Loading...
HomeMy WebLinkAboutSUS364Growth Rate and Body Cornposition of Fingerling Sockeye Salrnon, Oncorhynchus nerkao in relation to Ternperature and Ration Size .J. R. Bnnrr, J. E. SnBruounx, AND C. T. Snoop Fisheries Research Board o;f Canad.a B'iological Stotion, Nanoimo, B.C. BHnm, J. R., J. E. SnetnounN, eNo C. T. Snoop. 1969. Growth rate and body composition of fingerling sockeye salmon, Oncorhynchus nerko, in relation to temperature and ration size. J. Fish. Res. Bd. Canada 26: 2363-2394. The growth of young sockeye salmon (Oncorhyn,chus nerko) was studied at ternperatures ranging from 1 to 24 C in relation to rations of 0, 1.5, 3,4.5, and 67o of dry body weight per day, and at an "excess" ration. Optimum growth occurred at approxi- mately 15 C for the tlr.o highest rations, shifting progressively to a lower temperature at each lower ration. The naximum growth rate for sockeye 5-7 months old was 2.616/da,,t; that for fish 7-12 months old was 1.6c/6/day. At 1 C a ration of 1.S%/day was sufficieut to provide for a maximum growth rate of 0.23o/6/day. The maintenance ration was found to increase rapidly above 12 C, amounting to 2.6o/6,/day at 2O C. No growth took place at approximately 23 C despite the presence of excess food. Isopleths for gross and net food-conversion efficiencies were calculated. A maximun'r gross efficiency of 25/s occurred in a small area with a center at 11.5 C and a ration of 4.lo/o/d.ay; a maximum net efficiency of 40/s occurred within a range of 8-10 C for rations of. L5/6/d.ay down to O.8o/s/day, the maintenance level. Gross body constituents changed in response to the imposed conditions, varyirrg in extreme f.rom 86.9/6 water, 9.4a/6 protein, and 1.\ok fat for starved fish at 20 C to 71.3c/6 water, 79.70/6 protein, and 7.6/6 fat on an excess ration at 15 C. It is concluded on the basis of growth and food-conversion efficiency th.rt tenrper.ittLres front.5 to 17 Cl are most favorzrble for youue sor:keye, and th:ri a general physioloqical optimum oc:r:urs in the vicinitl' of 15 ll. Received llarch 1.3. 1969 INTR()DI]C'IION Tuts PaPER is the lirst of a series deaiing r,vith laboratory studies on the relation of environmental factors (abiotic and biotic) to the grou'th rate of ]'ounq socrke)'c salnroll, Oncorh-t,nchlrs ?terka, The ainr is to olrtain a broercl undcrstanding of tlie lrioenergetics of this species throughout its life histor-v. As such, thc present work constitutes an extension of the research on rnetabolic rate and performance of sockeye salmon for u'hich a considerable background of in- forrnation croncerning the effects of tenrperature. size, and su.imnring spceci is available (13rett, 7964, 1965, 1967). I3y relating food consumption, gro\vth rate, and metal)olic rate the energy budget can be determined, providing a measure of the elficiencl' of food con- version and of energy loss tl.rrough metabolism and excretion (see \'Vinberg, Printed in Canada (J1409)J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2361 Jor,rRN.\r- FrslIEnrES RESE-\RCII Bo-\RD oI.' c\N-l,D.\, vor-. 26, No. 9, 1969 1956; Warren and Davis, 1967). Studies of ttiis sort can be expected to contri- bute to an improved understanding of the food requirements of natural popu- lations as rvell as provide fundamental information for fish-cultural practices. \"Iany of the methods and approaches to animal energetics, conducted for years on u,arm-blooded vertebrates (e.g., Brody, 1945; Kleiber, 196I; Blaxter, 1965), n-ray be applied 'lr'ith profit to studies on fish energetics. The present experiments deal lvith the combined effects of temperature and ration on grou'th rate. Temperature has been shown repeatedly to be one of the nrost influential environmental factors affecting the grou'th of fishes (Baldu'in, 1957; Brolr.n, 1957; Donaldson and Foster, 1940; Haskell et al., 1956; Paloheimo and Dickie, 1966a, b; Strarvn, 1961; Srvift, 1964; West, 1966). When food is present in abundance an optimum temperature for growth l-ras been recorded, varying among species and in the case of the desert pupfish (Cyprinod,on m.acularius) significantly influenced by salinity (Kinne, 1960). Quantity and quality of food under ambient temperature conditions have been studied intensively, especially for artificial culturing of fish (e.g., Brown,1957 Halver and Shanks, 1960; Hatanaka and Takahashi, 1960; Phillips et al., 1966). Davis and Warren (1968) reported that young chinook salmon (Oncorhynchus tshawytscha) weighing 0. 6 g 'n ould consume a ration as high as 20/6 of their dry body r'veight per day. However, the in- leract'ing effect of ration and temperature on the growth rate of fish does not appear to have been the object of critical study except lvhere either maintenance or ad libitum rations were involved (Brown, 1946; Pentelow, 1939). It rvas l-rypothesized that the optimum temperature for growth r'vould drop as the ration decreased, accompanied by a reduction in conversion effi- ciency. This was based on the supposition that the decrease in maintenance metabolism that accompanies reduced temperature would permit comparatively better grou'th at lorver temperatures u,'hen the source of energy r'vas restricted. The assumption l'vould apply only for a poikilotherm if the temperature- dependent activity of digestive enzymes and growth processes did not exert such a controlling influence that the potential shift was inhibited. The presence oi sharp peaks in the activity of digestive enzymes in relation to temperature lrzrs lreen demonstrated for the brown bullhead (Ictalurus nebulosus) by Smit (1967). Hoar (1966) further confirms tl.ris as a general phenomenon among poikilotherms (see also Jennings, 1965). MATERIALS AND METHODS The program of study was conducted over a period of 3| years during which time some change in methods was adopted. The initial phase, dealing with the effect of temperature (5, 10, 15, 20, and 24 C) using excessive ration, lasted for 7 months (June 2, 1964-January 12' 1965). This series served to establish the pattern of the grorvth curve, to provide an estimate of the maximum growth rate at each temperature, and to assess the methods employed. Subsequent experiments involving reduced rations were perfornred from November 1,1965, to February 3, 1966, at temperatures of 5, 10, and 15 C, and from December Ll, 1966, to March 5' 1967, at temperatures of I and 20 C. The methodology for the last two experiments was essentially similar. A distinction will be made only between the two series - A group, on excess ration, and B group, including restricted rations - whenever this is pertinent to the presentation.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. tsRETT Ct AI.: SOCKEYE GRO\VTH AND COX{POSITION SouncB eNo Curruntxc on Ftsn 2365 All fish rvere shipped as eyed eggs from Scully Creek, Lakelse l-ake near Terrace, B.C. After hatching, the young were reared in circular or oval Fiberglas tanks at the Biological Statior-r, Nanaimo, B.C. l'hey were the progeny of a single cross in each of the 3 years. Subject to ambient water temperatures from a lake source, the fish lvere hand-fed on a diet of beef liver and Clark's pellet feed (J. R. Clark & Co., Salt l,ake City, Utah) supplemented with a weekly supply of frozen brine shrimp. No history of epidemic disease, mortality in excess of l07o,r or poor growth was present in any of the stocks tested. When transferred from the rearing tanks the fish were first screened for a fairly uniform, intermediate size (mean *1 so) and then distributed by random numbers to the experimental tanks (see Table 1, 3). The order of tanks according to ration was also assigned by random ntrm- bers. ExpnntlrnNrer Tenxs Twelve 197-liter (43-gal) oval tanks equipped with sr-rbmerged jets to provide circulation and efficient water exchange were used (Fig. 1, 2; see also Alderdice et a1., 1966). 'lhese were modified slightly by the addition of recirculating pumps and a gas-stripping column to suppllr approximately air-saturated water at each controlled temperature. In most instances a 12-inch, tubular air-breaker was provided as an additional source of oxygen, and served as an emergency stand-by. Water was flushed through the tanks at a rate of 240 liters/hr (53 gal/hr). At no tinre was the oxygen saturation found to exceed 1.067o or be less than 87/6. The average was 95/o' Cover was made available by a black coating over the center half of the Plexiglas top. Water velocity within the tank ranged from 9 to 15 cm/sec (0.3 0.5 ft/sec) for most of the oval path; lower velocities occurred in the vicinity of the central drain. Fish were observed swimming mostly in the main flow arvay from the areas of reduced velocity. The application of a low velocity current was considered desirable not only for its cleaning action but also to facilitate food presentation and to mairtain a greater uniformity of activity between tanks. Spontaneous activity is temperature-dependent in many fish including salmonids (Brown, 1957; Fisher, 1958). By inducing all fish to swim constantly reduction in this inherent variable was effected. TEuprnerunn AND Lrclrr By cross-mixing between temperature-controlled supply lines with hand valves, or by ttse of thennostatic valves, temperatures were maintained within +0.3 C (an average of +2 sl for all tanks, read from standardized thermographs - see Table 1). The extremes of experimental temperature (l and 24 C) 'lvere chosen to be approximately 1 degree C above and below the respective lower and upper lethal temperatures for underyearling sockeye in fresh water (Brett, 19s2). The A-group fish were transferred from culture tanks at an ambient temperature of 7 Ci to their assigned growth tanks. The temperature was raised at a rate of 1 degree C/day, and length-weight measurements were taken as soon as the prescribed temperature for all tanks was reached (June 2). Since a growth "adjustment" period of the order of 2 weeks was observed in the initial experiments (see Fig. 3), the B-group fish were provided with this further introductory period. Their ambient temperature ranged from 10 to 12 C. The fish tested at I C t'ere allowed an additional 2 weeks of acclimation at this low temperature after an intermediate 2 weeks at 5 C. Fluorescent lighting illuminated the tanks with a minimun-r of 15 ft-c at the water surface. 'lhis was controlled by a time switch set for a 9-hr day (0800-1700 hr, rsr). Natural light was reflected on the tanks from a louvre-shaded window providing normal photoperiod from spring to early fall. IFrom physical darnage, nipping, popeye, and accidental loss.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. fesl-E 1' T'anktemperatureaudvitalstatisticsoJ -r\-group sockeye fed on excess ration. Data for 1 C have beel ilclyded fr6m B gro'p. Variabiiitl, is recorded as *2 sn except for temperatures (t2 sn;. Acclimatior-r temperature (C) 24r.)10 Nlean temp (C) Test period (do1,s) Initial length (cn) Initial wt (g) Terminal length (cm)'Ierminal rvt (g) Initial growth rate (/s/d.ay) Intermediate growth rate (%/doy) Terminal growth rate (%/day) 1.09+0 16 83 8.66+0 11 6.04 +0 .22 9 .20 +0 12 7,69+0 30 0.23 +0 06 .1 80+0.47 2t9 5 4i +0.08 1 18+0.06 10.14+0 68 7 83+1.76 1.18+0.57 0.99+0.16 0 76+0.11 9.92+0.16 2t9 5.74+0.14 1.43 +0.10 16.78!r.14 46 40+10 50 2.97 +0.73 1.96+0.17 1.35+0.12 14.9t+0.12 220 5.65+0.12 1 .33 + 0.09 19.73+0.86 85.50 + 12. 70 4.24+0.57 2.60+0.18 1.42+0.11 20.09 +0 28 5. 81 + 0.05 1 40I0.08 16.85+0 47 51 .70+4.70 5.05 + 0. 65 1.81+0 17 1 2I+0.13 23 .68 +0 .44 44 5 .67 + 0.08 1 31+0 06 5.74+0.20 1.18+0.22 -0.34 + 0.35 -0. 20 + 0.66 (50% dead) oo o z,- a E a a , a o 0 o E o z F f c 9 *J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT ct al.: SOCKEYII cl{OWTH AND COX,IPOSi'1'ION tJo I FIc. 1. Typical arrangelnerlt of tanks and services assembled for growth experimeltts. b'tc. 2. Example of size difference in A-group fish after 3 months at 5 and 10 C. on excess ration. Note uniformity of size within tanks.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2368 JOURNAL FISHERIES RESEARCH BOARD OF- CANADA' VOL.26, NO.9' 1969 |' I F' I 9U- TIME - Doyt FIc, 3. l\lean weights of samples (n:20) of A-group fish held a,t five tem- peratures on exceti rations. Experiments starteil on June 2. .l'imits are ior *2 sE. Three stanzas of erowth are indicated accoiding to inflection of ;i;;. O* poinf (circled) for-fish at 5 C was discarded as being beyond the 95% confidence limit of the mean growth rate. Sellpr.trc Pnocroune A GROUP Samples of 20 fish were removed etery 2 weeks from an initial stock of 250 fish per tank. On each oicasion fish were herded into one end of the tank and caught by drawing a small dipnet vertically through the group. The fish were killed by anaesthetic (150 ppm M.5.222,\, and rolled in a slightly damp cloth to remove surface moisture. Fork length and weight measurements were taken to the nearest 0.1 cm and 0.01 g. The fish were then placed into a deep freeze for later analysis of gross body constituents. The possibility of sampling bias was tested by measuring atl the fish in a control tank in lots of 20, throughout the experimental period. No significant difference occurred between the sample and remaining population in seven such comparisons 1P ( 0.05, using the method of Hubbs and Hubbs, 1953). B GROUP To reduce chance differences between samples within each tank, later experiments were c<-rndtrcted with 25 fish only. These were lightly anaesthetized for measurements every 2 weeks and returned immediately to their respective tanks. Fish were not fed on weighing days. To check on the possible efiect of periodic handling and anaesthetic all the frsh in a tank of 250 were subjected to "standard treatment" once a month and their weight relations compared with a control group, No significant difference was apparent over a 3-month period although the anaesthetized group grew lsVofaster in the first 2 months, then slowed down slightly. Because of the rapid recovery of the fish and their excellent appetite, it was assumed that the effect of the biweekly treatment was unlikely to be different from the monthly check' Fooo eNn FBE,orNc Although essentially the same ingredients were used (see Appendix A) the composition of the diet differed between the two groups (Table 2). The moisture content was purposefullyJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION 2369 T.tern 2. Composition and calorific value of diet fed to the two groups of fish. Figures obtained from an average of 10 samples (A group) and 14 samples (B group). A groulr (1964-65)L3 group (1965-67) \\ret (%) I)ry (%) Wet (%) f)ry (lo) Cornponents \Vater Protein Fat Ash Fibera Carbohydrate" Kcal/g Bornb calorinreter Calculatedb ct 25 IJ A 1 6 .A 50 27 8 2 IJ 5.; 63 22 6 3 I o 2.0 2.0 60 lo 8 lA 5.4 'Approximate fiber content determined from constituents (Anon., i959) with carbohvdrate estimated bv difference. bCalorific value calculated from proportion of dry constituents using values of 5.7 (protein), 9.5 (fat), and 4.0 (carbohydrate) kcal/g. irrcreased from an average of 51o/s for A group to 63o/p for B group, for rvhich controlled rations were used, to facilitate dispensing. -fhe reduction in fat rvas the result of differences between years in the composition of the major ingredients - canned salmon and commercial pelleted feed. New food was prepared every 2-3 t'eeks as required and stored at -10 C, the noisture content being checked on each occasion. Proxirrate analyses of frozen spot-samples were performed at the end of the test period. The methods used were the same as those for analysing fish, described in the next section. The diet is one that has evolved over a number of years for use in culturing experimental fish in our iaboratorv. Obviously high in nutrient quality it has provided excellent growth, good condition, and no rnortalities. Since the difference in diet composition occr.rrred between A and, B groups, and not with'in the B group for rvhich restricted rations were used, the change did not result in any significant difference between the various excess feeding regimes, although there is a tendency for the A group to have somewhat higher mean growth rates (see Fig. 6). For the A group, food was dispensed by shredding frozen cnbes through a hand ricer. The trsh u'ere fed to satiation three times a day at 4-hr intervals, starting at 0830 hr. 'lhis was temred "excess." Depending on telnperatr-lre, the daily anlount presented during the ternriual grolvth stanza ranged from 10 to 744,/6 of the body weight on a dry rveight basis. The B group was fed with a small press that ejected a prescribed rveight in the lonn of nrLrlLi- ple short strings of unfrozen food. The restricted rations selected as a fraction of dry body weight were:0, 1.5,3,4.5, and.60/6, andexcess. Itwasdiscoveredthatasinglerationof 3(/6/d.ay,dis- pensed in about 5 min, was sufficient to allow each fish to obtain a "fair share" of the food. -l-he 1.5/6 d,aily ration was therefore presented as 3/6 every 2 days, and the 6/6 ration as 3lp twice a day. In the case of excess ration the uneaten particles were washed out of the tank within 3-8 min from the time of introduction. For the B group the small wastage involved was assessed, providing an estimate of the maximum daily intake.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2-170 TOI.'ITNAL FISHERIITS RESEAI{LIH IIOAI{D Ol'' CAN;\D\, VOL. 26' NO 9' 1e69 Sockeye are a rlatr.lrally schoo'lirlg lish. As underyearlings2 they do not display obvious agqressive or territorial behaviour when food is evenly dispersed and general uniformity of con- ditions prevail within holding tanks. If food is not corlsumed proportionatellr amongst all {ish, growth depelsation occrrrs, resulting in an increased size-differential with time (Magnuson, 1962). The variarce, as indicated in the grorvth relations of A-group fish (1og weight +2 ss, n : 20, see Fie. 3), did not differ in extreme instances by more tha:l 33/6 of the overall mean value, except at 24 C rvhere grorvth rvls inhibited, and for the last rveighing as a result of the smaller san'rple (n : 10). Less differeuce occLrrred in the B-gror.rp experitnents rvith the exception of three instauces rvhere there u,as a change in variance zrssociated with the accidental loss of fish under anaesthetic, and orle ir.r-.tance s,here maintenar-ice ration rvas involvecl. In general, disparity in consnmpticrtt of iood arnong the tlembers of each s:rrnple did not occur to any sigr1ifrcant degree' ItnottlrArtl AN.tr-lsrs AND CALoRIFIC \'TALUE At the beginning and end of each experiment gross body constituents were determined f rom sermples of 20-253 iish, starved for 24 hr. Subsamples of five fish were grortttd to a fine homog- enate and u'eighed-fractions $ere used to deterrnine'lvater, fat, and protein content in duplicate by the follou'ing methods. Woter- Approxirnately half the subsample was oven-dried at 105 C for 24 hr. A correction for an-v moistttre loss during deep-freezing was applied' Fr:orn the sum of all samples taken in the 1964 tests and those at the start of the 1965 tests ameanwatercontent ol 74.4+t.2% (+2se, n:21) wasobtained.TheBgroupaloneaveraged 7+ gr/c. As a basis for calculating the ration in terms of fractions of dry body weight a general level of 25t6 dry material u'as applied throughout. Protreiy. - Nitrogen content was deterrnined by the micro-Kjeldahl techniqtre outlined by the Associatiol of Of6cial Anal1'tical Chernists (1960). The valrte obtained rvas multiplied by 6.25 to obtain the average protein value. Nonprotein ni,trogett. .-. The difference between the acid-soluble and the acid-insoluble nitro- gen values (method of Association of Official Analytical Chemists, 1960) gave the nonproteitr r-ritrogen content. It averaged 0.2/6 of. dry weight (about 0.05/6 of wet weight) and varied from an undetectable amount to 1 .2ls of dry weight. The highest levels were associated with the highest temperatures, independent of ration except for starved fish, which contained consistently low levels. Since only major constitnents were being examined these small values have been omitted from the compilation in Table 3. Fat - Lipid analyses lvere first performed by the chloroform r-nethanol rvet extractiotr technique of Folch et al. (1957). This was later changed to hexane extraction from dried material (see Appendix B). A slight difference in values was ascribed to the possible carry over of moisture by the wet extraction method. An appropriate correction was therefore applied to bring all results to a common ba.is for dry extraction' Carbohydrote - No analysis of carbohl'drate was cor.rducted since this constituent does not amount to more than 0.5% of body weight (Vinogradov, 1953; Black, 1958). Highest con- celtration occurs in the liver where values from 3 to 4/6have been obtained for normal feeding fish (Hochachka and Sinclair, 1962). Cal,ori,metry - Calorific values were determined with a Parr bomb calorimeter (methods prescribed by the American Society for 'l'esting Materials, 1966). \\ihen the calorific content was calculated from proxirnate analysis, values of 9.5 kcal/g for fat and 5.7 kcal/g for protein u'ere applied (Brody, 1945; Kleiber' 1961). De,re CornpnATroN AND ANALYSIS The quantitative larvs governing growth have been treated as general concepts by such authors as Bertalanffy (1957) and Needham (1964), u,ith more specific consideration of the process 2Aggressive behaviour has been observed in yearling fish, lvith inteuse periods occurring soon after feeding. 3lixcept in a few cases listed in Table 3.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2372 JoURNAL FTsHERiES RESEARCH BoARD oF CANADA, voL.26, No. 9, 1969 in fish provided by Brown (1957),'faylor (1962), and Winberg (1956). llnder conditions of en- vironmental control and in the absence of limiting factors, growth is a mtrltiplicative process, which, irr the early stages of life, often follows an exponential curve. Under the influence of apparent inhibition with age its overall configuration tends to follow the sigmoid shape of a logistic curve. The instantaneous rate of gain in ir.eight per unit weight (per cent change in weight/time, or specific growth rate) declines progressively with increasing size. The general equation for exponential growth is w : beo', where zr,' is size and I is time. When logqz, is plotted against i, (Fig. 3) the slope ft X 100 is equivalent to the specific growth rate (G). This was determined by taking length-weight measurements every 2 weeks, usually for a 10- to 12-week period. An estimate of the error of ,b was computed from the pooled data of all measurements (*2 sn). 'lhe restricted ration was adjusted at the start of each biweekly period according to the new n'rean weight obtained, on a dry weight basis. The actual proportion of the ration was subsequently calculated according to the observed mean weight achieved by the fish during the biweekly interval, e.g., as the fish grew the actual fraction became slightly less than the ration initially prescribed. Because of the progressive change in growth rate that appears to occur in stanzas of decreas- ing slope (Fig.3), a period of relatively stable growth from October to February (Fig.5) was accepted as a suitable time for making comparisons betrveen treatments, spread over a number of years. RI'SUI-TS TBupnnlrunB X Excnss RrrroN - A Gnour As a result of the favorable change in density, diet, and feeding frequency that accompanied the transfer from culture to experimental conditions, all stocks shorved a marked increase in grorvth rate, except those at 24 C (Fig. 3, Table 1). This surge declined after an initial 2 r,veeks "adjustment" (possibly less) and the growth rate remained constant for the next 8 weeks. At this time (August 10) a synchronous decrease in slope occurred in all stocks despite the difference in acclintation temperatures of 5-20 C. No further change was noted up to the time of terminating the experiment 2I weeks later (January 12). Such seasonal variations in growth rate have been observed by Swift (1955) for 1'ratchery-reared bror,vn trout (.Salmo trutta) exposed to normal telnperature :rnd light. A reduced growth rate occurred in l,Iay and June despite increasing temperature and daylight. It is possible that changes in the production of growth hormone could account for such temperature- independent decreases in growth. The maximum growth rate recorded in the initial period \\ras 5.05%/day at 20 C, a doubling of rveight in 2 weeks. The relation bet\,veen temperature and maximum growth rate for the two subsequent stable periods ("intermediate" and "terminal" --'Iable 1) has been plotted in Fig. 4 including one instance where excess rat;on was fed in the B-group experinlents (1 C,'fable 3). At an age of 5-7 months, from the start of feeding, a pronounced optimurn temper- ature for grolt'tir occurred at 15 C. Although still apparentfor 7- to 12-month- old fish, the peaking of the relation is sulficiently reduced that no statistical difference exists between the grou'th rates at 10 and 15 C. With increasing age, growth continues to slow down so that optimunt temperatures become progressively Iess apparent. The expression of maxjmum growth rate in relationJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION 237 3 to acclimation temperature (Fig. 4) has been called the "scope for growth" by Warren and Davis (1967) by analogy u,ith the "metabolic scope," which h-ry Q9al used to describe tl're capercity to elevate n'retabolic rate above the maintetrance or standard level at various ternperatures. AGE - Months TEMPERATUF Frc. 4. Relation between temperature and growth rate (+2 se) of young sockeye salmon fed on excess ration. Age is in months fron start of feeding. At 24 C some of the young salmon would not accept food. trtlortalities cornmenced in the 1st week. The mean r'veight declined and the variance increased as a consequence of the variability in feeding response. Within 3 weeks one-third of the stock r'vere "pin-heads," w-hich died rvithin the month. By the 44th day over 50/6 were dead, at lvhich time the experiment was termi- nated. Similar problerns of feeding and subsequent mortality near the boundary of the upper tolerance limit for this species have beetr observed by Donaldson and lioster (1940) at 23 C, and rvere encountered by Brett (1952) r'vhen he attempted to acclimate 3-month-old sockeye to high temperatures, elevated by stages to 24 C. TBrrpBnarune X PnBscnrilED RArroN - B Gnoup An example of the effect of ration level on growth rate at a given temper- afure is illustrated in Iris. 5. It rvas characteristic of most of the data to exhibit o 0 u= :e I F;} o 0 I E 6U J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. a= E I U o z I o 9 U= 2374 JoURNAL FISHERIES RESEARCH BOARD oF CANADA, vor,.26, No. 9, 1969 Flc. 5. The effect of ration on growth rate of fingerling sockeye at 10 C. For ease in comparing rates (slopes) a common origin for all stocks has been used. Solid lines link each mean weight (n:251, limits:+2 sr). Broken lines represent the computed slopes for all data at each ration. Prescribed rations are indicated; actual consumed rations are oresented in Table 3. Note that the statistical separation of the fish on 3/p ration was not possible in less than 6 weeks. deviations along the path of exponential gro\'vth, and these deviations, together with the difference in rveight within each lot of 25 fish, contributed to the variance in slope (specific growth rate, Table 3). At high feeding rates, ap- proaching the maximum intake, it is apparent that the time required to demon- strate a significant difference in growth rate bet\'veen t\ /o levels of ration was of the order of 6 8 rveeks (".g.,3 and 4.5/6 ration at 10 C, Fig. 5). When growth rate uras plotted against ration for each temperature a series of curves was obtained, which change in form from a "logistic shape" at20 C to a "geometric shape" at lower temperatures (Fig. 6). In the absence of a generally suitable transformation a smooth curve has been fitted by eye to pass through the means *2 sB. Since various growth parameters can be obtained by interpolation all the computed points are presented graphically for inspection in Fig. 6. The parameters of particular interest were: (1) the maintenance ration, i.e., that ration that just maintains the fish without any lveight change; (2) the optimum ration, i.e., that ration that provides for the greatest grorvth for least intake (most efficient); and (3) the maximum ration, i.e., that ration that just provides for the maximum growth rate.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION zSlJ RATI0N - %WEIGHT/Doy c. 6. Relation of growth rate to ration at five temperatures. Points for A-group fish (O) have been rtted for an average ration of 1016, presented as -excess. \\rhere B-group fish were_ satiated by the prescribed ration, or fed a measured excess, the points have been circled' These may be derived geometrically (Fig. 7) (ct. Thompson, 1941). The fact that the optimum ration can be determined by drawing the tangent to the curve from the origin may be proven by calculating the ratio of grlwth rate to ratiln for a series of rations. The parameter most difficult to determineJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2376 JOURNAL FTSHERTES r{ESnARcH BoARD oI. CAN.{D-A., vol-.26, No.9, 196q F I tr= :i I U I =o E RATION - '/. WEIGHT / Doy Frc. 7. Geometric derivation of various parameters of growth with accom- panying ration, using the data for fish at 10 C as an example. accurately was the rnaximum ration, involving thc asymptote for nraximum grol'th rate, the values for rvhich \vere sul)ject to greater variability than most of the other rates. With increasing tenrperature there rvas a large increase in the ration required to meet the defined grou'th parameters (F-ig.8). The maintenance ration at20 C \,vas seven times that at 1 C; tl.re optinrurrl was five times greater for the same temperature difference; and the maximum just under three times greater. The maximum food intake per day \\ras approximately 8.76 of. the dry body $reight at 2A C. TI.re temperature X ration relations could be made linear (Fig. 8, equations), improving the basis for estimated values and providing for extrapolation to the extremes of temperature at u'hich food r'vas still accepted. Above 23.3 C loss of appetite and inefficiency of food conversionlvere respon- sible for the breakdor.r'n in the overall relation, and the ultimate death of the fish. Gnoss Boov ColrsrrruFrNrs Botl.r ration level and temperature had a considerable effect on all body constituents (Table 3). They ranged from 86.97o water, 9.44/6 protein, and L00/6 lat at 20 C for fish starved for 83 days, to 71 .37o water, 19 .7% protein, and 7. 6/o fat for fish on an excess ration lor 99 days at 15 C - the optimumJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH -{ND COMPOSITION 2377 oo Fr o U=s z 9 F Fre. 8. Relation of maintenance, oDtimum, and maxintum rations to temperature. Points were determined as outlined in text, except the tu/o circled points, which were derived by inspection from Fig. 11. temperature for grorvth. This dynamic and variable state of body constituents is apparent for all stocks and is represented by the progressive change in moisture content depicted in Fig. 9. At all temperatures a lolver u'ater fraction occurred with increasing ration, reaching a minimum of about 72/6 by the termination of the experilnents. Among the starved fish the moisture content varied according to tempera- ture, with the exception of the fish held at 1 C. Since the rate of change in composition will be dependent on some function of temperature and time' it would be expected that changes rvould proceed very slowly at 1 C. Comparison between experiments of equal duration cannot be made on common physio- logical grounds unless a static state of body composition has been reached. This is suggested b)' including a column in Table 3 for day-degrees, as one indication of physiological time. It may be seen that on this basis the exper- iments rallged from 83 to 1700 day-degrees. In addition it $ras an inevitable outcome that considerable difference in size would characterize the various groups at the terminal stage. Size has been shorvn to influence body composition of young pink salmon, Oncorltynchus gorbuscha (Parker and Vanstone, 1966). As water content increased, both fat and protein decreased.a In the light of this highly significant negative correlation (fat : -0. 89, P < 0.01; protein : -0.80, P < 0.01) it rvas possible to derive simple equations with appropriate aAbove the maintenance ration any decrease in protein is strictly relative; however, fat may show some absolute changes, MAXIMUM Y = 2.68 +4.05 IoSoX MAINTENANCE logoYr -0.505+0.046X U fe U I F - F 3 TEMPERATURE - C J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. )378 JOURNAL FISHERIES RESEARCH BOARD OF CANADA, VOL. 26, NO.9, 1969 t-_l RATION - %WEIGHT,/Doy F E E=s I z Fz U E f F Io Frc. 9. Moisture content of fish at the start (single circled point with +2 sE) and end of growth experiments. Each "eyed" curve represents the final percentage ofbody water in relation to daily ration for a given -temperature' \\tith the exceptioir of 1 C, it is likely that these represent equilibrium states for any ration and temperature. confidence limits relating water content to corresponding fat and protein levels (Fig. 10). Contributing to the variability in composition betrveen samples of fish are the direct and interacting effects of temperature, ration, size, grorvth rate, and conversion efficiency, coupled with a degree of curvilinearity produced by approaching threshold levels of these major constiLuents. No attempt has been made to sort these out. It is apparent, however, that by determining the moisture content only the total gross body constituents can be estimated fairly accurately from the above relations, assuming an average ash of 2/6 of wet 1\.eight (Table 3). DISCUSSION Soulrcrs oF ERRoR The experiments rvere designed to provide statistical sensitivity for determining the difference betueen treatments (temperature and ration) by reducing the variability w'ithin lots (tanks of fish). Reduction in genetic varia- bility was achieved by selecting a limited size range from the progeny of one female. Although this was undoubtedly an effective approach in any one year, it did not anticipate the subsequent need for comparison between years' as well as the desirable feature of obtaining a wider representation of the population. The limited gene pool can be expected to be one of the factorsJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. IIRETT et al.: SOCKEYE GROWTH AND COMPOSITION 2i79 F 6= F ;l I zu z 2 U FoE - o Fg FIc. 10. Relation of protein and fat content to water content of young sockeye. Points represent pooled data from all temperatures ind rations. Lines of best ht deter- mined by least squares, with 68/6 confidence limits (broken lines). 75 AO 85 WATER CONTENT- i TOTAL wEIGT{T contributing to the differences in maximum growth rate between years, as exemplified by the various stocks of fish on excess ration at any one temperature (Fig. 6). Yet it can be seen that the dissimilarity u/as not great, amounting, for example, to a difference in specific growth rate of about 0.2/s/day f.or stocks at 15 C. Since some change in diet occurred between A and B groups (Table 2), and it r'vas decided to present excess food twice a day rather than three times a day in later experiments, the question of maximum growth rate for the species at this age and size remains undetermined though probably not seriously underestimated. Indeed, with advances in nutrition and in making environmental conditions optimal, it is doubtful if the full potential for growth of any species including man has yet been assessed. It is apparent from the change in body constituents that accompanied the different treatments that the presentation of restricted rations on the basis of the initial moisture content would apply throughout the experimental period only to those instances that did not deviate significantly from the original state (signified by the points close to the dotted line for 750/6 moisture in Fig. 9). Except for starving and excess-fed fisir an error progressively creeps in for all other instances, reaching a maximum at the end of the test time. The extremes occur at lorv temperature on high ration, and at high temperature on lou' ration. For example. the fish held at .5 C on 3.5%" ration had a terminal J F 2 ! g = \.I J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2380 JoURNAL FTSHERTES RESEARCH BOARD oF C.\NADA, vol.. 26, No. s, 1e6e average moisture content of. 72.5/6. On the final day the prescribed ration of. 3.5/6 rvas actually 3.27a of the existing dry weight. Lacking knorvledge of the rate at rvhich the constituents changed, coupled u'ith the fact that it rvould apply in varying degrees up to the sort of error indicated, no correction rvas possible. The basis of conducting exacting experiments on bioenergetics and grolrth in fish is treated with insight by Warren and Davis (1967). From a bioenergetic point of view the conversion of food should be based on calorific rather than dry weight values because of the differences in the fat/protein ratio. When considering growth only, tl'rere is justification for calculations based on nitrogen content such as Gerking (1962) used in studies on food conversion of the bluegill, Lepom'is macroch,irus. Since the protein fraction of the dry rveight varied some'lvhat less than the fat fraction (Fig. 10), the inherent error indicated above rvould be reduced accordingly. It is worth stating here that subsequent experiments are being conducted by increasing the number of fish to allorv for bir'veekly subsamples to be witl'rdrawn for appropriate analysis. With the above limitations set forth it is possible to examine the combined effects of temperature and ration on the specific growth rate, and to proceed to a consideration of the efficiencv of food conversion. TEMpBneruRE-RATroN Rpr-erroNs From tl.re relation bett'een ration and gror,r,th rate for each temperature (Fig. 6), the form of the grorvth X temperature curves for each prescribed ration may be determined (Fig. 11). The temperature for optimum growth shifts progressively to the left as the ration is reduced, moving from 15 C on excess ration to approximately.5 C for a ration of. 1.5/6/day' The outer curve is the same as that describing the scope for grorvth for 7- to 12-month-old sockeye presented previously (terminal grou'th rate, Table 1, Fig.4). With the exception of the relation for starved fish, eacl'r of the remaining domed curves represents the scope for growtl-r cl-raracterizing eacl-r reduced level of available food. As the fish grou, older there u'ould be a corresponding flattening of this fleet of curves, approaching the ultimate minirnum grorvth rate for aged fish. The shunt in optirnum temperature supports the hypothesis that, for acclimated fish, digestion and groivth can still proceed effectively at low temperatures. It suggests that either the enzymes involved are relatively temperature-independent (wide spectrum), that a fleet of enzymes is present with a wide range of temperature reactivitlt, or that rapid changes in the metabolic system occur to meet the new situation. 'Ihe marked increase in energy requirements for maintenance tl'rat accompanies rising temperature (trig. 8) accounts for the decreasing scope for grou'th, to the point of losing rveight at low rations (Fig. 11). Fish feeding on a limited ration in nature rvould achieve better growth by moving along the temperature axis, torvards the appropriate optimum.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et a1.: SOCKEYE GROWTH .\ND COMPOSITION 2381 d2 - 9 =s - =o 0 9 E Frc. 11. Effect of reduced ration on the relation between growth rate (+2 se) and tem- perature, for 7- to 12-month- old sockeye. Points for excess ration or where a prescribed ratiorr turned out 1o be excessive are marked with an X. The broken line for starved fish is a provisional interpretation. Thus, at 15 C on a ration oI 1.Sa/o/day no growtli u'ould occur' rvhereas for the same ration at 5 C a size increase of 0.3,76/day lr'ottld be possible. The daily vertical movements of young socke-ye observed during the summer in Babine Lake (Narver, N{S,1967) involved changes of 10 C, from near-surface feeding at dusk follorved by a gradual retreat to depths of 100 ft or more shortly after dawn. This could be conceived as a mechanism evolved to improve growth in the presence of a lirnited food supply. Since yearling n'rigrants frequently average no more than 6 g (Johnson, 1965), whereas experimentally fed young sockeye under optirnum conditions may weigh as much as 80 g 4 months prior to the smolt phase, it is not hard to conceive that food is a limiting factor for this species in many lakes. The predictive value associated u'ith established food requirements is best illustrated by calculating the grou'th-rate isopleths (Fig. 12). The food ration necessary to provide for any given growth rate at any particular temper- ature may be ascertained readily by checking the appropriate combination. Since this applies to only one grou'th stanza, or age, it is possible to see that by a series of such experiments, conducted at selected intervals throughout the life history, a method of computing the total food requirements would be obtained. f t F F T =--- I I --__ T=---l- i-------_J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2382 JoURNAL FrsHERrEs RESEARCH BoARD oF cANADA, voL.26, No. 9, 1969 TEMPERATURE - C Frc. 12. Growth rate isopleths for yearling fingerling sockeye, showing the percentage gain in weight per day that would be expected for any combi- nation of ration and temperature. Sranv.trroN EppBcr The rate of loss of body weight for starved fish in relation to temperature r'vas difficult to interpret since the mean rate for 15 C (-0.64o/a/day) 'was unexpectedly greater than the mean for the 20 C (-0 507o/day). These bordered on being statistically different (P : 0.04). A provisional line was drawn in Fig. 11 to represent the most likely relation for increased temperature. The experiments for temperatures of 1 and 20 C were conducted a year later than those for 5, 10, and 15 C. Aside from a difference in the origin of stocks it was conceivable that spontaneous activity might have been highest at 15 C resulting in a generally higher metabolic rate than for the stock at 20 C. It should be noted, however, that the fish at 20 C had a terminal moisture content of 86.9/6, whereas the fish at 15 C were 83.2/6 water (Table 3). A check on the provisional interpretation was made by conducting a comparable experiment at temperatures of 5, 10, 15, and 20 C for t2 weeks ln 1967 68. At this time only fish of approximately twice the weight were available (12.0 + 0.3 g).The results gave confirmation for the rate of loss of rveight at 5 and 20 C but not for the intermediate points (Fig. 13). It must be concluded that greater variability is either natural or somehow induced during starvation experiments. The sample size should therefore be increased and the conditions governing excitation should be rigidly controlled, o F Io U= f, z F &J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT ec al.: SOCKEYE GROWTH AND COMPOSITION 2383 TEMPERATURE - C Frc. 13. Rate of loss of weisht for starved fish in relation to temDerature. *:r'*?i::?''ot"1'l".tJJ;:,"*:?.-'TlnSri'*"',"uT';l1h""J?3:uin?i:iiJ was 12.0 e. particularly regarding disturbances that may result inadvertently from the feeding of adjacent tanks. The mean for all data has been used in Fig. 13 as the best representation of the general temperature relation. CoN vBnsroN EprrcrBNcrrs GROSS EFFICIENCY Few studies of growth relations in fish irave been performed with sufficient accounting of food consumed and flesh produced to permit determining conver- sion efficiencies (Kinne, 1960). With a common unit of dry r,veight, nitrogen content, or calorific value, the gross efficiency (Er) may be calculated as the (; ratio of output to input by the formula e- : , X 10070, where Q- : growth, and I : food intake. This simple index provides one of the most revealing aspects in analyzing growth phenomena, not only by indicating the circum- stances under which the animal is most effective but also by providing a measure of the most economical use of food. In addition to the factors of temperature and salinity, gross efficiency has been shown to depend on the type of diet, the feeding interval, and the size or age of the fish (Paloheimo and Dickie, t966b; Pandian, t967a, b). Conversion values range from 8/6 for omnivorous adult carp feeding on detritus s | -.c o o F I il = -.8 J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2384 JOURNAL F'ISHERIES RESEARcH BOARD oF CANADA, voL.26, No. 9, 1969 and algae (Kevern, 1966) to 44/o f.or carnivorous young Ophiocephal'us striatus feeding on a diet of prawns and chopped fish (l'andian, 1967a). lvlev (1945) considered 35o/o to be a maximunr for young fish on a high ration. t,Tnder- yearling cutthroat trout, Salnto clarki, shoived an individual variation of 13 37 /6 in ability to convert a diet of housefly larvae and adults at a mean temper- ature of 8 . 5 C (Warren and Davis, 1967). The combined effect of temperature and ration provided a small maximum area of 25/6 effi,crency for fingerling sockeye, rvith a graphically determined center at IL5 C and ration of.4.0/6/day (Fig. 14).5 The mapped isopleths {or TEMPERATURE - C Frc. 14. Gross efficiencv of food conversion in relation to temperature and ration, drawn as isoplethj overlying the growth cltrves of Fig. 11 (broken lines)' decreasing efticiencies radiate outwards, with a "favourable axis" passing from the lower left corner diagonally uprvards through the various growth optima for each prescribed ration. In line with this axis a large oval area is encompassed by the 20/6 conversion contour. It is apparent that temperature may have as much effect as ration on conversion efficiency when the full ranges of both are considered. The conclu- sion of Winberg (1956) that temperature has virtually no effect appears to result from a lack of available evidence for temperatutes aboae the optimum' When food is present in abundance (ad libitum, or excess rations), temperatures 6As an aid to interpreting this graph it is worth pointing out that by following a vertical line at any one temperature, say 10 C, the conversion efficiency increases from zero at a main- tenance ration, passes through a maximum value at an intermediate ration, and then decreases reflecting some wasteful gorging at high rations (cf. Fig. 7). o F I 0 U= s U F - =o E 0 g o u o J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT er al.; SOCKEYE GROWTH AND COMPOSITION 2385 below the optimum are associated u'ith redtrced food intake resulting in a fairly constant conversion efficiency, such as that along the "fayourable axis" of Fig. 14. This interaction produces an apparent ternperature independence, which proves to be circumstantial rvhen tl.re n'hole phenomenon is examined. A similar sort of limitation appears to apply to the analyses of Paloheimo and Dickie (L966b) who concluded that at a given temperature increasing ration results in a decreasing conversion efficiency. Because of the paucity of available data on restricted rations the majority of instances cited by these authors relate to supraoptimum rations, which inevitably result in Iowered eificiency. The full range of response is clearly demonstrated in the case of young sockeye by applying the method of analysis of Paloheimo and Dickie (1966b) who determined that the logaritl-rm of the gross growth efficiency (log K1) decreased linearly with increasing ration. Evidence to support such an hypothesis could come only from rations that r,r.ere sufficiently above the maintenance level to provide for "excessive feeding." Efficiency of food assim- milation increased |rom 71.5/o to 86.5/6 f.or goldfish fed restricted rations near the maintenance level, at 21.5 c (Davies, 1963). Assuming a moisture content of 75/6 for the adult goldfish the respective rations ranged from about 0.5/6/day to 2/s/day on a dry weight basis. It is possible that in some species the position of maximum grorvth efficiency may occur at comparatively low rations, particularly under experimental conditions rvhere the work associated with food-finding is mini'ral. As is indicated for young sockeye (Fig. 15) the inflection of log K, to a negative slope occurred at progressively higher rations with increasing temperature. An optimum ration rvas not even reached at 20 C. The latter circumstance prevails because fish would not accept a sufficiently high ration at this temperature to surpass the point of maximum conversion. The small size of most yearling sockeye at the time of migration makes it iiighly unlikely tirat supraoptimum food conditions characterize this phase of life so that the use of log K1 as developed by Paloheimo and Dickie' would not be applicable during freshwater growth. NET EFF]CIENCY If the maintenance ration is known a further calculation may be made of the net elficiency (E") by subtracting the fraction of the total ration that is involved in maintenance (\"I), thereby deriving the efficiency of utilization of the fraction of food available for growth. In this instance E, ::=X n|(V".I-M ,. Because Bror'vn (1957) reported that the maintenance ration was difficult to determine directly, since the fish kept adapting their growth rate to com- petrsate for reduced rations, Pandian (1967a) dismissed the value of determining net elficiency. Paloheimo and Dickie (1966b) also considered that the difficulties of obtaining satisfactory maintenance estimates precluded useful application of net eflrciencies. Hou'ever, since it I.ras been shown that where stable growthJ. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2386 JoURNAL FIsHERIES RESEARcH BoARD oF cANADA' vor'.26' No. e' 1e6e RATI0N - groms d.y wl / fiEh / 2 weeks FIc. 15. Relation between logarithm of gross growth effrciency and ration at five different temDeratures. Clircled pointJ weretalculated from interpolated values (Fig. 6). The^efficiency, Kr, was determined in the manner of Paloheimo and Dickii: (1s66b) using ihe daily gain in dry rveight/dailv d.rv ration' Straight line projections beyong the optimum suggest the relatlon determlnecl bv these authors.'Each point was calculated from the total growth during a.10- or 12-week period. Variability within tanks is indicated-as *1 so of the birveekly .K1 values, which did not shorv any consistent trend in relation to increasing ration. Variance increases as K1 approaches zero. rates irave been established the maintenance ration can be determined quite accurately by interpolation (see Fig. 6 and 8; also Warretr and Davis, 1967), the case for net efficiency is worth further consideration. Brody (1945) recounts that net values were used primarily to compare the efifrciencies of farm produce (such as milk and eggs) without the maintenance costs for the different domestic animals being involved. Since it is a matter o{ producing flesh or converting energy in fish that is of concern it might seem reasonable to dismiss the net efficiency. However, it is important wherever possible to partition the energy relations to distinguish the relative contributions when comparing produce, species, rations, ages, or the effects of environmental factors. Since the estimated maintenance portion of the ration is subtracted from the total ration, it is obvious that tl-re net efficiency r'vill ahvays be greater than the gross efficiency. This difference is least at maximum growth rate on a high ration and greatest approaching the maintenance level, i.e., the utilization of that small fraction of the ration in excess of maintenance tends to be very efficient. By using actual and interpolated values from the curves relating growth rate and fation (Fig. 6) a set of isopleths were constructed fof net , I z g a o Y J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION 2381 TEMPERATURE - C Frc, 16. Net efficiency of food conversion in relation to temperature and ration, drawn as in Fig, 14. efficiency (Fig. 16). A maximum net effrciency, just bordering on 4070, occurred between 8 and 10 C for rations of 1.S/o/dav dorvn to 0.8/o/day, the main- tenance level. Minimum efficiency of 15/6 occurred at low temperatures (1-3 C) at all accepted rations, andat 23+ 0.5 C for rations in excess of 4/6/day. The net efficiency isopleths form a pattern opposite to those for gross efficiency, net efficiency decreas'ing with increasing growth rate (cf. Fig. 14 and 16).'Ihe "favourable axis" has a center in the same temperature region (9-10 C) but is rotated anticlockwise in a more vertical position with relation to the temperature axis. The advantage of low temperature for gross efficiency conversion is not present for net efficiency demonstrating that the process of growth, when separated from the interwoven complex of the energy re- quirements and food turnover associated with maintenance, has a narrower, different focal area in relation to temperature. The combined evidence from growth rates and conversion efficiencies indicates that temperatures betweeu 5 and 17 C provide young sockeye lvith the best environmental circumstance for elaborating cells and storing energy' Boov CouposrrroN Many factors have been shown to influence body composition (Love, 1957). In addition to quantity of food and environmental temperature, Parker and Vanstone (1966) reported that young pink salmon (O. gorbuscha) showed changes in composition with size, age, diet, and at particular stages of life (ontogenesis). Slight but significant changes in moisture content even occurred o - I U= * U F I 3o o L o U Iq J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2388 JoURNAL F-ISHERTES RESEARCH BOARD oF CANADA, vor,. 26, No. 9, 1969 diurnally. Various investigators noted that starvation is accompanied by an early decrease in fat followed by gradual protein depletion; loss of body weight is partially offset by an increase in the proportion of water (Idler and Clemens, 1959; Phillips et aI., 1966). At any one stage in life it can be expected that the balance of body constit- uents will be governed by the level of food intake and the rate of expenditure of energy. To a large extent energy expenditure is governed by activity and temperature. Since metabolic rate appears to be ciosely correlated to feeding rate (Paloheimo and Dickie, 1966a), it may be deduced that temperature will act as a naajor independent variable. It is therefore of interest to examine further some of the interrelations between food intake and temperature as tl'rey affect the body composition of young sockeye in an effort to distinguish the respective roles. FAT PROPORTION At all temperatures the most responsive constituent to food intake was fat, which varied from a low of 0.6/6 to a high of 9.5/s (Table 3). If the percentage composition of fat is considered in relation to ration (Fig. 17A) there is a direct proportionality at each level of temperature, with verl'little distinction between the fish held at 20 C and those at 15 C. At lorver temper- RATION Frc. 17. Relation of body fat to daily ration expressed as a fraction of dry body weight (A) and of the maximum food intake (B), showing the relative influence of temoerature. Trend lines for each temperature have been drawn by eye in A,. The solid line in B has been drawn for all points, with broken lines representing the constrained effect of temperature when body fat is related to Dercent maximum food intake. F n= F F I F J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION 2389 atures lrigh fat fractions of about 9/6 occur at lorver levels of daily ration, as might be expected from the reduction in maintenance associated with reduced temperature. A refinement of the above relation may be made by expressing the ration as a fracLion of the maximum daily food intake. Since the latter is exponentially temperature-dependent (see Fig. 8) a large proportion of the temperature influence is removed (Fig. 178). Only at 5 C does there appear to be sufficient difference to warrant separate representation of tle data. It is quite possible that the fish at 5 C had not reached a static balance of their body constituents at each ration, and that in time they would have been inseparable from the rest. Aside from demonstrating the dominant role that ration plays in defining the proportion of body constituents, it is apparent that a simple determination of the fat fraction is likely to reveal at what sort of level the fish is feeding in relation to its capacity, without particular concern for temperatures. As long as some stage of metamorphosis such as srnolt-transformation is not in progress the fat assessment should be of value to the ecologist as a means of assessing the general feeding level of any particular population. Since protein tends to be the most stable component of the gross body constituents it might be thought that the fat/protein ratio (or fat to fat-free dry material) would provide a better indicator of feeding intensity. However, it will be recalled (Fig. 10) that both fat and protein respond in a similar manner to reduced ration so that with the exception of the lowest rations and highest temperatures the fat/protein ratio does not provide a useful index of the relative feeding rate. \4/ATER CONTENT The starved and excess-fed fish provide examples of extremes of water content that can occur for a given species, age, and diet (71-86.97o, Table 3). Love (1957) records values for fish ranging from 53 to 89 . 3/6, but warns that determinations are 11ot strictly comparable because of different analytical techniques. Despite this wide variation (which is temperature dependent on a fired ration), wiren food is present at either the maintenance level or sufficient for maximum growth the body composition was found to be remarkably constant at all temperatures. Under maintenance conditions the water content ranged from 78.3 to 79.77o; on a maximum ration the values ranged from 72.0 to 73 .0/6. This demonstrates the sort of balance that is struck when particular feeding rations are considered. CONCLUDING COMMENTS The prime purpose was to determine how temperature affects the growth and efficiency of food conversion of young sockeye salmon at various levels of feeding intensity. The hypothesis that the optimum temperature for growth would shift to a lower temperature rvith a decrease in ration was supported.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. JOURNAL FISHERIES RESEARCH BOARD OF CANADA, VOL.26, NO.9, 1969 A change in the optimum from 15 to 5 C occurred when the ration lvas reduced from 6 to L|/e/day. The support for the hypothesis depended on maintaining a favorable balance between reduced maintenance costs and efficiency of growth processes (including digestion and transformation) associated with lower temperatures. The fact that a gross conversion efficiency of.20/o occurred at 5 C on a reduced ration lends confirmation to the conclusion that the saving in maintenance costs that low temperature affords is the main contributor to sustained efficiency. This is further resolved and supported by examining net efficiency, which indicates some interaction betr'veen temperature and food transformation in favor of sustained efficiency at reduced temperature, but not to the same extent as is the case for gross efficiency. This suggests that distinct relations exist between temperature and each of maintenance, metabolism, digestion, and transformation. Although not defined with precision, it is apparent from the higlr gross efficiency between 5 and 77 C that the enzymatic process of digestion is relatively temperature-independent. A decrease in the maximum growth rate with age supports the general concept that grorvth rate is size-specific at any given temperature. That this is not entirely so is indicated by the duration of the growth stanzas. Thus, at 15 C on excess ration sockeye shorved a relatively stable growth rate of | .7/o/day from August to January although changing in r,veight from 10 to 80 g. Since a variety of mathematical transformations have been applied to the growth process, it may be that use of the simple exponential equation is inadequate for a consideration of the effect of size. It is apparent from the sockeye data that the temperature optimum becomes broader with age, so that the significance of temperature to growth, in the presence of unrestricted food, is greatest in the very early stages. Additional significance may'be attached to the high efficiency potential that occurs between 5 and 17 C. This temperature range characterizes much of the lacustrine distribution of the species in its early freshwater life. Experimentally, the determination of food-conversion efficiency may well offer one of the greatest sources of insight concerning what governs the success of an organism in nature. The presence of an optimum for growth at 15 C (when fed to excess) coincides with optimum metabolic scope, greatest tolerance to oxygen-debt, and maximum sustained speed (Brett, 1964). It is apparent that a general physiological optimum occurs at this temperature in fresh water. The potential for increasing returns in the culturing of fish is self-evident by the presence of such a wide range of efficiency. By determining the particular combination of temperature and ration that produces naximum growth a gross efficiency of 25/6 is predictable. Finally, the insight gained by conducting growth X ration X temperature experiments, accompanied by determinations of body constituents, obviously offers a great deal to the general problem of interpreting ecological relations.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COI\IPOSITION 2391 ACKNOWI,EDGMENTS Messrs D. B. Sutherland and G. D. Heritage have assisted by providing the careful and continuous attention necessitated by these iong-term experiments. Their willing support and patience have done much to make the program possible. As u'ell as contributing a great deal of helpful advice, Dr 'I'. D. D. Groves and associates conducted over half the proximate analyses. It is a pleasure to acknowledge this assistance and the cooperative interest of the University of Victoria through its Department of Biochemistry under the direction of Dr A. T. Wood. REFERENCES Armnorcr, D. F., J. R. Bnnm, AND D. B. SurnBnr,-qNn. 1966. Design of small holding tank for fish. J. Fish. Res. Bd. Canada 23: 1447-1450. AlrBnrcaN Socrnrv FoR TESTTNG Memruers. 1966. N{ethod D.277, f.or Parr bomb calorim- eter. Am" Soc. Testing Materials 79: 42-46. ANoN. 1959. The Heinz handbook of nutrition. XrlcGraw-Hill Book Co., New York. 439 p. AssocrerroN oF OFFTcTAL Axlryrrc.e.r- Crrpursrs. 1960. Methods of analvsis. 9th ed. Assoc. Offic. Anal. Chemists, Washington, D.C. Brrowrx, N. S. 1957. Food consumption and growth of brook trout at different temperatures. Trans. Am. Fish. Soc. 86:. 323-328. BnntlreNrnv, L. vox. 1957. Quantitative laws in metabolism and growth. Quart. Rev. Biol. 32: 217-231. Brecr, E. C. 1958. Energy stores and metabolism in relation to muscular activity tn !s!res'p.51-67. In P. A. La;kin [ed.] The investigation of fish-power problems. H. R. Mac- Millan lectures in fisheries, Univ. of British Columbia, Vancouver, B.C. Blexrrn, K. L. [ed.] 1965. Energy metabolism. Academic Press Inc., New York. 450 p' Bnnrr, J. R. 1952. Temperature tolerance in young Pacific salmon, genus Oncorhynchus. J. Fish. [tes. Bd. Canada 9:265-323. 1964. The respiratory metabolism and swimming performance of young sockeye salmon. J. Fish. Res. Bd. Canada 21:1183-1226.- 1965. The relation of size to rate of oxygen consumption and sustained swimming speed of sockeye salmon (Oncorhynchus nerka). - J. fisn. Re;. Bd. Canada 22: l49I-I501. 1967. - Swimming perfoimance of sockeye salmon (Oncorhynchus nerka) in relation to fatigue time and teniperature. J. Fish. Res. Bd. Canada 24: 1731-1741. Bnoov, S. 1945. Bioenergetics and growth. Reinhold Publishing Corp., New York. 1023 p. Bnowx, M. E. 19+6. The growth of brown trout (Salmo trutta Linn.). IL Growth of two- year-old trout at a constant temperature of 11.5"C. J. Exptl. R\L22: 130-14+.- 1957. Experimental studies of growth, p. 361-400. In M. E. Brown [ed.] Physiology of fishes, Vol. 1. Academic Press, Inc., New York. Dnvres, P. M. C. 1963. Food input and energy extraction efficiency in Carass'ius auratus. Nature 4881: 707. D.nvrs, G. E., exo C. E. WlnnBN. 1968. Estimation of food consumption rates' p" 204-225. InW. E. Ricker [ed.] Methods for assessment of fish production in fresh waters. Intern. Biol. Programme Handbook 3. Blackwell Scientific Publications, Oxford and Edinburgh. Doner,osoll, L. R., ewo F. J. Fosren. 1940. Experimental study of the effect of various water temperatures on the growth, food utilization, and mortality rates of fingerling sockeye salmon. Trans. Am. Fish. Soc. 70: 339-346. Frsnrn, K. C. 1958. An approach to the organ and cellular physiology of adaptatio_n to tem- pe;ature in fish and smali mammals, p. 3-35. In C.L. Pros3ei led.l Physiological adaptation. Am. Physiol. Soc., Washington, D.C. Forcn, J., M. Lee, AND G. H. S. Suxrev. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226: +97-509.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2392 JOURNAL FISHERIES RESEARCH BOARD OF CANADA, VOL, 26, NO.9, 1969 Fnv, F. E. J. 1947. Effects of the environment on animal activity. Urriv. Toronto Studies, Biol. Ser. 55, Publ. Ont. Fish. Res. Lab. 68:1-62. Grnr<rNc, S. D. 1962. Production and food utilization in a population of biuegill sunlish. Ecol. Monographs 32: 31-78. Harvrn, J. E., er.rn W. E. Sn-q,Nr<s. 1960. Nutrition of salmonoid fishes. VIII. Indispensable amino acids for sockeye salmon. J. Nutr. 72: 340 346. Hl5r<er,l, D. C., L. E. \Vorr, AND L. BoucHARD. 1956. The effect of ternperature on the growth of brook trout. Nerv York Fish Game J. 3: 108-113. Het-rNaxa, A., eNo N{ase,o TerenesRl. 1960. Studies on the amounts of the anchovy consumed by the mackerel. Tohoku J. Agr. Res. 11: 83-100. Hoen, W. S. 1966. General and comparative physiology. Prentice-Hall, Inc., New Jersey. 6IJ D. Hoa"oa"*o, P. W., eNn A. C. SrNcl,lrn. 1962. Glycogen stores in trout tissues before and after stream planting. J. Fish. Res. Bd. Canada 19: 127-136. Hunns, C. L., exo C. Huans. 1953. An improved graphical analysis and comparison of series of samples. Systematic Zool. 2: 49 56. Inren, D. R., ewr W. A. CrnueNs. 1959. The energy expenditures of Fraser Rivcr sockeye salmon during the spawning migration to Chilko and Stuart Lakes. Intern. Pacific Salmon Fish. Comm. Progr. Rept. 80 p. IvLE\', V. S. 1945. [The biological productivity of waters.] Usp. Sovrem. Biol. 19:98-100. (In Russian; 'Iransl. in j. Fish. Res. Bd. Canada 23:1727-1'759.) JoNxrNcs, J. B. 1965. Feeding, digestion and assimilation in animals. Pergamon Press, London and New York. 228 p. Jouxsow, W. E. 1965. On mechanisms of self-regulation of population abundance in Oncor' hynchus nerka. Mitt. Intern. Ver. Limnol. 13:66-87. KrvnnN, N. R. 1966. Feeding rate of carp estimated by a radioisotopic method. 'frans. Am. Fish. Soc. 95: 363-371. Krwxn, O. 1960. Growth, food intake, and food conversion in a euryplastic fish exposed to different temperatures and salinities. Physiol. Zool. 33 288-317. KLBrsBn, l\4. 1961. The fire of life. An introduction to animeLl energetics. John Wiley & Sons, Inc., New York. 454 p. l.ove, R. M. 1957. The biochemical composition of fish, p.401 418. InM.E. Brown [ed.] Physiology of fishes, Vol. 1. Academic Press, Inc., New York. MacNusox, J. J. 1962. An analysis of aggressive behavior, growth and competition for_ food and space in medaka, Oryzias Lat'ipes (Pisces, Cyprinodontidae). Canadian J. Zoo1.40; 313-363. Nenven, D. W. MS, 1967. Diel vertical movement of pelagial sockeye salmon juveniles. Fish. Res. Bd. Canada, MS Rept. Ser. 949: 24-28. Nernnlu, A. E. t964. The growth process in animals, Sir Isaac Pitnam and Sons Ltcl., London, 522 p. Per-ouuuo, J. E., eNo L. M. Drcxrn. 1966a. Food and growth of fishes. II. Effects of food and tem-peraiures on the relation between metabolisi and body weight. J. Fish. Res. Bd. Canada 23:869-908. 1966b. Food and growth of fishes. III. Relations among food, body size, and growth efficiency. J. Fish. Res. Bd. Canada 23:1209-1248. PrNoreN, T. J. 1967a. Intake, digestion, absorption and conversion of food in the fishes Megalops cyprionides and O1'h,iocephalus slr'iatus. Marine Biol. 1: 16-32. tD67b. Transformationbf foo-d in the fish Megalops cypr'ionides. I. Influence of quality of food. Marine Biol. 1: 60 64. Penr<nn, R., -nxn W. E. VeNsroNr. 1966. Changes in chemical composition of central British Columbia pink salmon during early sea life. J. Fish. Res. Bd. Canada 23:1353-1384.J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. BRETT et al.: SOCKEYE GROWTH AND COMPOSITION 2393 Prwrnrow, F. T. K. 1939. The relation between growth and food consumption in the brown trout (Salmo lrutta). J. Exptl. Biol. 16: 446-473. Pnrrurs, A. M., Jn., D. L. LtvrncsroN, AND H. A. PosroN. 1966. The effect of changes.in protein qualitv, calorie sources and calorie levels upon the growth and chemical composition of brook'trouf. N. Y. State Dept. Conserv. trish.- Res. BuIl. 29l- 6-7. Sr'rrr, H. 1967. Influence of temperature on the rate of gastric juice secretion in the brown bullhead (Ictol,urus n'ebulosus). Comp. Biochem. Physiol. 2l: 125-132. Sruww, K. 196l. Growth of largemouth bass at various temperatures. Trans. Am' Fish. Soc. 90: 334-335. Srvrrr, D. R. 1955. Seasonal variations in the growth rate, thyroid gland activity and food reserves of brown trott (Sal,mo truttaLi'nn.). J. Exptl. Biol. 32: 751-76+. 1964. The effect of temperature and oxygen bn the growth rate of the Windermere char, Salaelinus alpinus (Willuglbii). Comp. Biochem. Physiol. 12:- 179-183. T.e.vror., C. 1962. Growth equations with metabolic parameters. J. Conseil, Conseil Perm. Intern. Exploration l,Ier 27 : 270-286. TnoupsoN, D. H. 194I. The fish production of inland streams and lakes. Symp. Hydrobiol'' Univ. Wisconsin Press, Madison, Wis. p. 206-217. Vryocuoov, A. P. 1953. The elementary chemical composition of marine organisms. Efron and Setlow ftranslators]. Yale Univ. Press, New Haven, Conn. p. 463-566. Wennry, C. E., eNo G. E. Davrs. 1967. I-aboratory studies on the feeding,-bio-ene.rgetics and growth of fish, p. 175-214. InS. D. Gerking [_ed-] The biological basis for freshwater fish production. Blackwell Scientific Publications, Oxford. Wnsr, B. W. 1966. Growth rates at various temperatures of the orange-throat darter Etheostoma spectabi!'is. Arkansas Acad. Sci. Proc. 20: 50-53. WrNeonc, G. G. 1956. Rate of metaboiism and food requirements of fishes. Belorussian State Univ., Minsk. 251 p. (Fish. Res. Bd. Canada Transl. Ser. 194') (Appendices A-B follow)J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only. 2394 JOURNAL FISHERIES RESEARCH BOARD OF CANADA, VOL.26, NO. 9, 1969 APPENDIX A. INcnnorrNrs eNo PnncrNreco CotrrpostuoN By Wrrcnr or DrBr UsBo rN SocxovB Gnowrn ExppnIuBms, 1964-67. Water Drained, canned salmon Clark's pellet feed Fresh beef liver Yeast Pablum mixed cereal Codliver oil Vitamin pack Gelatin Iodized salt % 37.2 29.7 11.1 10.0 J./ 2.2 2.2 1..) t.J 0.9 (J. R. Clark & Co., Salt Lake City, Utah) (Edward Dalton Co., Toronto, Ont.) (Nutritional Biochemicals Corp., Cleveland, Ohio) APPENDIX B. Pnocrnunr FoR DRy ExrnecrroN oF FAT6 Frozen fish were sliced longitudinally and their abdominal cavities cut open. The fish were then placed on tared aluminum foil trays and dried to constant v/eight in a forced air oven at 80 C. The fish were not left in the oven any longer than necessary in order to minimize fat oxidation. The dried fish were crushed and transferred, along with the aluminum foil, to a 500-ml Erlenmeyer fitted with an aluminum foil-rvrapped stopper. The material in the flask was extracted with 20Oml of z-hexane- The foil was then.e-o.'ed a.rd the sample was further extracted with three successive 100-ml portions of hexane. Each extraction proieeded luith occasional shaking for 24hr. After each extiaction the hexane was poured off thiough a small tared filter paper to prevent loss of susoended oarticulate material. Final traces of hexane were removed from the fish in a vacuum oven at 4b C. Subsequent ash (predominantly as oxides) and total protein determinations carried out on the fat-extracted material accounted for 99.57o of the sample. Vt._.1y";181 used by Dr T. D. D. Groves, Department of Biochemistry, University of Victoria,J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from nrcresearchpress.com by CSP Staff on 12/18/12For personal use only. J. Fish. Res. Bd. Can. Downloaded from www.nrcresearchpress.com by ARLIS - Alaska Resources Lib & Info Services on 09/23/13For personal use only.