TRENDS IN ANGLING AND PISCIVORE CONDITION FOLLOWING ELEVEN YEARS OF NUTRIENT ADDITIONS IN ARROW LAKES RESERVOIR (ARROW LAKES RESERVOIR CREEL SURVEY 2003 – 2009) Steve Arndt, M.Sc. Fish & Wildlife Compensation Program - Columbia Basin 103-333 Victoria St., Nelson, B.C. and Carl Schwarz, Ph.D, P. Stat. (SSC, ASA) Dept of Statistics and Actuarial Science Simon Fraser University, Burnaby, BC December 2011 Arrow Lakes Reservoir Creel Survey 2003-2009 Executive Summary The Fish and Wildlife Compensation Program - Columbia Basin (FWCP) is a joint initiative of the Province of British Columbia, BC Hydro, and Canada Fisheries and Oceans to conserve and enhance fish and wildlife populations affected by BC Hydro dams in the Canadian portion of the Columbia River basin. Arrow Lakes Reservoir (ALR) is influenced by Hugh Keenleyside dam (completed 1968) near the outlet of the original lakes, and by Mica (1973) and Revelstoke (1984) dams and their associated reservoirs upstream. The FWCP annually funds a large scale nutrient program and kokanee spawning channel on ALR as compensation for footprint dam impacts. Data collected through angler creel surveys are a key component of program evaluation. Access point angler surveys have been conducted at selected ALR locations as early as 1976 to monitor the effects of BC Hydro dams and fishery compensation efforts including Hill Creek kokanee spawning channel, started in the 1980s, and the nutrient program started in 1999. These surveys provide a valuable long term index of fishing effort and harvest starting from dam construction through to 11 years after the beginning of the nutrient program. In addition to monitoring angling trends, they are useful for evaluating the response of rainbow trout and bull trout populations, which have not been monitored systematically by other methods over this period. This report provides reservoir-wide estimates of angler effort, catch and harvest for 2003 to 2009, and summarizes longer term trends (1987-2009) at three access locations to assess the performance of FWCP compensation initiatives and in particular, to compare the pre-nutrient era (1987-1998) to the eleven years of the nutrient program. Trends in relative condition factor (K ) n are also examined to evaluate the suitability of feeding conditions for apex predators, bull trout and piscivorous rainbow trout, that feed primarily on kokanee. Creel survey data were collected at three primary access locations (Shelter Bay, Nakusp, Castlegar) for five days per month (three weekdays and two weekend/holidays). For estimation of total reservoir effort and catch from 2003 on, the monthly estimates from the three access locations were expanded using correction factors based on the ratio of access-sampled boats to total boats counted during 48 airplane flights made over the whole reservoir between April 2003 and March 2005. For longer term comparisons, the annual estimates specific to the three sampled access sites were used. Total annual effort (± 95% confidence limits) from 2003 to 2009 ranged from 14,500 (± 2400) to 17,600 (±3,500) angler days. Annual catch (all species including released fish) ranged from 12,000 to 25,000 with harvest ranging from 8 to 12 tonnes. Angling effort out of the Castlegar access has declined by approximately 50% in recent years due to a substantial decline in the kokanee fishery. Nakusp effort increased significantly shortly after the beginning of the nutrient program and remained above the 1987-1998 range up to 2009. Shelter Bay effort has remained relatively constant since 1987. Annual expenditures wholly attributable to the fishery are about $1 million based on daily values from a federal angler survey, or $3 million including purchases partly attributable to the fishery. Residents of British Columbia comprised about 90% of anglers in all years. Kokanee harvest from 2003-2009 ranged from 2,300 – 9,000 fish/year (300 - 1,800 kg/year) with average catch rates (CPUE) of < 0.6 fish/h. Recent harvest estimates at the monitored sites are Columbia Basin Fish & Wildlife Compensation Program ii Arrow Lakes Reservoir Creel Survey 2003-2009 less than 20% of those from 1990-1996 with associated declines in kokanee-directed effort. Kokanee effort and harvest since 1998 were positively related to mean size of retained kokanee, which was larger in the early years of the nutrient program. Bull trout catch from 2003-2009 ranged from 2,600 – 3,800 fish/year, about half of which were retained for harvests of 3,300 - 5,400 kg/year; CPUE ranged from 0.06 - 0.08 fish/h. Catch of bull trout increased sharply in 2001, three years after the beginning of the nutrient program, and remained at a higher level until 2005. The increase was concurrent with a large increase in kokanee abundance, in particular older age classes, during the early years of the nutrient program. Since then, catch has declined to levels similar to pre-nutrient years, coincident with declines in kokanee spawners. Rainbow trout catch from 2003-2009 ranged from 3,900 – 6,400 fish/year; about two thirds of which were retained for harvests of 2,500 - 4,400 kg/year; CPUE ranged from 0.06 - 0.08. Harvest of piscivorous rainbow trout over 50 cm was in the range of 200-500 fish/year with a catch trend similar to that of bull trout. Hatchery stocked (clipped) bull trout and rainbow trout made up less than one percent of the harvest in most years, suggesting low survival to catchable size. Burbot comprise a much smaller, but relatively stable, fishery with catch ranging from 400 - 700 fish/year and harvest ~700 kg/year. Mean annual condition factors (K ) of piscivorous rainbow trout and bull trout were closely n correlated (R2 = 0.81), as would be expected given the similarities in diet. A period of enhanced K from 2002 to 2005 paralleled catch trends. Mean K was also positively related to several n n measures of fishery quality (e.g., total annual catch, catch of larger fish, CPUE) supporting the notions that: recruitment and survival of piscivores are strongly influenced by suitable feeding conditions, and angler catch trends are valid indices of piscivore abundance. Mean K did not n appear to be related to hydroacoustic estimates of total density or biomass of all age classes of kokanee in the same year; however, relationships to kokanee spawner escapement were quite strong (R2 = 0.73 and 0.60 for bull and rainbow respectively) suggesting that optimal feeding conditions are related to the abundance of larger (older) kokanee. For piscivorous rainbows, there was also some evidence of a dome-shaped relationship to prey size after accounting for spawner abundance, with optimal K occurring when kokanee spawner size was ~ 23 cm. n Increases in piscivore catch rate, harvest, size and condition factor suggest a strong positive response to nutrient additions reaching to the upper trophic levels during the first seven years of the program, at least in the upper basin of the reservoir. Declines in more recent years suggest trophic efficiency has been reduced, for reasons that are as yet unclear. The recent decline in kokanee angling may be related at least partially to the reduction in size and vulnerability to angling that accompanies increased density. Reductions in kokanee catch limits (from 15 to 5 since 1995) may also be a factor in the reduced effort and harvest for this species. Columbia Basin Fish & Wildlife Compensation Program iii Arrow Lakes Reservoir Creel Survey 2003-2009 Acknowledgements Thanks to the anglers who contributed information to the surveys, and creel technicians Glen and Gail Olson in Nakusp, Deb and Lorne Imeson and staff at Scotties Marina, and Allison Alder at Shelter Bay for their efforts in collecting field data. Beth Woodbridge provided data entry and administrative support. The report was improved by review comments from Dale Sebastian and Jeff Burrows of the Ministry of Forests, Lands, and Natural Resource Operations (MFLNRO). Dale also provided hydroacoustic kokanee data, a summary of regulation changes over time, and other information relevant to interpretation of condition factor and angler effort trends. The Fish and Wildlife Compensation Program - Columbia Basin is a joint initiative of the Province of British Columbia, Canada Fisheries and Oceans, and BC Hydro to conserve and enhance fish and wildlife populations affected by BC Hydro dams in the Canadian portion of the Columbia River basin. Funding is provided by BC Hydro as a requirement of their water licences for dams in the region. Columbia Basin Fish & Wildlife Compensation Program iv Arrow Lakes Reservoir Creel Survey 2003-2009 Table of Contents Executive Summary ...................................................................................................................................... ii Acknowledgements ...................................................................................................................................... iv Table of Contents .......................................................................................................................................... v List of Tables and Figures ............................................................................................................................ vi 1.0 INTRODUCTION ................................................................................................................................. 1 2.0 METHODS ............................................................................................................................................ 2 2.1 Access Sampling ................................................................................................................................................. 2 2.2 Overflight Boat Counts ..................................................................................................................................... 5 2.3 Analyses .............................................................................................................................................................. 5 2.3.1 Logistical Constraints ................................................................................................................................... 5 2.3.2 Effort and Catch Estimation ......................................................................................................................... 6 2.3.3 Catch per Unit Effort and Fin Clips ............................................................................................................. 8 2.3.4 Condition Factor of Piscivorous Rainbow Trout and Bull Trout and Relationship to Attributes of the Kokanee Population .............................................................................................................................................. 8 3.0 RESULTS AND DISCUSSION ......................................................................................................... 10 3.1 Fishery Overview ............................................................................................................................................. 11 3.2 Angling Effort and Potential Expenditures ................................................................................................... 14 3.3 Harvest, Size Distribution, and Catch Rates ................................................................................................. 16 3.3.1 Bull Trout .................................................................................................................................................... 16 3.3.2 Rainbow Trout ............................................................................................................................................ 20 3.3.3 Kokanee ...................................................................................................................................................... 26 3.4.4 Burbot ......................................................................................................................................................... 30 3.4 Piscivore Condition Trends and Implications ............................................................................................... 34 3.4.1 Trends ......................................................................................................................................................... 36 3.4.2 Relationship to Kokanee Population Structure ........................................................................................... 37 3.5 Hatchery Contribution to the Fishery ............................................................................................................ 39 4.0 CONCLUSIONS ................................................................................................................................. 42 4.1 Recommendations ............................................................................................................................................ 44 5.0 REFERENCES .................................................................................................................................... 46 6.0 APPENDICES ..................................................................................................................................... 50 1. Estimating the Precision of the Creel Surveys on Arrow Lake. 2. Monthly correction factors and standard errors pooled over daytypes and all years from the ratio of overflight boat counts and interviewed anglers at 3 access points. 3. Angler residence composition from 1976 to 2009. 4. Arrow Lakes creel survey annual estimates using a common monthly correction factor applied to all zones to adjust for boats returning to non-sampled access sites 5. Arrow Lakes creel survey annual estimates using separate monthly correction factors for each site/zone combination to adjust for boats returning to non-sampled access sites. 6. Size statistics for bull trout, rainbow trout, kokanee, and burbot in the Arrow Lakes Reservoir creel survey from 1998 to 2009. 7. Locations of fishing boats on Arrow Lakes Reservoir for 48 flights made between April 2003 and March 2005. Columbia Basin Fish & Wildlife Compensation Program v Arrow Lakes Reservoir Creel Survey 2003-2009 List of Tables and Figures Table 1. Spatial and temporal strata for the Arrow Lakes Reservoir creel surveys from 2003 - 2009 .......................... 3 Table 2. Sampled rod-hours, harvest, catch, and catch rates (fish/hour) of boats returning to the Edgewood and Fauquier ramps compared to Castlegar and Nakusp access locations in 1998 .............................................................. 6 Table 3. Three measures of estimated angling effort ( 95% confidence limits) for Arrow Lakes Reservoir from 2003 to 2009 ................................................................................................................................................................ 15 Table 4. Bull trout angler catch and harvest ( 95% confidence limits) statistics for Arrow Lakes Reservoir from 2003 to 2009 ................................................................................................................................................................ 17 Table 5. Rainbow trout angler catch and harvest ( 95% confidence limits) statistics for Arrow Lakes Reservoir from 2003 to 2009 ....................................................................................................................................................... 21 Table 6. Kokanee catch and harvest ( 95% confidence limits) for Arrow Lakes Reservoir from 2003 to 2009 ....... 27 Table 7. Estimated number of burbot caught and retained ( 95% confidence limits), percentage of fish kept, and harvested weight for Arrow Lakes Reservoir from 2003 to 2009 ............................................................................... 32 Table 8. Burbot catch rate comparison from creel surveys in four Kootenay Region Lakes. ...................................... 32 Table 9. Comparison of Arrow Lakes Reservoir burbot harvest to other North American lakes ................................ 32 Table 10. Summary of the number and percent (upper 95% confidence limit) of hatchery-clipped bull trout and rainbow trout in Arrow Lakes Reservoir creel samples from 2003 to 2009 ................................................................ 40 Table 11. Sample location, reported clip, possible year of stocking and length of marked fish in 2003 to 2009 creel surveys on ALR ........................................................................................................................................................... 41 Table 12. Summary of fish harvest from Arrow Lakes Reservoir from 2003 to 2009 ................................................ 42 Fig. 1. Map of Arrow Lakes Reservoir showing three sampled access locations and zone boundaries used for assigning overflight boat counts .................................................................................................................................... 4 Fig. 2. Proportion of angling effort (sum of sampled angler days) by species sought ................................................. 11 Fig. 3. Catch of harvested and released fish by month and species at three access locations on Arrow Lakes Reservoir in 2009......................................................................................................................................................... 13 Fig. 4. Number of angler days by month at three access locations on Arrow Lakes Reservoir in 2009. .................... 14 Fig. 5. Trends in annual angler days for three access locations on Arrow Lakes Reservoir from 1987 – 2009. ......... 15 Fig. 6. Trends in (a) fish harvested (b) targeted rod-hours, and (c) catch rate for bull trout from three access locations in Arrow Lakes Reservoir from 1987 - 2009 ............................................................................................................... 18 Fig. 7. Fork length distributions of harvested bull trout from the Arrow Lakes Reservoir creel survey from 1998 to 2009 ............................................................................................................................................................................. 19 Fig. 8. Trends in the (a) number of fish kept, (b) directed rod-hours, and (c) catch rate for rainbow trout from three access locations in Arrow Lakes Reservoir from 1987-2009 ...................................................................................... 22 Fig. 9. Fork length distributions of harvested rainbow trout from the Arrow Lakes Reservoir creel survey from 1998 to 2009 ......................................................................................................................................................................... 23 Fig. 10. Number of rainbow trout greater than 50 cm sampled at three access locations during the Arrow Lakes creel survey from 1998 to 2009 ............................................................................................................................................ 25 Fig. 11. Number of rainbow trout weighed at Nakusp by size category from 1977 to 2009 ....................................... 25 Fig. 12. Trends in (a) number of fish kept, (b) directed rod-hours, and (c) catch rate for kokanee anglers from three access locations in Arrow Lakes Reservoir from 1987-2009 ...................................................................................... 28 Fig. 13. Fork length distributions of harvested kokanee from the Arrow Lakes Reservoir creel survey from 1998 to 2009 ............................................................................................................................................................................. 29 Fig. 14. Relationship between the amount of fishing effort for kokanee and length of harvested kokanee in Arrow Lakes Reservoir from 1998 to 2009 ............................................................................................................................ 30 Fig. 15. Trends in the (a) number of fish kept, (b) directed rod-hours, and (c) catch rate for burbot anglers from Nakusp access location in Arrow Lakes Reservoir from 1987-2009 ........................................................................... 31 Fig. 16. Length frequency distributions of burbot angled in Arrow Lakes Reservoir from 1998 to 2002 .................. 33 Fig. 17. Comparison of annual mean relative condition factor (K ) for bull trout and rainbow trout over 50 cm in n Arrow Lakes Reservoir from 1996 to 2009 ................................................................................................................. 34 Columbia Basin Fish & Wildlife Compensation Program vi Arrow Lakes Reservoir Creel Survey 2003-2009 Fig. 18. Graphs showing positive relationships between mean annual condition factor (K ) and catch rate , total n catch , and catch of larger fish for bull trout and piscivorous rainbow trout from Arrow Lakes Reservoir 1998-2009 ..................................................................................................................................................................................... 35 Fig. 19. Mean annual (± 95% confidence limits) condition factor (K ) relative to the average weight of pre-nutrient n fish (K =1)for all bull trout and for bull trout over 60 cm in Arrow Lakes Reservoir from 1991 to 2009 ................. 36 n Fig. 20. Mean annual (± 95% confidence limits) condition factor (K ) relative to the average weight of pre-nutrient n fish (K =1)for piscivorous rainbow trout in Arrow Lakes Reservoir from 1994 to 2009 ........................................... 37 n Fig. 21. Relationships between bull trout and piscivorous rainbow trout condition (K ) and kokanee population n structure ....................................................................................................................................................................... 38 Fig. 22. Residuals from linear regressions using kokanee spawner abundance to predict annual median K of bull n trout and piscivorous rainbow trout plotted versus the average size of kokanee spawners ......................................... 39 Columbia Basin Fish & Wildlife Compensation Program vii Arrow Lakes Reservoir Creel Survey 2003-2009 1.0 INTRODUCTION Arrow Lakes Reservoir (ALR) extends from Revelstoke to Keenleyside Dam near Castlegar in the West Kootenay Region of British Columbia. It has a total surface area (upper and lower basins) of 46,450 ha at full pool (Pieters et al. 2003), and is affected by three dams, one at the reservoir outlet (Hugh L. Keenleyside Dam) and two upstream (Mica and Revelstoke dams). Impacts to fish populations include the loss of stream spawning and juvenile rearing habitats in the lower reaches of tributaries and upstream of Revelstoke Dam, and changes in nutrient levels, water clarity and primary productivity due to upstream dams (Matzinger et al. 2007, Moody et al. 2007, Hagen 2008, Arndt 2009a, 2009b). The fish community in ALR includes 24 species (McPhail and Carveth 1992) with the most abundant in anglers’ catch being kokanee Oncorhynchus nerka, bull trout Salvelinus confluentus, rainbow trout Oncorhynchus mykiss, and burbot Lota lota. Rainbow trout occur in at least two different ecotypes, a slower growing (mostly insectivorous) form reaching a maximum length of about 45 cm, and a piscivorous ecotype that can exceed 90 cm (Arndt 2004b). Early attempts to compensate for dam impacts focused on replacing juvenile production from lost stream habitat. Hill Creek Spawning Channel, located north of Nakusp, was started in the early 1980s to provide spawning and juvenile rearing habitat for kokanee and rainbow trout (Lindsay 1982, Barney 2009). Annual kokanee fry production from the channel has ranged from less than 1 million to over 20 million (MOE/FWCP file data). The channel also has high use by rainbow trout for spawning and juvenile rearing (Porto and Arndt 2006). A fish hatchery, operated at the same site for production of bull trout and rainbow trout, was discontinued after 2000 due to poor survival of released fish (Arndt 2004a). However, small releases of piscivorous rainbow trout were made from 1995 to 2002 by Selkirk College, and from 2005 to 2009 the Freshwater Fisheries Society of BC has released triploid rainbow trout yearlings on an experimental basis. Since bull trout typically enter the fishery starting at age 5 and can live for more than 10 years (Sebastian et al. 2000), stocked fish of both species could be at large in the lake during the period covered by this study. In 1999, another large-scale compensation project commenced to address the issue of nutrient loss in upstream reservoirs. Limiting nutrients (phosphorus and nitrogen) are distributed in the upper basin during the growing season with the goal of increasing reservoir primary productivity (Pieters et al. 2003, Schindler et al. 2006). This in turn is expected to translate into higher kokanee production and improved growth and survival of bull trout and piscivorous rainbow trout in the lake phase of their life history. Productive fish stocks are expected to provide angling opportunities and economic benefits to the local communities and province. Funding for Hill Creek Spawning Channel and the nutrient program is provided by BC Hydro through the Fish and Wildlife Compensation Program - Columbia Basin (FWCP), a joint initiative of BC Hydro, the provincial government, and Fisheries and Oceans Canada. Angler surveys have been conducted annually at selected access points on ALR since the 1970s to monitor the effects of the dams on fish populations and recreational fisheries, and the success of compensation efforts. These surveys provide a valuable index of angling effort and harvest trends, and the only long-term data on piscivorous fish species in the reservoir. Sebastian et al. (2000) summarized creel trends to 1997 and Arndt (2002a, 2004b) from 1998 to 2002. This report provides detailed estimates of angler effort and catch from 2003 to 2009, and summarizes longer term trends starting in 1987 to allow a comparison of years before and after the beginning of the nutrient program. It differs from the previous reports in that the 2003 to 2009 access point data were Columbia Basin Fish and Wildlife Compensation Program 1 Arrow Lakes Reservoir Creel Survey 2003-2009 expanded to whole-reservoir estimates using a correction factor based on airplane counts of the total fishing boats.1 Two categories of data are provided by the survey: angler effort and catch, and population metrics (size structure, condition) of the harvested species. By combining the two (changes in catch/harvest and changes in size/condition), it is possible to test hypotheses about ecosystem function and limiting factors to better understand how to optimize compensation benefits. Questions and corresponding objectives include the following: Question Objective • What is the recreational, food, and approximate Provide annual estimates of angler effort, catch, economic value of the ALR fishery? and harvest in the reservoir as a measure of recent recreational, food and economic benefits, and as a baseline for comparison to other large lakes. • How has the fishery changed (catch, harvest, Evaluate the success of the major compensation effort, catch-per-unit-effort) since the beginning initiatives, particularly at the upper trophic of the nutrient program in 1999? levels, over the 11 years of the nutrient program in terms of the recreational fishery. • What fish population conditions contribute to Determine the relationships between angler increased angler use and community benefits, effort and fishery characteristics such as catch and are there other factors that could affect rate and size distribution of the harvest to assist angler behaviour and interpretation of before/ with adaptive management for increased after nutrient addition effects? benefits. Examine changes in regulations and the potential influence on angler effort (i.e., kokanee). • Have feeding conditions for apex predators Evaluate prey suitability for apex predators (bull improved since the beginning of the nutrient trout and piscivorous rainbow trout) since the program? beginning of the nutrient program. • What prey conditions seem optimal for transfer Investigate relationships between apex predators of nutrient benefits to upper trophic levels of (abundance, size, condition) and the abundance fish production? and population structure of kokanee prey to better understand what optimizes transfer efficiency to their trophic level. 2.0 METHODS 2.1 Access Sampling Anglers were interviewed by creel clerks at three primary access sites (Shelter Bay, Nakusp, Castlegar) for five days per month from January 2003 to December 2009 (Table 1; Fig. 1). Three weekdays (WD) and two weekend/holidays (WE) were sampled in each month. This provided coverage of approximately a sixth of the total days in the survey period including a quarter of 1 Earlier reports either did not expand the site estimates or expanded them using a correction factor based on professional judgement. Columbia Basin Fish and Wildlife Compensation Program 2 Arrow Lakes Reservoir Creel Survey 2003-2009 weekend days. Sampling was randomized within the day types and sampled days were the same at all access points so that aerial boat counts could be calibrated against boats returning to the three monitored sites. In keeping with past surveys, sampled weekdays included one randomly-selected Monday in each month, although for analysis purposes all weekdays were combined (Arndt 2002a). Creel clerks monitored the access points from two hours after official sunrise to one half hour after sunset, and the number of interviews was assumed to be the total effort for a given access point and day.2 Interviews took place at the completion of the fishing trip. Recorded information for each angling party included start and end time of the fishing trip, fish species sought, number of fish harvested and released (by species), number of anglers and rods used, and angler residence. Fork length (FL) and weight were recorded for a subsample of harvested kokanee with the stipulation that all fish from a given boat be measured. Size measurements were recorded for all harvested bull trout and rainbow trout and all were examined for the presence of hatchery fin clips (contingent on angler permission). At the Shelter Bay and Nakusp access sites, there is only one active boat ramp, and all returning angling parties can usually be contacted. At the Castlegar location near the south end of the reservoir, there were three ramps in close proximity up to 2008 (Scotties Marina, Syringa Marina, Syringa Provincial Park) and 2 ramps afterward (Syringa Marina closed). During winter months when angling effort is lower, complete coverage of all Castlegar ramps was possible by waiting in a vehicle at the most northerly ramp and following returning boats to other ramps if necessary. During summer high activity periods, the Castlegar clerk interviewed as many boats as possible, although it was not always possible to interview every returning party. Table 1. Spatial and temporal strata for the Arrow Lakes Reservoir creel surveys from 2003 - 2009. Access Location Description Sampling Frequency 1. Shelter Bay Shelter Bay Provincial Park boat ramp 5 days per month (3 weekdays (shifts to nearby ferry ramp if park ramp is and 2 weekend/ statutory inaccessible due to snow) holidays) 2. Nakusp Nakusp government wharf in the Town of Nakusp 3. Castlegar Scotties Marina, Syringa Marina, Syringa Park public boat launch at the south end of the lower basin of the reservoir 2 Start times were slightly different at the Shelter Bay boat ramp and park because it is located about one hour drive from the nearest community and required the clerk to arrive by hourly ferry. Survey start time was 8:00 AM from April to March when the Shelter Bay campsite is open, and 10:00 AM from November to March (campsite closed). Columbia Basin Fish and Wildlife Compensation Program 3