Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Littoral zone primary production in a coastal reservoir ecosystem Beer, Julie Ann 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2004-0364.pdf [ 19.44MB ]
JSON: 831-1.0099770.json
JSON-LD: 831-1.0099770-ld.json
RDF/XML (Pretty): 831-1.0099770-rdf.xml
RDF/JSON: 831-1.0099770-rdf.json
Turtle: 831-1.0099770-turtle.txt
N-Triples: 831-1.0099770-rdf-ntriples.txt
Original Record: 831-1.0099770-source.json
Full Text

Full Text

Littoral Zone Primary Production in a Coastal Reservoir Ecosystem  by  JULIE ANN BEER B . S c , University of British Columbia, 1993  A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES Resource Management and Environmental Studies  We accept this thesis as conforming to the required standard  UNIVERSITY OF BRITISH C O L U M B I A July 2004  © Julie A . Beer, 2004  THE UNIVERSITY OF BRITISH  F A C U L T Y OF G R A D U A T E  COLUMBIA  STUDIES  Library Authorization  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  \J"VA1J(^ Beef Name of Author (please print)  •Trfie of Thesis:  Degree: Department of  Date (ad/mm/yyyy)  Lifrorfll  ^IflSrCir !{{Y\  erf  Z<m-e.  P r o d u c t viUj  ir\ <x CoasYcJL  Year:  ScilvOO^  r^ese^y/ovY  CZJLaSYS:\rew\  J2Q34-  ~&S  The University of British Columbia Vancouver, B C  Canada  page 1 of 1  i  last updated:  31-Aug-04  Abstract  There has been little research examining littoral productivity in freshwater ecosystems. Previous studies have focused on pelagic production, largely because pelagic production was viewed to be the predominant source of carbon in aquatic ecosystems. More recently there has been some research to suggest that the productive capacity of littoral zones may be significant, especially in nutrient-poor ecosystems, where the transfer of carbon through the food chain is driven by microbial activity at the base of the food web. This may be particularly important in reservoir ecosystems, as there are concerns that water level fluctuations resulting from reservoir operations may seriously undermine aquatic function in the littoral zone.  To assess the impacts of fluctuating water levels, a field study was designed to measure primary productivity in the littoral zones of two coastal temperate hydro-electric reservoirs located in British Columbia. Both located on the Stave River, Stave Reservoir exhibited a pronounced annual water level fluctuation, while immediately downstream Hayward Reservoir maintained a relatively constant water level. Measurements of periphyton biomass were conducted to estimate primary productivity in the littoral zones of both reservoirs. Three sampling transects were established in Stave Reservoir and one in Hayward. Periphyton samples were collected by scuba divers from Plexiglas plates along transect gradients in each reservoir to a depth of 20 m. Samples were analysed using standard lab techniques to estimate and characterize production using measurements of Ash Free Dry Mass (accrual) and chlorophyll-a concentrations to quantify biomass, and species composition to quantify abundance, biovolume and dominance of the algal communities.  Primary productivity was estimated by integrating periphyton biomass accrual measurements over depth for each sampling period. Three year average production in Stave was 5.3 gC/m2/year and in Hayward was 12.3 gC/m2/year. Differences in production between the Stave and Hayward reservoirs were evaluated statistically using a two-way analysis of variance (ANOVA) to look for a lake effect and a seasonal effect. The results from the A N O V A found production in Hayward Reservoir to be significantly higher than Stave Reservoir (p = 0.05). A seasonal analysis indicated winter production was significantly lower than all other seasons in ii  both reservoirs.  Measurements of light, temperature, nutrients and water level were used to  interpret spatial and temporal variability in periphyton production.  Littoral and pelagic productivity measurements were used to estimate total aquatic production in Stave and Hayward reservoirs. Analysis results indicate that the littoral contribution to total aquatic production in the fluctuating reservoir environment (Stave) was only 4%, which is lower but still comparable to other oligotrophic temperate lakes. In Hayward (stable water-level environment) the littoral contribution was approximately 50% of overall aquatic production.  Reseach findings comparing Stave and Hayward suggest that littoral production is negatively impacted by reservoir operations but the importance of this is questionable when the littoral contribution is so low. Primary productivity in a large, power generating reservoir like Stave, appears to be pelagically driven. Nutrients dynamics tend towards recycling but are compliated by flow in the old river channel in the lower basin. Although it is a managed system, the contribution of littoral production was approximately 5%, which is comparable to other oligotrophic B C lakes. Littoral primary production in Hayward was high compared to oligotrophic temperate lakes. This is likely a function of the fast flushing, riverine character of this reservoir. This study did not examine other factors such as light, temperature and nutrient regimes in sufficient detail to directly compare these factors between the two systems. A clear indication of the effect of water level fluctuations on littoral zone productivity is complicated in this case by the fact that the two reservoirs studied exhibited significant differences in character which may also have accounted for differences in littoral zone productivity.  iii  Table of Contents  Abstract  u  Table of Contents  iv  Acknowledgements _ _ _  x  CHAPTER 1  1  1.0  BACKGROUND  -  1  1.1  R E S E R V O I R S AS A Q U A T I C R E S O U R C E S : P A S T A N D P R E S E N T V I E W S  1  1.2  T H E O N T O G E N Y OF IMPOUNDMENT  3  1.3  IMPACTS OF RESERVOIR FUNCTION ON T H E L I T T O R A L COMMUNITY  5  1.4  OBJECTIVES  9  CHAPTER 2  13  2.0  RESEARCH DESIGN  13  2.1  STUDY A R E A  2.1.1  Stave Reservoir  2.1.2  Hayward Reservoir  2.2  13  13 ;  14  METHODS  17  2.2.1  Study Design  17  2.2.2  Periphyton Sampling  2.2.3  Environmental Data  . 22  2.2.4  Laboratory Analysis  24  2.2.5  Pelagic Primary Production Sampling and Analysis  26  :  :  CHAPTER 3 3.0  RESULTS AND DISCUSSION  3.1  PHYSICO-CHEMICAL FEATURES OF T H E STAVE/HAYWARD ECOSYSTEM  21  .  28 28 28  iv  3.1.1  Light  28  3.1.2  Temperature  31  3.1.3  Nutrients  32  3.1.4  Water Level Fluctuation and Residence Times  36  3.1.5  Discussion of Environmental Data  38  3.2  3.2.1  BIOLOGICAL FEATURES OF T H E STAVE/HAYWARD LITTORAL ZONES  Species Composition Results and Discussion  45  45  3.2.3  Estimates of Littoral Periphyton Production: Accrual and Chl-a  52  3.2.4  Stave/Hayward: A Statistical Comparison of Littoral Production  58  3.2.5  Discussion of Littoral Production In Stave and Hayward reservoirs  62  3.4  STAVE/HAYWARD COMPARISON OF LITTORAL TO PELAGIC PRODUCTION  70  3.4.1  Pelagic Primary Production  70  3.4.2  Contribution of littoral production to Stave/Hayward Ecosystems  71  3.5  DISCUSSION O F T H E S T A V E A N D H A Y W A R D R E S E R V O I R E C O S Y S T E M  72  CHAPTER 4  84  4.0  CONCLUSIONS  84  4.1  MITIGATION AND RECOMMENDATIONS  87  REFERENCES  90  APPENDIX 1  97  Data Summary Table  . 97  APPENDIX 2__  101  Summary Table of Annual Littoral Production  101  Summary Table of Littoral Chlorophyll Production  102  v  APPENDIX 3  103  Hayward Annual Sample Period Irradiance over Depth  103  Stave North Sample Annual Sample Period Irradiance over Depth  104  Stave South Sample Annual Sample Period Irradiance over Depth  105  Stave West Sample Annual Sample Period Irradiance over Depth  106  APPENDIX 4  107  Hayward 2003 Temperature Profiles over Depth  107  Stave 2003 Temperature Profiles over Depth  109  Depth (m)  109  APPENDIX 5  111  Stave Reservoir: Summary Water Level Data  111  Stave/Hayward Plate Elevation Data  111  Hayward Reservoir: Summary Water Level Data  112  Hayward Reservoir: Summary Water Elevation June 2001 Drawdown  112  vi  List of Figures  Figure 1. Nutrient levels and productivity following impoundment in a north temperate reservoir (after Stockner et al., 2000). Figure 2.  5  Schematic diagram showing shifting environmental variables associated with water level fluctuations (modified from Bruce and Stockner, 2000).  7  Figure 3. Schematic cartoon of the linkages between the classic food chain and the microbial loop; the bottom periphyton box (red) showing the primary focus of this thesis Figure 4.  10  Stave/Hayward Hydroelectric Complex including Stave and Hayward reservoirs (modified from Kerr. 1999).  15  Figure 5 & 6. Stave Reservoir (main basin) at full pool (left) and during drawdown (right).  16  Figure 7 & 8. Hayward Reservoir at full pool (left) and during drawdown Oighf).  16  Figure 9.  20  Sampling block in substrata typical of the upper elevations in Stave  Figure 10. Transect layout - shows difference in elevation between blocks and plate  20  Figure 11. Sample plate with a 10x10 cm area scraped  20  Figure 12. Sampling of plate and field storage of samples  20  Figure 13. G V R D surface solar radiation data from Pt. Moodv, B C .  29  Figure 14. Average sampling period irradiance estimated from Pt. Moodv.  29  Figure 15. Stave/Havward percent light extinction.  30  Figure 16. Stave/Havward compensation depths.  30  Figure 17. Stave/Havward secchi depths.  30  Figure 18. Stave/Havward average 1996-2001 temperature regime.  31  Figure 19. Stave TP and TDP Concentrations.  33  Figure 20. Hayward TP and TDP Concentrations.  33  Figure 21. Stave/Havward integrated pelagic chlorophvll-a concentrations.  34  Figure 22. Stave/Havward integrated pelagic nitrate concentrations.  35  Figure 23. Stave/Havward NCh-N: TDP ratio.  35  Figure 24. Stave/Havward water levels (2000-2002).  37  Figure 25. Stave Reservoir epilirnnion volume replacements 1952-2002  37  Figure 26. Ten dominant species in Havward Reservoir bv abundance (cells/cm ). 2  47  Figure 27: Ten most dominant species by abundance (cells/cm ) for Stave Reservoir transects. 48 2  vii  Figure 28. Ten most dominant species by biovolume (mm ) for Stave (all transects) and 3  Hayward reservoirs.  49  Figure 29. Relationship of species composition samples to year, transect, season and depth using nonmetric multidimensional scaling. Stress of solution was 7.83 and instability was 0.0007. Only depth indicates a compositional grouping of species.  51  Figure 30. Annual Littoral Accrual over depth a) 2000, b) 2001, c) 2002.  53  Figure 31. Stave/Hayward Seasonal Production (2000-2002 average).  54  Figure 32. Average Littoral Production 2000-2001.  55  Figure 33. Annual Chlorophyll Concentration over depth a) 2000, b) 2001, c) 2002.  56  Figure 34. Stave/Hayward Littoral Chlorophyll Production.  57  Figure 35. Scatterplot of Chlorophyll and Accrual data for Stave and Hayward reservoirs.  58  Figure 36. Interaction plot showing factors of lake and season for Stave and Hayward.  61  Figure 37. Comparison of Water Level and Accrual Measurements Over Time.  63  Figure 38. Sample period accrual in Hayward Reservoir 2000, 2001, 2002.  64  Figure 39. Changes in total periphyton biomass and diatom density.  [  69  Figure 40. Estimates of daily C production.  70  Figure 41. 2003 N 0 : T D P Ratio.  75  Figure 42. Schematic diagram of the microbial loop for Stave and Hayward reservoirs.  79  3  Figure 43. Schematic diagram of periphytic matrix growing on substrata similar to those in Stave and Hayward (modified from Allan 1995).  ;  :  81  viii  List of Tables  Table 1.  Summary of important physical attributes of Stave and Hayward reservoirs.  14  Table 2.  Table of sampling periods, number of days per sample period and Season.  21  Table 3.  Abundance and biovolume estimates for species encountered at plate 4 (8 m below full pool) in Stave and Hayward  46  :  Table 4.  Stave/Hayward A N O V A Results  Table 5.  Results of Tukey's Test Comparing Seasonal Differences.  62  Table 6.  Stave/Hayward Average Annual Production  67  Table 7.  Summary of Stave and Hayward pelagic C production, chlorophyll, and production efficiency (AN).  61  :  ;  71  Table 8.  Littoral and pelagic production estimates for Stave and Hayward reservoirs  71  Table 9.  Summary of environmental data with key differences highlighted.  74  ix  Acknowledgements  This thesis would not have been possible without the foresight and funding of B C Hydro to embark upon collecting baseline data on reservoir ecosystems as part of their Water Use Planning Process. I am deeply indebted to many people who helped me along my journey to complete this thesis. This acknowledgement is unable to fully express the gratitude extended to each person who aided this process. First and foremost, I would like to thank the members of my academic committee: M y deepest gratitude goes to my supervisor, Dr. John Stockner for his unending patience, support and willingness to impart knowledge. I would also like to thank Chris Perrin (Limnotek) for his assistance with statistical analyses. From U B C , I would like to thank Doctors Les Lavkulich and Ken Hall who both supported this endeavour from start to finish. Finally I would like to thank Dr. Will Carr for his role as a mentor, his belief in my ability, his practical advice and his friendship. I would also be remiss not to include my other mentor, Anne Moody, for invoking a curiosity in the biological sciences and for making me appreciate the nature of reservoir ecosystems. M y most sincere thanks also go to those who helped with practical aspects of this project: To Dave Hunter (BC Hydro) for his hard work and professionalism in organizing field operations and logistics, and for coordinating all aspects of interaction with B C Hydro. As well, I would like to thank Foreshore Technologies Ltd. for their diving expertise, their creative problem solving and their willingness to work hard. To each member of this crew I am also indebted for your friendship, which made me look forward to days in the field. To Danusia Dolecki for her work identifying periphyton. To Shannon Harris, Sandra Wilson, and Dr. Rob Kozak for their help with various aspects of this project. I would also like to acknowledge my friends and family who supported me so whole-heartedly through this process. Thanks go to Nick Page, Lea Elliot, and Alena Fikart who paved the path ahead and made me the R M E S soul survivor, with a note of additional thanks to Nick for help with statistics and for editing. I would also like to thank Kate Schendel for always being willing to stop for a coffee or a treat, for making me swim so I could indulge guilt-free and for listening. To Liz Leboe for her editorial skills while driving. To the geo-ladies - you are my sanity. Special tribute goes to my family, especially to my parents Thomas and Gwyneth Beer, because you are always there when I (or any of us) need you and I love you dearly for teaching me what unconditional love is. Finally this thesis is dedicate to my husband, Kenneth Reid, for his patience, guidance and encouragement when things got tough and I couldn't see where to take my next step. Thank you for believing in me, even when I could not, for instilling confidence, for supporting me to do the other things I needed to do and for making me realize that I should worry less and cherish this time and education more. I love you.  x  CHAPTER 1  1.0  BACKGROUND  1.1  Reservoirs as Aquatic Resources: past and present views  Reservoirs have been built for thousands of years, serving a variety of purposes, including navigation, floodwater management, hydroelectric power generation, drinking water supply, industrial and agricultural uses (Baxter, 1977; Gunnison, 1985). Historically dam construction has been regarded from the vantage points of economic development, engineering ingenuity, and public service, with limited interest in ecological processes or impacts. Reservoirs resulting from the construction of dams have typically been disregarded as aquatic ecosystems or viewed to be ecologically synonymous with unproductive lakes (Hutchinson, 1957). This past view of reservoir ecosystems was fostered by the concept that they are semi-dysfunctional lakes, which typically exhibit lower overall productivity than natural lakes (Stockner et al., 2000). As a result, most freshwater biological research has focused on natural, undisturbed lake systems and there is a paucity of studies that address reservoir ecosystems.  Recently, and in response to the recognition of the ecological value of reservoirs, environmental awareness and multiple-use demands (including historic fisheries values, First Nations' archaeological sites, recreation and aesthetics, and aquatic health and function) have highlighted a need for more rigorous scientific studies of reservoirs. This initiative recognizes, in part, that furthering the understanding of reservoir function potentially provides valuable ecological information that may have practical implications for sound management  of freshwater  ecosystems. Reservoirs provide a unique opportunity to study littoral zone response to disturbance associated with reservoir creation and subsequent water management operating regimes. Gunnison (1985) points out that although similar to lakes, the ecological function of reservoirs is more complex, largely as a result of the water management regime under which they operate, creating unique research opportunities. Conducting baseline studies of aquatic function, specifically littoral response, in reservoir ecosystems provides a novel addition to freshwater research.  Chapter 1  1  Past views of reservoir ecosystems as dysfunctional may have been influenced by the fact that there is a lack of information regarding the biological, and especially the microbial communities within the littoral zones of these aquatic ecosystems. Microbial interactions form the base of the food chain for fish and other higher organisms, and are now being recognized for the integral role they play in nutrient transfers and primary productivity. Trophic transfers of carbon at the microbial level ensure the sustenance of zooplankton and macrozooplankton that in turn provide food for fish.  Primary productivity is commonly used in freshwater biology as an indicator of ecosystem function. Primary production, as defined by Allan (1995), is the formation of new energy by photosynthesis. Wetzel (1983) defined productivity as "the rate of formation of organic material averaged over some defined period of time, such as a day or a year." Gross primary production, as defined above, does not account for energy from primary production that is utilized by organisms, such as periphyton, bacteria and plants for their own metabolism. Net primary production, accounts for metabolic loss and is a measure of the energy remaining after expenditures by autotrophic organisms to sustain themselves. Measurements of both gross and net primary production are used to predict or indicate the level at which the system functions. Estimates of primary production typically involve measurements of biomass accrual, chlorophyll concentrations, rates of photosynthesis, or carbon uptake, and may also account for losses from death, respiration, or herbivory. Generally, low production values are associated with ecosystems that are nutrient poor (oligotrophic) or have some lack of functionality, while high productivity is associated with an adequate nutrient supply and frequently with an excess of nutrients (eutrophication). The most productive and functional ecosystems  are associated  with  mesotrophic environments where a balance between nutrient supply and uptake by organisms (eg. reduction) is achieved.  In the past, the focus of most limnological research has been on pelagic primary production, largely ignoring littoral function (Stockner, 1988). Pelagic and littoral food webs were typically approached as discrete interactions, with parallel processes for bacteria, primary producers and consumers and few studies attempting to integrate these interactions (Vadeboncoeur et al., 2002). Pelagic dynamics were viewed to be the driving force behind lentic food webs and pelagic phytoplankton studies were easier to conduct than studies of attached, microbial organisms and were therefore conducted more frequently. As a result, littoral habitats were viewed as a source Chapter 1  2  or sink of pelagic nutrients, energetic pathways were ill-defined, and littoral contribution to ecosystem production was often overlooked (Vadeboncoeur et al., 2002).  More recently, studies have suggested that littoral periphyton (attached algae) have a key role to play in primary production, nutrient recycling and food web interactions (Wetzel, 1983; Vadeboncoeur et al., 2002); and that littoral processes may be an even more significant contributor in aquatic ecosystems that are nutrient deficient, such as reservoirs (Stockner 2002, Pers. Com.). These gaps in our understanding of littoral function leave unknown the potential contribution littoral productivity has in the overall productivity of lakes and reservoirs, and indicate that there is a fundamental need for studies that attempt to quantify and integrate the role of littoral productivity in aquatic ecosystems. It is hypothesized that littoral productivity may be of particular benefit in ecosystems such as reservoirs, where overall aquatic productivity is typically low but at this time there are no studies that examine the value of littoral production in reservoirs.  1.2  The Ontogeny of Impoundment  To explore littoral function in reservoir ecosystems, it is important to understand the dynamic environment under which reservoirs operate. Coastal British Columbia hydroelectric reservoirs, which include riverine, lacustrine and terrestrial habitats, are created when a dam is built on a river or lake to impound water for electricity production. The resulting water body is lake-like in appearance, but typically retains some of its riverine flow characteristics. The reservoir receives runoff from the surrounding mountains, typically during two periods of recharge - one occurring during the fall rains and the other during spring snowmelt. Prior to each period of recharge, the reservoir is partially drained (or drawn down), in anticipation of incoming runoff.  Hydroelectric facilities run vast quantities of reservoir-stored water through turbines to produce the power to light and heat our homes and to facilitate our electrically dependent lifestyles. The amount of time it takes to replace the volume of water in a lake in a given year, or residence time can range from a period of days to years in reservoir systems. Residence time is often a function of the type of facility and hydroelectric demand (Baxter, 1977). In a run-of the-river facility, water is stored for short time periods usually to manage flooding downstream and is typically not used for power generation. In this type of reservoir, water moves through the system quickly, Chapter 1  3  creating a riverine-like habitat for organisms. In contrast, large storage reservoirs have traditionally been managed primarily for power production, but have more recently been recognized for their ecological importance. Characterized as lake-river hybrids (Benson 1982; Soballe and Kimmel 1984), these facilities typically have much longer residence times that result in the potential for more variable flow regimes, sediment and nutrient dynamics, and water level fluctuation. As a result, the variability associated with large storage facilities is generally thought to negatively impact the ability of reservoirs to support organisms (Baxter, 1977, 1980).  The process of creating large-scale hydroelectric reservoirs and the subsequent operating regime of drawdown and recharge result in the flooding of surrounding near-shore vegetation, which includes highly productive areas of wetland, riparian and forest habitats. The consequences and impacts associated with flooding of the existing landscape are well documented (Baxter, 1977; Gunnison, 1985; Thornton et al., 1990). When reservoirs are created they begin a cycle of upsurge and depression of nutrient production (Gunnison, 1985; Stockner et al., 2000). Typically reservoirs experience high productivity shortly after inundation occurring in response to nutrient loading from decaying vegetation, microorganisms, and the release of soil nutrients. The process results in a peak in nutrient release that occurs in the first decade following inundation (Figure 1). During this initial nutrient upsurge, the reservoir effectively becomes a storage facility for nutrients, notably phosphorus and nitrogen; the primary limiting nutrients in nutrient poor ecosystems (Baxter, 1977; Thornton et al., 1990; Marotz et al., 1994; Stockner et al. 2000). Following upsurge, the reservoir goes through a transitional period during which the system reequilibrates as the sources of nutrients are depleted.  Lower nutrient levels, lower biological productivity, and decreased carbon production characterize the subsequent decline in productivity (Stockner et al., 2000). These traits are likely attributable to increased sediment retention, water and nutrient discharge, and large, unnatural water level fluctuation. At the completion of the upsurge and depression cycle, the reservoir is, at best, a pseudo-stable environment for aquatic life.  The littoral community is most severely  impacted by this process of stabilization and by the ensuing water management regime that the reservoir operates under once created.  Chapter 1  4  I  Total Phosphorus  —  ,  y  i !  { 0  _  0  _  _  _  _  _  _  _  _  i  J Impo undment  Phase Year Production (gC/m2/year) Total P Trophic State Fish Prod. Transparency  Pre 50-70  'Boom' 4-5 100-120  Transition 5- 8 80-100  'Bust' 8-15 40-60  PostPerpetuity 30-50  2-4 Oligotrophic  10-14 Mesotrophic  6- 8 Oligotrophic  Low High  Moderate Low  Low Moderate  2-4 Ultraoligotrophic Very low High  1-5 Ultraoligotrophic Very Low Very high  Figure 1. Nutrient levels and productivity following impoundment in a north temperate reservoir (after Stockner et al., 2000).  1.3  Impacts of Reservoir Function on the Littoral Community  Water level fluctuation is the fundamental difference between natural lake systems and reservoir ecosystems. In large hydroelectric reservoirs, water level fluctuations are typically much more pronounced and frequently longer in duration than what is common in natural lakes (Gasith and Gafny, 1990). Flooding and drawdown in reservoirs can result in water level fluctuations that range from a few meters to twenty metres or more, therefore the impacts on littoral productivity can be severe. Water levels are usually altered to produce power, and can result in relatively fast flushing (i.e short residence time) of the water volume in a reservoir, which also has the potential to impact growth.  In natural ecosystems, littoral organisms are commonly adapted to tolerate moderate changes in water level; consequently wetlands, riparian areas and near-shore forests associated with littoral ecosystems are commonly thought of as rich, ecologically diverse communities that are critical components of fish and wildlife habitats (Carr and Moody, 2000). In a reservoir ecosystem, littoral communities are frequently affected by exaggerated water level fluctuation; the impacts of these fluctuations are directly related to their amplitude, frequency, and duration (Thornton et al., 1990). The amplitude of the fluctuation determines the area that is affected, while the Chapter 1  5  duration and frequency of occurrence determines the response time available to littoral organisms and biota.  The magnitude and extent of impacts associated with water level fluctuation in reservoirs are not well documented, but are thought to be important in defining why reservoirs are typically less productive than natural lake systems. Godshalk and Barko (1985) report that the impact of water level fluctuation may be beneficial or detrimental depending on the duration and the amplitude of the event. Generally it is established that brief periods of water level drawdown increases microhabitat complexity and species diversity (Gasith and Gafny, 1990). However, extreme, frequent fluctuations tend to stress aquatic organisms and plants, and in most cases result in a reduction in growth and productivity.  Superficially reservoirs appear similar to natural lakes, but draw down and flooding associated with power generation result in complex physical and biological dynamics that are associated with reservoir function. As water levels fluctuate, the littoral zone community undergoes a vertical shift, and the environmental variables that control growth undergo parallel changes that force organisms in the littoral community to adapt to survive (Figure 2). In this study, four factors that influence growth were examined; Light, temperature and nutrients were examined as the primary variables influencing growth, and water level and residence times were evaluated to help explain the impact water level operations has on each of these environmental variables.  Light is one of the key variables impacting periphytic production because of its fundamental link to photosynthesis and the photo-autotrophic existence of a large component of the periphytic community. The amount of photosynthesis that occurs in a given habitat is directly related to the amount of insolation received at the surface and attenuation through the water column. Variability in the amount and intensity of light potentially accounts for much of the variation seen in organic carbon production, which ultimately impacts growth and community structure of algae (Stevenson et al., 1996).  Temperature is another environmental factor that influences the growth of periphyton. Similar to light, temperature varies temporally following seasonal climatic trends. The energy available to warm the water is ultimately linked to the amount of solar radiation received, as well as, the ability of the lake to absorb heat. This variability is largely a function of factors such as latitude, Chapter 1  6  elevation, aspect, and morphology, all of which determine the general temperature regime of a given location. .  Figure 2. Schematic diagram showing shifting environmental variables associated with water level fluctuations (modified from Bruce and Stockner, 2000). In aquatic ecosystems, nutrients are important because algal growth is largely controlled by availability of specific nutrients. Growth is inhibited by shortages of one or more nutrients, therefore understanding nutrient dynamics of aquatic ecosystems is essential for interpreting growth patterns of periphytic communities. Nitrogen (N), phosphorus (P) and carbon (C) are the primary nutrients because they are the most likely to limit growth.  Surface solar radiation, light penetration through the water column and temperature are closely linked. As well, nutrient cycling and decomposition are largely driven by seasonal light and temperature cycles. When water levels are altered in a reservoir environment, light attenuation through the water column shifts vertically up and down coinciding with water levels, forcing aquatic organisms to adapt to changing light, temperature and nutrient regimes. Temperature changes are greatest at the water surface and are less intense with depth. Nutrient dynamics in aquatic ecosystems are more complex and the potential impacts associated with large changes in water surface elevation and reservoir function are not fully understood (Baxter, 1977, Borchardt, 1996), but it is believed that on-going water fluctuations result in a biological community that is  Chapter 1  1  in a continuous state of transition and that there is a likely to be a measurable impact on production.  While it is relatively well documented that fluctuating water levels, especially drawdown, negatively impact the overall productivity of reservoirs (Fraley et al., 1987; Thornton et al., 1990; Marotz et al., 1994; Northcote and Atagi, 1997), the impacts of these events on the biological productivity of the littoral zone are less well documented. In particular, there has been little research to examine and quantify primary productivity associated with reservoir littoral zones. Lanza and Silvey (1985) assert that studies that focus on microbial assemblages in reservoirs are generally lacking and those that do exist are, at best, qualitative. The majority of available literature focuses on macrophytic plants and phytoplankton, and much of the existing regional literature is more broadly focused on larger questions such as the reduction of fish stocks, primarily salmon (Slaney et al., 1996). This literature does little to address fundamental problems that are centred on microbial functioning of the littoral community.  During periods of flooding, littoral plant communities that were historically terrestrial, become predominantly aquatic in terms of how they function (Nilsson and Keddy, 1988; McLachlan, 1969). Flooding produces a condition for aquatic microorganisms, such as algae, bacteria and fungi to establish. These organisms form a slimy matrix and live in association with plants, rocks and sediments in the near-shore littoral zones of lakes and other water bodies. They are fundamental components of aquatic ecosystems, acting as the base of the food chain. Under conditions of periodic flooding, reservoir environments create an unstable habitat where neither the terrestrial plants nor the microbial community that is associated with them, reach their full growth potential. The terrestrial plants that serve as substrata for growth in the reservoir littoral zones are generally not well adapted to living under conditions of flooding, while the associated micro-organisms have difficulty establishing and adapting to alternating periods of flooding and desiccation associated with drawdown.  Also associated with reservoir function, and additionally detrimental to littoral function, are frequent and often lengthy periods of drawdown. Thornton et al., (1990) and Marotz et al. (1994) indicate that one of the primary causes of reduced productivity in reservoirs is desiccation during periods of exposure. Drawdown, depending on the duration and amplitude, has the potential to transform vast areas of created aquatic habitat back into a terrestrial environment for periods of Chapter 1  8  days to several months, disrupting the acquired aquatic function. Most severely affected by drawdown are the microbial communities associated with littoral habitat such as algae, bacteria, fungi, and picoplankton. These periphytic organisms are exposed to wind, light and other physical elements normally not experienced, and the ability of these organisms to tolerate extended periods of exposure is short-lived. Under these conditions periphyton desiccate and die within a period of hours to days, which is also believed to have a significant impact on productivity of the littoral community. The transition between aquatic and terrestrial habitats (flooding and drawdown) in the littoral zone is accompanied by a variety of disturbances that are thought to cause intensive physical, chemical, and biological change that may result in productivity losses that are difficult to fully recover (Fraley et al., 1987; Thornton et al., 1990; Marotz et al., 1994; Northcote and Atagi, 1997).  Also unknown is the relationship between littoral productivity and pelagic productivity of coastal temperate reservoirs. In natural lakes it is generally thought that littoral biomass production is low relative to the estimates of overall lake productivity. In fact, Shortreed et al. (1983) report that periphyton biomass in the littoral zones of 21 British Columbia lakes accounts for less than 1% of the average total algal biomass. What remains in question is whether or not this relationship holds true for an oligotrophic system where productivity is already negatively impacted by various disturbances. Is the contribution of carbon from the littoral zone community different in a nutrient-poor, functionally disrupted ecosystem, like a reservoir, than in a nutrient balanced, functional mesoeutrophic ecosystem or a naturally occurring oligotophic system? Is it possible to enhance productivity by expanding the productive capacity of the littoral zone interface community by reducing water level fluctuation?  1.4  Objectives  This thesis measured and compared periphyton growth in the littoral zone of two coastal temperate hydroelectric reservoirs that share that same water mass, but which operate under two contrasting water management regimes (stable and fluctuating). The study focused on periphyton growth as the link between the classic food chain and the microbial loop because it has been suggested that microbial activity is an important driver of production, and that it may be of particular importance in nutrient deficient ecosystems. Figure 3 below is a cartoon of the  Chapter 1  9  Chapter 1  10  possible linkage between the classic food chain and the microbial loop. This thesis focused on measuring and quantifying periphytic primary production in the reservoirs littoral zones (the lowest (red) box of the classic food chain). This study was a component of a British Columbia Hydro (BCH) initiative to examine at water management on the Stave River reservoir system from the perspective of multiple use interests. B C H ' s Stave River Water Use Planning (WUP) process identified a need to conduct an extensive and specific littoral zone study to address concerns regarding the impact of water level fluctuation on littoral zone function and in turn the impacts to fish health of the Stave Reservoir ecosystem. Specifically the W U P questioned whether changes to the littoral zone have significantly impacted the overall biological (carbon) production of the ecosystem. While it was clear from existing literature and observational site visits that fluctuating water in Stave reservoir is likely to have had an impact, the extent and magnitude of impacts remained uncertain. Also unclear was the relationship and magnitude of littoral productivity to pelagic zone productivity.  Following the Stave W U P recommendations, the Stave River Littoral Zone Monitor (SLZM) was completed over a three-year period to address the concerns and questions raised as part of the W U P process. The S L Z M consisted of two parts, a pelagic and a littoral component, which were carried out concurrently. The study design for this project was based on the S L Z M document (Bruce and Stockner, 2000). This thesis focused on the littoral component of the S L Z M and used components of the pelagic study to carry out a preliminary assessment of littoral contribution to reservoir productivity.  Stave and Hayward reservoirs are biogeochemically similar ecosystems; located downstream of one another on the same river system and sharing the same climatic regime, geology and morphology. The study aimed to compare and quantify the magnitude and impact of water level fluctuation on littoral carbon production. In addition, the study also proposed to estimate and compare the ratio of littoral productivity to pelagic productivity to determine the relative contribution of littoral carbon in reservoir food webs. Specifically this thesis addresses the following objectives: 1. To estimate periphyton production in the littoral zone of Stave (fluctuating) and Hayward (stable) reservoirs.  Chapter 1  11  2. To compare differences in biomass production (accrual) between Stave and Hayward reservoirs. 3. To estimate and compare the ratio of littoral productivity to pelagic productivity (carbon) on an areal basis for Stave and Hayward reservoirs. To address objective 1, periphyton accrual in the littoral zone of each reservoir was measured as an indicator of primary production in the littoral zone. Periphyton was measured from sampling transects established in Stave and Hayward reservoirs, then analyzed in the lab to characterize the littoral community and to estimate littoral productions based on the following parameters: ash-free dry weight (AFDW), chlorophyll-a concentrations (chl-a), and species composition.  Objective 2 was addressed using estimates of A F D W integrated over depth to estimate primary production of the littoral zone. The littoral zone is typically defined as the near-shore area where plants and organisms grow because they have sufficient light and nutrients to support growth. In this study, the littoral zone was defined to be 20 m below full pool. This depth was chosen to ensure that even under conditions of draw down sampling was completed to the 1% light level (compensation depth). Littoral zone primary production estimates of Stave and Hayward reservoirs were compared descriptively and statistically using an analysis of variance as a way to compare between a fluctuating and a stable reservoir environment.  The final objective, to estimate the contribution of littoral production within each reservoir, was addressed using data collected as part of the pelagic part of the S L Z M . Pelagic (overall reservoir) production was estimated using C uptake in the deep water of Stave and Hayward reservoir. 1 4  Annual littoral and pelagic production measurements were calculated using estimates of reservoir area provided by B C Hydro.  For the purpose of this study and because of the reservoirs' similarities, the measurements of periphyton accrual collected from Hayward Reservoir were used to represent productivity of a non-fluctuating, pseudo-lake-like environment. While it is acknowledged that there are some factors that arguably make Stave and Hayward reservoirs less than directly comparable, for this thesis it has been assumed that the physical proximity of these two systems, the similarity of water chemistry and the lack of water level fluctuation in Hayward form a reasonable basis for comparison to Stave.  Chapter 1  12  CHAPTER 2  2.0  R E S E A R C H DESIGN  2.1  Study Area  Stave and Hayward reservoirs are located approximately 75 km east of Vancouver, in the Coast Mountains of British Columbia. The surrounding landscape is dominated by mountains and valleys in the Coastal Western Hemlock biogeoclimatic zone. Surrounding peaks range from 1500 to 2600 m (a.s.l.). Mild temperatures and abundant rainfall (>2000 mm) characterize the local climate. Surrounding bedrock geology is predominantly quartz diorite with exposed cliffs that are up to 15 m in height. Overburden materials consist of glacial till overlain by sand gravel and boulders. Both reservoirs are situated on Stave River, which flows south from its headwaters in Garibaldi and Golden Ears Provincial Parks to its mouth at the Fraser River, near Mission, B C (Figure 4). Important physical attributes of each reservoir are summarized in Table 1.  2.1.1  Stave Reservoir  Stave Reservoir, created in the 1920s with the construction of Stave Falls dam, flooded nearly 2000 ha of adjacent lowland and raised the original lake level by 12 m to a maximum depth at full pool of 101 m above sea level (a.s.l.) (Jackson, 1994). The reservoir is 25 km long and covers a surface area of nearly 60 km . Approximately half of the upper basin of Stave Reservoir 2  was originally Stave Lake, while the lower basin was formed when the existing river and surrounding riparian habitat was flooded. As a result Stave Reservoir is characterized by both lake and riverine characteristics of sedimentation, nutrient dynamics and water retention.  Operating as a hydroelectric storage facility, Stave Reservoir typically operates on a dual cycle of drawdown (i.e. partially drained twice per year). Traditionally this has meant water levels in Stave Reservoir are maintained near full pool (82.1 m a.s.l.) during the summer to accommodate recreational use and during the winter when energy demands are the highest (Figure 5 & 6). In the spring and fall, reservoir levels are drawn down by as much as 9 m (73.0 m a.s.l.) to prepare for inflows from fall and winter rainfall and spring snowmelt. More recently, the Stave reservoir Chapter 2  13  operating regime has been modified to follow guidelines set by the Stave River WUP, which suggests that water levels be maintained slightly lower than full pool and with less extreme annual variation associated with drawdown (Kerr, 1999).  2.1.2  Hayward Reservoir  Hayward reservoir, situated approximately 5.5 km south of Stave Falls dam lies in a relatively small watershed and is only 5 km long. Hayward reservoir, built in the 1930s with the completion of Ruskin dam, is operated as a run-of-river facility whose main purpose is to control flow down stream. Consequently, little water is impounded by this system and water levels typically remain within a meter of mean surface water elevation. The normal operating range for Hayward reservoir is between 41 m and 43 m a.s.l (Figure 7 & 8) (Jackson, 1994).  Table 1.  Summary of important physical attributes of Stave and Hayward reservoirs.  Variable Surface Area (km ) Volume (m xlO ) Mean Depth (m) Length (km) Drainage Basin (km ) Max/Min water elevation (m a.s.l.) Rainfall (cm) Average Discharge (nrVs) Epilirnnion Flush (years) 2  j  b  2  Chapter 2  Stave Reservoir 58 2,040 35 25 1,170 82.1-73.0 230 130 0.22  Hayward Reservoir 2.9 42 14.5 5.6 953 42.9-33.0 230 145 0.005  14  i, 0 9 ,  Mount  !  "B RI  ISM  I  O L U M B  C  A  Golde Ears «t *v * \ / Provinc al Park Dktt Lal{ ^Horth Vancouver \ \Lake Crown Mountain  L o Q K L  .® Vancouver  k e  -Burnaby ^  te Richmond^ *N W „ ^ 0  frT^-. MiSCE S u r r e y * " ^ ^ngley"  White Rock  *  MT J U D G E HOWRY PROVINCIAL PARK  /  ?  Matsqui  ^  1Q200D M c r o s o f t G o r p . a n d / o r its, s u p p l i e r s \ M  rights:res.erved.  GOLDEN EARS PROVINCIAL PARK  Mission (2 km)  Vancouver (75 km)  Legend N  Figure 4.  Transect locations on Stave and Hayward reservoirs  Stave/Hayward Hydroelectric Complex including Stave and Hayward reservoirs (modified from Kerr, 1999).  Chapter 2  15  Figure 5 and 6. Stave Reservoir (main basin) at full pool (left) and during drawdown (right).  2.2  Methods  2.2.1  Study Design  Periphyton growth was measured seasonally along transect gradients in Stave Reservoir (highly fluctuating) and Hayward Reservoir (stable) for a three year period. Accumulated periphyton biomass was estimated, using the following variables: •  Ash-Free Dry Weight (AFDW) (accrual of biomass);  •  Chlorophyll-a concentrations (chl-a) (algal biomass accrual); and,  •  Species composition  Periphyton growth was measured by repeated sampling of periphyton biomass from artificial substrata on an approximate six-week interval throughout the primary growing season (February - November) from 2000 to 2002. Sampling was not carried out during the winter months (November - February) when low light levels and cold temperatures result in lower growth. This is also the period of lowest water levels, which may have been an oversight in the sampling protocol of this project. In addition, the study used  1 4  C measurements of pelagic production  collected in 2003, to compare littoral production to pelagic carbon production of the Stave/Hayward ecosystem.  The following physico-chemical data were also collected as part of the S L Z M (2000-2003 data) and used to help with interpretation in this study: •  Light intensity (Photosynthetically Active Radiation (PAR);  •  Water temperature as a function of depth;  •  Nutrient data: Total Phosphorus (TP-P0 ); Total Dissolved Phosphorus (TDP); and Nitrate 4  (N0 -N), Chlorophyll-a (Chl-a). 3  A statistical analysis of these physical variables was considered, but consultation with Dr. A . Kozak (UBC) indicated that data was not collected frequently enough or in enough detail to conduct a regression analysis of the physical data for this study.  Chapter 2  17  Sampling Locations Stave Reservoir is larger than Hayward Reservoir with potentially higher spatial differences in periphyton growth due to more physical variability of the site. To account for variation, two transects were established on the eastern shores of the main basin in Stave Reservoir and one transect was established on the western shore of Hayward Reservoir. Difficulty was encountered establishing a location on the eastern side of Hayward reservoir due to steep gradients and numerous tree stumps in the littoral zone. The Hayward site was therefore established on the western shore and is assumed to reasonably represent periphyton growth within the reservoir. To ensure comparability a third transect was established on the western shore of Stave reservoir in spring 2001, to provide a more directly comparable site to the one established in Hayward Reservoir. A total four transects were established (three in Stave and one in Hayward) in mid February of 2000 (Figure 4). Ten 100 pound cement sampling blocks (Figure 9) were located at each transect in Stave to a depth of 20 m (64 m a.s.l.) and eight blocks were located at the Hayward transect to a depth of 16 m (28 m a.s.l.). Sampling blocks were positioned at approximately 2 m vertical intervals along a permanently established transect starting at 80 m a.s.l. in Stave Reservoir and at 42 m a.s.l. in Hayward Reservoir (Figure 10). The range of depth for each transect extended from the splash zone to at least 1 m below the mean light compensation depth of the reservoir at low pool.  Difficulty was encountered establishing the five deepest anchor blocks at all transect locations; soft bottom sediments of the reservoir were unable to support the weight of the cement block. To compensate for this problem a floating tray with an attached Plexiglas® plate was suspended from the anchor block at the appropriate elevation to maintain the 2 m transect depth interval. It is acknowledged that the floating trays potentially introduce differences in colonization and algal growth associated with their distance from the reservoir bottom and their ability to move with water currents when compared to the sampling blocks. Due to operational limitations of the project there was no simple way to account for this difference in the sampling program, therefore floating trays were adjusted to reproduce the height and conditions of the blocks as best as possible and it was assumed that the floating trays are reasonably comparable to the cement block sampling apparatus.  Chapter 2  18  Use of Artificial Substrata For the purpose of this study, the use of artificial substrata with known sampling size areas was beneficial to quantitatively estimate periphyton accrual. In addition, depth required the use of scuba divers to retrieve the sampling plates, therefore it was imperative to employ an efficient sampling protocol. Also in support of using artificial substrata in this study Cattaneo and Amireault (1992) report that for algal assemblages with a dominance of diatoms are relatively well reproduced on artificial substrata.  An artificial substrate was used to provide a standardized surface for periphyton growth, that was also easy to sample. It consisted of Plexiglas® sampling plates mounted horizontally above the anchor block with a 2.5 cm spacer to allow free flow of water around the plate (Figure 9 & 10). The plate surface was roughened with sand paper to create a suitable growing site for algal attachment. Plates were scored into quadrants to permit accurate quantitative sampling of periphyton (see inset, Figure 10). A 10x10 cm sample area has been reported to supply sufficient periphyton material to easily detect temporal changes in biomass and species composition in most oligotrophic aquatic systems (Shortreed et al., 1984).  The use of artificial substrata in contrast to natural substrata for sampling periphyton growth is common in limnological studies (Burkholder and Wetzel, 1990; Shortreed and Stockner, 1983; Evans and Stockner, 1972; Stockner and Armstrong, 1971). Cattaneo and Amireault (1992) addressed the usefulness of artificial versus natural substrata in periphytic studies by reviewing a variety of periphyton studies. They found that some groups, such as the green and blue-green algae are typically underrepresented by artificial substrata, while diatoms and whole periphyton assemblages are relatively well reproduced by artificial substrata experiments. Colonization times were found to be an important factor influencing how well the artificial substrata mimicked a natural substrata, be it epilithic (growing on rock), epiphytic (growing on plants) or epipelic (growing on sediment). For whole periphyton assemblage comparisons between rock and artificial substrata Cattaneo and Amireault (1992) indicated 83% taxonomic similarity with a colonization period of 31-60 days.  Chapter 2  19  Chapter 2  20  2.2.2  Periphyton Sampling  Periphyton samples were collected from each site during the primary growth seasons, from February 2000 to November 2002. Sampling began in late February or early March and continued at approximately six-week intervals, finishing in late October or early November. Sampling through the winter months, when periphyton production is reduced in response to low light, colder temperatures and lower water levels was conducted as one long sample period. Twenty sampling trips were completed in total, with six trips in the first year and seven trips in each subsequent year. Typically the first sample was collected in February with periphyton accumulation from the last sampling trip the previous fall (=Winter), two samples were collected in spring (Sp = March-May), two in summer (S = June-August) and two in fall (F = SeptemberNovember). On average there were 43 days between sampling trips (Table 2).  Table 2.  Table of sampling periods, number of days per sample period and Season.  Time (SP) TO T1 T2 T3 T4 T5 T6  Sampling Date  T7 T8 T9 T10 T11 T12 T13 T14 T15 T16 T17 T18 T19 T20  Chapter 2  Number of days / Sample Period  15-Feb-00  Season Start date - plates installed  12-Apr-00  56  23-May-00  41  Spring Spring  7-Jul-00  45  Summer  15-Aug-00  39  Summer  27-Sep-00  43  Fall  18-Nov-00  52  Fall  20-Feb-01  95  Winter  4-Apr-01  43  Spring  24-May-01  50  Spring  5-Jul-01  44  Summer  30-Jul-01  25  Summer  21-Sep-01  53  Fall  1-Nov-01  41  Fall  25-Feb-02  116  Winter  3-Apr-02  37  Spring  13-May-02  40  Spring  24-Jun-02  42  Summer  22-Jul-02  28  Summer  4-Sep-02  43  Fall  25-Oct-02  51  Fall  21  A scuba dive team operating from a boat removed plates. Two divers retrieved five plates at a time and placed them into a receiving box specifically designed to firmly hold the plates in place to reduce disturbance while the plates were collected and brought to the surface. Once in the boat, the plates were stored in the retrieval box and covered to reduce exposure to sun and drying. Each plate was removed from the box in numbered order and a predetermined area (100 cm ) scraped and rinsed (for species composition, Chi a, and A F D W ) into a labeled jar for transport to the laboratory (Figures 11 & 12). Scraping was done using a glass microscope slide and all samples were kept cold until lab processing.  2.2.3 Environmental Data Light Solar radiation data from Port Moody was used to estimate surface irradiance at the sampling sites and to estimate sample period solar radiation. Surface solar radiation for the study area was estimated using hourly measurements of global radiation (sum of direct and diffuse solar radiation) collected by the Greater Vancouver Regional District (GVRD) at Port Moody using a LI-COR pyranometer (LI-200SA, Lincoln, NE). Solar radiation data collected in this manner includes most of the visible light spectrum and some infrared wavelengths (400 - 1100 nm), and therefore includes a slightly wider range of radiation than is typically used in limnological studies (400-700 nm). Surface solar radiation data from G V R D was used in this study to provide a continuous record of solar radiation. Although the site is located some distance (~ 40 km) from the sampling sites, it is assumed that this data is representative of the solar radiation reaching the surface of both Stave and Hayward reservoirs since this is the only complete record available. The Port Moody site was chosen from several locations where the G V R D collects data based both on proximity and topographical similarity to the sampling site.  In addition to surface solar radiation data, measurements of light at 1 m depth intervals were made using a LI-COR submersible quantum sensor (LI-250, Lincoln, NE), at each transect on the day of sampling as a method to calculate light extinction coefficient (k) and compensation depth of Stave and Hayward reservoirs which was in turn used to define the littoral area of each reservoir.  Chapter 2  22  Measurement of attenuated solar radiation was collected in units of Photosynthetically Active Radiation (PAR). P A R is a direct measure of irradiance in the 400-700 nm range. Photosynthesis occurs only in the photic zone, which is the layer of the water column that receives sufficient light energy to support photo-autotrophic production (i.e. photosynthesis > respiration). The depth at which light is diminished to one percent of its surface value defines the lower boundary of the photic zone; this is known as the compensation depth (photosynthesis = respiration). In near-shore areas the compensation depth and photic zone can be thought to define the littoral zone area of aquatic ecosystems.  Temperature Water temperature measurements collected by B C H at several stations and depths on Stave Reservoir and at Ruskin dam tailrace between 1996-2001 were used to estimate an average annual temperature profile at 3 m below the surface for Stave Reservoir and at Ruskin dam tailrace for Hayward Reservoir. In addition, temperature profiles at 1 m depth intervals were collected to a depth of 20 m at each transect on the date of sampling. These profiles were used to assess stratification periods and thermocline depth on the two reservoirs.  Nutrients Nutrient measurements from the water column for the Stave/Hayward system were made the same day that biological sampling from each transect took place during the years 2001 and 2002. In 2000, three nutrient samples were collected on separate dates from when sampling was completed. Nutrient samples were collected as a composite sample from 1,3, and 5 m depth at each station. Samples collected in Hayward Reservoir, the main basin of Stave Reservoir, at the Allouette facility discharge and at Ruskin power plant were analysed at Cultus Lake Laboratory (Cultus Lake, BC), for total concentration of phosphorus (TP), total dissolved phosphorus (TDP), nitrates (nitrite plus nitrate) and chlorophyll-a concentrations (Chl-a). This study will focus on data collected from Hayward Reservoir and from the main basin in Stave Reservoir to help explain patterns in periphyton growth and nutrient dynamics.  Chapter 2  23  Water Level Fluctuations  Daily average surface water elevations collected by B C H at each facility were provided for the duration of the study. A n average sample period water elevation was used to analyse the impact of changing water levels in the reservoirs.  2.2.4  Laboratory Analysis  Periphyton samples collected at Stave and Hayward reservoirs were stored cold and dark during collection and until analysis was completed. Laboratory analyses were completed at U B C within 24 hours of sampling.  Ash-Free Dry Weight  Periphyton samples scraped from one of the quadrants of the sampling plate were filtered at low vacuum pressure onto a pre-ashed Whatman glass fibre filter (GFF). Filters were placed in an aluminium weigh boat and dried to a constant weight (DW = dry weight) in an oven at 100°C for 12-24 hours and weighed. Filters were then ashed at 500°C in a muffle furnace for 5 hours and then reweighed (AFDW = ash-free dry weight). The difference between the D W and A F D W for the sample period divided by the number of days in the sample period, or periphyton accrual is expressed in mass of organic matter per unit area per day (mg/cm /day). The carbon (C) 2  component of periphyton accrual is calculated as 45% of the sample organic matter (Stockner and Armstrong, 1971).  Chlorophyll  Chlorophyll (Chl-a) samples were filtered onto a 47 mm diameter, 0.45 pm porosity, Whatman GFF filters. Chlorophyll samples from one quadrant (100 cm ) were poured into a Millipore 2  filtration funnel and rinsed twice with de-ionized distilled water using a low-pressure vacuum (<20 cm Hg). The filtration system was turned off immediately once the sample filtered through. Filters were folded and stored frozen in labeled aluminium weigh boats until they were extracted and analysed.  Chl-a extractions were carried out by placing each filter into a test tube where 8-15 ml of 90% acetone was added and the sample sonicated in an ice bath for 7-10 min. The test tubes were stored in the dark and frozen (-20°C) while the extraction took place (20-25 h). The extracted Chapter  2  24  liquid was then poured into culture tubes that were inserted into a Turner Fluorometer and fluorescence readings taken before and after three drops of 10% Hydrochloric acid (HC1) was added. These measurements were then converted into measurements of Chl-a (mg/m ) using the 2  following equation (Parsons et al., 1984):  Where: Fo = Fluorescence reading before acidification Fa = Fluorescence reading after acidification v = Volume of acetone extract V = Volume of sample filtered 1.79 = Acid ratio correction factor  1.79 x 1.128 (Fo-Fa) x (v/V)  For Stave and Hayward reservoir Chl-a samples, the volume of the sample filtered was corrected because a known area was used rather than a volume.  Species Composition Algal species composition was evaluated for two depths (plate 4 = 8 m and plate 8 = 16 m, below full pool) at each transect for the duration of the study. Species abundance data (cells/cm ) 2  from this analysis was summarized into dominant species at the depth of peak production (4 m below average water elevation) for each transect.  Species composition was evaluated using flat-bottomed settling chambers and a Nikon inverted microscope employing a modified version of the well known Utermohl technique that is described in detail in Beer and Dolecki (2000). Algal abundance was expressed as number of cells per square centimetre (cells/cm ) and as biovolume (mm ). 2  3  Additionally data was analysed using nonmetric multidimensional scaling (NMS), a data reduction tool commonly used in ecology, to provide a preliminary evaluation of the relationships between species composition and individual samples. N M S analyses of species data were undertaken using PC-ORD software following the recommendations of McCune and Grace (2002). Stress and instability of 142 samples was evaluated for solutions in one to four dimensions, using the Sorensen distance measure, a random starting configuration and 15 runs of real data. Stress of the real data solutions was compared to those generated by random configuration of the data using 30 Monte Carlo simulations. Based on the initial test a 2dimensional solution was recommended and a final analysis was rerun for this dimensionality Chapter 2  25  using a single run with the same distance measure, the same starting configuration, and 63 iterations to assess stability.  2.2.5 Pelagic Primary Production Sampling and Analysis Pelagic primary production was estimated using a carbon-14 method (Steemann Nielsen, 1952). This method used the uptake of a isotope marker to estimate the amount of in-situ microbial activity. Water samples were taken at 1, 3, 5, and 7 m in both Hayward and Stave reservoirs and an additional 10 m sample was taken in Stave reservoir using a Van Dom water sampler to obtain water samples from discrete depths. The depths were selected to represent a profile from the surface to a depth that approximates the compensation - depth (1% light level) in each reservoir. One dark and two light samples were transferred directly from the Van Dom bottle into acid washed B O D bottles at each depth; these were stored in a light-tight box, until all samples were collected. In addition, water samples for analysis of chl-a concentration were collected at each depth and water samples for analysis of alkalinity were collected from the top and bottom of each profile. Care was taken to limit light exposure during all parts of the collection and storage procedure until the start of incubation, usually within 1 h of sampling.  Once collected, samples were inoculated with 0.185 M B q (5 uCi) of and rates of primary productivity were determined from the amount of  1 4  C incorporated into particulate organic  carbon retained on a filter (Steemann Nielsen, 1952). Inoculation of the dark bottle with  1 4  C  determined non-photosynthetic C incorporation by the sample. Each of the three sample bottles 1 4  from their respective depths was attached to an acrylic plate designed to hold the bottles in a horizontal plane at right angles to each other and then re-suspended to the original sampling depth. Samples were then left to incubate in-situ for 2-4 hours, generally between 10 a.m. and 2 p.m. After incubation, samples were retrieved and placed back into light-tight boxes for immediate transport back to the laboratory.  The incubations were terminated by filtration through each of a 0.2, 2.0 and 20.0 pm 47mm polycarbonate filter using < 10 cm Hg vacuum differential (Joint and Pomroy, 1983) and each filter was placed into a 7 ml scintillation vial. Samples were stored in the dark until processing. 200 p L of 0.5 N HC1 was added to each vial to eliminate the unincorporated inorganic  Chapter 2  26  NaH CC>3 and the vials left uncapped in the fumehood to dry (approximately 48 hours). Once 14  dry, 5 ml of Ecolite® scintillation fluor was added to each filter and stored dark for at least 24 hours. Samples were then counted for 10 minutes using a Beckman® Model #LS6500 liquid scintillation counter operated in an external standard mode to correct for quenching (Pieters et al., 2000).  The specific activity of the stock was determined by adding 100 p L C-bicarbonate solution to 14  scintillation vials containing 100 pL of ethoanoalamine and 5 ml Ecolite®scintillation cocktail and counted using the same counter. This assay determined the total activity added (DPM aOtot  Hourly primary productivity rates were calculated according to that described by Parsons et al. (1984). Daily primary productivity was obtained by dividing the primary productivity rate during the incubation by the ratio of the incubation period irradiance to the total daily irradiance. Primary productivity rates were vertically integrated by averaging the measured primary productivity between two depths, then multiplying by the depth interval (Ichimura et al., 1980). The assimilation rate is the photosynthetic rate per unit of chlorophyll and was calculated by dividing the daily productivity rates (mg Cl m /d) by the Chl-a concentrations (mg Chl-a/ m ). 3  Chapter 2  3  27  CHAPTER 3  3.0  R E S U L T S AND DISCUSSION  3.1  Physico-Chemical Features of the Stave/Hayward Ecosystem  3.1.1 Light Inferred surface solar radiation for Stave and Hayward reservoirs followed a pronounced seasonal pattern. Winter values were in the 25-50 W / m range. Spring values increased steadily by month to a summer peak in the 225-300 W / m range. Surface solar radiation in autumn 2  followed a declining trend, steadily decreasing each month until it reached a low in early January (Figure 13). Surface solar radiation measured on the day of sampling and averaged for the duration of the sampling period followed a similar pattern with winter averages remaining in the 25-50 W / m , while peak summer values averaged during sample periods were slightly lower (150-210 W / m ) than the daily Port Moody data (Figure 14). 2  Sample period light extinction over the course of this study did not vary greatly either by season or by location. Light extinction values among the three Stave Reservoir transects were similar, therefore, these sites were combined into an overall Stave Reservoir average extinction value. In Stave Reservoir, extinction coefficient values ranged from 0.30 to 0.52, with an average value of 0.40. Extinction coefficient values in Hayward Reservoir were more variable ranging from 0.34 to 0.70, with an average value of 0.44 (Figure 15).  The depth at which light is diminished to 1% of its surface intensity is called compensation depth. It is considered the critical value for growth in limnological ecosystems, defining the depth where production is equal to respiration. Seasonal patterns in compensation depth varied from 7 to 15 m in depth. Attenuation data were more variable and given the lack of data from winter months, no strong pattern was observed (Figure 16). It appears that compensation depth from late fall through early spring was approximately 8-10 m, while in summer months it was 12-14 m.  Chapter 3  28  GVRD Solar Radiation (Port Moody) 300 250 200 CM  -Li  J  ft  100  Ji  50  0 J-00  1 in  I  \\ Af 'il j \s  E 150 5  I i\.  J  I  1  1  M-00 S-00 J-01 M-01 S-01 J-02 M-02 S-02  7-day running average  Time  Figure 13. GVRD surface solar radiation datafromPt. Moody, BC.  Port Moody Surface Solar Radiation Sample Period Average 300  200 CM  E 150 5  2001  2000  250  2003  / \\  100 /  50 0  1  1  1  1  — i — i — i — i — i — i — — i — i — i — i — i — i —  1 2 3 4 5 6 7 8 9 1011121314151617181920  Sample Period  Figure 14. Average sampling period irradiance estimated from Pt. Moody.  The compensation depth in Stave Reservoir during the spring (May) tended to be higher than any other time of the year and higher than in Hayward Reservoir at the same time. The spring-time compensation depth in Stave Reservoir was typically 14-16 m. Morefrequentmeasurements of light during this study would have helped significantly to define the light profile of the two reservoirs.  Chapter 3  29  Stave & Hayward Extinction Coefficient  0.80  Jan  May Sep Jan May Sep Jan May  Sep  l Hayward  I Stave (3 transect average)  Figure 15. Stave/Hayward percent light extinction. S t a v e & Hayward C o m p e n s t a t i o n D e p t h 2000 Jan  2001  May S e p J a n  2003  May S e p J a n  I Stave (3 transect average)  May  Sep  • Hayward  Figure 16. Stave/Hayward compensation depths. (Depth of 1% light level) Stave and Hayward Secchi Depths 2002/2003 2002 Jan  Apr  2003  Jul  Oct  Jan  0  1  I i &  2  3  •r  5  6 7 I Stave  l Hayward  Figure 17. Stave/Hayward secchi depths.  Chapter 3  Apr  Jul  Oct  Light penetration depths in Stave and Hayward were generally quite comparable. Stave Reservoir exhibited a maximum secchi reading of 6.2 m in July of 2003 and a minimum value of 1.0 m in October 2003 shortly after the storm event. Hayward Reservoir exhibited maximum secchi depth of 6.3 m and a minimum of 0.6 m that occurred in the same months. In general, Hayward Reservoir secchi depths were slightly lower than those in Stave Reservoir, with average values of 3.9 m in Hayward and 4.1 m in Stave (Figure 17). Secchi depth measurements from Stave and Hayward reservoirs indicate that lower transparency occurred in late fall and winter months, largely in response to lower light levels. Lower transparency may also be the result of storm events that result in particulate matter in the water column, as was the case for the fall 2003 measurements. Light levels dropped dramatically following a large storm event in 2003.  3.1.2  Temperature  Water temperature measurements taken at 3 m below surface water elevation in Stave Reservoir from 1996-2001 indicates that average winter temperature at full mixing were approximately 5°C, increasing in the spring with epilimnetic temperature in the 10-12°C range. Surface temperature typically peaked in late August with average temperatures reaching 16-18°C and declined in autumn to 10°C (Figure 18). This temperature regime is similar to other coastal temperate lakes (Shortreed et al., 1983; Stockner and Shortreed, 1985).  Stave/Hayward R e s e r v o i r s A v e r a g e T e m p e r a t u r e 1996-2001 20 ,  0 J  F  M  A  M  J  J  A  S  O  N  D  Month — « — Stave (3m)  — • — Hayw ard (Ruskin Tailrace)  Figure 18. Stave/Hayward average 1996-2001 temperature regime.  Chapter 3  31  Water temperature measurements taken at 1 m depth intervals in 2003 (the most complete year of data) indicated that the thermocline in Stave reservoir begins to develop in May at approximately 6-7 m depth (Appendix 4). Surface temperatures at the start of stratification were in the mid-teens, dropping 2-3 degrees to approximately 12°C at the base of the thermocline. By mid-August the thermocline was more strongly developed and deepens slightly. Surface water temperatures peak at approximately 22°C and water column temperature remains relatively constant to a depth of approximately 6 m (the top of the thermocline). Below 6 m temperature decreases more rapidly through the thermocline layer reaching 17°C at 9 m (the base of the thermocline). Hypolimnetic temperature was 8°C. In fall, ambient air temperatures begin to drop, accompanied by episodic events of wind and wave activity that mix the water column. By September, the thermocline becomes less pronounced and deeper (11-15 m) as epilimnetic water becomes cooler and denser as the lake undergoes turnover. Isothermal conditions characterize water temperatures in Stave Reservoir by mid-October or early November, depending on seasonal weather.  Although Stave and Hayward reservoirs share the same water mass, the temperature regime in Hayward tends to differ from Stave in two ways. Water temperatures in Hayward Reservoir tended to be slightly higher on average than Stave Reservoir. In winter, average 3 m water temperature in Hayward Reservoir was 6°C, approximately one degree warmer than the average Stave Reservoir temperature. Average springtime temperatures were in the 12-14°C range, approximately 1-2°C higher than measured in Stave Reservoir. Summer temperatures were more similar between the two systems and were commonly in the 20-22°C range. The second and more significant contrast in temperature regimes between Stave and Hayward reservoirs was that Hayward Reservoir does not develop a thermocline (Appendix 4). Hayward Reservoir remained mixed (isothermal) throughout the year. This is significant because surface water temperatures in Hayward Reservoir tended to extend deeper in the water column, which has the potential to facilitate decomposition and in turn promote growth.  3.1.3 Nutrients Phosphorus concentration in Stave and Hayward were measured as both total phosphorus (TP) and total dissolved phosphorus (TDP), with TDP representing the closest approximation of Chapter 3  32  bioavailable phosphorus measured in this study. Comparing TP and TDP concentrations between Stave and Hayward reservoirs indicate that TDP is generally 25-40% less than TP values in most reservoirs, which has been found to be a typical relationship between the two measures (Stockner, 2003 pers. comm.). Both Stave and Hayward reservoirs had very low concentrations of phosphorus, exhibiting TP concentrations ranging from 2.0-6.8 pg/L and TDP concentrations in the <1.0-5.3 pg/L range (Figures 19 and 20). Stave TP and TDP 8.0  2002  2001  2000  7.0 6.0 _,5.0 o) 4.0  lit  t:  11 Ii  Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct l Stave TP D Stave TDPJ  Time  Figure 19. Stave TP and TDP Concentrations.  Hayward TP and TDP 8.0 7.0 6.0 -.5.0 oi4.0 3.0 2.0 1.0 0.0  2002  2001  2000  —  3  1 m  II  ii  —  —k  it 1  Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct l Hayward TP  • Hayward TDP 1  Time  Figure 20. Hayward TP and TDP Concentrations.  Seasonal patterns in phosphorus measured at Stave and Hayward reservoirs were variable throughout the study, making interpretation of results more difficult. Concentrations of TP and TDP in Hayward Reservoir seem to exhibit more of a seasonal pattern than Stave Reservoir. In general, TDP in both reservoirs decreased from winter into late spring, increased into summer  Chapter 3  33  and then decreased to a fall low that tended to coincide with periods of highest production (Figure 21).  Chlorophyll 1.5  2000  1.3  d D>  ^  o  2002  2001  1.0 0.8 0.5 0.3 0.0  JL_J  1 Jan  i Stave  ill  Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct • Hayward |  Time  Figure 21. Stave/Hayward integrated pelagic chlorophyll-a concentrations.  In Hayward, TDP exhibited peak values of 3.5 pg/L in July 2001 then decreased steadily through fall (October). Following a similar trend as in 2001, 2002 peak values occurred in both May and July (5.3 pg/L) and declined to just over 1.0 pg/L by September. Hayward TP followed a similar partem to TDP in 2001, but in 2002 Hayward TP exhibited a different pattern increasing steadily from winter to a peak of 6.8 pg/L in late spring. Stave Reservoir TP and TDP concentrations were less variable and did not exhibit a distinguishable seasonal pattern. TP concentrations in Stave remained in the 2.0-4.0 pg/L range and TDP concentrations were in the 1.0-3.0 pg/L range throughout the study.  Peak nitrate (NO3-N) concentrations in Stave and Hayward reservoirs ranged from 160-180 pg/L between October and April, declining to a late summer low of as little as 40 pg/L. Nitrate measurements generally exhibited a more seasonal pattern than phosphorus (Figure 22). Peak values occurred in fall and winter months (October-March), concentrations declined into spring and exhibited a summer low, generally occurring in August or September. Average nitrate values observed during May through September in Stave Reservoir were 98.2 pg/L and 137.6 pg/L  Chapter  3  34  from October through April. A similar trend is evident in Hayward Reservoir with average nitrate values of 100.00 pg/L in summer and 143.25 pg/L through fall and winter.  Ratios of NO3-N: TDP for Stave and Hayward reservoirs exhibit a partem with low values occurring from June through September and high values occurring from October through May (Figure 23). Maximum values in both Stave and Hayward occur in May (129.8:1 and 181.2:1, respectively). Minimum values occurred consistently in July in Hayward Reservoir with values of 24.5:1 in 2001 and 27.3:1 in 2002. In Stave Reservoir, minimum values generally occurred later (late summer) with values as low as 16:1. Stave/Havward Nitrate Concentration  Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct • Stave  oHaywardl  Time  Figure 22. Stave/Hayward integrated pelagic nitrate concentrations.  Stave & Hayward Ratio of N 0 3 - N : T D P 200 OL  •  2000  i  150  2001  2002  100 O  50 0  niHiiL  II Hi X t. Ill I  Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan 1 Stave  • Hayward  Time  Figure 23. Stave/Hayward NO3-N: TDP ratio.  Chapter 3  35  3.1.4 Water Level Fluctuation and Residence Times Hayward Reservoir operates as a run-of-the-river hydro-electric facility whose main purpose is to control flooding downstream of Ruskin Dam. As a result water levels in this system are generally held constant (figure 24). For the duration of this study (Feb 2000-October 2003), Hayward Reservoir water surface elevations remained constant, operating between 41.41 to 42.8 m a.s.l. with an average water level of 40.30 m a.s.l. As a requirement to complete maintenance on Ruskin Dam, Hayward Reservoir experienced one significant and unexpected period of drawdown during the study in June 2001. Drawdown occurred rapidly, falling to a low of 33.12 m a.s.l. and refilling 41.98 m a.s.l. over a period of nine days. Although the objective of this study was to compare between a fluctuating (Stave Reservoir) and a stable reservoir system (Hayward Reservoir), the drawdown on Hayward provided a unique opportunity to evaluate the impact of a single, intense drawdown event on periphyton accrual. Outside of this period of drawdown for maintenance, average water fluctuations between sample periods in Hayward Reservoir was 0.08 m.  In contrast to Hayward Reservoir, Stave Reservoir undergoes more frequent and extreme periods of drawdown (figure 24). Stave Reservoir has the potential to fluctuate by up as much as 9 m, but for the duration of this study followed the operating regime proposed by the Stave Reservoir WUP, resulting in a more stable water management regime (< 6 m fluctuation) than what was previously typical. Regardless of adopting the new W U P water management regime, Stave Reservoir continues to undergo more extreme water level fluctuations than Hayward Reservoir.  The cycle of drawdown in Stave Reservoir generally follows an annual pattern; lowest water surface elevations occur in late fall, winter and early spring, increasing to a summer maximum (usually occurring in July), then gradually declining again through the late-summer and fall. This cycle allows for reservoir recharge, during late fall and from spring rain and snowmelt, and facilitates recreational-use such as fishing and boating during the summer months. During this study Stave Reservoir experienced a maximum water surface elevation of 82.02 m a.s.l. and a minimum of 73.15 m a.s.l. Average water fluctuation was 1.7 m between sample periods. Since both systems are operated as hydroelectric storage facilities, water resident times are short. This is especially true in Hayward Reservoir where residence times are measured in days, and also in  Chapter 3  36  the epilimnion of Stave where volumes can be replaced up to 5 times during the growing season (April-October) (Figure 25).  Average Sample Period Water Level Fluctuation 2000-2002  Date  Figure 24. Stave/Hayward water levels (2000-2002).  Figure 25. Stave Reservoir epilimnion volume replacements 1952-2002 (analysis and graph provided by B C Hydro).  Chapter  3  37  3.1.5 Discussion of Environmental Data Light Light extinction values in Hayward Reservoir indicate that under specific climatic conditions, such as storm events, light penetration through the water column is substantially inhibited relative to Stave Reservoir. Hayward is a smaller, shallower reservoir, with a littoral zone characterized by fine sediments (silts and clays) that are easily suspended in the water column. In addition, organic matter on surfaces and old stumps in the littoral area were noted to cloud the water column if disturbed. These characteristics explain, at least in part, the higher light extinction values exhibited occasionally in Hayward Reservoir. It is interesting that higher than average light extinction coefficients were not exhibited shortly after the drawdown event that occurred in summer 2001. Stave Reservoir is much larger and divided into an upper and lower basin by distinct physiographic and surficial geology differences. The upper, main basin of Stave is characterized by a rocky shoreline and in many areas by steep bedrock walls. Along the littoral gradient the character shifts to a muddy, fine sediment lake-bottom that is deep enough (8-10 m) to not be eroded during occasional drawdown events. As a result, secchi depth transparency was greater in Stave Reservoir than in Hayward.  The lower reach of Stave Reservoir draws its character from the old Stave River channel. Nearshore areas are large expanses of flooded riparian and forest habitats that are relatively flat, with a defined river channel that is evident during periods of drawdown. These areas are more susceptible to erosion and may act as a source of particulate and organic matter that in turn attenuate irradiance.  Light attenuation with depth follows an exponential decay in the water column. When light penetrates the water column it interacts with water itself, phytoplankton, particulate matter suspended in the water column, or dissolved organic matter particularly dissolved carbon that act to absorb, reflect and scatter light. The extinction coefficient is a measure that describes the amount of light extinguished at a given level within the water column. No seasonal patterns in light extinction were discernable from data collected in this study. This is most likely a function of the frequency of measurement. Observations were made only at the time of sampling (once Chapter 3  38  per 6 weeks, roughly), which provides only a brief snapshot of the variability in light attenuation in each reservoir. Conditions for light absorption and scatter can change relatively quickly in lake systems and reflect many things about the condition for growth. These changes are likely to occur even more rapidly and frequently in reservoir ecosystems where water levels are manipulated in response to water management needs, but sampling frequency was not intense enough to reveal trends in the data.  Temperature  Seasonal fluctuation of air and water temperature created by changes in duration and angle of the sun determine the degree of thermal stratification in lakes. Heating of the water body occurs when solar energy is transferred at the air-water interface. This increase in surface temperature plays an important role in determining both the stratification and the circulation of lakes. Surface heating often creates a stably stratified water column (thermocline layer) due to differences in water density. Episodes of wind and turbulence with dense cooler water below warmer less dense water, causes mixing that acts to transport oxygen and nutrients through the water column and results in an isothermal temperature profile.  There are several concepts that offer explanation for differences in temperature regime between Stave and Hayward reservoirs. Hayward Reservoir is much smaller and shallower than Stave Reservoir (5 km vs. 25 km, respectively), hence it is likely that the smaller volume of water in Hayward Reservoir is warmed more readily than the waters of Stave Reservoir.  Another factor impacting water temperature is depth of water intake at Stave Falls dam. The depth at which water is withdrawn from Stave Reservoir before it is passed through the turbines determines the temperature of the water entering Hayward Reservoir. Intake level at Stave Falls dam is 68 m a.s.l., measured at the mid-point of the penstock. Based on average surface water elevation of Stave Reservoir, water entering Hayward Reservoir is, on average, 11m deep and based on the temperature profiles taken during this study, would be approximately 12.4°C. Thus, water entering Hayward Reservoir is relatively warm metalimnetic water derived from lower Stave Reservoir.  Chapter  3  39  Finally the most obvious explanation for the difference in water temperature between Stave and Hayward reservoirs is that Hayward Reservoir is operated as a run-of-the-river facility. Water within the system flows and is refreshed in a period of days, resulting in continuous mixing of the water mass, thus isothermal conditions prevail (no thermocline). Surface water is continually renewed and warmed resulting in an overall warmer profile with depth, especially over the period that would typically be stratified in a larger reservoir with shorter residence times (MayOctober).  Nutrients Nitrogen and phosphorus concentrations were variable through the duration of the study. Water samples were taken on the same 6-week interval as the periphyton sampling completed on Stave and Hayward reservoirs, therefore, the frequency of nutrient data is not well defined throughout each sample period. In addition, the samples taken were a composite of 1, 3, and 5 m water depths and as a result, nutrients cannot be used to interpret depth effects on periphyton growth. However, the data provides valuable information regarding the general patterns and concentrations influencing littoral production of Stave and Hayward reservoirs.  The phosphorus concentration findings are important because they characterize Stave and Hayward reservoirs as oligotrophic ecosystems. Wetzel (1983) defines oligotrophic lakes as those with total phosphorus concentrations between 5-10 ptg/1 and the Washington State Administrative Code (2003) defines ultra-oligotrophic ecosystems as those with 0-4pg/l TP and oligotrophic systems with 4-lOjug/l TP. Under these definitions Stave Reservoir would be classified as an ultra-oligotrophic ecosystem, while Hayward Reservoir would be classified as an oligotrophic system.  Low phosphorus values such as those measured in Stave and Hayward reservoirs during this study are consistent with other nutrient deficient coastal temperate lakes. Shortreed et al. (1984) conducted a nutrient response study on 21 British Columbia lakes and found average growing season TP levels, including both treated and untreated lakes, ranging from 1.5-6.9 ju.g/1. Mesooligotrophic Castle Lake in northern California is reported to have TP concentration near to 5 /ig/1 and TDP concentrations of approximately 1 /xg/1 (Axler et al., 1996). In a study conducted on coastal temperate lakes in British Columbia, Shortreed et al. (1978) reported phosphate (PO4-  Chapter 3  40  P) concentrations of 1.8 jU.g/1 in an unfertilized area of Great Central Lake and 3.6 /xg/1 in the main arm of Kennedy Lake.  In addition, it is interesting to compare nutrient concentrations reported for Carnation Creek on the west coast of Vancouver Island to the reservoir nutrient levels measured in the two reservoirs of this study. This comparison is germane to a reservoir ecosystem such as Stave because of the complex relationship to Stave River from which Stave and Hayward reservoirs were created and because of the short residence times in this system. Shortreed et al. (1983) report pre-logging phosphorus concentrations ranging from <l-9 /ug/l measured at Carnation Creek, which are similar to the range exhibited in Hayward Reservoir that has retained more riverine characteristics, as a result of short residence time and run-of-the-river management regime.  Peak phosphorus values in spring are likely associated with spring run off, allochthonous nutrient inputs, increased solar radiation and subsequent warming of the water column. Although chl-a concentrations measured concurrently to nutrients in Stave and Hayward reservoirs do not provide a continuous record of production, higher chl-a values measured during the study appear to correspond with periods of lower TDP, particularly in Hayward Reservoir indicating uptake of phosphorus by organisms in the water column.  Similar to phosphorus, the range of nitrate concentrations measured on Stave and Hayward reservoirs are typical of coastal temperate lakes. Stockner et al. (1985) report nitrate concentrations on 13 coastal lakes in British Columbia ranging from 17-148 mg/m . While the 3  measured values for Stave and Hayward were low, they are not as low as nitrate concentration measured on some coastal B C lakes. Shortreed et al. (1978) report mean nitrate concentrations in the unfertilized area of Great Central Lake, British Columbia of 1.0 /xg/L and maximum levels in the study measured at Long Lake, British Columbia of 13.2 jug/L. Castle Lake, California is reported to have dissolved inorganic nitrogen concentration of < 10 fig/L (Axler et al., 1996).  The decreased nitrate concentration during the stratified period in Stave Reservoir may, in part, be explained by increased chl-a values that represent increased nutrient uptake by organisms. As noted earlier, Hayward Reservoir remains mixed (isothermal) even during the summer months when Stave Reservoir, like many coastal lakes, becomes stratified. Surface temperatures in  Chapter 3  41  Hayward can be as high as 22°C in the peak of summer and decline gradually with depth. Hayward temperature does not exhibit the sharp decrease associated with thermocline, therefore summer temperatures in Hayward are likely to be warmer with depth than temperatures in Stave Reservoir. Under conditions of high water clarity it is possible that warmer temperatures with depth may allow more periphyton growth to occur in Hayward Reservoir than in Stave Reservoir. Higher biomass in Hayward is evident in the Chl-a concentrations measured in Hayward Reservoir as part of the nutrient study, as well as in the biomass measurements made from the sampling blocks.  Nutrient Ratios  Comparing of nutrient ratios, namely nitrogen to phosphorus, is the most common criteria aquatic scientists use to determine nutrient limitation. Unfortunately few studies examine growth-limiting concentrations of nutrients for benthic algae in lentic ecosystems. The majority of nutrient limitation studies have focused on streams and rivers and of the studies conducted, most fail to quantify the concentration of nitrogen and phosphorus that are growth limiting. Borchardt (1996) lists and summarizes benthic algal nutrient studies conducted in the last half century, and confirms the Redfield ratio C: N : P (106:16:1) as the benchmark for optimum nutrient levels for assessing nutrient limitation at both the community and cellular level in freshwater ecosystems. Studies on benthic algae (periphyton) conducted by Shanz and Juon (1983) indicate that for benthic algae (periphyton) ratios of ambient N:P of 20:1 are considered phosphorus limited, while ratios of 10:1 are nitrogen limited and ratios ranging from 10-20:1 represent the transition between N and P limitation and are not definitive.  When examining unproductive temperate lakes, geographic generalizations regarding the pattern of nitrogen or phosphorus limitation have been suggested. Borchardt (1996) references studies that indicate that northern U S A is likely to be phosphorus limited, whereas the southern and southwestern U S A is more likely to be nitrogen limited. In the Pacific Northwest, nitrogen and phosphorus are both reported to be growth limiting. Axler et al. (1996) assert that lakes with low external loading of nitrogen and phosphorus are likely to exhibit either nitrogen limitation or colimitation with phosphorus. Stockner et al. (1978) indicate that when phosphorus limited lakes are augmented via phosphorus enrichment, nitrogen is frequently found to be secondarily limiting. Based on the N:P ratios calculated for Stave and Hayward reservoirs it would appear  Chapter 3  42  that both systems tend towards phosphorus limitation, with ratios near to the 20:1 benchmark. While nitrogen concentrations were found to be near limiting levels by the end of the growing season, nitrogen limitation does not appear to be a key factor controlling nutrient dynamics in Stave and Hayward reservoirs.  Water Level Fluctuation Reservoir littoral zones are altered in area and in vertical profile by water management regimes. When drawdown occurs the most obvious change is exposed shoreline. Exposed to direct solar radiation, wind and other elements, the upper areas of the littoral habitat dry-out and the periphytic community desiccate. The photic zone shifts downward along the near shore gradient and coinciding with this downward shift, the new water surface experiences increased light and temperature; the zone of highest productivity also shifts to a lower elevation. As water levels change sediment and woody debris are frequently disturbed. Logs and rocks move with the force of the water, eroding sediment and destroying established microbial communities. Eroded sediments can be entrained in the water column resulting in a reduction of light and nutrients may be released from bottom sediments. Remaining organisms in the vertically shifted littoral zone are forced to adapt to a new and potentially less hospitable environment that is typically less productive than natural lakes.  One possible theory that may explain the generally unproductive nature of reservoir ecosystems is the concept of "pulse stability" introduced by Odum (1971). This concept describes the situation where ecosystems experiencing repetitive or periodic physical perturbations are maintained in a state of intermediate succession. In other words, these communities exist in a less than optimal productive state and are unlikely to ever reach climax succession. In the case of large-scale coastal temperate reservoir ecosystems, it is often water level fluctuation that is the primary disturbance to inhibit productivity of the littoral community.  Cycles of wetting and drying associated with water management in reservoirs, like Stave, have been found to have significant adverse effects on biological function, particularly on the interface communities that exist at the edge of the vertically shifting littoral zone. Normal seasonal growth cycles are frequently interrupted and unable to re-establish after periods of wetting and drying to the detriment of littoral organisms. Ongoing fluctuation, combined with  Chapter 3  43  flushing events directly associated with power generation, manipulate the aquatic community to the extent that it exists in a state of perpetual disruption (i.e. constant state of early succession). The pattern of water management in reservoirs is in contrast to natural ecosystems where gradual less intense periods of inundation and drying promote diversity and habitat complexity. Thus, pulse stability in the littoral zones of reservoirs is driven by the disequilibrium and dysfunction that characterizes them as ecosystems.  Chapter 3  44  3.2  Biological features of the Stave/Hayward Littoral Zones  3.2.1 Species Composition Results and Discussion For the purpose of this study, species composition data from plate 4 (8 m depth, or approximately peak biomass) and plate 8 (16 m depth) from all three years were analysed. While many more samples were evaluated and enumerated and may provide additional information, this assessment provided preliminary findings of dominant species (ten most abundant species based on #cells/cm and biovolume) and possible patterns in species composition for Stave and 2  Hayward reservoirs. Algal species composition was similar among all four transects, with diatoms (Bacillariophyta) dominating both abundance and biomass of the littoral communities. A total of 67 species were identified. Of these, nine were considered to be pelagic, occurring as resting cells in the samples. These cells were rare and did not account for appreciable biomass. Excluding the pelagic species, it was found that diatoms accounted for 72% of the total, green algae for 16%, bluegreen algae for 9% and red algae for 2%. Table 3 provides a list of littoral species and average abundance and biovolume estimates.  Analysis of algal abundance (cells/cm ) indicated that the same species dominate both Stave and 2  Hayward reservoirs. The ten most abundant species occurred at all four transects but in varying proportions. One exception to this finding was the occurrence of Frustrulia rhombiodes, which was found to be the 10 most dominant species at Stave north, excluding Eunotia lunaris at this th  site. A l l but one of the ten dominant species were diatoms. The blue-green alga Pseusdo anabaena was also abundant at all four transects. The 10 most abundant species are shown in Figure 26 for Hayward Reservoir and Figure 27 for Stave Reservoir.  Chapter 3  45  CM I•tf  LU ICO  O O  CO O) CO  Is  o E cn  co  LO  CO N  CM  i  CO LO  •tf LO o I  LO I  CM O  o o o o  co CM CO CM  co  o o o o  o o o o  o o  00 LO  0  0  0  0  co o7 N 00 O LO CM  •tf  o o o o  •tf  •53  o o  10  s  N  o o o o  CO CO  CM  0  •tf ICO  w  o o o o  co  s  •tf co co co CO  10  10  o> •tf CM  o  o  o  N  00  00 •tf  co  CO CM  LO CO CO  ICM CO s  co  0  s 3  co  9  —  CC IB LU  LU  <  > o  CO  LU CC  I co o j o E co o E 0  • < z  3 CC I-  a 5  o  CO ca  o o o  I -  LO CO CM  I  LU  z  C5 >-  CO CO  &oCO  _N  §'  •tf o  O  o o ol  •tf  LO co  LO LO CM CO  s  o  s  3  Ico O) LO LO  s  <  < _i  LU _l  0 •tf LO  < CO  LU <  co Z < o CC o ca 3  CC  o  CO  co  ILU  3  < X CL CC o CO z o I0. LU s < <_ z |L0 o I CO  CC 3 CO  LU  s  CM CM •tf 0 ) LO CO CO  •tf  o o o  o  CO CO  o  0  CO CO CO  •tf co  co O) CD  o LO cn  CO  s  CM ICM  •tf  CO  LO CM CO  CM  CO co LO CO CO CO o CO CM  cn  •tf N cn co LO cn  CO CM  cn  LU  < ca <  z <  -I LU  _o s LO  •<t  o o  CM  LO CM LO  ICM CM Is  s  CO  co LO o  •tf  o o cn o •tf  co co CM 10  Is  o  |co  CO  CO  E  CD CO  CO  co  CO a LU c a. co I O CO LU LU CL. z CO h-  Z <  I Chapter 3  Q  a LU < a. o o o LU i o CC  cn Iw  CO  LO I  <  CD D)  ~S CO  < _l _l LU I-  o -I o >o  co  •g 'ra 3  Q. CO I  <  £ LU  Z  o x a E o|  a  co a. a CO  < lO  z  3 LU  CO LU  z o LU o  E  l£ £  2 c H 2 CO o  5  D  Z <  LL  LU  ~  O CO  !s I  46  Hayward Dominant S p e c i e s EUNOTIA tenelten  ANABAENA  CYCLOTELLA stelligera  Figure 26. Ten dominant species in Hayward Reservoir by abundance (cells/cm ).  The three sites from Stave Reservoir were averaged into one group for biovolume assessments. Figure 28 shows the ten most abundant species by biovolume. Species composition based on biovolume estimates resulted in a different grouping of dominant species than for cell abundance. Achnanthes, Cyclotella and Pseudo ananbaena dominated algal abundance in Stave and Hayward, whereas, Fragilaria acus and Tabellaria fenestra dominated biovolume species composition. While Achnanthes, Cyclotella and Pseusdo anabaena are present in large numbers at the two sites, they do not account for a large volume due to their small size. By biovolume, Fragilaria  and Tabellaria account for 41 and 44% of the ten dominant species in Stave and  Hayward, respectively. Generally species composition based on biovolume was slightly more variable than species composition based on measures of abundance in the two systems.  Chapter 3  47  Stave South dominant species FRAGILARIA  • 1  acus  • 2  ANEMOENBSsp^ EUNTOTIA tenella  ACHNANTHES  .3  spp  • 4  GOMFHONF_w\  .5  sp  • 6  EUNOTIA lunarisl  PSUEDO  ACHNANTHES^  • 7  ANABAENA  • 8  rrinutissima NAVICULA spp  • 9  CYCLOTELLA  • 10  stelligera  Stave North dominant snecies FRUSTRULIA rhorrtookJes  •1 • 2  FRAGILARIA acus  • 3  PSEUDO f ANABAENA  EUNTOTIA tenella AI_MOENBSsp  • 4 • 5  GOM^HONEMA sp  • 6 • 7  CY CLOTH-LA ACHNANTHES  stelligera  • 8  spp NAVCULA spp  • 9  ACHNANTHES  • 10  rrinutissima  Stave West dominant species •1 • 2  EUNOTIA tenella ANEMOENEB s p , FRAGLAR1A acus  s  • 3  ACHNANTHES spp  • 4 • 5  BJNOTIA lunaris, GOMPHONEMA sp NAVCULA spp PSEUDO ANABAENA  • 6 CYCLOTELLA  • 7  stelligera  • 8 • 9  ACHNANTHES rrinutissima  • 10  Figure 27: Ten most dominant species by abundance (cells/cm ) for transects in Stave Reservoir. 2  Chapter 3  48  Hayward Dominant S p e c i e s : Biovolume (mm ) 3  FRUSTRULIA  CY MB EL LA  rhomboides  RNNULARIA sp  Figure 28. Ten most dominant species by biovolume (mm ) for Stave (all transects) and 3  Hayward reservoirs. Chapter 3  49  In addition to the evaluation of dominant species, a preliminary analysis was conducted on Stave and Hayward species composition data from 8 and 16 m, using nonmetric multidimensional scaling (NMS). In this study, N M S was used to examine the relationship between samples and species abundance  in Stave and Hayward reservoirs. N M S iteratively searches for a  multidimensional solution with the least difference or stress between distance or dissimilarity in the original data set and those generated for the same samples in the reduced ordination space of the final solution (Page, 2003). McCune and Grace (2002) consider N M S the ordination method of choice in ecology because of mathematical robustness and biological meaningfulness.  Results of the N M S analysis examined the relationship between species abundance and variables of sample year, transect (Stave N , S, W and Hayward), season (spring, summer, fall, winter) and depth (8 m and 16 m). Stress (difference) in the analysis was relatively low (7.82) indicating that the difference between the original data and the final solution was small and the results are relatively reliable for interpretation. The cumulative explanation of variance for axis 1 was r = 2  0.77 and for axis 2 was r = 0.86. Scatter plots of the four variables indicated that there are no 2  patterns in species composition (by abundance) based on year, site, transect or season, while there is a clear grouping of species by depth (8 m and 16 m samples) (Figure 29).  The results of the species composition analysis indicate that species composition and dominance are similar between Stave and Hayward. There were no distinguishable patterns or groupings of species relative to transect, year or season indicating that these factors were not important determinants of species groupings. By depth species composition resulted in an obvious grouping of species by abundance. Similar results were achieved using biovolume data in the same analysis with PC-ORD suggesting that comparisons between Stave and Hayward reservoirs are valid since the resulting differences are probably not due to compositional differences. In turn, the patterns of species composition either do not exist or are not evident by this analysis. It is clear from this analysis that there are notable differences in species composition by depth that warrant further study. As well, there may be opportunities to expand the analysis of site and seasonal variables with additional environmental data that would allow a more complete analysis.  Chapter 3  50  Transect  Year •  2000 A 2001 • 2002  • 1 A 2  ft A  •  • 3 • 4 CM CO  '5 <  • •-  •  l=Stv. North, 2=Stv. South, 3=Stv. West 4=Hayward Axis 1  Axis 1  Depth  Season  ft  • 1 A 2  ft  • 8 A  16  • 3 • 4 CM (0  CM (0  '5 <  "x  <  - A *  •  !-A A  A ^  A  A A  A A  l=Spring 2=Summer 3=Fall 4=Winter Axis 1  \ A  A  Axis 1  Figure 29. Relationship of species composition samples to year, transect, season and depth using nonmetric multidimensional scaling. Stress of solution was 7.83 and instability was 0.0007. Only depth indicates a compositional grouping of species.  Chapter 3  51  3.2.3  Estimates of Littoral Periphyton Production: Accrual and Chl-a  Annual accrual over depth Average annual accrual at depth represents periphyton biomass growing in the littoral zone of Stave and Hayward reservoirs, and is the primary data used in this thesis to estimate littoral production (Figure 30). In general, Hayward Reservoir exhibited higher peak accrual with depth than Stave Reservoir. In Hayward Reservoir peak accrual generally occurred approximately 4 m below full pool (plate 2), while peak accrual among transects in Stave Reservoir were deeper and more variable occurring 6-10 m below full pool. In 2000, peak accrual in Hayward was 141.48 mgC/m /day and occurred 4 m below full pool 2  (approximately 2 m below the average surface water elevation). In contrast, Stave north and Stave south transects exhibited peaks of 63.91 and 48.98 mgC/m /day occurring 8 m below full 2  pool (Figure 30a).  In 2001, accrual in Hayward was less than in 2000 at 109.12 mgC/m /day, but was still higher 2  than all sites on Stave. Stave north exhibited highest accrual values that occurred in Stave in the duration of the study at values of 77.60 mgC/m /day. Stave south exhibited biomass accrual 2  similar to values for Stave in 2000 (53.62 mgC/m /day). Stave west, established early in 2001, 2  exhibited peak accrual of only 27.19 mgC/m /day. Peak accrual at all sites in Stave were at least 2  2 m below the peak in Hayward, occurring 6-8 m below full pool in Stave (Figure 30b).  In 2002, accrual was the lowest of all sampling years at all 4 transects, but the patterns remained similar to previous years with Hayward exhibiting the highest accrual and Stave transects substantially lower accrual and occurring at deeper depths (Figure 30c). In all years, accrual was less than 10 mgC/m /day at 14 m and approached zero values at the deepest point of 2  measurement at all transects.  Chapter 3  52  Figure 30. Annual Littoral Accrual over depth a) 2000, b) 2001, c) 2002.  Seasonal Accrual Seasonal accrual in Stave and Hayward reservoirs was estimated as an average based on all three years of sampling (Figure 31). The three transects on Stave were combined as an average to provide a single estimate of seasonal accrual with depth for Stave Reservoir. Hayward Reservoir exhibited higher seasonal accrual with depth in all seasons relative to Stave Reservoir. Highest seasonal accrual in Hayward Reservoir occurred in spring at the rate of 150.59 mgC/m /day, summer and fall exhibited similar accrual at the rate of 93.58 and 109.99 mgC/m /day, respectively. Lowest seasonal accrual occurred in winter. In Stave Reservoir, seasonal accrual exhibited a different partem than Hayward Reservoirs. Peak production occurred in fall at a rate of 64.95 mgC/m /day. Spring and summer production rates in Stave were similar to each other at rates of 31.67 and 32.45 mgC/m /day, respectively. Similar to Hayward, lowest accrual rates in Stave occurred in winter. Also notable, average seasonal water level fluctuations on Stave were only 2 m, ranging from approximately 80 m a.s.l. to 78 m a.s.l. Daily water level fluctuations are smoothed out by averaging over season and sample period, which has been done to facilitate analysis for all aspects of this thesis. This limitation was particularly evident when averaged over season, where  Chapter 3  53  the change in water elevation appears to be 2 m when in reality water levels varied by nearly 6 m over the course of the year.  This is important when interpreting the impact of water level  fluctuation on periphyton growth in this study.  2000-2002 Seasonal Production (mgC/m /day) 2  Spring 0  100  Summer 200  0  100  Fall 200  0  100  Winter 200  rj  100  200  Figure 3 1 . Stave/Hayward Seasonal Production (2000-2002 average). Note: Stave profiles reflect average water levels in each season.  Annual production Estimates of annual production (biomass accrual) in Stave and Hayward reservoirs were calculated by integrating sample period accrual (mgC/m /day) measurements collected from each transect for the duration of this study. Area under the accrual curve was calculated using a weighting factor to compensate for variability in the depth between sampling blocks. In reservoirs with fluctuating water management regimes, accrual curves shift vertically up and down as surface water elevations change. For this study it was assumed that by sampling to 20 m depth, the periphyton sample collected always included the base of the accrual curve where it approaches zero growth, regardless of surface water elevation. Accrual values at 20 m depth were generally in thousandths of a gram and were commonly zero values, which was within  Chapter 3  54  experimental error for measurement of accrual in this study. Also supporting this assumption, light (measured as PAR) was found to be extinguished or very low at 20 m depth regardless of the water surface elevation, indicating that growth at 20 m is minimal. Stockner et al. (1971) report that benthic algal growth in most lakes in the Experimental Lakes Area, Ontario was negligible at depths greater than 10 m, depending on light penetration, confirming that the compensation depth (i.e. photosynthesis = respiration) defines the level at which growth is negligible.  Comparing average littoral production among years (Figure 32), the first year of sampling (2000) exhibited maximum production (122.16 mgC/m /day) in August on Hayward Reservoir. In 2  2000, Hayward experienced two peaks in production, one in May followed by the maxima in August, while Stave Reservoir exhibited only one peak that was drawn out from August through September. Generally peak production in Stave Reservoir tended to be of lower magnitude and for longer duration than Hayward. Average 2000 production in Stave was estimated to be 5.3 9  9  gC/m /year and in Hayward was 12.7 gC/m /year.  Littoral Accrual Production  1 § o 5 " O  Feb-00 Stave North Stave West  Jun-00  Oct-00  Feb-01  Jun-01  Stave South Hayward  Oct-01  Feb-02  Jun-02  Oct-02  Time  Figure 32. Average Littoral Production 2000-2001. (3 Stave transects and Hayward) In 2001, Hayward production did not exhibit as large a peak as in 2000. Peaks occurred in May, and August in the 60-80 mgC/m /day range. In Stave Reservoir two peaks were evident in 2001, one occurring in May and a second in September at Stave south transect and in November on the north transect in Stave. Also notable, the North transect exhibited maximum production that occurred on Stave over the duration of the study with a value of 62 mgC/m /day. Average 2001 9  9  production in Stave was estimated to be 6.6 gC/m /year and in Hayward was 12.0 gC/m7year. Chapter 3  55  In 2002, production was lower in both Stave and Hayward than in the other two years. In addition, neither reservoir exhibited the strong summer low in production that was evident in both 2000 and 2001, which appears to be the result of higher than previous summer production values. Otherwise production in 2002 followed a similar pattern to that exhibited in 2000. Stave Reservoir exhibited a small peak in April and a second peak in September, but was generally lower than in the other two sampling years. In 2002 average production in Stave was estimated to be 4.7 gC/m /year and in Hayward was 12.1 gC/m7year.  Chlorophyll-a Chlorophyll-tf concentrations were also measured as part of the primary data of this thesis, and represent an estimate of living algal biomass in these systems (Figure 33). In 2000, peak chlorophyll concentrations at all sites (Stave north, south and Hayward) were comparable ranging from 4.99 mg/m to 5.16 mg/m . Hayward exhibited two peaks in chlorophyll that 2  2  occurred approximately 4 m and 8 m below full pool water levels. Transects in Stave exhibited only single peaks in chlorophyll occurring at 8 and 10 m below full pool water elevations.  a  )  2002 Chlorophyll a  2000 Chlorophyll a  Chi a (mg/m2)  Chi a (mg/m2) 0.0  2.0  4.0  6.0  0.0 OH  2.0  4.0  1  Stave North Stave West  1  6.0 1  - • — Stave South • Hayward  Figure 33. Annual Chlorophyll Concentration over depth a) 2000, b) 2001, c) 2002.  Chapter 3  56  In 2001, peak chlorophyll concentrations at depth were lower at all sites. However, of the four transects, Hayward Reservoir still exhibited the highest concentration with a single peak of 3.85 mg/m  occurring 6 m below full pool. Concentrations in Stave north, south and west were  similar to one another ranging from 2.48-2.76 mg/m occurring 6-8 m below full pool.  The lowest chlorophyll concentrations of the study occurred in 2002 in Stave Reservoir. Peak concentrations in Stave Reservoir ranged from 0.78 -1.75 mg/m . The peak in Stave north 2  occurred 6-10 m below full pool, while the peak in Stave south occurred 4-6 m below full pool elevation. Stave west exhibited the lowest concentrations with no evident peak. Peak chlorophyll in Hayward Reservoir remained high at concentrations of 4.17 mg/m . Also notable peak concentrations in Hayward occurred consistently 4 m below full pool.  Littoral chlorophyll biomass differences over the complete study time series were less pronounced on both Stave and Hayward reservoirs (Figure 34) compared to measurements of accrual (Figure 32). Still, Hayward chl-a concentrations remained consistently higher than chl-a concentrations estimated for Stave Reservoir. Highest chlorophyll concentrations occurred in 2000 on both Hayward and Stave exhibiting a maximum of 4.80 mg/m on Hayward and 3.95 mg/m  on Stave south. Hayward exhibited a dual peak in 2000, while the two transects on Stave  exhibited only a single peak.  Littoral Chlorophyll Biomass  Feb-00  Jun-00  - Stave North Stave West  Oct-00  Feb-01  Jun-01  Oct-01  Feb-02  Jun-02  Oct-02  Stave South Hayward  Figure 34. Stave/Hayward Littoral Chlorophyll Production.  Chapter 3  57  In 2001, chlorophyll concentrations for all 4 transects were lower than in 2000. A l l sites exhibited a dual peak, the first occurring in late spring (May) and the second in fall. The fall peak in Stave Reservoir occurred later than the peak in Hayward. Peak chlorophyll 2  2  concentrations in 2001 were 3.66 mg/m in Hayward and 2.47 mg/m in Stave north. Stave west exhibited lower concentrations than at all other sites. Lowest chlorophyll concentrations for the duration of the study occurred in 2002, which was also reflected in the annual results for estimates of accrual. Average annual chlorophyll in Stave was estimated to be 6.68 mg/m /year 2  •  m  7  m  while in Hayward values were 14.22 mg/m /year. A regression analysis was conducted between chlorophyll a and accrual measurements collected for this study with the aim of determining the relationship between these two measurements of biomass (Figure 35). The analysis found that Chi a and accrual measurements were not overly well correlated with an r -value of 0.40.  Scatter Plot Chi a vs. AFDW 400 | £ 350  0.0  *  2.0  4.0  6.0  8.0  10.0  Chi a (mg/m2)  12.0 y  14.0 =  1 2  -  6 1 x +  16.0 -  6 3 8 2  R = 0.3989 2  Figure 35. Scatterplot of Chlorophyll and Accrual data for Stave and Hayward reservoirs.  3.2.4 Stave/Hayward: A Statistical Comparison of Littoral Production A univariate (two-way) analysis of variance ( A N O V A ) was conducted on average periphyton production data to evaluate differences in littoral accrual between Stave and Hayward reservoirs testing for a lake effect and a seasonal effect.  Chapter 3  58  Assumptions  In order to conduct an analysis of variance on production measurements from Stave and Hayward reservoirs, the four assumptions of ANOVA.were examined. A density of sampling sites (transects) that was approximately similar between the two reservoirs was randomly established to achieve a normal distribution from both systems. In total four sampling transects, three in Stave Reservoir and one in Hayward Reservoir were established to measure periphyton accrual. Transects were established at locations within each reservoir that were considered to have similar gradient, exposure, substrate and access. Each of the four transects was assumed to be representative of the periphyton communities within the littoral zones of each reservoir.  Measurements of littoral accrual were login transformed to improve normality. The logarithmic transformation greatly improved the approximation of a normal distribution for both Stave and Hayward Reservoir production data. It was also noted that A N O V A is relatively robust with regard to meeting the assumption of a normal, bell-shaped distribution of data. Stave and Hayward transformed distributions are viewed to meet the assumption of normality relatively well.  To meet the assumption of equivalent variances for A N O V A , two tests were conducted. Levene's test of equality of error variances yielded a p-value of 0.55 indicating that populations being examined have equal variance (P<0.05 is taken to mean that populations do not have equal variance (Townsend, 2002)). Additionally residual plots were plotted which also indicate that the variances were similar between the two sites.  The most critical assumption to meet in order to conduct an A N O V A was that of independent measures. Samples are considered independent when individual samples within samples behave the same way, regardless of which other samples were included in the sample. One could argue in the case of this study, that because samples were taken from within the same reservoir that they are pseudoreplicates, but of course it is impractical to replicate a reservoir. Within the practical limitations of this study, it can be argued that the three transect locations in Stave Reservoir are sufficiently removed from one another to be considered independent. Although Hayward did not have replicated transects within the reservoir, it is assumed that the collection site represented the whole of the system.  Chapter 3  59  Within individual transects, sampling sites were located at 2 m vertical intervals. Given the natural gradient of the littoral zone in both Stave and Hayward reservoirs, it is also arguable that individual sampling blocks were sufficiently separated from one another to be considered independent. Sampling blocks are separated by enough distance, especially when considering that the base line measurements of the study is attached microscopic algae, that the water column surrounding each site is sufficiently different to not impact how any other site behaves.  In addition each time plates were collected for sample removal, periphyton was removed from the plate such that growth could be considered "new" for each sampling period. None-the-less, it remains that sampling transects were interconnected by being part of the same environment. Although the sampling regimen may not have been statistically ideal, lines of evidence are available that indicate that the sites were sufficiently removed from one another to be considered independent when given consideration at a practical level.  Related to the idea of independent measures is the issue of repeated measures. Repeated measures, or sampling repeatedly from the same location, are another cause of nonindependence. It could be argued that samples were collected from the same reservoir and furthermore, because the water from Stave drains into Hayward that samples from both reservoirs are repeated measures. In this study, measurements were made over three year time period and investigated differences between seasons. Measures of littoral periphyton accrual were averaged over depth to provide an estimate of littoral production for each season (i.e. ten data points into one measure of production). By summarizing individual measurements at each of 10 (Stave) and 8 (Hayward) depths it is possible to use A N O V A to compare mean production in each season, despite the issue of repeated measures (Thomson et al., 1996; Townsend, 2002). Additionally seasonal data were averaged over the three years of sampling and a similar analysis conducted which yielded similar results, indicating that repeated measures in this experiment have been appropriately considered.  A N O V A Results The results of the A N O V A are presented in Table 4. The lack of interaction between lake and season indicates that factors of the A N O V A can be individually explored. The results of the  Chapter 3  60  A N O V A indicate that there was a lake effect present, which is to say there was a significant difference between average littoral production between Stave and Hayward reservoirs. From the graph of the interaction between Stave and Hayward it is apparent that production in Hayward was greater in all seasons than production in Stave (Figure 36). Table 4.  Stave/Hayward A N O V A Results  Tests of Between-Subjects Effects Dependent Variable: Log 10 transformed data Type III Sum of Squares 4.871  Source Corrected Model Intercept  69.708  7  Mean Square .696  F 9.957  Sig. .000  1  69.708  997.479  .000  df 3  LAKE  1.097  1  1.097  15.700  .000  SEASON  2.115  3  .705  10.090  .000  .220  3  .073  1.049  .377  Error  4.542  65  .070  Total  131.694  73  LAKE * S E A S O N  Corrected Total  9.413  72  a. R Squared = .517 (Adjusted R Squared = .465)  Interaction Plot of Stave/Hayward Littoral Production 1.8  o  —  O  —  5  -if  E  Q U!  o  bo  -o  d  Hayward  Figure 36. Interaction plot showing factors of lake and season for Stave and Hayward. No significant interaction is present and Hayward Reservoir production was higher than Stave Reservoir production (periphyton accrual) in all seasons.  Chapter 3  61  The A N O V A also indicates that there was a significant seasonal effect on production in the two systems. Post hoc analyses of seasonal effects, using Tukey's test, indicated that production in winter was significantly different (lower) than in other seasons (Table 5).  Table 5.  Results of Tukey's Test Comparing Seasonal Differences. * Indicates there is a significant difference between winter and all other seasons. Multiple Comparisons Dependent Variable: LoglO transformed data Tukey HSD  (I) Season Fall  Winter  Spring  Summer  (J) Season Winter Spring Summer Fall Spring Summer Fall Winter Summer Fall Winter Spring  Mean Difference d-J) .6899* .2127 .0417 -.6899* -.4771* -.6482* -.2127 .4771* -.1710 -.0417 .6482* .1710  Std. Error .12781 .09508 .09361 .12781 .11257 .11132 .09508 .11257 .07140 .09361 .11132 .07140  Sig. .000 .124 .970 .000 .000 .000 .124 .000 .088 .970 .000 .088  95% Contidence Interval Lower Bound Upper Bound .3529 1.0269 .4634 -.0380 -.2051 .2885 -.3529 -1.0269 -.7740 -.1803 -.9417 -.3546 -.4634 .0380 .1803 .7740 -.3593 .0172 -.2885 .2051 .9417 .3546 -.0172 .3593  Based on observed means. *• The mean difference is significant at the .05 level.  3.2.5  Discussion of Littoral Production In Stave and Hayward reservoirs  Measurements of accrual over depth reveal that periphyton growth in Hayward Reservoir is more similar to a stable lake habitat, with peak biomass occurring approximately 2 m below the average water surface elevation, which is likely in the zone of optimum light, temperature and nutrients. Although it has been shown that available nutrients in reservoirs that are located down stream of one another generally declines (Thornton, 1990), the higher accrual is not unexpected given that Hayward is a run-of-the river reservoir with relatively short residence time and stable water levels (Allan, 1995). Accrual over depth in Stave Reservoir exhibited lower biomass with more variable peak levels. This is also an expected result that is attributable to fluctuating water levels that disrupt cycles of growth by impacting the zone of optimal conditions for growth.  Chapter 3  62  It is of interest to assess littoral accrual relative to water level variation over the study period. Water levels in Hayward were constant (with the exception of July 2001), while levels in Stave typically cycle from a winter low to a summer high, decreasing through fall, with an overall water level difference of approximately 6 m. Comparing water levels to the accrual curves over depth (Figure 37), the stable water regime in Hayward offers one explanation for higher growth occurring closer to the water surface.  Average Sample Period Water Level Fluctuation 2000-2003  82 80 . Ill 78 ro -S 76  43  1>  j'  m  * v  41 40  * 2000  2000 Littoral Accrual  mgC/mVday 0  » --•—•—A~m\ 11- 42  25 50 75 100 125150  2002  2001  2001 Littoral Accrual  2002 Littoral Accrual mgC/m /day  mgC/m /day  2  2  25 50 75 100 125 150  0  25 50 75 100 125 150  -Stave North Stave West  -Stave South Hayward  Figure 37. Comparison of Water Level and Accrual Measurements Over Time.  In 2000, Stave and Hayward both exhibited typical water level regimes; Hayward was constant, while Stave underwent the dual-cycle of drawdown described above. Accordingly, Hayward accrual is much higher than that in Stave and peak accrual in Stave occurs at lower depths in response to the downward shift of the littoral zone through the most productive growth period of the year. Compared to the stable system, accrual values in Stave are diminished. Stave biomass Chapter 3  63  accrual is only 38% of accrual in Hayward, which is largely due to organisms having to respond to desiccation or flooding (i.e. altered conditions of growth).  In 2001, Hayward Reservoir undergoes a short but intense period of drawdown, providing a unique opportunity to examine the effect of drawdown, which theoretically should result in a •y  decrease in littoral accrual. Hayward accrual in 2001 is 23% (30 mgC/m /day) lower than in 2000. Additionally, summer time solar radiation in 2001 was higher, which would be expected to yield higher accrual. Examining sample period growth curves for Hayward in 2001 indicate that for the three sample periods prior to the drawdown (February (T7), April (T8), and May (T9)), Hayward exhibited •y  peak accrual values up to 189.90 mgC/m /day, but following the July drawdown levels dropped to 34.77 mgC/m /day (T10) the lowest values o f the year (Figure 38). The post drawdown 2  accrual levels potentially indicate the degree to which periphyton growth is set back when water levels fluctuate and desiccation occurs.  Hayward Sample Period Accrual (mg C/m2/day)  (mg C/m2/day) 100  200 300 400  2002  2001  2000 0  100 200 300 400  (mg C/m2/day) 100  200 300 400  0  77 -•—T8 T9 - X — T10 - * - T11 -•—T12 -I— T13  -•—TO -•—T1 T2 HX—T3 -*— T4 -•—T5 -I—T6  Figure 38. Sample period accrual in Hayward Reservoir 2000,2001,2002.  Chapter 3  64  In 2002, peak littoral accrual in Hayward was of similar magnitude to 2001 despite stable water levels, which may be indicative of the resilience of the 2001 communities ability to recover and yield peak accrual post drawdown. It may also indicate that factors other than water level fluctuation, such as solar radiation and nutrient dynamics, may play an important role in driving periphyton growth. Stave 2002 accrual values were lower than in other years of sampling, which may be the result of the higher water levels at the start of the year and the relatively sharp decline in water level in the fall.  Climatic variables and nutrient dynamics are complex and difficult to evaluate because more than one factor can vary at the same time. In this study most of the physico-chemical variables were not measured frequently enough to fully explore what was occurring in these systems. One variable that stood out and may offer explanation to the lower accrual in 2002 relative to 2000 was the unusually high Hayward light extinction coefficient (0.70) measured in February and April of 2002, which was about 45% higher than the average (0.39 when excluding the spring 2002 values). While it is expected that there would be a negative correlation between light extinction and benthic algal biomass (Hansson, 1988), Hansson (1992) found that this relationship only held true at extinction coefficients > 1.5m in oligotophic lakes in Sweden and Antarctica. Below this level biomass increased, which Hansson argued was owing to optimal light and nutrient conditions. In the case of Hayward Reservoir it is difficult to draw firm conclusions regarding the impact of individual variables, but it in reasonable to think that lower light levels may explain, at least in part, the lower and more variable accrual values exhibited in 2002.  Production in Hayward Reservoir was higher than production estimated in Stave Reservoir for each of the three sampling years. Peak production in Hayward was generally in the 60-80 mgC/m /day range, while Stave production was lower and more variable (3 transects), with peak production values in the range of 20-40 mgC/m /day. Estimates of production in Stave Reservoir 2  are similar to those reported by Shortreed et al. (1984) where periphyton accrual values ranging from 1-47 mgC/m /day in 21 B C lakes, some which were fertilized and others that were not. 2  Hayward Reservoir production values appear to be higher than what is typical in oligotrophic coastal lakes. Stockner and Armstrong (1971) report periphyton growth rates taken at 1 m depth  Chapter 3  65  in four lakes in northern Ontario ranging from 27 mgC/m /day during the initial colonization 2  period to a maximum of 250 mgC/m2/day during periods of maximum production. Fox et al. (1969) reported average periphyton growth rates of 67 mg organic/m /day in Stony Point Bay on 2  Lake Superior. These studies suggest that production in both Stave and Hayward reservoirs are within a similar range as other oligotophic temperate lakes.  Although the timing of peak production was variable in both reservoirs, it is notable that when comparing peaks annually between Stave and Hayward, Hayward peaks consistently occurred earlier than those on Stave. Peak values in Hayward generally occurred in early May and again in August, while production peaks in Stave lagged behind occurring late in May (or early June) and September. This result is explained by the slightly warmer temperatures and by the riverine flow characteristics, which typify Hayward Reservoir. Allan (1995) supports this result suggesting that production in flowing environments, where nutrients continually spiral, tend to result in higher production earlier in the season than in other freshwater environments. In contrast, Stave Reservoir takes longer for production to peak due to its large size, cooler water and due to the recycling of nutrients that are typical in this system.  Also of interest are low production values that occurred consistently both in Hayward and Stave in July, between peak production periods. Bimodal annual production in both Stave and Hayward reservoirs is explained in part by increased chlorophyll values and low N:P ratios measured during the summer and early fall months. It is likely that periphyton growth is limited in summer months when increased pelagic biological activity utilizes available ambient nutrients. The secondary and often higher peak occurs when N:P ratios increase again with fall rains, autumnal turnover and allochthonous inputs from surrounding streams and tributaries (autumnal pulse) (Cummins et al., 1989).  Evaluated seasonally, Hayward Reservoir exhibited peak values in spring. Periphyton grow throughout the year in coastal temperate ecosystems, but peak accrual in spring is common in lakes (Wetzel 1983b, Cattaneo, 1983) and streams (Power 1990a, Lamberti et al., 1989). Lamberti (1996) explains that conditions during spring such as high irradiance and nutrient loading often yield maximum primary production in spring.  In Stave, the fall maxima in  seasonal accrual are not as easily explained, but it was observed that water levels in Stave were  Chapter 3  66  highest in summer and into the fall but were lowest from winter into spring, offering one possible explanation for the different pattern of accrual. A second explanation is that accrual differences are the result of mixing periods (turnover.) that transport nutrients to the littoral community.  Average annual production values estimated based on measurements of accrual in Stave and Hayward for 2000, 2001 and 2002 are presented in Table 6. Averaged over the three year sampling period average annual production in Stave was estimated to be 5.3 gC/m /year and in 2  Hayward 12.3 gC/m /year. Table 6 also illustrates that between years littoral production was 2  relatively consistent in both reservoirs.  Table 6. Year  Stave/Hayward Average Annual Production Stave Average Littoral Production (gC/m /year)  Hayward Average Littoral Production (gC/m /year)  5.3 6.6 4.7 5.3  12.7 12.0 12.1 12.3  2  2000 2001 2002 3-year average  2  Chlorophyll-a Generally chlorophyll concentrations at depth compared between Stave and Hayward followed a similar pattern to measurements of accrual taken in this study. Similar to measurements of accrual, chlorophyll values in Hayward were generally higher than those measured in Stave, but the differences were not as pronounced as for accrual. The depths at which peak chlorophyll values occurred were more variable than was the case with accrual, especially among transects in Stave Reservoir, making generalizations and interpretation of patterns more difficult. Despite the more variable results, peak chlorophyll concentrations in Stave Reservoir were typically deeper and of longer duration than peaks occurring in Hayward which, again was similar to patterns of accrual at the two locations.  Chapter 3  67  Also of interest was that chlorophyll concentrations remained at detectable levels to at least 14 m at all sites, in all years for both reservoirs. Hayward concentrations followed a more consistent pattern dropping to about 0.5 mg/m by 14 m and nearing zero by 16 m. Chlorophyll 2  concentrations at all three transects in Stave Reservoir during 2000 and 2001 were approximately 1 mg/m at 14 m and neared zero by 20 m depth. In 2001, Stave north exhibited concentrations of 2 mg/m at 14 m, but also declined to zero at 20 m. This result is expected given that 1% light levels measured in this study typically occurred between 12-14 m depth. These results are in contrast to measurements of accrual with depth for this study that found accrual to decline towards zero at shallower depths than chl-a. Accrual values were consistently less than 10 mgC/m /day by 14 m and near zero by 16 m at all sites for all years.  Although a closer correlation may have been the expected result of the regression analysis between Chl-a and accrual, there are several possible factors that offer explanation for the low r -value (0.40). In this study, sampling was on average conducted once every 6-weeks. 2  Periphyton growth follows a pattern with a lag time while the periphyton becomes established, then a period of exponential growth, gradually levelling off to an asymptote. Following the initial growth phase (i.e. > 45 days), it is common to see sloughing, grazing and self-shading playing greater roles in reducing biomass. In this study, it is possible that the sampling frequency was too long. In a study conducted by Stockner and Armstrong (1971) it was shown that the most rapid period of periphyton growth occurs between 30 and 57 days, beyond which growth begins to decline (Figure 39). In this study, the average length of time plates were submerged was 42 days, not including the winter months when plates were left submerged through the entire season (3+ months). It is likely that sampling, particularly during winter, was not frequent enough to avoid sloughing and grazing effects. No measurements were taken to quantify biomass losses during this period. A l l comparisons of production between Stave and Hayward will focus on measurements of accrual from this point forward.  Chapter 3  68  Figure 39. Changes in total periphyton biomass and diatom density (a) Initial colonization period; and (b) period of most rapid growth between 30-57 days accumulation with growth decreasing after this time due to effects of grazing and sloughing (after Stockner and Armstrong 1971). A N O V A Discussion An analysis of variance on transformed sample period production found that there is a significant difference in the average production between Stave and Hayward. There was no interaction between Stave and Hayward indicated by the A N O V A . Post Hoc multiple comparison test (Tukey) of season for both reservoirs indicate that winter in significantly different from all other seasons. Sample period production measurements were lower in winter months, which were expected given the colder, darker climatic condition. Although sample period measures were normalized by the number of days per sample period, it is apparent that winter measurements may have been influenced by other processes related to the extended sample period length (3+ months) in winter. During the winter months, processes such as sloughing, grazing and selfshading, which are well documented in the literature but were not accounted for in this study, have high potential to impact growth given the duration of our sample period. On this basis, the multiple comparison analysis of season will not be taken further.  Chapter 3  69  3.4  Stave/Hayward Comparison of Littoral to Pelagic Production  3.4.1  Pelagic Primary Production  Pelagic production of carbon (C) in Stave and Hayward reservoirs followed a similar seasonal trend with little variability between systems (Figure 40).  50.0 >; 45.0 ^ 40.0  April I Stave  May  July  June  August  September October November  2003  • Hayward  Figure 40. Estimates of daily C production ( C method) in Stave and Hayward reservoirs in 2003. l4  The pattern was bimodal with low periods in April/May and October/November and a late spring peak (June) followed by lower July values in both systems, then rising to maximum values in August in Hayward and both August and September in Stave. In both reservoirs production values ranged from <5 to >40 mgC/m /day and averaged 16.4 mgC/m /day in Stave and 14.9 2  2  mgC/m /day in Hayward. Average integral Chl-a values were also slightly higher in Stave, but 2  assimilation rates (AN) were the same in both systems. Table 7 below summarizes pelagic production, chlorophyll and production efficiency in both reservoirs.  Chapter 3  70  Summary of Stave and Hayward reservoirs pelagic C production, chlorophyll, and production efficiency (AN).  Table 7.  Stave  Hayward  Date  mgC/m /day  mgChl/m  14-Apr 26-May 18-Jun 18-Jul 13-Aug 12-Sep 30-Sep 29-Oct 25-Nov  2.3 5.9 19.8 10.7 35.6 34.7 34.4 2.2 1.9  AVERAGE  16.4  3.4.2  2  AN/day  mgC/m /day  mgChl/m  8.7 10.3 11.2 14.8 11.3 6.5 6.3 3.4 2.5  0.26 0.57 1.76 0.72 3.1 5.33 5.46 0.65 0.76  4.7 9.6 23.5 10.9 44.9 19.6 17.9 1.4 2.4  7.1 7.3 19.5 8.7 8.5 3.6 5.2 3.3 1.8  0.66 1.31 1.2 1.25 5.28 5.44 3.44 0.42 1.33  8.3  2.1  14.9  7.2  2.26  2  2  2  AN/day  Contribution of littoral production to Stave/Hayward Ecosystems  As part of the littoral zone monitor for Stave Reservoir (SLZM), littoral areas were calculated by B C Hydro based on the limit of the photic zone for biotic growth (i.e. production = respiration), which was arbitrarily defined as a 6 m vertical elevation band based on estimates from previous experience. While this estimate was the best available approximation of littoral areas in Stave and Hayward reservoirs, it is acknowledged that the littoral areas defined in this study extended to 20 m and 16 m respectively. As mentioned, the definition of littoral areas for this thesis, were defined to account for water level fluctuation. The definition of littoral areas as a six meter elevation band include peak production, but may underestimate littoral production as defined in the thesis, therefore this analysis serves only as a preliminary estimate of total annual littoral production. These estimates of annual littoral production can then be compared to pelagic carbon estimates that were measured as part of the S L Z M program.  Table 8 displays littoral C  production estimates, estimates of pelagic production and percentage contribution of littoral production to total aquatic production.  Table 8.  Stave Hayward  Chapter 3  Littoral and pelagic production estimates for Stave and Hayward reservoirs Littoral (tC/year)  Pelagic (tC/year)  % littoral contribution (tC/year)  94 6  2108 12  4 50  71  3.5  Discussion of the Stave and Hayward Reservoir Ecosystem  In fresh water ecosystems, including reservoirs, primary producers include periphyton, bryophytes and vascular plants (Lamberti, 1996). In this study, periphyton was viewed to be the dominant primary producer and, as a result, the study focused on measurements of periphyton biomass (accrual and chl-a) as a way to examine and potentially quantify littoral productivity. The study maintained a focus on evaluating whether water management regimes in reservoirs negatively impact biological (carbon) productivity in the littoral zone and to the pelagic reservoir, as an approximation of whole reservoir production.  Among the reasons that periphyton was viewed to be the most telling and important measure of primary production, a few are worth elaboration. As reservoir ecosystems, a dominant feature characterizing these systems is the lake-river character of water flow, both from a historical stand point of how the system was created, as well as how short residence times and nutrient dynamics impact productivity and carbon flows. In this study, Hayward Reservoir is highly riverine in its character, while Stave is a truly hybrid complex of lake (main basin) and riverine (south, old river basin) characteristics in terms of its limnology and is further complicated by water level fluctuations that have been hypothesized to impact the overall biological (carbon) production of the ecosystem. While measurements of periphyton accrual and estimates of production are common in stream ecosystems where periphyton is recognized to dominate production (Hynes, 1970, Minshall, 1988), it has not been studied as extensively in lake habitats where estimates of productivity have concentrated on pelagic phytoplankton studies.  Periphyton along with bacteria, fungi and pico-size plankton actively function at the base of aquatic food webs, both pelagic and littoral. Periphyton is believed to be a key resource in nutrient deficient aquatic ecosystems, and is likely to have an even more critical role in a reservoir ecosystem like Stave where normal cycles of growth are disrupted by water management for power generation. Periphyton is a valuable food resource that is consumed by organisms at nearly all trophic levels (herbivores and secondary consumers) (Gregory, 1983), and when in short supply has the potential to limit consumer populations (Lamberti, 1996). Under conditions of disturbance, such as drawdown, aquatic organisms such as periphyton and grazers become stressed and populations tend to decline. Peterson et al. (1990) found that on horizontal surfaces in lentic systems, early-successional periphyton communities are easily Chapter 3  72  disrupted by turbulence, but typically recover faster than consumer (grazer) population (Steinman and Mclntire, 1990). In Stave Reservoir, frequent drawdown events are likely to maintain the periphyton community in a perpetual state of early development (pulse stability), and in addition quick recovery of the community is jeopardized by the lengthy duration of drawdown. Comparing littoral periphyton production between Stave and Hayward, which share similar physical and chemical properties but operate under different water management regimes allows for inferences to be made regarding the impact of disturbance on primary production.  The objectives of this thesis were to characterize and quantify littoral biomass accrual, to estimate littoral carbon production and to compare littoral and pelagic carbon production. This thesis looked at both accrual (biomass) and Chl-a concentrations of periphyton to assess littoral production, but focused on accrual as the primary measure of evaluation. On the basis of the A N O V A and descriptive statistics it is clear that there are differences in production between Stave and Hayward reservoirs. Three year average annual production values in Stave and Hayward were calculated as 5.3 and 12.3 gC/m /year, respectively. Directly comparing all 2  aspects of the Stave/Hayward ecosystem, periphyton as a food resource in Hayward is 2.3 times greater than that of Stave. Statistical analysis of production data for the two reservoirs found that Hayward Reservoir consistently produced more littoral carbon than Stave Reservoir in all seasons. Mean differences in production data ranged from 0.2 gC/m day in fall to 0.5 gC/m /day 2/  2  in spring. Comparisons between seasons found that winter was significantly different (at 0.5 level) than all other season, but due to differences in the sampling regime in winter months data from this research did not support taking evaluations of winter differences further.  Physico-chemical contribution to production differences Physico-chemical variables explored in this study offer some explanation of the differences in productivity between Stave and Hayward (Table 9). Based on proximity, light, temperature and nutrient regimes that are largely controlled by local climate and geology, Stave and Hayward reservoirs can be viewed to be similar, but reservoir management strategies alter the way these environmental variables interact within the aquatic environment resulting in some notable differences between the two systems.  Chapter 3  73  Table 9.  Summary of environmental data with key differences highlighted. Light  Surface solar radiation: Winter Surface solar radiation: Summer Extinction coefficient Compensation Depth: Winter Summer  Stave  Hayward  25-50 W / m * 150-210 W / m  2  25-50 W/m2 150-210 W / m  2  0.30-0.52 8-10 m ^  8-10 m 12-14 m  12-14 m  Temperature Stratification Average temperature  1 C warmer S  Nutrients - 1, 3, 5 m composite Nutrient classification N:P Ratio (max) (min) TP TDP Nitrate ( N 0 ) 3  130:1 16:1 2.0-4.5 /jg/L 1.0-3.0 jjg/L 40-160 /yg/L  181:1 25:1 2.0-6.8 /yg/L <1.0-5.3/yg/L 40-160/yg/L  Water Level Fluctuation Magnitude of fluctuation Residence times  80 i  l 4 rt  Among these differences a few emerge as important in characterizing the differences between systems. First and foremost is the impact of a fluctuating water regime in Stave relative to the generally stable water regime in Hayward. The impacts of fluctuation are many but desiccation, shifting light and temperature profiles, and erosion standout in terms of their observed impact on periphyton growth.  A second difference is that Stave stratifies and develops a thermocline during the summer months, while Hayward does not. Allan (1995) indicates that this result is not unusual, stating that in shallow reservoirs with short residence times thermal stratification is less likely to develop in summer months. Residence times in both reservoirs are short, but by its run-of-theriver nature Hayward exchanges its full volume of water almost daily. The fast flushing, continual flow results in mixing that inhibits stratification and provides a continual renewal of nutrients that spiral rather than recycle within the water column, therefore pelagic production in Hayward Reservoir is simply washout from Stave and no difference in production is expected. On the other hand, the fast flushing nature of Hayward Reservoir provides a continual source and renewal nutrients, which may result in higher littoral productivity. Chapter 3  74  A third prominent feature is low nutrient levels in both systems. The principal nutrient limiting growth through much of the growing season is dissolved phosphorus (PO4), which declines to undetectable levels in both Stave and Hayward. Nitrate concentrations approach limiting levels in the fall, but appear to play a less significant role in controlling nutrient dynamics of the systems. The most comprehensive record of nutrient balance for Stave and Hayward is found in a supplemental year of nutrient sampling that occurred in 2003 which shows the rapid decline in the NC»3:TDP ratio in fall when production was highest (Figure 41).  Nitrate :TDP Ratio 150 125  Q  • ~i r i 1  1 00 00 1  75 CO  50 25 0  1 I I 11  Jan  Feb Mar Apr May Jun  I Stave  ! Hayward  Jul  Aug Sep Oct  Nov Dec  2003  Figure 41. 2003 N 0 : T D P Ratio 3  Note: declining levels in fall. More closely examining the lake-river hybrid complexity of Stave Reservoir reveals that the main basin of Stave, once Stave Lake, is predominantly lake-like in its function. The shoreline is characterized by steep quartz diorite bedrock, boulders and gravel that are slow to weather, and low ambient nitrogen and phosphorus inputs into the reservoir. Notably absent from most areas of the upper Stave basin are other primary producers such as macrophytic plants, making the role of periphyton even more critical in the Stave ecosystem.  While some sections of the lower basin of Stave Reservoir are characterized by bedrock walls, similar to the upper basin, the dominant feature of the lower basin is the old Stave river channel that is exposed during periods of drawdown. Stave Reservoir shoreline has been heavily Chapter 3  75  impacted by flooding and drawdown, but the impacts are more prominent in the lower basin where vast, low-lying areas of aquatic vegetation and riparian shrubs and the remnant channel of the Stave River are periodically exposed. Although biomass accrual was not measured in the lower basin, based on observations of periphyton desiccation on exposed sampling substrata, it is likely that periphytic communities growing in association with macrophytic plants are severely impacted by the drying and erosion that occurs as a result of drawdown.  Further complicating production dynamics, it is theorized that during drawdown and water transport nutrients are released from sediments, old stumps and also from revegetation in vast 'old' riparian areas, albeit at relatively low levels. It is likely that ambient nutrients in the lower Stave basin tend to spiral through the pseudo-riverine system and given the relatively fast flushing nature (i.e. monthly residence time) of Stave Reservoir, the majority of these nutrients and carbon are quickly flushed into Hayward Reservoir. Conversely, in the main basin, where the function is generally lake-like, there is potential for sedimentation and nutrient recycling to occur.  Exploring the role of the Microbial Loop Comparing production estimates of Stave and Hayward provides an opportunity to examine some of the linkages between attached algae (periphyton) and the microbial food web that in the past were largely disregarded as quantitatively significant to production processes in aquatic science (Wetzel and Sondergaard, 1997). Recent limnological studies assert that littoral production has the potential to contribute significantly to the overall primary production of freshwater ecosystems, particularly those that are nutrient deficient (Wetzel and Sondergaard, 1997; Wetzel 1983). The nutrient-poor Stave/Hayward ecosystems where small 'opportunistic' species prevail provide a unique opportunity to assess the role of the microbial loop in reservoir ecosystems and more specifically to the littoral food web.  Until recently little was know about the importance of the microbial loop in terms of both trophic dynamics and carbon flows. The classic food chain first introduced by Lindeman (1942), described the aquatic food web as a simple linear, unidirectional trophic level transfer of carbon, and this view was the accepted theory until 1970's and early '80's. Decomposers in this model were generally poorly understood, largely because techniques for identification of these tiny organisms had not yet been developed. These organisms were treated as a "black box" and Chapter 3  76  broadly viewed as reducers of organic waste. More recently, new techniques for examining bacteria and plankton in aquatic ecosystems have revealed that pico-sized organisms (<2[i) are far more abundant than originally thought and that they have a fundamental role in the transfer of carbon in aquatic food webs (Pomeroy 1974; Azam et al., 1983). Estimates of bacterial cell abundance were later found to be in the 10 -10 cells per ml range in the open ocean and 5  6  potentially up to an order of magnitude higher for estuaries, lakes and coastal waters (Bergman, 1990). Bergman (1990) also estimated that carbon production associated with bacteria accounts for 20-40% of total carbon production.  Shortly after the importance of bacteria was realized came the discovery of picocyanobacteria. Picocyanobacteria are similar to bacteria in terms of their size (1-2JU) in aquatic ecosystems, but they differ because bacteria because they are less abundant by an order of magnitude (10 -10 ) 4  5  and because they are heterotrophs that rely on D O M and P O M as their food source, while picocyanobacteria  are  autotrophs,  thereby  relying  on  photosynthesis.  Bacteria  and  picocyanobacteria, also called autotrophic picoplankton (APP) compete with each other actively feeding on D O M , P O M and other nutrients such as N , P, phosphate (PO4), and ammonia (NH ). 3  Bacteria, because of their high abundance and their ability to survive even under conditions of darkness, often out compete APP. However, A P P are at an advantage when nutrients are low and there is an abundance of light. Together picoplankton (bacteria and APP) form the primary carbon template that are preyed upon by protozoan consumers (primarily nanoplanktors such as flagellates and ciliates), who in turn are grazed by large rotifers, nauplii and other small microzooplankton. Picoplankton are now recognized as the most abundant primary producers in aquatic environments, accounting for up to 90% of total phytoplankton production in oligotrophic ecosystems and potentially up to 50% in eutrophic environments (Stockner and Antia, 1986; Stockner, 1988, 1991; Stockner and Shortreed, 1989, 1991).  The discovery of picoplanktors brought with it a paradigm shift in aquatic science regarding material cycling and carbon transfers within the food web. This discovery shed light on what occurs in the black box that is now commonly referred to as the microbial loop (Azam 1983). Carbon transfers at the bottom of the food chain are now viewed as integral to nutrient dynamics because at all levels of predation nutrients and D O M are released into the water column by the processes of excretion and sloppy feeding, a term first introduced by Pace et al. (1990), that  Chapter 3  77  defined a commensalistic feedback loop that is the central link between microbial processes and the classic food chain (Sherr et al., 1988).  The microbial loop theory has brought with it the realization that food web function in ecosystems is dependant upon the overall nutrient supply of that ecosystem. In oligotrophic and ultra-oligotrophic environments, nutrient limitation diminishes the success of keystone predators, such as Daphnia and other smaller Cladocerans {eg. Bosmina). Keystone grazers are capable of feeding over a relatively wide range of organisms and under a more balanced nutrient regime facilitate the transfer of carbon, but under conditions of nutrient deficiency their existence is difficult or impossible as the ecosystem struggles to support phyto and zooplankton populations.  Oligotrophic conditions, like those in Stave Reservoir, generally elongate the food web by increasing the number of trophic level transfers, overall making them less efficient ecosystems with high respiratory costs (CO2 loss). Under these conditions the microbial loop becomes the primary source of nutrients and the main pathway to transfer carbon through the food web (Figure 42). Bacteria and pico-sized algae possess high mean abundance but low mean biomass owing to their small size (Stockner and Antia, 1986). As a result, they are the most well adapted scavengers to take advantage of the few nutrients that are available, largely due to their ubiquitous presence and their ability to recycle nutrients and organic matter from higher tophic levels.  Data from the pelagic component of the Stave/Hayward monitoring program confirm the importance of microbial function in coastal reservoir ecosystems. Size-fractioned primary production in the pelagic monitor of Stave confirmed that pico (0.2-2.0/*)- and nano (2-20/*)sized fractions dominated pelagic production in both Stave and Hayward reservoirs. Bacteria counted as part of the pelagic monitor in Stave and Hayward reported in Stockner and Beer (2004) averaged 700,000 cell/mL, values which were considered typical for coastal temperate B C lakes and oligotrophic boreal lakes. APP densities in the same study were reported to be very low relative to other coastal B C lakes with populations in the 7-9000 cell/mL range.  Chapter 3  78  Microbial function in the Stave/Hayward littoral food web Although picoplankton densities were not measured as part of the littoral monitoring program, the results of the pelagic study (Stockner and Beer, 2004) suggest that microbial processes are likely to dominate material cycling and energy flows in the littoral zones of Stave and Hayward reservoirs, as they were found to in the pelagic. Following the paradigm shift to microbial food web theory, limnologists have also come to recognize the potentially significant role of littoral communities, in terms of both the nutrient dynamics and contribution to whole-lake carbon production. Wetzel (1983a) found that littoral production accounted for 1-62% of whole-lake primary productivity, indicating that the contribution of littoral microorganisms can be significant. It is now being acknowledged that picoplankton readily use periphytic algae and macrophytic plants as substrates for colonization. The relationship between macrophytes, periphyton and picoplankton are interconnected by grazing, which is thought to account for the high productivity of periphytic algae in littoral communities.  Wetzel and Sondergaard (1997) assert that bacterial productivity is associated with submerged macrophytes and other available surfaces and that there is a direct coupling between bacteria and the productivity of epiphytic algae. High productivity rates of periphytic algae in littoral communities are now viewed to be a function of intensive internal recycling of nutrients, including carbon and gases within attached microbial communities (Wetzel, 1983b).  The littoral food web of Stave and Hayward appears to be driven by the ubiquitous presence of periphyton, bacteria and A P P living on and in association with most available surfaces in the reservoir including sediment, rocks, plants, detritus and stumps. Together they form a matrix (slime on rocks) of nutrient rich food that is consumed over all trophic levels (Lamberti 1996) (Figure 43). Stave is largely devoid of nutrients, therefore microorganisms in this periphytic matrix are best positioned to successfully compete and take up D O M and nutrients that drive primary productivity in the littoral food web. The majority of picoplankton in Stave and Hayward reservoirs are likely to be consumed by nanoplanktors such as ciliates, flagellates and dinoflagellates, as well as, some larger zooplankton such as rotifers and Bosmina (Stockner, pers. comm.). Notably lacking from the littoral food web is an abundant and dominating presence of keystone grazers, such as chironomids (Chironomus  Chapter 3  spp.), dragonflies (Odonata), bugs  80  (Hemiptera) or crayfish (Decapoda), although these organisms undoubtedly exist within the ecosystem nutrient supply is insufficient for them to have a dominant role.  In Stave Reservoir, algal biomass (primary production) is regulated not only by nutrient limitation but also by disturbance associated with water management. Higher primary production in Hayward Reservoir relative to Stave implies that reservoir management for power generation has a significant negative impact on littoral function and its ability to contribute to primary production of the system as a whole.  Calculations of littoral contribution to the aquatic ecosystem in this study are approximations based on littoral area data (BC Hydro), which arbitrarily defined the littoral zones of Stave and Hayward as a 6 m bathymetric band. Although the areas are only estimates, they serve as a starting point to understanding the potential impact that drawdown and hydropower operations may have on the overall biological productivity of reservoir ecosystems.  o£° 0  g +  BACTERIA CYANOBACTERIA  X  FUNGAL HYPHA MATRIX  ENZYMES  Figure 43. Schematic diagram of periphytic matrix growing on substrata similar to those in Stave and Hayward (modified from Allan 1995).  Chapter 3  81  While recent studies have found that algal biomass production in littoral habitats are high and have the potential to contribute significantly to overall ecosystem production (Wetzel and Sondergaard (1997), comparisons between Stave and Hayward reservoirs indicate that water level fluctuation combined with short residence time and lake-river hybrid nutrient dynamics deprive Stave Reservoir of this potential. Estimates of littoral to pelagic carbon production from this thesis indicated that areal littoral carbon production in Stave accounted for only 5% of total aquatic carbon production, while areal littoral carbon production in Hayward accounted for about 50% total aquatic carbon production. Littoral habitat in Hayward was found to be more riverine and the high proportion of littoral carbon production is comparable to flowing water habitats of large rivers and slower reaches of steams that are capable of supporting a viable phytoplankton pelagic community (Allan, 1995). Littoral vs. pelagic production comparison in Stave Reservoir is very low when compared to values reported in Shortreed et al. (1984) where littoral production seldom accounted for more than 1-5% of total lake carbon production (Stockner and Beer, 2004).  Overall Stave and Hayward reservoirs are ultra-oligotrophic and estimates of carbon production, dissolved phosphorus and periphyton accrual are among the lowest, if not the lowest in any B C freshwater ecosystem (Stockner and Beer, 2004). Low production in both the littoral and pelagic components of Stave and Hayward signals the dominant role of microbial food webs, the result of which is a less productive ecosystem that has difficulty supporting a functional zooplankton population, let alone a functional fish population.  Two-fold lower annual littoral production in Stave relative to Hayward implies that water fluctuations, short residence times and nutrient dynamics associated with lake circulation patterns and power generation have a significant impact on littoral carbon production. Additionally, striking differences found in littoral carbon contribution to aquatic production between Stave and Hayward reservoirs reinforces that reservoir operations have a significant impact on reservoir function. These differences in production between Stave and Hayward reservoirs appear to be largely attributable to differences in water management, residence times and nutrient dynamics.  When compared to natural systems, it is apparent that the character of Hayward reservoir is more closely affiliated to a riverine ecosystem that is characterized by continual mixing, spiraling of Chapter 3  82  nutrients, and high export of particulate and organic matter associated with near to daily exchanges of water volume, the result of which is a more productive littoral food web.  In this study, littoral carbon contribution in Stave Reservoir was estimated to be among the lowest levels observed in coastal B C lakes and/or reservoirs, suggesting that water management regimes in hydroelectric reservoirs exacerbate the productive capacity of these ecosystems resulting in low annual production and an inability to support zooplankton and fish further up the food chain. In systems that are already low in nutrients, reservoir function lowers littoral biomass production, but because production is already very low the difference is not as evident as for a nutrient balanced ecosystem. The combined effect of low ambient nutrients, short residence times and high export of nitrogen, phosphorus and carbon is to eliminate benthic-pelagic-littoral coupling of nutrients, which ensures the predominance of the microbial food web and low biotic productivity in the system as a whole. In Stave Reservoir, reservoir function combined with a naturally nutrient-poor environment maintain microbial populations in a state of early succession, ensure the sustenance of oligotrophy and drive carbon production to levels that are among the lowest observed in the literature for British Columbia lakes.  Chapter 3  83  CHAPTER 4  4.0  CONCLUSIONS  The objectives of this thesis were three-fold: to estimate periphyton production in the littoral zone of Stave and Hayward reservoirs; to compare differences in biomass production between the two reservoirs; and to compare the ratio of littoral productivity to pelagic productivity on an areal basis for each reservoir.  Findings from this research indicate that Stave Reservoir is characterized by low periphytic accrual relative to other coastal temperate lakes, while Hayward Reservoir is low but slightly higher when compared to other coastal temperate lakes. Both reservoirs exhibited a similar pattern of accrual with peak biomass occurring in the early spring and again in early fall. Although peak accrual in Stave occurred at approximately the same times of year and followed a similar pattern as Hayward, Stave can be differentiated from Hayward by a longer time over which peaks occurred and by its particularly low rates of biomass accrual. When examining accrual data by depth, Stave Reservoir exhibited more variability. Peak accrual in Stave generally occurred at deeper depths but with more variation between depths and among transects and was again of lower magnitude. In contrast to Stave, the more stable water management regime in Hayward resulted in what can be considered a more typical pattern of accrual for lakes or large rivers, with peak values occurring relatively consistently at 2-4 m below the average water surface elevation.  Chlorophyll concentrations were measured as a secondary indicator of production in this study. Analyses of chlorophyll concentrations were more variable than measures of accrual, but generally indicated similar trends in production between Stave and Hayward. Hayward reservoir was found to out-produce Stave Reservoir, but overall chlorophyll production in both reservoirs was found to be low and consistent with comparisons to other temperate freshwater oligotrophic ecosystems.  Statistical analysis comparing production between Stave and Hayward and amongst seasons, found that Stave was significantly different than Hayward and seasonally winter was found to be Chapter 4  84  significantly different than the other three seasons. While there are many practical explanations to explain why winter would exhibit significantly lower production, such as lower light and temperature to slow growth, this analysis was not explored extensively because the study used a different interval for sampling during winter months that are likely to have influenced the data and which complicate the ability to statistically interpret the outcomes.  Differences in production between Stave and Hayward were explored by examining the physical and chemical data that was collected concurrent with littoral periphyton sampling in this study. Temperature and light data were generally similar between Stave and Hayward, partly because measurements were relatively infrequent. Temperatures in Hayward were found to be slightly warmer in summer months, which may be due to the smaller and shallower volume of water that is warmed during transit through the system, and also due to mixing that results in isothermal conditions throughout the year. Therefore although temperatures were generally within a degree or two between Stave and Hayward throughout the year, the non-stratified character of Hayward was found to be a key difference between the two systems.  Light measurements were used to provide a more complete record of solar irradiance during the study. The moderately clear, oligotophic nature of both Stave and Hayward waters, as well as the seasonality and the potential for quick temporal changes in incident light due to cloud cover and climatic variation, indicate that light is likely to be a key variable controlling periphytic growth in these systems. A shortcoming of this research project was the lack of a more continuous and complete record of solar irradiance at each location, which may have allowed for a more comparative analysis of light and its influence on growth among depths and between transects. From the light data that was assessed in this study, it was found that there was some variability from transect to transect within Stave Reservoir, particularly between east and west sites. Variability in light among the three transects in Stave Reservoir is likely to be impacted by the fact that two of the transects were located on the east side of the reservoir and one site was located on the west side, which has the potential to impact the angle of incidence of the light as well as differences in aspect and shading associated with the surrounding mountains of this area that are near to 2000 m elevation. Stave was generally found to have moderately clear water with light penetration sometimes reaching 10-15 m.  Chapter 4  85  Light extinction in Hayward Reservoir exhibited brief periods of higher extinction than in Stave Reservoir, which may be attributable to finer sediments that characterize the littoral area of Hayward. Hayward light extinction was more variable due to the fine sediments (clays and silts) that characterized the reservoir bottom and tended to remain suspended in the water column once disturbed due to isothermal conditions.  Overall nutrients in Stave and Hayward were found to be very low defining Stave as an ultraoligotrophic ecosystem and Hayward as oligo- to ultra- oligotrophic. Phosphorus was inferred to be the primary limiting nutrient, frequently approaching limiting level in late summer and early fall during periods of maximum plant growth. Nitrogen was also found to be low but not to the same limiting extent as phosphorus. Nutrient loading from the watershed surrounding Stave and Hayward reservoirs  are not likely to contribute significantly  to either of these aquatic  ecosystems. The surrounding geology is quartz diorite that is slow to weather  and the  surrounding landscape is dominantly forest covered with relatively few large tributary creeks and little human settlement.  The key factors controlling the physico-chemical nature of Stave and Hayward ecosystems are water level fluctuation and water residence times associated with reservoir management for hydropower.  Stave  and Hayward  are similar ecosystems  sharing the  same  water-mass,  surrounding topography and geology. The results of the statistical analysis of the two reservoirs clearly show that Hayward is more productive than Stave Reservoir, with the one clear difference between these two reservoirs being the way that water is managed. Stave reservoir undergoes extreme water level fluctuation that has been shown in this study to negatively impact the growth of periphyton and in turn littoral productivity of the ecosystem as a whole. Higher production in Hayward is thought to be largely attributable to two key factors:  stable water  regime within this reservoir, and to extremely short residence time.  Furthermore, the assessment of littoral vs. pelagic production in Stave and Hayward found that littoral production in Stave contributed only 5% the overall productivity of the ecosystem while Hayward, the stable, run-of-the-river managed system contributed about 50%. Stave Reservoir was found to have the lowest production ever published in the primary literature for freshwater lakes. While it is acknowledged that the findings of this research are only estimates, they may be  Chapter 4  86  of particular importance to the management of power generating reservoirs with respect to other interests, such as fisheries values.  The comparison of Stave and Hayward reservoirs conducted in this thesis, suggests that reservoir operations for power generation negatively impacts littoral function resulting in lower biological carbon production. The main impacts appear to be associated with water level fluctuation and short water residence times that in turn influence other variables controlling growth. Short water residence times are generally thought to lower productivity in oligotrophic ecosystems, but in the case of Hayward, fast flushing water appears to ensure a continuous supply of nutrients, which in turn promotes productivity. In Stave, nutrient cycling in a system deprived of nutrients, the relatively short residence times (albeit longer than in Hayward -months vs. days), and extensive water fluctuation lower carbon production and sustain it as an ultra-oligotrophic system.  In Stave Reservoir the impact of reservoir function on littoral production is considerably less pronounced, than under conditions of mesotrophy. In a system with a more balanced nutrient regime, the impact of drawdown, variable residence times, vertically shifting physical and chemical variables and erosion associated with reservoir function are likely to be more profound than in a poor production system like Stave, and would likely exhibit an even larger impact on whole-ecosystem production.  4.1  Mitigation and Recommendations  The potential to enhance or mitigate the impacts of reservoir function on littoral productivity is possible. A n obvious although not always practical solution is to reduce the frequency and amplitude of reservoir fluctuation. Under the Stave River W U P , a less intense cycle of drawdown was found to be an agreeable outcome for all groups involved in the round table process. Evaluating all potential options and seeking input from other users can be effective and result in win-win outcomes.  Another potential strategy to help mitigate the impact of reservoir function is to increase macrophytic vegetation in the littoral habitat of reservoirs. While macrophytic plants in littoral zones are generally not considered to be overly productive, it is now being acknowledged that the importance of plants growing in nearshore littoral zones is to provide three-dimensional Chapter 4  87  growing surfaces for algal colonization and to potentially provide important nutrients for periphyton growth. Plant foliage provides diverse microhabitat and increases the surface areas available as a growing medium, which enhances the exchange of dissolved gases, such as oxygen from photosynthesis and carbon dioxide from decomposition. Perennial macrophytes also take up nutrients from interstitial hydrosoils that are primarily derived from decomposing plant tissue and attached algae. Nutrient transfer and uptake by flooded macrophytes facilitates a significant recycling of the pool of nutrients available to support periphytic growth. Additionally plant decomposition during flooding leaks nutrients into the water column that may also become available to other organisms such as periphyton. Vegetated littoral zones are now recognized to be highly productive habitat for periphyton growth, which in turn supports other microbial communities such as those for bacteria and A P P that form the base of the microbial food web. Research currently being conducted at U B C , in association with B C H , has the potential to shed light on the significance of these interactions  There is a distinct possibility to improve microbial function and primary productivity through the enhancement of macrophytic growth in the littoral zones of freshwater lakes and reservoirs. Planting programs have already been successfully implemented to control erosion in Arrow Reservoir, near Revelstoke, British Columbia, with spin-off benefits of expanding native wetland species within the littoral zone. Sedges, rushes and grasses have multiplied in existing areas of the drawdown zone and in some areas have colonized lower elevations that were previously barren. While the focus of the Arrow project was to control dust, similar methods could potentially be employed to enhance and expand microhabitats for periphytic colonization.  While it is noted that Arrow is an Interior reservoir with different annual water level regimes than a coastal reservoir such as the Stave/Hayward system, it is still indicative of the possibility to enhance or restore a more functional food web in reservoirs. Some reservoirs may offer greater restoration potential than others, depending upon individual site characteristics, water management regime and length of the growing season. Outcomes of the Arrow project indicated that establishment is key to the plants being able to tolerate water level fluctuation, especially at lower elevations within the basin. As a result of the establishment period, the implementation of specific water management regimes may be required during establishment or there may be a need to focus on higher elevations of the basin that experience less flooding. Once established,  Chapter  4  88  increased colonization of macrophytic plants may greatly enhance aquatic functioning of the ecosystem, particularly in the littoral community.  Nutrient supplementation is another potential to improve productivity in oligotrophic reservoir ecosystems like Stave. Nutrient addition has been used with great success in other aquatic ecosystems (Stockner and Maclssac, 1996) and in Allouette Reservoir in the adjacent watershed to Stave. The result of nutrient enrichment programs has resulting in increased productivity and in some cases the ability to support a viable fishery.  Overall the research conducted in this thesis shows foresight on the part of BCH to conduct baseline research that has advanced the understanding of reservoirs as functional ecosystems. The research concretely and conceptually forwarded our understanding and suggested ways to mitigate reservoir dysfunction, which could impact other values of the ecosystem as a whole. As ecosystems, reservoirs have the potential to provide many resources including hydro-electric power, fisheries and recreation. In a society that continually demands more from our resources, research such as that conducted in this thesis aims to balance use, management and function of the ecosystem as a whole by incorporating sustainability principals alongside reservoir operations.  Chapter 4  89  REFERENCES  ALLAN,  J. D. 1995. Stream Ecology: Structure and function in running waters. Chapman and Hall.  AXLER,  R. P., A N D J. E. R E U T E R . 1996. Nitrate uptake by phytoplankton and periphyton: Wholelake enrichments and mesocosm-15N experiments in an oligotrophic lake. Limnology and Oceanography 41: 659-671.  A Z A M , F., T. F E N C H E L , J. G . F I E L D , J. S. G R A Y , L - A - M E Y E R - R E I L A N D F. T H I N G S T A D . 1983.  The  ecological role of water-column microbes in the sea. Mar. Ecol. Prog. Ser. 10: 257-263. R. M . 1977. Environmental effects of dams and impoundments. Annual Review of Ecological Systems 8: 255-283.  BAXTER,  R. M . , A N D P. G L A U D E . 1980. Environmental effects of dams and impoundments in Canada: experience and prospects. Dept. of Fisheries and Oceans.  BAXTER,  BC  1994. Stave Falls Powerplant Replacement Project Environmental Impact Management Plan, p. 83. B C Hydro -environmental affairs.  HYDRO.  BEER,  J. A., A N D D. D O L E C K I . 2003. Stave River Littoral Zone Monitor Phase 1 - Stave/Hayward Reservoirs Periphyton Species Identification Report, p. 14. B C Hydro. N . G . 1982. Some observations of the ecology and fish management of reservoirs in the United States. Canadian Journal of Water Resources 7: 2-25.  BENSON,  BERGMAN, BIRD,  T. 1990. Large Lakes: Ecological Structure and Function. Springer.  D. F., A N D J. K A L F F . 1986. Bacterial grazing by planktonic algae. Science 231: 493-495.  BORCHARDT, BRUCE,  M . A . 1996. Nutrients, Algal Ecology. Academic Press.  J., A N D J. S T O C K N E R . 2000. Stave River Littoral Zone Monitor: Design of a Long Term Plan. B C Hydro. J. M . , R . G . W E T Z E L . 1990. Epiphytic alkaline phosphatase on natural and artificial plants in an oligotrophic lake: Re-evaluation of the role of macrophytes as a phosphorus source of epiphytes. Limnol. Oceanography 35.  BURKHOLDER,  C A N A D I A N W A T E R RESOURCES ASSOCIATION.  1976. Canadian water resources journal. Revue  canadienne des resources en eau. CARR,  W. W., A N D A . M O O D Y . 2000. Strategic Plan for Reservoir and Drawdown Zone Revegetation within B C Hydro. B C Hydro.  References  90  CATTENEO,  A . 1983. Grazing on epiphytes. Limnol. Oceanography 28: 124-132.  A., A N D M A R I E C H R I S T I N E A M I R E A U L T . 1992. How artificial are artificial substrata for periphyton? Journal of the North American Benthological Society 11: 244-256.  CATTENEO,  C U M M I N S , K . W., M . A . W I L Z B A C H , D . M . G A T E S , J.B. P E R R Y , A N D W . B . T A L I A F E R R O .  1989.  Shredders and riparian vegetation. Bioscience 39: 24-30. ENNIS,  G. L . 1975. Distribution and abundance of benthic algae along phosphate gradients in Kootenay Lake, British Columbia. Verh. Int. Ver. Thoer. Angew. Limnol. 19: 562-570.  EVANS,  D., A N D J.G. S T O C K N E R . 1972. Attached Algae on Artificial and Natural Substrates in Lake Winnipeg, Manitoba. Journal of Fisheries Research Board of Canada 29: 31-44.  Fox, J. L., T.O. O D L A U G , A N D T.A. O L S O N . 1969. The ecology of periphyton in western Lake Superior. Bull. Water Resources Res. Center University of Minnesota 14: 99. J., B. M A R O T Z , J. D E C K E R - H E S S , W. B E A T T I E , A N D R. Z U B I K . 1989. Mitigation, Compensation and Future Protection for Fish Populations Affected by Hydropower Development in the Upper Columbia System, Montana, USA. Regulated Rivers: Research and Management 3: 3-18.  FRALEY,  GASITH,  A., A N D S A R I G G A F N Y . 1990. Effects of Water Level Fluctuation on the Structure and Function of the Littoral Zone, p. P156-171. In M . M . a. C. S. Tilzer [ed.], Large Lakes Ecological Structure and Function - Chapter 8. Springer- Verlag. G. L . A . J. W. B. 1985. Vegetative succession and decomposition in reservoirs, p. 5978. In D. Gunnison [ed.], Microbial processes in Reservoirs. Dr. W. Junk, Kluwer Academic Publishers group.  GODSHALK,  S. V . 1983. Plant-herbivore interactions in stream systems, p. 157-189. In J. R. B. a. G. W. Minshall [ed.], Stream Ecology.  GREGORY,  GUNNISON,  D. 1985. Microbial Processes in Reservoirs. Dr. W. Junk Publisher  L . - A . 1988. Effects of competitive interaction on the biomass development of planktonic and epiphytic algae in lakes. Limnol. Oceanography 33: 121-128.  HANSSON,  —. 1992. Factors regulating periphytic algal biomass. Limnol. Oceanography 37: 322-328. G. E., and Y . H. E D M O N D S O N . 1957. A treatise on limnology: Vol. 1. Geography, physics, and chemistry. Wiley.  HUTCHINSON,  HYNES,  H. B. N . 1970. The Ecology of Running Waters. University of Toronto Press.  References  91  S., T.R. P A R S O N S , M . T A K A H A S H I , A N D H. S E K I . 1980. A comparison of four methods for integrating C-primary productivity measurements per unit area. Journal of Oceanogr. Soc. Jap. 36: 259-262.  ICHIMURA,  14  J. L. 1994. Stave Falls Powerplant Replacement Project Environmental Impact Management Plan, p. p83. British Columbia Hydro Environmental Affairs.  JACKSON,  JOINT,  I. R., A N D A.J. P O M R O Y . 1983. Production of picoplankton and small nanoplankton in the Celtic Sea. Marine Biology 77: 19-27.  KERR,  I. 1999. Stave River Water Use Plan, p. 33. B C Hydro. A . 1996. Role of Periphyton in Benthic Food Webs, p. 533-572. In R. J. Stevenson, M . L . Bothwell, and R.L. Lowe [ed.], Algal Ecology. Academic Press, Inc.  LAMBERTI, G .  L A M B E R T I , G . A . , S. V . G R E G O R Y , L.R.  A S H K E N A S , A.D.  STEINMAN AND C D .  MCINTIRE.  1989.  Stream ecosystem recovery following catastrophic debris flow. Can. J. Fish. Aquat. Sci. 48: 196-208. LINDEMAN,  R. L . 1942. The trophic-dynamic aspect of ecology. Ecology 23: 399-418.  E. A., K . S . S H O R T R E E D , A N D J . G . S T O C K N E R . 1981. Seasonal distribution of bacterioplankton numbers and activities in eight fertilized or untreated oligotrophic British Columbia lakes. Can. Tech. Rep. Fish. Aquat. Sci. 994: 43.  MACISAAC,  B. L., D. G U S T A F S O N , A N D C. A L T H E N . 1994. Model Development to Establish Integrated Operational Rule Curves for Hungry Horse and Libby Reservoirs (Draft). Bonneville Power Administration.  MAROTZ,  MCCUNE,  B.  A.  J. B. G . 2002. Analysis of Ecological Communities. M j M Software.  A. J. 1969. The effect of aquatic macrophytes on the variety and abundance of benthic fauna in a newly created lake in the tropics. Archiv fur Hydrobiologie 66: 212231.  MCLACHLAN,  W. 1988. Stream Ecosystem Theory: A global perspective. J. North American Benthological Society 7: 263-288.  MINSHALL, G .  C. A . P. A . K . 1988. Predictability of change in shoreline vegetation in a hydroelectric reservoir, northern Sweden. Can. J. Fish. Aquat. Sci. 45: 1896-1904.  NILSSON,  T. G . 1975. Periphytic and planktonic algae of the lower Fraser River in relation to water quality conditions. University of British Columbia.  NORTHCOTE,  T. G . , A N D D.Y. A T A G I . 1997. Ecological Interactions in the flooded littoral zone of reservoirs: The importance and role of submerged terrestrial vegetation with special reference to fish, fish habitat and fisheries in the Nechako Reservoir of British Columbia, Canada, p. 1-64. Ministry of Environment, Lands and Parks.  NORTHCOTE,  References  92  O D U M , E . P. 1971. Fundamentals of Ecology, 3rd ed. Saunders. P A C E , M . L . , G . B . M C M A N U S , A N D S . E . G . F I N D L A Y . 1990. Plankton community structure determines the fate of bacterial production in a temperate lake. Limnol. Oceanography 35: 795-808. P A G E , N . A . 2003. Community and Regional Scale Patterns of Native and Exotic Plant Species in Sand Beaches of Vancouver Island, British Columbia, p. 79, Resource Management and Environmental Studies. University of British Columbia. P A R S O N S , T . K . , Y . M A I T A A N D C . M . L A L L I . 1984. A Manual of Chemical and Biological Methods for Seawater Analysis. Pergamon Press. P E R R I N , C . J . , M . L . B O T H W E L L , A N D P . A . S L A N E Y . 1987. Experimental enrichment of a coastal stream in British Columbia: Effects of organic and inorganic additions on autotrophic periphyton production. Can. J . Fish. Aquat. Sci. 44: 1247-1256. P E T E R S O N , C . G . , K . D . H O A G L A N D A N D R . J . S T E V E N S O N . 1990. Timing of wave disturbance and the resistance and recovery of a freshwater microalgal community. J . North American Benthological Society 9: 54-67. P I E T E R S , R., L . T H O M P S O N , L . V I D M A N I C , M . R O U S H O R N E , J . S T O C K N E R , K . H A L L , M . Y O U N G , S. POND, M . D E R H A M , K. A S H L E Y , B. LINDSAY, G. LAWRENCE, H . ANDRUSAK, D. SEBASTIAN, G . S H O L T E N . 2000. Arrow Reservoir Fertilization Experiment Year 1 (1999/2000) Report. Fisheries Branch Ministry of Environment, Lands, and Parks. P O M E R O Y , L . R. 1974. The ocean's food web: a changing paradigm. Bioscience 24: 499-504. P O M E R O Y , W . M . 1977. Benthic algal ecology and primary pathways of energy flow on the Squamish River Delta, British Columbia. University of British Columbia. P O W E R , M . E . 1990. Effects offish in river food webs. Science 250:  811-814.  P R E S C O T T , G . W . 1978. How to Know the Freshwater Algae, 3d - ed. W . C . Brown. R O U N D , F . E . 1973. The Biology of the Algae, 2 ed. St. Martin's Press. S H A N Z , F . , A N D H . J U O N . 1983. T w o different methods of evaluating nutrient limitations periphyton bioassays using water from the Rhine and eight of its tributaries. Hydrobiologia 102:  187-195.  S H E R R , B . F . , E . B . S H E R R , A N D T . B E R G M A N . 1988. Trophic interactions within pelagic microbial communities: indications of feedback regulation of carbon flow. Hydrobiologia 159:  19-  26.  References  93  S., A . C . C O S T E L L A , A N D J . G . S T O C K N E R . 1 9 8 4 . Periphyton biomass and species composition in 2 1 British Columbia lakes: seasonal abundance and response to wholelake nutrient additions. Canadian Journal of Botany 62: 1 0 2 2 - 1 0 3 1 .  SHORTREED, K .  S., A N D J . G . S T O C K N E R . 1 9 7 8 . Response of Attached Algae to Whole-Lake Fertilization Experiments in Five British Columbia Coastal Lakes, p. 2 5 . Department of Fisheries and the Environment.  SHORTREED, K .  —. 1 9 8 1 . Limnological results from the 1 9 7 9 Lake Enrichment Program. Canadian Technical Report of Fisheries and Aquatic Science 995: 7 1 . —. 1 9 8 3 . Periphyton biomass and species composition in a coastal rainforest stream in British Columbia: effects of environmental change caused by logging. Canadian Journal of Fisheries and Aquatic Sciences 40: 1 8 8 7 - 1 8 9 5 . T. L., K . D. H Y A T T , T. G . N O R T H C O T E , A N D R. J. F I E L D E N . 1 9 9 6 . Status of anadromous salmon and trout in British Columbia and Yukon. Fisheries Habitat-Perspective 21: 2 0 -  SLANEY,  35.  D. M., A N D B.L. K I M M E L . 1 9 8 7 . A large-scale comparison of factors influencing, phytoplankton abundance inrivers,lakes and impoundments. Ecology 68: 1 9 4 3 - 1 9 5 4 .  SOBALLE,  L., J.E R E U T E R A N D CR. G O L D M A N . 1 9 8 3 . Littoral zone production of oligotrophic lakes: The contributions of phytoplankton and periphyton, p. 1 6 1 - 1 6 7 . In R. G . Wetzel [ed.], Periphyton of Freshwater Ecosystems. Dr. W Junk.  STANFORD,  E. 1 9 5 2 . The use of radioactive carbon ( C) for measuring organic production in the sea. Journal du Conseil Interm. I'Explor. de la Mer 18: 1 1 7 - 1 4 0 . 14  STEEMAN NIELSEN,  A. D. A . C. D. M. 1 9 9 0 . Recovery of lotic periphyton communities after disturbance. Environmental Management 14: 1 7 - 2 6 .  STEINMAN,  R. J., M. L. B O T H W E L L , and R. L. L O W E . ecosystems. Academic Press.  STEVENSON,  1996.  Algal ecology: freshwater benthic  J. G . 1 9 8 8 . Phototrophic picoplankton: An overview from marine and freshwater ecosystems. Limnol. Oceanography 33: 7 6 5 - 7 7 5 .  STOCKNER,  —. 1 9 9 1 . Autotrophic picoplankton in freshwater ecosystems: The view from the summit. Int. Revue ges Hydrobiologia 76: 4 8 3 - 4 9 2 . J. G . , A N D E.A. M A C I S A A C . 1 9 9 6 . The British Columbia Lake Fertilization Program: overview after two decades of Salmon Enhancement. Regulated Rivers: Research and Management 12: 3 4 4 - 3 5 6 .  STOCKNER,  J. G . , A N D F. A. J. A R M S T R O N G . 1 9 7 1 . Periphyton of the Experimental Lakes Area, Northwestern Ontario. Journal of Fisheries Research Board of Canada 28: 2 1 5 - 2 2 9 .  STOCKNER,  References  94  J. G . , A N D J.A. B E E R . 2 0 0 4 . The Limnology of Stave/Hayward Reservoirs: With a perspective on carbon production, p. 3 6 . B C Hydro.  STOCKNER,  J. G . , A N D K. S. S H O R T R E E D . 1 9 8 9 . Algal picoplankton production and contribution to food webs in oligotrophic British Columbia lakes. Hydrobiologia 173: 1 5 1 - 1 6 6 . —. 1 9 9 1 . Autotrophic picoplankton: community composition abundance and distribution across a gradient of oligotrophic British Columbia and Yukon lakes. Int. Revue ges. Hydrobiologia 76: 5 8 1 - 6 0 1 . STOCKNER,  J. G . , A N D K.S. S H O R T R E E D . 1 9 7 8 . Enhancement of autotrophic production by nutrient addition in a coastal rainforest stream on Vancouver Island. Journal of Fisheries Resource Board of Canada 35: 2 8 - 3 4 .  STOCKNER,  —. 1 9 8 5 . Whole-Lake Fertilization Experiments in Coastal British Columbia Lakes: Empirical Relationships between Nutrient Inputs and Phytoplankton Biomass and Production. Canadian Journal of Fisheries and Aquatic Sciences 42: 6 4 9 - 6 5 8 . J. G . , A N D N.J. A N T I A . 1 9 8 6 . Algal picoplankton from marine and freshwater ecosystems: a multidisciplinary perspective. Canadian Journal of Fisheries and Aquatic Sciences 43: 2 4 7 7 - 2 5 0 3 .  STOCKNER,  J. G . , E. R Y D I N , A N D P. H Y E N S T R A N D . 2 0 0 0 . Cultural Oligotrophication: Causes and Consequences for Fisheries Resources. Fisheries Habitat-Perspective 25: 7 - 1 4 .  STOCKNER,  M . , J. G . T U N D I S I , and A . D U N C A N . 1 9 9 3 . Comparative Reservoir Limnology & Water Quality Management. Kluwer Academic.  STRASKRABA,  V . 1 9 8 9 . Proceedings of the International Conference on Reservoir Limnology and Water Quality. W. Schweizerbart'sche Verlagsbuchhandlung.  STRASKRABOVA,  S T R A S K R A B O V A , V., and J. F. T A L L I N G . 1 9 9 4 . Proceedings of the Second International Conference on Reservoir Limnology and Water Quality. E. Schweizerbart'sche Verlagsbuchhandlung.  S Z E , P. 1 9 8 6 . A Biology of the Algae. W.C. Brown. J. D . , G . W E I B L E N , B.A. T H O M S O N , S. A L F A R O , A N D P. L E G E N D R E . 1 9 9 6 . Untangling multiple factors in spatial distributions: lilies, gophers, and rocks. Ecology 77: 1 6 9 8 -  THOMSON,  1715.  K. W. 1 9 9 0 . Perspectives on Reservoir Limnology. In K. W. Thornton, Bruce L . Kimmel, and Forrest E. Payne [ed.], Reservoir limnology: ecological perspectives. Wiley.  THORNTON,  K. W., B R U C E L. K I M M E L , A N D F O R R E S T E. Limnology: Ecological Perspectives. Wiley.  THORNTON,  References  PAYNE,  E. (ED.).  1990.  Reservoir  95  TILZER,  M . M . , A N D C O L E T T E S E R R U Y A . 1990. Large Lakes Ecological Structure and Function. Springer-Verlag. J. 2002. Practical Statistics for Environmental and Biological Scientists. John Wiley and Sons Ltd.  TOWNSEND,  1952. Neue Wege in der quantitativen Erfassung des Planktons (mit besonderer Berucksichtigung des Ultraplanktons). Verh. Int. Ver. Thoer. Angew. Limnol. 5: 567596.  UTERMOHL, H .  D. S T E I N M A N . 2002. Periphyton function in lake ecosystems. The Scientific World 2: 1449-1468.  VADEBONCOEUR, Y . , AND A L A N  M . J A K E V A N D E R Z A N D E N , A N D D A V I D M . L O D G E . 2002. Putting the lake back together: reintegrating benthic pathways into lake food web models. Bioscience 52: 44-54.  VADEBONCOEUR, Y . ,  C , D. G . M A N N , A N D H A N S M A R T I N Phycology. Cambridge University Press.  V A N DEN H O E K ,  JAHNS.  1995. Algae: An Introduction to  2003. Water quality standards for surface waters of the State of Washington. Washington State Administrative Code.  WASHINGTON STATE.  WETZEL,  R.  G.  1983a. Limnology, 2nd ed. Saunders.  —. 1983b. Periphyton of freshwater ecosystems: Proceedings of the First International Workshop on Periphyton of Freshwater Ecosystems, held in Vaxjo, Sweden, 14-17 September 1982. Dr. W. Junk Publishers —. 1990. Reservoir Ecosystems: Conclusions and Speculations. Wiley. R. G . , A N D M . S O N D E R G A A R D . 1997. Role of Submerged Macrophytes for Microbial Communities and Dynamics of Dissolved Organic Carbon in Aquatic Ecosystems, p. 133-148, The Structuring Role of Submerged Macrophytes in Lakes. Springer-Verlag.  WETZEL,  References  96  APPENDIX 1 Data Summary Table SP 1 Hayward Accrual (mg C/m2/day) Chi a (mg/m2) Average Water Level (m) a.s.l. LIGHT Total Sample Period Surface Irradiance (Pt.Moody) (W/m2) Average Sample Period Surface Irradiance (W/m2) Extinction coefficient Secchi Depths TEMP Average S P Temperature Ruskin Tailrace Nutrient Total Phosphate (TP) Total Disolved Phosphate (TDP) Nitrates (N03-N) N 0 3 - N / TDP Stave Common Data Average Water Level (m) a.s.l. LIGHT Total Sample Period Surface Irradiance (Pt.Moody) (W/m2) Average Sample Period Surface Irradiance (W/m2) Secchi Depths Temperature Average S P Temperature (C) (1996-2001 data) Nutrient Total Phosphate (TP) Total Disolved Phosphate (TDP) Nitrates (N03-N) N 0 3 - N / TDP  SP 2  5P4  5P3  SP 5  Feb-Apr Apr-May May-Jul Jul-Aug Aug-Sep 2000 2000 2000 2000 2000 12.58 64.37 15.84 122.16 62.00 0.99 2.90 2.62 2.60 4.80 42.30 42.32 42.15 42.14 42.39 5055.0  5498.0  7016.0  5150.0  5677.0  87.16  134.10  163.16 0.34 3.60  183.93 0.38 4.70  132.02  5.82  9.86  11.92  16.13  16.32  3.20 2.80 131.6 47  3.30 2.60 122.7 47.19  76.51  78.93  81.51  80.49  78.37  5055.0 87.16  5498.0 134.10  7016.0 163.16 4.60  5150.0 183.93 5.25  5677.0 132.02  5.24  8.60  11.36  14.94  16.61  2.90 2.70 142.40 52.74  3.30 3.10 157.20 50.71  4.71 0.25  21.92 0.66  13.29 1.08  34.72 2.50  39.42 2.72  0.35  0.37  7.61 0.75  20.65 2.36  0.34  0.39  Slave North Accrual (mg C/m2/day) Chi a (mg/m2) LIGHT Extinction coefficient Stave South Accrual (mg C/m2/day) Chi a (mg/m2) LIGHT Extinction coefficient  4.16 0.22  5.21 0.23  59.16 3.95  Stave West Accrual (mg C/m2/day) Chi a (mg/m2) LIGHT Extinction coefficient Notes: water level averages for Stave sites are all the same Nurtient data for stave is the same at each site taken from Stave Upper basin data Surface Solar Radiation is taken at Pt. Moody, data supplied by G V R D S S R measured with a LI-200SA pyronometer which measures global radiation (sumof both direct and diffuse solar radiation) in the 400-1100 nm .(most visible, some IR and no UV) PAR 360-750 nm.  Appendix 1  97  SP 8 SP9 SP 6 5P7 5P16 5PH 5P12 5P13 Sep-Nov Feb-Apr Apr-May May-Jul Jul-Aug Aug-Sep Sep-Nov Nov-Feb 2000 2001 2001 2001 2001 2001 2001 2001  24.08 1.96 42.36  11.85 1.03 42.28  43.81 1.11 42.26  63.56 3.66 42.44  24.08 1.90 40.81  76.12 3.56 42.44  19.04 1.53 42.28  45.21 0.91 42.28  3053.0  3552.0  4070.0  8245.0  8592.0  4832.0  7077.0  2557.0  58.71  37.79  94.65 0.54  164.90 0.37  204.57 0.38  193.28 0.38  133.53  62.37 0.40  12.59  5.22 1.85 0.79 142.53 181.41  2.96 1.87  5.22 0.10 85.65 856.50  2.39 5.34  155.11 67.71  3.26 2.13 153.88 72.39  84.53 15.83  4.20 1.13 149.78 132.05  76.71  76.88  77.09  79.60  81.51  81.42  80.39  77.87  3053.0 58.71  3552.0 37.79  4070.0 94.65  8245.0 164.90  8592.0 204.57  4832.0 193.28  7077.0 133.53  2557.0 62.37  10.98  10.47  5.20  8.26  11.30  13.67  16.79  12.44  2.29 2.29 150.36 65.64  2.29 2.02 147.10 72.85  2.19 1.12 145.64 129.76  2.66 1.97 105.64  1.98 0.10 81.35 813.46  4.17 2.22 83.93 37.73  3.67 1.33 128.94 96.62  10.83 0.53  25.66 0.27  35.63 2.02  14.15 2.47  34.14 1.18  45.66 1.54  59.96 0.00  0.51  0.34  0.35  0.36  8.76 0.28  14.27 2.17  16.15 1.08  26.34 1.33  0.51  0.32  0.36  0.37  2.19 0.03  8.51 1.09  9.79 1.00  14.88 0.51  0.51  0.32  0.44  0.36  4.18 2.29  12.00 1.21  20.19 2.76  Appendix 1  5.28 0.42  105.78  0.49 45.27 2.17  38.57 1.54 0.42  11.16 1.05  19.07 1.50 0.49  98  SP14  Nov-Feb  Appendix 1  SP 15 Feb-Apr  5P-I6  SP17  Apr-May  May-Jun Jun-Jul Sep-Oct Jul-Sep 2002 2002 2002 2002 35.75 40.17 43.14 72.63 3.18 2.80 1.93 2.36 42.34 42.33 42.42 42.36  SP 18  5P19  S P 20  2002 5.96 0.52 42.29  2002 40.43 0.83 42.40  2002 49.69 1.47 42.41  3047.0 26.27 0.70  3115.0 84.19 0.69  5203.0  7280.0  4877.0  7226.0  4822.0  130.08 0.38 2.80  173.33 0.38 3.33  174.18 0.41 3.21  164.23 0.36 5.30  94.55 0.39 4.90  5.44 0.31 147.69 472.73  6.36 2.23 161.12 72.27  6.77 5.31 144.66 27.26  5.72 5.31 104.05  2.72 1.20  4.94 1.38 42.60 35.57  65.91 47.81  79.55  76.55  78.13  81.05  81.23  80.62  76.19  3047.0 26.27  3115.0 84.19  5203.0 130.08 3.30  7280.0 173.33 3.70  4877.0 174.18 3.86  7226.0 164.23  4822.0 94.55 5.50  11.23  5.33  7.85  10.55  12.74  16.76  13.97  3.69 0.44 144.33 329.99  3.22 2.97 151.05 50.82  0.77 0.10 144.08 1440.76  4.22 4.29 71.10 16.59  3.51 2.82 45.27 16.04  2.07 1.61 65.91 40.98  3.30 0.23  19.07 0.31  15.42 0.63  13.74 0.58  7.54 0.34  35.27 2.50  22.03 1.73  0.38  0.43  0.31  0.43  0.53  0.44  0.32  1.89 0.24  28.40 0.33  18.27 0.39  16.77 0.58  19.16 0.58  16.66 1.59  15.45 1.21  0.46  0.44  0.41  0.42  0.44  0.44  0.31  1.67 0.21  4.88 0.12  10.61 0.27  13.43 0.42  13.98 0.32  10.95 0.98  36.33 1.27  0.40  0.37  0.27  0.46  0.48  0.44  0.39  108.78  99  2606  2001  2002  All SP  SPAVG  SPAVG  SPAVG  TOTAL  All 5P Average  50.17 2.64 42.28  40.53 1.96 42.11  41.11 1.87 42.37  872.48 42.66 845.02  43.62 2.13 42.25  5241.5 126.51 0.36 4.15  5560.7  5081.4 120.97 0.47 3.91  105944.0 2496.98 6.10  5297.2 124.85 0.44 3.98  127.30 0.41  77.87  Appendix 1  3.25 2.70 127.15  3.44 1.95 125.32  5.32 2.62 111.01  62.50 34.78 1797.59  4.17 2.32 119.84  78.76  79.25  79.05  0.00 1580.63  79.03  5241.5 126.51 4.93  5560.7 127.30  5081.4 120.97 4.09  105944.0 2496.98  5297.2 124.85 4.37  3.10 2.90 149.80  2.75 1.58 120.42  2.91 2.04 104.36  42.92 29.09 1873.08  2.86 1.94 117.07  21.01 1.40  32.29 1.114  16.62 0.90  468.44 22.75  23.42 1.14  0.36  0.41  0.40  5.61  0.40  19.50 1.71  22.09 1.29  16.66 0.70  388.22 24.18  19.41 1.21  0.36  0.40  0.42  5.62  0.40  10.93 0.86  13.12 0.51  157.45 8.77  12.11 0.67  0.43  0.40  4.94  0.41  100  APPENDIX 2 Summary Table of Annual Littoral Production Littoral Production: Sample period production  Sampling Sample Period  0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  Date  Accrual integrated over depth STVN  STVS  STVW  Hayward  15-Feb-00 12-Apr-00 23-May-00 7-Jul-00 15-Aug-00 27-Sep-00 18-Nov-00 20-Feb-01 4-Apr-01 24-May-01 5-Jul-01 30-Jul-01 21-Sep-01 1-Nov-01 25-Feb-02 3-Apr-02 13-May-02 24-Jun-02 22-Jul-02 4-Sep-02 25-Oct-02 (mgC/m2/yr) (gC/m2/yr)  0.00 4.71 21.92 13.29 34.72 39.42 12.00 10.83 25.66 35.63 14.15 34.14 45.66 59.96 3.30 19.07 15.42 13.74 7.54 35.27 22.03 6944.88 6.94 5.32  0.00 4.16 5.21 7.61 20.65 59.16 20.19 5.28 8.76 14.27 16.15 26.34 45.27 38.57 1.89 28.40 18.27 16.77 19.16 16.66 15.45 5625.20 5.63  0.00 2.19 8.51 9.79 14.88 11.16 19.07 1.67 4.88 10.61 13.43 13.98 10.95 36.33 3386.67 3.39  0.00 12.58 64.37 15.84 122.16 62.00 24.08 11.85 43.81 63.56 24.08 76.12 19.04 45.21 5.96 40.43 49.69 35.75 40.17 72.63 43.14 12292.06 12.29  2000  (mgC/m2/yr)  5188.25  0.00  12739.26  2001  (mgC/m2/yr)  6941.03  2696.09  12013.11  2002  (mgC/m2/yr)  5433.49 5310.87 10267.67 6634.93 5133.50 4652.36  4746.32  4077.25  12123.83  3 year average  Stave Transect Avg.  Appendix 2  101  Summary Table of Littoral Chlorophyll Production  Littoral Production: Sample Period Chlorophyll Average C H L a Sample Period  0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 3 year average  Sampling Date  15-Feb-00 12-Apr-00 23-May-00 7-Jul-00 15-Aug-00 27-Sep-00 18-Nov-00 20-Feb-01 4-Apr-01 24-May-01 5-Jul-01 30-Jul-01 21-Sep-01 1-Nov-01 25-Feb-02 3-Apr-02 13-May-02 24-Jun-02 22-Jul-02 4-Sep-02 25-Oct-02 mg/m2/year  STVN  0.00 0.25 0.66 1.08 2.50 2.72 1.21 0.53 0.27 2.02 2.47 1.18 1.54 0.00 0.23 0.31 0.63 0.58 0.34 2.50 1.73 347.59  STVS  0.00 0.22 0.23 0.75 2.36 3.95 2.76 0.42 0.28 2.17 1.08 1.33 2.17 1.54 0.24 0.33 0.39 0.58 0.58 1.59 1.21 368.86  STVW  0.00 0.03 1.09 1.00 0.51 1.05 1.50 0.21 0.12 0.27 0.42 0.32 0.98 1.27 201.34  Hayward  0.00 0.99 2.90 2.62 2.60 4.80 1.96 1.03 1.11 3.66 1.90 3.56 1.53 0.91 0.52 0.83 1.47 3.18 2.80 1.93 2.36 628.78  305.93 2000  mg/m2/year  2001  mg/m2/year  2002  mg/m2/year  331.09 343.33 427.28 382.79 205.78 272.98  355.57  634.51  576.85  144.23  745.46  339.40  273.75  528.27  stave transect avg.  Appendix 2  102  APPENDIX 3 Hayward Annual Sample Period Irradiance over Depth Hayward 2 0 0 0 : PAR vs Depth Irradiance (umol quanta/m2/sec) 0  400  800 1200 1600 2000  0  Hayward 2 0 0 1 : PAR vs Depth Irradiance (umol quanta/m2/sec) 0  400  800  1200  1600  2000  24  4 6 8 10 OL Q  12 14 16 18  -7-Jul  ^30-Jul  - 15-Aug  -•—1-Nov  20  Hayward 2002: PAR vs Depth Irradiance (umol quanta/m2/sec) 0  400  800  1200 1600 2000  10 - e — 25-Feb 12 14 16 ' 18  -•— 4-Apr -*— 13-May 24-Jun - X — 22-Jul -*— 4-Sep -•—• 25-Oct  20  Appendix 3  103  Stave North Sample Annual Sample Period Irradiance over Depth Stave North 2000: P A R vs. Depth  Irradiance (umol quanta/m2/sec)  Irradiance (umol/m2/sec) 0  500  1000  1500  Stave North 2001: P A R v s Depth  2000  0  400  800  1200 1600 2000  Q. O  a  -•— 4-Apr -•— 24-May -#-7-Jul 18 » 20  I  -m—7-Aug  5-Jul ^-30-Jul -•—1-Nov  Stave North 2002: P A R v s Depth Irradiance (umol quanta/m2/sec) 0 0 9-9  Appendix 3  400 L  800  1200 1600 2000  J"—  104  Stave South Sample Annual Sample Period Irradiance over Depth Stave South 2000: PAR vs. Depth  Stave South 2001: PARvs Depth Irradiance (umol quanta/m2/sec)  Irradiance (umol/m2/sec) 500  1000  1500  2000  0  400  800  1200  1600 2000  |  -•—4-Apr -m— 24-May 5-Jul  • 7-Jul  30-Jul  • 15-Aug  -•—1-Nov  Stave South 2002: PARvs Depth Irradiance (umol quanta/m2/sec) 0  400  800  0 W—9  '  zM.  Appendix 3  1200  1600  2000  = **>  105  Stave West Sample Annual Sample Period Irradiance over Depth  Stave West 2001: PAR vs Depth  Stave West 2002: PAR vs Depth  Irradiance (umol quanta/m2/sec) 0  400  800  1200  1600  2000  Irradiance (umol quanta/m2/sec) 0 0 »  400 KP  1  800  1200  1600  2000  TM—-am  - • — 4-Apr -•—25- May 5-Jul 30-Jul -•—1-Nov  Appendix 3  106  APPENDIX 4 Hayward 2003 Temperature Profiles over Depth Hayward 9-May-03  Hayward 27-Mar-03 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 • 16 17 • 18 • 19 20 -  2  4  6  8 10 12 14 16 18 20 22 24  0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  2  4  6  • • • • -  0  2  4  6  8 10 12 14 16 18 20 22 24  0  2  4  6  Appendix 4  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  2  4  6  0  2  4  6  8 10 12 14 16 18 20 22 24  0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  • • • • • • •  2  4  6  8 10 12 14 16 18 20 22 24  10 12 14 16 18 20 22 24  Hayward 9-Sep-03 0  2  4  6  8 10 12 14 16 18 20 22 24  Hayward 29-Oct-03  Hayward 30-Sep-03  8 10 12 14 16 18 20 22 24  8  • • -  Hayward 7-Aug-03  Hayward 12-Sep-03 0i 1 2 3 4 5 6 7 8 9 10 11 J 12 J 13 i 14 -J 15 16 17 18 19 20 -  0  8 10 12 14 16 18 20 22 24  I  Hayward 18-Jul-03  Hayward 18-Jun-03  0  2  4  6  8 10 12 14 16 18 20 22 24  123•4 56789 10 11 12 13 14 15 16 17 18 19 20 -  107  Hayward 10-NOV-03 0  0  2  -t—  1 2 3 4 5 6 7 8 9  1  4 1  6  8 '  1  14 16 18 1  1  1  1  20  22 24  1  • C I  '> 1 .  11  • • 11  12  ••  20 J  1  <>  ••  13 14 15 16 17 18 19  1  0  10 11  10 12  •>  < 1 <> <> • 1 •> •• • 1  Appendix 4  4  1  108  Stave 2003 Temperature Profiles over Depth Temperature (°C) Stave 27-Mar-03 0 2  4  6 8 10 12 14 16 18 20 22 24  0  2  4  6  Stave 18-Jul-03 1  1  8 10 12 14 16 18 20 22 24  0  2  4  6  0  2  4  6  8 10 12 14 16 18 20 22 24  8 10 12 14 16 18 20 22 24  Stave 13-Aug-03  Stave 7-Aug-03  0 2 4 6 8 10 12 14 16 18 20 22 24 1  Stave 18-Jun-03  Stave 9-May-03  0  2  4  6  8 10 12 14 16 18 20 22 24  1  • ••  /-  1  =  4  \  =  Stave 9-Sep-03 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  0  2 4 6 8 10 12 14 16 18 20 22 24  • -  Appendix 4  Stave 18-Sep-03  Stave 12-Sep-03 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  • • • • • •  2 4  6 8 10 12 14 16 18 20 22 24  0  2  4  6  8 10 12 14 16 18 20 22 24  0 1 2  34-  4 56 • 7 8 9 10 11 • 12 13 14 • 15 • 16 17 18 19 • 20 •  109  Temperature (°C) Stave 30-Sep-03 10 12 14 16 18 20 22 24  Stave 29-Oct-03 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -  Appendix 4  2  4  6  8 10 12 14 16 18 20 22 24  Stave 10-Nov-03 0  2 4 6 8 10 12 14 16 18 20 22 24  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -  110  APPENDIX 5  Stave Reservoir: Summary Water Level Data  Average water level during sampling period (m asi)  Sample Date  14-Feb 12-Apr 23-May 7-Jui 15-Aug 27-Sep 18-Nov 20-Feb 4-Apr 24-May 5-Jul 30-Jul 21-Sep 1-Nov 25-Feb 3-Apr 13-May 24-Jun 22-Jul 4-Sep 25-Oct  plate instal  2000-2 2000-3 2000-4 2000-5 2000-6 2000-7 2001-1 2001-2 2001-3 2001-4 2001-5 2001-6 2001-7 2002-1 2002-2 2002-3 2002-4 2002-5 2002-6 2002-7  77.3 76.51 78.93 81.51 80.49 78.37 76.71 76.88 77.09 79.60 81.51 81.42 80.39 77.87 79.55 76.55 78.13 81.05 81.23 80.62 76.19  AVERAGES  Spring  77.72  Summer  81.00  Fall Winter  77.54 76.88  Spring  78.35  Summer  81.47  Fall Winter  79.13 79.55  Spring  77.34  Summer  81.14  Fall  78.41  water level change  -0.79 2.42 2.59 -1.02 -2.12 -1.66 0.17 0.22 2.51 1.91 -0.09 -1.03 -2.52 1.68 -3.00 1.57 2.92 0.19 -0.61 -4.43  Stave/Hayward Plate Elevation Data Mate elevation (m a.s.l.) Stave Average  80 78 76 74 72 71 69 67 65 63  Appendix 5  North  South  West  Hayward  80.1 77.8 76.5 74.3 72.5 70.7 69.3 67.4 65.5 63.1  79.1 77.8 76.3 74.3 72.4 71.8 69.5 67.7 65.8 64.9  79.5 77.8 76.3 73.7 71.9 70.1 67.7 65.8 63.7 61.9  42.1 40.3 38.8 36.3 34.5 33.3 30.9 28.9  111  Hayward Reservoir: Summary Water Level Data Average water level during sampling period (m asi)  Sample Date  14-Feb 12-Apr 23-May 7-Jul 15-Aug 27-Sep 18-Nov 20-Feb 4-Apr 24-May 5-Jul 30-Jul 21-Sep 1-Nov 25-Feb 3-Apr 13-May 24-Jun 22-Jul 4-Sep 25-Oct  42.22 42.30 42.32 42.15 42.14 42.39 42.36 42.28 42.26 42.44 40.81 42.44 42.28 42.28 42.29 42.40 42.41 42.34 42.33 42.42 42.36  plate instal  2000-2 2000-3 2000-4 2000-5 2000-6 2000-7 2001-1 2001-2 2001-3 2001-4 2001-5 2001-6 2001-7 2002-1 2002-2 2002-3 2002-4 2002-5 2002-6 2002-7  AVERAGES  Spring  42.31  Summer  42.15  Fall Winter  42.38 42.88  Spring  42.35  Summer  41.62  Fall Winter  42.28 42.29  Spring  42.40  Summer  42.34  Fall  42.39  water level change  0.08 0.02 -0.17 -0.01 0.25 -0.03 -0.08 -0.02 0.18 -1.64 1.64 -0.16 0.00 0.01 0.10 0.02 -0.07 -0.01 0.08 -0.05  Hayward Reservoir: Summary Water Elevation June 2001 Drawdown  drawdowns Hayward Draw Down: June 2001 Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir Hayward Lake Reservoir  APPENDIX 5  9/Jun 10/Jun 11/Jun 12/Jun 13/Jun 14/Jun 15/Jun 16/Jun 17/Jun 18/Jun 19/Jun  42.546 42.507 42.042 42.249 42.37 41.784 41.54 41.54 42.048 41.442 41.794  2001 41.813 38.56 35.316 33.254 33.391 33.46 33.118 34.163 37.759 41.989 42.595  42.288 42.35 42.438 42.498 42.216 42.294 42.268 42.368 42.569 42.658 41.987  112  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items