physiological status of naturally reared juvenile spring … 2010/beckma… · (yearling smolt...

727

Transactions of the American Fisheries Society 129:727–753, 2000American Fisheries Society 2000

Physiological Status of Naturally Reared Juvenile SpringChinook Salmon in the Yakima River: Seasonal Dynamics

and Changes Associated with Smolting

BRIAN R. BECKMAN*Integrative Fish Biology Program,

Northwest Fisheries Science Center, National Marine Fisheries Service,2725 Montlake Boulevard East, Seattle, Washington 98112, USA and

School of Fisheries, University of Washington, Seattle, Washington 98195, USA

DONALD A. LARSEN

Integrative Fish Biology Program,Northwest Fisheries Science Center, National Marine Fisheries Service,

2725 Montlake Boulevard East, Seattle, Washington 98112, USA

CAMERON SHARPE1

Oregon Cooperative Fishery Research Unit, Biological Resources Division,U.S. Geological Survey, Oregon State University, Corvallis, Oregon 97331, USA

BEEDA LEE-PAWLAK


2725 Montlake Boulevard East, Seattle, Washington 98112, USA

CARL B. SCHRECK

Oregon Cooperative Fishery Research Unit, Biological Resources Division,U.S. Geological Survey, Oregon State University, Corvallis, Oregon 97331, USA

WALTON W. DICKHOFF


2725 Montlake Boulevard East, Seattle, Washington 98112, USA andSchool of Fisheries, University of Washington, Seattle, Washington 98195, USA

Abstract.—Two year-classes of juvenile spring chinook salmon Oncorhynchus tshawytscha fromthe Yakima River, Washington, were sampled from July (3–4 months postemergence) through May(yearling smolt out-migration). Physiological characters measured included liver glycogen, bodylipid, gill Na1-K1 ATPase, plasma thyroxine (T4), and plasma insulin-like growth factor-I (IGF-I). Distinct physiological changes were found that corresponded to season. Summer and fall werecharacterized by relatively high body lipid and condition factor. Winter was characterized bydecreases in body lipid, condition factor, and plasma hormones. An increase in condition factorand body lipid was found in February and March. Finally, April and May were characterized bydramatic changes characteristic of smolting, including increased gill Na1-K1 ATPase activity,plasma T4, and IGF-I and decreased condition factor, body lipid, and liver glycogen. These resultscreate a physiological template for juvenile spring chinook salmon in the drainage that providesa baseline for comparison with other years, populations, and life history types. In addition, thisbaseline provides a standard for controlled laboratory experiments and a target for fish culturistswho rear juvenile spring chinook salmon for release from conservation hatcheries. The implicationsof these results for juvenile chinook salmon ecology and life history are discussed.

At an organismal level, one must understand an

* Corresponding author: [email protected] Present address: Washington Department of Fish and

Wildlife, Kalama Research Station, 804 Allen Street,Number 3, Kelso, Washington 98626, USA.

Received June 25, 1999; accepted November 29, 1999

animal’s physiological processes if one seeks tounderstand growth, survival, and reproductive suc-cess (Kitchell 1998). Physiological processes inPacific salmon Oncorhynchus spp. have receiveda great deal of attention (Groot et al. 1995). How-ever, the physiological status of free-living juve-niles has been relatively unexplored; previous

728 BECKMAN ET AL.

work has primarily reported on fish reared in eitherproduction or experimental hatcheries. Cultureconditions are usually different from those foundin streams and rivers with regard to seasonal tem-perature profiles, photoperiod, nutrition, and socialinteractions. These differences might lead to dif-ferences in biological characters, resulting in al-teration of developmental timing, growth rate, sizeat age, and body composition. Thus, it may bemisleading to apply results from cultured fishes tonaturally reared juveniles because the physiolog-ical status of naturally reared fish may not corre-spond to that of hatchery reared fish.

A better understanding of physiological pro-cesses in naturally reared juvenile salmon mayhave two broad implications. First, it will allowbetter assessment of how environmental variationor perturbation affects individual survival and thuspopulation size. Second, knowledge of these phys-iological processes may also give us insight intohow to rear salmon more successfully in artificialculture. Logic suggests that if we want hatcheryfish to match the performance of wild fish, theyshould resemble wild fish, both physically andphysiologically. This similarity is particularly im-portant for supplementation programs for chinooksalmon O. tshawytscha, which are receiving wide-spread support and development in the ColumbiaRiver basin (Sterne 1995; Bugert 1998). Unlikeproduction hatcheries, supplementation projectsare not designed solely to produce fish for harvest;rather, the goal of supplementation programs is toincrease the size of distinct ‘‘wild’’ populations ofsalmon while minimizing effects on either theirgenotypic or phenotypic characteristics. Thus, onemight use a physiological template obtained fromnaturally reared fish to guide the rearing of fish insupplementation facilities.

Growth and physiological status are associatedwith the timing and intensity of smolting. Smoltingis an important developmental transition in anad-romous salmonids that both allows and stimulatesjuvenile fish to undertake a freshwater to oceantransition. Age, size, and season of smolting varieswidely within and among chinook salmon popu-lations (Healey 1991), and this variation has beenattributed to both genetic and environmental fac-tors (Taylor 1990; Clarke et al. 1994; Beckmanand Dickhoff 1998). Examination of smolting innaturally reared fish will allow us to better un-derstand this life history variation. In addition,smolting is a critical process for cultured salmon.If salmon are released from culture facilities assmolts (having initiated the smolting process) they

are more likely to show a rapid directed migrationto the ocean (Zaugg 1981, 1989; Muir et al. 1994).A rapid migration may both increase survival ofjuvenile salmon and decrease interactions betweencultured fish and wild fish. Thus, the process ofsmolting in naturally reared chinook salmon ju-veniles is of particular interest.

The objective of this study was to examine theendocrine and physiological status of naturallyreared spring chinook salmon juveniles in the Yak-ima River, a medium-sized tributary of the Colum-bia River in south-central Washington. YakimaRiver fish belong to a metapopulation of springchinook salmon in the Columbia River that typi-cally smolt as yearling fish in April–May (Myerset al. 1998). Smolting after a year of freshwaterrearing is a common characteristic of ‘‘stream-type’’ chinook salmon of Alaska, British Colum-bia, Washington, and Oregon (Healey 1991). Webegan sampling in July, several months post-emergence, and continued sampling at 2–4 week-intervals through smolting in the following spring.In this manner we attempted to construct a thor-ough seasonal profile of the endocrine and phys-iological status of juvenile chinook salmon, withspecial attention paid to the period preceding andincluding smolting. We report here on 2 years ofstudy.

Methods

Study area.—The Yakima River (total length,349 km; drainage area, 15,900 km2) is in centralWashington State, USA (Figure 1). The headwa-ters drain the eastern escarpment of the CascadeMountain Range (altitude, 2,440 m) and flowsoutheasterly through the Columbia Plateau to itsconfluence with the Columbia River (altitude, 104m). The Yakima River traverses three distinct rivervalleys that are characterized by irrigated agricul-tural areas. Important geographical features in-clude the 40-km-long Yakima River Canyon (riverkilometer, rk, 235 from the confluence with theColumbia River), the confluence with the NachesRiver (drainage area, 2,860 km 2) at the town ofYakima (rk 187), and the confluence with the Co-lumbia River near the town of Richland (rk 0).Flow in the Yakima River system is regulated bya series of irrigation reservoirs. Kacheelus, Ka-chess, and Cle Elum lakes on the upper YakimaRiver supplement flows from March to August,and Bumping and Rimrock lakes on the NachesRiver supplement flows during September and Oc-tober. Temperature and flow data (Figure 2) for theYakima River reported herein were obtained from

729CHINOOK SALMON SMOLT CHANGES

FIGURE 1.—Map of the Yakima River basin. Circled numbers indicate sample sites for chinook salmon juveniles.

the U.S. Bureau of Reclamation Hydromet data-base.

The Yakima River basin contains three closelyrelated, yet distinct, subpopulations of spring chi-nook salmon: American River, Naches River andtributaries (excluding the American River), andYakima River main stem and its tributaries. Al-though populations are distinct, the difference be-tween them is very small, and they cluster togethervery tightly within the overall upper ColumbiaRiver spring chinook salmon metapopulation(Busack and Marshall 1991; Myers et al. 1998).Spring chinook are quite distinct temporally andspatially from fall chinook in the Yakima River,and there is little or no possibility that we includedfall chinook juveniles within our sample collec-tions. Approximately 60% of spring chinook salm-on spawning in the Yakima River basin occursbelow Easton Dam and in the Cle Elum River be-low Cle Elum Lake in the main-stem Yakima Riv-er. The remaining 40% of spawning occurs in theNaches River and its associated tributaries (Fastet al. 1991).

Sampling sites.—Our sampling was constrainedby seasonal changes in abundance and distributionof chinook salmon as described by Fast et al.(1991). In the spring and summer months, the ma-jority of juvenile salmonids in the main-stem Yak-ima River are distributed above Roza Dam (Figure1). In the late summer to early fall, seasonallydeclining temperatures (Figure 2) allow fish to oc-cupy the river below the town of Yakima. Finally,in April of their second year, the majority of ju-veniles out-migrate past Chandler Dam, which isin the town of Prosser. Using these general distri-butional trends, our sampling sites and timing weredivided into two broad areas: above Yakima (June–April) from below Easton Dam (rk 324) to RozaDam (rk 204) and below Yakima (September–May) from the Naches River confluence (km 187)to Zillah (rk 150; Table 1). In addition, migrantsmolts were collected from the bypass facilities atChandler Dam (April–May) in Prosser (rk 75; Fig-ure 1).

Capture methodology.—Electrofishing (SmithRoot model 12-A POW backpack electrofisher,

730 BECKMAN ET AL.

FIGURE 2.—Daily water temperatures measured in the Yakima River from June 1993 through May 1995 (upperpanel). Filled circles show temperatures from Cle Elum (upper river), and open squares show temperatures measuredat Parker (lower river). The lower panel shows mean daily flow in the Yakima River from June 1993 to May 1995,measured at Cle Elum in the upper river (filled circles) and Parker in the lower river (open squares). Asterisksdenote flooding events in which flow exceeded 400 m3/s.

Smith Root, Inc.,2 Vancouver, Washington) wasused to collect fish both above and below Yakima,primarily from around boulders in rip-rapped banksor among woody debris. Beach seining (30-m 3 2-m seine with 1-cm mesh wings and a 0.3-cm meshbag) was conducted primarily below Yakima at sitescharacterized by slack-water eddies below largegravel bars. On certain dates during high-water con-

2 Reference to trade name does not imply endorsementby the National Marine Fisheries Service or the NationalOceanic and Atmospheric Association.

ditions, fish in the river below Yakima were col-lected by electrofishing rather than beach seine (Ta-ble 1). During the fall of 1994, we captured fall-migrating fish in a fyke net set in a side channelnear the town of Zillah. Fish species commonlycaptured along with chinook salmon above Yakimaincluded juvenile rainbow trout O. mykiss and scul-pins Cottus spp. Fish commonly collected belowYakima included redside shiner Richardsonius bal-teatus, dace Rhinichthys spp, chiselmouth Acroch-eilus alutaceus, northern pikeminnow Ptychocheilusoregonensis and peamouth Mylocheilus caurinus.


TABLE 1.—Number of spring chinook salmon sampled at a given site on a given date in the Yakima River, Washington.Underscore indicates fish were measured for length and weight only. See Figure 1 for location of sampling sites. Sitesare as follows: 1 5 Easton Dam, 2 5 Cle Elum, 3 5 Ellensburg, 4 5 Yakima Canyon, 5 5 Union Gap, 6 5 Sunnyside,7 5 Zillah, 8 5 Zillah migrant fish, and 9 5 Chandler migrant fish. Fish were collected by electroshock (*), beachseine (**), fyke net at sample site 8, removal from Chandler bypass holding tank (***), or from separator screens atthe Chandler bypass (site 9).

Date

Site

1 2 3 4 5 6 7 8 9

Brood year 1992

7 Jun 1993 27* 40* 16*26 Jul 199318 Aug 199315 Sep 1993

5 Oct 199312 Oct 199325 Oct 1993

15*14 (16)*15*15*

15*

13*12**11*12*

12*00

8 Nov 199316 Nov 199330 Nov 199316 Dec 199310 Jan 199431 Jan 199422 Feb 1994

8 Mar 199422 Mar 1994

4 Apr 1994

15*

15*15*15*15*15*14*7*

12*

7*12*15*15*7*

15 (49)3

84

2

6 Apr 199413 Apr 199418 Apr 1994

13***1511

Brood year 1993

22 Jun 199412 Jul 1994

4 Aug 199425 Aug 1994

8 Sep 199414 Sep 1994

6 Oct 1994

15*

10*

15*15*15*15*

15*13*

8*

6*

8*

15 (8)**9**

9**

19 Oct 199427 Oct 1994

2 Nov 199415 Nov 199420 Dec 199425 Jan 1995

8 Feb 19951 Mar 1995

15*

6*

15*

15*13*10*10*6*

15*

8*

15 (16)**

15 (26)**15*

15*

11*

00

3

16 Mar 199529 Mar 1995

6 Apr 199520 Apr 199526 Apr 1995

1 May 1995

7*12* 6*

12*10*11**15**

13**15 (27)**15**

15**

151515

Physiological sampling.—On each sampling daywithin a given stretch of river, 6–15 fish were col-lected within 2 h. All tissue samples were obtainedon site within 90 min of capture of the last fish.Fish were individually anesthetized (buffered0.05% tricaine methanesulfonate, MS-222; ArgentChemical Laboratories, Redmond, Washington),measured for fork length (mm) and weight (g), andvisually assessed for smolt development (1 5 parr,2 5 transitional, 3 5 smolt; modified from Gorb-

man et al. 1982). Blood samples were collectedfrom severed caudal vessels (heparinized Natelsontubes; VWR Scientific), centrifuged, and stored ondry ice (as were all tissue samples). Gill tissue wascollected as described by Zaugg (1982). The liverwas removed, weighed (for calculation of hepa-tosomatic index, HSI), and flash frozen. The sexand stage of sexual development was noted, andstomach fullness was graded on a scale from 0(empty) to 7 (distended). Occasional precocious

732 BECKMAN ET AL.

mature male parr were found in the fall, data fromthese fish are not included in the present report.Carcasses were individually bagged and stored.All samples were stored at 2808C until analyzed.

Laboratory analysis.—Plasma thyroxine (T4)values were determined according to the methodof Dickhoff et al. (1982). Plasma insulin likegrowth factor-I (IGF-I) was quantified as describedby Moriyama et al. (1994). Gill Na1-K1 ATPase(hereafter ATPase) activities were measured as de-scribed by Schrock et al. (1994). Liver glycogenwas measured as described by Plisetskaya et al.(1994). Whole body lipid levels were determinedby the method of AOAC (1975), using methylenechloride for extraction.

Data analysis.—Condition factor was calculatedas [weight (g)/length (mm)3] 3 100,000. Poly-nomial regression was performed on the mean val-ue of characters for each sampling date and lo-cation according to Ryan (1997), using StatviewII (Abacus Concepts, Cupertino, California). Alldata were loge transformed before analysis; sig-nificance levels and correlation coefficients are re-ported in figure legends. The polynomial equationcalculated for each character was determined byan iterative process, starting with lower-orderequations and advancing to higher-order equationsuntil a nonsignificant result was obtained (t-test).The regression equation produced for each char-acter was strongly influenced by our sampling dis-tribution and frequency throughout each season.We know that sampling frequencies were not even-ly distributed through time or space and, addi-tionally, varied between years. Thus, we are notoverly concerned with the specific regressionequation. We are more interested in inflectionpoints, which we interpret as a dynamic change instatus for that character. In addition, we did notperform an outlier analysis. Individual points, es-pecially those obtained during the start or end ofthe sampling season, may have leveraged the re-gression equation unduly. Our analysis has not re-moved these influences,so we are cautious aboutinterpretation of inflection points found at the be-ginning or end of a sampling season.

Results

Temperature and Flow

The Yakima River underwent distinct seasonalchanges in temperature and flow, with variationbetween years (Figure 2). Temperature reached17–188C in July and August and steadily declinedthrough the fall, reaching seasonal lows of 08C in

December–February. Temperatures increased ineach year, beginning in March and continuingthroughout July. Flow at Cle Elum was marked byhigh seasonal discharge in June–September, pri-marily due to water released from upstream res-ervoirs (Figure 2). Flow decreased markedly inSeptember. Flow at Parker was stable throughmuch of 1993 and 1994. The winter and spring of1995 were marked by relatively high flow and sev-eral large flooding events.

Size and Morphology

Length and weight were both modeled by third-order polynomials (r2 . 0.70; Figures 3, 4). Sizeincreased from June to October, changed littlefrom October through February, and then in-creased again from March through May. Migrantscaptured at Chandler Dam in April–May (.120mm and .20 g) were considerably larger than oth-er fish captured by electrofishing or seine in thespring (100–120 mm and 10–17.5 g). October–February sizes were similar between years, rang-ing from 90 to 110 mm and from 7.5 to 15 g.

Condition factor showed a complex pattern ofchange that was best modeled with a fourth-orderequation (r2 5 0.45). Condition factors were highin August–September (1.10–1.25) and decreasedto low values in January (1.00–1.10; Figure 5).Condition factors increased in some samples takenin March–April (.1.15). The condition factors ofChandler Dam migrants were uniformly low(,1.05).

Energy Reserves

Whole body lipid showed a seasonal profile sim-ilar to that of condition factor (Figure 6) and wasbest modeled with either a third-order or fourth-order equation (r2 5 0.53, 1993–1994; r2 5 0.64,1994–1995). Values were highest in late summer(August–September; 5–8%) and decreased throughthe winter (January–February), reaching lows of2–3.5%. In both winters, the polynomial regres-sion indicated an inflection in January, with av-erage body lipid levels increasing to almost 4% inFebruary–March at several sites during both years.Average body lipid levels of Chandler Dam mi-grants were always less than 3%.

The stomach fullness data did not fit any re-gression equation well (Figure 7; r2 5 0.30, 1993–1994; not significant, 1994–1995). There was anapproximate seasonal pattern, with highest stom-ach fullness indices found in late summer (Au-gust–September) and spring (March–April). Forwinter of 1994–1995 (15 November through 15


FIGURE 3.—Length of juvenile chinook salmon captured in the Yakima River during June 1993–May 1994 (upperpanel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collected at the following locations:filled circles 5 above the city of Yakima, open squares 5 below the city of Yakima, open triangles 5 from theChandler Dam bypass, and open diamonds 5 from a fyke net near Zillah (1993–1994 only) (P 5 0.0001, r2 50.79, 1993–1994; P 5 0.0001, r2 5 0.73, 1994–1995).

734 BECKMAN ET AL.

FIGURE 4.—Weight of juvenile chinook salmon captured in the Yakima River during June 1993–May 1994 (upperpanel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collected at the locations definedin Figure 3 (P 5 0.0001, r2 5 0.76, 1993–1994; P 5 0.0001, r2 5 0.74, 1994–1995).

February) 71 of 136 individuals sampled (52%)had stomach fullness scores of 2 or less, and only18 individuals had scores of 5 or higher (13%). Incontrast, for spring (16 February through 15

April), 30 of 154 individuals (20%) had scores of2 or less and 48 individuals had scores of 5 orbetter (31%). Generally, there was a great deal ofvariation in stomach fullness in the spring, and


FIGURE 5.—Condition factor of juvenile chinook salmon captured in the Yakima River during June 1993–May1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collected at thelocations defined in Figure 3 (P 5 0.009, r2 5 0.36, 1993–1994; P 5 0.0001, r2 5 0.43, 1994–1995).

stomachs were mostly empty or near empty in thewinter.

Liver glycogen values were quite variable, andthere was no consistent seasonal pattern. It is note-worthy that all samples taken from smolts in thelower river and at Chandler Dam in April–May of

1995 had quite low values (,5 mg/g; Figure 8). Nosignificant regression model could be found for the1993–1994 data, and a binomial model was producedfor 1994–1995, but the fit was poor (r2 5 0.3).

Hepatosomatic indices showed a complex patternof change in each year that was best modeled with

736 BECKMAN ET AL.

FIGURE 6.—Body lipid composition of juvenile chinook salmon captured in the Yakima River during June 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collected atthe locations defined in Figure 3 (P 5 0.001, r2 5 0.53, 1993–1994; P 5 0.0001, r2 5 0.64, 1994–1995).


FIGURE 7.—Stomach fullness (0 5 empty, 7 5 distended) of juvenile chinook salmon captured in the Yakima Riverduring June 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values forfish collected at the locations defined in Figure 3 (P 5 0.008, r2 5 0.30, 1993–1994; not significant, 1994–1995).

a fourth-order polynomial (Figure 9; r2 5 0.40). Highvalues were found in the spring (March–April), andlowest values were found in the winter (December–January) for the 1993–1994 data or in late fall–winter(August–October) for the 1994–1995 data.

Appearance and Smolt PhysiologyBody appearance changed dramatically in the

spring and was best modeled with a third-order orfourth-order equation (Figure 10; r2 5 0.50, 1993–1994; r2 5 0.84, 1994–1995). All migrants taken

738 BECKMAN ET AL.

at Chandler Dam were scored as 3s, meaning theywere strongly silvered and had no parr marks, col-orless peripheral fins, and strong black margins onthe caudal fin. This appearance was in contrast tosamples taken in midwinter when parr marks weredistinctly visible and peripheral fins were brightyellow. In each year, body appearance alsochanged in the fall; some fish collected at that timehad silvery bodies and parr marks were obscured;however, peripheral fins never cleared and no blackfin margins were observed.

Seasonal plasma thyroid hormone profiles weredifferent between years (Figure 11). A four-factorpolynomial was fit to the 1993–1994 data, with apoor fit to the equation (r2 5 0.34); a three-factorpolynomial was fit to the 1994–1995 data, with abetter fit (r2 5 0.54). The main difference betweenthe years lies in the high T4 values found in fishin the spring of 1995 in the lower river and at theChandler Dam bypass compared with those sam-pled in 1994. Furthermore, an additional inflectionpoint was found in the 1993–1994 data in March,which appears to be strongly driven by one of thethree data points generated by sampling fish at theChandler Dam bypass. Finally, for 1994–1995,fish sampled below Yakima in the spring had muchhigher T4 values than those sampled above Yaki-ma.

Plasma IGF-I levels increased more than two-fold from February through May, and mean valueswere best modeled with either a third-order orfourth-order equation (Figure 12; r2 5 0.45, 1993–1994; r2 5 0.54, 1994–1995). The highest valueswere associated with samples taken from migrantsat the Chandler Dam bypass. The model for 1994–1995 displayed an additional inflection point inJuly–August, which was strongly influenced byone data point in June.

The gill ATPase data were fit with either a sec-ond-order or third-order polynomial and displayeda poor to moderate fit (Figure 13; r2 5 0.34, 1993–1994; r2 5 0.57, 1994–1995). Gill ATPase in-creased strongly in the spring of each year, withvalues increasing from less than 4 mmole PO4 ·mg protein21 · h21 to more than 10 mmole PO4 ·mg protein21 · h21 for fish sampled at the ChandlerDam bypass. Values of samples taken from Chan-dler Dam migrants were greatly elevated over oth-er samples in 1993–1994. In 1994–1995, valuesfrom Chandler Dam migrants were similar to thatof samples taken from fish seined in the lowerriver. There appeared to be some fish displayingelevated ATPase values in the summer–fall of

1994–1995; however, these levels did not ap-proach those found in the spring.

Discussion

We conducted a simple descriptive study ex-amining endocrine and physiological characters ofjuvenile spring chinook salmon in the Yakima Riv-er. However, analysis of the data was not simple.These fish displayed distinct seasonal physiolog-ical changes; however, the response of differentphysiological characters varied in direction (pos-itive or negative), magnitude, and temporal pat-tern. Overall, we are most interested in comparingand discriminating similar patterns of change be-tween different physiological characters and ex-amining the relevance of these patterns within anorganismal context. We tried a number of differentmethods of grouping and discriminating betweenfish collected at different places and on differentdates. Based on these analysis, we chose to modelthe data as representative of one population re-acting to rather large seasonal variations in pho-toperiod and temperature. The fit of the data to thepolynomial regression model may be regarded asan assessment of how well our assumption wasmet. Some characters, such as stomach fullnessand liver glycogen, did not show a significant re-lation to season. These characters were probablyresponding to small-scale variation that we couldnot identify. Undoubtedly, there are environmentaland genetic differences between fish rearing in dif-ferent parts of the river; however, because our sam-pling design gave us no power to discriminatethese differences, we will spend little effort in dis-cussing them.

Energy Reserves, Growth, and the Winter–SpringTransition

There was little or no growth from Octoberthrough February, as might be expected based onthe temperature profile and previous work on sal-monids (Figure 2; Elliott 1994; Weatherley and Gill1995). Similarly, Levings and Lauzier (1991) foundlittle change in size of yearling chinook salmonduring winter in the Fraser River. Studies of otherfree-living juvenile salmonids have found seasonalgrowth curves that closely follow seasonal tem-perature changes (Jensen 1990; Elliott 1994; Lo-bon-Cervia and Rincon 1998). Mean size of out-migrant smolts found in this study are similar tothose reported by Major and Mighell (1969) andFast et al. (1991) in the Yakima River. Although itis difficult to compare results among studies of dif-ferent years and with different sampling methods,


FIGURE 8.—Liver glycogen content of juvenile chinook salmon captured in the Yakima River during June 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collected atthe locations defined in Figure 3 (not significant, 1993–1994; P 5 0.0001, r2 5 0.30, 1994–1995)

out-migrating Yakima River smolts appear to belarger than smolts from other interior ColumbiaRiver chinook salmon populations, where smoltsare generally less than 120 mm (Bjornn 1978; Lind-say et al. 1986, 1989; Burck 1993).

The effects of winter conditions or differencesin winter conditions on metabolism, energy re-serves, and mortality of juvenile chinook salmonare relatively unexplored. Several studies of ju-venile salmonids have reported decreases in con-

740 BECKMAN ET AL.

FIGURE 9.—Hepatosomatic index (HSI) of juvenile chinook salmon captured in the Yakima River during June1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collectedat the locations defined in Figure 3 (P 5 0.007, r2 5 0.40, 1993–1994; P 5 0.03, r2 5 0.40, 1994–1995).


FIGURE 10.—Appearance (1 5 parr, 3 5 smolt) of juvenile chinook salmon captured in the Yakima River duringJune 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fishcollected at the locations defined in Figure 3 (P 5 0.0004, r2 5 0.50, 1993–1994; P 5 0.0001, r2 5 0.84, 1994–1995).

dition factor or lipid through the winter (Cunjakand Power 1986, 1987; Berg and Bremset 1998).Rogers et al. (1989) reported extractable lipidlevels (dichloromethane: hexane) of 1.7–6.0%from chinook salmon juveniles captured in theFraser River in December, values comparable tothose found in the Yakima River. Studies of ju-

venile rainbow trout and brook trout Salvelinusfontinalis and Atlantic salmon Salmo salar sug-gest that winter mortality may be greater than50% (Meyer and Griffith 1997; Cunjak et al.1998). Our data suggest that free-living juve-niles lose energetic reserves during winter andimply that severe winters or scarce food supplies

742 BECKMAN ET AL.

FIGURE 11.—Plasma thyroxine concentration of juvenile chinook salmon captured in the Yakima River duringJune 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fishcollected at the locations defined in Figure 3 (P 5 0.01, r2 5 0.34, 1993–1994; P 5 0.0001, r2 5 0.54, 1994–1995).


FIGURE 12.—Plasma insulin-like growth factor-I (IGF-I) concentration of juvenile chinook salmon captured inthe Yakima River during June 1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbolsrepresent values for fish collected at the locations defined in Figure 3 (P 5 0.001, r2 5 0.45, 1993–1994; P 50.0001, r2 5 0.54, 1994–1995).

during the previous summer–fall could affectoverwinter survival.

We found low stomach fullness scores in thewinter, including a rather substantial proportion of

empty stomachs. Because stomach clearance timesslow down at low temperature, it is not unreason-able to suggest that these fish are not feeding muchin the winter, supporting our supposition that Yak-

744 BECKMAN ET AL.

FIGURE 13.—Gill Na1-K1 ATPase activity of juvenile chinook salmon captured in the Yakima River during June1993–May 1994 (upper panel) and June 1994–May 1995 (lower panel). Symbols represent values for fish collectedat the locations defined in Figure 3 (P 5 0.004, r2 5 0.34, 1993–1994; P 5 0.0001, r2 5 0.57, 1994–1995).


ima River chinook salmon juveniles did not growmuch and experienced an energetic deficit duringthe winter. During other seasons, stomach fullnesswas highly variable, which may be due to dailysampling time; several studies have noted feedingactivity of juvenile salmonids is variable throughthe day (Johnson and Johnson 1981; Sagar andGlova 1988). The logistics of field collections atseveral sites eliminated the possibility of samplingat a set time of day for fish at all locations. Inaddition, there were almost certainly differencesin the quantity and quality of food available atdifferent sites due to longitudinal differences inphysical environment, prey availability, and po-tential competitors in the Yakima River (Fend andCarter 1995; Leland 1995). Overall, there appearto be broad seasonal changes in the feeding ofjuvenile chinook salmon in the Yakima River, butour data are neither accurate enough nor preciseenough to support further conclusions.

It is apparent from the shape of the regressionmodels that there is little concordance betweenmost of characters associated with feeding, me-tabolism, and energy storage (gut fullness, liverglycogen, HSI, body lipid, and condition factor).Given differences in daily sampling time andstomach fullness and based on other studies, it isnot unexpected that we should find large variationin liver glycogen and HSI values (Boujard andLeatherland 1992). Liver glycogen values foundin this study were variable, and no seasonal trendwas apparent. However, the very low values foundin migrating smolts at Chandler Dam are notable.A seasonal pattern in HSI was apparent, with peaksin the fall and spring. Hepatosomatic indices arevaluable indicators of energetic status in Atlanticcod Gadus morhua (Couture et al. 1998), which,in contrast to salmonids, store much of their lipidreserves in their livers (Zhou et al. 1996). Thus,HSI may not provide a good signal of energeticstatus for juvenile salmon.

One of the most striking, yet unexpected, find-ings from this study was an increase in anaboliccharacters in the spring (February–April). The cat-abolic response of juvenile chinook salmon in thewinter was expected (the result of low tempera-tures and little food); in turn, an anabolic increasewas expected in the late spring following increasesin temperature and food supply. Surprisingly, bodylipid, condition factor, and size increased in Feb-ruary–March, when water temperatures were stilllow (58C) and when we anticipated that food sup-plies would be limited. However, stomach fullnessappeared to increase in February–March, sug-

gesting that increased condition factor and bodylipid values were reflective of changes in feedingactivity and energetic status. In addition, increasesin body size suggest that growth had occurred.Finally, it is unlikely that the increase found inplasma IGF-I (early April) could occur in fish thatwere not nutritionally replete (Duan et al. 1995;Perez-Sanchez et al. 1995; Beckman et al. 1998).

There are few other data on the anabolic statusof free-living juvenile salmonids during the spring.Cunjak and Power (1986) found an increase inlipid between March and May in juvenile Ontariobrook trout. Berg and Bremset (1998) found anincrease in body lipid from April to June in ju-venile Atlantic salmon and brown trout Salmo trut-ta in Britain. In addition, Tveiten et al. (1996)found increases in appetite and growth rate of year-ling Arctic char Salvelinus alpinus in April andMay, even though the fish were held at a constant48C. Koskela et al. (1997) found that brown troutwill feed at 28C and that growth was significantlyincreased at 48C and 68C, temperatures similar tothat in the Yakima River in spring. Finally, Fors-berg (1995) reported on seasonal growth of At-lantic salmon held in land-based culture facilitiesin Norway. He found significant increases ingrowth during February, March, or April at dif-ferent facilities and suggests that ‘‘growth wasmore closely linked to seasonal changes in pho-toperiod than to water temperature.’’ Together,these studies suggest that salmonids may initiateanabolic processes at low temperatures in thespring in response to changing photoperiod.

Growth hormone (GH) has been shown to bephotoperiod responsive (Bjornsson et al. 1989,1995; McCormick et al. 1995), and increases inGH have been found as early as March in fish heldunder natural photoperiods. In addition, an in-crease in plasma insulin has been found in cohosalmon Oncorhynchus kisutch as early as February(Plisetskaya et al. 1988) and in Atlantic salmonsmolts in March (Mayer et al. 1994). Increases ineither GH or insulin have been suggested to pro-mote growth (Mommsen 1998). Thus hormonescapable of inducing growth increase in the springin reared fish in captivity and may also occur infree-living fish.

We do not know whether an increase in energyresources in the spring is a situation unique to theYakima River in the 2 years we sampled or whetherit is a common feature in free-living chinook salm-on juveniles. There are several other possibilitiesthat might explain our findings. There may havebeen changes in fish behavior or habitat preference

746 BECKMAN ET AL.

associated with the change in season that resultedin fish with different physiological status becom-ing more available for sampling in February–April. We have no data with which to dispute thissupposition. However, it seems unlikely that therewas an area in the river in which fish did not ex-perience ‘‘winter-like’’ conditions; thus, we wouldexpect decreased lipid and condition factor in allfish. In addition, increases in body lipid and con-dition factor appear before the increase in size inspring. If fish in ‘‘better condition’’ were becomingmore accessible to our capture methods in March,we might expect that they would be larger.

Perhaps combining data from fish rearing in up-per- and lower-river areas drives the dynamicfound in our data. In most cases, values were dis-tributed both above and below the regression linein February–March, suggesting that samples fromboth locations contributed to increases in conditionfactor and lipid in March. It is apparent that somefish that are not necessarily geographically seg-regated had higher lipid levels and condition fac-tors than fish found in midwinter. Finally, the lipiddepletion rate found from October through Feb-ruary could not be sustained through April as bodylipid composition would approach zero, a physi-ological impossibility. Similarly, the decrease incondition factor levels off in February and March,regardless of a few high values that may cause theregression model to show an actual increase. Wethus discount the possibility that sampling meth-ods or local geographic differences in fish sampledfrom the river generated the inflection in the poly-nomial regression found in February and March.The data support an increase in condition factorand lipid, suggesting that juvenile salmon in theYakima River shift from a catabolic to an anabolicstate in the early spring.

Spring Smolting

This is the first report of plasma IGF-I valuesfrom free-living juvenile salmonids. We found alarge increase in IGF-I in the spring associatedwith smolting and downstream migration. Valuesincreased 200–300% from resident parr in Feb-ruary to migrating smolts in April–May. Thesechanges are much larger than have been previouslydocumented in relation to smoltification in chi-nook salmon reared in captivity, either in hatcheryor laboratory situations (Beckman and Dickhoff1998; Beckman et al. 1998, 1999; Silverstein etal. 1998). Komourdjian et al. (1976) and Dickhoffet al. (1997) discussed the possibility that GH orthe GH–IGF-I endocrine axis plays a key role in

stimulating and coordinating the process of smolt-ing. McCormick and Bjornsson (1994) reportedhigh plasma GH levels in free-living, migrant At-lantic salmon smolts, and several authors have de-scribed seasonal increases in GH associated withsmoltification (Bjornsson et al. 1989, 1995; Mc-Cormick et al. 1995). The IGF-I levels measuredin this study, coupled with the GH levels measuredby McCormick and Bjornsson (1994), suggest thatthe GH–IGF-I endocrine axis is highly stimulatedin migrating smolts and reinforce the notion thatthis axis and growth are important elements ofsmolting (Dickhoff et al. 1997; Beckman et al.1998).

The spring changes in plasma T4, IGF-I, and gillATPase were relatively small in fish taken aboveYakima compared with migrating smolts at Chan-dler Dam. Conversely, condition factor, liver gly-cogen, and body lipid values were much lower infish sampled at Chandler Dam than in fish sampledabove Yakima. We attribute these changes to thesmolting process; however, we did not attempt todiscriminate differences between parr and smolts.Numerous studies have demonstrated that parr andsmolts differ in almost any physiological characterone examines. Instead, our data, grouping parr andsmolt, demonstrates that smolting in naturallyrearing fish is a temporal process, requiring weeksto accomplish (at least in a populational view). Inaddition, this study, along with several others(Zaugg et al. 1985; Muir et al. 1994; Schrock etal. 1994; Haner et al. 1995), shows that smoltingis a spatial process, with physiological changesintensifying as fish move downriver. We have thusattempted to define the temporal and spatial tra-jectory of physiological changes accompanyingsmolting. Our analysis suggests that endocrine andphysiological changes occur in concert with mi-gration in naturally reared salmon, rather than pre-ceding migration. This conjunction of behavioraland physiological change found in naturally rear-ing and migrating fish may make comparisons withfish held in hatchery raceways or laboratory tankstenuous.

Energy Reserves and Smolt Migration

Our observations of migrating smolts ended atthe Chandler Dam bypass, 75 km from the con-fluence of the Yakima River with the ColumbiaRiver and a further 540 km from the Pacific Ocean.Smolts captured at the Chandler Dam bypass dis-played depleted liver glycogen and body lipidstores after they had traveled no further than 250km (Easton to Chandler Dam), which was at most


30% of their journey. It is difficult to understandhow such a short migration could result in severedepletion of energy stores and how the fish couldsupport continued metabolic demand through theremainder of their journey. Vijayan et al. (1993)found only moderate decreases in liver glycogenfor juvenile coho salmon that were continuouslyswam and fasted for 3 weeks. The nearly unmea-sureable liver glycogen levels found in migrantsmolts at the Chandler Dam bypass suggest thatexercise alone may not explain these reduced gly-cogen levels.

A number of reports have documented increasedmetabolic rates in smolting salmon (for review seeHoar 1988). This increased metabolic rate mightbe directly associated with an increase in circu-lating hormones with catabolic properties. Severalexperiments have shown that GH treatments in sal-monids results in reduced lipid levels or conditionfactor (Danzmann et al. 1990; O’Connor et al.1993). In addition, thyroid hormone and cortisolare lipolytic in salmonids (Sheridan 1986). Thecombined effect of high plasma GH, cortisol, andthyroid hormone is inevitably lipid depletion, re-gardless of the activity level of the fish. The highplasma levels of lipolytic hormones measured inmigrating smolts (McCormick and Bjornsson1994), coupled with the quite low lipid and gly-cogen values found in migrating Yakima Riversmolts, suggests that the lipolytic status of smoltshas been underestimated in many laboratory ex-periments. These observations suggest we have notyet determined the causal mechanisms (endocrineor exercise) leading to the high metabolic rate andcatabolic status of migrating smolts, leaving it dif-ficult to assess the physiological impacts of en-vironmental alterations on these animals.

Fall Migration

An autumnal migration from headwater areas tolower river reaches or main-stem areas appears tobe characteristic of many interior populations ofColumbia River and Fraser River chinook salmon(Bjornn 1971, 1978; Lindsay et al. 1986, 1989;Murray and Rosenau 1989; Levings and Lauzier1991; Burck 1993). In addition, several studiesdemonstrated that juvenile chinook salmon maysmolt in the fall (Ewing et al. 1980; Beckman andDickhoff 1998). Our data suggest that appearance,T4, IGF-I, and gill Na1-K1 ATPase all show somesigns of increased values in September–October,either by an inflection point in the regression mod-el or by increased variation in data around theregression line. However, these data are difficult

to evaluate; it is not clear whether part of the pop-ulation is smolting or whether there is some sea-sonal physiological change which may be asso-ciated with fall redistribution movements.

The majority of Yakima juveniles are not smolt-ing and emigrating to the ocean in the fall. Thelarge out-migration of yearling smolts the follow-ing spring clearly argues against this (Fast et al.1991). In addition, no adult spring chinook salmonfrom the Yakima River (.1,000 samples) werefound to possess ocean-type scale characteristicsfound in fish that migrate to the ocean as subyear-lings (Knudsen 1991). These data provide little orno evidence for fall smolting in naturally rearingYakima chinook salmon; thus, we cannot ascribefall changes in physiology to smolting.

Fall redistributions have been described as asearch for appropriate overwintering habitat(Bjornn 1971; Hillman et al. 1987), though this be-havior may not be universal within a population. Inthe Yakima River, some fish rear year-round at orabove Cle Elum, while others move several hundredkilometers downriver in the fall (Fast et al. 1991).Migratory behavior may be genetically fixed; Brad-ford and Taylor (1997) showed that individual chi-nook salmon fry display either migratory or non-migratory behavior in the spring following emer-gence and, further, that individuals from differentpopulations possessed different tendencies to mi-grate. However, fall migration might also be a con-ditional strategy, dependent on habitat or the fishes’physiological status. We did not find dramatic phys-iological differences between fish sampled aboveYakima and below Yakima (which had presumablyrecently migrated downstream) in the fall. Fish cap-tured in the lower river in the fall did tend to belarger and have greater lipid deposits than fish col-lected in the upper river, suggesting that fall mi-grants were not energetically stressed. Thus, whileit appears to be a relatively common trait in juvenilespring chinook salmon, a physiological relation tofall in-river migration is not apparent.

Annual Variation

The winter of 1993–1994 was relatively warm,with relatively low flow. In contrast, the winter of1994–1995 was colder and punctuated with sev-eral major floods. We did not conduct a formal testfor differences in physiology between years as ourdata provided little power to determine year-to-year variation because of differences in samplingmethod, date, location, and the small number offish per sample. Holtby et al. (1989) found dif-ferences in median date of smolt migration for

748 BECKMAN ET AL.

→

FIGURE 14.—The seasonal anabolic–catabolic status of juvenile chinook salmon rearing in the Yakima River1993–1994 (upper panel) and 1994–1995 (lower panel).

coho salmon in Carnation Creek (Vancouver Is-land, British Columbia) of up to 21 d over 17 yearsof study. This variation was highly related to an-nual differences in temperature. Similarly, Fast etal. (1991) found differences of up to 18 d in dateof 50% passage at Chandler Dam for chinooksalmon in the Yakima River (1983–1990). Giventhe large role that temperature plays in the phys-iology of poikilotherms, we might expect physi-ological differences associated with annual vari-ation in temperature. However, to discern differ-ences would require multiyear sampling and largenumbers of fish sampled at a common physiolog-ical points (midwinter, smolt out-migration).

There was some variation in absolute levels ofgill ATPase, T4, and IGF-I measured betweenyears. However, the same trend was evident in eachyear: values were low in the winter and increasedin the spring. The slope of the ATPase curve ap-pears shallower in the spring of the 1994–1995rearing period compared with 1993–1994 period,which is probably due to the greater number ofsamples obtained in the lower river in the secondyear. The shape of the T4 curve in 1993–1994 ap-pears to be radically different than in 1994–1995.The apparent decrease in 1993–1994, found inApril, is due to the leverage produced by one sam-ple with a low mean value. As stated previously,we attach little physiological significance to in-flection points found at the beginning or end orthe sampling profile. The important physiologicalpoint is that values increased from winter to springin each year. The IGF-I profiles are essentiallyidentical between years; however, the absolute val-ues measured in 1994–1995 are roughly half thosefound in 1993–1994. We cannot at this time ascribeparticular significance to differences in absolutevalues of blood hormone levels. As was shown forthyroid hormones in relation to smolting, the pres-ence of a seasonal change is more important thanthe absolute level measured (Dickhoff et al. 1982).

Growth and Adiposity of Naturally RearingVersus Hatchery Chinook Salmon

The seasonal growth patterns of naturally rearingfish may be different from that of cultured fish. Aprevious study (Beckman et al. 1999) of juvenilechinook salmon from Columbia River hatcheriesusing various water sources showed three different

seasonal growth patterns: little growth from No-vember through April (release); continued growthfrom November through February, with an in-creased growth rate from February to April; andlittle growth through the winter, then increasedgrowth February through April. In that study, thespring growth rate was positively related to smoltdevelopment and subsequent smolt-to-adult surviv-al. Given the generally better smolt-to-adult returnof wild chinook salmon as compared with hatcherysmolts (Lindsay et al. 1989; Fast et al. 1991), thesignificant early spring growth found in wild fishreinforces the hypothesis that spring growth ratemay be important for smolt development and per-formance. However, it is not yet clear whether ahigh growth rate promotes smolting. Wild fish gofrom no growth in the winter to more rapid growthin the spring. It may be the dynamic change ingrowth rather than the rate of spring growth that isimportant to wild fish performance.

Lipid levels in naturally reared Yakima Riverfish were strikingly lower than in Yakima Riverfish reared in an experimental hatchery and fedcommercial feed. The cultured fish had lipid levelsranging from 11% to 8% in January through March(Beckman et al. 1998), levels which are commonlyfound in fish fed commercial feeds (Shearer et al.1997). The high energy density of artificial feedsand high feeding level in culture promote highadiposity. The significance of the differences inbody lipid level between cultured and free-livingfish has not been established; thus, what consti-tutes an obese fish in not known. However, thebiomedical field has clearly shown the hazards ofobesity in humans (Wickelgren 1998), and thissuggests that investigation into the effects of highadiposity on the physiology, health, and devel-opment of captive salmonids may be of interest.

Integrated Seasonal Pattern

On a seasonal basis, juvenile chinook salmon inthe Yakima River appear to go through four dis-tinct physiological states: an anabolic phase insummer–fall, in which fish increase in size andstore energy as lipid; a catabolic phase that extendsthrough the winter, in which fish cease growth anddeplete energy reserves; a second anabolic phasein February–March, where fish once again beginto grow and replenish energy reserves; and, finally,

750 BECKMAN ET AL.

a period (March–April) that exhibits both anabolicand catabolic elements as body size continues toincrease, yet energy reserves are depleted. Forcontinued reference within this paper, we will referto this final state as the smolt-associated metabolicstate. These phases are illustrated in Figure 14 byoverlaying the regression model for condition fac-tor, lipid, weight, and IGF-I for each year-classexamined. In both years the same overall patternwas seen.

The smolt-associated metabolic state is clearlya product of the developmental process of smolt-ing. As seen in the present investigation, chinooksalmon smolts typically deplete liver glycogen andlipid and lose condition factor (Hoar 1988). Dueto limited plasma volume, we did not assess GHlevel, but increased plasma GH is characteristic ofsmolting under natural photoperiods (Bjornsson etal. 1989, 1995; McCormick et al. 1995). In thisstudy, we found that smolting chinook salmon ju-veniles have high plasma IGF-I levels at the sametime they exhibit depleted energy stores. In a lab-oratory study, we also found that chinook salmonsmolts in the spring have depleted glycogen storeswhile maintaining high plasma IGF-I levels (Beck-man et al. 1998). It is unusual for animals to dis-play both elevated GH and IGF-I. Elevated GHlevels are characteristic of fasting animals (Far-bridge and Leatherland 1992), and the high GHlevels act to deplete lipid and glycogen while IGF-I levels are low (Duan and Plisetskaya 1993; Duanet al. 1995). In rapidly growing animals, GH levelsare generally low, IGF-I levels are high, and lipidand glycogen levels are maintained or increase(Storebakken et al. 1991; Perez-Sanchez et al.1995). Thus, we may ascribe the smolt-associatedmetabolic state to the combination of high plasmaGH and IGF-I, with GH acting to deplete glycogenand lipid while IGF-I supports continued muscleand bone growth.

Smolting has often been viewed as a number ofloosely organized endocrine, physiological, andbehavioral events that may not be directly inter-related. However, such a view may have evolvedfrom an exclusive focus on laboratory or hatcheryexperiments in which fish may not have receivedcoordinated seasonal signals. In natural popula-tions, spring-smolting salmonids respond to sev-eral seasonally coupled environmental signals (in-creasing photoperiod, temperature, food supply).Presumably, this response has been refined throughselection to produce an adaptive process that al-lows juvenile salmonids to effectively make thefreshwater to ocean transition typified by their life

history. While acknowledging that smolting en-compasses a broad number of physiological andbehavioral changes, it is also evident that springsmolting occurs within a discrete seasonal period,suggesting that in naturally rearing salmon, thevarious components of smolting are temporallylinked. This study’s uniqueness lies in that it com-bined winter and spring sampling in a continuousseries allowing us to observe (a) the strongly cat-abolic state of wild chinook salmon in the winter,(b) the anabolic increase found in February–Marchand (c) the subsequent strong depletion of energyreserves found during smoltification that appearsto be accompanied by continued bone and musclegrowth (smolt-associated metabolic state). Theseobservations suggest that the smolting processmay be an integrated product of a seasonal seriesof environmental stimuli and endocrine–physio-logical responses. Continued investigation into therelation of winter–spring feeding conditions, met-abolic status, and growth may yield important re-sults with regard to the control of smolting.

Acknowledgments

This study was supported by the National Ma-rine Fisheries Service, BPA contract 92-022, andUSDA grant 94-37206-1096. The Oregon Coop-erative Research Unit is supported jointly by theBiological Resources Division of the U.S, Geo-logical Survey, Oregon State University, and theOregon Department of Fish and Wildlife. We thankGeoff McMichael and Todd Pearsons for sharinglocal knowledge on the habits of wild spring chi-nook salmon, Karl Sheares for lipid analysis, andKim Larsen for the loan of vital sampling equip-ment.

References

AOAC (Association of Official Analytical Chemists).1975. Official methods of analysis, 12th edition.AOAC, Washington, D.C.

Beckman, B. R., and W. W. Dickhoff. 1998. Plasticityof smolting in spring chinook salmon: relation togrowth and insulin-like growth factor-I. Journal ofFish Biology 53:808–826.

Beckman, B. R., D. A. Larsen, B. Lee-Pawlak, S. Mo-riyama, and W. W. Dickhoff. 1998. Insulin-likegrowth factor-I and environmental modulation ofgrowth during smoltification of spring chinooksalmon, (Oncorhynchus tshawytscha). General andComparative Endocrinology 109:325–335.

Beckman, B. R., and ten coauthors. 1999. Growth, smol-tification, and smolt-to-adult return of spring chi-nook salmon (Oncorhynchus tshawytscha) fromhatcheries on the Deschutes River, Oregon. Trans-


actions of the American Fisheries Society 128:1125–1150.

Berg, O. K., and G. Bremset. 1998. Seasonal changesin the body composition of young riverine Atlanticsalmon and brown trout. Journal of Fish Biology52:1272–1288.

Bjornn, T. C. 1971. Trout and salmon movements in twoIdaho streams as related to temperature, food,stream flow, cover and population density. Trans-actions of the American Fisheries Society 100:423–438.

Bjornn, T. C. 1978. Survival, production, and yield oftrout and chinook salmon in the Lemhi River, Idaho.University of Idaho, College of Forestry, Wildlifeand Range Sciences, Bulletin 27, Moscow.

Bjornsson, B. T., S. O. Stefannson, and T. Hansen. 1995.Photoperiod regulation of plasma growth hormonelevels during parr-smolt transformation of Atlanticsalmon: implications for hypoosmoregulatory abil-ity and growth. General and Comparative Endocri-nology 100:73–82.

Bjornsson, B. T., H. Thorarensen, T. Hirano, T. Ogasa-wara, and J. B. Kristinsson. 1989. Photoperiod andtemperature affect plasma growth hormone levels,growth, condition factor and hypoosmoregulatoryability of juvenile Atlantic salmon (Salmo salar)during parr-smolt transformation. Aquaculture 82:77–91.

Boujard, T., and J. F. Leatherland. 1992. Circadian pat-tern of hepatosomatic index, liver glycogen and lip-id content, plasma non-esterified fatty acid, glucose,T3, T4, growth hormone and cortisol concentrationsin Oncorhynchus mykiss held under different pho-toperiod regimes and fed using demand-feeders.Fish Physiology and Biochemistry 10:111–122.

Bradford, M. J., and G. C. Taylor. 1997. Individual var-iation in dispersal behaviour of newly emerged chi-nook salmon (Oncorhynchus tshawytscha) from theupper Fraser River, British Columbia. CanadianJournal of Fisheries and Aquatic Sciences 54:1585–1592.

Bugert, R. M. 1998. Mechanics of supplementation inthe Columbia River. Fisheries 23(1):11–20.

Burck, W. A. 1993. Life history of spring chinook salm-on in Lookingglass Creek, Oregon. Oregon De-partment of Fish and Wildlife, Information Report94-1, Corvallis.

Busack, C., and A. Marshall. 1991. Genetic analysis ofYFP chinook salmon stocks. Pages 2–45 in C. Bus-ack, C. Knudsen, A. Marshall, S. Phelps, and D.Seiler. Yakima Hatchery experimental design. Bon-neville Power Administration, Report DOE/BP-00102, Portland, Oregon.

Clarke, W. C., R. E. Withler, and J. E. Shelbourne. 1994.Inheritance of smolting phenotypes in backcrossesof hybrid stream-type 3 ocean-type chinook salmon(Oncorhynchus tshawytscha). Estuaries 17:13–25.

Couture, P., J-D. Dutil, and H. Guderley. 1998. Bio-chemical correlates of growth and condition in ju-venile Atlantic cod (Gadus morhua) from New-foundland. Canadian Journal of Fisheries andAquatic Sciences 55:1591–1598.

Cunjak, R. A., and G. Power. 1986. Seasonal changesin the physiology of brook trout, Salvelinus fontin-alis (Mitchell), in a sub-Arctic river system. Journalof Fish Biology 29:279–288.

Cunjak, R. A., and G. Power. 1987. The feeding andenergetics of stream-resident tout in winter. Journalof Fish Biology 31:493–511.

Cunjak, R. A., T. D. Prowse, and D. L. Parrish. 1998.Atlantic salmon (Salmo salar) in winter: ‘‘the seasonof parr discontent’’? Canadian Journal of Fisheriesand Aquatic Sciences 55(supplement 1):161–180.

Danzmann, R. G., G. J. Van Der Kraak, T. T. Chen, andD. A. Powers. 1990. Metabolic effects of bovinegrowth hormone and genetically engineered rain-bow trout growth hormone in rainbow trout (On-corhynchus mykiss) reared at a high temperature.Canadian Journal of Fisheries and Aquatic Sciences47:1292–1310.

Dickhoff, W. W., B. R. Beckman, D. A. Larsen, C. Duan,and S. Moriyama. 1997. The role of growth in en-docrine regulation of salmon smoltification. FishPhysiology and Biochemistry 17:231–236.

Dickhoff, W. W., L. C. Folmar, J. L. Mighell, and C. V.W. Mahnken. 1982. Plasma thyroid hormones dur-ing smoltification of yearling and under-yearlingcoho salmon and yearling chinook salmon and steel-head trout. Aquaculture 28:39–48.

Duan, C., and E. M. Plisetskaya. 1993. Nutritional reg-ulation of insulin-growth factor-I mRNA expressionin salmon tissues. Journal of Endocrinology 139:243–252.

Duan, C., E. M. Plisetskaya, and W. W. Dickhoff. 1995.Expression of insulin-like growth factor I in nor-mally and abnormally developing coho salmon (On-corhynchus kisutch). Endocrinology 136:446–452.

Elliott, J. M. 1994. Quantitative ecology and the browntrout. Oxford University Press, New York.

Ewing, R. D., and five coauthors. 1980. Influence ofsize, growth rate, and photoperiod on cyclic changesin gill (Na1 K1) -ATPase activity in chinook salm-on (Oncorhynchus tshawytscha). Canadian Journalof Fisheries and Aquatic Sciences 37:600–605.

Farbridge, K. J., and J. F. Leatherland. 1992. Temporalchanges in plasma thyroid hormone, growth hor-mone and free fatty acid concentrations, and he-patic. 5-monodeiodinase activity, lipid, and proteincontent during chronic fasting, and re-feeding inrainbow trout (Oncorhynchus mykiss). Fish Physi-ology and Biochemistry 10:245–257.

Fast, D., J. Hubble, M. Kohn, and B. Watson. 1991.Yakima River spring chinook enhancement study.Bonneville Power Administration, Report DOE/BP-39461-9, Portland, Oregon.

Fend, S. V., and J. L. Carter. 1995. The relationship ofhabitat characteristics to the distribution of Chiron-omidae (Diptera) as measured by pupal exuviae col-lections in a large river system. Journal of Fresh-water Biology 10:343–359.

Forsberg, O. I. 1995. Empirical investigations on growthof postsmolt Atlantic salmon (Salmo salar L.) inland-based farms. Evidence of a photoperiodic in-fluence. Aquaculture 133:235–248.

752 BECKMAN ET AL.

Gorbman, A., W. W. Dickhoff, J. L. Mighell, E. F. Pren-tice, and F. W. Waknitz. 1982. Morphological in-dices of developmental progress in the par–smoltcoho salmon, Oncorhynchus kisutch. Aquaculture28:1–20.

Groot, C., L. Margolis, and W. C. Clarke. 1995. Phys-iological ecology of Pacific salmon. University ofBritish Columbia Press, Vancouver.

Haner, P. V., J. C. Faler, R. M. Schrock, D. W. Rondorf,and A. G. Maule. 1995. Skin reflectance as a non-lethal measure of smoltification for juvenile sal-monids. North American Journal of Fisheries Man-agement 15:814–822.

Healey, M. C. 1991. Life history of chinook salmon(Oncorhynchus tshawytscha). Pages 311–394 in C.Groot and L. Margolis, editors. Pacific salmon lifehistories. University of British Columbia Press,Vancouver.

Hillman, T. W., J. S. Griffith, and W. W. Platts. 1987.Summer and winter habitat selection by juvenilechinook salmon in a highly sedimented Idahostream. Transactions of the American Fisheries So-ciety 116:185–195.

Hoar, W. S. 1988. The physiology of smolting salmo-nids. Pages 275–344 in W. S. Hoar and D. J. Rand-all, editors. Fish physiology, volume 11. The phys-iology of developing fish, part B. Academic Press,San Diego,California.

Holtby, L. B., T. E. McMahon, and J. C. Scrivner. 1989.Stream temperatures and inter-annual variability inthe emigration timing of coho salmon (Oncorhyn-chus kisutch) smolts and fry and chum salmon (O.keta) fry from Carnation Creek, British Columbia.Canadian Journal of Fisheries and Aquatic Sciences46:1396–1405.

Jensen, A. J. 1990. Growth of young migratory browntrout Salmo trutta correlated with water temperaturein Norwegian rivers. Journal of Animal Ecology 59:603–614.

Johnson, J. H., and E. Z. Johnson. 1981. Feeding pe-riodicity and diel variation in diet composition ofsubyearling coho salmon, Oncorhynchus kisutch,and steelhead, Salmo gairdneri, in a small streamduring summer. Fishery Bulletin 79:370–376.

Kitchell, J. F. 1998. Physiological ecology. Pages 163–198 in S. I. Dodson and seven coauthors. Ecology.Oxford University Press, New York.

Knudsen, C. 1991. Scale pattern and age/length analysisof 1989 and 1990 Yakima River adult spring chi-nook. Pages 101–121 in C. Busack, C. Knudsen, A.Marshall, S. Phelps, and D. Seiler. Yakima Hatcheryexperimental design. Bonneville Power Adminis-tration, Report DOE/BP-00102, Portland, Oregon.

Komourdjian, M. P., R. L. Saunders, and J. C. Fenwick.1976. Evidence for the role of growth hormone asa part of a ‘light-pituitary axis’ in growth and smol-tification of Atlantic salmon (Salmo salar). Cana-dian Journal of Zoology 54:544–551.

Koskela, J., J. Pirhonen, and M. Jobling. 1997. Growthand feeding responses of a hatchery population ofbrown trout (Salmo trutta L.) at low temperatures.Ecology of Freshwater Fish 6:16–121.

Leland, H. V. 1995. Distribution of phytobenthos in theYakima River basin, Washington, in relation to ge-ology, land use, and other environmental factors.Canadian Journal of Fisheries and Aquatic Sciences52:1108–1129.

Levings, C. D., and R. B. Lauzier. 1991. Extensive useof the Fraser River basin as winter habitat by ju-venile chinook salmon (Oncorhynchus tshawyts-cha). Canadian Journal of Zoology 69:1759–1767.

Lindsay, R. B., and five coauthors. 1986. Study of wildspring chinook salmon in the John Day River sys-tem. Bonneville Power Administration, ReportDOE/BP-39796-1,Portland, Oregon.

Lindsay, R. B., B. C. Jonasson, R. K. Schroeder, and B.C. Cates. 1989. Spring chinook salmon in the Des-chutes River, Oregon. Oregon Department of Fishand Wildlife, Information Report Series Fisheries89-4, Corvallis.

Lobon-Cervia, J., and P. A. Rincon. 1998. Field as-sessment of the influence of temperature on growthrate in a brown trout population. Transactions of theAmerican Fisheries Society 127:718–728.

Major, R. L., and J. L. Mighell. 1969. Egg to migrantsurvival of spring chinook (Oncorhynchus tshaw-ytscha) in the Yakima River, Washington. U.S. Fishand Wildlife Service Fishery Bulletin 67:347–359.

Mayer, I., B. Borg, and E. M. Plisetskaya. 1994. Plasmalevels of insulin and liver glycogen contents in one-and two-year old Atlantic salmon (Salmo salar L.)during the period of par–smolt transformation. FishPhysiology and Biochemistry 13:191–197.

McCormick, S. D., and B. T. Bjornsson. 1994. Physi-ological and hormonal differences among Atlanticsalmon parr and smolts reared in the wild and hatch-ery smolts. Aquaculture 121:235–244.

McCormick, S. D., and five coauthors. 1995. Increasedday length stimulates plasma growth hormone andgill Na1 K1-ATPase in Atlantic salmon (Salmosalar). Journal of Comparative Physiology B 165:245–254.

Meyer, K. A., and J. S. Griffith. 1997. First-winter sur-vival of rainbow trout and brook trout in the HenrysFork of the Snake River, Idaho. Canadian Journalof Zoology 75:59–63.

Mommsen, T. P. 1998. Growth and metabolism. Pages65–100 in D. H. Evans, editor. The physiology offishes. CRC Press, Boca Raton, Florida.

Moriyama, S., and six coauthors. 1994. Developmentof a homologous radioimmunoassay for coho salm-on insulin-like growth factor-I. General and Com-parative Endocrinology 96:149–161.

Muir, W. D., W. S. Zaugg, A. E. Giorgi, and S. Mc-Cutcheon. 1994. Accelerating smolt developmentand downstream movement in yearling chinooksalmon with advanced photoperiod and increasedtemperatures. Aquaculture 123:387–399.

Murray, C. B., and M. L. Rosenau. 1989. Rearing ofjuvenile chinook salmon in nonnatal tributaries ofthe lower Fraser River, British Columbia. Trans-actions of the American Fisheries Society 118:284–289.

Myers, J. M., and ten coauthors. 1998. Status review of


chinook salmon from Washington, Idaho, Oregon, andCalifornia. NOAA Technical Memorandum NMFS-NWFSC-35 (also available at http://www.nwfsc.noaa.gov/pubs/tm/tm35/index.htm).

O’Connor, P. K., B. Reich, and M. A. Sheridan. 1993.Growth hormone stimulates hepatic lipid mobili-zation in rainbow trout, Oncorhynchus mykiss. Jour-nal of Comparative Physiology B 163:427–431.

Perez-Sanchez, J., H. Marti-Palanca, and S. J. Kaushik.1995. Ration size and protein intake affect circu-lating growth hormone concentration, hepaticgrowth hormone binding and plasma insulin-likegrowth factor-I immunoreactivity in a marine tele-ost, the gilthead sea bream (Sparus aurata). Journalof Nutrition 125:546–552.

Plisetskaya, E. M., T. W. Moon, D. A. Larsen, G. D.Foster, and W. W. Dickhoff. 1994. Liver glycogen,enzyme activities, and pancreatic hormones in ju-venile Atlantic salmon (Salmo salar) during theirfirst summer in seawater. Canadian Journal of Fish-eries and Aquatic Sciences 51:567–576.

Plisetskaya, E. M., P. Swanson, M. G. Bernard, W. W.Dickhoff. 1988. Insulin in coho salmon (Oncor-hynchus kisutch) during the parr to smolt transfor-mation. Aquaculture151–164.

Rogers, I. H., C. D. Levings, W. L. Lockhart, and R. J.Norstrom. 1989. Observations on overwintering ju-venile chinook salmon (Oncorhynchus tshawytscha)exposed to bleached kraft mill effluent in the upperFraser River, British Columbia. Chemosphere 19:1853–1868.

Ryan, T. P. 1997. Modern regression methods. Wiley-Interscience, New York.

Sagar, P. M., and G. J. Glova. 1988. Diel feeding pe-riodicity, daily ration and prey selection of a riv-erine population of juvenile chinook salmon, On-corhynchus tshawytscha (Walbaum). Journal of FishBiology 33:643–653.

Schrock, R. M., J. W. Beeman, D. W. Rondorf, and P.V. Haner. 1994. A microassay for gill sodium, po-tassium-activated ATPase in juvenile Pacific sal-monids. Transactions of the American Fisheries So-ciety 123:223–229.

Shearer, K. D., J. T. Silverstein, and W. W. Dickhoff.1997. Control of growth and adiposity of juvenilechinook salmon (Oncorhynchus tshawytscha).Aquaculture 157:311–323.

Sheridan, M. A. 1986. Effects of thyroxin, cortisol,growth hormone, and prolactin on lipid metabolismof coho salmon, Oncorhynchus kisutch, during smol-tification. General and Comparative Endocrinology64:220–238.

Silverstein, J. T., K. D. Shearer, W. W. Dickhoff, and E.

M. Plisetskaya. 1998. Effects of growth and fatnesson sexual development of chinook salmon (Oncor-hynchus tshawytscha) parr. Canadian Journal ofFisheries and Aquatic Sciences 55:2376–2382.

Sterne, J. K. 1995. Supplementation of wild salmonstocks: a cure for the hatchery problem or moreproblem hatcheries? Coastal Management 23:123–152.

Storebakken, T., S. S. O. Hung, C. C. Calvert, and E.M. Plisetskaya. 1991. Nutrient partitioning in rain-bow trout at different feeding rates. Aquaculture 96:191–203.

Taylor, E. B. 1990. Environmental correlates of life his-tory variation in juvenile chinook salmon, Oncor-hynchus tshawytscha (Walbaum). Journal of Fish Bi-ology 37:1–17.

Tveiten, H., H. K. Johnsen, and M. Jobling. 1996. In-fluence of maturity status on the annual cycles offeeding and growth in Arctic char reared at constanttemperature. Journal of Fish Biology 48:910–924.

Vijayan, M. M., A. G. Maule, C. B. Schreck, and T. W.Moon. 1993. Hormonal control of hepatic glycogenmetabolism in food-deprived, continuously swim-ming coho salmon (Oncorhynchus kisutch). Cana-dian Journal of Fisheries and Aquatic Sciences 50:1676–1682.

Weatherley, A. H., and H. S. Gill. 1995. Growth. Pages101–158 in C. Groot, L. Margolis, and W. C. Clarke.Physiological ecology of Pacific salmon. Universityof British Columbia Press, Vancouver.

Wickelgren, I. 1998. Obesity: how big a problem? Sci-ence 280:1364–1367.

Zaugg, W. S. 1981. Relationships between smolt indicesand migration in controlled and natural environ-ments. Pages 173–183 in E. L. Brannon and E. O.Salo, editors. Proceedings of the salmon and troutmigratory behavior symposium. University ofWashington, Seattle.

Zaugg, W. S. 1982. A simplified, partially purified prep-aration for adenosine triphosphatase determinationin gill tissue. Canadian Journal of Fisheries andAquatic Sciences 39:215–217.

Zaugg, W. S. 1989. Migratory behavior of underyearlingOncorhynchus tshawytscha and survival to adult-hood as related to prerelease gill (Na1 2 K1)-ATPase development. Aquaculture 82:339–353.

Zaugg, W. S., E. F. Prentice, and F. W. Waknitz. 1985.Importance of river migration to the developmentof seawater tolerance in Columbia River anadro-mous salmonids. Aquaculture 51:33–47.

Zhou, S., R. G. Ackman, and C. Morrison. 1996. Ad-ipocytes and lipid distribution in the muscle tissueof Atlantic salmon (Salmo salar). Canadian Journalof Fisheries and Aquatic Sciences 53:326–332.

physiological status of naturally reared juvenile spring … 2010/beckma… · (yearling smolt...

Documents