Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Physical habitat below a hydropeaking dam : examining progressive downstream change Winterhalt, Lesley Marie 2015

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

Item Metadata


24-ubc_2015_september_winterhalt_lesley.pdf [ 18.27MB ]
JSON: 24-1.0166537.json
JSON-LD: 24-1.0166537-ld.json
RDF/XML (Pretty): 24-1.0166537-rdf.xml
RDF/JSON: 24-1.0166537-rdf.json
Turtle: 24-1.0166537-turtle.txt
N-Triples: 24-1.0166537-rdf-ntriples.txt
Original Record: 24-1.0166537-source.json
Full Text

Full Text

Physical habitat below a hydropeaking dam: Examiningprogressive downstream changebyLesley Marie WinterhaltB.Sc. (Hons), McGill University, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENTOF THE REQUIREMENTS FOR THE DEGREE OFMaster of ScienceinTHE FACULTY OF GRADUATE AND POSTDOCTORALSTUDIES(Geography)The University of British Columbia(Vancouver)August 2015c© Lesley Marie Winterhalt, 2015AbstractThis study examines the short-term physical habitat conditions at four sites on theKananaskis River, Alberta, where a hydropeaking dam was installed in 1955. Thisdam imposes both the approximate pre-dam minimum flow, and the pre-dam flood(from a small flood year) on a daily basis. The purpose of this study was to ex-amine the extent of daily changes in physical habitat conditions that organisms inthe stream would have to endure, and the extent to which these fluctuations mightbe reduced downstream due to distance from the dam and unregulated tributaryinfluence. Physical habitat conditions monitored over low flow and high flow damreleases were: velocity; depth; bed mobility; ramping rates; and total suspendedsolids. River2D was used to calculate weighted usable area and potential habitat forBrown Trout (fry, juveniles and adults) and Mountain Whitefish (fry, juveniles andadults) at the low and high flow conditions. Of the factors examined, only rampingrates and total suspended solids showed signs of downstream attenuation. Differ-ences in depth, velocity, weighted usable area, and potential habitat between lowflow and high flow dam releases were variable, and showed no downstream pattern.Between low and high flow releases, significant (p = 0.05) changes in depth wereobserved at all sites, and significant changes (p = 0.05) in velocity were observed atall but the second site. The second site also saw the smallest changes in measuresof habitat between low flow and high flow dam releases; however, all other sitessaw median differences of 48.1% to 170.9%. Percent differences in habitat be-tween low and high flow dam releases ranged from 2.6% (second downstream site,juvenile Brown Trout) to 193.3% (third downstream site, adult Mountain White-fish). These habitat changes happen more often than before the dam was installed(many times weekly vs. about once a year during the spring freshet) and they oc-iicur more rapidly. Because these changes happen at times of the year that are out ofsynchronization with the biota of the river, and as these changes are extreme, thisimplies challenging physical habitat conditions for indigenous stream biota.iiiPrefaceB. Eaton, M. Lapointe and the HydroNet research group selected the river and damfor study. B. Eaton assisted with designing field and data analysis methods. Allother work was completed by the Author, L. Winterhalt.ivTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixGlossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Hydropeaking flows . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Hydraulic modelling . . . . . . . . . . . . . . . . . . . . . . . . 91.2.1 Hydraulic models and habitat suitability indices . . . . . . 111.3 Downstream trends on impounded rivers . . . . . . . . . . . . . . 121.4 Research objectives . . . . . . . . . . . . . . . . . . . . . . . . . 132 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142.1 Kananaskis River . . . . . . . . . . . . . . . . . . . . . . . . . . 142.1.1 Dams on Kananaskis River . . . . . . . . . . . . . . . . . 142.1.2 Geography . . . . . . . . . . . . . . . . . . . . . . . . . 142.1.3 Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.1.4 Flow regime . . . . . . . . . . . . . . . . . . . . . . . . 17v2.2 Study sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Changes in physical stream characteristics . . . . . . . . . . . . . . 243.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.1.1 Serial discontinuity concept . . . . . . . . . . . . . . . . 243.1.2 Bed mobility . . . . . . . . . . . . . . . . . . . . . . . . 253.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263.2.1 Instream measurements . . . . . . . . . . . . . . . . . . . 273.2.2 River2D modelling . . . . . . . . . . . . . . . . . . . . . 313.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343.3.1 Stream hydrology . . . . . . . . . . . . . . . . . . . . . . 343.3.2 Bed mobility . . . . . . . . . . . . . . . . . . . . . . . . 383.3.3 Modelled depth and velocity . . . . . . . . . . . . . . . . 383.3.4 Total suspended solids . . . . . . . . . . . . . . . . . . . 483.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493.4.1 Ramping rates . . . . . . . . . . . . . . . . . . . . . . . 493.4.2 Bed mobility . . . . . . . . . . . . . . . . . . . . . . . . 533.4.3 Depth and velocity . . . . . . . . . . . . . . . . . . . . . 543.4.4 Total suspended solids . . . . . . . . . . . . . . . . . . . 543.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554 Downstream fish habitat . . . . . . . . . . . . . . . . . . . . . . . . . 564.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.2.1 Habitat modelling . . . . . . . . . . . . . . . . . . . . . 574.2.2 Potential habitat . . . . . . . . . . . . . . . . . . . . . . 584.2.3 Change in habitat between flow conditions . . . . . . . . 584.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.4.1 Habitat reductions . . . . . . . . . . . . . . . . . . . . . 614.4.2 Serial discontinuity concept . . . . . . . . . . . . . . . . 624.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64vi5.1 Observed longitudinal trends . . . . . . . . . . . . . . . . . . . . 645.2 Elements not displaying downstream longitudinal trends . . . . . 655.3 River recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67viiList of TablesTable 1.1 Summary of the effects of rapidly varying flow on fish. . . . . 3Table 2.1 Indicators of hydrologic alteration for Kananaskis River follow-ing flow regulation. . . . . . . . . . . . . . . . . . . . . . . . 18Table 3.1 River2D model parameters . . . . . . . . . . . . . . . . . . . 33Table 3.2 Initial up-ramping and down-ramping rates during a high flowrelease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Table 3.3 Up-ramping and down-ramping rates for an entire high flow re-lease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Table 3.4 Estimates of bed mobility after one flood cycle. . . . . . . . . 38Table 3.5 Median grain size for study sites along Kananaskis River . . . 39viiiList of FiguresFigure 2.1 Kananaskis Lakes . . . . . . . . . . . . . . . . . . . . . . . . 15Figure 2.2 Average monthly air temperature and precipitation values forKananaskis Pocaterra weather station, 1981-2007. . . . . . . 16Figure 2.3 All daily discharges at Water Survey of Canada gauge Pocaterrafor pre-dam (1931-1935) and post-dam (1975-2009) years. . . 19Figure 2.4 Daily flows averaged for pre-dam and post-dam conditions. . . 20Figure 2.5 Annual peak discharges for 1931 to 2009. . . . . . . . . . . . 21Figure 2.6 Annual minimum discharges for 1931 to 2009. . . . . . . . . 22Figure 2.7 Study sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Figure 3.1 Water level logger. . . . . . . . . . . . . . . . . . . . . . . . 26Figure 3.2 Well with water level logger. . . . . . . . . . . . . . . . . . . 28Figure 3.3 Surveying stream bed topography at the Opal site. . . . . . . . 29Figure 3.4 Surveying stream bed topography at the Ribbon site. . . . . . 29Figure 3.5 Velocity measurements made by ADCP at the Opal site. . . . 30Figure 3.6 Hydrograph for each study site of a high flow release from thePocaterra Dam on August 5-6, 2011. . . . . . . . . . . . . . . 35Figure 3.7 Summer 2011 hydrographs for all study sites. . . . . . . . . . 37Figure 3.8 2D velocity map for the Pocaterra study site. . . . . . . . . . 40Figure 3.9 2D depth map for the Pocaterra study site. . . . . . . . . . . . 41Figure 3.10 2D velocity map for the Opal study site. . . . . . . . . . . . . 42Figure 3.11 2D depth map for the Opal study site. . . . . . . . . . . . . . 43Figure 3.12 2D velocity map for the Galatea study site. . . . . . . . . . . 44Figure 3.13 2D depth map for the Galatea study site. . . . . . . . . . . . . 45Figure 3.14 2D velocity map for the Ribbon study site. . . . . . . . . . . . 46ixFigure 3.15 2D depth map for the Ribbon study site. . . . . . . . . . . . . 47Figure 3.16 Boxplot of depths at high and low flow for the four study sites. 49Figure 3.17 Boxplot of the changes in depth between low and high flow atthe four study sites. . . . . . . . . . . . . . . . . . . . . . . . 50Figure 3.18 Boxplot of velocities at high and low flow for the four studysites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Figure 3.19 Boxplot of the changes in velocity between low and high flowat the four study sites. . . . . . . . . . . . . . . . . . . . . . 52Figure 3.20 Total suspended solids at all study sites, at high and low flow. 53Figure 4.1 Potential habitat per m stream length - Mountain Whitefish. . 59Figure 4.2 WUA per m stream length - Mountain Whitefish. . . . . . . . 59Figure 4.3 Potential habitat per m stream length - Brown Trout. . . . . . 60Figure 4.4 WUA per m stream length - Brown Trout. . . . . . . . . . . . 60Figure 4.5 Percent difference of habitat between the high and low flowconditions. Each site represents the fry, juvenile and adult lifestages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61xGlossaryADCP acoustic doppler current profilerPVC polyvinyl chlorideD50 median grain sizeTSS total suspended solidsSDC serial discontinuity conceptIHA indicators of hydrologic alterationWUA weighted usable areaIFIM instream flow incremental methodologyUSGS United States Geological SurveyHSI habitat suitability indexPHABSIM physical habitat simulationxiAcknowledgmentsFirstly, I would like to thank Dr. Brett Eaton for his endless support. His technicalexpertise, generosity with his time, and encouragement made him an excellentmentor, without whom this project would not have been possible. Thanks also tocommittee member Dr. John Richardson, whose advice, comments and guidancehelped bring important ideas and direction to this project.Thanks are due to field and lab assistants Aaron Tamminga, Tyler McDivittVandermolen, Alistair Davis and Byeong Kim; and the rest of Brett Eaton’s labgroup, Dan McParland, Holly Buehler and Sarah Davidson for their help alongthe way. I would like to acknowledge the rest of the UBC Geography cohort whoprovided much needed technical R, Matlab and LaTex help, and Shreejoy Tripathyfor the statistics crash course.I would also like to acknowledge friends and family for their emotional support.My parents, for high expectations and raising me in an environment that valued andencouraged higher education. Paul and Elaine for providing me with a beautifulplace to write. Alyssa Stryker, Alyssa Salaciak, Lauren Boivin, and Andrea Reidfor commiserating with, and encouraging me. Sara Hopkins for help during thefinal stages. Ilana Klinghoffer, for helping me in every way that you possiblycould, through every single stage of this project. And finally to Tyler, for theinfinite patience, love, encouragement and perspective - thanks for never giving upon me.Funding for this project was provided by the HydroNet Consortium NSERCresearch network.xiiChapter 1IntroductionFish living downstream of hydropeaking dams are affected by dam operations. Hy-dropeaking dams (dams with rapidly varied flow, adjusted to meet consumer energydemands) change the flow regime of a river drastically, modifying the magnitude,frequency, duration, timing and rate of change of flows. Physical habitat is altered,and the relationship between fish population fitness and available habitat has beenwell studied. To date, the study of hydropeaking power has been overwhelminglycentered on the biological responses, with less study on how short-scale tempo-ral variations in flow may influence habitat. Further, how progressive downstreamunregulated tributary inflow may alter fish habitat under hydropeaking regimes,can help to understand how organisms respond to regulated rivers at increasingdistances downstream.1.1 Hydropeaking flowsThe discharge fluctuations observed below hydropeaking dams are extreme, andwhen dealing with river restoration efforts, it is common to focus on reducing theextent of hydropeaking. However, at present, direct empirical support of its effec-tiveness is limited (Korman and Campana, 2009). Hydropeaking dams affect eachaspect of the natural flow regime. On average, they: (1) reduce the volume of thehighest, and possibly the lowest, discharge events; (2) increase the frequency whiledecreasing the duration of extreme flows; (3) obscure the seasonal timing of key1flow rates; and (4) increase the incidence of rapid rates of change to a near-dailyoccurrence. The following will review the effects of highly varied flow on fish andaquatic invertebrates.Many studies have examined the effects of rapidly varying flow on fish popu-lations. Many of these have specifically examined fish below hydropeaking dams(e.g. Heggenes, 1988; Bunt et al., 1999; Korman and Campana, 2009) while othershave examined flood flows (e.g. Seegrist and Gard, 1972; Erman et al., 1988). Allthese studies examining small, young fish found that they were more susceptibleto large, rapidly varying flows, and often sought refuge in woody debris or thesubstrate. Generally, detrimental effects of hydropeaking dams were more notice-able directly below the dam, with effects diminishing downstream. When exam-ining physical habitat features in relation to fish populations, depth was the mostcommonly studied feature, while cover type, wetted area, velocity, and substratesize were also commonly investigated features. Physical habitat made referenceto hydraulic geometry relationships between discharge, depth, velocity and wettedperimeter. However, most studies did not make mention of other geomorphologicproperties, mainly sediment transport and entrainment of bed material under highlyvaried flows. An exception to this is the Erman et al. (1988) study that linked mo-bility of the bed to the mechanical crushing of age-0 fish. Table 1.1 provides asummary of the results of selected studies.2Table 1.1: Summary of the effects of rapidly varying flow on fish.Methods and Relevant results Studyhabitat variablesMeasure 1. Age-0 rainbow trout did not maintain their position within Korman and Campana (2009)near shore immediate shoreline areas when flows were high.habitat use 2. Catch rates 2 to 4 times higher at daily minimum flow comparedto daily maximum flow in nearshore areas.Rainbow 3. Otolith growth larger on Sundays, where low flows allowed fishtrout (age-0) to be near shore areas with higher water temperatures and lowervelocities.Reviewed 326 1. Modified flows affect fish and habitat, but the sign of the Murchie et al. (2008)articles on flow, correlation and the severity vary from study to and fish 2. Newly emerged fry more vulnerable to stranding becausehabitat of use of substrate as cover in shallow water habitats andlimited swimming ability.3. Thermal relationship examined in only 43.3% in studies(temperature is known to vary with flow).Examined use of 1. Use of cover and pools increased as flow increased. Bunt et al. (1999)(1) cover, (2) pools, 2. Similar depths were used at high and low flow (63 cm).Continued on next page3Table 1.1 – continued from previous pageMethods and Relevant results Studyhabitat variables(3) wetted area and 3. Used lower surface water velocities at high than low(4) depth by Brown flow.Trout 4. Woody debris and roots used more during high flows.5. Woody debris most common cover type, and appears toprovide refuge from large fluctuating flows.Reference 1. Fish no more common after reductions in extremes from Cowx et al. (1998)conditions: hydropeaking. The lack of a significant difference in fishHydropeaking numbers may be due to the relatively short study period.(1992, year prior totreatment).Treatment:reduction in extentof hydropeaking(measured twoyears post, in1994).Examined fish 1. Small fish did worse in hydropeaking environments. Bain et al. (1988)Continued on next page4Table 1.1 – continued from previous pageMethods and Relevant results Studyhabitat variablescommunity 2. Species using broad range of habitat higher in density instructure as a regulated vs control system.function of:(1) Water depth(2) Current velocity(3) Substratecoarseness(4) SubstrateheterogeneityExamined effects 1. Peaking flows did not displace Brown Trout downstream Heggenes (1988)of (1) Depth 2. Coarse substrate thought to be crucial in creating low water(2) Cover velocity microniches.(3) Velocity(4) Substrate onBrown TroutExamined effects 1. Hydraulic refugia from high flows crucial. Valentin et al. (1996)of (1) Depth 2. WUA does not follow strict relationship with discharge.Continued on next page5Table 1.1 – continued from previous pageMethods and Relevant results Studyhabitat variables(2) Velocity Varies greatly spatially and temporally under hydropeaking(3) Weighted dams.usable area onBrown TroutObserved fish 1. Fewer fish immediately downstream of dam (within first Moog (1993)fauna composition few km).immediately below 2. Fish numbers low further downstream (20-40 km) but not asa dam (a few km) severely reduced, however fewer than reference stream.and further 3. Increasing discharge range corresponded with significantdownstream (20-40 decrease in fish effect of 1. Brook Trout and Paiute Sculpin age-0 individuals killed Erman et al. (1988)winter floods on during winter floods due to mobility of bed causingBrook Trout and mechanical grinding or crushing.Paiute Sculpin age-0 class. (Nothydropeaking, butContinued on next page6Table 1.1 – continued from previous pageMethods and Relevant results Studyhabitat variableslarge, rapid changein flow)Examined effects 1. Winter floods decimated developing eggs of fall spawning Seegrist and Gard (1972)of winter floods on Brook Trout.Rainbow and 2. Spring floods destroyed spring-spawned Rainbow TroutBrook Trout. (Not eggs.hydropeaking, but 3. Influence of floods on adults less pronounced andlarge, rapid change flow)Review articles on 1. More than 10 studies reviewed observed organism (fish Cushman (1985)effects of rapidly or invertebrates) stranded as water levels drop.varying flow Subsequently, organisms may die due to dewatering ofpools, lack of food, low DO, high temperatures, andpredation.2. Fish eggs may be in dewatered areas, and may perish dueto thermal stress, insufficient oxygen or desiccation.3. Cushman highlights need to know channel morphometryContinued on next page7Table 1.1 – continued from previous pageMethods and Relevant results Studyhabitat variablesso that velocity, depth, width and wetted perimeter can bedetermined.4. Discusses need to understand channel composition sothat erosion and sediment potential can be made. Alsodiscusses bank erosion.Examined effect of 1. Fish more resilient to flash floods with increased Pearsons et al. (1992)flash flood on hydraulic complexity.structural fish 2. Fish may be washed out under flood flows.assemblages8Prior to the 1980s, little focus was given to the importance of channel mor-phology in determining the ecological integrity of riverine systems (Nowell andJumars, 1984). Recently, more research has been conducted in this area, in a shifttowards more holistic and environmentally sensitive river management (Maddock,1999). Studies have attempted to determine the effect of geomorphology on fish,with descriptors of geomorphology including such factors as elevation, channelslope, wetted width, depth, proportion of reach with deep-pool habitat, and percentcomposition of various substrate materials (e.g. Bain et al., 1988; Valentin et al.,1996; Bunt et al., 1999; Quist et al., 2004). While the effects of many geomorphicfactors on fish have been studied, there are discrepancies in terms of the relative im-portance of such factors when different fish, rivers, and experimental methods areconsidered. However, of the geomorphic factors considered, the most commonlyexamined and agreed upon aspects are the velocity, depth, substrate material, andaccess/proximity to cover.Relationships between depth, velocity, substrate and discharge have been largelystudied, and while these interactions are still not perfectly understood, models arebeing developed to help to quantify these relationships (e.g. Eaton and Church,2007; McParland et al., in press). Furthering these models is useful for predict-ing fish habitat, the relative role of dams and the quantity and quality of availablehabitat.In addition to factors previously discussed, flow alterations from a dam canaffect nearly all aspects of a river. Those that are often studied include: tempera-ture (e.g. Camargo and Voelz, 1998; Cereghino et al., 2002; Paller and Saul, 1996;Preece and Jones, 2002; Saltveit et al., 1994); physicochemical properties (e.g.Byren and Davies, 1989; Storey et al., 1991; O’Keeffe et al., 1990); algae, inver-tebrate and fish abundance and species composition (e.g. Rehn, 2009; Cereghinoet al., 2002; Patterson and Smokorowski, 2011; Jones, 2013; Cortes et al., 2002).1.2 Hydraulic modellingWhile the uses of freshwater resources span a large spectrum in Canada, two per-vasive and conflicting goals are to maximize power generation profits for economicbenefit and to maintain the ecological integrity of a system for biological and soci-9etal good. In order to balance these goals, managers have sought tools which enablethem to quantify habitat preservation at different flows in order to decide the min-imum flow necessary for the ecosystem, and then to quantify the ‘surplus’ waterthat can be used for economic activities. Largely, the instream flow incrementalmethodology (IFIM) and hydraulic habitat modelling are used to set minimum bi-ological flow requirements. As this method is one of the most commonly applied,it is important to understand the methodology, along with shortcomings, if thesemethods are used to make important water management decisions. This sectionwill examine the IFIM and hydraulic modelling methodologies.Implemented 1970, the American National Environmental Policy Act forcedregulators to go beyond the previous minimum flow requirements in regulatedrivers, and instead examine the relationship between discharge and life-stage-specifichabitat needs (Stalnaker et al., 1995). The IFIM was subsequently developed by theU.S Fish and Wildlife Service as a framework for quantifying this relationship.The ultimate goal of the IFIM was to make more informed management decisionsin terms of the allocation of water resources among different users. Like all models,IFIM simplifies true conditions, and as such certain assumptions are made in orderfor the model to be applicable. One important assumption is that physical habi-tat is the key limiting factor for fish, and that any changes in physical habitat aresolely related to changes in flow (Maughan and Barrett, 1992). This premise facil-itates the further study of physical habitat, which is divided into macro- and micro-habitat variables. Under IFIM, the micro-habitat variables measured are: watervelocity; water depths; instream objects such as cover; and bottom substrate mate-rials (Bovee, 1982). The important macro-habitat variables are: water temperature;dissolved oxygen; total alkalinity; turbidity; and light penetration through the wa-ter column (Bovee, 1982). Both macro- and micro- habitat variables are influencedby discharge, and these changes can be assessed through modelling. According toUnited States Geological Survey (USGS) guidelines for IFIM, the amount of usablehabitat is determined by a) the length of usable stream habitat (length of stream fit-ting required macro-habitat variables) and b) the habitat area per stream length (asdetermined by micro-habitat variables). IFIM does not provide a method to assessmicro-habitat, but rather establishes the framework for precise modelling tools tobe employed. It is often hydraulic models that are used to investigate macro-habitat10elements as a function of discharge (Bovee, 1982).1.2.1 Hydraulic models and habitat suitability indicesHydraulic models are a logical second step following IFIM. They allow physicalchannel characteristics (depth, velocity, substrate) to be calculated as a function ofdischarge. However, modelling these characteristics alone does not allow a man-ager to determine the available aquatic habitat. To do so, physical hydraulic mod-els must be combined with biological indicators of acceptable physical habitat. Ahabitat suitability index (HSI) is often used to link these biological needs with phys-ical habitat. HSI measures range from 0 for unsuitable habitat to 1 for ideal habitatconditions. Each fish species and life stage usually have different physical habitatneeds, and therefore will have their own specific HSI.A HSI may be developed using one of three different methods: a review ofthe literature; empirical frequency analyses; or habitat preference curves (Stier andCrance, 1985). Literature reviews will examine peer-reviewed articles pertainingto the target species at the target life stage. Ideally, HSIs developed from litera-ture reviews will preferentially use case studies from areas near that of interest, asdifferent populations within the same species may exhibit different habitat prefer-ences. Reviews often seek to identify important habitat variables in relation to (1)where the target species was found, (2) where it was in highest densities and (3)where it grew the fastest (Stier and Crance, 1985). This method assumes that moresuitable habitat will have target fish present, support higher densities and supporthigher growth rates.The second method, empirical frequency analyses, involves investigating thefrequency of fish observed at sites within the stream of interest and may be prefer-ential to literature reviews. This method indicates the habitat that is chosen whengiven other habitat options (while literature reviews may not make mention of otheravailable habitats). Empirical frequency analyses may also be referred to as utiliza-tion curves. The third method, which is to develop preference curves, is similar toempirical frequency analysis. However, corrections are made for environmental bi-ases. This is done by taking into account not only a species’ preference for a givenhabitat type, but also the availability of that habitat in the environment. For exam-11ple, if a river is composed of 5% pools by volume, and target species are sightedin pools 50% of the time, preference curves would correct for the low percentageof pools. A simple frequency analysis of this situation might indicate that there isno preference for pools (50% in pools versus 50% not in pools). However, the rareoccurrence of pools in the stream would indicate that there is a strong preferencefor this habitat type, and a HSI developed based on a preference curve corrects forthis.Different hydraulic models exist to map suitable habitat and determine overallweighted usable area (WUA). Most common in the United States is the physicalhabitat simulation (PHABSIM) model, developed in conjunction with IFIM. Thismodel is also commonly used in Canada, Europe and New Zealand (e.g. Ghanemet al., 1996; Moir et al., 2005; Hudson et al., 2003). While developed to workalongside IFIM, PHABSIM models often focus solely on the implementation ofHSIs relating to preferences in depth, velocity and substrate while ignoring macro-habitat variables (Spence and Hickley, 2000).1.3 Downstream trends on impounded riversBelow hydropeaking facilities, there exists a range of physical habitat conditions,as peak discharge pulses progressively disperse downstream and unregulated trib-utaries play increasing roles. Downstream variability in invertebrate drift and fishpopulations below hydropeaking facilities have been examined. For example, lon-gitudinal changes in fish communities, from bottom-dwelling fish in fast-flowingwaters to more generalist species downstream have been observed (Vehanen et al.,2005). In contrast, a 9-fold increase in drift with 7-fold increase in flow was ob-served 250 m downstream from a hydropeaking dam, with similar drift levels 8km downstream (Bruno et al., 2010). However, little research has been conductedto quantify a gradient of physical habitat downstream, particularly in reference tounregulated tributary influence. A study by Valentin et al. (1996) refers to the im-portance of specific site morphology in terms of predictions of changes in physicalhabitat, but this study does not address spatial distribution of morphologic differ-ences.121.4 Research objectivesThe first goal of this research is to quantify the scope of physical changes in theriver, as defined by depth, velocity, bed mobility and total suspended solids (TSS)between low and high flow pulses from a hydropeaking dam and how progressivedistance downstream from the dam may alter these changes. The second goal ofthis research is to quantify changes in WUA between the low and high flow condi-tions for specific fish species, with increasing distance below a hydropeaking damand increasing unregulated tributary inputs. This research will help to understandthe mechanisms by which organisms are affected by hydropeaking operation strate-gies, and if distance downstream and unregulated tributaries decrease variations inphysical habitat conditions between low flow and high flow dam releases.13Chapter 2Study area2.1 Kananaskis River2.1.1 Dams on Kananaskis RiverThe section of Kananaskis River between Pocaterra Dam and Barrier Reservoir isthe subject of this study. Kananaskis River flows 74 km in length from its headwa-ter lakes in the Rocky Mountains to its confluence with Bow River in the foothills,upstream of Calgary, Alberta. There are four dams that impact the Kananaskis Sys-tem: the Interlakes Dam (1955) between the Upper and Lower Kananaskis Lakes;Pocaterra Dam (1956) downtream of the Lower Kananaskis Lake and at the begin-ning of the Kananaskis River; Barrier Dam (1947); and Kananaskis Dam (1913)which dams the Bow River immediately downstream of its confluence with theKananaskis. The Pocaterra hydro plant is capable of producing 15 MW of power,and produces 29,000 megawatt hours per year (TransAlta, 2014). Releases fromthe Pocaterra Dam are 0.5 m3/s at base flow and 20 to 23.6 m3/s during releases.2.1.2 GeographyThe Kananaskis catchment is very mountainous, with many features evident of pastglaciation, including cirques, U-shaped valleys, and moraines. While small, a fewglaciers currently remain in the basin. The rock type is predominantly limestone,14Figure 2.1: Upper (right) and Lower (left) Kananaskis Lakes, with the Inter-lakes Dam between the two.with large thrust faults as occur elsewhere in the Canadian Rockies. The LowerKananaskis Lake, from which the Kananaskis River begins, sits at an elevation of1680 m asl. Towards the lower end of the river, Barrier Lake is at an elevationof 1380 m asl, and the mouth of the river is at 1280 m asl. Mountain Peaks inthe catchment can reach elevations of 3000 m asl. The Kananaskis River varies insinuosity through its course from tortuous meanders to irregular sinuosity. On oc-casion it is confined by debris fans or bedrock, but also has large sections of widervalley with wetlands immediately adjacent to the river. The valley bottom consistsmostly of forest, while high elevation areas are largely exposed alpine rock, withsome alpine meadows above treeline. The Kananaskis River has a drainage basinarea of 933 km2, of which 899 km2 lies above the Barrier Dam, and with 362 km2accounted for above Pocaterra Dam.15Figure 2.2: Average monthly air temperature and precipitation values forKananaskis Pocaterra weather station, 1981-2007. Precipitation meansrepresented by dark blue bars. Monthly maximums, means and mini-mums represented by red, green and blue lines respectively.2.1.3 ClimateClimatic variables are measured near the Pocaterra Dam at the Environment Canadaweather station Kananaskis Pocaterra (ID 3053604). The climate station is locatedat 50◦ 42’ 45.020” N, 115◦ 07’ 12.060” W at 1610 m asl. This climate stationprovides information on valley bottom conditions. The average minimum dailytemperature for the month of January, the coldest month of the year, is -16.5 ◦C.The average maximum daily temperature for July, the hottest month of the year, is20.6◦C (see Fig. 2.2). Average annual precipitation is 568 mm, with 255 mm ofthat falling as snow (Environment Canada, 2015). In the winter, more snowfall isrecorded at higher elevations than in the valley bottom, so annual precipitation forthe entire basin is somewhat higher.162.1.4 Flow regimeRiver regulation by hydroelectric dams can have profound effects on the physicaland biological components of aquatic ecosystems. One mechanism by which thesechanges are induced is the flow regime. The flow regime consists of five key el-ements: the magnitude; timing; duration; frequency; and rate of change. Each ofthese five aspects has important implications for the physical structure and biotaof a river ecosystem. For example, high flows are important for flushing fine grainsediment from spawning habitat of fish (Beschta and Jackson, 1979) and the timingof these high flows may also be an important biological cue for spawning (Mont-gomery et al., 1983) and egg hatching (Naesje et al., 1995). Poff et al. (1997)provides a summary of the multitude of ecological responses to the five aspects ofthe flow regime.Richter et al. (1996) identified five groups of indicators of hydrologic alter-ation (IHA). These five groups are: (1) Magnitude of monthly water condition; (2)magnitude and duration of annual extreme water conditions; (3) timing of annualextreme water conditions; (4) frequency and duration of high and low pulses; and(5) rate and frequency of water condition changes.Pre- and post-Pocaterra DamIHA were used to assess changes on the Kananaskis River below Pocaterra Dam.Mean daily discharge data was acquired online from the Water Survey of Canadafor the Pocaterra gauging station (station ID 05BF003). Discharge data coversa pre-dam period (1931-1955) and a post-dam period (1975-2009) with a 20-yeargap following the construction of the dam. Because data are averaged daily, groups4 (frequency and duration of high and low pulses) and 5 (rate and frequency ofwater condition changes) of the IHA are not used for this assessment. Due to thehydropeaking operation of the Pocaterra Dam, daily mean data are not sensitiveenough for all suggested analyses.The operation of the Pocaterra hydropeaking dam has had significant effects tothe hydrologic regime of the Kananaskis River. A summary of the results of theIHA is found in Table 2.1. Differences from pre- to post- dam conditions rangedfrom -183% (1-day minimum discharge) to 67% (mean March discharge). Of IHA17Table 2.1: Indicators of hydrologic alteration for Kananaskis River followingflow regulation.Streamflow m3/sPre-dam Post-dam DifferenceGroup 1: Monthly magnitudeJanuary 7.75 12.08 44%February 6.76 11.35 51%March 4.91 9.88 67%April 3.34 5.91 65%May 6.53 3.29 -66%June 14.06 2.51 -139%July 13.61 3.70 -115%August 9.14 5.16 -56%September 5.99 5.29 -12%October 3.93 6.23 45%November 7.19 8.80 20%December 8.87 10.95 21%Group 2: Magnitude andduration of annual extremes1-day minimum 1.39 0.07 -183%3-day minimum 1.41 0.08 -180%7-day minimum 1.5 0.23 -146%30-day minimum 2.22 0.51 -125%90-day minimum 4.07 2.24 -58%1-day maximum 24.78 20.29 -20%3-day maximum 23.05 18.94 -20%7-day maximum 20.81 16.66 -22%30-day maximum 17.1 14.14 -19%90-day maximum 11.8 11.87 1%that were calculated, all but the 90-day maximum showed a significant difference(p = 0.05) in pre-and post- dam states. Statistics from the monthly magnitudesshowed a marked shift in seasonality. Seasonality was drastically changed. Aprilthrough October showed increased discharge from 20% to 67% after damming andSeptember through May experienced markedly decreased discharge from -12% to18Figure 2.3: All daily discharges at Water Survey of Canada gauge Pocaterrafor pre-dam (1931-1935) and post-dam (1975-2009) years.-139% after damming. This shift in seasonality is depicted in Fig. 2.3 and Fig. 2.4depicts the change in seasonality.The average annual 1-, 3-, 7-, and 30- day minimum and maximum dischargeswere all significantly lower after the installation of the Pocaterra Dam, with de-creases between 19% and 183% (see Table 2.1). The mean 1-day maximum flowevent after damming is below the mean prior to damming, however all values fallwithin 1-standard deviation of the pre-dam data (see Fig. 2.5). This range is sug-gested by Richter et al. (1997) to be an appropriate initial target for the range ofpost-dam flows, in order to maintain ecosystem integrity. In contrast, 1-day mini-mum flows post-dam fall far below the suggested range (see Fig. 2.6).2.2 Study sitesFor this research, there are four study sites on the Kananaskis River (see Fig. 2.7).The first site, called Pocaterra, is at 1.5 km from Pocaterra generating station. Site2, called Opal, is 20.3 km downstream. Site 3, called Galatea, is 24.9 km down-stream. Ribbon, the fourth site, is 35.4 km downstream of the generating station.The inflow end of Barrier Lake (the next downstream reservoir) is 46.0 km down-stream of Pocaterra generating station.19Figure 2.4: Daily flows averaged for pre-dam (blue solid line) and post-dam(red solid line) conditions. Dashed blue line represents the mean annualflow prior to dam construction and the dashed red line, the mean annualflow after dam construction.20Figure 2.5: Annual peak discharges for 1931 to 2009. The blue solid linerepresents annual peak discharge prior to dam construction, with thedashed line representing the mean. The red solid line represents annualpeak discharge after dam construction, with the dashed line representingthe mean. The dashed black lines represent one standard deviation fromthe mean pre-dam average.21Figure 2.6: Annual minimum discharges for 1931 to 2009. The blue solidline represents annual minimum discharge prior to dam construction,with the blue dashed line representing the mean. The red solid linerepresents annual minimum discharge after dam construction, with thered dashed line representing the mean. The dashed black lines representone standard deviation from the mean pre-dam average.22Figure 2.7: Kananaskis River study sites in relation to the hydropeaking dam.Flow is northward.23Chapter 3Changes in physical streamcharacteristics3.1 IntroductionHydropeaking dams result in significant alterations to the natural flow regime ofa river. These alterations can have profound effects on the biota of the aquaticecosystem and are realized by changes to physical habitat conditions. Understand-ing and quantifying these physical changes is the first step to understanding the bi-ological changes that result from hydopeaking dam operation. Additionally, quan-tifying downstream trends, as compared to examining single sites below dams, aidsin the understanding of the geographic extent of these physical habitat changes.3.1.1 Serial discontinuity conceptEarly in the study of regulated rivers, Ward and Stanford (1983) proposed the se-rial discontinuity concept (SDC), the idea that as distances downstream from damsincrease, a river’s ecological properties become more like those of unregulatedsystems. Fundamental to this theory is the idea that dams act as discontinuitieson a river, interrupting natural gradients in properties such as nutrient or thermalregimes. The SDC assumes there should be a downstream recovery of biophys-ical factors with increasing distance from a dam, due to lateral (e.g. tributary),24and vertical (i.e. hyporheic) inputs, referred to as the recovery distance or zoneof influence. While many studies, and even entire journals have been devoted tothe study of the effects of dams on rivers (e.g. Regulated Rivers: Research andManagement), relatively few studied longitudinal trends and recovery distances,specifically testing the SDC (Stanford and Ward, 2001). Ellis and Jones (2013)and Stanford and Ward (2001) reviewed previous studies in attempts to summa-rize findings related to the SDC. These review articles found studies that somerecovery gradients occur quickly (e.g. within 3.5 km for temperature (Cereghinoet al., 2002), within 3 km for nutrient levels (O’Keeffe et al., 1990)), some re-quire much greater distances (e.g. Plecoptera recovering over 80 km (Stanford andWard, 1989), Trichoptera recovering within 60-80 km (Hauer et al., 1989)) andother variables and river system combinations may show no longitudinal trends ofrecovery (e.g. invertebrates on 387 km of the Colorado River (Stevens et al., 1997),physicochemical properties over 74 km of study (Byren and Davies, 1989)). Thetrend towards recovery distances was not necessarily related to the factor studied(for example, sometimes invertebrates showed recovery and other times they didnot) and different variables may present different patterns depending on type andoperation of the dam and the downstream inputs into the river.3.1.2 Bed mobilityInterstitial space within the bed of a river is an important habitat area for manyaquatic species. Moog (1993) identified the importance of this interstitial spaceas habitat refugia against high flows for invertebrates, and Liebig et al. (2001) didthe same for fish during high-flow events. Erman et al. (1988) linked bed mobil-ity to the mechanical crushing of age-0 fish, and Death and Winterbourn (1995)linked bed instability with decreases in the number of invertebrate species. It istherefore likely that organisms are sensitive to bed mobilization during high-flowevents. Grain size distributions, and patterns of degradation and agradation belowdams are common areas of research focus (e.g. Weston, 2013), but the influence ofhydropeaking operations on bed mobilization is less commonly studied. Bed mo-bility was a variable examined by Bruno et al. (2010) below a hydropeaking dam,but in the 8 km studied, no bedload was observed.25Figure 3.1: Water level logger pre-installation.3.2 MethodsTo assess the range of change in physical properties along the length of the KananaskisRiver below the Pocaterra Dam, field measurements were conducted in the sum-mer of 2011 between May 15 and Aug 15. Velocity was measured along crosssections at the four main study sites (see Fig. 2.7). To gain a more complete under-standing of depth and velocity characteristics throughout the full lengths of fourstudy reaches, River2D (a 2-dimensional hydraulic modelling program) was usedto combine survey data with velocity and discharge data to model depths and ve-locities throughout the entirety of each site. The following will first describe theinstream measurements and secondly, the modelling process in River2D.263.2.1 Instream measurementsStream HydrologyTo address the main elements of depth and velocity as habitat variables, multiplefield techniques were used. As the physical properties of streams are highly relatedto stream flow, loggers were installed at various sites downstream of the dam totrack depth throughout the summer. Depth data could then be correlated with other,non-continuous data that were collected to extrapolate these data for the entiresummer period.Prior to spring freshet, polyvinyl chloride (PVC) wells were installed into thesubstrate and Schlumberger Mini-Diver depth loggers were deployed (see Figs. 3.1, 3.2),recording pressure readings on a 15-minute interval. Logger elevation was deter-mined using a Garmin GPSmap 76S by calibrating it at known elevation (1390 masl) at the University of Calgary Biogeoscience Institute. Another logger was in-stalled in the Kananaskis River valley at the University of Calgary BiogeoscienceInstitute to act as atmospheric barometric compensation. Instream depth loggerpressure data were compensated with the barometric pressure reader using the datacompensation feature in the Schlumberger Water Services Diver-Office program2011.1 (Version given the difference in pressure readings and the givenelevations, by the following formula from the diver manual:PH = P0 ∗ e−(M∗g∗H)/(R∗T ) (3.1)where PH is the atmospheric pressure at elevation height H, P0 is the atmosphericpressure at the reference height, M = 28.8 ∗ 10−3 kg/mol (molecular mass of air),g = 9.81 m/s2 (standard gravity), H = height in metres, R = 8.314 J/mol/K (gasconstant), and T = temperature in Kelvin. Barometric compensation data wereunavailable from the period of July 15, 2011 to Aug 15, 2011, therefore instreamdepth loggers from this period contain a much greater degree of uncertainty due tothe range of barometric pressure changes.27Figure 3.2: PVC piping well with water level logger.Bed topography surveysThe four study reaches were surveyed using a Leica TCR805 total station. Pocaterra,the most upstream study reach was surveyed for a length of approximately 300 m atapproximately 1 m intervals. Opal, the second study reach, had six cross sectionstaken for a reach length of approximately 150 m, with each cross section at ap-proximately one channel width distance downstream of the previous (see Fig. 3.3).Galatea, the third study reach, was surveyed similarly to Opal, but with five crosssections, for a reach length of approximately 100 m. The furthest downstreamstudy reach, Ribbon, was surveyed similarly to the first, over a length of approx-imately 300 m at 1 m intervals (see Fig. 3.4). When surveying streams for twodimensional (2D) hydraulic modelling, ten cross sections of complete surveys arerecommended, and in this regard the middle two reaches have fewer cross sec-tions than recommended. However, for similar 2D hydraulic habitat modellingprograms, five cross sections in a stream have been used for habitat modelling pur-poses (e.g. Rosenfeld et al., 2005). Justification for less intensive surveying relatesto the fact that these two reaches had more uniform bed topography.VelocityTo calculate discharge and measure velocity, an acoustic doppler current profiler(ADCP) was set up at each of the four study sites. The ADCP was used to takevertical velocity profiles at 1 m intervals across the channel (see Fig. 3.5). Exact28Figure 3.3: Surveying stream bed topography at the Opal site.Figure 3.4: Brett Eaton and Aaron Taminga surveying using a Leica TCR805total station and survey rod at the Ribbon site.29Figure 3.5: Velocities measurements made by ADCP at the Opal site.positioning of the ADCP during recording events was determined using the totalstation. When ADCP measurements were taken along a cross section, water surfaceelevation at the cross section and at the depth logger for that site were also recorded.Raw data exported from the ADCP were processed using a script to truncate data atthe appropriate depth based on bed topography and water survey elevation, as wellas average the forwards and backwards facing beams. The velocity profiles werethen integrated over depth and used to calculate discharge and velocity at the givendepth at the time of recording.Grain size and tracer rocksBasic pebble counts of 100 stones were completed at each study reach along crosssections until the requisite number of stones were collected and the cross sectionwas completed. Once the median grain size (D50) had been established, D50 sizedstones were collected and painted blue to be installed as tracer rocks. They werereplaced, slightly embedded, along one cross section at each study site. They werere-examined after one flood pulse to determine whether or not the bed had beenmobilized during the flood pulse.Total suspended solidsThe final element of physical conditions that was examined was the level of TSS.A depth-integrated sampler was used to collect three water samples for each site.30For study sites near tributary creeks (Galatea and Ribbon) samples were taken up-stream and downstream of the tributaries. One additional site named Evan Thomaswas added upstream of Ribbon Creek, and samples were taken there immediatelyupstream of the Evan Thomas Creek. Samples were then filtered using Whatman934-AH glass fiber filters (particle retention of 1.5 µm) , dried and then weighed inthe lab to measure TSS.3.2.2 River2D modellingMeasuring velocity and depth along cross sections provides important informationon the physical habitat conditions in a stream; however, this information appliesonly to the cross section, representing an infinitely small area of the stream chan-nel. To understand the physical conditions experienced by aquatic organisms inthe Kananaskis River, channel topography and flow data were used to model thetwo-dimensional hydraulic conditions over each study reach. Modelling was doneusing River2D (version 0.95a, January 15, 2010) a 2D hydrodynamic finite elementmodelling program developed by P. Steffler, A. Ghanem, J. Blackburn, and Z. Yangin conjunction with the University of Alberta, Fisheries and Oceans Canada, andthe United States Geological Survey. River2D is capable of outputting 2D maps ofhydraulic conditions, including such information as water surface elevation, veloc-ity magnitude, Froude number, depth and shear velocity magnitude. Of interest inthis study are 2D maps of depth and velocity.Model inputsTo run the model, certain physical conditions of the stream, measured in the field,were used as input parameters. First, bed roughness height had be to determinedfor the study reaches. Detailed roughness characteristics were not recorded in thefield, and as such, study reaches were given a single value for bed roughness height.The average reach bed roughness height was determined using the D84 for the studysite by the following equation (Hey, 1979):K = 3.5∗D84 (3.2)Secondly, a mesh had to be overlaid onto the study sites. Survey data were31input into River2D, and study sites were defined by external boundaries. Two-dimensional meshes were created by inputting uniform nodes throughout the studyreaches. The meshes were then triangulated and smoothed by making alterationsto the mesh triangulation, until they had a quality index 0.30 or higher. For eachmodel run, discharge was input along with starting inflow and outflow water sur-face elevations. River2D was run under steady flow (as opposed to the transientflow option). Inflow and outflow elevations were altered slightly to achieve solu-tion convergence. Solutions were deemed acceptable when the inflow and outflowdischarges were equal within plus or minus 5% (often within 1%). For specificmodel parameters, see Table 3.1.32Table 3.1: River2D model parametersSite Flow (high or low) Q inflow (cms) Q outflow (cms) Roughness (K) Mesh spacing (m)Pocaterra Low 1.18 1.18 0.26 2, 3Pocaterra High 20.40 20.40 0.26 2, 3Opal Low 8.24 8.24 0.13 0.3, 1Opal High 20.60 20.60 0.13 2Galatea Low 5.52 5.52 0.5 0.5Galatea High 16.80 16.80 0.5 0.3Ribbon Low 3.35 3.27 0.266 2Ribbon High 25.97 25.86 0.266 2333.3 Results3.3.1 Stream hydrologyDepth loggers recorded water level fluctuations in response to environmental flowchanges and flow releases from the Pocaterra Dam along the Kananaskis River.Hydrographs for the summer 2011 period are displayed in Fig. 3.7. Each discretepeak represents a high flow release from the Pocaterra Dam. Peak freshet flowsoccurred in late June, and were most clearly seen at Ribbon. High flows from thedam were released on semi-regular intervals throughout the summer.The flood hydrograph for single flood events varied between sites. However,the second, third and fourth sites had much more similar characteristics to oneanother than the most upstream site. A single flood pulse of 10.75 hrs duration onAugust 5, 2011 was identified as a typical flood, and the hydrographs from eachsite are depicted in Fig. 3.6. The rising limb of the hydrograph at the Pocaterra sitereached its maximum depth in 1 hour at a rate of 57.3 cm/hr, and achieved half ofits maximum depth within approximately 16 minutes at a rate of 107.4 cm/hr. AtOpal, the maximum depth was achieved within 9.5 hours (the slowest of the foursites) at a rate of 4.0 cm/hr and half of the maximum depth increase occurred within27 minutes at a rate of 42.6 cm/hr. The next site, Galatea, showed a similar overallup-ramping rate of 3.9 cm/hr requiring 8.5 hours to reach its maximum depth, butwith a much slower increase to reach 50% of its maximum depth at a rate of 21.3cm/hr doing so in approximately 47 minutes. Ribbon, the most downstream sitealso required 8.5 hours to reach its maximum depth and did so with an up-rampingrate of 2.5 cm/hr. Ribbon obtained half of its depth increase in approximately 23minutes at a rate of 25.8 cm/hr. Also of importance are down-ramping rates, whichwere similar, but lower than up-ramping rates (see Tables 3.2 and 3.3). It appearsthat the flood pulse becomes sufficiently dispersed, that with a 10.75 hr flood, it isonly the first site, Pocaterra, that reaches a steady stage. If the peaking release fromthe dam were to have a duration of greater than 10.75 hrs, it is likely that the Opal,Galatea and Ribbon sites would continue to experience rises in stage and wouldreach a maximum stage after a greater period of time than described for the 10.75hr flood. Releases of longer than 10.75 hrs are likely to have minimal effects on34Pocaterra OpalGalatea Ribbon204060204060Aug 05 06:00Aug 05 12:00Aug 05 18:00Aug 06 00:00Aug 06 06:00Aug 05 06:00Aug 05 12:00Aug 05 18:00Aug 06 00:00Aug 06 06:00Date and timeDepth (cm)Figure 3.6: Hydrograph for each study site of a high flow release from thePocaterra Dam on August 5-6, 2011.stage at the Opal site, due to the rapidity with which it attains maximum stage.35Table 3.2: Up-ramping and down-ramping rates for the first 50% of depthincrease or decrease during a 10.75 hr high flow release on August 5,2011.Site Up-ramping rate Down-ramping rate(cm/hr) (cm/hr)Pocaterra 107.4 77.5Opal 42.6 15.0Galatea 21.3 8.5Ribbon 25.8 5.3Table 3.3: Up-ramping and down-ramping rates for the entire duration of a10.75 hr high flow release on August 5, 2011.Site Up-ramping rate Down-ramping rate(cm/hr) (cm/hr)Pocaterra 57.3 43.9Opal 4.0 6.7Galatea 3.9 5.2Ribbon 2.5 3.836Figure 3.7: Hydrographs for all study sites. The most upstream site is Pocaterra, followed by Opal, Galatea and lastlyRibbon. Red lines represent the date and water level when low flow velocity readings were taken. Blue linesrepresent the date and water level when high flow velocity readings were taken. Barometric calibration wasunavailable from July 15, 2011 to Aug 15, 2011.37Table 3.4: Estimates of bed mobility after one flood cycle. Not all displacedtracer rocks were recovered, and as such, the percentage displaced repre-sents a conservative estimate.Site Tracer rocks displaced (%)Pocaterra 16Opal 31Galatea US 20Galatea DS 16Evan Thomas US 22Ribbon US 12Ribbon DS 203.3.2 Bed mobilityTracer rocks were installed to determine if the bed was mobile, and not to necessar-ily precisely measure that mobility. As such, exact distance and number displacedwere not recorded. However, a general survey was done to count the displacedrocks. Only rocks that had been displaced and retrieved were counted, so therecorded percentage of displaced rocks is likely to be an underestimate of the totalnumber displaced. The D50 from each study site was mobilized after one eight-hour flood flow release from the dam. Tracer rock displacement ranged from 12%to 31%, and after one eight hour flood pulse some tracer rocks were found as far as9 m downstream from their original placement (see Table 3.4).3.3.3 Modelled depth and velocityRiver2D depth and velocity maps demonstrate the spatial distribution of physi-cal characteristics at low and high flows. Each site had between 1760 and 28500nodes where depth and velocity were modelled, and used for statistical interpreta-tion. Depth and velocity maps for each study site can be seen in Figs. 3.8 to 3.15.In Fig. 3.8 and Fig. 3.9, the maps of depth and velocity at both low and high flowsfor the Pocaterra study site can be seen. In this first 300 m section, there is a sidechannel bar, a point bar and a mid channel bar at a low flow of 1.18 m3/s. Whensubjected to high flow conditions of 20.4 m3/s, these bars become submerged. Thenext site, Opal, had somewhat less distinct gravel bars, and as a result, a gen-38Table 3.5: Median grain size for study sites along Kananaskis RiverSite D50 (mm)Pocaterra 44.0Opal 19.7Galatea US 52.3Galatea DS 78.7Evan Thomas US 48.3Ribbon US 31.3Ribbon DS 32.0eral widening is viewed between the low and high flow conditions (Fig. 3.10 andFig. 3.11). At the Opal site at low flow, the regions of high velocity occurred overriffles. At high flow, the pools and riffles became less distinguishable as the watersurface slope became more even over the site, accounting for the lower velocitiesseen at high flow.At the Galatea site, the stream has little influx of sediment, and as a result hasa large D50 and the channel is somewhat incised (see Table 3.5). It is relativelystraight, with few exposed bars at low flow. This site experienced the smallestchange in depth of all the sites, which may be attributed to the higher channel slope.The Galatea site also experienced the smallest range in changes (see Fig. 3.12 andFig. 3.13), likely due to the uniformity of the channel.Ribbon, the furthest downstream site, experienced flooding of the side and midchannel bars at high flow, with an especially large point bar at the left side of thechannel becoming submerged at high flow (Fig. 3.14 and Fig. 3.15). The highestdepths were observed towards the left bank mid-way along the study section, wherea large pool was found.39Figure 3.8: 2D velocity map for the Pocaterra study site. (a) Low flow release from dam. (b) High flow release fromdam.40Figure 3.9: 2D depth map for the Pocaterra study site. (a) Low flow release from dam. (b) High flow release from dam.41Figure 3.10: 2D velocity map for the Opal study site. (a) Low flow release from dam. (b) High flow release from dam.42Figure 3.11: 2D depth map for the Opal study site. (a) Low flow release from dam. (b) High flow release from dam.43Figure 3.12: 2D velocity map for the Galatea study site. (a) Low flow release from dam. (b) High flow release fromdam.44Figure 3.13: 2D depth map for the Galatea study site. (a) Low flow release from dam. (b) High flow release from dam.45Figure 3.14: 2D velocity map for the Ribbon study site. (a) Low flow release from dam. (b) High flow release fromdam.46Figure 3.15: 2D depth map for the Ribbon study site. (a) Low flow release from dam. (b) High flow release from dam.47Depth and velocity data were not normally distributed, and as such, a non-parametric test was performed. Samples were assumed independent for statisticalpurposes, but may not be truly independent; they can be approximated as suchgiven that discharge had nearly doubled between samples and depth (velocity) iscommonly considered dependent on discharge, velocity (depth), channel slope andfrictional forces. The non-parametric test showed a significant (p=0.05) differencebetween depth at the high and low flow conditions at all sites. Increases in depth athigh flow ranged from 0.29 m at the Galatea site (the third downstream) up to 0.53m at the Pocaterra site (the most upstream site), which correspond to increases indischarge of 11.2 m3/s and 19.2 m3/s respectively (see Fig. 3.16 and Fig. 3.17).Velocities at three of the four sites showed significant increases from the low tothe high flow conditions. However, the second site, Opal, experienced no statisti-cally significant change in velocity between low and high flow events (see Fig. 3.18and Fig. 3.19).3.3.4 Total suspended solidsTSS changed between low and high flow dam releases (see Fig. 3.20). At all but twosites, TSS were found to be significantly higher during high flow releases from thedam, using a t-test with a corresponding p value of less than 0.05 (three samples persite, per flow condition). The highest TSS at high flow were found at the Pocaterrasite (42 mg/L) and were more than eight times higher than levels seen at highflow at any other site. At Evan Thomas Creek, 32 km from the dam, there wasno statistically significant difference in TSS between low and high flow. However,the average TSS were higher at low flow at this site. Unlike high levels of TSSobserved elsewhere, the higher sediment load at low flow at Evan Thomas wasassociated with a few large suspended grains (1 to 2 mm); whereas elsewhere,suspended solids were much finer (less than 1/4 mm) and uniform throughout thewater samples. As water samples were taken during one day, it is likely that thesefew large grains were indicative of a small localized event, and not representativeof regular levels of suspended solids at low flow. This is further supported by thefact that high flow had lower levels of TSS. This site experienced higher dischargerates and greater depths at high flow. Sediment transport theory would anticipate48Figure 3.16: Boxplot of depths at high (teal) and low (beige) flow for the fourstudy sites.higher shear stresses at higher discharge rates and therefore greater, not less, levelsof suspended solids. This is why it is suspected that the large grains present in thesample were from a small, local disturbance. Ribbon Creek is one of the largesttributaries over this stretch of the Kananaskis River. Downstream of this creek, nostatistical difference was observed between the levels of TSS at high and low flow.3.4 Discussion3.4.1 Ramping ratesIn this study, rates of flow change rapidly due to the hydropeaking nature of thedam. Most of the literature on the effects of ramping focuses on the rate of fishstranding in rapidly dewatering areas. Scientific studies have also been conductedon the effects of flow fluctuations on fish growth, and have concluded slight posi-49Figure 3.17: Boxplot of the changes in depth between low and high flow atthe four study sites. The distribution represents the changes that oc-curred at individual locations throughout the study site, over an evenlyspaced grid.50Figure 3.18: Boxplot of velocities at high (teal) and low (beige) flow for thefour study sites.tive (Smokorowski et al., 2011; Crisp et al., 1983), neutral (Almodovar and Nicola,1999) and negative effects (Baran et al., 1995; de Crespin de Billy et al., 2002;Weyers et al., 2003). It is unknown if fish stranding occurs on the KananaskisRiver. Preliminary fish surveys completed in the summer of 2012 do indicatethat fewer fish are located upstream where ramping rates are highest, comparedto downstream sites on the Kananaskis River (Macnaughton, 2013, pers. comm.).A study using an artificial stream by Halleraker et al. (2003) showed that strandingof juvenile Brown Trout was significantly decreased by reducing down-rampingrates from 60 cm/hr to 10 cm/hr. Pocaterra, 2 km downstream from the dam ex-perienced the greatest down ramping rates of the study sites, of up to 77.5 cm/hr.With large bars at low flow that then become submerged at high flow, the Pocaterrasite has potential to experience fish stranding. Down-ramping rates at the otherthree sites ranged from 3.8 to 15.0 cm/hr. Potential for stranding is present at thesesites, especially when the down-ramping rate exceeds 10 cm/hr (Halleraker et al.,51Figure 3.19: Boxplot of the changes in velocity between low and high flowat the four study sites. The distribution represents the changes that oc-curred at individual locations throughout the study site, over an evenlyspaced grid.52010203040PocaterraOpalGalatea_USGalatea_DSEvan_ThomasRibbon_USRibbon_DS Total suspended solids (mg/L)Low flowHigh flowFigure 3.20: Total suspended solids at all study sites, at high and low flow.2003; Saltveit et al., 2001), and occurs during daylight hours (Halleraker et al.,2003) (as is often the case for the Kananaskis). Stranding potential is likely higheron the Kananaskis River, as down-ramping continues until minimum flow levelsare achieved. Where down-ramping occurs rapidly, but to a moderate flow level,stranding is less likely (Tuhtan et al., 2012).3.4.2 Bed mobilityA thorough analysis of bed mobility is not possible here, due to limitations fromsampling procedures whereby not all rocks were recovered as they were not fittedwith tracer magnets. As a result, percentages of rocks displaced are underesti-mated. The 12% to 31% + of tracer rocks that were mobilized at high flow indicate53that the D50 is mobile at all sites at high flow. The percentage of displaced rockswas approximately within one order of magnitude across sites, indicating that sim-ilar levels of bed mobility are observed throughout the Kananaskis River duringhigh flow events.3.4.3 Depth and velocityFluctuations in depth and velocity were extreme at most sites between high andlow flow. Changes in velocities and depths were shown to modify some salmonidbehaviour, potentially with energetic costs (e.g. Korman and Campana, 2009; Geistet al., 2005; Scruton et al., 2003, 2005). However, it is difficult to determine the neteffect, as other studies have shown positive growth with hydropeaking. At Opal,there was a wide range of change in velocities at any specific point in the site;however, an average of these changes actually resulted in a decrease in velocity(although this change was not statistically significant). This is partly explainedby high water surface slope riffles becoming flooded at high flow, resulting in adecrease in slope and therefore decrease in velocity over the riffles. This behaviour,while reasonable given the morphology, is odd as generally depths and velocitiesincrease with increasing discharge.3.4.4 Total suspended solidsIgnoring the anomaly at Evan Thomas where a few large grains increased the TSS,differences in TSS between low and high flows were seen only at Pocaterra, thefirst site within 1.5 km of the dam. When high levels of TSS are seen, it suggeststhat shear stresses are high enough to erode banks during flood flows immediatelydownstream of the dam. Suspended load originates largely as erosion of the banksand from overland flow. Because climatic conditions were fairly dry when thesamples were taken, all suspended solids in the stream are likely to have originatedfrom within the channel.The flood wave is likely responsible for the increased erosion at Pocaterra re-sulting in higher TSS. Downstream, while depths and velocities (and thereforeshear stress) were high at high flow releases, there was a much reduced level ofsolids in suspension, indicating that the rapid up-ramping rate and flood wave ex-54perienced upstream may be responsible for the solids in the water. Slight overtopping of the banks at high flow between the first and second study sites couldresult in deposition of suspended particles, and in part, explain lower levels of sus-pended solids observed at high flow at the downstream sites. TSS concentrationsmay also be decreased progressively downstream by the influx of clearer waterfrom tributaries and groundwater flow.In a summary report on the effects of suspended solids on salmonids, sus-pended solids were reported to cause physiological (more acute) and behavioural(more long-term) changes (Bash et al., 2001). Low level thresholds leading to be-havioural changes were usually measured in hundreds of mg/L, which is more thandouble the highest level recorded on the Kananaskis. Levels as low as 1.5 mg/Lresulted in adverse health conditions in some Chinook Salmon fry (Newcombe andMacDonald, 1991). It is possible then that TSS at high flow on Kananaskis Riverhave some adverse effects on fish, with sites between 5.1 mg/L and 6.8 mg/L (ex-cept Pocaterra’s high of 42.3 mg/L).3.5 ConclusionsMost of the results of the changes in physical condition point to the idea that longi-tudinal position has little influence on how a site is affected by the upstream dam.The dominant driving factor as to how physical characteristics will change betweenhigh and low flow seems to be site-specific morphology. Of indicators measured,only ramping and TSS relate to distance from the dam. TSS were high 1.5 kmdownstream of the dam and appeared to reach an equilibrium by the second studysite, 20.3 km from the dam. This equilibrium, however, is above the level of TSSobserved at low flow.55Chapter 4Downstream fish habitat4.1 IntroductionHow ecosystems respond to river regulation plays a large role in how dams are op-erated. Environmental needs are considered along with human needs (hydropower,flood control, etc.) when regulators decide how a dam may be operated. Whileentire ecosystems are affected by river regulation, it is riverine species that aremost affected and most often considered in environmental flow needs. As they aregenerally the most valuable economic and recreational animal in streams, fish areusually given the most consideration. In the provincial instream flow guidelinesfor British Columbia, two separate guidelines exist: those for fish-bearing streams,and those for non fish-bearing streams (MWLAP, 2004). It is important to have astrong scientific understanding of how fish respond to changes in flow, so that thebest management decisions can be made.Many studies have been conducted below dams to assess how they impact fish.They generally focus on fish, or fish habitat. When examining fish habitat variables,often a single site below a dam is studied (e.g. Garcia et al., 2011). Fish abundanceand diversity below dams may be studied at progressive distances below dams,with little quantifiable data on habitat (e.g. Vehanen et al., 2005). Rarely do studiesinclude detailed, progressive longitudinal habitat data, below dams. The study byValentin et al. (1996) does address habitat at multiple sites below a dam, but thisarea of study is uncommon. This chapter seeks to examine habitat at four sites56downstream of a hydropeaking dam, at the high and low flow releases. It alsoexamines how differences in usable habitat change with increasing distance fromthe dam and with increasing unregulated tributary influence.4.2 MethodsTo assess changes in fish habitat on the Kananaskis River, field studies were con-ducted in the summer of 2011. Pressure transducers were installed at four sitesalong the river downstream of Pocaterra Dam. Sites were located 1.5 km to 35.4km downstream of the dam (see Fig. 2.7). The next reservoir on the river is 46 kmdownstream from Pocaterra Dam.From these sites, stream bed topography, depth, and velocity measurementswere taken between May 15 and August 15, 2011. Field data were used as inputparameters to model depths and velocities for 100 m to 400 m at each study site atboth the low and high flow dam releases. River2D, a 2-dimensional hydrodynamicfinite element modelling program, was used. For a detailed explanation of fieldmethods, River2D and model parameters, please refer to the Methods section ofChapter Habitat modellingRiver2D was used to model fish habitat at high and low flow on Kananaskis Riverbelow the hydropeaking Pocaterra Dam. Depth and velocity estimates presented inChapter 3 of this thesis were combined with habitat preference curves for knownfish species in the Kananaskis River to generate 2D WUA estimates. Courtney et al.(1998) conducted an instream flow requirement study for fish in the KananaskisRiver; they developed habitat preference curves for Brown Trout and MountainWhitefish, the two main species in the Kananaskis River. Habitat preference curvesfor fry, juvenile and adult life stages were used for both Brown Trout and MountainWhitefish. For measures of WUA in River2D, the minimum suitability given byeither depth, velocity or channel index was used. Using surveyed bed topography,ArcGIS 10.3 was used to measure the stream length of each site. WUA data foreach life stage, fish species, site, and flow condition were divided by stream lengthto get a standardized measure of WUA so that sites of different lengths could be57more easily compared.4.2.2 Potential habitatIn addition to WUA, this study also looked at potential habitat, which was definedas any portion of the 2D model that had a habitat value greater than zero. Thisapproach attempts to simplify the model in hopes that a binary measure of ‘habit-able’ or ‘non-habitable’ may carry a higher degree of confidence, if less detail. Allanalyses were performed for measures of WUA and potential habitat.4.2.3 Change in habitat between flow conditionsEach life stage for fish has different habitat requirements. In general, fry fare betterin low compared to high flow conditions, with adults having more habitat availableat higher flows. As it would be difficult to measure a general increase or decreasein habitat across multiple life stages from the low flow to high flow condition,absolute changes in habitat were measured for each life stage and species. Becauseboth the low and high flow condition are experienced on multiple occasions perweek, the minimum habitat (experienced at either high flow or low flow) is assumedto be a limiting factor. Percent differences of habitat were calculated between theoptimal and limiting flow conditions for each site, life stage and species, in orderto determine the extent that the hydropeaking flows limit habitat.4.3 ResultsWUA ranged from a low of 0 m2/m stream length (adult Mountain Whitefish atGalatea during low flow) to 17.7 m2/m stream length (Mountain Whitefish fry atOpal during low flow). Potential habitat varied from 1.0 m2/m stream length (adultMountain Whitefish at Ribbon during low flow) to 31.9 m2/m stream length (adultBrown Trout at Pocaterra during high flow).Mountain Whitefish adults had the least amount of habitat compared to otherlife stages as measured by WUA and potential habitat (at 14 site/flow/habitat com-binations of 16), except for potential habitat at the Galatea and Ribbon sites athigh flow. For Mountain Whitefish, fry commonly had the most WUA and potentialhabitat (in 10 of 16 cases). When considering potential habitat, Brown Trout adults58Figure 4.1: Potential habitat per m stream length - Mountain Whitefish.Figure 4.2: WUA per m stream length - Mountain Whitefish.had the most available habitat. For WUA, Brown Trout adults had the most habitatat high flow, whereas juveniles had the most WUA at low flow.Comparing across species, fry of Mountain Whitefish always had more habi-tat (WUA and potential) than Brown Trout. The opposite is true of adults, whereBrown Trout had more available habitat. For juvenile life stages, no one speciesconsistently had more available habitat. For a full summary of potential habitat andWUA, see Figs. 4.1, 4.2, 4.3, and 4.4.When examining the percent difference between the high and low habitat con-dition, a consistent trend is visible across both species by measures of WUA andpotential habitat (see Fig. 4.5). The smallest difference in available habitat be-tween the two flow conditions occurs at Opal for both Brown Trout and Mountain59Figure 4.3: Potential habitat per m stream length - Brown Trout.Figure 4.4: WUA per m stream length - Brown Trout.Whitefish. Downstream of Opal, this difference increases at the Galatea site, andincreases further at the Ribbon site. At Pocaterra, the most upstream site, the dif-ference between the two flow conditions is similar to that at the Ribbon site.60Brown trout potential habitat Brown trout WUAWhitefish potential habitat Whitefish WUA050100150200050100150200PocaterraOpalGalateaRibbonPocaterraOpalGalateaRibbonSitesDifference (%)Figure 4.5: Percent difference of habitat between the high and low flow con-ditions. Each site represents the fry, juvenile and adult life stages.4.4 Discussion4.4.1 Habitat reductionsWinter habitat has been indentified as a bottleneck for salmonids (Cunjak, 1996;Huusko et al., 2007). In a stream restoration study in Norway by Koljonen et al.(2013), it was determined that while restoration efforts increased summer habitatby 20%, they had no significant effect on winter habitat. This high interannual61variability in discharge and WUA negated restoration efforts and resulted in noincrease in salmonid numbers. While this study notes the wintertime low in habi-tat as a bottleneck to salmonid numbers, we speculate that in the hydropeakingKananaskis River, daily low habitat values (occuring at the high or low flow damrelease) observed year round are likely habitat bottlenecks.Habitat bottlenecks on Kananaskis River for different life stages, species, sitesand habitat measures have between a 2.6% (juvenile Brown Trout at the Opal site,potential habitat) and 193% (adult Mountain Whitefish at the Galatea site, WUA ashabitat measure) difference in habitat between the two extreme flow conditions. Onaverage, looking at all variables, there is an 84% difference between the flow con-dition with the most and least habitat. This suggests that the current hydropeakingflow regime is having severe effects on fish in the river.4.4.2 Serial discontinuity conceptThe SDC presents the idea that dams on regulated rivers interrupt theoretical con-tinuous gradients of natural flowing systems (Ward and Stanford, 1983). Naturaldownstream gradients include biotic (eg. periphyton, invertebrate species composi-tions) and abiotic (eg. temperature, velocity) factors. Frequently, the SDC examinesrecovery distances, refering to the distance downstream of a dam that is influencedby the dam. Beyond the recovery distance, the river is thought to behave as if thedam were not there, or where downstream changes have equalized.Throughout the 35 km of of Kananaskis River examined in this study, thereis no apparent recovery. Fluctuations in habitat availability were as extreme im-mediately below the dam as they were a further 35 km downstream. Along withthe lack of recovery, there is no downstream trend in how habitat varied betweenhigh and low flow conditions. Rather, extreme fluctuations were seen for most lifestages at most sites. These findings contrast those of Vehanen et al. (2005) wheredownstream gradients of physical conditions and fish assemblages were observed.The Vehanen et al. (2005) study examined a much shorter section of river (8.7 km)and the downstream reach included the impounded area immediately above a dam,perhaps explaining the differences in results.Due to the strong range of flow conditions under the hydropeaking regime, it62is likely that no true recovery can occur between the Pocaterra Dam and where itis dammed a further 51 km downstream, as this distance is too short.4.5 ConclusionChanges in habitat between high and low flow dam releases were extreme. Thesechanges happen regularly, and pose challenges to fish. The Scruton et al. (2003)study of the movement of Atlantic salmon and trout experiencing hydropeakingflows, showed that fish did not move large distances, and usually stayed within thestudy reach. In this study, trout moved more during peaking events than at steadyflows. This increased movement is likely displayed by fish on the Kananaskis Riveras well, and has energetic costs associated with almost daily peaking events. It isnoteworthy that no discernible downstream trend was noticed in habitat. However,this may relate more to site morphology, as the most upstream and most down-stream sites were more heterogeneous than the two intermediary sites.63Chapter 5ConclusionsChanges in flow, ramping rates, bed mobility, TSS, depth, velocity, and habitat (po-tential and WUA) were investigated for this study. These factors were examinedfor changes between high flow (23 m3/s) and low flow (0.5 m3/s) dam releases,which both occur every week of the year. Changes between the high and low flowconditions were then examined for downstream trends. Of the factors measured,few showed any sign of downstream attenuation, in either absolute terms or indifferences between the high and low flow conditions. A review article on longi-tudinal trends in regulated river by Ellis and Jones (2013) describes how studiesoften identify changes caused by dams, but more rarely report on any longitudinalchanges.5.1 Observed longitudinal trendsHydrologic characteristics (including flood pulse attenuation and ramping rates)and TSS were two examined factors that demonstrated a longitudinal trend down-stream of the Pocaterra Dam on the Kananaskis River. The downstream attenuationof the flood pulse was described in Chapter 3. This information is an important ad-dition to discussions on longitudinal patterns below dams, as it is rarely reported(Ellis and Jones, 2013). Ramping rates were highest immediately below the dam,and attenuated with increasing distance downstream. Ramping rates were alsosubject to local site morphology, as demonstrated by the lowest ramping rates oc-64curing at the third, and not fourth site. This is likely related to the high slope atthe Galatea site, with more of the additional discharge being accommodated by in-creased velocities as opposed to increased depth. TSS appears to somewhat followthe predictions of the SDC. At the first site 1.5 km below the dam, TSS were high(42 mg/L) at the peaking discharge. At all other downstream sites, the high flowvalue appears relatively constant. It appears then, that a steady state is reached ata distance between Pocaterra at 1.5 km and Opal at 20.3 km. However, this steadystate does not imply a full recovery, as TSS levels at low flow were, on average, at1.0 mg/L (excluding the outlier Evan Thomas site) compared to 6.1 mg/L for highflow (excluding the outlier Pocaterra site).5.2 Elements not displaying downstream longitudinaltrendsThe D50 was mobile at all sites, but not enough precision was employed in thefield methods to determine any downstream trend. A future study of longitudinalpatterns of bed mobility below a hydropeaking dam would benefit from an exami-nation of multiple size classes and a proper recovery of tracer rocks. Resources didnot permit for such an indepth procedure for this study, and therefore only a binarymobile or non-mobile test was completed.Depth, velocity, and habitat did not display a downstream longitudinal trend.Neither did the changes in these factors between low and high flow. According tothe SDC, the relative change in these factors between the two flow releases shoulddiminish downstream, as more tributaries flow into the river, supposedly dampen-ing the flood signal.5.3 River recoveryStream characteristics downstream of hydropeaking dams have been greatly stud-ied, and to a lesser extent, the longitudinal trends below these dams. Longitudinaltrends commonly examined are temperature (Camargo and Voelz, 1998; Cereghinoet al., 2002; Paller and Saul, 1996; Preece and Jones, 2002; Saltveit et al., 1994),invertebrates (Rehn, 2009; Cereghino et al., 2002; Patterson and Smokorowski,2011; Jones, 2013; Cortes et al., 2002), and periphyton (Rader and Ward, 1988;65Morley et al., 2008; Rehn, 2009).Recovery distances for habitat features (either fish or invertebrate) are lesscommonly studied than animal abundances or diversity. Katano et al. (2009) foundthat the influence of a tributary downstream of the dam helped in recovering theriver from the influence of the dam, which counters what was found in this study,where tributaries and downstream distance did little to reduce the effects from thedam. It is possible that due to the extreme fluctuations in flow on Kananaskis River,the distance between Pocaterra Dam and the subsequent impoundement is too shortfor recovery to occur.66BibliographyA. Almodovar and G. Nicola. Effects of a small hydropower station upon browntrout (Salmo trutta l.) in the River Hoz Seca (Tagus basin, Spain) one year afterregulation. Regulated Rivers: Research and Management, 15(5):477–484,1999. → pages 51M. B. Bain, J. T. Finn, and H. E. Booke. Streamflow regulation and fishcommunity structure. Ecology, 69(2):382–392, 1988. → pages 4, 9P. Baran, M. Delacoste, F. Dauba, J.-M. Lascaux, A. Belaud, and S. Lek. Effectsof reduced flow on brown trout (Salmo trutta l.) populations downstream damsin French Pyrenees. Regulated Rivers: Research and Management, 10(2-4):347–361, 1995. → pages 51J. Bash, C. Berman, and S. Bolton. Effects of turbidity and suspended solids onsalmonids. Technical report, Center for Streamside Studies, University ofWashington, 2001. → pages 55R. Beschta and W. Jackson. The intrusion of fine sediments into a stable gravelbed. Journal of the Fisheries Research Board of Canada, 36(2):204–210, 1979.→ pages 17K. D. Bovee. A guide to stream habitat analysis using the Instream FlowIncremental Methodology. Technical report, US Fish and Wildlife Service,1982. → pages 10, 11M. C. Bruno, B. Maiolini, M. Carolli, and L. Silveri. Short time-scale impacts ofhydropeaking on benthic invertebrates in an alpine stream (Trentino, Italy).Limnologica, 40(4):281–290, 2010. → pages 12, 25C. M. Bunt, S. J. Cooke, C. Katopodis, and R. S. McKinley. Movement andsummer habitat of brown trout (Salmo trutta) below a pulsed dischargehydroelectric generating station. Regulated Rivers: Research & Management,15(5):395–403, 1999. → pages 2, 3, 967B. A. Byren and B. R. Davies. The effect of stream regulation on thephysico-chemical properties of the Palmiet River, South Africa. RegulatedRivers: Research & Management, 3(1):107–121, 1989. → pages 9, 25J. A. Camargo and N. J. Voelz. Biotic and abiotic changes along the recoverygradient of two impounded rivers with different impoundment use.Environmental Monitoring and Assessment, 50(2):143–158, 1998. → pages 9,65R. Cereghino, P. Cugny, and P. Lavandier. Influence of intermittent hydropeakingon the longitudinal zonation patterns of benthic invertebrates in a mountainstream. International Review of Hydrobiology, 87(1):47–60, 2002. → pages 9,25, 65R. M. V. Cortes, M. T. Ferreira, S. V. Oliveira, and D. Oliveira. Macroinvertebratecommunity structure in a regulated river segment with different flow conditions.River Research and Applications, 18(4):367–382, 2002. → pages 9, 65R. F. Courtney, G. L. Walder, and R. L. J. Vadas. Report on instream flowrequirements for fish in the Kananaskis River. Technical report, Fisheries andRecreation Enhancement Working Group, 1998. → pages 57I. G. Cowx, K. T. O’Grady, P. Parasiewicz, S. Schmutz, and O. Moog. The effectof managed hydropower peaking on the physical habitat, benthos and fish faunain the river bregenzerach in austria. Fisheries Management and Ecology, 5(5):403–417, 1998. → pages 4D. Crisp, R. Mann, and P. Cubby. Effects of regulation of the River Tees upon fishppopulation below cow green reservoir. Journal of Applied Ecology, 20(2):371–386, 1983. → pages 51R. Cunjak. Winter habitat of selected stream fishes and potential impacts fromland-use activity. Canadian Journal of Fisheries and Aquatic Sciences, 53(1):267–282, 1996. → pages 61R. M. Cushman. Review of the ecological effects of rapidly varying flowsdownstream from hydroelectric facilities. North American Journal of FisheriesManagement, 5(3A):330–339, 1985. → pages 7V. de Crespin de Billy, B. Dumont, T. Lagarrigue, P. Baran, and B. Statzner.Invertebrate accessibility and vulnerability in the analysis of brown trout(Salmo trutta l.) summer habitat suitability. River Research and Applications,18(6):533–553, 2002. → pages 5168R. G. Death and M. J. Winterbourn. Diversity patterns in stream benthicinvertebrate communities: The influence of habitat stability. Ecology, 76(5):1446–1460, 1995. → pages 25B. C. Eaton and M. Church. Predicting downstream hydraulic geometry: A test ofrational regime theory. Journal of Geophysical Research-Earth Surface, 112(F3):353–364, 2007. → pages 9L. E. Ellis and N. E. Jones. Longitudinal trends in regulated rivers: a review andsynthesis within the context of the serial discontinuity concept. EnvironmentalReviews, 21(3):136–148, 2013. → pages 25, 64Environment Canada. Canadian climate normals 1981-2010 station data, 2015.URL normals/index e.html. → pages 16D. C. Erman, E. D. Andrews, and M. Yoderwilliams. Effects of winter floods onfishes in theSierra-Nevada. Canadian Journal of Fisheries and AquaticSciences, 45(12):2195–2200, 1988. → pages 2, 6, 25A. Garcia, K. Jorde, E. Habit, D. Caamano, and O. Parra. Downstreamenvironmental effects of dam operations: Changes in habitat quality for nativefish species. River Research and Applications, 27(3), 2011. → pages 56D. R. Geist, R. S. Brown, V. Cullinan, S. R. Brink, K. Lepla, P. Bates, and J. A.Chandler. Movement, swimming speed, and oxygen consumption of juvenilewhite sturgeon in response to changing flow, water temperature, and light levelin the Snake River, Idaho. Transactions of the American Fisheries Society, 134(4):803–816, 2005. → pages 54A. Ghanem, P. Steffler, F. Hicks, and C. Katopodis. Two-dimensional hydraulicsimulation of physical habitat conditions in flowing streams. Regulated Rivers:Research & Management, 12(2-3):185–200, 1996. → pages 12J. H. Halleraker, S. J. Saltveit, A. Harby, J. V. Arnekleiv, H.-P. Fjeldstad, andB. Kohler. Factors influencing stranding of wild juvenile brown trout (Salmotrutta) during rapid and frequent flow decreases in an artificial stream. RiverResearch and Applications, 19(5-6):589–603, 2003. → pages 51, 53F. Hauer, J. A. Stanford, and J. V. Ward. Serial discontinuities in a RockyMountain river. II. Distribution and abundance of Trichoptera. RegulatedRivers: Research and Management, 3(1):177–177, 1989. → pages 2569J. Heggenes. Effects of short-term flow fluctuations on displacement of, andhabitat use by, brown trout in a small stream. Transactions of the AmericanFisheries Society, 117(4):336–344, 1988. → pages 2, 5R. D. Hey. Flow resistance in gravel-bed rivers. Journal of the HydraulicsDivision, 105(4):365–379, 1979. → pages 31H. R. Hudson, A. E. Byrom, and W. L. Chadderton. A Critique of IFIM: InstreamHabitat Simulation in the New Zealand Context. Department of Conservation,2003. → pages 12A. Huusko, L. Greenberg, M. Stickler, T. Linnansaari, M. Nykanen, T. Vehanen,S. Koljonen, P. Louhi, and K. Alfredsen. Life in the ice lane: The winterecology of stream salmonids. River Research and Applications, 23(5):469–491,2007. → pages 61N. E. Jones. Spatial patterns of benthic invertebrates in regulated and naturalrivers. River Research and Applications, 29(3):343–351, 2013. → pages 9, 65I. Katano, J. N. Negishi, T. Minagawa, H. Doi, Y. Kawaguchi, and Y. Kayaba.Longitudinal macroinvertebrate organization over contrasting discontinuities:effects of a dam and a tributary. Journal of the North American BenthologicalSociety, 28(2):331–351, 2009. → pages 66S. Koljonen, A. Huusko, A. Mki-Petys, P. Louhi, and T. Muotka. Assessinghabitat suitability for juvenile atlantic salmon in relation to in-streamrestoration and discharge variability. Restoration Ecology, 21(3):344–352,2013. → pages 61J. Korman and S. E. Campana. Effects of hydropeaking on nearshore habitat useand growth of age-0 rainbow trout in a large regulated river. Transactions of theAmerican Fisheries Society, 138(1):76–87, 2009. → pages 1, 2, 3, 54H. Liebig, S. Mastrorillo, P. Gaudin, P. Lim, and A. Belaud. Microhabitat use by0+ brown trout (Salmo trutta l.) in a mountain stream affected by hydropeakingoperations (Pyrenees, Southeast France). Archiv fuer HydrobiologieSupplement, 135(2-4):203–218, 2001. → pages 25I. Maddock. The importance of physical habitat assessment for evaluating riverhealth. Freshwater Biology, 41(2):373–391, 1999. → pages 9O. E. Maughan and P. J. Barrett. An evaluation of the instream flow incrementalmethodology (IFIM). Journal of the Arizona-Nevada Academy of Science,24/25:75–77, 1992. → pages 1070D. McParland, B. Eaton, and J. Rosenfeld. At-a-station hydraulic geometrysimulator. River Research and Applications, in press. → pages 9H. Moir, C. Gibbins, C. Soulsby, and A. Youngson. Phabsim modelling of atlanticsalmon spawning habitat in an upland stream: testing the influence of habitatsuitability indices on model output. River research and applications, 21(9):1021–1034, 2005. → pages 12W. L. Montgomery, S. D. McCormick, R. J. Naiman, F. G. Whoriskey, and G. A.Black. Spring migratory synchrony of salmonid, catostomid, and cyprinidfishes in Riviere-a-la-Truite, Quebec. Canadian Journal of Zoology-RevueCanadienne De Zoologie, 61(11):2495–2502, 1983. → pages 17O. Moog. Quantification of daily peak hydropower effects on aquatic fauna andmanagement to minimize environmental impacts. Regulated Rivers: Research& Management, 8(1-2):5–14, 1993. → pages 6, 25S. A. Morley, J. J. Duda, H. J. Coe, K. K. Kloehn, and M. L. McHenry. Benthicinvertebrates and periphyton in the Elwha River basin: current conditions andpredicted response to dam removal. Northwest Science, 82(sp1):179–196,2008. → pages 66K. J. Murchie, K. P. E. Hair, C. E. Pullen, T. D. Redpath, H. R. Stephens, and S. J.Cooke. Fish response to modified flow regimes in regulated rivers: Researchmethods, effects and opportunities. River Research and Applications, 24(2):197–217, 2008. → pages 3MWLAP. Instream flow thresholds for fish and fish habitat as guidelines forreviewing proposed water uses : synopsis. Technical report, British ColumbiaMinistry of Water, Land & Air Protection, Victoria, 2004. → pages 56T. Naesje, B. Jonsson, and J. Skurdal. Spring flood: A primary cue for hatching ofriver spawning Coregoninae. Canadian Journal of Fisheries and AquaticSciences, 52(10):2190–2196, 1995. → pages 17C. Newcombe and D. MacDonald. Effects of suspended sediments on aquaticecosystems. North American Journal of Fisheries Management, 11(1):72–82,1991. → pages 55A. R. M. Nowell and P. A. Jumars. Flow environments of aquatic benthos. AnnualReview of Ecology and Systematics, 15(1):303–328, 1984. → pages 971J. H. O’Keeffe, R. W. Palmer, B. A. Byren, and B. R. Davies. The effects ofimpoundment on the physicochemistry of two contrasting southern Africanriver systems. Regulated Rivers: Research & Management, 5(2):97–110, 1990.→ pages 9, 25M. Paller and B. Saul. Effects of temperature gradients resulting from reservoirdischarge on Dorosoma cepedianum spawning in the Savannah River.Environmental Biology of Fishes, 45(2):151–160, 1996. → pages 9, 65R. Patterson and K. Smokorowski. Assessing the benefit of flow constraints on thedrifting invertebrate community of a regulated river. River Research andApplications, 27(1):99–112, 2011. → pages 9, 65T. N. Pearsons, H. W. Li, and G. A. Lamberti. Influence of habitat complexity onresistance to flooding and resilience of stream fish assemblages. Transactionsof the American Fisheries Society, 121(4):427–436, 1992. → pages 8N. L. Poff, J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter,R. E. Sparks, and J. C. Stromberg. The natural flow regime. Bioscience, 47(11):769–784, 1997. → pages 17R. M. Preece and H. A. Jones. The effect of Keepit Dam on the temperatureregime of the Namoi River, Australia. River Research and Applications, 18(4):397–414, 2002. → pages 9, 65M. C. Quist, W. A. Hubert, and D. J. Isaak. Fish assemblage structure andrelations with environmental conditions in a rocky mountain watershed.Canadian Journal of Zoology-Revue Canadienne De Zoologie, 82(10):1554–1565, 2004. → pages 9R. B. Rader and J. V. Ward. Influence of regulation on environmental conditionsand the macroinvertebrate community in the upper Colorado River. RegulatedRivers: Research & Management, 2(5):597–618, 1988. → pages 65A. C. Rehn. Benthic macroinvertebrates as indicators of biological conditionbelow hydropower dams on west slope Sierra Nevada streams, California, USA.River Research and Applications, 25(2):208–228, 2009. → pages 9, 65, 66B. Richter, J. Baumgartner, R. Wigington, and D. Braun. How much water does ariver need? Freshwater Biology, 37(1):231–249, 1997. → pages 19B. D. Richter, J. V. Baumgartner, J. Powell, and D. P. Braun. A method forassessing hydrologic alteration within ecosystems. Conservation Biology, 10(4):1163–1174, 1996. → pages 1772J. S. Rosenfeld, T. Leiter, G. Lindner, and L. Rothman. Food abundance and fishdensity alters habitat selection, growth, and habitat suitability curves forjuvenile coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheriesand Aquatic Sciences, 62(8):1691–1701, 2005. → pages 28S. Saltveit, J. Halleraker, J. Arnekleiv, and A. Harby. Field experiments onstranding in juvenile Atlantic salmon (Salmo salar) and brown trout (Salmotrutta) during rapid flow decreases caused by hydropeaking. Regulated Rivers:Research & Management, 17(4-5):609–622, 2001. → pages 53S. J. Saltveit, T. Bremnes, and J. E. Brittain. Effect of a changed temperatureregime on the benthos of a Norwegian regulated river. Regulated Rivers:Research & Management, 9(2):93–102, 1994. → pages 9, 65D. Scruton, L. Ollerhead, K. Clarke, C. Pennell, K. Alfredsen, A. Harby, andD. Kelley. The behavioural response of juvenile Atlantic salmon (Salmo salar)and brook trout (Salvelinus fontinalis) to experimental hydropeaking on aNewfoundland (Canada) river. River research and applications, 19(5-6):577–587, 2003. → pages 54, 63D. Scruton, C. Pennell, M. Robertson, L. Ollerhead, K. Clarke, K. Alfredsen,A. Harby, and R. McKinley. Seasonal response of juvenile Atlantic salmon toexperimental hydropeaking power generation in Newfoundland, Canada. NorthAmerican Journal of Fisheries Management, 25(3):964–974, 2005. → pages 54D. W. Seegrist and R. Gard. Effects of floods on trout in Sagehen Creek,California. Transactions of the American Fisheries Society, 101(3):478–&,1972. → pages 2, 7K. Smokorowski, R. Metcalfe, S. Finucan, N. Jones, J. Marty, M. Power, R. Pyrce,and R. Steele. Ecosystem level assessment of environmentally based flowrestrictions for maintaining ecosystem integrity: a comparison of a modifiedpeaking versus unaltered river. Ecohydrology, 4(6):791–806, 2011. → pages 51R. Spence and P. Hickley. The use of PHABSIM in the management of waterresources and fisheries in England and Wales. Ecological Engineering, 16(1):153–158, 2000. → pages 12C. Stalnaker, B. L. Lamb, J. Henriksen, K. Bovee, and J. Bartholow. The instreamflow incremental methodology: a primer for IFIM. Technical report, DTICDocument, 1995. → pages 10J. A. Stanford and J. Ward. Revisiting the serial discontinuity concept. RegulatedRivers: Research & Management, 17(4-5):303–310, 2001. → pages 2573J. A. Stanford and J. V. Ward. Serial discontinuities in a Rocky Mountain river. I.Distribution and abundance of Plecoptera. Regulated Rivers: Research andManagement, 3(1):169–169, 1989. → pages 25L. E. Stevens, J. P. Shannon, and D. W. Blinn. Colorado River benthic ecology inGrand Canyon, Arizona, USA: dam, tributary and geomorphological influences.Regulated Rivers: Research & Management, 13(2):129–149, 1997. → pages 25D. J. Stier and J. H. Crance. Habitat suitability index models and instream flowsuitability curves: American shad. Technical report, US Fish and WildlifeService, 1985. → pages 11A. W. Storey, D. H. Edward, and P. Gazey. Recovery of aquatic macroinvertebrateassemblages downstream of the canning dam, western australia. RegulatedRivers: Research & Management, 6(3):213–224, 1991. → pages 9TransAlta. Plants in operation, 2014. URL → pages 14J. A. Tuhtan, M. Noack, and S. Wieprecht. Estimating stranding risk due tohydropeaking for juvenile european grayling considering river morphology.KSCE Journal of Civil Engineering, 16(2):197–206, 2012. → pages 53S. Valentin, F. Lauters, C. Sabaton, P. Breil, and Y. Souchon. Modelling temporalvariations of physical habitat for brown trout (Salmo trutta) in hydropeakingconditions. Regulated Rivers: Research & Management, 12(2-3):317–330,1996. → pages 5, 9, 12, 56T. Vehanen, J. Jurvelius, and M. Lahti. Habitat utilisation by fish community in ashort-term regulated river reservoir. Hydrobiologia, 545:257–270, 2005. →pages 12, 56, 62J. V. Ward and J. A. Stanford. Dynamics of Lotic Ecosystems, chapter The serialdiscontinuity concept of lotic ecosystems., pages 29–42. Ann Arbor SciencePublishers: Ann Arbor, MI, 1983. → pages 24, 62M. Weston. A profile of the physical attributes of the Little Bear River in thecontext of the serial discontinuity concept. Natural Resources andEnvironmental Issues, 18(1):3, 2013. → pages 25R. Weyers, C. Jennings, and M. Freeman. Effects of pulsed, high-velocity waterflow on larval robust redhorse and v-lip redhorse. Transactions of the AmericanFisheries Society, 132:84–91, 2003. → pages 5174


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