would provide a more complete view of the temporal variability of chrysophyte species assemblages. Moreover, an analysis of time periods that were on average warmer than the present-day like the mid-Holocene could provide a perspective on the likely species composition of Adirondack lakes in a 182 significantly warmer environment. In 2012, a piston core whose sedimentary record covers the entire Holocene was collected from Wolf Lake. Analysis of this core will provide what is, to our knowledge, the first low-elevation Holocene-perspective on an Adirondack Lake. Analysis of the Wolf Lake core, and other Holocene records from the Adirondacks, will be able to provide us with a critical insight not only into millennial-level changes in species composition in Adirondack lakes, but also changes in regional vegetation and precipitation patterns. As always, the long-term perspective of paleolimnological work will be critically important in evaluating the effects of ongoing global environmental change on aquatic ecosystems. Literature Cited Beier CM, Stella JC, Dovčiak M, McNulty SA. (2012) Local climatic drivers of changes in phenology at a borealtemperate ecotone in eastern North America. Climatic Change 115:399-417. Bennion H, Battarbee RW, Sayer CD, Simpson GL, Davidson TA. (2011) Defining reference conditions and restoration targets for lake ecosystems using palaeolimnology: a synthesis. Journal of Paleolimnology 45:533-544. Cumming BF, Davey KA, Smol JP, Birks HJB. (1994) When did acid-sensitive Adirondack lakes (New York, USA) begin to acidify and are they still acidifying?. Canadian Journal of Fisheries and Aquatic Sciences 51:1550-1568. Daufresne M, Lengfellner K, Sommer U. (2009) Global warming benefits the small in aquatic ecosystems. Proceedings of the National Academy of Sciences 106:12788-12793. Dixit SS, Smol JP, Charles DF, Hughes RM, Paulsen SG, Collins GB. (1999) Assessing water quality changes in the lakes of the northeastern United States using sediment diatoms. Canadian Journal of Fisheries and Aquatic Sciecnes 56:131-152. Garmo ØA, Skjelkvåle BL, de Wit HA, Colombo L, Curtis C, Fölster J, Hoffmann A, Hruška J, Høgåsen T, Jeffries DS, Keller WB, Krám P, Majer V, Monteith DT, Paterson AM, Rogora M, Rzychon D, Steingruber S, Stoddard JL, Vuorenmaa J, Worsztynowicz A. (2014) Trends in surface water chemistry in acidified areas in Europe and North America from 1990 to 2008. Water, Air, and Soil Pollution 225:1880-1-1880-14. 183 Gray DK, Arnott SE. (2009) Recovery of acid damaged zooplankton communities: measurement, extent, and limiting factors. Environmental Reviews 17:81-99. Jeziorski A, Paterson AM, Smol JP. (2012) Crustacean zooplankton sedimentary remains from calcium-poor lakes: complex responses to threshold concentrations. Aquatic Sciences 74:121-131. Jeziorksi A, Yan ND. (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 ND, Paterson AM, DeSellas AM, Turner MA, Jeffries DS, Keller B, Weeber RC, McNicol DK, Palmer ME, McIver K, Arseneau K, Ginn BK, Cumming BF, Smol JP. (2008) The widespread threat of calcium decline in fresh waters. Science 322:1374-1377. Leira M, Jordan P, Taylor D, Dalton C, Bennion H, Rose N, Irvine K. (2006) Assessing the ecological status of candidate reference lakes in Ireland using palaeolimnology. Journal of Applied Ecology 43:816-827. Likens GE, Driscoll CT, Buso DC. (1996) Long-term effects of acid rain: response and recovery of a forest ecosystem. Science 272:244-245. Martin SL, Hayes DB, Rutledge DT, Hyndman DW. (2011) The land-use legacy effect: adding temporal context to lake chemistry. Limnology and Oceanography 56:2362-2370. Moore MV, Folt CL, Stemberger RS. (1996) Consequences of elevated temperatures for zooplankton assemblages in temperate lakes. Archiv für Hydrobiologie 135:289-319. Quinlan R, Hall RI, Paterson AM, Cumming BF, Smol JP. (2008) Long-term assessments of ecological effects of anthropogenic stressors on aquatic ecosystems from paleoecological analyses: challenges to perspectives of lake management. Canadian Journal of Fisheries and Aquatic Sciences 65:933-944. Renberg I, Bigler C, Bindler R, Norberg M, Rydberg J, Segerström U. (2009) Environmental history: a piece in the puzzle for establishing plans for environmental management. Journal of Environmental Management 90:2794-2800. Skjelkvåle BL, Stoddard JL, Jeffries DS, Tørseth K, Høgasen T, Bowman J, Mannio J, Monteith DT, Mosello R, Rogora M, Rzychon D, Vesely J, Wieting J, Wilander A, Worsztynowicz A. (2005) Regional scale evidence for improvements in surface water chemistry 1990–2001. Environmental Pollution 137:165-176. 184 Soranno PA, Wagner T, Martin SL, McLean C, Novitski LN, Provence CD, Rober AR. (2011) Quantifying regional reference conditions for freshwater ecosystem management: a comparison of approaches and future research needs. Lake and Reservoir Management 27:138-148. United States Environmental Protection Agency (USEPA). (2011) National Lakes Assessment 2012: a fact sheet for communities. USEPA Office of Water, Monitoring Branch, Washington, DC, EPA 841-F-11-007. Verdonschot PFM, Spears BM, Feld CK, Brucet S, Keizer-Vlek H, Borja A, Elliott M, Kernan M, Johnson RK. (2013) A comparative review of recovery processes in rivers, lakes, estuarine and coastal waters. Hydrobiologia 704:453-474. Winder M, Sommer U. (2012) Phytoplankton response to a changing climate. Hydrobiologia 698:5-16. 185 APPENDICES Appendix A - Figure A1. Histogram plots of Adirondack Lakes Survey (ALS) variables. Morphological variables are elevation (elv, m) maximum depth (max depth, m), surface area (Area, ha), and watershed area (WArea, ha). Chemical variables are field pH, SO42-, F-, Cl-, Na+, K+, Ca2+, Fe2+ (mg L-1), DIC (dissolved inorganic carbon, mg L-1), DOC (dissolved organic carbon, mg C L-1) TP (total phosphorus, mg L-1) and specific conductance (SCONDUCT, µmhos cm-1). Transformations are listed in histogram plots. Data were log (x + 1) transformed if minimum value in ALS survey > 1; log (x + ½ minimum value) if 1 > minimum value > 0.1; log (x + 0.001) if smallest non-zero recorded value > 0.001; and, log (x + 0.0001) if smallest nonzero recorded value > 0.0001. If a variable included negative values after log transformation, a constant (the absolute value of the minimum log-transformed data) was added to the log-transformed results. 186 Appendix B - Table B1. List of variables measured in 2010 and 2011 water chemistry samples from reference lakes. Methods and instruments used are given. Analyte pH Method EPA 150.1 ANC Modified Gran analysis technique Cations (Na, Mg, P, K, Ca, Si, Mn, Fe, Zn) Anions (F, Cl, SO4, NO3) Total Nitrogen EPA 200.8 8 Dissolved Organic Carbon Dissolved Inorganic Carbon EPA 300.0 EPA 415.3 (Modified for TN) EPA 415.3 EPA method 160.2 Monomeric Aluminum McAvoy et al. 19928 Instrument Brinkman Metrohm 799 DMS Titrino and the Brinkmann Metrohm 748 autosampler; and the Brinkmann Titrino Brinkman Metrohm 799 DMS Titrino and the Brinkmann Metrohm 748 autosampler; and the Brinkmann Titrino ICP-MS, PE ELAN 6000 Dionex DX-500 ion chromatography system Teledyne Tekmar Apollo 9000 analyzer with TN module Teledyne Tekmar Phoenix analyzer Teledyne Tekmar Apollo 9000 analyzer with TN module Bran & Luebbe Auto Analyzer 3 McAvoy DC, Santore RC, Shosa JD, Driscoll CT. (1992) Comparison between pyrocatechol violet and 8hydroxyquinoline procedures for determining aluminum fractions. Soil Science Society of America Journal 56: 449-455. 187 Appendix C - Table C1. Location information for 31 Adirondack reference lakes. Lake name, code, latitude and longitude, and Adirondack Park Agency (APA) land classification are provided. Lake Name Arbutus Lake Bass Lake Bessie Pond Boottree Pond Cascade Lake Cascade Pond Challis Pond Clamshell Pond Clear Pond Copperas Pond Deer Pond Eagle’s Nest Lake East Pine Pond Fish Pond Grizzle Ocean Gull Pond Island Pond Little Fish Pond Long Pond Lower Sargent Pond Lydia Pond Middle Branch Lake Moss Lake Nellie Pond Rock Pond Round Pond Sampson Lake Streeter Lake Upper Spectacle Pond Wilcox Lake Wolf Lake Code ArbL BassL BessP BootP CascL CascP ChalP ClamP CleaP CoppP DeerP EagNL EPineP FishP GrizO GullP IslaP LitFP LongP LSargP LydiP MidBL MossL NellP RockP RounP SampL StreL UpSP WilcL WolfL Latitude 43º 59’ 14’’ 43º 58’ 25’’ 44º 22’ 51’’ 44º 14’ 24’’ 43º 47’ 21’’ 43º 49’ 46’’ 43º 59’ 00’’ 44º 23’ 00’’ 44º 33’ 30’’ 44º 19’ 45’’ 44º 13’ 47’’ 43º 45’ 50’’ 44º 20’ 23’’ 44º 23’ 42’’ 43º 49’ 22’’ 43º 49’ 51’’ 43º 40’ 50’’ 44º 23’ 42’’ 44º 18’ 59’’ 43º 51’ 34’’ 44º 23’ 45’’ 43º 41’ 52’’ 43º 46’ 52’’ 44º 23’ 02’’ 44º 11’ 26’’ 44º 07’ 26’’ 43º 34’ 43’’ 44º 06’ 39’’ 43º 48’ 46’’ 43º 23’ 59’’ 44º 01’ 01’’ Longitude 74º 14’ 28’’ 73º 40’ 39’’ 74º 23’ 14’’ 74º 39’ 20’’ 74º 48’ 46’’ 74º 26’ 12’’ 73º 40’ 14’’ 74º 21’ 54’’ 74º 46’ 21’’ 73º 53’ 54’’ 74º 39’ 47’’ 74º 43’ 42’’ 74º 24’ 44’’ 74º 22’ 42’’ 73º 35’ 51’’ 73º 42’ 29’’ 73º 39’ 17’’ 74º 23’ 04’’ 73º 55’ 45’’ 74º 34’ 02’’ 74º 23’ 46’’ 75º 06’ 08’’ 74º 51’ 11’’ 74º 23’ 28’’ 74º 18’ 10’’ 73º 43’ 57’’ 74º 34’ 33’’ 75º 04’ 17’’ 73º 42’ 17’’ 74º 09’ 26’’ 74º 13’ 13’’ 188 APA classification Private (Huntington Wildlife Forest) Hammond Pond Wild Forest Area St. Regis Canoe Area Private (Massawepie Scout Camp) Pigeon Lake Wilderness Area Blue Ridge Wilderness Area Hammond Pond Wild Forest Area St. Regis Canoe Area White Hill Wild Forest Sentinel Range Wilderness Private (Massawepie Scout Camp) Moose River Wild Forest River Saranac Lakes Wild Forest Area St. Regis Canoe Area Pharaoh Lake Wilderness Area Pharaoh Lake Wilderness Area Lake George Wild Forest Area St. Regis Canoe Area McKenzie Mountain Wilderness Area Sargent Ponds Wild Forest Area St. Regis Canoe Area Ha-De-Ron-Dah Wilderness Area Moose River Wild Forest Area St. Regis Canoe Area High Peaks Wilderness Area Dix Mountain Wilderness Area West Canada Lakes Wilderness Area Aldrich Pond Wild Forest Area Pharaoh Lake Wilderness Area Wilcox Lake Wild Forest Area Private (Huntington Wildlife Forest) 189 Appendix D - Figure D1. Summary diagram of species assemblages in 31 Adirondack reference lakes. Data are provided as relative abundance (%). Top (present-day) assemblages are show in solid black bars and bottom (pre-1900) assemblages are show in open bars. Samples are organized in descending order based on PCA-1 axis scores of 31 present-day samples (data not shown). Chrysophyte species are split between colonial and unicellular taxa. Chrysophyte-inferred pH (CI-pH) values are provided. Note that 5 reference lakes had scale concentrations too low to count in the pre-1900 sample and so only a present-day sample is provided. Appendix E – Figure E1. Relationship between colonial taxa and DOC in 31 reference lakes. Relative abundance (% RA) of 3 colonial taxa (S. petersenii, S. sphagnicola, S. echinulata) plotted against DOC (mg C L-1) and depth of 1% PAR irradiance (m). Depth 1% PAR based on Bukaveckas and Robbins-Forbes (2000). Gray dotted line denotes a DOC concentration > 5.2 (DOC 5.3 mg C L-1), which corresponds to a 1% PAR depth < 5.2 m (5.1 m). Note decreased abundance of S. petersenii when DOC > 5.2 mg C L-1 whereas S. sphagnicola and S. echinulata occur at > 10% RA when DOC > 5.2 mg C L-1. 190 Appendix F - Table F1. Stocking history, recent netting survey results, land use characteristics, and watershed disturbance histories for South Lake, Queer Lake, Moss Lake, Arbutus Lake, and Wolf Lake. See footnotes for citation information. Stocking history is summarized and recent netting survey results are provided with species listed alphabetically by scientific name. Current land use information (e.g. forest preserve, cottage development, etc.) is provided. Disturbance history includes information on logging events, and watershed impact of large windstorms in 1950 and 1995. An ice storm in 1998 and forest fires in 1903 and 1908 had no known impact on the study sites. Lake South Lake9 Queer Lake10 Moss Lake11 Stocking History Stocked with lake trout, brook trout, and other piscivorous fish from 1993-2006 Date 2002 Stocked with brook trout from 1942 to 1975 and from 19952006; currently managed as a cold-water fishery 1999 Earliest records of stocking of brook trout and lake trout in 1898. Stocked with lake trout and brook trout from 19861998. 2000 Netting Survey Survey Result - Couesius plumbeus - Ictalurus nebulosus - Phoxinus eos - Rhinichthys atratulus - Salmo salar - Salvelinus fontinalis - Semotilus atromaculatus - Catostomus commersonii - Fundulus diaphanus - Ictalurus nebulosus - Lepomis gibbosus - Notemigonus crysoleucas - Salvelinus fontinalis - Salvelinus namaycush - Catostomus commersonii - Ictalurus nebulosus - Lepomis gibbosus - Luxilus cornutus - Micropterus salmoides - Notemigonus crysoleucas - Perca flavescens - Salvelinus fontinalis - Semotilus atromaculatus - Umbra limi Watershed land-use Disturbance History Logging History Windstorms Pre-1890: some denuded - 1950: 50-100% area blowdown in 65.8% of watershed; 25-50% 1890-1920: logged for in 34.2% of the spruce & hardwoods watershed 1916: 100% of the - 1995: No impact watershed either virgin or 2nd growth forest No known history of - 1950: minimal impact logging - 1995: 0-30% change 1916: 100% of the in tree crowns in watershed either virgin or 100% of watershed 2nd growth forest - 56% protected ‘wild forest’ preserve since 1972 - 40% privately owned with little or no development - 4% developed, with cottages, private camps, and a gravel road on the NW lakeshore - 100% of watershed is protected ‘wilderness’ preserve since 1972 - Primitive campsites along lakeshore; no cottages - - A road paved bisects the watershed (originally built 1895-1896) - 1923-1973: 612 acre camp surrounding lake - Watershed protected in a ‘wild forest’ and ‘wilderness’ preserve since 1973 - Primitive campsites along lakeshore; no cottages - Pre-1890: no known logging - 1916: 93.7% of the watershed described as either virgin or 2nd growth forest 9 - - - 1950: 50-100% blowdown in 13.2% of watershed; 25-50% blowdown in 3.5% of watershed - 1995: 0-30% change in tree crowns in 93.2% of watershed; 30-60% change in 6.8% of watershed Stocking history, netting survey data, land-use, windstorms: Roy et al. 2011; logging history: Roy et al. 2011, Sullivan et al. 1999, McMartin 1994 Stocking history, netting survey data, land-use, windstorms: Roy et al. 2011; logging history: Roy et al. 2011, Sullivan et al. 1999, McMartin 1994 11 Stocking history, netting survey data, windstorms: Roy et al. 2011; land-use: Roy et al. 2011, McMartin 2004, Charles et al. 1987; logging history: Roy et al. 2011, McMartin 1994 10 191 Table F1. Continued Arbutus Lake12 Stocking history is unknown; lake treated with rotenone in 1973 and stocked with brook trout (dominant fish taxa were brown bullhead, yellow perch, bass, and ‘suckers’ – species not specified) 2001 - Ictalurus nebulosus Phoxinus eos Rhinichthys atratulus Salvelinus fontinalis Wolf Lake13 - No known history of stocking; repeated surveys until 1980s found no nonnative fish taxa 1997 - Salvelinus fontinalis Ictalurus nebulosus Semotilus atromaculatus Lepomis auritus Catostomus commersonii Other minnows (species unspecified) - Private lodge & camp in late 1800s - Part of a private research property established in 1932; 2000 acres around Arbutus Lake added in 1939 - Watershed largely undeveloped - Cabins & unpaved roads on southern lakeshore - Part of a private research property established in 1932 - Watershed largely undeveloped - Unpaved road (built in late 1930s) within 200-300 m of shoreline; 1 cabin near lakeshore (abandoned) - Pre-1890: no known logging - 1916: 87.9% of watershed classified as either virgin or 2nd growth forest. - Watershed logged in late 1910s - 1960-1961: 162 ha of softwoods logged - 1950: no impact - 1995: no impact - Pre-1890: no known logging - Selective logging between 1957-1991 - 1950: minimal impact - 1995: no impact Literature Cited: Charles DF., Whitehead DR, Engstrom DR, Fry BD, Hites RA, Norton SA, Owen JS, Roll LA, Schindler SC, Smol SP, Uutala AJ, White JR, Wise RJ. (1987) Paleolimnological evidence for recent acidification of Big Moose Lake, Adirondack Mountains, N.Y. (USA). Biogeochemistry 3:267-296. Chen LM, Driscoll CT. (2004) Modeling the response of soil and surface waters in the Adirondack and Catskill regions of New York to changes in atmospheric deposition and historical land disturbance. Atmospheric Environment 38:4099-4109. Jenkins J., Keal A. (2004) The Adirondack atlas: a geographic portrait of the Adirondack Park. Syracuse University Press, Syracuse, New York, 296 p. McMartin B. (1994) The great forest of the Adirondacks. North Country Books, Utica, New York, 254 p. McMartin B. (2004) The privately owned Adirondacks. Lake View Press, Caroga, New York, 304 p. Roy K., Houck N, Hyde P, Cantwell M, Brown J. (2011) The Adirondack long-term monitoring lakes: a compendium of site descriptions, recent chemistry and selected research information. New York State Energy Research and Development Authority, Albany, 298 p Stager JC, Sanger T. (2003) An Adirondack "heritage lake". Adirondack Journal of Environmental Studies 10:6-10. Sullivan TJ, Charles DF, Bernert JA, McMartin B, Vache KB, Zehr J. 1999. Relationship between landscape characteristics, history, and lakewater acidification in the Adirondack Mountains, New York. Water, Air, and Soil Pollution 112:407-427. 12 Netting survey data, windstorms: Roy et al. 2011; stocking history: C.L Demers & R.J. Pearl, personal communication, 2010; land-use: Roy et al. 2011, McMartin 2004; logging history: Roy et al. 2011, Chen & Driscoll 2004, McMartin 1994 13 Stocking history and netting survey data: Stager & Sanger 2003; land-use: Stager & Sanger 2003, McMartin 2004; logging history: Stager & Sanger 2003, McMartin 1994; Windstorms: Stager & Sanger 2003, Jenkins & Keal 2004 192 Appendix G – Table G1. Seasonal Kendall τ results for long-term monthly chemistry data. The Seasonal Kendall τ is a non-parametric test that identifies if a data series shows a monotonic trend overtime (i.e. generally increasing or generally decreasing) (Helsel and Hirsch 2002). The test is recommended for data with serial dependence, such as monthly water chemistry measurements. Tests were performed using the computer program of Helsel et al. (2006). Tests were run on long-term chemistry data from South Lake (1992-2010), Queer Lake (1992-2010), Moss Lake (1982-2010), and Arbutus Lake (1983-2010). Chemistry variables are: sum acid anions (SO42- + NO3-, in μeq L-1), acid neutralizing capacity (ANC, in μeq L-1), pH, inorganic monomeric aluminum (Ali, in μg L-1), dissolved organic carbon (DOC, in mg C L-1), and specific conductance (Sp. Cond., in μS cm-1). The correlation coefficient (τ), test statistic (S), and p-value adjusted for seasonal dependence of chemistry data (p adj) are provided. Chemistry variables that show a significant (p adj < 0.05) monotonic trend overtime (generally increasing, positive τ; generally decreasing, negative τ) are highlighted in bold. Lake South Lake Queer Lake Moss Lake Arbutus Lake τ -0.81 0.53 0.52 -0.47 0.32 -0.73 -0.84 0.63 0.59 -0.45 0.37 -0.80 -0.82 0.22 0.16 -0.01 0.20 -0.56 -0.72 -0.01 -0.01 0.03 0.07 -0.53 Variable SO42- + NO3ANC pH Ali DOC Sp. Cond. SO42- + NO3ANC pH Ali DOC Sp. Cond. SO42- + NO3ANC pH Ali DOC Sp. Cond. SO42- + NO3ANC pH Ali DOC Sp. Cond. S -1574 1035 1006 -908 617 1425 -1602 1193 1117 -854 702 -1522 -3666 989 695 -53 858 -2499 -2999 -33 -34 133 274 -2221 padj <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.002 0.020 0.845 0.003 <0.001 <0.001 0.917 0.932 0.563 0.427 <0.001 Literature Cited: Helsel DR, Hirsch RM (2002) Statistical methods in water resources techniques of water resources investigations, book 4, chapter A3. U.S. Geological Survey, pp. 209-220. Helsel DR, Mueller DK, Slack, JR. (2006) Computer program for the Kendall family of trend tests. U.S. Geological Survey Scientific Investigations Report 2005-5275, U.S. Geological Survey, Reston, Virginia, 4 p. 193 Appendix H – Figure H1. Stratigraphies of chrysophyte relative abundance in acidified lakes. Long-term (preca. 1900-present) stratigraphies of the relative abundance of chrysophyte taxa (%) are provided for South Lake (A) and Queer Lake (B). pH optimum is given in brackets after species name (Cumming et al. 1992a). Chrysophyte-inferred pH (CI-pH) is based on a calibration set of 71 Adirondack lakes (Cumming et al. 1992a). CI-pH calculated by a weighted-averaging function with tolerance down-weighting, inverse deshrinking and a ln (x + 1) species transformation (r2 boot = 0.75, RMSEP = 0.51). The total relative abundance of colonial chrysophyte taxa (% Colonial) is shown. 194 Appendix H – Figure H2. Stratigraphies of chrysophyte relative abundance in reference lakes. Long-term (pre-ca. 1900-present) stratigraphies of chrysophytes relative abundance (in %) are provided for Moss Lake (A), Arbutus Lake (B), and Wolf Lake (C). pH optimum is given in brackets after species name (Cumming et al. 1992a). Chrysophyte-inferred pH (CI-pH) is based on a calibration set of 71 Adirondack lakes (Cumming et al. 1992a). CI-pH calculated by a weighted-averaging function with tolerance down-weighting, inverse deshrinking and a ln (x + 1) species transformation (r 2 boot = 0.75, RMSEP = 0.51). The total relative abundance of colonial chrysophyte taxa (% Colonial) is shown (note change of scale for Arbutus Lake and Wolf Lake). 195 Appendix I – Table I1. List of taxon codes used in nMDS ordination diagram. Taxa are listed alphabetically by name and a code for each taxon is given. The taxon’s form (either unicellular or colonial) is also provided. Taxon Chrysodidymus synuroides Chrysosphaerella spp. Mallomonas acaroides Mallomonas caudata Mallomonas crassisquama Mallomonas duerrschmidtiae Mallomonas elongata Mallomonas hindonii Mallomonas hamata Mallomonas lychenensis Mallomonas pseudocoronata Mallomonas punctifera Mallomonas ‘small’ Synura echinulata Synura petersenii Synura spinosa Synura sphagnicola Synura uvella Synura sp Code C syn Chrysos M acar M caud M crass M duerr M elong M hind M hama M lych M pseudo M punct M small S echin S peter S spin S spag S uvella S sp 14 Form 14 Colonial 15 Colonial Unicellular Unicellular Unicellular Unicellular Unicellular Unicellular Unicellular Unicellular Unicellular Unicellular 16 Unicellular Colonial Colonial Colonial Colonial Colonial Colonial Chrysodidymus synuroides forms 2-celled colonies Likely includes both Chrysosphaerella longispina and Chrysosphaerella brevispina 16 Group of Mallomonas scales too small to distinguish between using light-microscopy 15 196 Appendix J – Figure J1. Principal Components Analysis (PCA) results for acidified and reference lakes. Long-term (pre-ca. 1900-present) PCA results of chrysophyte species assemblages in South Lake (A), Queer Lake (B), Moss Lake (C), Arbutus Lake (D), and Wolf Lake (E) are provided. Species assemblages are divided into 4 time periods: pre-disturbance (pre-1900, □), early disturbance (1900 to pre-1950, Δ), acidification (1950 to 1995, ♦), and post-1995 implementation of the US Acid Rain Program (post-1995, ●). 197