S. L. Schroder3, A. Fritts3, M. V. Johnston2, Trends in Demographic and Phenotypic Traits of Hatchery- and Natural-Origin Upper Yakima River Spring Chinook Salmon C. M. Knudsen1, W. J. Bosch2, S. L. Schroder3, A. Fritts3, M. V. Johnston2, and D. E. Fast2 1 Oncorh Consulting 2 Yakama Nation 3 Washington Department of Fish and Wildlife

Objectives How has the mean size-at-maturity of UY natural and hatchery populations changed over time relative to Naches wild control? Has the proportion of natural or hatchery origin jacks changed over time? Does the proportion of jacks naturally spawning or used as broodstock influence future jack production? Does the sex composition of natural and hatchery populations differ? Has mean natural and hatchery origin egg size changed over time? Do PIT and non-PIT SARs still differ?

Definitions Natural Origin (NO)– progeny of naturally spawning parents. Parents could be natural or hatchery origin. Hatchery Origin Supplementation Hatchery (SH) Origin – Parental broodstock of natural origin only, one generation of domestication. Supplement naturally spawning population, integrated population. Hatchery Control (HC) Origin – Parental broodstock of hatchery origin only. Multiple generations of domestication. Are not allowed to naturally spawn, segregated hatchery population. Wild Control – Naches population

The YKFP spring chinook hatchery program was designed to minimize domestication effects use only natural-origin broodstock take no more than 50% of the NO returns into the hatchery utilize factorial crosses during artificial matings limit the proportion of NO jacks in the broodstock Mean = 12.7% (range 6.0 – 29.2) randomly mate individuals use “best culture practices” such as low rearing densities volitionally release juveniles at sizes larger than, but comparable to, wild-origin smolts

Yearling Smolt Fork Length at Chandler 2006 (BY2004) (LOWESS trend lines shown) 83 93 103 113 123 133 143 Julian date 80 100 120 140 160 180 Fork length (mm) Hatchery origin (○) Natural origin (x) LOWESS trend lines were used to indicate trends over time.

Yearling Smolt Fork Length at Chandler 2007 (BY2005) 160 150 Hatchery origin (○) 140 130 Fork length (mm) 120 110 Natural origin (x) 100 90 88 98 108 118 128 138 148 Julian date

Yakima River Basin

Roza Dam Adult Monitoring Facility All adults passing upstream are examined, hatchery fish identified by marks, and natural origin broodstock collected and transported to the Cle Elum Hatchery Facility..

Trends In Traits Over Broodyears 900 Naches 850 NO 800 SH Phenotypic trait 750 HC 700 LOWESS trend lines - For each X value where a Y value is to be calculated, the LOWESS technique performs a regression on points in a moving range around the X value, where the values in the moving range are weighted according to their distance from this central X value. 650 600 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Broodyears

Size-at-Age Mean POHP Differences 2000-2004: age 3: 2.7 cm; age 4: 1.7 cm; age 5: 2.7 cm Mean ByWt Differences 2000-2004: age 3: 0.3 kg; age 4: 0.3 kg; age 5: 0.8 kg Representing a divergence in body size of between 0.5 and 1.0 SD

Age 3 Jacks POHP length (cm) Brood Year UpYak NO Δ Hat Control Supp. Hatch □ 46 45 44 43 POHP length (cm) 42 41 40 39 1995 2000 2005 2010 Brood Year

Linear Trends in Length - Age 3 46 NO 45 Regression (SH and NO) r2>0.48, p<0.02 ANCOVA Equal slopes p=0.565 44 SH HC 43 POHP (cm) 42 41 What’s missing? No Naches wild control samples. Of the past 12 years 10 had sample sizes of just 0 or 1. 40 SH NOR HC 39 38 1995 2000 2005 Brood Year

Jack Size-at-Age Age 3 jacks have significantly increased in size over time in both hatchery and natural origin UY populations at the same rate (cm/yr) NO jacks are significantly larger than SH jacks and have been since the first generation of hatchery jack production Yet, SH smolts are equal to or larger than NO smolts, typically leading to larger size-at-age but younger age-at-maturation

Trends in Length – NO, Naches Age 4 2000 2002 2004 2006 1998 Brood Year 58 60 62 64 66 POHP length (cm) SH Nach NO HC

Linear Trends in Length - Age 4 1998 2000 2002 2004 2006 Brood Year 57 58 59 60 61 62 63 64 65 66 POHP (cm) Regression r2=0.06, p>0.84 ANCOVA Equal slopes p=0.947 Naches NO

No significant linear trend in SH Nach NO HC 66 No significant linear trend in Naches, UY NO, or SH (p=0.32) 64 62 POHP length (cm) 60 58 1998 2000 2002 2004 2006 Return Year

Age 4 Size-at-Age While differing in mean size, Naches (a true wild control population) and U Yakima NO age 4 fish exhibited the same lack of trend in size over time SH origin population also showed no significant trend over time, but is converging on the NO population No apparent negative effect (decrease in mean size) on NO fish despite 56% of naturally spawning fish are hatchery origin on average

Gender Identification From 1997 to 2009 we estimated gender ratios from fish collected at RAMF and then taken to CESRF and held to maturity All adipose fin clipped fish were inspected and gender identified visually, but errors were approximately 20-30% for males and 10% for females Beginning in 2010, an ultrasound device (Honda HS-101V) was used on all fish passing RAMF increasing gender-specific data by more than an order of magnitude

Sampling At Roza Adult Trap Video provided by P. Huffman, YN Joe Hoptowit Checked for ad clip, elastomer and CWT tags Weighed and measured length Identified gender PIT tagged DNA sampled Put a caudal fin clip to ID as HC or SH Video provided by P. Huffman, YN

Sampling At Roza Adult Trap 28 seconds total time, ~1 second gender ID Checked for ad clip, elastomer and CWT tags Weighed and measured POHP and FL Identified gender PIT tagged DNA sampled Joe Hoptowit Checked for ad clip, elastomer and CWT tags Weighed and measured length Identified gender PIT tagged DNA sampled Put a caudal fin clip to ID as HC or SH

Gender Identification 2010 There were a total of 624 fish classified to gender using ultrasound and then held to maturity at CESRF 621 (99.5%) were corrected classified All 9,749 fish passing RAMF were classified to gender in 2010

Gender Comparison BY2006 (Ages 3 and 4) 30 40 50 60 70 Female Male Percent HC (n=1137) SH (n=7325) NO (n=3470) Pearson X2 = 134.01, df = 2, p<0.001

Gender Comparison BY2006 (Age 4 only) Pearson X2 = 5.21, df = 2, p = 0.074 20 30 40 50 60 70 Female Male Gender Percent HC (n=743) SH (n=5774) NO (n=2734)

Why aren’t the F:M ratios of age 4 hatchery fish skewed toward females more than Natural Origin fish? BY 2006 had 40-54% minijack production in SH and HC males BY2006 SH and HC jack production was ~50% greater than NO fish (36% vs 22%) Yet, there was <3% difference in the proportion of adult age 4 females and males and no significant difference even with very large sample sizes More work is needed to understand this issue What are wild precocious male production rates? Compare Female recruits/Female spawner rates

Trends In Jack BY Proportions

0.4 0.3 Prop Jacks 0.2 0.1 0.0 Brood year UpYak NO Δ Hat Control ○ Supp. Hatch □ 0.4 0.3 Prop Jacks 0.2 0.1 0.0 1997 1999 2001 2003 2005 2007 Brood year

UpYak NO Δ Hat Control Supp. Hatch □ 0.4 RY2009 0.3 Broodyear 2006 (RY 2009) was significantly higher for both NO and SH populations. Prop Jacks 0.2 0.1 Broodyear 2006 was significantly higher for both NO and SH populations. 0.0 1997 1999 2001 2003 2005 2007 Brood year

UpYak NO Δ Hat Control Supp. Hatch □ 0.4 0.3 Broodyear 2006 (RY 2009) was significantly higher for both NO and SH populations. Prop Jacks 0.2 0.1 0.0 1997 1999 2001 2003 2005 2007 Brood year

Proportion Age 3 Jacks Over Time SH NO HC 0.4 Natural Origin 0.3 R2 = 0.486, p=0.01 Proportion Jacks 0.2 0.1 Significant trend drivcen by the 2006 point. Remove that and the regerssion is no longer significant and line is flat. 0.0 1998 2000 2002 2004 2006 Brood Year

Proportion Age 3 Jacks Over Time SH NO HC 0.4 SH Origin 0.3 Proportion Jacks 0.2 R2 = 0.581, p=0.01 0.1 0.0 1998 2000 2002 2004 2006 Brood Year

Proportion Age 3 Jacks Over Time SH NO HC 0.4 HC Origin 0.3 Proportion Jacks 0.2 R2 = 0.002, p=0.91 0.1 0.0 1998 2000 2002 2004 2006 Brood Year

Proportion Jacks Spawned vs Proportion Jacks Produced 0.1 0.2 0.3 0.4 0.5 Proportion Naturally Spawning Jacks (Return Year) Prop. jacks produced (Brood year) Natural Origin Jacks ( ) BY = (0.0195(RY)) + 0.1204 R 2 = 0.010, p=0.783, n=10 Natural Spawning 2009

Prop. jacks produced (Brood year) Proportion Jacks Spawned vs Proportion Jacks Produced Artificial Spawning Hatchery Origin Jacks (Δ) BY = 0.253 - (0.415(RY)) R 2 = 0.302, p=0.100, n=10 0.4 2009 0.3 0.2 Prop. jacks produced (Brood year) 0.1 0.1 0.2 0.3 0.4 0.5 Proportion Jacks Broodstock (Return Year)

Linear Trends in Spawn Timing

Linear Trends in Spawn Timing Regression R2<0.16, p>0.17 ANCOVA Equal slopes p=0.872 Nach<NO Hat=NO R NO hatchery 2.3 days earlier than NO river Based on redd counts NO river 270 NO hatchery Median Julian Spawn Date Naches 260 NO Hatchery median is 2.3 days earlier than NO river (redds) Naches is 12.3 days earlier than NO river 250 1995 2000 2005 2010 Return Year

Linear Trends in Spawn Timing Naches SH HC NO hatch NO river 270 NO hatchery Median Spawn Date HC 260 SH NO Hatchery median is 2.3 days earlier than NO river (redds) 250 1995 2000 2005 2010 Return Year

Trends In Egg Size (Mass)

Trends In Egg Size (Mass) 0.22 SH NO 0.21 HC 0.20 Mean Egg wt (g) 0.19 0.18 0.17 1998 2000 2002 2004 2006 2008 2010 Return year

r2 > 0.8, regression p<0.001 Slopes equal p=0.177 Nach 2010 6.1 6.0 NO 5.9 ln(Fry body mass) 5.8 NO SH Nach HC 5.7 ANCOVA r2 > 0.8, regression p<0.001 Slopes equal p=0.177 Origin effects P<0.001 5.6 Why should we care about egg sizes anyway? What makes them important and are they biologically important here? Fry body size is tightly correlated with egg size and early growth rate, dominance, and survival are positively correlated with fry body size. Shows significant differences between our wild control, Naches, and our NO up. Yakima population. A locally adapted phenotypic trait. 5.5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 ln(Egg mass)

HC SH NO ln(Fry body mass) ANCOVA Slopes equal p=0.032 6.1 HC 6.0 SH NO 5.9 ln(Fry body mass) 5.8 NO SH Nach HC 5.7 5.6 ANCOVA Slopes equal p=0.032 P<0.001, all r2 >0.8 The rate at which fry body mass increases relative to the increase in egg weight is greatest for HC fry and lowest for NO fry. Again, why should we care? When we combined these relationships with the mean egg mass values for 2010 I pointed out in an earlier slide we can calculate what the expected mean fry body weight will be. 5.5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 ln(Egg mass)

Estimated Fry Mass 2010 Due to combination of larger HC eggs and Greater “Fry body mass/mg egg mass” 400 HC 12.3% larger HC 15.2% larger SH 2.6% larger Fry Body Mass (mg) 300 HC fry are estimated to be 15% heavier that NO fry in 2010. Due to a combination of larger HC egg mass and the greater rate of fry body mass increase per unit egg mass increase. I would guess that in a paired test of dominance the fish with 15% greater body weight would dominate the majority of the time. If that is true, then this is a biologically significant difference between NO, SH and HC fry. 200 HC SH NO Origin

Comparison of PIT vs Non-PIT SARs 1997-2001 broods mean PIT tag loss = 18.4% (17.2% -19.5%) Mean PIT tag induced mortality = 10% What’s happened since then (2002-2006 broods)?

Comparison of PIT vs Non-PIT SARs Brood Year PIT SAR PIT SAR corrected Non-PIT SAR Ratio corrected PIT/Non-PIT SAR 1997 1.50% 1.84% 1.81% 1.015 1998 1.06% 1.30% 1.31% 0.994 1999 0.06% 0.07% 0.11% 0.694 2000 0.40% 0.49% 0.48% 1.027 2001 0.22% 0.27% 0.32% 0.839 2002 0.25% 0.31% 0.28% 1.103 2003 0.08% 0.10% 0.21% 0.456 2004 0.34% 0.42% 0.61% 0.687 2005 0.35% 0.43% 0.79% 0.544 2006 1.16% 1.42% 1.001

Conclusions – Size-at-age Jacks Age 3 SH and HC jacks were significantly smaller than NO fish, but all populations were increasing in size over time at the same rate. SH and HC jacks NS difference in size most years. Unable to use age 3 Naches to compare to UY fish to show and compare trends over time due to low sample sizes.

Conclusions – Size-at-age Age 4 Age 4 Naches and UY NO fish show the same trends over time indicating that naturally spawning SH fish have not impacted NO size-at-age. Age 4 populations did not increase significantly in size over time, but SH and HC fish are now closer in size to NO fish, though still significantly smaller.

Conclusions – Jack Production Trends Hatchery jack production is significantly greater than NO jack production (11% vs 23% before BY2006) Jack production increased significantly in BY2006 in both SH and NO fish Perhaps due to the increasing size-at-age of jacks which is likely driven by marine environmental conditions

Conclusions – Gender Differences When all ages are analyzed, hatchery origin fish have a significantly higher proportion of males than NO fish If just age 4 fish are examined (which make up 70-95% of a cohort on average) there is no significant difference between NO, SH and HC gender proportions More work needed here

Conclusions – Effects of Spawning Jacks The proportion of SH males maturing as jacks was not significantly effected by the proportion of jacks naturally spawning (range 3% to 50%) The proportion of NO males maturing as jacks was not significantly effected by the proportion of NO jacks used as broodstock (range 3% to 29%) Managing the proportion of jacks naturally spawning or in broodstock is not likely to have significant effects on subsequent jack production in UY spring Chinook salmon These results do not necessarily hold for other spring or fall Chinook populations

Conclusions – Other In recent years HC and SH egg mass has been larger than NO egg mass In 2010 HC fry were 15% larger than NO fry due to larger eggs and higher Egg Mass-to-Fry Mass conversion rate Spawn timing of hatchery fish at CESRF was significantly earlier than NO fish, but no population showed significant temporal trends PIT tagged SARs were significantly lower than Non-PIT tagged SARs in 5 of 10 brood years

Acknowledgements Yakama Nation Roza Adult Monitoring Facility Crew Cle Elum Supplementation Research Facility Crew WDFW Ecological Interactions Crew Bonneville Power Administration for providing funding through the Yakima/Klickitat Fishery Project Monitoring and Evaluation Program I’d like to acknowledge the hard work and efforts by the Yakama Nation’s RAMF crew and the CESRF crew. Their cooperation and help was critical in making this work possible. And Patricia Smith (COTR) and BPA for funding support.