Project History

This project #200203100 "Growth Modulation in Salmon Supplementation" derives it's title from the fact that growth is a central driver of life-history decisions in salmonid fishes and highlights the importance of understanding the causes and effects hatchery rearing protocols have on the quality and demographics of the fish they produce.  This project has a significant historical base in the Yakima River basin, but has broad scope and application in hatchery reform throughout the Columbia and Snake River Basins. 

History in the Yakima Basin

The Yakima Supplementation Program rears Yakima River spring Chinook salmon to supplement the natural population occurring in the river at the Cle Elum Supplementation and Research Facility (CESRF) (Fast and Berg 2001).  This extensively monitored program serves as one of the central tests of the efficacy of using "Supplementation" rearing strategies for hatchery production as well as recovery of listed stocks throughout the Columbia River Basin.  Fish rearing in this system was initiated with the collection of broodstock in 1997 (smolts first released in 1999) and continues to the present. We (Larsen, Beckman, and Dickhoff) initiated BPA funded studies (#199202200) examining growth and smolting of wild Yakima River spring Chinook salmon in 1993, prior to any hatchery activity (Larsen et al. 2003; Beckman et al. 2000).  We have monitored both hatchery and wild salmon continuously since the initiation of hatchery releases in accordance with APRE guidelines (NWPCC 2004), again supported by BPA funding.  While intimately associated with the Yakima Supplementation Project, our funding has been separate and independent as our findings have basin-wide significance with regard to hatchery rearing and supplementation programs.   Our expertise in physiology and endocrinology, coupled with our laboratory and experimental fish culture facilities, have provided us with a unique capability to assess fish in multiple hatchery programs within the basin and this ability is not solely limited to the Yakima Supplementation Program.

Initial Major Finding #1: Early detection of high minijack rates

The reproductive androgen, 11-ketotestosterone, is instrumental in initiating the maturation process in teleost fish and can be measured in fish plasma to screen a population prior to release for minijack rates (Larsen et al. 2004).  We compared the physiology and development of naturally rearing wild and hatchery-reared spring Chinook salmon in the Yakima River Basin, and found substantial differences. The most serious difference was an average 41%  (range 20-72%) incidence of early maturation of Yakima Hatchery-reared males over the past 11 years (age-2 minijacks) (Larsen et al. 2004; Larsen et al. 2010a) (Table 1).  This is approximately 9 times our estimate of early male maturation in wild spring Chinook salmon in the Yakima River (Larsen et al. 2010a).  Apparently, hatchery rearing practices promote early male maturation.  Hundreds of thousands of the early maturing hatchery males may residualize in the Yakima and/or Columbia River basins after release and cause potential negative genetic and ecological impacts (Beckman and Larsen 2005).  The ecological concerns include competition for space and food, food depletion and predation on emerging salmonids and other species.  Furthermore, early male maturation translates into as much as a 36% (depending on population and year) reduction in anadromous adult production.  Finally, the release of significant numbers of minijacks from hatchery programs seriously compromises the accuracy of smot-to-adult return estimates (SAR's) as these fish are not in fact smolts being released, but small maturing adult males that rarely return to the spawning grounds (Pearsons et al. 2010).  This major finding has historically served as an anchor point to this project and represents an important long term data monitoring objective for this research.  Starting with this major finding our research progressed through an iterative 3-step process as outlined below:

Step 1)  Study wild fish living in the Yakima River to develop a template for supplementation rearing of spring Chinook salmon (Beckman et al. 2000, previous BPA project # 199202200, Larsen et al. 2003, 2004). 

Step 2   Assess fish reared in the supplementation program according to APRE guidelines  (Document 2004-17, 30 November 2004) as they relate to size, growth rate, body proximate composition and life-history pattern. 

Step 3) Conduct controlled laboratory experiments to test what aspects of hatchery rearing have the greatest effect on hatchery produced fish conforming with APRE guidelines (wild fish template for size, growth rate, life-history composition).

Finding #2:  The wild fish template and APRE guidelines: early male maturation and seasonal growth.

We’ve shown that seasonal growth profiles of wild and hatchery spring Chinook from the Yakima River differ and this supports the contention that high rates of autumnal growth in the hatchery environment may promote high rates of age-2 maturation. (Beckman et al. 2000, Larsen et al. 2004, Larsen et al. 2006; Table 1). Wild Yakima River fish typically show high growth in the summer and reduced growth in the fall as water temperature decreases.  In contrast, despite some variation in absolute size between brood years, Yakima hatchery fish have relatively moderate growth in the summer following later emergence and comparatively high growth in the fall.  These data demonstrated that the Yakima program was outside of APRE guidelines with regard to growth rate and life-history pattern relative to wild Yakima spring Chinook salmon.  Accordingly, these results answered one of the key uncertainties regarding Hatcheries posed by the Draft Columbia River Research Plans (NWPCC, 2005):  Yes, hatchery rearing conditions can alter life-history patterns.  Moreover, these results suggest that lowering autumn growth rates might reduce early male maturation (see Findings #4 and #5 below)

 Finding #3:  Directly compare age-2 male maturation rates of wild and hatchery Yakima River spring Chinook

The Chandler smolt by-pass facility at Prosser Dam, in the lower Yakima River, provides the best venue for selectively collecting both hatchery and wild fish during the spring smolt migration.  Despite conventional thinking that most minijacks remain in headwaters in search of spawning opportunities the following Autumn, this project documented that a significant number of minijacks migrate large distances downstream in a "smolt-like" manner before attempting to return to their natal spawing grounds (Beckman and Larsen 2005).  The physiology of this unique phenotype, which appears maladapted for either smoltification or maturation, was recently described in detail in Larsen et al. (2010b).  Because of the migratory tendency of many of the minijacks in the Yakima Program over the past 7 years this project has documented that precocious male maturation rates of the migrating Yakima hatchery stock are consistently higher than that of the migrating wild Yakima stock (by approximately 9 fold), (Figure 1) and the rate of minijack production from the Yakima CESRF is significantly correlated with the rate of hatchery minijack collection at Prosser Dam (Figure 2) (Larsen et al. 2010a).  These data further confirmed that the program was outside APRE guidelines with regard to life-history composition.

Finding #4:  Laboratory experiment I: seasonal growth and early male maturation 

Null Hypothesis:

H₀₁: Growth rate in the first Summer-Autumn has no effect on physiological development and minijack rates of experimental Yakima hatchery spring Chinook salmon.

Based on numerous studies (Hopkins and Unwin 1997, Silverstein et al. 1997; Shearer et al. 2006) we hypothesized that the high autumn growth rate of the Yakima hatchery fish provided the energetic signal to initiate maturation in a significant proportion of the male fish.  We proposed that by altering growth rate during the critical summer-autumn period, through manipulation of feed ration, to more closely approximate growth of wild Yakima River fish, the incidence of precocious maturation could be reduced to a level more similar to that of the wild stock.  Growth modulation reduced the precocious male maturation rate by 39% among experimental treatments and 21% between CESRF production and experimental fish (Larsen et al. 2006).  However, maturation rates and whole body lipid levels of hatchery fish still differed markedly from that of their wild cohorts suggesting that more dramatic alterations of rearing regime may be required in subsequent experiments to further reduce the prevalence of this phenotype in the hatchery stock.  This experiment demonstrated the importance of the APRE guideline suggesting that growth patterns of hatchery fish match wild fish.  Because hatchery fish grow faster and have higher whole body lipid stores in the autumn than wild fish, the proportion of males maturing at age-2 in the hatchery population is much higher than that found in wild fish.  This experiment resulted in the Yakima Program initiating the first production scale growth modulation experiment aimed at reducing rates of age-2 male maturation (described in Finding #5).

Finding #5:  Production-scale hatchery experiment: seasonal growth modulation

Null Hypothesis:

H₀₁: Growth rate in the Autumn has no effect on physiological development and minijack rates of production Yakima hatchery spring Chinook salmon.                

In response to the “best available data” from our first laboratory based study (Finding #4), the Yakima Project co-managers implemented a three year hatchery production scale growth modulation study at the CESRF aimed at reducing the rate of minijack production (brood years 2002-2004 released from 2004-2006).  Half of the production fish were grown to achieve a mid-October target weight of 15 grams (Conventional Treatment a.k.a. High Growth) and the other half grown to achieve a mid-October target weight of 10 grams (Modified Treatment a.k.a. Low Growth). It should also be noted that since BY 2002, one pair of raceways at the Clark Flat acclimation site each year contain fish that are part of a long term experiment investigating the effects of domestication vs. supplementation on survival and demographics of the Yakima population (see Pearsons et al. 2006).  Two raceways designated Hatchery x Hatchery (HH) contain fish that are part of a “domesticated” population that is uniquely marked such that at return as adults these fish are only crossed with one another.  By contrast, all other raceways at Clark Flat (4 raceways), Easton (6 raceways) and Jack Creek (6 raceways) use untagged and non-adipose fin clipped natural source spawners also referred to as Supplementation Natural (SN) fish for production broodstock.  Several parameters were monitored throughout this investigation to compare physiological responses to the two treatments.

As expected, compared to the Modified (Low) growth treatment, the Conventional (High) growth fish were significantly heavier (p<0.0001), longer (p<0.0001) and had higher condition factors (p<0.0143). Comparisons between High and Low Growth fish for other physiological parameters including the growth and development regulating hormones, thyroxine (T4) and insulin like-growth factor-I (IGF-I), and gill Na+/K+-ATPase activity (an indicator of smolt development) were presented in Larsen et al. (2005).  All of those data indicated that the Low Growth treatment had no significant detrimental effect on smolt associated physiology of those fish.  In three years of testing, the Modified (Low) growth treatment reduced the level of precociously maturing males compared with the Conventional (High) growth treatment by 33% (43 vs. 29%) in BY 2002, 52% (27 vs. 13%) in BY 2003 and 46% (16 vs. 30%) in BY 2004 (Figure 3).  However, survival data compiled by the Yakima Klickitat Fisheries Project (YKFP) indicated that in all three years of this experiment both juvenile and adult (Figure 4) survival  was significantly lower for the Low Growth treatment compared to the High growth treatment suggesting that losses due to small smolt size outweigh losses due to precocious male maturation.  Size of fish at release has been shown to be an important factor influencing SAR’s in both hatchery fish (Bilton 1984; Martin and Wertheimer 1989; Virtanenen et al. 1991; Farmer 1994; Lundqvist et al. 1994) and wild fish (Ward and Slaney 1988; Ward et al. 1989; Henderson and Cass 1991). However, it should be noted that losses due to precocious male maturation might be very significant when compared to that of wild fish (Figure 1).  Furthermore, alterations in life-history composition of the supplemented population compared to the wild stock may increase the potential for undesirable fitness loss through domestication selection in this stock.  

In conclusion, this production scale test of modulating growth in the Yakima Spring Chinook to 1) reduce rates of age-2 male maturation and 2) improve adult survival appeared to have only succeeded with regard to the first factor.  A more appropriate test of the effect of growth modulation would have increased growth in the spring in the Low Growth treatment or a delayed release date to allow for additional growth such that size at release of the two treatments would be equivalent.  Cold water temperatures (2-6 C) at the acclimation sites in the spring prevented attaining significant growth in either group prior to the start of volitional release. Furthermore, after much discussion, the YKFP made the programmatic decision not to delay release in the spring in an attempt to increase growth rate of the Low Growth treatment.  It should be noted that Yakima Spring Chinook salmon are perhaps the healthiest, highest quality fish this research group has examined over the past 20+ years of this field of study and SAR's for this program are comparatively quite high. Thus, it is understandable why the YKFP co-managers would choose this course of action. The long term genotypic effects these high levels of minijack production may have on selection for this population remain to be seen as discussed in the Background section of this proposal and are being monitored and addressed in Objective 3 (Long-term minijack monitoring in Yakima Spring Chinook). In the current proposal, effort will be made in the rearing regime from Dec to the time of release to equalize release sizes through ration modification in order to reconcile this important confounding factor for the Umatilla Fall Chinook program.

Finding #6:  Laboratory experiment II: emergence timing, lipid level and early male maturation 

Null hypotheses:

H₀₁: Emergence timing has no effect on physiological development and rates of age-1 and -2 early male maturation in experimental Yakima hatchery spring Chinook salmon.

H₀₂: Dietary lipid level has no effect on physiological development and rates of age-1 and -2 early male maturation in experimental Yakima hatchery spring Chinook salmon.

The aim of our second laboratory based experiment was to further reduce the rate of age-2 male maturation by exploring the combined effect of altering dietary lipid level and fry emergence timing (Larsen et al. 2005).  As mentioned above (Finding #4) we have found that hatchery (and experimental) fish have higher lipid levels than wild fish (Larsen et al. 2006).  This is important from two perspectives.  First, it demonstrated, once again, that the supplementation program was outside of APRE guidelines for proximate body composition and second, previous work has shown that larger salmon males with higher fat levels tend to mature at an early age at a higher rate (Silverstein and Shimma 1994; Silverstein et al. 1997; 1998; Shearer and Swanson 2000; Shearer et al. 2006).

Results from this experiment provided the following conclusions (Larsen et al. 2005): 1) low growth in their first summer reduced rates of age-2 precocious male maturation in all treatment groups (range 15-19% of males).  2) In contrast with previous studies (see above) the low lipid diet had no significant effect on precocious male maturation rate.  3) Altering fry emergence timing by manipulating egg incubation temperature produced age-1 precocious parr (1-4%) for the first time in an experimental group of Yakima spring Chinook.  It is worth noting that while these rates of age-1 maturation were minimal, they were observed in fish grown to a relatively small size in their first autumn (8-9gms).  Any increase in autumn growth rate beyond the level in this experiment, coupled with the earlier emergence timing, may significantly increase the prevalence of age-1 “precocious parr”.  4) In this study we did not originally plan to produce fish of such a small size since program mangers view a parr less than 10 grams by mid-October as too small to tag.  However, this work emphasized that growth pattern had a more significant effect on determining early male maturation than whole body lipid level (at least in small fish) and demonstrated that rearing fish at a small size could result in smolts with a more “wild-like” life history composition.  5) These data once again demonstrated a conflict between the current design of the supplementation program and the ultimate APRE guideline.  Results from this study lead us to ask the question: How significant is the effect of emergence time and growth rate on both age-1 (precocious parr) and age-2 (minijack) maturation rates in Yakima hatchery spring Chinook?  This was a timely question as the Yakima hatchery program planned to move the emergence/pond date of the production stock to an earlier date to improve early spring growth in their fish for ease of tagging in the Autumn.  A laboratory scale experiment #3 (Finding #7 explored this question).

Finding #7:  Laboratory Experiment III, emergence timing, growth rate and early male maturation

Null Hypotheses:

H₀₁: Emergence timing has no effect on incidence of age-1 and age-2 precocious male maturation rates of experimental Yakima hatchery Spring Chinook salmon.

H₀₂: Growth rate has no effect on incidence of age-1 and age-2 precocious male maturation rates of experimental Yakima hatchery Spring Chinook salmon.

This experiment investigated the interaction between early and later emergence time and high and low growth on rates of age-1 and age-2 precocious male maturation in Yakima Spring Chinook salmon (Larsen et al. 2007).  The primary focus of this experiment was to test the limits of expression for early male maturation in this stock of fish to help make future predictions about the extent of impact manipulation of these factors might have.  Egg incubation temperatures were manipulated to induce emergence in early-January and early February of 2006, and fish from each emergence group were reared under either High (0.5% above maximum feed manufacturers recommended level) or Low (50% of manufacturers level) ration during the first year post-emergence. The four treatments (emergence/ration) in the 2x2 factorial design were Early/High, Early/Low, Late/High, and Late/Low were each reared in triplicate tanks. On the final sampling date of the experiment precocious male maturation rates ranged from 32% precocious parr and 40% minijacks (Total 72%) in the Early/High treatment to 11.4% precocious parr and 22% minijacks (Total 33%) in the Late/Low treatment. Precocious maturation rates were intermediate in the Early/Low (23%PP, 25% MJ, Total 48%) and Late/High treatments (16.6% PP, 31% MJ, Total 47%).  Growth modulation experiment #3 successfully demonstrated the significant capacity for precocious male maturation in Yakima Spring Chinook salmon. While precocious male maturation as age-2 minijacks is relatively common in the CESRF stock, maturation at age-1 has only been documented in two production fish among thousands examined since the establishment of the CESRF program.  However, when emergence time is manipulated through egg incubation temperature, as shown in the current investigation, precocious parr maturation rates (age-1) can be significant, ranging from 11 to 32%of males depending on treatment. When both precocious parr and minijacks rates are combined early male maturation ranged from 35 to 72% of males. Early emergence allows culture facilities like the CESRF to produce larger fish for tagging and larger smolts at the time of release. Our previous production scale growth modulation experiment at the CESFR reduced rates of minijack production at the expense of smolt release size. However, that study (Pearsons et al. 2006) and numerous studies in hatchery and experimental fish (Martin and Wertheimer 1989; Virtanen et al. 1991; Farmer 1994; Lundqvist et al. 1994) and wild fish (Ward and Slaney 1988; Henderson and Cass 1991) have suggested that large smolt size is advantageous for survival. Clearly, results from the current study demonstrate the limits to increasing smolt size excessively.

 Finding #8:  The extent of age-2 precocious male maturation throughout the Columbia and Snake River basins, or "How big is this issue?"

In an effort to determine the extent of minijack production throughout the Columbia and Snake River hatchery system we used information produced from (1) large-scale passive integrated transponder (PIT)-tagging programs at these hatcheries and (2) the installation of PIT tag detectors in the fish passage ladders built into dams on these rivers to investigate the occurrence, behavior, distribution, and origin of minijacks (Beckman and Larsen, 2005).  Minijacks were found in every hatchery release group of more than 7,500 PIT-tagged Chinook salmon smolts reported (2002, 13 groups; 2003, 15 groups).  Detection rates of minijacks at all dams ranged from 0.02% to 0.8% of the total smolts released for different hatchery groups (Figure 5).  While these percentages may seem modest, expansion of this rate, based on the relation between minijack estimates at Cle Elum Hatchery (before release, Larsen et al. 2004) and detections at Bonneville Dam (after release), suggests that minijack abundance in the Columbia and Snake River basins exceeds hundreds of thousands of fish.  External characteristics of maturation (olive pigmentation, deep body shape, and dark fin margins) as well as very large white gonads typical of the later stages of maturation may not be evident in these fish until mid-summer prior to autumn spawning.  These obvious external signs of early maturity are often not seen in the hatchery since most spring Chinook hatcheries release fish in March and April.  This fact may explain why, prior to the use of the new diagnostic tools employed by this project, this issue has historically received only modest attention.  These data show that the Yakima Supplementation Program is not unique.  Many programs appear to produce significant numbers of early maturing males.  We have used the basic research tools and knowledge obtained through this project to collaborate on research starting in 2009 with state and tribal co-managers in the Hood River Basin Spring Chinook salmon supplementation program where very high minijack rates have historically been documented.  Furthermore, we are also collaborating since 2008 with the Upper Columbia River Public Utility Districts (Chelan, Douglas, Grant) exploring the extent of minijack maturation in Summer Chinook stocks as well.  There is evidence of even higher rates among Fall Run Chinook in the Columbia Basin as part of the this current proposal (see Objective 1 and 2 of this proposal).  Clearly, this work has basin-wide significance.

Finding #9:  Assess fish reared in other Columbia basin hatcheries according to APRE guidelines.

In spring 2005 we initiated pre-release minijack surveys at Leavenworth, Entiat, and Winthrop hatcheries in the mid-Columbia River to further confirm our findings from the Yakima program and the PIT-tag database.  In subsequent years these surveys were opportunistically expanded to other spring Chinook populations at Lookingglass Hatchery in NE, Oregon and Summer Chinook populations in the Upper Columbia River (Larsen 2010a) (Table 2).  These surveys have demonstrated: 1) As documented in Beckman and Larsen (2005) minijacks are ubiquitous throughout the Columbia and Snake River basins.  2) Newer supplementation programs focused on localized broodstock that have not been in culture for several generations may be highly susceptible to high minijack rates (i.e. CESRF -41%, Imnaha - 51.7%, Lostine - 52%).  3) However, even stocks that have been in culture for over 20 generations have significant minijack rates (i.e. Leavenworth - 20%).  So, this life-history strategy is genetically retained to some degree even though these fish are not used for broodstock in any hatchery program.  4)  This issue of generations in culture and early male maturation has contributed to the decision to conduct a "Common Garden" rearing experiment described  in Finding #10.

Finding #10:  The effect of hatchery domestication on early male maturation   

Related to the conclusions above in Finding #10, a second significant finding observed in Figure 3 concerns comparisons of minijack rates between the Supplementation Natural (SN) line and the Hatchery Control (HH) line of Yakima Spring Chinook salmon reared at the Clark Flat Acclimation site.  Minijack rates were consistently lower in the HH line under the High or Low Growth treatments used in BY 2002-2004 fish.  Statistical comparisons between the SN and HH lines were not conducted within years in 2002-2004 due to lack of replication with only one High and one Low Growth raceway each (Figure 3).  However, when minijack rates are compared with years (BY 2002-2004) serving as replicates, they were significantly lower in the HH line than the SN line under both the High and Low growth regime (Figure 6).  These data suggested that domestication selection for age of maturation may occur very rapidly; i.e. reduce rates of expression within 1 or 2 generations in culture in traditional segregated hatchery programs that continually recycle the population through the hatchery compared with integrated supplementation programs designed to return hatchery adults to the spawning grounds.  These results lead to the implementation of a "Common Garden" rearing experiment designed to explore the effect of generations in culture on the expression of early male maturation in spring Chinook salmon.

Null Hypothesis: 

H₀₁: Increasing the number of generations of hatchery propagation (range 0 to 2) will have no effect on the incidence of precocious male maturation in Yakima hatchery spring Chinook salmon

The YKFP, Cle Elum Spring Chinook Salmon Supplementation Project is an integrated hatchery program (Mobrand et al. 2005) designed to test whether artificial propagation can be used to increase natural production and harvest opportunities while limiting ecological and genetic impacts.  Potential genetic risks of integrated hatchery programs have received significant attention in recent years (Hard 1995; Ford 2002; Goodman 2005; Berejikian and Ford 2004; Araki et al. 2007).  Demographic differences including reduced body size and earlier spawn timing in hatchery fish compared with wild fish have been documented in Yakima Spring Chinook salmon after a single generation of hatchery culture (Knudsen et al. 2006).  We have been monitoring the incidence of precocious male maturation at age-2 (minijacks) in the Yakima stock since the initiation of the program in 1997 and found rates ranging from 20 to 54% among males prior to release (Larsen et al. 2004, Finding #1).   While comparative estimates for wild fish are difficult to obtain, relative rates for co-migrating wild and hatchery fish captured at Prosser Dam in the spring are approximately 9 fold lower for wild fish compared to hatchery fish (Finding #3), suggesting that hatchery rearing may be influencing age of male maturation in this stock (Larsen et al. 2007).

In male salmonids the physiological decision to mature as a full sized anadromous adult (age-3-5) or a precociously mature "sneaker" parr or minijack (age 1-2) is commonly described as a conditional strategy in which variation in life-history type is under polygenic control and expression of a phenotype depends on exceeding some threshold condition (Roff 1996).  In Atlantic salmon, studies have shown that individuals whose size, growth rate or energetic condition exceeds some genetically determined threshold are more likely to mature precociously (Myers and Hutchings 1986; Aubin-Horth and Dodson 2004; Piche et al. 2008).  Piche et al. (2008) modeled the threshold graphically as a norm of reaction that expresses how the phenotypic value or a trait changes with the environment (and affects growth and energetics).  In this investigation we used this logistic approach to examine whether there is any evidence of evolutionary change in the threshold for minijack maturation as a function of size in three lines of Yakima spring Chinook with varying levels of domestication ranging from 0 to 2 generations in culture. We hypothesize that under traditional hatchery protocols (which do not use minijacks for broodstock production) there may be strong selection acting to reduce the incidence of the minijack phenotype with years in culture.  

The YKFP spring Chinook program includes a domestication monitoring effort that compares several traits of the supplemented natural (SN) population with those of a hatchery control (HH) line founded from first generation hatchery adults starting with BY 2002 (Knudsen et al. 2006).  As part of our monitoring effort we have estimated minijack rates of the SN and HH lines from brood years 2002 to 2007.  Minijack rates were lower in the HH line compared with the SN line in 4 of 6 years, equivalent in one year (BY 2005) and higher in one year (BY 2006).  These data initially suggested that while minijack rates were quite high in progeny of first generation Yakima hatchery fish, domestication selection against the minijack phenotype may be occurring relatively rapidly in the HH line.  The reversal in trend in the BY 2006 fish emphasized the need for a controlled, common garden experiment to clarify this issue.   Salmonid age of maturation is influenced by both environmental and genetic factors and the aim of this experiment was to improve our understanding of the rates at which artificial propagation in supplementation programs may influence this important life-history trait: age of male maturation.                  

Methods

 Fertilized eggs were obtained from three distinct lines of BY 2007 Yakima spring Chinook salmon: the SN line (untagged returning adults), HH line (uniquely tagged line described above), and the Supplementation Hatchery (SH) line (non-HH tagged returning adults).  Now that the Yakima Supplementation Program is rearing its third generation in culture these three lines have experienced varying degrees of hatchery influence and for some lines the exact amount is no longer known.  However, it is known that at the time of this experiment the SN line had either 0 or 1 generations in culture, the SH line had either 1 or 2 generations in culture, and the HH line had 2 generations in culture.  Thus, a potential “hatchery domestication” continuum ranging from 0 to 2 generations in culture among the lines essentially existed.  A mixture of eggs was obtained from each line depending on availability and with an emphasis on maximizing representation across multiple families as follows: SN line-2000 eggs from 80 families (25 eggs/family), SH line-approximately 4000 eggs from 45 potential families constructed from 15 females and 5 males, HH line-3100 eggs from 31 families (100 eggs/family).  The egg lots were transported to the Northwest Fisheries Science Center, Seattle, and were incubated at 5°C until emergence.  Fry were ponded (400/rep.) in quadruplicate 1m diameter circular tanks under an identical moderate growth regime for each line (Total 12 tanks). Growth was regulated by ration manipulation to achieve smolt sizes (April-May) comparable to the High (25 g) growth regimes used for production experiments conducted previously at the Cle Elum Hatchery (Larsen et al. 2006a).  Size was monitored monthly for tracking growth and adjusting rations.  The experiment was terminated after 18 months of rearing.  Minijack rates were determined by visual examination and gonadasomatic index (GSI) of maturing testes in all males in late Spring 2009.  Data were analyzed by ANOVA comparing genetic lines and via logistic regression comparing weight versus minijack proportion using the R statistical package (http://www.r-project.org/).

Growth rates and final weights were nearly identical between the HH, SH, and SN lines throughout the experiment (Figure 7). Minijack rates at the end of the experiment were very high; exceeding 50% of all males in each genetic line.  However, the HH line had a significantly lower proportion of minijacks (58.6%) than either the SH line (70.3%) and SN line (68.3%) (R² = 0.78, p=0.003, Figure 8a).  This observation was consistent with the hypothesis that since minijacks are not used for broodstock in the HH line this life-history may be reduced in the population.  Precocious parr rates were low (<2%) for all lines but were greatest in the SH line (Figure 8b) .

More importantly, logistic regression analysis was conducted comparing the threshold norms of reaction between fish size (weight) and minijack proportion for each sampling date at the end of the experiment in May.  The reaction norm is shifted significantly to the right (higher weight) for the HH line compared with the SN and SH lines (Figure 9). The reaction norms can be used to statistically compare the inflection point at which 50% of all males are maturing as minijacks  (Piche et al. 2008).  Using tanks as replicates, a One-Way ANOVA shows that the HH line is statistically different in the threshold weight needed to achieve 50% maturity (R² = 0.67, p=0.01, Figure 10) being approximately 37 grams for the HH line vs. 31 grams for either the SN or SH lines.  In the HH line breeding occurs in a closed loop where minijacks are not used as sires.  The SN and SH lines are natural origin stock that may have some hatchery influence and natural influence, most notable for the SN line.  Thus, these fish may be exposed to some degree of genetic input from precocious males on the spawning grounds.  Results from this study are consistent with the prediction that increases in the number of generations in hatchery culture can lower minijack rates by shifting the critical size threshold for physiological decision to mature to a larger body size.  These results are relatively striking when one considers that the threshold size for maturity as measured in May of age-2 shows a 19% change in size after only approximately 1-2 generations in culture.

From a hatchery management perspective results from this investigation support the following conclusions:

1) These data provide evidence that the Hatchery control (HH) line at the Cle Elum Hatchery is undergoing domestication selection after 1-2 generations in the size threshold for minijack maturation.

2) With regard to this important demographic trait (age of maturation), these data would suggest that an integrated hatchery strategy (SN line) may help reduce or slow the rate of selection on this and associated traits.

3) Over the long-term, segregated hatchery rearing (HH line) likely results in lower minijack rates at a larger threshold size for initiation of early male maturation.  This observation is supported by evidence from the Leavenworth hatchery on the neighboring Wenatchee River where that spring Chinook stock has been in culture for approximately 20 generations.  The average minijack rate of fish from BY's 2003-2007 is 16% at Leavenworth hatchery and 35% at Yakima Hatchery.  This represents an approximate 50% reduction in the rate of early male maturation in fish that are comparable if not larger in size at release in the spring.

Current Research

Our most current research has 4 areas of focus: 1) Continue to monitor minijack rates in the Yakima Program as part of a long-term Yakima M and E effort.  2) Expanded minijack monitoring of spring Chinook stocks at additional hatchery facilities in the Columbia Basin 3) a laboratory scale growth modulation experiment exploring the post maturation physiology of age-1 precocious males from the Yakima CESRF.  There is evidence to suggest that these fish survive after maturation.  This study is designed to characterize potential for smoltification and/or rematuration of these fish under variable growth regimes.  4) Complete two draft manuscripts for peer reviewed publication related to this study.  In FY 2011 we intend to complete this study and intitiate Objective 1 of the current proposal exploring URB Fall Chinook from the Umatilla Hatchery program.