Lewis River Fish Passage Program 2014 Annual Report (30-day review Draft)

Transcription

1 Lewis River Fish Passage Program 2014 Annual Report (30-day review Draft) FERC Project Nos. 935, 2071, 2111, 2213 Merwin Fish Collection Facility Adult Holding Tank 2014 Photo by Chris Karchesky PacifiCorp Energy & Public Utility District No.1 of Cowlitz County February 2015

2 Table of Contents 1.0 INTRODUCTION PASSAGE FACILITIES SWIFT RESERVOIR FLOATING SURFACE COLLECTOR MERWIN UPSTREAM COLLECTION FACILITY DOWNSTREAM COLLECTION AND PASSAGE METRICS NUMBER OF JUVENILES ENTERING SWIFT RESERVOIR OVERVIEW RESULTS/DISCUSSION NUMBERS COLLECTED AT THE FSC OVERVIEW RESULTS/DISCUSSION JUVENILE MIGRATION TIMING OVERVIEW RESULTS/DISCUSSION FSC COLLECTION EFFICIENCY OVERVIEW RESULTS/DISCUSSION SWIFT FSC INJURY AND SURVIVAL OVERVIEW RESULTS/DISCUSSION SWIFT POWERHOUSE ENTRAINMENT EVALUATION OVERALL DOWNSTREAM SURVIVAL (ODS) OVERVIEW RESULTS/DISCUSSION UPSTREAM COLLECTION AND PASSAGE METRICS SUMMARY ADULT PASSAGE SURVIVAL OVERVIEW RESULTS/DISCUSSION ADULT TRAP EFFICIENCY OVERVIEW SPAWNING TIMING, DISTRIBUTION, AND ABUNDANCE OF TRANSPORTED FISHES OVERVIEW RESULTS/DISCUSSION OCEAN RECRUIT ANALYSIS OVERVIEW RESULTS/DISCUSSION PREFORMANCE MEASURES FOR INDEX STOCKS OVERVIEW RESULTS/DISCUSSION REINTRODUCED AND RESIDENT FISH INTERACTIONS OVERVIEW RESULTS/DISCUSSION LITERATURE CITED

3 1.0 INTRODUCTION The North Fork Lewis Hydroelectric Project begins about 10 miles east of Woodland, Washington (Figure 1.0-1), and consists of four impoundments. The sequence of the four Lewis River projects upstream of the confluence of the Lewis and Columbia rivers is: Merwin, Yale, Swift No. 2, and Swift No.1. These four projects are licensed separately by the Federal Energy Regulatory Commission (FERC). Merwin (FERC No. 935), Yale (FERC No. 2071), and Swift No. 1 (FERC No. 2111) are owned and operated by PacifiCorp Energy (PacifiCorp). Swift No. 2 (FERC NO. 2213) is owned by Public Utility District No. 1 of Cowlitz County (Cowlitz PUD) and is operated by PacifiCorp under contract with Cowlitz PUD in coordination with the other hydroprojects. Combined, the Lewis River project has a generation capacity of 606 megawatts. On 26 June 2008, PacifiCorp and Cowlitz PUD were issued an Order by FERC approving the Settlement Agreement and granting new licenses for the North Fork Lewis River Hydroelectric Projects. Among the conditions contained in each License was a requirement for reintroducing anadromous salmonids and providing fish passage upstream of Merwin Dam and downstream of Swift No. 1 Dam. The overarching goal of this comprehensive reintroduction program is to achieve genetically viable, self-sustaining, naturally reproducing, harvestable populations of anadromous salmonids upstream of Merwin Dam. The target species identified in the Settlement Agreement for reintroduction are spring Chinook salmon (Oncorhynchus tshawytscha), early-run (S-type) coho salmon (O. kisutch), and winter steelhead (O. mykiss). The Settlement Agreement called for a phased approach for reintroduction that occurs over a seventeen year period following issuance of the new Licenses. The phased approach provides for a carefully devised plan to protect the listed species and to verify effectiveness of the passage facilities while allowing for the reintroduction program to take effect. Among the tasks identified for Phase I of the reintroduction plan was establishing a downstream passage facility in the forebay of Swift No.1 Dam and making upgrades to the existing adult fish capture facility at Merwin Dam. Subsequent phases would establish facilities for both upstream and downstream passage at Merwin, Yale, and Swift No.1 dams, with fish ultimately spawning and rearing naturally throughout the project area. A decision on whether subsequent phases are implemented is anticipated in The Lewis River Aquatic Monitoring and Evaluation (M&E) Plan (PacifiCorp and Cowlitz PUD 2010) was developed as part of the Settlement Agreement to evaluate performance measures outlined in the new Licenses. The primary focus of the plan is to provide methods for monitoring and evaluating the fish passage program. In accordance with the Settlement Agreement, the Licensees shall Consult with the Aquatic Coordination Committee (ACC) as necessary, but no less often than every five years, to determine if modifications to the M&E Plan are warranted (SA 9.1). Revisions to the M&E Plan are scheduled to be completed by the end of The purpose of this report is to document results of the field assessments associated with implementation of the fish passage program in the existing M&E Plan during

4 Figure 1.0-1: An overview of key features of the North Fork Lewis River Hydroelectric Project area located in Southwest, Washington. 2

5 2.0 PASSAGE FACILITIES 2.1 Swift Reservoir Floating Surface Collector The Swift Reservoir Floating Surface Collector (FSC) began daily operations on December 26, The facility is located at the south end of Swift Dam near the turbine intake (Figure 2.1-1), and consists of three primary structures: Fish Collection Barge Truck Access Trestle and Mooring Tower Barrier Net and Net Transition Structure Figure 2.1-1: Aerial photo of the Swift Floating Surface Collector. The Swift Floating Surface Collector is a floating barge that measures 170 feet long, 60 feet wide and 53 feet tall. The purpose of the FSC is to provide attraction flow at the surface of the reservoir where juvenile salmonids are migrating. Fish enter the FSC via the Net Transition Structure (NTS), which funnels water and fish into an artificial stream channel created by electric pumps. The stream channel then entrains and guides fish into the collection facility that automatically sorts fish by life-stage (i.e., fry, smolt, and adult) and then routes them to holding tanks for biological sampling and transport downstream 1. The artificial stream channel is 1 Following transport downstream, smolts are release into release ponds located near Woodland, WA. Fish are held in these ponds for 24-hours before being allowed to volitionally enter the river. As of December 2014, these ponds have not been constructed. Fish transported downstream in 2014 were released directly in the lower river. 3

6 maintained at a capture velocity of approximately 7 feet per second (fps) with 600 cubic feet per second (cfs) attraction flow during normal operations (80% of full flow capacity). The purpose of the 660-foot access trestle is to provide fish transport trucks access to the 280 foot tall mooring tower. The mooring tower doubles as a hopper-to-truck fish transfer structure, allowing operators to move fish from the FSC to the truck across a broad range of reservoir surface elevations 2. The portion of the exclusion net that is located perpendicular to the front of the FSC is approximately 1,700 feet long and consists of three distinct vertical panel materials. The upper section of the net consists of a solid material running 0-15 feet below the surface. The middle net section (15-30 feet) consists of a fine net material (Dyneema ) with 1/8-inch mesh opening. The lower most section (30 feet and beyond) is also constructed of Dyneema with 3/8-inch mesh opening. In addition to the forward-facing exclusion net, there are two side nets that begin at each of the turning points and extend to shore. Each side net is constructed of nylon material. The upper portion (0-15 feet) of the net has a mesh opening of 1/8-inch and the lower portion (15 feet and beyond) has a mesh opening of 3/8-inch. Soon after the FSC began operation in late December 2012, it was determined that the exclusion net sustained damage during severe weather conditions. The extent of this damage was evaluated with a number of dive and ROV surveys of the net beginning in early February It was determined that the net separated at both north and south turning points. These tears compromised the effectiveness of the net throughout the 2013 migration season. Efforts to repair the net began in December 2013 and were completed by April During this repair period, the FSC was turned off. The FSC resumed operation on April 1, The FSC was operated 24-hours a day through 2014 except during periods when it was necessary to shut the facility down due to power outages, facility modification, or scheduled maintenance (Table 2.1-1). Table Scheduled FSC outages in Outage Duration Purpose December 21 st, 2013 March 31 st Scheduled repairs - barrier net July 15 th July 16 th Unscheduled repairs rocker gate repair July 29 th July 31 st Scheduled repairs - net transition structure August 8 th August 22 nd Scheduled repairs barrier net August 29 th October 9 th Scheduled maintenance period deballasting December 24 th December 31 st Unscheduled repairs smolt water supply valve repair 2 The Swift FSC has an operation range of 120 feet in reservoir elevation change. 4

7 Fish Crowder & Lift Assembly Conveyance Flume Presort Pond & Sorting Building Fishway Entrance No. 1 Auxiliary Water Supply Station & Conveyance Pipe Figure 2.2-1: Merwin Sorting Facility, May Merwin Upstream Collection Facility The new upstream collection and transport facility (Figure 2.2-1) was considered substantially complete in April The intent of the modifications made to the existing collection facility at Merwin Dam were to provide safe, timely, and effective passage of adult salmonids being transported upstream. The new facility is designed to be constructed in phases, offering the ability to incrementally improve fish passage performance (if needed) in the future to meet biological performance goals. Depending on the biological monitoring of the facility s performance (which is scheduled to begin spring 2015), there are up to four additional phases that will increase flow into the fishway attraction pools, and add a second fishway with additional attraction flow, if necessary (per the Lewis River Settlement Agreement Section ). Phase I represents the initial construction, consisting of four major features: Auxiliary Water Supply Pump Station and Conveyance Pipe Fishway Entrance Number 1 5

8 Lift and Conveyance System Sorting Facility The auxiliary water supply (AWS) system provides pumped water from the tailrace to the fishway entrance pools to attract fish from the tailrace. This system uses hydraulic turbines to power attraction water pumps. Tailrace water is utilized (as opposed to reservoir water) to allow generation with the attraction flow with the high head dam prior to the water s use in the fishway. The AWS system also includes a 108-inch pipeline and conveyance conduits to deliver the water from the tailrace to the lower fishway entrance pools (Pool 1-1). The AWS system has a maximum flow capacity of 400 cfs attraction flow (Phase 1). The entrance of Fishway 1 is located in the tailrace of Merwin Dam adjacent to the discharge of Turbine Unit 1 in the south corner of the powerhouse. The entrance pool (Pool 1-1) contains flow diffusers that introduce the AWS attraction water flow along the walls. The diffusers are made of construction pickets with 7/8-inch clear spacing, with baffle panels mounted immediately upstream of the diffusers to dissipate energy and provide uniform flow across the diffusers. Upstream of the lower entrance pool (Pool 1-1) are a series of ladder steps. The ladder has two intermediate pools (Pool 1-2 and Pool 1-3) leading to a loading pool (Pool 1-4). The fish ladder is designed to operate at 30 cf, and is a vertical slot style fish ladder. Water is supplied from hatchery return line (HR) (~11 cfs) and the ladder water supply (LWS) system (~19 cfs). The vertical slots allow the pool levels to self-regulate the water surface elevations. Depending on tailwater elevation, the designed water elevation changes between pools ranges from 0.25 to 1.0 foot. The loading pool (Pool 1-4) is the last in the fishway, and contains the fish crowder which automatically loads fish into the hopper of the lift and conveyance system. The lift and conveyance system then transports fish from the fish ladder over to the sorting building. Fish are transported from the top of the elevator shaft to the pre-sort pond by the 16-inch diameter conveyance flume (Figure ). Fish are held in the Pre-sort Pond until they are sorted by biologist on a daily basis. 6

9 Hopper sump Pool 1 4 Pool 1 2 Pool 1 3 Pool 1 1 Fishway Entrance 1 Figure 2.2-2: Merwin Sorting Facility ladder entrance and pool configuration. All fish sorting is preformed manually on the sorting table located within the sorting building. Fish are moved from the Pre-sort Pond into the sorting building via a false weir and crowder system. An electro-anesthseia (EA) system temporarily anesthetize the fish to allow easier handling by staff, and to reduce the stress of handling on the fish during sorting. Once sorted, fish are routed into holding tanks for transport by truck to their final destination (i.e., transported upstream, to the hatchery, or returned to the lower Lewis River). The Merwin Fish Collection Facility was operated 24-hours a day through 2014 after final completion except during periods when it was necessary to shut the facility down due to facility modifications, scheduled maintenance, or repairs (Table 2.1-1). 7

10 Table Scheduled MFCF outages occurring after substantial completion (April), Outage Duration Purpose a April 8 th Fish truck occupied Released smolts at acclimation sites a May 1 st June 1, 8, 9, 11, 12 June 19 th June 20 th June 26 th July 2 nd July 13 th July 15 th July 22 nd July 24 th September 4 th Scheduled maintenance - Road paved to Merwin Fish Collection Facility Unscheduled repairs lift and conveyance system repairs Scheduled maintenance attraction water supply repairs Unscheduled repairs - third party damage to fish hopper support beam Unscheduled repairs lift and conveyance system repairs Scheduled maintenance service water supply line repair Unscheduled repairs lift and conveyance system repairs September 10 th September 12 th Scheduled maintenance lift and conveyance system repairs maintenance September 13 th September 15 th Scheduled facility modification presort pond vertical crowder September 26 th September 28 th Scheduled facility modification fish hopper October 27 th October 28 th Scheduled facility modification fish hopper brush install November 25 th November 27 th Scheduled facility modification fish hopper brush install a The fish crowder, lift assembly and pre-sort pond remained operational - only the sorting facility was not operated. 3.0 DOWNSTREAM COLLECTION AND PASSAGE METRICS 3.1 Number of Juveniles Entering Swift Reservoir Overview Developing an annual estimate for total number of juveniles entering Swift Reservoir is required under section of the Settlement and identified as Objective 7 of the M&E Plan. In spring 2014, a single screw trap was installed in the mainstem of the North Fork Lewis River just upstream of the head of Swift Reservoir near the Forest Road 90 Bridge. As outlined in the M&E Plan, subsets of juvenile salmonids collected at the trap daily were marked using Passive Integrated Transponder (PIT) tags. These fish were then either released at the trap to continue their migration downstream into Swift Reservoir, or transported back upstream and released to estimate screw trap efficiency. 8

11 Following the M&E Plan, estimates of emigration were to be developed on a weekly basis for juvenile spring Chinook, coho and steelhead over the out-migration period. Estimating the numbers of juveniles entering Swift Reservoir ( ) was calculated for each species by dividing the total weekly catch by the respective weekly trap efficiency (Equation 3.1-1). Equation N Ent = Total fish of a given species entering Swift Reservoir for the respective week; S sp = Total number of fish of a given species captured in screw trap for the respective week; η sp = Screw trap efficiency for respective week and species. Where weekly trap efficiencies ( ) were calculated as: Equation η sp = Screw trap efficiency for respective week and species; R sp = Total number of recaptured fish for respective week and species; T sp = Total number of released tag fish for respective week and species Results/Discussion In 2014, the screw trap at the upstream end of Swift Reservoir was in operation from March 18 th to June 12 th. During that 12-week period, a total of 189 coho, 214 spring Chinook, and 96 juvenile steelhead were captured (Table 3.1-1). Of these fish, 74% (n = 139) of the coho, 94% (n = 201) of the spring Chinook and 60% (n = 58) of the juvenile steelhead captured were tagged and transported upstream to estimate trap efficiency. A reliable estimate of total number of coho, spring Chinook and steelhead juveniles entering Swift Reservoir could not be calculated in This was the result of low migration numbers of each species combined with very low recapture rates. Changes to the current screw trap study design that should be considered are: Release location of trap efficiency fish: In 2014, trap efficiency fish were released approximately 200 feet upstream of the screw trap. Because of this, it is possible that fish did not distribute evenly, which violates the assumption that smolt releases be 9

12 representative of naturally migrating smolts. An alternative release location further upstream may be needed. Low migration numbers: The low number of smolts migrating past the screw trap was a significant problem for calculating a reliable estimate for upper basin escapement in It is expected that migration numbers will increase as reintroduction efforts continue. Estimating confidence around point estimates: The current M&E Plan does not provide calculations for estimating confidence around point estimates. Because of the inherent variation associated with weekly estimates of smolts entering Swift Reservoir, future studies should incorporate a calculation of variation. 3.2 Fish Numbers Collected at the FSC Overview Section of the Settlement requires PacifiCorp to enumerate the number of salmonids collected at FSC (FSC COL ) by species and life-stage. This requirement is identified as Objective 6 in the M&E Plan. The M&E Plan states that the number of juvenile fish entering the FSC will be calculated through both subsampling and by automatic fish counters. During development of the M&E Plan the accuracy of the automatic fish counters were unknown, thus conducting both methods of enumeration was recommended initially. SUBSAMPLING COUNTS Diversion gates on the FSC allow for smolts to be diverted into either a subsample tank or a general population tank. The diversion gates operate on a time-driven interval within a ten minute time frame (i.e., during a 10 percent sample period the diversion gate would operate one minute out of every ten minute cycle). The intent is that during periods of low migration the sampling rate is set to 100% and all fish collected are processed. When capture rates increase (i.e., during peak outmigration), the samples can be varied between 10, 25, and 50%. That is, only a portion of fish are sampled and the rest are diverted to the general population tanks. As described in the M&E Plan, the daily subsample totals if expanded could then be expanded to estimate the total daily number of fish collected by: Equation FSC COL = Number of fish by species collected each day by the FSC; N SUB = Number of fish by species subsampled each day; S DIV = Diversion gate sampling rate for respective day. 10

13 Table 3.1-1: Estimated number of smolts entering Swift Reservoir during spring 2014 migration season. Tagged refers to the total amount of fish that received a PIT tag and were released upstream of the screw trap to evaluate screw trap efficiency. Week Coho SPCH STLHD Captured Tagged Recap Efficiency Estimate Captured Tagged Recap Efficiency Estimate Captured Tagged Recap Efficiency Estimate N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Total N/A N/A N/A N/A N/A N/A 11

14 However, because daily fish collection numbers remained low in 2014, sample rates were continuously set to 100%. On a daily basis, fish in the subsample tanks were anesthetized, identified to species, measured for length (mm), and enumerated. AUTOMATIC FISH COUNTERS There are a total of six automatic fish counters on the FSC (per design specification AquaScan CSE-1600). They are located in the sorting area on the subsample and general population smolt flumes (port and starboard) and fry flumes. After extensive testing, it was determined that the automatic fish counters are unreliable for long term daily operation Results/Discussion A total 13,855 salmonids were captured by the FSC in 2014 (Table 3.2-1). Of these fish, 11,155 were transported and released downstream of Merwin Dam (Table 3.2-2). Juvenile coho accounted for the highest proportion of the overall catch (68%), followed by spring Chinook (19%), coastal cutthroat trout (8%), and steelhead (5%). A total 648 hatchery rainbow trout and six bull trout were collected in 2014 and returned to the reservoir. ALTERNATIVE METHODS FOR ANALYZING SUBSAMPLED COUNTS Because the automatic fish counters have been shown to be unreliable, PacifiCorp has placed emphasis on expanding the current methods outlined in the M&E Plan for subsampling counts to better estimate total number of fish trapped by species in Based on the methods outlined in the M&E Plan, subsample counts are to be linearly expanded by the respective sampling interval (e.g., if 20 coho are subsampled in a day during 10 percent sampling, it is assumed 200 total coho were trapped (i.e., 20/.1). The major assumption with this expansion is that for any one sampling time frame (current operations use a revolving 10 minute frame) fish concentrations entering the FSC must be constant and made up of a homogenous species composition. We found that this assumption was regularly violated in 2013 and again in 2014 and recommend addressing the issue with statistical modeling. The mechanical design of the FSC provides a unique opportunity to analyze discrepancies between subsample and total counts. The FSC sorting area is essentially a mirror image of itself from port to starboard side in that each side has a smolt sample tank and a general population tank (not sampled). Diversion gates can send fish either to the sample tanks or the general population. At any one time water is diverted to only a single side of the sorting area (i.e., port or starboard sides) by what is known as the rocker gate. The rocker gate is located beneath the smolt sorting bars and has the ability to switch flow routes between port and starboard on a time driven basis. If the diversion gates on both port and starboard sides are set to pass fish into the subsample tanks 100 percent of the time and the rocker gate is set on revolving interval (e.g., 10 percent of the time to port side and 90 percent of the time to starboard side) then normal operations of 12

15 subsample to general population diversions would be mimicked. Hence, the starboard side subsample tank (90 percent of time) would be considered the general population and the port side (10 percent of time) considered the subsample. During periods of the season when handling 100 percent of the daily catch is feasible, PacifiCorp plans to continue utilizing this method by running the rocker gate at different intervals (10, 25, and 50 percent mimicked sampling). Using this information, PacifiCorp plans to develop a model that will be used to estimate total counts of each target species. It is suspected that variations between subsample and total counts will be caused by two main variables, the sampling interval and the amount of fish entering the FSC. The intent is that during normal operations biologists will be able to input each day s subsample counts (by species) into the model and get an estimated total trapped for the day, with associated confidence intervals. An example of what the model is expected to produce is shown in Figure In 2014, PacifiCorp staff began efforts to develop the statistical model for estimating total fish captured by the FSC. In 2014, the rocker gate was operated on a 50% time interval on 25 separate days. At a 50% sampling rate the linear expansion would assume both tanks have an equal amount of fish per species. This assumption was found to be false. The results of this initial analysis indicated that the subsample would result in an overestimation of fish being collected by approximately 6% (Figure 3.2-4). These efforts will continue in 2015 and a model will be developed once enough data is available for reliable statistical power. 13

16 Table 3.2-1: Estimated monthly and annual totals of all species collected at the FSC. Month Coho Chinook Steelhead Cutthroat Fry Smolt Adult Fry Smolt Adult Fry Smolt Adult Kelt Fry < 13 in. > 13 In. Bull Trout Rainbow Trout Total Trapped January February March April May June July August September October November December Annual Total Table 3.2-2: Estimated annual totals of species transported downstream. Coho Chinook Steelhead Cutthroat Fry Smolt Adult Fry Smolt Adult Fry Smolt Adult Kelt Fry < 13 in. > 13 In. Bull Trout Rainbow Trout Total Fish Transported Downstream 0 7, , ,155 14

17 Figure 3.2-3: An example of what the expected outcome of the statistical model for estimating total trapped fishes. **The graphics displayed were generated solely for viewing purposes and have no scientific significance. 500 COHO 50% rocker gate interval Actual Count y = x R² = Linear Approximation Figure 3.2-4: An example of the error associated with the linear expansion used to estimate total fish captured. The y-axis is the actual total coho captured in a day while the x-axis is the linear approximation of total coho captured that same day based solely on the subsample. If the linear expansion were valid the slope of the fitted line (m=0.988) at 50% sample interval would be equal to one. During these 25 days the linear expansion would have overestimated coho counts by 5.6%, or 2,094 coho rather than 1,976 a difference of +118 fish. 15

18 3.3 Juvenile Migration Timing Overview In accordance with section of the Settlement Agreement, PacifiCorp is required to determine natural juvenile migration timing by tracking abundance at the FSC each year. This task was identified as Objective 8 in the M&E Plan. Following the M&E Plan, an index of juvenile migration was developed by tracking the number of fish captured each day at the FSC over time. The number of fish collected each day at the FSC (FSC col ) was calculated by equation , and plotted on a daily basis. In addition to monitoring migration timing, PacifiCorp also monitored smolt fork lengths to describe, temporally, the size (or life-stage) of fish entering the FSC. Size distributions for coho, spring Chinook and steelhead were calculated on a seasonal basis for the periods April-June and October-December. Size distributions were not calculated for January-March as the FSC was not operating during this time period in Size distributions were also not calculated for July-September since few fish were captured during that time period Results/Discussion Overall, the run timing in 2014 followed a normal spring time distribution for rivers west of the Cascade Crest. The peak spring out-migration period generally occurred from the first of April through mid-july. Within this time frame, 79% of the coho, 39% of the spring Chinook, 94% of the steelhead and 73% of the cutthroat were collected relative to the total annual catch (Figures through ). In addition to the spring out-migration period, a large number of spring Chinook were also collected in the fall. This out-migration period occurred primarily from mid- October through late-december, and accounted for approximately 59% of the total annual number of Chinook collected in This trend in Chinook smolts out-migrating in the fall was also observed in 2013, although proportionally a greater number of smolts were seen outmigrating in the fall in 2014 than during the previous year. Coastal cutthroat followed a similar out-migration trends a coho with the majority of fish passing in April and May along with a smaller component of fish out-migrating in the fall. Similar to 2013, a large number of cutthroat fry were collected in late-july. These fish ranged in length from mm and were returned to the reservoir. COHO SIZE DISTRIBUTIONS A bimodal size distribution was observed for juvenile coho collected at the FSC throughout the year, however the mean length of each mode varied by season. In the spring, coho size distribution followed a trimodal pattern; with the new presence of naturally produced fry, which 16

19 were assigned to the 0-60 mm interval (Figure ). During the spring major and minor modes were 165 mm and 275 mm, respectively. The largest size difference occurred during the fall from October through December when the major mode was at approximately 105 mm and the minor mode was at about 275 mm (Figure ). SPRING CHINOOK SIZE DISTRIBUTIONS A bimodal size distribution of spring Chinook was observed in the spring with major and minor modes of 165 mm and 225 mm (Figure ). The 165 mm distribution is thought to be spring Chinook originating from upriver acclimation plants that occurred in the spring of The 225 mm distribution is suspected to be spring Chinook originating from upriver acclimation plants during the spring of 2013 of which, during migrations seasons prior to the spring of 2014, did not successfully emigrate. These multiple distributions of spring Chinook fork lengths were not observed during the spring of From October through December spring Chinook displayed a more normal distribution skewed slightly left with a median of 204 mm (Figure ). Review of spring Chinook data captured at the collector in 2014 reveals size class distribution patterns that positively correlate with hatchery smolt releases. This suggests the majority of spring Chinook collected by the FSC in 2014 originated from the acclimation pond plants. However, adult spring Chinook reintroduction above Swift Dam began in the spring of 2013 and in 2014 a small amount (n=23) of naturally produced spring Chinook parr and fry were collected by the FSC. Naturally produced spring Chinook smolts are expected to arrive at the FSC in the spring of STEELHEAD SIZE DISTRIBUTIONS In the spring, the median steelhead fork length was 201 mm (Figure ). The few steelhead that were captured in the fall (n=30) displayed a variety of life-stages (Figure ). 17

20 Figure 3.1-5: Estimated daily counts of coho smolts captured at the FSC. Figure 3.1-6: Cumulative coho migration timing. 18

21 Figure 3.1-7: Estimated daily counts of spring Chinook smolts captured at the FSC. Figure 3.1-8: Cumulative spring Chinook migration timing. 19

22 Figure 3.1-9: Estimated daily counts of steelhead smolts captured at the FSC. Figure : Cumulative steelhead migration timing. 20

23 Figure : Estimated daily counts of cutthroat captured at the FSC. 21 Figure : Cumulative cutthroat migration timing.

24 30 Coho Size Distribution April June Percentage of Sample FL Interval (mm) n=5,547 Percentage of Sample Coho Size Distribution Oct Dec FL Interval (mm) n=1,527 Figure : Size distribution for coho captured in

25 30 Spring Chinook Size Distribution April June 25 Percentage of Sample FL Interval (mm) n= Spring Chinook Size Distribution Oct Dec Percentage of Sample FL Interval (mm) n=1,290 Figure : Size distribution for spring Chinook captured in

26 30 Steelhead Size Distribution April June Percentage of Sample FL Interval (mm) n= Steelhead Size Distribution Oct Dec Percentage of Sample n= FL Interval (mm) Figure : Size distribution for steelhead smolts captured in

27 3.4 FSC Collection Efficiency Overview The use of radio telemetry to measure collection efficiency (P CE ) of juvenile salmonids at the FSC was evaluated in spring This evaluation was in accordance with Section of the Settlement Agreement and based on recommendations from the 2013 pilot study (PacifiCorp 2014). The M&E Plan defines P CE as the percentage of juvenile salmonids emigrating from Swift Reservoir that is available for collection and that is actually collected. A juvenile that is available for collection is one that is detected within the zone of influence (ZOI); the area 150 feet diameter by 20 foot deep section immediately outside the exclusion net that is influenced by flow entering the FSC. A performance standard of 95 percent or greater for out-migrating smolts was agreed upon for P CE Results/Discussion In total, 193 fish were radio tagged and release approximately 2 miles upstream of Swift Dam. Of these fish, 38 were detected near the entrance of the FSC at the ZOI and 10 were successfully collected for an overall collection efficiency of 26.3% (10 of 38). There were no significant effects of original release location or tagging technique (i.e., surgical vs. gastric) for those fish that successfully passed (p>0.05). However, reservoir temperatures were significantly related to fish passage rate with more fish passing during cooler periods. Regardless of release location (i.e., north or south shoreline), most fish approached the FSC from the southern shoreline where more detections occurred than along the northern shoreline. While overall and species specific collection efficiency rates (P CE ) measured for this study were lower than the designated performance standard of 95%, small sample size, fish stress levels and non-naïve study fish may have impacted these estimates. However, these results using radio telemetry are consistent with detections of PIT tagged-only fish released in the upper basin. A detailed report describing the methods and results of the 2014 effort can be found in Appendix A. 3.5 Swift FSC Injury and Survival Overview Injury and survival of captured salmonid smolts, cutthroat, bull trout, and steelhead kelts were monitored daily on the FSC during 2014 in accordance with Objectives 4 and 5 of the M&E Plan and Section of the Settlement Agreement. As outlined in the M&E Plan, smolt injury and survival was evaluated based on fish collected in the subsample tanks. The methods outlined in the M&E Plan assume that rates of fish injury and 25

28 mortality found in subsampled fish would be representative of the general population. Survival and injury standards that PacifiCorp is required to achieve are displayed in Table Each day the FSC was operational, biologists anesthetized smolts in the subsample, enumerated fish by species, measured fork length, and inspected them for injury or mortality. Classifications for injury types were grouped into three categories: recordable injuries or injuries caused by collection practices that may substantially decrease the chance of surviving; non-recordable injuries or injuries caused by collection purposes that likely will not decrease the chance of survival; and non-trap related injuries or injuries from natural occurrences prior to fish entering the FSC (Table 3.5-2). Table Specified injury and survival standards. Species and Life Stage Recordable Injury Rate Survival Rate Chinook, Coho, Steelhead, Cutthroat Smolts (>60mm) Chinook, Coho, Steelhead, Cutthroat Fry (<60mm) 2.0% 99.5% 2.0% 98.0% Bull Trout 2.0% 99.5% Table Categories used for documenting visible injury at the FSC. Recordable Injury Non-Recordable Injury Hemorrhaging Open Wound (No Fungus) Open Wound (Fungus) Gill Damage Bruising > 0.5 cm diameter Bruising < 0.5 cm diameter Loss Of Equilibrium Descaling > 20% Descaling < or = 20% Any mortality observed in the subsample tank was also recorded. Mortality was classified into two categories, trap related and non-trap related. Biologists utilized various signifiers to determine whether or not mortality was caused by collection practices. Signifiers included presence of fungus, gill coloration, inspection for cause of death (i.e., descaling, brain trauma, predation, hook & line injury), and rigormortis. As specified in the current M&E Plan, injury and survival rates were calculated daily and are shown in Equation and Equation 3.5-2, respectively. Equation R Inj = Observed daily injury rate per species; SS inj = Number of injured fish per species in subsample, mortalities are not included; 26

29 SS Total = Total number of fish per species in subsample, mortalities are not included. Equation CS = Observed collection survival rate per species; M SS = Number of mortalities of a particular species and age class in the subsample; SS Total = Total Number of fish of a particular species and age class in the subsample Results/Discussion INJURY RATE Annual injury rates for target species passing through the FSC ranged from 0 to 2.4 percent (Table 3.5-3). Chinook smolt had the highest overall injury rate (2.4%), followed by steelhead smolt (2.2%), coho smolt (0.9%) and cutthroat (0.8%). Injuries were not observed on any of the bull trout collected. Descaling accounted for the greatest proportion of the injuries observed (greater than 50%) in all species, followed by bruising (14.8%), and eye hemorrhaging (4.9%; Figure 3.5-2). No injuries were observed among coho fry (n=1,520), cutthroat fry (n=150), and steelhead fry (n=3). Overall, annual injury rates for all combined salmonid species (1.3%) met the required performance standard of two percent (Table 3.5-3). PacifiCorp continued to address the causes of injury throughout Debris loading on the fry and smolt separator bars proved to be an ongoing cause of fish injury. As an attempt to reduce this problem, PacifiCorp staffed the FSC around the clock to clear debris from the separator bars during peak migration periods. Table Annual injury rates for smolts. 1 Mortalities with injuries are not assigned as injured fish; they are assigned to mortality totals. 2 The number sampled for injury rate calculations does not include mortalities Species No. Injured 1 No. Sampled 2 Injury Rate (%) Coho smolts 71 7, Chinook smolts 52 2, Steelhead smolts Cutthroat Bull Trout Overall:

30 Figure 3.5-2: Composition of injury type occurrences by species. Percentages reflect numbers collected that are referenced in Table

31 SURVIVAL RATE In the absence of juvenile Release Ponds, annual survival rates were based solely on collection survival (S COL ) because the Release Ponds were not yet constructed in Transported fish were directly released into the Lewis River below Merwin Dam and consequently, a true estimate of transport survival (S TRAN ) was not possible. Annual survival rates among target species passing through the FSC ranged from 98.6 to 100 percent (Table 3.5-4). Juvenile steelhead and bull trout had the highest survival rate (100%), followed by cutthroat (99.7%), coho (99.1%), and spring Chinook (98.6%). No mortality was observed among coho fry (n=1,520), cutthroat fry (n=150), and steelhead fry (n=3). In addition, all seven steelhead kelts that were captured and transported downstream of Merwin Dam survived. Overall, the combined survival rate (S COL ) for all salmonid species (99.1%) was slightly lower than the required performance standard of 99.5 percent (Table 3.5-4). However, only Chinook and coho smolts experienced survival rates lower than the required performance standard. Table Annual survival rates for smolts. Species No. of Mortalities No. Sampled Survival% (CS) Coho smolt Chinook smolts Steelhead smolts Cutthroat Bull Trout Overall: 99.1 While the combined annual survival rate for out-migrants do not meet the performance standard, a temporal change in survival was observed particularly for fish being collected in the spring and early summer. Fish collected earlier in the year had lower mortality rates (i.e., higher survival) than those collected later in the spring (Figure 3.5-2). This was attributed to reservoir water temperatures increasing in late spring and creating less favorable passage conditions. Mean water temperature recorded at the FSC increased from 12 o C in April to 20 o C by mid-july. During these warmer conditions, fish were noticeably stressed and more susceptible to handling mortality. Future operational changes that result in the FSC being turned-off under these warm water conditions are currently being discussed by the Aquatic Coordination Committee (ACC).. 29

32 12 10 FSC Collection Mortality Rate May July, 2014 COHO (n=6,126) n=623 Mortality Rate (%) SPCH (n=905) CT (n=512) ST (n=853) BT (n=5) n=1,078 n=3,394 n=3297 Mortality Standard 0.5% 0 April May June July Month 2014 Figure Monthly mortality rates recorded for all target species at the Swift FSC during spring The dotted line represents the mortality standard of 0.5%. 3.6 Swift Powerhouse Entrainment Evaluation Assessing the proportion of fish entering the intake of the Swift No.1 Powerhouse is required under section of the Settlement Agreement and identified as Objective 3 of the M&E Plan. However, this M&E Objective will not be quantified until downstream passage facilities are installed at Yale and Merwin Dams. 3.7 Overall Downstream Survival (ODS) Overview An estimate of overall downstream survival (ODS) is required under section of the Settlement Agreement and is identified as Object 1 of the M&E Plan. An ODS rate of greater 30

33 than or equal to 80 percent 3 is required. ODS is defined in Section of the Settlement Agreement as: The percentage of juvenile anadromous fish of each of the species (i.e., Chinook, steelhead, coho, and cutthroat) that enter the reservoirs from natal streams and survive to enter the Lewis River below Merwin Dam by collection, transport and release via the juvenile fish passage system, passage via turbines, or some combination thereof, calculated as provided in Schedule That is, ODS is the percentage of the fish entering the Project that migrate, or are transported to the lower Lewis River (i.e., downstream of Merwin Dam) and released successfully (i.e., alive). Initially, pending the development of passage facilities for Merwin and Yale projects, ODS is defined as the survival of anadromous fish from the head of Swift Reservoir to the Lewis River below Merwin Dam immediately at the exit of the Release Ponds. Estimates of ODS are to be initially developed for out-migrating juvenile coho, spring Chinook, and steelhead. An estimate of ODS will also be developed for sea-run cutthroat trout if data indicate that this cutthroat life history is present in the upper Lewis River basin and the number of juveniles produced and collected at the FSC is sufficient for a meaningful estimate. The current plan states that fish will be PIT tagged and released at the head of Swift Reservoir to collect ODS data. The Plan also states that these PIT tagged fish can be collected from screw trap operations or, if needed, from fish collected at the FSC. However, it was suggested that fish already collected at the FSC may have a higher propensity to avoid the FSC if used for ODS data, possibly biasing results. In 2014 only fish collected at the screw trap were PIT tagged for ODS purposes. PIT-tag detectors are located on the FSC and, eventually, the exit of the Release Ponds will be used to monitor passage of tagged fish to estimate ODS. Dead tagged fish found in the FSC and Release Ponds would be assigned to collection loss (S COL ) and transport loss (S TRAN ), respectively. Because the Release Ponds were not yet constructed in 2014, transported fish were directly released into the Lewis River below Merwin Dam. Consequently, a true estimate of S TRAN was not possible. The M&E Plan also calls for 50 dead PIT-tagged fish being released into the FSC over the course of the season as a check on the ability of the biologists to detect and recover dead fish at the Release Ponds. (These actions were also not conducted in 2014). Ultimately, if tag recoveries are less than 100%, estimates of ODS will be adjusted based on the calculated error rate. 3 An ODS of greater than or equal to 80 percent is required until such time as the Yale Downstream Facility is built or the Yale In Lieu Fund becomes available to the Services, after which ODS shall be greater than or equal to 75 percent. 31

34 ODS estimates are to be developed on a weekly basis during the migration season and then expanded to an annual estimate. These estimates will be based on pooling release recapture data over the season. Because a proportion of tagged fish are likely to overwinter in the reservoir, any fish captured in subsequent years will be added to the ODS estimate for their release year. The ODS calculation under the intended operations (i.e., after completion of the Release Ponds) is shown in Equation The ODS calculation used in the 2014 study (absent of S TRAN ) is shown in Equation Equation (with release ponds) S RES = Survival probability through reservoir; S COL = Survival probability through the collector; S TRAN = Survival probability through the smolt transport system. Equation (without release ponds ) S RES = Survival probability through reservoir S COL = Survival probability through the collector S TRAN = Survival probability through the smolt transport system RESULTS/DISCUSSION Only PIT tag interrogations at the FSC recorded on or before December 31 st, 2014 were included in the ODS calculations (Table 3.7-1). No dead PIT tagged fish (pertaining to the ODS study) were found in the FSC. Hence, S COL was considered 100 percent for each species during Since S TRAN was not calculated and assumed to be 100 percent in 2014, thus ODS estimates during the 2014 study were equal to S RES. The M&E Plan calls for 996 tagged fish per species to be released over a six week period during the particular species respective run-timing in order to achieve the desired statistical power. No species received the required 996 tags during a six week period; during the study period only 140 coho, 201 Chinook, and 57 steelhead were released (Table 3.7-1). SUGGESTED STUDY DESIGN CHANGES FOR ODS IN 2014 With few smolts being captured by the screw trap in 2014 there was little confidence in the calculated ODS values. Only smolts captured by the screw trap were used as ODS test fish. This was done as an attempt to reduce biasing results by using smolts already captured by the FSC as test fish. For 2015, PacifiCorp plans to hold importance of achieving proper sample size ( 996 per species in a six week period) over the possibility of biasing results from using FSC captured smolts as test fish. Meaning, if needed, PacifiCorp will PIT tag smolts captured in the FSC and release them back at the head of Swift Reservoir to achieve the desired sample size, 32

35 despite the possibility of biasing results. This practice will continue until there are adequate numbers of out-migrants produced in the upper watershed. The M&E Plan states that screw trap operations would continue into the summer or fall if the Week Coho Chinook Steelhead Released Recaptured SRES% Released Recaptured SRES% Released Recaptured SRES% N/A N/A N/A Annual pilot study indicates that juvenile run-timings extend into such seasons. The pilot study data indicated that juvenile run-timing for all species appears to come to an end during the latter part of June. Therefore, screw trap operations are expected to run from April 1 st to June 30 th in the coming years. Table 3.7-1: Weekly and annual ODS data for each species (functionally S RES ). ODS performance standard for all species is 80 percent. 33

36 4.0 UPSTREAM COLLECTION AND PASSAGE METRICS 4.1 Summary The historic adult fish trap at Merwin Dam was operated by PacifiCorp staff until June 28 th, 2013, when it was decommissioned for construction activities associated with the new passage facility. The new upstream sorting facility at Merwin Dam was put into routine service in April and actively ran throughout the remainder of All adult salmonids collected were identified to species and sorted by origin (i.e., hatchery or wild), broodstock (i.e., hatchery or supplementation), or as upstream target species. A total 31,944 fish were captured (recaptured fish counts include 4,817 hatchery steelhead, 96 wild early run coho, 3 wild late run coho, and 3 wild fall Chinook) at the Merwin Trap in 2014 (Table 4.1-1). Among the species collected, hatchery summer steelhead accounted for the majority of fish captured (n=12,994) followed by early run coho (n=12,136), late run coho (n=3,221), winter steelhead (n=1,601), spring Chinook (n=934), various resident fishes (n=692), and fall Chinook (n=366). A total of 6,464 hatchery summer steelhead were captured at Merwin Trap and marked with a caudal clip. These fish were transported and released back into the lower Lewis River as part of the Washington Department of Fish and Wildlife Fish Recycle Program. A total 4,817 were recaptured at Merwin Trap, which produced a 75 percent recapture efficiency in A total of 1,033 winter steelhead 4, 9,179 early coho, and 42 cutthroat were transported upstream and released above Swift Dam as part of the reintroduction program in 2014 (Table 4.1-2). Of the 1,033 winter steelhead, 1,032 were captured at Merwin Trap, one fish was captured at the Lewis River Hatchery (LRH). Of the 9,179 early coho that were transported upstream, 3,164 were collected at Merwin Trap, and 6,015 were collected at Lewis River Hatchery. All 42 cutthroat transported upstream were collected from Merwin Trap. 4.2 Adult Passage Survival Overview Section of the Settlement Agreement require that upstream passage survival (UPS) of adult salmonids and bull trout being equal to or greater than 99.5%. The methods to calculate adult passage survival are outlined in Objective 9 of the M&E Plan. Adult bull trout and cutthroat trout are defined as fish with fork length greater than 13 inches (330 mm). UPS is defined as the 4 To distinguish the complete winter steelhead run, counts include only winter steelhead transported above Swift Dam in that run year. 34

37 survival from the time adult target species enter the adult upstream facility to their release above Swift Dam. UPS is calculated based on Equation 4.2-1: Equation N = Number of total adults collected AD TRAP = Number of dead adults in trap AD REL = Number of dead adults at release site 35

38 Table 4.1-1: Total fish collected at Merwin Trap during Resident sockeye, cutthroat and miscellaneous fish were not gendertyped. Characteristic AD Clip CWT Wild Wild Recap Wild BWT Recap Misc Species M F J M F J M F J M F J M F M F Not sexed Total % Spring Chinook Fall Chinook Early Coho 3,799 4, , Late Coho 1,187 1, , Summer Steelhead , Winter Steelhead ,601 5 Sockeye Chum 0 0 Salmonid Smolts Cutthroat (>13 inches) Cutthroat (< 13 inches) Rainbow (< 20 inches) Bull Trout (> 13 inches) 0 0 Bull Trout (< 13 inches) 0 0 Whitefish Sucker Pikeminnow Chiselmouth Total 31,

39 Table 4.1-2: Total fish transported above Swift Dam in 2014 (winter steelhead counts include one BWT steelhead captured and transported upstream from LRH). Species Male Female Jack Not sexed Female:Male Ratio Jack:Adult Ratio Total Spring Chinook 0 Early Coho 4,788 4, ,179 Winter Steelhead ,033 Cutthroat >13'' Bull Trout >13'' 0 Total 10,254 37

40 4.2.2 Results/Discussion A total 10,254 adult salmonids (9,179 early coho, 1,033 winter steelhead, and 42 cutthroat) were transported upstream throughout the migration period in Four early run coho were found dead upon release above Swift Dam, resulting in a percent UPS for all transported species. All wild winter steelhead and cutthroat survived the trapping and transport processes resulting in a UPS of 100 percent. No spring Chinook were transported upstream in Adult Trap Efficiency Overview Adult trap efficiency (ATE) is defined in Section of the settlement agreement as: The percentage of adult Chinook, coho, steelhead, bull trout, and sea-run cutthroat that are actively migrating to a location above the trap and that are collected by the adult trap at Merwin Dam. Based on the methods outlined in Objective 10 of the M&E Plan, radio telemetry is to be used to assess ATE. A total of 150 adults from each target species (i.e., spring Chinook, coho, and winter steelhead) will be tagged and released downstream of Merwin Dam. These fish will be monitored at multiple receiver sites as they enter the tailrace and approach the entrance of the new sorting facility. This behavioral information will be used to determine whether any modifications are needed to improved trap efficiency. This study will be conducted over a two year timeframe beginning in early Spawn Timing, Distribution, and Abundance of Transported Fishes Overview Section of the Settlement Agreement identified the need to determine the spawn timing, distribution, and abundance for transported anadromous species that are passed upstream of Merwin Dam. The primary objective of this task is to identify preferred spawning areas in order to: 1) inform revisions to the Hatchery and Supplementation Plan (H&S Plan; PacifiCorp and Cowlitz PUD 2009) and the Upstream Transport Plan (PacifiCorp 2009); and 2) guide the ACC in determining how to direct restoration efforts with the Aquatics Fund. Two methodologies for determining spawn timing, distribution, and abundance of transported fishes were developed. For adult Chinook and coho salmon, comprehensive spawning ground surveys were conducted in the potentially accessible river and stream reaches upstream of Swift Dam. Due to limited access and anticipated heavy snow accumulations during the spawning season for winter steelhead, a combination of aerial radio telemetry surveys, fixed-station radio 38

41 antennas, aerial red counts, and single pass electrofishing surveys for young-of-the-year steelhead (during the following summer) were conducted. A detailed description of each method is outlined in Objective 15 of the M&E Plan Results/Discussion Data collection on the spawn timing, distribution, and abundance of transported fish was completed in mid-november, At the time of this initial review draft, PacifiCorp has not received the results of this 2014 effort. When complete, the results will be attached as Appendix B to this report. 5.0 OCEAN RECRUIT ANALYSIS 5.1 Overview An analysis of ocean recruitment is stipulated in the Settlement Agreement to determine when the hatchery and natural adult production targets established for the upstream passage program were met. These targets were defined in Section 8.1 of the Settlement Agreement and described as: total escapement (fish that naturally spawned above Merwin Dam and hatchery fish) plus harvest (including ocean, Columbia River, and Lewis River Harvest). For this analysis, the average number of ocean recruits over a five-year period will be evaluated. That is, five consecutive brood years. These data will be evaluated to determine if and when hatchery production levels should be altered. A detailed description of the methodology for this analysis is outlined in Objective 12 of the M&E Plan. 5.2 Results/Discussion Ocean recruit analysis was initiated in fall of 2013 and continued through the rest of the year. Half-way through the process of determining a methodology, investigators realized that the use of coded-wire tags (CWT) and the Regional Mark Information System (RMIS) does not account for CWT detection in fish that still have their adipose fin. It was recommended that PacifiCorp come up with an alternative method to determine Ocean Recruits. PacifiCorp is working towards initiating a new contract to develop an applicable method for monitoring Ocean Recruits in PREFORMANCE MEASURES FOR INDEX STOCKS 6.1 Overview The H&S Plan (PacifiCorp and Cowlitz PUD 2009) recommends that other Lower Columbia River stocks be used as index groups to determine whether the success or failure of the Lewis River reintroduction program is the result of in-basin or out-of-basin factors. This would be 39

42 determined by comparing the survival rates of hatchery and natural-origin fish produced in other basins (such as the Cowlitz River) with releases made in the Lewis River. 6.2 Results/Discussion Since adult returns of natural-origin fish from the upper Lewis River have not occurred in numbers large enough for meaningful analysis, this metric will be postponed until larger adult returns are realized. 7.0 REINTRODUCED AND RESIDENT FISH INTERACTIONS 7.1 Overview As called for in Section 9.7 of the Settlement Agreement, PacifiCorp will monitor the interaction between reintroduced anadromous salmonids and resident fish species. Of specific interest to the Settlement parties was the possible effect resident trout released in Swift Reservoir may have on reintroduced salmonids and the effect of anadromous fish introductions on the kokanee populations in Yale Lake. Additionally, concern was expressed that anadromous fish may impact the health of ESA listed bull trout populations. This task is one of the assignments of the Fish Passage Feasibility Study being conducted by the US Geological Survey and University of Washington, Department of Fisheries. 7.2 Results/Discussion In order to provide a meaningful estimate of resident/anadromous interactions, the USGS/UW groups are assessing interactions through stable isotope analysis and analysis of diet from samples within the basin reservoirs and tributaries. There is still a great deal of data collection and sample analyses that need to occur in order to provide some meaningful results. Expectations are that there will be considerable information generated in 2014 that will help to answer the many questions surrounding this topic. In addition, the spawning surveys occurring in Swift and Yale tributaries have and will provide visual observations to assist with this assessment in terms of spawn timing overlap and redd superimposition. 40

43 8.0 LITERATURE CITED PacifiCorp and Cowlitz PUD Lewis River Hatchery Supplementation Plan. Prepared by Jones and Stokes for PacifiCorp Energy and Public Utility District No.1 of Cowlitz County. December PacifiCorp and Cowlitz PUD. 2010, Aquatic Monitoring and Evaluation Plan for the Lewis River. Prepared by ICF, Bioanalysts, Meridian Environmental, Inc. R2 Resources Consultants, and Skalski Statistical Services for PacifiCorp Energy and Public Utility District No.1 of Cowlitz County. June PacifiCorp Lewis River Upstream Transport Plan (Interim Final). Prepared by PacifiCorp. December 18, PacifiCorp Lewis River Fish Passage Program: 2014 Annual Report. FINAL. Prepared by PacifiCorp. April

44 APPENDIX 42

45 APPENDIX A SWIFT RESERVOIR FLOATING SURFACE COLLECTOR JUVENILE SALMON COLLECTION EFFICIENCY 43

46 Cramer Fish Sciences 600 NW Fariss Road Gresham, OR SWIFT RESERVOIR FLOATING SURFACE COLLECTOR JUVENILE SALMON COLLECTION EFFICIENCY 2014 Annual Report Memo - Final Prepared for: Prepared by: Dana Stroud, Forrest Carpenter and Peter Stevens - Cramer Fish Sciences December 2014 Population Recovery and Conservation Habitat Restoration Water Use and Hydropower Forests and Fish

47 EXECUTIVE SUMMARY Cramer Fish Sciences 600 NW Fariss Road Gresham, OR PacifiCorp and Cramer Fish Sciences jointly completed the second year of a study designed to quantify the collection efficiency of a new state-of-the-art floating surface collector (FSC) designed to collect out-migrating juvenile salmonid from Swift Reservoir on the North Fork of the Lewis River, Washington. The Lewis River is a major tributary of the Columbia River- roughly 133 km upstream from the mouth- and it supports anadromous populations of fall Chinook, spring Chinook, coho, and chum salmon, as well as summer and winter steelhead and sea-run cutthroat trout. In 2008, PacifiCorp began releasing adult stocks of winter steelhead, spring Chinook, and coho into Swift Reservoir with the goal of re-establishing natural runs in the upper Lewis Basin. This report summarizes the results from year two of a telemetry study, designed to address the needs of Section 2.2. of the Lewis River Aquatic Monitoring and Evaluation Plan (hereafter M&E Plan ; PacifiCorp and Cowlitz PUD 2010). The M&E Plan describes the need to quantify collection efficiency of the FSC with a target of 95% for juvenile anadromous fish that are available for collection. Following methods outlined in Section 2.2. and recommendations made from the 2013 pilot study (Courter et al. 2013), coho, Chinook and steelhead smolts were initially captured at the FSC, tagged with radio and PIT tags, released back upstream into the reservoir at one of three sites, and then monitored using a fixed-site radio telemetry array designed to detect fish as they made their way back downstream towards the FSC. Aerial and underwater radio antennas were placed within and around the FSC to detect smolt attraction and passage rates. We report on: 1) the attraction and collection efficiency of coho, Chinook, and steelhead smolts at the FSC; 2) preferred approach behaviors of smolts, 3) possible tagging effects on smolts; and 4) potential environmental effects on passage success rates. In total, 193 fish were radio tagged of which 38 fish were detected near the entrance of the FSC at the zone of hydraulic influence (ZOI) and 10 were successfully collected by the FSC for an overall capture efficiency of 26.3% (10 of 38). There were no significant effects of original release location or tagging style for those fish that successfully passed (p>0.05). However, reservoir temperatures were significantly related to fish passage rate with more fish passing during cooler periods. Regardless of release location (i.e. north or south shore), most fish approached the FSC from the southern shoreline where more detections occurred than along the northern shoreline. As a side experiment, eight fish were surgically tagged and released 14.5 km upstream at a rotary screw trap. Seven of the eight (87.5%) fish were detected in the forebay of Swift Reservoir within days following their release, all of which approached on the southern shoreline. None of these fish passed. Cramer Fish Sciences 2

48 METHODS Cramer Fish Sciences 600 NW Fariss Road Gresham, OR The FSC is a floating barge (50 x 20 x 8m) that has been equipped for collection, handling, and transport of juvenile and adult fish located at the south end of the Swift Dam, Lewis River, Washington, USA. Attached to the front of the barge is a metal structure called a Net Transition Structure (NTS), with barrier nets that extend to either shoreline to funnel water and fish towards the FSC. The barge captures fish through an artificial stream channel created by pumps that create a maximum velocity of 7 feet per second with approximately 600 cfs of flow during normal operations (80% of full capacity; R2 and Alden 2013). The channel then guides fish from the NTS into the FSC for safe transport downstream. The Zone of Influence (ZOI) is defined as a roughly 50-meter by 6-meter diameter deep section of water in front of the structure which is influenced by positive flow, designed to be attractive for emigrating fish (PacifiCorp and Cowlitz PUD 2010). Similar to methods described in M&E Plan, a radio telemetry array comprised of five fixed-site stations equipped with Lotek SRX-400 receivers were used to monitor the structure s efficacy. We determine the proportion of radio-tagged fish that enter the ZOI that were later collected by the FSC. Two fixed-site telemetry stations with eight-element antennas were installed on the north and south shores of the reservoir upstream of the FSC to determine out-migrating smolts approach behavior prior to interacting with the FSC and the ZOI (Figure 1). To create a field of detection roughly the same size as the ZOI, a fixed-site telemetry station equipped with two six-element aerial antennas was installed on the gangway at the front of the NTS ( ZOI array ). The ZOI array was range tested and tuned slightly inward to cover the entrance area in front of the FSC, while minimizing coverage of surrounding areas. A fixed-site telemetry station with four underwater antennas was also installed at the entrance to the FSC, at the upstream end of the NTS ( entrance array). Moving closer to the FSC, another station with three underwater antennas was installed in the artificial stream channel between the NTS and the collection facility ( channel array), and proved useful for examining whether fish that entered the FSC were successfully retained. Once successfully captured, a PIT tag array detected fish after they were sorted within the barge. The majority (n=149) of smolts were collected and tagged at the FSC. In addition, a small number (n=8) were collected and tagged at a rotary screw trap (RST) at the east end of the reservoir in an attempt to sample naïve fish in good condition. Fish were held in re-circulating tanks within the FSC, or within the live well of the RST for about 24-hours prior to surgery. In addition to the surgical candidates, a small number of tags were gastrically implanted at the FSC in order to employ all remaining radio tags quickly as we neared the end of the study window. Following tagging and holding at the FSC, fish were transported upstream were approximately 3.2 km east of the entrance to the FSC and released. Detailed tagging methods, pre and posttagging evaluations, and details regarding sedation techniques are provided in Appendix A. Transfer tanks used to move tagged fish from the FSC back to the reservoir were filled with water pulled directly from the reservoir s surface, thus the temperature was representative of Cramer Fish Sciences 3

49 Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Figure 1. Diagram of the study area depicting the five receiver stations comprising the Swift Reservoir floating surface collector telemetry array. Approximate detection ranges are indicated in yellow. the surface conditions. Care was taken to ensure that transport tanks remained within 2 C of the holding tanks at all times to reduce thermal stress. Water-to-water transfers were the preferred technique when moving fish. Tagged fish were divided evenly between two release sites, one on the north side of the reservoir and the other on the south, to determine the effects of release site on approach behaviors. One final release group (n=8 fish) was released at the eastern most end of the reservoir, 14.5 km east (upstream) from the FSC, out of a WDFW-operated rotary screw trap holding pen. There were no appropriate facilities (small circular raceways with adequate long-term holding environment) on site to perform a paired holding study, as specified in the M&E plan. Therefore the telemetry study did not quantify tag failure, tag loss, decay rate or determine post-release mortality. Cramer Fish Sciences 4

50 Water Quality Conditions Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Water temperature within the water tanks at the FSC increased across the study period. On 1 June, the FSC water tanks were an average of 12.3 C, and by 30 June they were an average of 16.3 C. Temperatures within the FSC first reached 18 C on 8 July and 20 C on 12 July. Water temperatures rose on a 24-hour period until approximately 11pm, when water temperatures reached their daily maximums. As reservoir levels fluctuated across the study, we calculated daily average depth of each temperature logger based on its fixed height against the daily reservoir height. We then averaged the temperatures of loggers at each day that were exposed to depths between 2 and 11ft to represent the reservoir conditions experienced by fish detected in this study. Reservoir temperatures captured directly behind the FSC increased across the study, but cooler than the temperatures recorded in the FSC. The temperatures between 2 and 11 ft depth averaged 10.6 C on 1 June 14.9 C on 30 June. Reservoir surface temperatures first reached 18 C (avg 8 ft depth) on 11 July and 20 C (avg 7 ft depth) on 16 July. The difference in daily average temperature between the surface reservoir water versus the FSC ranged from -0.9 to 2.3 C depending on day (average 1.0 C warmer in FSC than surface water, se 0.06). PacifiCorp reported no significant power outages during the course of the study. Analytical Approach An automated computer script was used to filter all capture histories for detections both prior to an individual tag s release and for detections after a confirmed capture at the FSC. Detection histories used for further statistical analysis were comprised of first and last detections at each receiver site (i.e. detection bins), plus PIT tag confirmations when fish entered the FSC. All records were manually reviewed for inconsistencies and then quality assurance/quality control checked by a second biologist. The collection efficiency (P CE ) metric is a component of the whole reservoir survival estimate (S RES ), specified in Section 2.2 of the M&E Plan. For this study, P CE was defined as the percentage of juvenile anadromous fish that were available for collection (i.e. detected at the ZOI) that were subsequently detected within the FSC. A fish was considered detected by the FSC if its detection history showed a logical, plausible progression through the array (e.g. detected at south, ZOI/entrance and channel receivers in sequence over a reasonable period of time) and/or was confirmed by a PIT tag detection within the FSC. As stated in the Lewis River Settlement Agreement, the performance standard for P CE is 95% or greater for smolts. To test the effects of release site and environmental characteristics on P CE, we performed a series of two sided t-tests in Systat 13. Cramer Fish Sciences 5

51 RESULTS Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Collection Efficiency (CE) In total, 193 radio tagged fish were released into the reservoir. Of these, 60 fish were logged for an average of 122 (se, 38) detections. The remaining fish were only detected between 1 to 3 times, most frequently on the south shoreline. Over half (53%, n=102) of the 193 total study fish were released on the south shoreline (Sou), 43% (n=83) were released on the north shoreline (Nor) and 4% (n=8) were released 14.5km upstream at a rotary screw trap (RST) (Table 1). In total, 38 fish were detected by the ZOI receivers at the FSC (31 coho, 3 spring Chinook, 4 steelhead). Of the fish that were released and subsequently detected at the ZOI, 9 coho, 1 steelhead and 0 spring Chinook passed for an overall collection efficacy (P CE ) of 26.3% (10 fish of the 38 detected in the ZOI) (Table 2). PIT tag detections within the FSC were reviewed through mid-november 2014 for fish that may have passed in a fall pulse, but no study fish were identified during this time period. Those smolts that were collected by the FSC were detected passing between 13 May and 20 June Six of the 10 moved into the FSC between the hours of 1007 and 1805 and four were collected between 2158 and 2345 hours. Table 1. Release of tagged fish, organized by species, release site (Nor, Sou, RST) indicating total number and average length. Release Coho Salmon Spring Chinook Steelhead Nor Sou RST n TL (mm) Nor Sou RST n TL (mm) Nor Sou RST n TL (mm) 4/ / / / / / / / / / / Total Cramer Fish Sciences 6

52 Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Table 2. The total number of tagged fish attracted to collection efficacy versus collected by the FSC and subsequent collection efficacy (P CE) calculation by species. Metric Coho Salmon Spring Chinook Steelhead Total Total tagged (n) Detected at ZOI Captured at FSC Collection Efficacy (PCE) 29.0% 0.0% 25.0% 26.3% Release Location & Approach There was no significant difference between passage rates based on release site (Mann-Whitney rank test; U = 2,554; p=0.121). However, this result was likely confounded by small sample size. Two percent (2 of 83) of fish released on the northshore and eight percent (8 of 102) of fish released on the south-shore were collected by the FSC. However, regardless of release location, 80% of fish were first detected on the south shore. There were significantly more detections along the south-shoreline (94%, SD 24%) than the north shoreline (21%, SD 41%) (t-test; p<0.001). Seventy percent of fish that eventually passed through the FSC approached along the southern shoreline, not enough to detect a significant difference by release site (t-test; p = 0.372). Rotary Screw Trap (RST) release Of the eight fish that were released at the RST approximately 14.5 km upstream (east) of the FSC, all but one (7 of 8) were detected by the radio antennas within 15 to 37 days following their release 6 of the 7 that were detected approached the FSC from the southern shoreline. Two of the seven (28.6%) were detected at either the ZOI or entrance at least once, however none of these fish passed. Tagging Method Seventy-one fish were tagged with gastric tags (45%), whereas eighty six (55%) were surgically tagged. Of the eight fish that passed through the FSC, five were surgically tagged and three were gastric tagged. This sample size was too small to conduct any meaningful statistical testing to confirm or reject a difference in passage rate based on tagging method. The average passage rate for fish tagged gastrically was 4.2% (SD 20.3%) and 5.8% (SD 23.5%) for fish tagged surgically. Temperatures Water temperature was monitored in Swift Reservoir just downstream of the FSC at fixed depths of roughly 1, 5, 10, 20, 40, 60, 80, 100, and 122 feet, which were corrected based on reservoir elevation. For the purpose of comparing water temperature to fish passage rates, we assumed that collected fish were primarily active at five feet below the surface because of the confirmed detection ranges of the installed radio telemetry array were less than five feet (range testing at a 10ft depth was not successful at this site). Water temperatures within the reservoir were related to passage success rate (t-test; p=0.008) (Figure 1). The majority of fish passage occurred when surface temperatures remained below 10 C (mid to late May). Temperatures were generally warmest within the upper 20ft (average 10 C in May, Cramer Fish Sciences 7

53 Cramer Fish Sciences 600 NW Fariss Road Gresham, OR increasing to 15 and 16 C by late June). Cooler water (under 10 C) was consistently available below a depth of 60-80ft. Temperature within the FSC holding tanks where fish were collected for surgeries also increased across the study. FSC temperatures were significantly warmer than reservoir surface temperatures (range -0.9 to +2.3 C, avg +1.0 C warmer) (two sample t-test, p value ), likely adding increasing stress to the tagging experience. Figure 2. Reservoir water temperature profiles and fish passage frequency at the FSC. Cramer Fish Sciences 8

54 DISCUSSION Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Overall and species specific collection efficiency (P CE ) rates measured for this study were lower than the designated performance standard of 95%. Small sample size, fish stress levels and nonnaïve study fish may have impacted these estimates. However, the results of this evaluation are consistent with previous upper basin studies. For example, a 2013 pilot study was conducted to approximate the survival of emigrating smolts through Swift Reservoir (S RES ) (PacifiCorp 2014). Naïve smolts (i.e. had not been previously captured by the RST nor the FSC) were captured, tagged, and released via rotary screw trap operations located at the head of Swift Reservoir. Tagged smolts entering the FSC were then detected by an automated PIT tag detector. Reservoir survival estimates were then made by the ratio of recaptured/released PIT tags. Thus, the reservoir survival estimate takes place on a spatial scale from a point at the head of Swift Reservoir downstream to the capture tanks within the FSC. That is, this reservoir survival estimate inherently includes the collection efficiency metric sought after in Section 2.2 of the M&E Plan. Mathematically, the reservoir survival (S res ) can be separated into the probability that a fish released at the screw trap migrates downstream to a point inside the ZOI (P res ) times the probability that a fish within the ZOI is subsequently collected by the FSC (P CE ): S res = P res P CE. By definition, P res - the probability of fish emigrating from the head of the reservoir downstream to the ZOI must always be less than or equal to 1.0, or 100%. Thus, the true value of S res should never be greater than the true values of P CE. However, in 2013 S res was estimated to be 18% for coho which is greater than the estimated P CE value for coho of 6% calculated from the 2013 Swift FSC Pilot Study (Courter et al. 2013). This counterintuitive result of S res greater than P CE could indicate that there was a significant difference in movement rates and survival of naïve PIT tagged in the upper basin versus non-naïve fish tagged at the Swift FSC, or the RT study was more conservative in its interpretation of passed fish and/or that the ZOI was liberal in its detection range. Low sample size is a known complication with both studies (Lewis River Fish Passage Program Annual Report 2013) and could also explain the illogical result. The results of the 2014 Swift reservoir smolt passage study were not available at the time of this writing. However, the current year s study estimated P CE values for coho and steelhead of 29% and 25%, respectively, for If the results of the 2014 Swift Reservoir smolt passage study remain similar to those from 2013, S res would be less than P CE as predicted indicating that one of the above complications was likely the cause of the aberrant 2013 result. Further, such a result would demonstrate that the passage rates measured in this study are logically consistent with the results of similar studies using alternate tracking technologies. In our study salmon and steelhead smolts both showed a distinct preference to approach the FSC along the south shore. This shore is more heavily shaded and possesses a more readily Cramer Fish Sciences 9

55 Cramer Fish Sciences 600 NW Fariss Road Gresham, OR accessible littoral zone likely providing escape cover to migrating smolts. This lends support to speculation that fish rely upon the cover provided by this shaded corridor and/or water hydrology in the Swift forebay and will move to the south side regardless of their original release location. Fish tracking studies assume that there will be no tagging effects (no difference in survival, swimming speeds, timing of migrations, passage rates or entrainment, and risk for predation between tagged and untagged fish), which may be relevant for this study. Many fish showed evidence of stress (i.e. scale loss, lethargy, bruising, parasitism) from their out-migration and capture prior to our surgical and gastric tagging. Some fish were discovered moribund and/or dead at the time of capture, which would be expected to lead to increased post-tagging mortality (Liedke et al. 2012). Despite attempts to limit stress during the tagging process, mortality was still witnessed both before and after tagging. Project protocols attempted to limit stress and tagging effects by carefully adhering to meticulous tagging and evaluation protocols developed by leading surgical tagging entities (Brown et al. 2006, 2010; Axel et al. 2011; Liedke et al. 2012; Wargo Rub 2014). However, we still witnessed mortality before, during, and after tagging events. Of note, 17 individuals (most frequently steelhead and spring Chinook) that we examined had exterior and/or interior copepods and some had considerable gill degradation. Many other fish (coho, Chinook and steelhead) were unusually skinny and had evidence of predatory scarring and associated scale loss. Despite our surgical standards, the temperature difference between the reservoir and the holding tanks within the FSC already was likely a source of stress for these fish. It is likely that a tagging design using fish naïve to the FSC may have different outcomes. Further, water temperatures were consistently warmer in the FSC when compared to the reservoir - by up to 2.3 C - so study organisms likely experienced a greater total thermal change than we were aware of during the course of this study. In the future, consideration should be given to reevaluating the holding design to a location away from the FSC to eliminate this risk. Eighty-eight percent of fish (7 of 8) inadvertently released at the rotary screw trap 14.5 km away from the FSC were detected within the array at some point and two were detected by the ZOI or entrance receivers. This lends further support to the idea that the stress of navigating the reservoir, changes in water temperature from the reservoir to the FSC, and being crowded within the FSC before being tagged may be impacting either the physiology or behavior of smolts. Delayed mortality as a result of accumulated stress from out-migration, collection and tagging is something to consider for future evaluations. Also, behavioral avoidance of trapping facilities by non-naïve adult salmonids is suspected (Keefer et al. 2002). Avoidance behavior would be expected to reduce passage rates for all subsequent passages. These effects whether delayed mortality or behavioral avoidance by non-naïve fish may be impacting passage rate estimates. Future studies have already been planned to incorporate this new information and study fish will be collected from the rotary screw trap for tagging and release. Cramer Fish Sciences 10

56 REFERENCES Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Axel G., J. Beman, R. Brown, B. Eppard, S. Fielding, E. Hockersmith, T. Liedtke, C. Peery, and C. Woodley Surgical protocols for implanting JSATS transmitters into juvenile salmonids for studies conducted for the U.S. Army Corps of Engineers. 21 pp. Brown R.S., D.R. Geist, K.A. Deters, and A. Grassell Effects of surgically implanted acoustic transmitters >2% of body mass on the swimming performance, survival and growth of juvenile sockeye and Chinook salmon. Journal of Fish Biology 69: Courter, I., T. Garrison, F. Carpenter Swift Reservoir Floating Surface Collector Juvenile salmon Collection Efficiency Pilot Study. Report to Pacific Power (PacifiCorp). 3 pp. Keefer, M.L, T.C. Bjornn, C.A. Peery, K.R. Tolotti, R.R. Ringe, and P.J. Keniry Migration of Adult Steelhead Past Columbia and Snake River Dams, Through Reservoirs and Distribution Into Tributaries, Idaho Cooperative Fish and Wildlife Research Unit, University of Idaho. 183 pp. Lewis River Fish Passage Program. Draft 2013 Annual Report. PacifiCorp Energy and Public Utility District No. 1 of Cowlitz County. 39 pp. Liedke, T.L., and A.M. Wargo-Rub Techniques for telemetry transmitter attachment and evaluation of transmitter effects on fish performance, in Adams, N.S., Beeman, J.W., and Eiler, J.H., eds., Telemetry techniques A user s guide for fisheries research: Bethesda, Maryland, American Fisheries Society, p PacifiCorp and Cowlitz PUD. 2010, Aquatic Monitoring and Evaluation Plan for the Lewis River. Prepared by ICF, Bioanalysts, Meridian Environmental, Inc. R2 Resources Consultants, and Skalski Statistical Services for PacifiCorp Energy and Public Utility District No.1 of Cowlitz County. June PacifiCorp Lewis River Fish Passage Program 2013 Annual Report. Prepared by PacifiCorp. April R2 Resources and Alden Research Laboratory Hydraulic Evaluation of Swift Reservior Fish Screens. Prepared for PacifiCorp. June Wargo Rub, M.W., N. Jepsen, T. Liedke, M. L. Moser, and E.P.S. Weber, III Surgical insertion of transmitters and telemetry methods in fisheries research. 15 pp. Cramer Fish Sciences 11

57 APPENDIX I: TAGGING TECHNIQUES Cramer Fish Sciences 600 NW Fariss Road Gresham, OR Pre-Evaluation Surgical outcomes can have significant effects on study results including but not limited to: entrainment, predation, lethargy, morbidity and mortality. Therefore, we used surgical protocols developed by leading fish surgical entities protocols developed by leading Pacific Northwest tagging entities (Brown et al. 2006, 2010; Axel et al. 2011; Liedke et al. 2012; Wargo Rub 2014) to guide our approach. For instance, if fish were in such poor condition that shortterm survival was questionable, then those fish ere not be considered viable surgical candidates. A rejection rate or holding mortality approaching or exceeding 1% for 2-5 d should be an indication of a larger fish health issue and should be evaluated before proceeding with tagging. The rejection protocol that we used for this study was based on standard protocols shared by PNNL and USGS. Fish were inspected pre- anesthesia and post-anesthesia and were rejected from the study if one of the following conditions existed: (1) >20% descaling, (2) significant injury (open wound that would lead to immediate tag loss, significant deformity, bleeding), (3) missing or mostly missing opercles, (4) fungus, (5) moribund from existing disease (such as BKD or GBT), (6) fish that are already tagged, (7) >3 parasites on gills and/or visible gill damage from parasites, and/or (8) condition factor below the 5 th percentile for average length class. Fungus on caudal fin, copepods on gills and operculum & cysts / furuncles: Bacterial kidney disease (BKD),~15% scale loss & skeletal deformation: Anesthesia When working with a new anesthetic protocol or species it is advisable to anesthetize a few fish and follow them through full recovery to insure that drug dosages and techniques are safe and Cramer Fish Sciences 12