MIGRATION, SURVIVAL, GROWTH, AND FATE OF HATCHERY JUVENILE CHINOOK SALMON RELEASED ABOVE AND BELOW DAMS IN THE WILLAMETTE RIVER BASIN

Transcription

1 Work Completed for Compliance with the 2008 Willamette Project Biological Opinion, USACE funding: MIGRATION, SURVIVAL, GROWTH, AND FATE OF HATCHERY JUVENILE CHINOOK SALMON RELEASED ABOVE AND BELOW DAMS IN THE WILLAMETTE RIVER BASIN Prepared for U. S. ARMY CORPS OF ENGINEERS PORTAND DISTRICT WILLAMETTE VALLEY PROJECT 333 S.W. First Ave. Portland, Oregon Prepared by Marc A. Johnson, Thomas A. Friesen, Paul M. Olmsted, and Jason R. Brandt Oregon Department of Fish and Wildlife Willamette Research, Monitoring, and Evaluation Program Corvallis Research Laboratory Highway 34 Corvallis, Oregon Task Order Numbers W9127N , 0009, 0018, 0025, and 0034 September

2 Summary Upper Willamette River (UWR) spring Chinook salmon Oncorhynchus tshawytscha are listed as a threatened species under the U.S. Endangered Species Act (NMFS 1999; NMFS 2005). In its 2008 Biological Opinion (BiOp) for the Willamette River Basin Flood Control Project ( Project ), the National Marine Fisheries Service identified lack of fish passage at Project dams as a major limiting factor to the viability of UWR Chinook salmon (NMFS 2008). The BiOp directed actions to identify, address and reduce impacts from existing dam passage conditions for adult and juvenile Chinook salmon, including continued trap, transport and release of adult salmon into above-dam habitats (Reasonable and Prudent Alternative RPA 4.1), and studies to investigate the feasibility of improving downstream fish passage at Project dams (RPA 4.12) (NMFS 2008). Together, these actions are intended to re-establish viable populations of naturally spawning UWR Chinook salmon in their historical habitats above Willamette Project dams. In 2011, we began releasing PIT-tagged juvenile Chinook salmon at sites above and below Project dams in the Middle Fork Willamette (MFW) River and using detections of these fish at Willamette Falls and other locations to estimate the effects that residence and passage through reservoirs and dams have on the growth, movement patterns and survivorship to adulthood of this species. In 2012, we initiated a parallel study in the North Santiam (NS) River. Results for fish released in were described by Brandt et al. (2015). Here we report results from subyearling Chinook salmon released into the Middle Fork Willamette and North Santiam rivers in 2014, and relate preliminary findings for adult returns from previous years releases. On 5 June 2014, we released 134,785 PIT-tagged hatchery Chinook salmon into the MFW, roughly divided into four equally-sized groups and released at: Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter Dam tailrace (DEX TR) here listed in order from upstream to downstream sites. The proportion of fish detected at Willamette Falls for each of these tagged groups differed in all pairwise comparisons (pairwise z-tests, P < 0.001) and was highest for the LOP TR group (0.0178) and lowest for LOP HOR (0.0009). Median travel rates to Willamette Falls for above-dam groups ( km/d) did not significantly differ from one another (P > 0.05), but all were significantly less than the median travel rate of DXT TR fish (7.0 km/d) (P < 0.05). Mortalities of MFW tagged fish (N = 525) were recorded by other field researchers, with most of these (N = 456) attributable to gillnet captures in reservoirs, followed by avian (N = 43) and piscine (N = 20) predation. Individual growth rates varied greatly for all groups immediately after release, but stabilized by day 30 at mm/d for most individuals observed. For most time periods, growth rates could not be compared among groups due to absence of data for one or more groups. On 9 July 2014, we released 100,426 PIT-tagged hatchery Chinook salmon into the NS, roughly divided into three equally-sized groups and released at: Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR) here listed in order from upstream to downstream sites. The proportion of fish detected at Willamette Falls for each of these tagged groups differed in all pairwise comparisons (pairwise z-tests, P < 0.001) and was highest for the BC TR group (0.0233), lowest for the DET HOR group (0.0091) and intermediate for the DET FB group (0.0167). The number of detections at Stayton Canal and upper Bennett 2

3 Dam (NSS and NSB; total N = 4,789) exceeded that observed at Willamette Falls (SUJ; N = 1,643), and the order of group-proportions detected (greatest to least) at NSS-NSB was consistent with that observed at SUJ, though the ratio of BC TR:DET HOR was markedly greater at NSS-NSB (5.84:1) than at SUJ (2.56:1). Release group travel rates differed between the SUJ and NSS detection sites. Detections at NSS suggested that the median travel rate for the BC TR group (1.26 km/d) was significantly greater than that of above-dam release groups (DET HOR = 0.50 km/d, DET FB = 0.66 km/d), whereas detections at SUJ suggested that the median rate for the BC TR group (1.7 km/d) was significantly less than that of above-dam release groups (DET HOR = 2.10 km/d, DET FB = 2.40 km/d). These disparate travel rates to NSS and SUJ for the BC TR group can be explained by greater representation of yearling smolts at SUJ (45.4% of detections) than at NSS (2.1% of detections), which may suggest that a large fraction of subyearling Chinook salmon migrated past NSS to rear in the lower Santiam or mainstem Willamette rivers, then outmigrated past Willamette Falls as yearling smolts. Alternatively, detection probabilities for juvenile salmon of different age classes may differ greatly between SUJ and NSS, and thereby influence our results. Despite the earlier downstream movements and greater detection frequencies for juvenile Chinook salmon released below Project dams, preliminary counts of adult returns to the upper Willamette River for above-dam release groups matched or exceeded those of below-dam groups in all years and both subbasins. These results suggest some survival advantage for fish that rear in and successfully exit Project reservoirs, perhaps conferred through faster growth and greater survivorship at saltwater entry. However, our growth data cannot directly test this hypothesis and we emphasize that current adult return data are preliminary and do not yet include all age classes for most release groups. 3

4 Table of Contents Summary... 2 Introduction... 5 Study Objectives... 9 Methods... 9 Results Middle Fork Willamette River releases Detections of juvenile Chinook salmon at Willamette Falls Additional detections Willamette Falls detections and movement rates Dam operations Mortalities and live captures Growth rates Survival to adulthood North Santiam River Detections of juvenile Chinook salmon at Willamette Falls Willamette Falls detections and movement rates Additional detections and movement rates Dam operations Mortalities and live captures Growth rates Survival to adulthood Discussion Future Plans and Recommendations Acknowledgments References Addendum: Migration and Survival of Juvenile Steelhead Released above and below Dams in the North Santiam River Appendix: Biomark Tagging Report,

5 Introduction In 1999, the Upper Willamette River (UWR) spring Chinook salmon Oncorhynchus tshawytscha Evolutionarily Significant Unit (ESU) was listed as threatened under the U.S. Endangered Species Act (NMFS 1999) and this status was reaffirmed in 2005 (NMFS 2005). Historically among the most productive of the ESU, the Middle Fork Willamette River (MFW) population suffered a precipitous decline during the past century, primarily caused by impacts from Willamette Valley Project (WVP) dams that block adult migrations to historical spawning grounds (Hutchison et al. 1966; NMFS 2008; Keefer et al. 2010). The National Marine Fisheries Service (NMFS) concluded in their 2008 Willamette Project Biological Opinion (BiOp) that the continued operation and maintenance of the WVP would jeopardize the continued existence of UWR Chinook salmon and winter steelhead O. mykiss (NMFS 2008). Reasonable and Prudent Alternatives (RPAs) and in the BiOp, among others, address downstream fish passage concerns. Four WVP dams, operated by the U.S. Army Corps of Engineers (USACE), are present in the MFW subbasin (Figure 1). The MFW watershed encompasses 3,509 km 2 and joins the mainstem Willamette River at river kilometer (rkm) 299. Transport and release of adult Chinook salmon into historical MFW spawning grounds above Dexter and Lookout Point (LOP) reservoirs began in 1993 (NMFS 2008). Although these actions were originally intended to provide forage for native bull trout Salvelinus confluentus (Johnson and Friesen 2010), the ancillary benefit of augmenting natural production of Chinook salmon in the subbasin soon became a priority (NMFS 2008). Successful spawning above LOP Reservoir was particularly encouraging in light of high pre-spawn mortality and low egg survivorship observed below Dexter Dam (McLaughlin et al. 2008; NMFS 2008; Keefer et al. 2010). However, major challenges accompanied this novel approach towards recovery. For example, high but variable rates of pre-spawn mortality were observed for fish released above LOP Reservoir (Keefer et al. 2010). Perhaps most importantly, downstream passage through LOP and Dexter dams is thought to cause unacceptably high levels of juvenile mortality (NMFS 2008). Similar to LOP and Dexter dams on the MFW, the construction of Detroit and Big Cliff dams (Figure 2) on the North Santiam River (NS) blocked access to an estimated 71% of historic Chinook salmon spawning habitat (Mattson 1948). The NS watershed encompasses 1,980 km 2 and flows into the Willamette River at rkm 174 after joining the South Santiam River. The Oregon Department of Fish and Wildlife (ODFW) began outplanting adult hatchery Chinook salmon above Detroit Dam in 2000 and recent studies have confirmed that natural production occurs there (Monzyk et al. 2011; Romer et al. 2012). As with juvenile Chinook salmon that outmigrate from habitats above LOP Reservoir, fish from the upper NS must pass through two WVP reservoirs and dams (Detroit and Big Cliff) before reaching the mainstem Willamette River. Little is known about the risks associated with reservoir and dam passage on the NS. However, research by Normandeau Associates, Inc. (2010) found that passage through the hydroelectric turbines of Detroit Dam resulted in significantly higher mortality for rainbow trout than other at-dam passage routes. Coded-wire tag data demonstrated that juvenile Chinook salmon released into Detroit Reservoir were generally recovered at lower rates (as adults) than those released below Big Cliff Dam, though direct comparisons were hindered by differences in liberation date and fish size (ODFW, unpublished data). 5

6 Figure 1. The Middle Fork Willamette River and reservoirs associated with US Army Corps of Engineers-operated dams. Black crosses indicate juvenile Chinook salmon release sites for 2014: LOP HOR=Lookout Point head of reservoir, LOP FB = Lookout Point Dam forebay, LOP TR = Lookout Point Dam tailrace and DXT TR = Dexter Dam tailrace. 6

7 Figure 2. The North Santiam River and reservoirs associated with US Army Corps of Engineers-operated dams. Black crosses indicate juvenile Chinook salmon release locations for 2014; DET HOR=Detroit head of reservoir, DET FB=Detroit Dam forebay, and BC TR=Big Cliff Dam tailrace. 7

8 Although adult collection, transport and release appear to provide a means for reestablishing natural Chinook salmon production above WVP dams, juvenile downstream passage conditions at LOP, Dexter, Hills Creek (HCR), Detroit, and Big Cliff dams represent a potentially serious threat to outmigrating juvenile salmonids in the NS and MFW subbasins (NMFS 2008). Currently, fish must pass through precarious hydroelectric turbines or regulating outlets (Čada 2001; Muir et al. 2001; Ferguson et al. 2006). In addition, known and potential predators of juvenile salmonids, including northern pikeminnow Ptychocheilus oregonensis, largemouth bass Micropterus salmoides, walleye Sander vitreus, rainbow trout, and cutthroat trout O. clarkii are present in the study area and may represent a substantial risk (Monzyk et al. 2011; Monzyk et al. 2012). Diet samples collected from crappie (Pomoxis spp.) in HCR and LOP reservoirs anecdotally suggested high levels of predation on PIT-tagged fish released for this study in Residualism of Chinook salmon in WVP reservoirs may also disrupt natural life histories and elevate risks during at-dam passage through size-dependent turbine or spill mortality. Recognizing these potential threats, NMFS (2008) recommended that Action Agencies assess juvenile fish passage through WVP reservoirs (RPA 4.10) and dams (RPA 4.11), as initial steps toward assessing and improving juvenile downstream passage. One alternative to direct passage through reservoirs and dams is to collect juvenile fish at the head of reservoirs and transport them below dams for release. The effectiveness of such an approach in the Willamette Basin, relative to at-dam passage, is unknown. Accordingly, NMFS (2008) recommended that the Action Agencies work to assess the feasibility of collecting juvenile Chinook salmon at the head of WVP reservoirs for subsequent transport and release below dams and plan, design, build, and evaluate a prototype head-of-reservoir juvenile collection facility above either Lookout Point or Foster reservoir, with these actions being preceded by feasibility studies (RPA 4.9). NMFS (2008) further stated that the Action Agencies will investigate the feasibility of improving downstream fish passage at Lookout Point Dam, beginning no later than 2012 (RPA ) and investigate the feasibility of improving downstream fish passage at Detroit Dam, beginning no later than 2015 (RPA ). However, collection, transport and downstream release of juvenile salmon could present new risks and challenges. For example, Keefer et al. (2008) provided convincing evidence that when juvenile salmon are subjected to collection and transport, they tend to stray at higher rates during adult spawning migrations. Head-of-reservoir collection and release operations would also likely place juvenile salmonids in highly degraded rearing habitats below dams (NMFS 2008), where mortality might exceed that of reservoir rearing and at-dam passage. This study was developed to compare the behavior, performance, and survival of hatchery Chinook salmon that pass through USACE reservoirs and dams with those that are released below the dams. This information can be used to estimate the benefits of a head-ofreservoir collection and transport scenario whereby fish are transported around the dams and reservoirs as opposed to the status quo of unassisted passage through the projects. Our study is designed to assess the effectiveness of alternate passage routes on the MFW and NS rivers, as these may influence outmigration behavior and survivorship. Specifically, we will measure and report on the effects that LOP, Dexter, HCR, Detroit, and Big Cliff reservoirs and dams have on Chinook salmon outmigration behavior and survivorship, as contrasted with downstream passage that does not involve these migration barriers. We will determine the effectiveness of these passage options through estimates of successful outmigration past Willamette Falls and 8

9 survivorship to adulthood. Results from this research will provide valuable information for assessing the feasibility and development of head-of-reservoir collection and at-dam passage facilities and address RPA 4.10 (NMFS 2008), which states that, The Action Agencies will, in coordination with and review by the Services, assess juvenile fish passage through the following Project reservoirs: 1) Cougar, 2) Lookout Point and Dexter, 3) Detroit and Big Cliff, 4) Green Peter and Foster, 5) Fall Creek, 6) Hills Creek. This research also addresses RPA 4.11, as it will assess passage survival and efficiency through all available downstream routes of Lookout Point, Dexter, Detroit, and Big Cliff dams. Finally, our research will generate basic information regarding survivorship and outmigration of juvenile Chinook salmon released below WVP dams, thereby providing requisite information for RPA 4.9 (NMFS 2008). This report is primarily relates information for juvenile fish released in 2014 and adult returns of tagged fish in previous years. Movement patterns, growth rates and other information relating to tagged juvenile fish released in previous years are presented or referenced in this report for purposes of comparison, but described in detail by Brandt et al. (2015). Study Objectives 1. Estimate the effect that passage through WVP dams and reservoirs has on outmigration success (counts) and rate (distance/time) by juvenile Chinook salmon in the NS and MFW rivers. 2. Estimate the effect that passage through WVP dams and reservoirs has on survivorship to adulthood for juvenile Chinook salmon in the NS and MFW rivers. 3. Where possible, determine growth rate of recaptured PIT-tagged Chinook salmon and describe differences in growth among release groups (by subbasin). 4. Where possible, describe the fate of PIT-tagged and CWT fish (e.g. lost to predation, captured in fisheries). Methods This work was implemented in the MFW and NS subbasins of the UWR. We PIT tagged juvenile hatchery Chinook salmon and released them at select above- and below-dam locations to estimate the separate effects from reservoir and at-dam passage on outmigration timing, success and, ultimately, survivorship to adulthood. This approach was chosen to provide information relevant to alternate downstream passage strategies (i.e., HOR collection, transport and below-dam release vs. volitional passage) for Chinook salmon produced naturally above WVP dams. Specific release locations for 2014 are provided in Table 1 and indicated in Figures 1 and 2. Tagging and Releases Although our study was designed to provide information relevant to the downstream passage of naturally-produced juvenile salmon, we used sub-yearling, hatchery-origin Chinook salmon for all releases due to the predictable availability of large numbers of fish. Though hatchery- and natural-origin Chinook salmon in our study area are genetically similar (Johnson and Friesen 2014), differences in the morphology, life history, and behavior are known to exist (Billman et al. 2014). However, we assumed that the hatchery-origin fish we used were adequately similar to natural-origin fish so as to provide relevant information. 9

10 Table 1. Information for juvenile hatchery spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in Release locations in the MFW were Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter Dam tailrace (DXT TR). NS release locations were Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). Tags and marks included adipose fin clip (AD) and passive integrated transponder tag (PIT). Temperatures (C ) for rearing ponds, transport trucks and release sites are provided. Release Site Name Middle Fork Willamette Release Location Release Date Release Number Mark and Tag Pond Temp. Truck Temp. Site Temp. Hampton Boat Ramp LOP HOR 5 June ,769 AD PIT Lookout Point Forebay LOP FB 5 June ,544 AD PIT Lookout Point Tailrace LOP TR 5 June ,952 AD PIT Dexter Tailrace DXT TR 5 June ,520 AD PIT North Santiam Hoover Boat Ramp DET HOR 9 July ,478 AD PIT Detroit Dam Forebay DET FB 9 July ,485 AD PIT Packsaddle Park BC TR 9 July ,463 AD PIT

11 Fish were produced and hand tagged at the Willamette (MFW releases) and Marion Forks (NS releases) hatcheries; all were reared in a common environment with the same water source and fed equal rations of food until tagging commenced. All study fish released in 2014 were tagged with 12-mm PIT tags (Biomark Inc., Boise, ID). Length distributions, condition, tagging methods and other biologically relevant information were recorded by Biomark and are included in the Appendix. For some 2011 MFW and 2012 NS releases, fish were tagged with coded-wire tags (CWTs, Northwest Marine Technology Inc., Shaw Island, WA). Protocols for PIT tagging followed those suggested by Prentice et al. (1990), and all fish released, except those in the MFW in 2011, were adipose (AD) fin clipped to identify their hatchery origin. Fork length (FL) data for all PIT-tagged fish were collected by Biomark staff at time of tagging. Chinook salmon should be a minimum 65 mm FL for PIT tagging (R. Richmond, Biomark Inc., personal communication), which is approximately 25-50% larger than the mean length of naturally-produced spring Chinook salmon in the MFW and NS rivers at time of reservoir entry (Monzyk et al. 2011; Romer et al. 2012). Accordingly, fish released for this study were slightly larger than average for the same cohort of naturally-produced fish present in the study areas. Once fish were tagged, hatchery rearing ponds with known tag groups were assigned release locations. Prior to release, ODFW personnel collected length data from at least 100 fish from each release group and verified PIT tag codes. All shed tags were collected from the rearing ponds holding study fish. Fish were released approximately 7-10 days after tagging. In 2014, PIT tagged Chinook salmon were released on June 5 th in the MFW and July 9 th in the NS. As in previous years, release dates were chosen to follow natural reservoir recruitment and outmigration as closely as possible, within the constraints imposed by hatchery operations, tagging and production (see Brandt et al. 2015). In 2014, a total 134,785 and 100,426 PIT tagged Chinook salmon were released for this study into the MFW and NS, respectively. The numbers of fish released at each of the different locations in 2014 were near equal within and between both subbasins (Table 1). Data Collection and Analysis Individual FLs, release dates and locations for all PIT-tagged fish were uploaded to the PIT Tag Information System (PTAGIS; shortly after the releases occurred. Tag detection data were primarily collected at Willamette Falls (PTAGIS site SUJ; Willamette rkm 43) and, for recent NS releases, at Stayton Canal and Upper Bennett Dam (respectively PTAGIS sites NSS and NSB; Santiam rkm 27 and 30). Additional PIT tag detections were recorded at Tryon Creek Mouth (TCM; Willamette rkm 32), lower Columbia River fixed interrogation sites (PD7; Columbia rkm 70) and the NMFS trawl mobile array (TDX) operated near Columbia rkm 75. Schematics of stationary PIT interrogation sites are available through the PTAGIS website ( After release, data for some PIT-tagged fish (i.e. capture date, FL, location) were collected and reported to PTAGIS by ODFW, USACE and other field research crews that encountered them in the Willamette and lower Columbia rivers during the course of their work. These data allowed us to estimate growth and movement rates. In some cases, mortalities from depredation were recorded. 11

12 On January 4, 2016, we queried the PTAGIS database for information from PIT tag detections at the Willamette Falls juvenile bypass and adult fishway (PTAGIS site SUJ and WFF). We used these data to evaluate outmigration success and timing for juveniles released in 2014 and survival to adulthood for Chinook salmon released as juveniles in We used two sample Kolmogorov-Smirnov tests to compare the FL distributions of all fish in each release group with those of fish detected at Willamette Falls for each group. We plotted the cumulative and daily number of PIT tag detections at Willamette Falls (through time) to provide a visual representation of juvenile outmigration success and timing for each release group, and we used pairwise z-tests to compare the proportions of fish detected at Willamette Falls for each release group. Detections at Willamette Falls were also used to estimate movement rates (rkm/d postrelease), so as to evaluate possible dam and reservoir effects on migration timing. Movement rates were calculated as the distance (rkm) traveled from release location to detection at the falls, divided by the number of days at large between release and detection. Shapiro-Wilk test results indicated movement rate data were non-normal or lacked homogeneity of variances (or both) among all release groups. We therefore used Mann-Whitney rank sum and Kruskal-Wallis oneway ANOVA tests, with Dunn s method used for pairwise comparisons, to test for movement rate differences among release groups. Some research activities not associated with this project, particularly those by ODFW that used gillnets in reservoirs, captured and in some cases incidentally killed study fish. Biometric data for these captures were often collected, uploaded to PTAGIS and used here to generate growth rate estimates (FL mm/d). We estimated growth rates by dividing the difference between FLs on the day of tagging and day of capture (or mortality) by the number of days at large. We plotted individual growth rates against days at large to investigate differences in growth rate among release groups in each subbasin. Using data for PIT tag detections in the adult fishway at Willamette Falls, we compared the number of adult returns for release groups within years and subbasins. Because these adult return data are preliminary and do not yet include expected returns for all adult age classes, we did not subject them to formal statistical analyses. Some emergent patterns are, nevertheless, present. In all cases, we performed statistical analyses to compare metrics for releases groups only within release years and subbasins, and we used an a priori critical value of α = 0.05 to qualify the significance of all test results. We used SigmaPlot version 12.5 (Systat Software, San Jose, CA) and S+ v.8.2 (TIBCO Spotfire, Inc.) software for all statistical analyses and to construct figures. Results Middle Fork Willamette River releases Detections of juvenile Chinook salmon at Willamette Falls The median FL of all tagged fish released into the MFW in 2014 was 73 mm (N=134,785), with a median FL of 75 mm for fish released in the LOP TR (N=33,952), 71 mm 12

13 for fish released at the DXT TR (N=33,520), 74 mm for fish released at LOP HOR (N=32,769), and 74 mm for fish released in the LOP FB (N=34,544) (Table 2). The median FL at tagging of MFW fish detected at Willamette Falls was 74 mm (N=1,226), with a median FL of 75 mm for LOP TR-released fish (N=603), 72 mm for DXT TR-released fish (N=462), 79 mm for fish released at LOP HOR (N=28), and 74 mm for fish released into the LOP FB (N=133) (Table 2). Among all 2014 MFW release groups, only DXT TR fish presented a significant difference between the FL distributions for the entire release group and those subsequently detected at Willamette Falls (ks = , P = ; Table 2), characterized by a shift toward longer fish in the detection group relative to the release group (Figure 3). Of the 134,785 PIT-tagged Chinook salmon released into the MFW on 5 June 2014, a total of 1,226 (N = 603 LOP TR; N = 462 DXT TR; N = 28 LOP HOR; and N = 133 LOP FB) were detected in the juvenile bypass at Willamette Falls between 13 June 2014 and 4 January 2016 (Table 2; Figures 4 and 5). The proportions of LOP TR-, DXT TR-, LOP HOR- and LOP FB-released fish that were subsequently detected at Willamette Falls differed significantly in all pairwise comparisons (P < 0.001) and the ratio of proportions for DXT TR to LOP HOR detections at Willamette Falls was 15.3:1 (Table 3). Peak daily detections at Willamette Falls occurred earlier for DXT TR (8 July 2014; N = 38/d) than for LOP TR (28 September 2014; N = 14/d), LOP FB (12 September 2014; N = 6/d) and LOP HOR releases (22 November 2014; N = 4/d) (Figure 4), with 95% of detections for these groups occurring by October 18 th, November 21 st, November 22 nd and December 3 rd, respectively. Additional detections In addition to those tagged fish detected at the Willamette Falls juvenile bypass, three Chinook salmon released at LOP TR in 2014 were detected descending the adult fishway at Willamette Falls (WFF) between 18 and 23 November Another fish released at LOP TR in 2014 was detected in the Columbia River estuary (PD7) on 25 October 2014, as was a DXT TR-released fish (TWX on 9 July 2014). Finally, on 28 July 2015, a single fish released at LOP TR in 2014 was detected in the Middle Harrison side channel (HN2) tributary of the Entiat River in central Washington State, presumably as a mini-jack struck by wanderlust. Willamette Falls detections and movement rates The median duration between date of release and detection at Willamette Falls was 111 d for Chinook salmon released at LOP HOR, 106 d for the LOP FB release, 108 d for the LOP TR release and 41 d for the DXT TR release (Table 4). The median movement rate for fish released at DXT TR (7.0 km/d) was significantly greater than the rates of all above-dam release groups (P < 0.05), among which median travel rates ( km/d) did not differ significantly (P > 0.05; Table 4; Figure 6). 13

14 Table 2. Median and mean fork lengths (FL) at time of tagging for all juvenile hatchery spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in 2014, and of the subset subsequently detected at Willamette Falls. Release locations in the MFW were Lookout Point head of reservoir (LOP HOR), forebay (LOP FB), tailrace (LOP TR) and Dexter tailrace (DXT TR). Release locations in the NS were Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). FL distributions that differed between release and detection groups are indicated with asterisks (two sample Kolmogorov-Smirnov test; P<0.05). Released FL (mm) Detected FL (mm) Release Location N Mean SE Median Range N Mean SE Median Range MFW 2014 LOP HOR 32, LOP FB 34, LOP TR 33, DXT TR* 34, NS 2014 DET HOR 33, DET FB 33, BC TR 33,

15 0.18 Relative frequency (proportion of total) DXT TR Release FLs DXT TR Detect FLs Fork length (mm) Figure 3. Relative frequencies for the fork lengths (FLs) of juvenile spring Chinook salmon released at the Dexter Dam tailrace (DXT TR) in 2014 and FLs at tagging of those DXT TR-released fish detected at Willamette Falls. 15

16 40 30 Lookout Point Head of Reservoir Lookout Point Forebay Lookout Point Tailrace Dexter Tailrace Daily Detections Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Month ( ) Figure 4. Number of daily detections at Willamette Falls for juvenile spring Chinook salmon released at four locations of the Middle Fork Willamette River on 5 June

17 800 Lookout Point Head of Reservoir Lookout Point Forebay Lookout Point Tailrace Dexter Tailrace Cumulative Detections Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Month ( ) Figure 5. Cumulative number of detections at Willamette Falls for juvenile spring Chinook salmon released at four locations of the Middle Fork Willamette River on 5 June

18 Table 3. For tagged juvenile spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in 2014, the proportion of each release group detected at Willamette Falls. Release locations for MFW include Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter Dam tailrace (DXT TR). Release locations for NS include Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). All proportions within both subbasins differed significantly (Pairwise z-tests; P<0.001). Release Location MFW 2014 LOP HOR LOP FB LOP TR DXT TR NS 2014 DET HOR DET FB BC TR

19 Table 4. The mean and median travel times (days) and movement rates (km/day), as determined from PIT tag detections at Willamette Falls, for juvenile spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in Release locations for the Middle Fork Willamette (MFW) River include Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter Dam tailrace (DXT TR). Release locations for the North Santiam (NS) River included Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). Median movement rates within subbasins that do not share the same superscript differed significantly (Mann-Whitney rank sum and Kruskal-Wallis one-way analysis of variance on ranks, P < 0.05). Travel time (d) Movement rate (km/d) Release Location N Mean SE Median Range Mean SE Median Range MFW 2014 LOP HOR a LOP FB a LOP TR a DXT TR b NS 2014 DET HOR c DET FB c BC TR d

20 40 30 Movement rate (km/d) DXT TR LOP TR LOP FB LOP HOR Figure 6. Downstream movement rates (km/d), as determined from PIT tag detections at Willamette Falls, for juvenile hatchery spring Chinook salmon released into the Middle Fork Willamette River on 6 June 2014 at Dexter tailrace (DXT TR), Lookout Point tailrace (LOP TR), Lookout Point forebay (LOP FB) and Lookout Point head of reservoir (LOP HOR). 20

21 Dam operations Similar to that observed in previous years by Brandt et al. (2015), LOP discharge and spill data revealed a prolonged period of increased spill shortly after PIT-tagged Chinook salmon were released into the MFW in 2014 (Figure 7), which may have facilitated some successful downstream passage. However, dam operations in 2014 were more similar to those of 2013, when spill rarely exceeded 1 kcfs during summer months, contrasting with operations of 2011 and 2012, which included periods of > 3 kcfs spill during June and July (Brandt et al. 2015; Figure 7). For LOP Reservoir elevations below 889 ft, spill is achieved through the regulating outlet and not over the surface spillway. However, reservoir elevations exceeded this threshold on all dates depicted in Figure 7, such that all spill plotted was surface spill. Mortalities and live captures A total N = 525 mortalities were recorded for juvenile Chinook salmon released at MFW sites in 2014, and most of these (N = 456) resulted from gillnet captures by ODFW field research crews (Table 5a). Forty-three mortalities were attributed to avian predators, which in some cases transported PIT tags from below-dam release groups (i.e. DXT TR) to above-dam sites. Among piscine predators, ten, eight, and two tags were recovered from white crappie P. annularis, walleye, and largemouth bass, respectively. One hundred eleven tagged Chinook salmon released at MFW sites in 2014 were captured alive by field research crews (Table 5b). Beach seining in reaches below Dexter Dam produced the greatest number of live captures, followed by in-reservoir box traps. Growth rates Fork length data were collected from PIT tagged fish that were either captured alive or as mortalities (N = 536 FL records from 531 fish) by other researchers in the field. These data were then uploaded to PTAGIS. To investigate differences in growth rate patterns for fish released at different sites, we plotted each captured individual s growth (FLCapture FLTagging) against days at large (DateCapture - DateTagging). In our previous report (Brandt et al. 2015), we tested for and reported differences among the median growth rates of fish released at different sites within subbasins and years. However, we observed from our 2014 growth data that captures of fish released at different sites tended to occur on markedly different dates (Figure 8), such that the growth rates observed for different release groups reflected, or were at least heavily influenced by, seasonal patterns of growth that likely affected all release groups. In short, temporal sampling biases for different release groups precluded direct comparisons of growth rates for spring Chinook salmon released at different MFW sites in Reported differences for growth rates of fish released at different MFW in previous years (Brandt et al. 2015) should also be interpreted with caution, as a preliminary review of these results revealed similar temporal relationships for capture rates of different release groups. Nevertheless, FL data for juvenile Chinook salmon captured after their release in 2014 provided valuable growth rate information, primarily for fish released into above-dam habitats (Figure 8). Our data suggest that soon (0-20 days) after release on 5 June 2014, juvenile salmon 21

22 growth rates were lower and more variable (median 0.46 mm/d, SD = 0.59) than observed later in the season (days : median 1.18 mm/d, SD = 0.17). Only four DXT TR release group fish were observed after 20 d post-release (all on day 26), and these presented individual growth rates of 1.08, 1.08, 1.04 and 0.79 mm/d LOP Total Discharge LOP Spill Release Date Flow (KCFS) X 0 May Jun Jul Aug Sep Month Figure 7. Total discharge and spill (kcfs) for Lookout Point Dam (LOP), Middle Fork Willamette River (MFW), for May-September, 2014 (Data from US Army Corps of Engineers, Dashed line indicates the release date (5 June 2014) of tagged juvenile spring Chinook salmon at four MFW sites. 22

23 Table 5a. Mortality information for juvenile spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in Release locations in the MFW were Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter tailrace (DXT TR). NS release locations were Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). Mortalities occurred from ODFW (gillnet, box trap, screw trap, electrofishing, juvenile bypass, and beach seine) and US Army Corps of Engineers (screw trap) research activities, as well as avian and piscine predation. Predatory fish species included walleye (WAL, Sander vitreus), largemouth bass (LMB, Micropterus salmoides), northern pike minnow (NPM, Ptychocheilus oregonensis), white crappie (WCR, Pomoxis annularis), black crappie (BCR, Pomoxis nigromaculatus), cutthroat trout (CT, Oncorhynchus clarkii), and rainbow trout (RBT, Oncorhynchus mykiss). Predation Research Release Location Avian WAL LMB NPM WCR BCR CT RBT Gillnet Box Trap Screw Trap E-fish Beach Seine Juv Bypass MFW 2014 DXT TR LOP TR LOP FB LOP HOR MFW Total NS 2014 BC TR DET FB DET HOR NS Total

24 Table 5b. Non-mortality capture information for PIT-tagged juvenile spring Chinook salmon released into the Middle Fork Willamette (MFW) and North Santiam (NS) rivers in Release locations in the MFW were Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB), Lookout Point Dam tailrace (LOP TR) and Dexter tailrace (DXT TR). NS release locations were Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). Live captures were recorded during ODFW (gillnet, box trap, screw trap, beach seine and Willamette Falls juvenile bypass), US Army Corps of Engineers (screw trap) and other (Columbia River survey) research activities. Release Location Gillnet Box Trap Screw Trap Beach Seine Juv Bypass Col. R. Survey MFW 2014 LOP HOR LOP FB LOP TR DXT TR MFW Total NS 2014 DET HOR DET FB BC TR NS Total

25 DXT TR LOP TR LOP FB LOP HOR Growth rate (mm/day) Days Post-release Dexter Tailrace LOP Head of Reservoir LOP Forebay LOP Tailrace Figure 8. Upper: Box and whisker plot depicts 25 th -75 th percentiles for capture dates (days post-release) of juvenile spring Chinook salmon released on 5 June 2014 into the Middle Fork Willamette River at the Dexter tailrace (DXT TR), Lookout Point tailrace (LOP TR), LOP forebay (LOP FB) and LOP head of reservoir (LOP HOR); within-box lines indicate median capture date for each group and whiskers indicate 90 th and 10 th percentiles. Lower: Individual growth rates of spring Chinook salmon released at different MFW sites, as determined through the difference in fork length at time of tagging and day of capture (i.e., days post-release). 25

26 Survival to adulthood As expected, we have not yet recorded any adult returns for PIT-tagged Chinook salmon released in However, Brandt et al. (2015) previously reported and we expand on preliminary results for adult returns of fish released in At the time of our data query (4 January 2016), a total of 106 PIT-tagged Chinook salmon from this study had ascended the Willamette Falls adult fish ladder (Table 6). Of these fish, 30 had been released at LOP HOR, 4 at LOP FB, and 14 at DXT TR. Interestingly, the number of adult returns for LOP HOR releases met or exceeded the number of adult returns for DXT HOR releases for every release year ( ). Note, however, that these data are preliminary and do not yet include all adult age classes for the larger 2012 and 2013 release groups. North Santiam River Detections of juvenile Chinook salmon at Willamette Falls The median FL of all PIT tagged fish released into the NS in 2014 was 74 mm (N=100,426), with a median FL of 74 mm for fish released in the BC TR (N=33,463), 74 mm for fish released at the DET FB (N=33,485), and 73 mm for fish released at DET HOR (N=33,478) (Table 2). The median FL at tagging for NS fish that were detected at Willamette Falls was also 74 mm (N=1,643), with a median FL of 74 mm for all release groups (Table 2). Two sample Kolmogorov-Smirnov tests provided no evidence for significant difference between the distributions of fork lengths for any release group and its subset detected at Willamette Falls (P > 0.05). Of the 100,426 PIT-tagged Chinook salmon released into the NS on 9 July 2014, a total of 1,643 (N = 780 BC TR; N = 560 DET FB; N = 303 DET HOR) were detected at Willamette Falls between 28 July 2014 and 12 November 2015 (first and last detection records) (Table 2; Figures 9 and 10). The proportions of BC TR-, DET FB- and DET HOR-released fish that were subsequently detected at Willamette Falls differed significantly in all pairwise comparisons (P < 0.001; Table 3), with the largest proportion observed for the BC TR group (0.0233) and the smallest for the DET HOR group (0.0091). Accordingly, the ratio of proportions for BC TR to DET HOR detections at Willamette Falls was 2.56:1. Peak daily detections at Willamette Falls occurred earliest for the DET FB group (6 October 2014; N = 31), followed by the BC TR group (12 October 2014; N = 30), then the DET HOR group (31 October 2014; N = 11) (Figure 10). However, the BC TR group recorded many detections in the following spring (15 March 2015; N = 28) that were not accompanied by similar numbers of detections for other release groups (Figure 9). Juvenile Chinook salmon released into the NS in 2014 were also detected in large numbers at the Stayton Canal (NSS; Figures 11 and 12) and upper Bennett Dam (NSB) PIT tag interrogation sites of the North Santiam River (Table 7). The number of unique detections at 26

27 Table 6. Number of adult returns for PIT-tagged Chinook salmon that were released as juveniles into the Middle Fork Willamette and North Santiam rivers in Release locations for MFW include Lookout Point head of reservoir (LOP HOR), Lookout Point Dam forebay (LOP FB) and Dexter Dam tailrace (DXT TR). Release locations for NS include Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR). For each release year, the set of adult age classes expected to have returned at the time of our data query are indicated, and are incomplete for release years 2012 and Middle Fork Willamette River North Santiam River Release Year Adult Ages LOP HOR LOP FB DXT TR DET HOR DET FB BC TR , 4, , Table 7. Median travel rates (km/day) for juvenile Chinook salmon released into the North Santiam River at Detroit head of reservoir (DET HOR), Detroit Dam forebay (DET FB) and Big Cliff Dam tailrace (BC TR) in 2014, as determined from PIT tag detections at upper Bennett Dam (NSB) and Stayton Canal (NSS) (PD7). Median travel rates differed significantly (P < 0.05) for all value pairs within rows and columns, except where indicated by shared superscript. NSB NSS Release location n Median n Median DET HOR b DET FB a a BC TR b 2,

28 Detroit Head of Reservoir Detroit Forebay Big Cliff Tailrace Cumulative Detections Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Month ( ) Figure 9. Cumulative detections at Willamette Falls of PIT-tagged juvenile hatchery spring Chinook salmon released on 9 July 2014 at three locations of the North Santiam River. 28

29 40 Detroit Head of Reservoir Detroit Forebay Big Cliff Tailrace 30 Daily Detections Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Month ( ) Figure 10. Daily detections at Willamette Falls of PIT-tagged juvenile spring Chinook salmon released on 9 July 2014 at three locations of the North Santiam River. 29

30 500 Cumulative Detections Detroit Head of Reservoir Detroit Forebay Big Cliff Tailrace Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Month ( ) Figure 11. Daily detections at Upper Bennett Dam and Stayton Canal for PIT-tagged juvenile Chinook salmon released into the North Santiam River in 2014 at the Big Cliff Dam tailrace. 30

31 Upper Bennett Dam (NSB) Stayton Canal (NSS) Daily Detections Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month ( ) Figure 12. Daily detections at Upper Bennett Dam and Stayton Canal for PIT-tagged juvenile Chinook salmon released into the North Santiam River in 2014 at the Big Cliff Dam tailrace. 31

32 these two sites 1 was least for the DET HOR group (n = 551), greatest for the BC TR group (n = 3,130) and intermediate for the DET FB group (n = 930). Thirty-five fish were detected at both NSS and NSB. The proportions of each release group detected at NSS and NSB (combined) were DET HOR = , DET FB = and BC TR = ; these proportions all being notably greater than observed for these groups at Willamette Falls (Table 3). Twenty-five additional PIT tag detections were made at other sites, including six detections in the Willamette Falls adult fishway (DET HOR n = 3; DET FB n = 2; BC TR n = 1), attributable to outmigrating juveniles. These six fish were mostly detected in late October and November of 2014, except for a single BC TR detection that occurred on 15 March Eight tagged fish, also released at BC TR, were detected at a new detection site (est. 2014) in the mouth of Tryon Creek near Lake Oswego, between 29 October 2014 and 22 January Four of these fish were detected multiple times at this site and, for two of these, a period >20d separated the first and second detections. Ten yearlings were detected by the Columbia River estuary towed array (TWX). Of these, four from the BC TR group were detected in mid-april 2015, three from the DET HOR group were detected in mid-may 2015 and the remaining three fish, released at DET FB, were detected on 8 and 30 May and 3 June of A single fish, released at BC TR, was detected in the lower Columbia River estuary (PD7) on 11 October Willamette Falls detections and movement rates The median travel time to Willamette Falls for all PIT-tagged juvenile Chinook salmon released into the NS in 2014 was 108 d (n=1,643) with median travel times of 114 d (n=303), 96 d (n=560) and 130 d (n=780) for DET HOR, DET FB and BC TR-released fish, respectively (Table 4; Figure 13). The median movement rates for above-dam release groups (DET HOR 2.1 km/d; DET FB 2.4 km/d) did not significantly differ from one another, but both were significantly greater than the below-dam release group (BC TR 1.7 km/d; Table 4), which included 4.37 times as many yearling outmigrants (N = 354) as both above-dam groups combined (N = 81; Figure 9). Additional detections and movement rates Numerous detections of tagged fish at NSS and NSB allowed us to estimate and compare movement rates of NS release groups to these within-subbasin PIT interrogation sites. Median travel rates to NSB and NSS for DET HOR, DET FB and BC TR release groups are presented in Table 7. Interestingly, while travel rates to NSB and NSS did not differ significantly between the DET FB group (0.68 and 0.66 km/d, respectively), the least (0.32 km/d) and greatest (1.26 km/d) travel rates observed at NSB and NSS, respectively, were for the BC TR group. The greater (faster) travel rate by BC TR fish to NSS, relative to NSB, could not be explained by a significantly greater proportion of yearling smolts among NSB detections (z = , P = 0.208), as yearlings comprised 1.99% of NSS detections and 3.26% of NSB detections. Instead, simply more subyearlings were detected at NSS than NSB, especially during the first three months of detections. The relatively low maximum number of daily detections and variance at 1 Here, only the first detection for each fish was considered 32

33 12 10 Movement rate (km/d) BC TR DET FB DET HOR Figure 13. Downstream movement rates (km/d) of juvenile spring Chinook salmon released into the North Santiam River on 9 August 2014 at Big Cliff tailrace (BC TR), Detroit Dam forebay (DET FB) and Detroit head of reservoir (DET HOR). 33

34 NSB (maximum = 9, SD = 1.43), as compared to NSS (maximum = 111, SD = 19.70), suggest that the latter site may better measure large fluctuations in juvenile salmon abundance. Dam operations On the day of release (9 July 2014) and until 1 September 2014, spill at Detroit Dam was held at or near 500 cfs, with total discharge at or near 1,000 cfs (Figure 14). On 2 September 2014, spill was decreased, then eliminated on 30 September 2014 and did not resume until 17 October Most Chinook salmon that exited Detroit Reservoir did so during these latesummer and fall dam operations, as relatively few fish released above the dam were detected after December (Figure 9). Cessation of spill in September similarly occurred at Detroit Dam in 2012 and 2013 (Brandt et al. 2015). It should be noted that at Detroit Reservoir elevation 1,543 ft and below, spill is achieved through the upper regulating outlet and not through surface spill. Detroit Reservoir fell below this elevation on 25 September 2014 and remained below 1,543 ft throughout the rest of the period plotted in Figure 9. Mortalities and live captures Relatively few mortalities were recorded for juvenile Chinook released at NS sites in 2014, likely due to minimal in-reservoir research activities that might have captured tagged Chinook salmon or their piscine predators. Nevertheless, 17 total mortalities were recorded for the 2014 NS releases (N = 8 BC TR group; N = 9 DET FB group) from screw trapping below Big Cliff Dam (Table 5a). Screwtrapping in the North Santiam River also captured live, tagged Chinook salmon (Table 5b), which provided length-at-date data used for growth rate estimates. Additional live captures of NS-released Chinook salmon were recorded at Willamette Falls and in the Columbia River (Table 5b). Growth rates Similar to growth rate data for MFW-released Chinook salmon, temporal biases associated with captures of different release groups precluded direct comparisons of growth rates among groups. Nevertheless, an interesting pattern of declining growth rate was observed throughout the course of nearly a year for Chinook salmon release at DET FB and BC TR release sites (Figure 15). Survival to adulthood Of the 124,084 PIT-tagged juvenile Chinook salmon we released in to the North Santiam River, a total of 58 had returned as adults to the upper Willamette River by 4 January Similar to results from the MFW, fish released at the head of reservoir site (DET HOR) were represented by as many or more adult returns as juvenile fish released in the same year at below dams and reservoirs (BC TR) (Table 6). We re-emphasize that these adult return data are preliminary, do not include all adult age classes and should be interpreted with caution. 34

35 10 8 DET Total Discharge DET Spill Release date Flow (KCFS) X 0 Jun Aug Oct Dec Feb Apr Jun Month ( ) Figure 14. Discharge and spill (kcfs) information for Detroit Dam in the North Santiam (NS) River for June 2014 May 2015 (Data from US Army Corps of Engineers, Beginning on 25 September 2014, spill plotted in this figure occurred through the upper regulating outlet, and not through surface spill. Dashed line indicates the release date (9 July 2014) of tagged juvenile spring Chinook salmon at various NS sites. 35

36 BC TR DET FB Growth Rate (mm/day) Days Post-release Big Cliff Tailrace Detroit Forebay Figure 15. Upper: Box and whisker plot depicts 25th-75th percentiles for the capture dates (days post-release) of juvenile spring Chinook salmon released into the North Santiam River on 9 July 2014 at Detroit Dam forebay (DET FB) and Big Cliff tailrace (BC TR); within-box lines indicate median capture date for each group and whiskers indicate 90th and 10th percentiles. Lower: Individual growth rates of spring Chinook salmon released at DET FB and BC TR sites of the North Santiam River, as determined through the difference in fork length at time of tagging and on day of capture (i.e., days post-release). 36

37 Discussion In general, the patterns and magnitude of juvenile detections at Willamette Falls for the 2014 release groups were similar to those observed in previous years (Brandt et al. 2015). In both basins there were large and statistically significant differences in the proportion of TR fish detected relative to the above-dam releases; this effect increased as releases moved upstream. The one exception to this pattern was for the 2014 LOP TR group, which outperformed the DXT TR group in terms of total detections at Willamette Falls. Without more detailed information on when these fish passed the dam (as obtained through active tags or at-dam PIT detection arrays), we can only speculate that natural mortality in Dexter Reservoir was low during the time at which that group occupied it (compared to river conditions below), or perhaps river conditions in the Middle Fork or Willamette mainstem improved after the initial pulse of DXT TR fish outmigrated. As in previous years, the discrepancy between Willamette Falls detections of above-dam and below-dam releases was much stronger in the MFW, where only 28 DXT TR fish were detected and the proportional HOR:TR ratio (15.3:1) was the highest for any two groups observed in any year of the study (excluding the Hills Creek Reservoir releases; see Brandt et al. 2015). While our study was not designed to identify the specific sources of presumed mortality, we can infer from other work that in-reservoir predation and impaired rearing conditions in the MFW are the most likely causes of the differential detection rates. Predator abundance is known to be much greater in MFW reservoirs and in-reservoir mortalities related to piscine predation on our tagged fish have only been observed for MFW releases (though monitoring efforts directed towards predation were lower in Detroit Reservoir). Monzyk et al. (2014) estimated juvenile Chinook salmon losses to northern pikeminnow in Lookout Point Reservoir were about 101,995 from April-June Tag recoveries from and anecdotal reports of predation by walleye, largemouth bass, and crappie also suggest frequent predation on juvenile Chinook salmon in the Middle Fork Willamette basin, possibly exacerbated by our large releases of tagged fish. A single crappie examined at Hills Creek Reservoir contained 13 PIT tags from one of our previous release groups (Kelly Reis, ODFW, personal communication). Numerous studies within the Willamette basin (Keefer et al. 2012; Keefer et al. 2013) and throughout the Pacific Northwest (e.g., Raymond 1979; Raymond 1988; Mathur et al. 1996; Muir et al. 2001; Ferguson et al. 2006) have examined the impacts of dam passage on outmigrating juvenile salmonid survival, with results consistently demonstrating direct and indirect dam passage effects. In our study, issues related to direct dam passage certainly appeared to be compounded for fish passing through multiple dams, as evidenced by (for example) the extremely low numbers of fish detected from the Hills Creek Reservoir releases we conducted in 2012 and 2013 (Brandt et al. 2015); to outmigrate successfully, these fish were required to traverse three reservoirs and dams as opposed to two (LOP HOR releases), one (LOP TR) or none (DEX TR). In addition, detections from our FB releases were always greater than HOR releases and lower than our TR releases, further suggesting a relationship between survival and the number of dams and reservoirs passed. As in previous years (Brandt et al. 2015), the lentic reservoir environment also appeared to affect juvenile Chinook salmon movement rates and timing in the MFW, with reservoir 37

38 releases having significantly slower movement rates and later Willamette Falls detection peaks than observed for the TR release. North Santiam movement rates were confounded somewhat because substantial numbers of the BC TR group did not outmigrate until the spring following their release, which resulted in the overall movement rate being significantly slower than for the above-dam releases that generated few yearling migrants. Movement rates we estimated for both basins continue to be substantially slower than those observed in previous studies of Chinook salmon in the Willamette (Friesen et al. 2007; Schroeder et al. 2008) and Columbia (Dawley et al. 1986; Bell 1991; Giorgi et al.1997) rivers, though the importance of this remains unclear. Protracted outmigration has generally been described as unfavorable for Chinook salmon, with delay resulting in prolonged exposure to freshwater predation, disease, parasites, and high temperatures, especially in reservoirs (Park 1969; Raymond 1988). In the Willamette basin, Monzyk et al. (2015) described significantly higher rates of parasitism by copepods on juvenile Chinook salmon rearing in reservoirs compared to those rearing in streams, though the associated impact on survival remains unknown. However, delayed migration could be beneficial when downstream freshwater conditions are poor, as in an extreme drought, or when arrival to the Columbia estuary would coincide with poor ocean conditions. Life-history expression may also play an important role. Schroeder et al. (2016) classified naturally-produced UWR juvenile spring Chinook salmon into at least six distinct life-history types and more generally into two groups, movers (those that migrate as subyearlings) and stayers (which migrate as yearling smolts). Restricted or delayed passage that compels movers into a reservoir-type (stayer) life history could result in residualism, mis-timed physiological development (e.g., smolting) or even conversion to an adfluvial life history, as observed in Green Peter Reservoir by Romer and Monzyk (2014). We suggest that improved downstream passage would favor the fullest possible expression of the range of spring Chinook salmon life history types, supporting population resiliency. Our revised analysis of growth, based on lengths of tagged fish captured after release, suggested that temporal sampling biases for different release groups precluded direct comparisons of growth rates for spring Chinook salmon released at different NS and MFW sites in Nevertheless, other studies have documented high growth rates in reservoirs relative to streams and suggested that these may confer a survival advantage through accelerated growth and high condition factor (ISRP 2011; Monzyk et al. 2013). The benefits of increased growth due to reservoir rearing are debatable. Large fish may have an advantage in competing for food and space, but we are unaware of any studies documenting that these resources are limited in the Willamette basin. In addition, larger body size may result in lower survival during dam passage. In a study conducted at Lookout Point Dam, Keefer et al. (2012) concluded that the risk of dam passage-related mortality to juvenile salmonids increased significantly with fork length, as had been previously documented by Taylor (2000) at the nearby Cougar Dam on the South Fork McKenzie River. Finally, larger outmigrants may return as younger-aged adults, which could potentially reduce population fitness. For example, Tattam et al. (2015) demonstrated that the probability of wild spring Chinook salmon returning as age-3 adults increased with smolt length, and the probability of age-5 maturation was inversely related to smolt condition. Despite the consistently greater detection frequencies and generally faster downstream movement for juvenile Chinook salmon released below Project dams, preliminary counts of adult returns to the upper Willamette River for above-dam release groups matched or exceeded those of below-dam groups in all years and both subbasins to date. A logical working hypothesis is 38

39 that there is a compensatory survival advantage for fish that rear in and successfully exit Project reservoirs, perhaps conferred through faster growth and greater survivorship at saltwater entry (Zabel and Achord 2004). However, we emphasize that current adult return data are preliminary and do not yet include even a single complete cohort. We acknowledge that the limited PIT tag detection infrastructure and uncertainties associated with the fate of released fish in our study create difficulties in making definitive determinations about reservoir and dam effects on outmigrating juvenile Chinook salmon. Willamette Falls is a substantial distance from release locations in the MFW and NS, and no specific efforts were made to collect Chinook salmon between release sites and the falls. The NSS and NSB detection sites may prove to alleviate this concern somewhat; we detected a large number of fish at NSS and proportional group detections were similar to those observed at SUJ. However, neither site has been operational for the full duration of the study, and their detection efficiencies need to be fully assessed. An unknown number of fish released above dams may have passed through them unharmed only to succumb to predation or other stressors later during downstream migration in the MFW, NS, or Willamette rivers. Those would be assumed mortalities from reservoir and dam passage for our analysis, which could inflate perceived effect sizes. The TR releases would also probably experience similar downstream effects, but it is important to recognize that the detection proportions used to evaluate reservoir and dam impacts on juvenile Chinook survival are indices that likely include some mortality not related to dam and reservoir passage. Another element of uncertainty is the detection efficiency at Willamette Falls; mean river discharge can vary widely among seasons and years, affecting the entrainment and detection of tagged fish at SUJ. Using our 2014 MFW releases as an example, these releases were made during a relatively low flow period for the mainstem Willamette River (late spring). The DXT TR group, unimpeded by dams, migrated quickly during a period of presumed high detection efficiency (summer), while the above-dam groups generally lagged behind by several months and passed Willamette Falls during the fall, when flows were higher and detection efficiency was presumably lower. In this case the disparity between TR and HOR releases would be less than we reported using unadjusted counts. Schroeder et al. (2016) developed expansion factors to provide adjusted tag detection counts; we will apply these expansions to our data in future versions of this report, and have suggested additional research below that may help confirm their validity. Despite these uncertainties, results from our study continue to provide evidence that WVP dams and reservoirs affect juvenile Chinook salmon survival, movement rate, and timing in the MFW and NS basins. Though our study was not intended to provide robust estimates of juvenile survival, the results have been very similar among years. Fish released below WVP dams were detected at a higher rate at Willamette Falls in all cases, and successful outmigration declined with the number of dams and reservoirs encountered. Based solely on unexpanded detection counts of similarly-sized release groups, the Projects appear to have an effect on outmigration success that ranges from about 35% to >90% depending on release location, with the MFW consistently having a larger impact than the NS. We acknowledge that many sources of bias are possible, but suggest that these would need to be large and have the same directionality every year to substantially influence the patterns we observed. 39

40 Future Plans and Recommendations We released about 98,800 PIT-tagged juvenile Chinook salmon in the NS in 2015, completing four consecutive years of releases in each basin (MFW = ; NS = ). Due to the extreme drought conditions in 2015, the release planned for DET FB was moved to DET HOR. Initial results from the 2015 Chinook releases will be available in summer In addition, we completed the second of four planned releases of PIT-tagged winter steelhead on 16 November 2015, with about 9,600 fish released at both DET HOR and BC TR. We present results to date from the 2014 steelhead releases in the addendum to this report. A fifth release year for juvenile Chinook salmon is currently planned for 2016; approximately 100,000 fish divided among the three release sites used in In addition, we will conduct the third year of winter steelhead releases, and focus on refining analyses of juvenile detections, outmigration behavior, and adult returns. Importantly, the 2016 run year for Chinook salmon will constitute the return of our first complete cohort (age-6 MFW fish released in 2011) and we can begin calculating juvenile-to-adult survival estimates. Unlike the uncertainties involved with Willamette Falls juvenile fish passage, detection efficiency at the adult fish ladder is close to 100% and should provide reliable estimates of survival to adulthood. However, 2011 was the pilot year for this study; the release group sizes were small and few age 3-5 adult fish have returned (n = 4), so meaningful results are not expected until the 2017 run year. All returns will be complete by 2021 assuming no further releases are conducted. We again recommend conducting research to estimate the entrainment rate of PIT-tagged fish to the Willamette Falls north fish bypass in relation to river discharge, which could provide retrospective estimates of the total number of fish from each release group that successfully reached Willamette Falls, and verify the expansion calculation of Schroeder et al. (2016), thereby improving juvenile survival estimates. Improvements to the existing PIT-tag detection infrastructure or additional monitoring stations would greatly enhance the usefulness of future releases, especially in providing more robust juvenile survival estimates than is currently possible with the existing infrastructure (see Skalski 2016). Additional detection arrays at or near Project dams would also help determine precisely when tagged fish pass the dams and where mortality is most likely to occur; important factors that could not be fully addressed by this study. Acknowledgments We thank Dan Peck and his staff (Willamette Hatchery) and Greg Grenbemer and his staff (Marion Forks Hatchery) for providing and rearing the fish used in this study. Kirk Schroeder (retired) and Suzette Savoie of ODFW contributed to previous iterations of the analyses. Tagging services were coordinated by Ryan Richmond of Biomark, Inc. We thank the many researchers from NOAA, ODFW, Oregon State University, the University of Idaho, USACE, and others who dutifully reported tag detections to PTAGIS. This work was funded by the USACE under Task Order W9127N , administered by Ricardo Walker and Richard Piaskowski. William Muir of NOAA (retired) was instrumental in developing the original project concept. 40

41 References Bell, M. C Fisheries handbook of engineering requirements and biological criteria. Fish Passage Development and Evaluation Program. North Pacific Division, U.S. Army Corps of Engineers, Portland, Oregon. Billman, E. J., L. D. Whitman, R. K. Schroeder, C. S. Sharpe, D. L. G. Noakes, and C. B. Schreck Body morphology differs in wild juvenile Chinook salmon Oncorhynchus tshawytscha that express different migratory phenotypes in the Willamette River, Oregon, U.S.A. Journal of Fish Biology 85: Brandt, J. R., T. A. Friesen, M. A. Johnson, and P. M. Olmsted Migration, survival, growth, and fate of hatchery juvenile Chinook salmon released above and below dams in the Willamette River Basin. Draft annual report to the U.S. Army Corps of Engineers, Portland District. Task Order W9127N Oregon Department of Fish and Wildlife, Corvallis. Čada, G The development of advanced hydroelectric turbines to improve fish passage survival. Fisheries 26(9): Dawley, E. M., R. D. Ledgerwood, T. H. Blahm, C. W. Sims, J. T. Durkin, R. A. Kirn, A. E. Rankis, G. E. Monan, and F. J. Ossiander Migrational characteristics, biological observations, and relative survival of juvenile salmonids entering the Columbia River estuary, Final Report. Bonneville Power Administration and National Marine Fisheries Service, Portland, Oregon. Ferguson, J. W., R. F. Absolon, T. C. Carlson, and B. P. Sandford Evidence of delayed mortality on juvenile Pacific salmon passing through turbines at Columbia River dams. Transactions of the American Fisheries Society 135: Friesen, T. A., J. S. Vile, and A. L. Pribyl Outmigration of juvenile Chinook salmon in the lower Willamette River, Oregon. Northwest Science 81: Giorgi, A. E., T. W. Hillman, J. R. Stevenson, S. G. Hays, and C. M. Peven Factors that influence the downstream movement rates of juvenile salmon and steelhead through the hydroelectric system in the mid-columbia River basin. North American Journal of Fisheries Management 17: Hutchison, J. M., K. E. Thompson, and J. D. Fortune, Jr The fish and wildlife resources of the Upper Willamette basin, Oregon, and their water requirements. Oregon State Game Commission, Portland, Oregon. Project F-69-R-3, Job No.1. Independent Scientific Review Panel (ISRP) Review of the Research, Monitoring, and Evaluation Plan and Proposals for the Willamette Valley Project. Northwest Power and Conservation Council, Portland, Oregon. Johnson, M. A., and T. A. Friesen Spring Chinook salmon hatcheries in the Willamette basin: existing data, discernable patterns and information gaps. Final Report to the U.S. Army Corps of Engineers, Task Order NWPPM-09-FH

42 Johnson, M. A., and T. A. Friesen Genetic diversity and population structure of spring Chinook salmon from the upper Willamette River, Oregon. North American Journal of Fisheries Management 34: Keefer, M. L., C. C. Caudill, C. A. Peery, and S. R. Lee Transporting juvenile salmonids around dams impairs adult migration. Ecological Applications 18: Keefer, M. L., G. A. Taylor, D. F. Garletts, G. A. Gauthier, T. M. Pierce, and C. C. Caudill Prespawn mortality in adult spring Chinook salmon outplanted above barrier dams. Ecology of Freshwater Fish 19: Keefer, M. L., G. A. Taylor, D. F. Garletts, C. K. Helms, G. A. Gauthier, T. M. Pierce, and C. C. Caudill Reservoir entrapment and dam passage mortality of juvenile Chinook salmon in the Middle Fork Willamette River. Ecology of Freshwater Fish 21: Mattson, C. R Spawning ground studies of Willamette River spring Chinook salmon. Fish Commission Research Briefs, Fish Commission of Oregon 1: McLaughlin, L., K. Schroeder, and K. Kenaston Interim activities for monitoring impacts associated with hatchery programs in the Willamette basin. USACE funding: NWPOD-07-FH-02. Oregon Department of Fish and Wildlife, Salem, Oregon. Monzyk, F. R., J. D. Romer, R. Emig, and T. A. Friesen Life-history characteristics of juvenile spring Chinook salmon rearing in Willamette Valley reservoirs. Annual Report to U.S. Army Corps of Engineers, Cooperative Agreement W9127N , Task Order Oregon Department of Fish and Wildlife, Corvallis. Monzyk, F. R., J. D. Romer, R. Emig, and T. A. Friesen Life-history characteristics of juvenile spring Chinook salmon rearing in Willamette Valley reservoirs. Annual Report to U.S. Army Corps of Engineers, Portland District, Task Order W912N Oregon Department of Fish and Wildlife, Corvallis. Monzyk, F. R., J. D. Romer, R. Emig, and T. A. Friesen Life-history characteristics of juvenile spring Chinook salmon rearing in Willamette Valley reservoirs. Annual Report to U.S. Army Corps of Engineers, Portland District, Task Order W912N Oregon Department of Fish and Wildlife, Corvallis. Monzyk, F. R., R. Emig, J. D. Romer, and T. A. Friesen Life-history characteristics of juvenile spring Chinook salmon rearing in Willamette Valley reservoirs. Annual Report to U.S. Army Corps of Engineers, Portland District, Task Order W912N Oregon Department of Fish and Wildlife, Corvallis. Monzyk, F. R., T. A. Friesen, and J. R. Romer Infection of juvenile salmonids by Salmincola californiensis (Copepoda: Lernaeopodidae) in reservoirs and streams of the Willamette River basin, Oregon. Transactions of the American Fisheries Society 144: Muir, W. D., S. G. Smith, J. G. Williams, and B. P. Sandford Survival of juvenile salmonids passing through bypass systems, turbines, and spillways with and without flow 42

43 deflectors at Snake River dams. North American Journal of Fisheries Management 21: NMFS (National Marine Fisheries Service) Endangered and threatened species: Threatened status for three Chinook salmon Evolutionarily Significant Units in Washington and Oregon, and Endangered status of one Chinook salmon Evolutionarily Significant Units in Washington; Final rule partial 6-month extension on final listing determinations for four Evolutionarily Significant Units of West Coast Chinook salmon; proposed rule. Federal Register 64(56): NMFS (National Marine Fisheries Service) Endangered and threatened species: final listing determinations for 16 Evolutionarily Significant Units of West Coast salmon, and final 4(d) protective regulations for threatened salmonid Evolutionarily Significant Units. Federal Register 70(123): NMFS (National Marine Fisheries Service) Willamette River basin project biological opinion. NOAA s National Marine Fisheries Service, Northwest Region, Seattle, Washington. F/NWR/2000/ Normandeau Associates, Inc Estimates of direct survival and injury of juvenile rainbow trout (Oncorhynchus mykiss) passing spillway, turbine, and regulating outlet at Detroit Dam, Oregon. Draft Technical Report to the U.S. Army Corps of Engineers, Portland District-Willamette Valley Project, Portland, Oregon. Park, D. L Season changes in downstream migration of age-group 0 chinook salmon in the upper Columbia River. Transactions of the American Fisheries Society 98: Raymond, H. L Effects of dams and impoundments on migrations of juvenile chinook salmon and steelhead from the Snake River, 1966 to Transactions of the American Fisheries Society 108: Raymond, H. L Effects of hydroelectric development and fisheries enhancement on spring and summer chinook salmon and steelhead in the Columbia River basin. North American Journal of Fisheries Management 8:1-24. Romer, J. D., F. R. Monzyk, R. Emig, and T. A. Friesen Juvenile salmonid outmigration monitoring at Willamette Valley project reservoirs. Annual Report to U.S. Army Corps of Engineers, Cooperative Agreement W9127N , Task Order Oregon Department of Fish and Wildlife, Corvallis, Oregon. Romer, J. D., and F. R. Monzyk Adfluvial life-history in spring Chinook Salmon from Quartzville Creek, Oregon. North American Journal of Fisheries Management 34: Schroeder, R. K., K. R. Kenaston, and L. K. McLaughlin Spring Chinook salmon in the Willamette and Sandy rivers. Progress Reports , Fish Research Report F-163- R-11/12. Oregon Department of Fish and Wildlife, Portland, Oregon. 43

44 Schroeder, R. K., L. D. Whitman, B. Cannon, and P. Olmsted Juvenile life-history diversity and population stability of spring Chinook salmon in the Willamette River basin, Oregon. Canadian Journal of Fisheries and Aquatic Sciences 73. DOI dx.doi.org/ /cjfas Skalski, J. R Review of tagging study designs to estimate reservoir passage survival in the Willamette Valley Project. Final Report to U.S. Army Corps of Engineers, Portland District. University of Washington, Seattle. Tattam, I. A., J. R. Ruzycki, J. L.McCormick, and R. W. Carmichael Length and condition of wild Chinook Salmon smolts influence age at maturity. Transactions of the American Fisheries Society 144: Taylor, G. A Monitoring of downstream fish passage at Cougar Dam in the South Fork McKenzie River, Oregon Oregon Department of Fish and Wildlife technical report available at (30 March 2016) Zabel R. W. and S. Achord. Relating size of juveniles to survival within and among populations of Chinook salmon. Ecology 85(3):

45 Addendum Work Completed for Compliance with the 2008 Willamette Project Biological Opinion, USACE funding: Migration and Survival of Juvenile Steelhead Released above and below Dams in the North Santiam River 45

46 Work Completed for Compliance with the 2008 Willamette Project Biological Opinion, USACE funding: MIGRATION AND SURVIVAL OF JUVENILE STEELHEAD RELEASED ABOVE AND BELOW DAMS IN THE NORTH SANTIAM RIVER Prepared for U. S. ARMY CORPS OF ENGINEERS PORTAND DISTRICT WILLAMETTE VALLEY PROJECT 333 S.W. First Ave. Portland, Oregon Prepared by Marc A. Johnson, Thomas A. Friesen, and Paul M. Olmsted Oregon Department of Fish and Wildlife Willamette Research, Monitoring, and Evaluation Program Corvallis Research Laboratory Highway 34 Corvallis, Oregon September