Transactions of the American Fisheries Society 126:259-272, 1997© Copyright by the American Fisheries Society 1997

The Metabolic Costs and Physiological Consequences toJuvenile Rainbow Trout of a Simulated Summer Warming

Scenario in the Presence and Absence of Sublethal AmmoniaTYLER K. LINTON, SCOTT D. REID,' CHRIS M. WOOD

Department of Biology, McMaster UniversityHamilton, Ontario L8S 4K1, Canada

Abstract.—Quantitative bioenergetic and physiological measurements were made on juvenilerainbow trout Oncorhynchus mykiss exposed over summer (June-September 1993) to a simulatedsummer warming scenario of +2°C in the presence and absence of 70 fimol total ammonia/L(nominal; equivalent to 0.013 mg NH^-N/L at 15°C, pH = 7.6) to determine the metabolic costsand physiological consequences associated with their growth in a warmer, more polluted envi-ronment. With unlimited food, fish exposed to +2°C show better energy conversion efficiency andincreased nitrogen retention at a metabolic cost equivalent to the base temperature group. Metabolicfuel use appears to have been optimized to support the bioenergetic demands imposed duringmaximum summer water temperatures. Low-level ammonia enhances nitrogen and energy con-version efficiency by stimulating protein retention, which ultimately results in the most cost-effective growth. However, in the +2°C ammonia treatment, the stimulatory effect of low-levelammonia is lost during mid to late summer due to the greater energy demands when fish are forcedto cope with the additional stress of a small further increase in temperature.

Temperature is the single most important envi-ronmental factor affecting animal activity, and theeffects of temperature change are especially pro-nounced for aquatic ectotherms such as fish. Asidefrom natural diel and seasonal elevations in watertemperature, fish face increased water tempera-tures from the discharge of cooling water frompower plants, water diversion for crop irrigation,and canopy removal from forest clear-cutting(Sprague 1985; Thomas et al. 1986). Water tem-peratures are also likely to rise in association withglobal warming. Simulation modeling suggeststhat a doubling of the atmospheric CO2 concen-tration will increase the Earth's average tempera-ture by 1.5 to 4.5°C in the next 50 years (Hansenet al. 1988). Schindler et al. (1990) have alreadyreported a 2°C increase in mean air and lake tem-peratures over the past 20 years in northern On-tario, Canada. Despite a large body of literatureaddressing the effects of elevated temperature onfish, large uncertainties in the magnitude and tim-ing of potential warming impacts still exist (Rubinet al. 1992). To date relatively few studies haveaddressed the effects of chronic small temperatureelevations over the annual cycle on fish.

The metabolic and physiological effects on fishof small temperature elevations above present nat-ural fluctuations in water temperatures could be

1 Present address: Department of Biology, OkanaganUniversity College, 3333 College Way, Kelowna, BritishColumbia V I V 1V7, Canada.

large. Many species inhabit waters encompassinga wide range of temperatures, but each typicallyhas a narrow range at which growth and survivalare maximized (Fry 1967; Magnuson et al. 1979;Elliot 1982; Trippel et al. 1991). Even smallchanges in local temperature may alter fishes' en-ergetic requirements (Dawson 1992; Mehner andWeiser 1994).

The goal of the present paper is to quantify themetabolic costs and physiological consequences tojuvenile trout associated with living in warmer,marginally polluted environments. The thermaloptimum for growth and food utilization in rain-bow trout Oncorhynchus mykiss is ordinarily 15°C(Cho and Kaushik 1990), and the species' upperlethal limit is approximately 26°C (Bidgood andBerst 1969; Kaya 1978). Reid et al. (1995) com-pared juvenile rainbow trout held in hard and softwater during the summer of 1993 to fish held 2°Cabove the natural water temperature cycle (char-acteristic of the inshore region of Lake Ontario)and found an average 20% reduction in growth,appetite, gross conversion efficiency and proteinturnover during peak summer water temperatures(20-24°C). In this paper, we address the energeticand physiological effects of the above thermal re-gime as well as the effect of a low, but environ-mentally realistic, level of ammonia (a 70-|xmol/Ldilution of total ammonia [NH^ + NH^], equiv-alent to 0.013 mg un-ionized ammonia/L at 15°Cand pH 7.6). The thermal and elevated ammoniaregimes are assessed alone and in combination,

259

260 LINTON ET AL.

which requires use of some data (growth, appetite,and conversion efficiency for the ammonia-freetreatments only) originally reported by Reid et al.(1995).

Ammonia is an important pollutant in aquaticenvironments. It is highly toxic to fish and ubiq-uitous in surface waters (Thurston and Russo1983; Erickson 1985; Russo 1985). More mole-cules of ammonia are manufactured each year thanany other industrial chemical (Atkins 1987). Am-monia also is a problem in fish culture where high-density stocking may result in an accumulation ofammonia via excretion from the fish. Larmoyeuxand Piper (1973) found that growth of juvenilerainbow trout raised in hatchery water-reuse sys-tems was significantly reduced in tanks withun-ionized ammonia loads greater than 0.01 mg/L.The lowest lethal concentration of un-ionized am-monia found for salmonids is 0.083 mg/L (Thurs-ton et al. 1984), but sublethal effects have beenreported at concentrations even lower than 0.01mg/L (Rice and Bailey 1980; EPA 1985). This lowchronic value is often exceeded in many urban andindustrial areas including Hamilton Harbor (LakeOntario) and its tributaries, where summer totalammonia concentrations have been recorded ashigh as 66 jimol/L at pH = 7.3, equivalent to 0.012mg/L un-ionized ammonia (B. Crosbie, McMasterUniversity, personal communication).

We used several physiological indices includingfood intake, growth, metabolic rate, nitrogen bal-ance, and proximate composition to assess thelong-term consequences and metabolic costs forjuvenile trout exposed to a small increase in tem-perature and an environmentally relevant concen-tration of priority pollutant, alone and in combi-nation. We asked three questions: (1) What phys-iological changes occur in juvenile rainbow troutgrowing under the normal, fluctuating summerthermal regime? (2) What is the effect of +2°Csuperimposed on this regime? (3) What is the ef-fect of 70 iJimol ammonia/L on both of the above?Our data suggest that additional temperature (2°C)and ammonia (70 (jimol/L) are beneficial up to athreshold temperature above which loss of appetiteand increased maintenance costs greatly reducegrowth and nitrogen transformation efficiencies.

MethodsAnimals.—Juvenile rainbow trout (2-5 g) were

obtained from Rainbow Springs Trout Farm,Thamesford, Ontario, on 19 April 1993 and al-lowed to acclimatize for 6 weeks prior to testing.The trout were held in a 600 L aerated polyeth-

-IjiII

25

20

15

1030 60 90

Time (days)FIGURE 1.—Daily water temperature profile as mea-

sured each day over 90-d exposures from 17 June to 15September 1993. Juvenile rainbow trout were exposedto either ambient laboratory water temperatures (base,solid line) or to ambient water temperatures + 2°C(base + AT, dashed line).

ylene tank receiving 2.5-L/min flow of dechlori-nated Hamilton tap water ([Ca2+] = 0.98 ±0.11mmol/L, [Na+] = 0.56 ± 0.03 mmol/L, [C1-] =0.73 mmol/L; pH = 7.57 ± 0.26) at an ambientwater temperature of 12 ± 1°C. They were fed amaintenance ration equivalent to 1% body weightper day (wet basis) of Zeigler's Trout Starter num-ber 3 (50% protein, 15% lipid, 12% moisture) andkept under the natural photoperiod for Hamilton,Ontario, throughout the experiment.

Experiments.—Groups of approximately 150rainbow frout were randomly distributed to eightexposure tanks (270 L) representing four treatmentconditions (N = 300 fish per treatment). The tanksreceived 2.5-L/min flows (95% replacement time,4.2 h) of either ambient temperature water (base= City of Hamilton tap water taken from the in-shore region of Lake Ontario) or ambient temper-ature water plus 2°C (base+AT), each with or with-out an additional 70 fxmol total ammonia/L (TAmm= NH4+ + NH3; base+Am and base+AT+Am,respectively). Water temperatures of base -I- 2°Cwere achieved with a heat exchanger. The actualmean difference between temperature treatmentswas 1.90 ± 0.03°C (Figure 1). The desired TAmmconcentration was achieved by delivering the re-quired amount of (NH4)2SO4 stock solution viamariotte bottles (Mount and Brungs 1967). Themean concentrations of total ammonia in the waterof each tank along with the pH and partial pressureof oxygen are listed in Table 1.

During the experiment, the fish were fed to sa-

METABOLIC COST OF HIGH TEMPERATURE AND AMMONIA 261

TABLE 1.—Mean (±SE) ammonia concentrations, pH, and partial oxygen pressures in tanks during 90-d exposuresof rainbow trout to +2°C (AT) and 70 p,mol ammonia/L (Am).

Total ammonia

Treatment

Base + Am

Base

Base + AT

Base + AT 4- Am

Repli-cate

12I212

12

Tank mean(ixmol/L)

65.8 r 6.358.2 ± 5.66.2 ± 1.65.8 ± 0.94.5 ± 1.07.1 ± 1.5

78.6 ± 7.672.8 ± 6,6

N

1010

98

99

1010

Treatmentmean (p.mol/L)

61.5 ±3.8

6.0 ± 0.2

5.8 ± 1.3

75.7 ± 2.9

PH

Mean N

7.67.6

7.67.6

7.6

0.0.

0.0.

0.7.5 0.

7.5 ± 0.7.6 ± 0.

1414

1413

14131414

Po2

Mean (torr)

131.9 ±4.5134.3 ±4.1132.9 ±3.5134.6 ± 3.9

132.2 ±3.5124.3 ± 4.7

121.3 ±4.1123.3 ±4.3

N

1414

151515151515

tiation twice daily (at 0830 and 1630 hours) frompreweighed bags of Zeigler Trout Starter so wecould monitor appetite, following the methods ofWilson et al. (1994a). Samples of fish from eachexposure were taken immediately after the test be-gan and approximately every 30 d thereafter.

Each sampling period began by measuring rou-tine in-tank oxygen (02) consumption and nitrog-enous (N) waste excretion (T^mm + urea) rates inone tank per treatment simultaneously. Fish werefed at their regular times throughout the samplingperiod. To measure C>2 consumption, samples weretaken every other hour over an 8-h period begin-ning at 0700 and ending at 1500 hours. The tankwater and air supplies were closed, the water sur-face was sealed with a transparent lid fitted snuglyto the tank walls, and magnetic drive pumps (LittleGiant, 1 -EUAA-MD) were activated to recirculateand mix the water. Water samples were taken every10-20 min for a total of 30 min to 1 h. The partialpressure of C>2 (Pc>2) of the water samples wasmeasured with a temperature-equilibrated Cam-eron C>2 electrode connected to a Cameron OM200 oxygen meter. After the final PO2 measure-ment, the water and air supplies were re-openedfor 1 h prior to the next sampling period to allowflushing and re-aeration. The C>2 consumption ratesfor each closed sample period were calculatedfrom the mean decline in PO2 over time (10-20min readings) and the water oxygen solubility co-efficients of Boutilier et al. (1984). The Po2 of thewater in any one tank was not allowed to dropbelow 90 torr. Mean O? consumption rates wereplotted for each closed cycle and the area underthe curve determined with a digitizing tablet (Sig-ma Scan) to give a single integrated C>2 consump-tion rate representative of the 8-h interval between0700 and 1500 hours.

Nitrogenous waste excretion was measured the

next day. Tank flow and aeration were maintained.Approximately 50 mL of influent and effluent wa-ter were collected simultaneously every 4 h for 24h, and the samples were frozen immediately forfuture analysis of total ammonia-N and urea-N.The area under the curve was determined as above.All O2 consumption and N-waste excretion datawere corrected for differences in fish size with theweight exponent 0.824 determined for rainbowtrout by Cho (1990). The contribution of bacteriato in-tank routine Ch consumption and N-wasteexcretion was not estimated during the experiment;however, an estimate ("worst case scenario") wasobtained following the experiment. Then, 25-grainbow trout (the approximate size of the fish after60 d of exposure) held at 21°C (equivalent to thehighest temperature reached in the ambient tem-perature treatments over summer) were removedfrom their tank (270 L) just after they had fed tosatiation, and the rates of bacterial 02 and TAmmconsumption were measured over 24 h. The con-tribution of bacterial respiration to C>2 consump-tion was determined to be no greater than 5%. Theinfluence of bacteria on TAmm consumption wasnegligible.

Following the in-tank measurements of routineO2 consumption and N-waste excretion, food waswithheld from the fish for the next 48 h, duringwhich blood and tissue sampling were conducted.Ten fish from each tank were killed for whole-bodyproximate analysis (protein, lipid, carbohydrate,and ash) and blood ions (plasma TAmm and so-dium) and another 10 were sacrificed for the es-timation of TAmm, urea, and glutamine in whitemuscle and liver. The fish were sampled by nettinga number of fish from which one was randomlytaken; the procedure was repeated 10 times at5-min intervals. With this approach, there was noeffect of sample order on any of the measured

262 LINTON ET AL.

variables. For whole-body composition and bloodsampling, the fish were killed by a quick blow tothe head, blotted dry, and measured for wet weightand total length. A terminal blood sample was col-lected via caudal severance, the plasma was sep-arated, and the carcass was freeze-clamped withaluminum tongs precooled in liquid N2. Both plas-ma and carcass were stored at — 70°C until furtheranalysis.

The fish sampled for TAmm» urea, and glutaminein white muscle and liver were handled in a similarmanner, but were sacrificed in a bucket of watercontaining an overdose of tricaine (MS-222, 1 g/L)buffered with NaHCO3 (2 g/L) to avoid TAmmbuild-up in muscle due to struggling (Wang et al.1994). The liver and a small portion of white mus-cle anterior to the dorsal fin and above the lateralline were excised and freeze-clamped immediatelyin liquid N2. Each tissue was individually wrappedin aluminum foil and stored at — 70°C for furtheranalysis.

Analysis.—Ammonia concentrations were de-termined in water by the salicylate-hypochloritemethod of Verdouw et al. (1978) and in plasma bya commercial enzymatic kit (Sigma 170-UV). Thetwo assays were cross-validated. Water urea sam-ples were freeze-concentrated 5-fold by lyophili-zation before assay by the modification of the di-acetyl monoxime method described by Lauff andWood (1996). The concentration of Na+ in waterand plasma was determined by atomic absorptionspectrophotometry (Varian AA 1275).

For the determination of proximate composi-tion, frozen whole bodies kept at -70°C wereground into a fine powder with a grinding mill(IKA-M10/M20) cooled to -40°C by a methanol-dry ice mixture, and then the powder was lyoph-ilized for 72 h at -55°C. A small portion of thefrozen ground tissue was withheld and dried in anoven at 80°C for 48 h to obtain the water content.The protein content of the lyophilized tissue wasmeasured by the Lowry method as modified byMiller (1959); glucose, glycogen, and lactate (car-bohydrate) were determined as described by Berg-meyer (1985); and lipid, after extraction in a 2:1chloroform:methanol mixture, was measured asdescribed by Folch et al. (1957).

Frozen white muscle and liver were ground intoa fine powder in an insulated mortar cooled withliquid N2. Subsamples (100 mg) were deprotein-ized in 1 mL of 8% perchloric acid (PCA) andmeasured for ammonia as described by Kun andKearney (1971). Similar deproteinizing procedu-res were used to measure white muscle and liver

urea as described by Crocker (1967). Conversely,for glutamine measurement we used approximate-ly 50 mg of frozen white muscle and liver in 250uJL of 3% PCA and assayed as described by Berg-meyer (1985).

Calculations. — Daily food intake (g) was deter-mined as the difference in food bag weight at thebeginning and end of feeding divided by the num-ber of fish in the tank (grams per fish per day).Cumulative food intake was calculated as the sumof the daily food intakes over each 30-d period.

Specific growth rates (SGR, %/d) were deter-mined for 30 fish removed from each tank formonthly sampling (N = 60 per treatment) and cal-culated by the standard formula

SGR = 100(logey2 - logfXO/te ~ *i),where Y\ and YI are tne mean wet weights of fishat times t\ and r2- Condition factors were deter-mined as the quotient of the wet weight of the fishand its total length cubed, multiplied by 100.

The nitrogen quotient (NQ), or the extent ofaerobic protein catabolism, was calculated as theratio of moles of N produced to moles of O2 con-sumed (Kutty 1972). The proportion of oxygenconsumption dedicated to protein catabolism wasmeasured as the ratio of the NQ at each respectivetime period to the maximum aerobic value (0.27;Kutty 1972).

The nitrogen budget followed the equation sum-marized by Birkett (1969):

where / is the total N consumed, FN is the N lostin fecal and unaccounted material (mucus, etc.), Ais the N absorbed from food, R is the nitrogenretained in body materials, and E is the nitrogenlost via branchial and urinary excretion. Measure-ments of N consumption, retention, and excretionallowed for estimation of N lost through feces (andunaccounted N) and the N absorbed. Protein wasmeasured as a surrogate for N; the percentage Nby weight in the protein of the food consumed andin the protein of the fish was taken as the standardvalue, 16% (Jobling 1980; Soderberg 1995).

The energy budget was derived from the equa-tion

C = Uwhere C is the total energy consumed from food,FE is the energy lost in feces, U is the energy lostvia excretion (branchial and urinary), AB is theenergy stored in body materials, and /?tol is thetotal metabolic energy lost as heat. The following

MKTABOLIC COST OF HIGH TEMPERATURE AND A M M O N I A 263

energy values were assumed: 23.6, 39.5, and 17.2kJ/g for protein, lipid, and carbohydrate, respec-tively (Braefield and Llewellyn 1982), 24.9 kJ/gfor nonfecal nitrogen (Cho and Kaushik 1990), and13.6 kJ/g for O2 (Elliott and Davison 1975).

A new and relatively simple index, the N-costindex (the total moles of C>2 consumed per moleof nitrogen stored), was used to compare the totalmetabolic expenditure associated with the incor-poration of nitrogen into body material under thestated conditions.

Statistics.—The values for the nitrogen and en-ergy budgets arc reported on a per-fish basis fromthe one experimental tank in each treatment whereO2 consumption and N-waste excretion were mea-sured. Because N = 1 for these parameters, nostatistical comparisons could be employed. In ad-dition, cumulative food consumption was calcu-lated on a per-fish basis. Here, N = 2 (two tanksper treatment), and again, statistics were not em-ployed. All other data are expressed as means ±1 SE from individual samples pooled together fromthe two tanks per treatment. One-way analyses ofvariance with SAS JMP (SAS Institute Inc., Ver-sion 2.0.5) followed by Tukey-Kramer honestlysignificant difference multiple-means comparisontests were used to distinguish statistically signif-icant differences among the four treatment groupswithin each sample period (30, 60, and 90 d, re-spectively). The level of statistical significance forall analyses was P < 0.05.

Results and DiscussionPhysiological Responses Due to the BaseSummer Thermal Regime

Juvenile rainbow trout (2-5 g) exposed to thenatural, or base, thermal regime and fed to satia-tion twice a day for 90 d exhibited an overall spe-cific growth rate of 3.02%/d. Despite several daysof fluctuating water temperature about a mean of20.5°C between days 60 and 90 (Figure 1), our fishcontinued to feed and grow with no apparent lossin appetite (Figure 2). Thus, maximum energy in-take was greater than the maintenance energy usedduring this period. In contrast, Elliott (1982)showed that brown trout Salmo trutta do not growat water temperatures above 19.5°C when fed max-imum rations of Tubifex sp.

The metabolic cost associated with growth ofour trout was high. Routine 62 consumption bythese fish (Figure 3A), after size correction (seemethods), was approximately 75% of the maxi-mum size-corrected O? consumption determined

oo

E

3

60

50

40

30

20

10

0

25

20

15

105

E23 dark hatched = base+Am• solid = baseCD open = base+ATEZ light hatched = base+AT+Am

Initial 30 60 90

0-30 30-60 60-90

Time (days)FIGURE 2.—Effects of +2°C (AT) and 70 p.mol

ammonia/L (Am) on (A) growth and (B) cumulative foodintake of juvenile rainbow trout fed to satiation twicedaily. Growth was measured as the wet weight of thefish sampled initially (N = 20) and after 30, 60. and 90d of exposure (N = 60 per treatment per time period).Cumulative food intake is presented on a per-fish basisand represents the cumulative food eaten over the re-spective time periods; each value represents the meanof the two replicate tanks per treatment (N = 2). Col-umns with the same letter on the same sampling day arenot significantly different (P > 0.05).

for juvenile rainbow trout held at 15°C in our lab-oratory (Wilson et al. I994b). These high valueswere most likely associated with the high protein(50%) and lipid (15%) contents of their diet andwith the intensive feeding regime (twice daily tosatiation). Soofiani and Hawkins (1982) showedthat for Atlantic Cod Gadus morhua, elevation inmetabolic rate associated with feeding depends onration size, increasing linearly as food intake in-creases. They also showed that the costs of food

264 LINTON ET AL.

0.3

0.2

0.1

0.0

82 dark hatched = base+Am• solid-baseCD open • base+ATG3 Rght hatched = base+AT+Am

Initial 30 60 90

Time (days)FIGURE 3.—Effects of +2°C (AT) and 70 \Lmo\

ammonia/L (Am) on (A) routine in-tank oxygen con-sumption (N = 5 measurements on the same tank), (B)routine in-lank N-waste excretion (ammonia -I- urea N;N = 6 measurements on the same tank), and (C) thenitrogen quotients of juvenile rainbow trout fed to sa-tiation twice daily. In (A) and (B) the data have beenscaled for fish weight. Measurements were made on onetank of individuals (N = 1) per treatment. Error barsrepresent measurement SE (of the mean 02 consumptionand N-waste excretion rates calculated over the respec-tive sample intervals) and cannot be used for statisticalcomparisons between treatments.

assimilation were highest for fish fed to satiationand that the cost increases with increasing tem-perature.

The type of food and feeding regime also haveconsiderable effect on the rate of N-waste excre-tion in fish. Wood (1995), in his review of theroutes and rates of excretion in salmonids, con-cluded that the single most important factor af-fecting N-waste excretion is feeding. He reportedthat for rainbow trout fed to satiation, total N ex-cretion rate generally ranges from 1.0 to 2.0junol'g"1-!!"1 (0.6 to 1.4 when scaled for fish size),but pointed out from Fromm's (1963) data that itmay drop to 0.5-1.0 junol-g-'-h-1 (0.07-0.35) af-ter 5 d of starvation. The scaled N-waste excretionrates in our satiation-fed fish dropped from 1.5fjimol-g'^h'1 initially to 0.8 jjimol-g'^h'1 by day90 (Figure 3B). Because appetite was not inhibitedin this study, the decline in N-waste excretion ratemay be the consequence of the metabolic parti-tioning of food energy into growth. For example,our results are consistent with the responses ofother salmonids fed high-lipid diets; increased lip-id in the diet acts to spare protein for growth (Ath-erton and Aitken 1970), thus reducing the amountof exogenous N-waste excretion (Jayaram andBeamish 1992; Agradi et al. 1995) and ultimatelyincreasing the proportion of ingested N retainedas growth (Beamish and Tandler 1990). This wouldexplain the observed decrease in protein catabo-lism. Our fish showed 39-55% decreases in aer-obic protein catabolism (NQ) between the initialvalue and those on days 30, 60, and 90 (Figure3C), hence comparable reductions in the use ofprotein for energy.

Rainbow trout under base conditions showedlarge increases of whole-body lipid from initialvalues, indicating an excess of dietary energy be-yond their needs for maintenance and metabolism,though lipid values declined slightly during highsummer temperatures after day 60 (Table 2; Figure1). During the 90-d exposure period, condition fac-tors increased from 0.87 to 1.22 in accord withthis observed increase in whole-body lipid level(Table 2). It is believed that deposition of energyinto lipid may be one of the most critical com-ponents of fish survival (Adams and Breck 1990).Swift (1955) found that fish typically accumulatelipid in the summer and fall and use lipid in thewinter when food consumption is minimal. Con-versely, the percentage of whole-body carbohy-drate (very small in comparison to protein andlipid) showed the opposite trend (Table 2). Car-bohydrate did not change through day 30, but by

METABOLIC COST OF HIGH TEMPERATURE AND AMMONIA 265

TABLE 2.—Proximate chemical composition and condition factors of juvenile rainbow trout fed to satiation andexposed to a +2°C warming scenario (AT) and 70 jimol ammonia/L (Am). Fish were sampled initially and after 30,60, and 90 d of exposure. Results are expressed as means ± SE, based on wet weight. For each sampling day separately,values within a column without a letter in common are significantly different (P ̂ 0.05). Sample sizes are in parentheses.

Sampling dayand treatment

Initial

Day 30Base + Am

Base

Base + AT

Base + AT + Am

Day 60Base + Am

Base

Base-*- AT

Base + AT + Am

Day 90Base + Am

Base

Base + AT

Base+AT+Am

Protein(mg/100 mg)

10.7 ± 0.3(9)

14.3 ± 0.2 z(20)

13.1 ±0 .2y(20)

13.5±0.2y(20)

14.6 ± 0.3 z(19)

13.8±0.2z(20)

12.8±0.3zy(20)

12.2±0.3y(20)

15.5 ±0.6x(20)

15.5 ±0.4z(20)

15.6±0.4z(19)

13.8±0.3y(15)

15.0±0.4z(20)

Lipid(mg/100 mg)

4.7 ± 0.2(10)

9.0 ± 0.2 z(20)

8.4 ± 0.4 z(20)

9.0 ± 0.3 z(20)

9.4 ± 0.2 7.(18)

10.4 ± 0.2 zy(19)

9.4 ± 0.2 z(20)

10.8±0.2y(19)

10.6±0.4y(20)

8.8 ± 0.2 z(18)

8.8 ± 0.3 z(19)

9.8 ± 0.3 y(16)

10.5 ±0 .2y(20)

Carbohydrate(mg/100 mg)

0.45 ± 0.03(10)

0.40 ± 0.02 z(20)

0.43 ± 0.04 z(10)

0.36 ± 0.03 zy(20)

0.30 ± 0.02 y(19)

0.20 ± 0.01 z(20)

0.20 ± 0.01 z(20)

0.25 ± 0.01 y(10)

0.23 ± 0.02 zy(6)

0.35 ± 0.03 z(10)

0.27 ± 0.01 y(13)

0.28 ± 0.02 y(4)

0.25 ± 0.01 y(10)

Conditionfactor

0.87 ± 0.01(20)

1.20±0.01z(20)

1.17 ±0.01 zy(20)

1.18 ±0.02 z(20)

I .13±0.01y(20)

1.11 ±0.01 z(20)

l . I4±0.02z(20)

1.12±0.01z(20)

1.12 ±0.01 /(20)

1.24±0.02z(20)

1.22±0.02z(20)

1.16 ±0.02 /(20)

1.19±0.04z(20)

day 60, when temperature had begin its summer September was 25.9% (Table 3). This value isincrease (Figure 1), carbohydrate had decreased to somewhat lower than those reported by Birkettless than half its the initial amount. Energy in ex- (1969), who found that gross N conversion effi-cess of maintenance requirements can be stored as ciency (synonymous with our N retention effi-lipid or glycogen (Brett and Groves 1979; Wood ciency) ranged from 27.9 to 49% over three spe-1993). However, during times of increased energy cies. Our estimated fecal N losses, regarded asdemand, such as stress, these energy stores are undigested or unabsorbed diet and all N not ac-mobilized (Mayer et al. 1992). counted for, are quite high (over 48%) (Table 3).

Total ammonia levels were initially four times Subsequent analysis of similar water samples inhigher in liver than in white muscle (Figure 4A, later experiments to estimate N-waste excretion by5A). However, over time, less ammonia was pres- the Kjeldahl method (Skoog and West 1982) in-ent in the liver and more glutamine accumulated dicated that total N-waste excretion may be asthere (Figure 5). These trends may have resulted much as 34% greater than we estimated from totalfrom increased glutamine synthetase production ammonia and urea only. The missing nitrogenousand incorporation of ammonia into protein. Reid waste components probably include mucoproteinset al. (1995) reported an increase in liver protein and amino acids (Olson and Fromm 1971). Nev-synthesis rates over the duration of this experi- ertheless, feeding to satiation may have also con-ment, the most dramatic increase occurring over tributed to the low N absorption, because there wasthe first 30 d. a strong correlation in the present study between

Crude nitrogen and energy budgets were con- the amount of N consumed and the amount of Nstructed to assess the costs of growth incurred un- lost in feces before the large temperature increaseder these conditions. Nitrogen retention efficiency at day 60 (r2 = 0.92; df = 7; P = 0.0001).for the 90-d period of growth between June and The energy budget tended to correlate with the

266 LINTON ET AL.

ii

EZ3 dark hatched = base+Am• solid = baseCU open = base+ATZ2 light hatched = base+AT+Am

E2 dark hatched = base+Am• solid - baseO open =base+ATEl light hatched = base+AT+Am

12

Initial 30 60

Time (days)

90

FIGURH 4.—Effects of + 2°C (AT) and 70 u,molammonia/L (Am) on (A) total ammonia (N = 6-20) and(B) glutamine (N = 5-10) concentrations in white mus-cle of juvenile rainbow trout fed to satiation twice daily.Data are means + SE. Columns with the same letter onthe same sampling day are not significantly different (P> 0.05).

results of the nitrogen budget (Table 4). Gross con-version efficiency over the 90-d study period was38% under base conditions. It was associated withhigh food intake, relatively high energy loss viaexcretion (FE, 22.5%; U, 2.9%), and a high lossvia heat (/?, 36%). The results reported here areconsistent with previous reports under comparableconditions (Brett and Groves 1979). Cho et al.(1982) report a conversion of energy into bodymaterials of 46%, an energy loss in excreta of 23%(fecal, 15%; nonfecal 8%), and a final 30% lossvia heat production. To summarize, for juvenilerainbow trout fed to satiation and growing underthe base thermal regime, a relatively high propor-tion of gross energy intake was being stored inbody materials (especially lipid) but at the expenseof nitrogen retention efficiency and energy con-

Initial 30 60 90

Time (days)FIGURE 5.—Effects of +2°C (AT) and 70 p,mol

ammonia/L (Am) on (A) total ammonia (N = 6-20) and(B) glutamine (N = 5-10) concentrations in liver ofjuvenile rainbow trout fed to satiation twice daily. Dataare means + SE. Columns with the same letter on thesame sampling day are not significantly different (P >0.05).

version efficiency. Their estimated cost of growth,or N-cost index, was 9.66 moles of C>2 consumedper mole of N stored.

Effects of an Additional 2°CThe high temperatures reached in the last 30 d

of the experiment approached the upper thermallimit of juvenile rainbow trout, 26°C (Bidgood andBerst 1969; Kaya 1978). Despite the higher tem-perature, the 90-d growth rate of fish held at +2°C(3.01 ± 0.05 %/d) was nearly identical to that ofthe base temperature group. However, the growthrate of the base+AT fish was actually higher thanthat of base temperature fish during the first 60 d,before the sharp temperature increase, as indicatedby the statistically significant increase in wetweight (Figure 2A), but lower during the last 30d. The latter decline stemmed in part from a lev-

METABOLIC COST OF HIGH TEMPERATURE AND AMMONIA 267

TABLE 3.—Nitrogen budgets for juvenile rainbow trout fed to satiation for 90 d and exposed to a +2°C wanningscenario (AT) and 70 p,mol ammonia/L (Am).

Interval andtreatment

Days 0-30Base + AmBaseBase* ATBase + AT + Am

Days 30-60Base + AmBaseBase + ATBase + AT + Am

Days 60-90Base + AmBaseBase 4- ATBase + AT -1- Am

Days 0-90Base + AmBaseBase + ATBase + AT + Am

Nitrogen (mmol/fish)a

/

34.936.641.840.4

61.976.666.874.8

84.8110.368.183.1

181.7223.5176.8198.3

E

14.011.710.016.4

23.617.917.826.2

24.727.622.020.4

62.357.149.863.0

R

15.514.515.317.5

17.617.219.012.7

48.526.329.621.9

81.658.063.852.1

^N

5.510.516.66.5

20.841.530.035.9

40.456.516.640.9

37.8108.563.183.3

A

29.526.225.333.9

41.235,136.838.9

73.253.951.542.3

143.9115.1113.6115.0

Retentionefficiency6

(%)

44.439.536.543.4

28.322.428.416.9

57.223.943.426.3

44.925.936.126.3

Absorptionefficiency0

(%)

84.371.460.483.8

66.545.855.152.0

86.448.875.750.9

79.251.564.358.0

a / = total N consumed; E = N lost via branchial and urinary excretion; R = N retained in body materials; F^ = N lost in fecal materialand all unaccounted N; A = N absorbed from food.

b Retention efficiency = lOOtf/7.c Absorption efficiency = 1004/7.

TABLE 4.—Energy budgets for juvenile rainbow trout fed to satiation for 90 d and exposed to a +2°C warmingscenario (AT) and 70 |jimol ammonia/L (Am).

Interval andtreatment

Days 0-30Base-*- AmBaseBase + ATBase + AT -t- Am

Days 30-60Base + AmBaseBase + ATBase + AT + Am

Days 60-90Base + AmBaseBase + ATBase+AT+Am

Days 0-90Base + AmBaseBase + ATBase -I- AT + Am

Energy (kJ/fish)a

C

108.5113.6129.7125.3

192.1237.7207.4232.0

263.0342.3211.3257.9

563.7693.6548.4615.2

U

4.94.13.55.7

8.26.26.29.1

8.69.67.77.1

21.719.917.422.0

Afl

80.068.679.886.0

83.894.6

105.9108.3

160.0101.595.291.4

323.8264.7280.9285.7

R

35.238.740.744.7

70.293.590,899.0

105.3120.6120.6148.0

210.7252.8252.0291.8

FE

-11.62.25.8

-11.1

29.943.34.4

15.5

-10.8110.6

-12.211.4

7.5156.1-1.915.8

Conversionefficiency5

(%)

73.860.461.568.6

43.639.851.146.7

60.829.745.135.4

57.438.251.246.4

a C = total energy consumed from food; U = energy lost via excretion (branchial and urinary); A/? = energy stored in body materials;R - total metabolic energy lost as heat; /•£ = energy lost in feces.

h Conversion efficiency = 100AB/C.

268 LINTON ET AL.

eling off of food intake (Figure 2B) and from aconcomitant change in energy metabolism.

A 30% reduction in food intake (relative to base)during days 60-90 (mean temperature, 22°C) wasperhaps the most significant effect of the 2°Cwarming on juvenile rainbow trout. This effect isnot uncommon for salmonids; as water tempera-ture nears the upper thermal tolerance of a species,food consumption declines precipitously (Jobling1993). However, up until day 60, additional tem-perature had little effect on food consumption.Wurtsbaugh and Davis (1977) observed that forfish fed higher ration levels, elevated temperatureup to 17°C increased growth rate, but that whenthe temperature reached 22.5°C in their summerexperiment, the fish ate less than those at 19.5°C.

The additional 2°C did not affect routine 03 con-sumption rates even when food intake was sup-pressed (Figure 3A). However, N-waste excretionwas markedly reduced at this time (Figure 3B).Protein utilization depends on water temperature(Steffens 1981). Reid et al. (1995) reported a 20%decrease in protein turnover in these last 30 d andproposed that metabolic fuels (protein, lipid, andcarbohydrate) may have been diverted fromgrowth to supply energy for maintenance and otherhomeostatic processes. Because N-waste excretionwas reduced at that time, protein probably was nota major energy source. Indeed, whole-body proteinincreased between days 60 and 90 (Table 2), sug-gesting that trout exposed to an additional 2°Cretained more N for growth (Table 3). However,whole-body lipid decreased (Table 2) and mayhave been used to meet the increased energeticdemands above 20°C, as argued earlier. We suggestthat during the last 30 d, change in fuel use allowedthe fish exposed +2°C to meet the increased main-tenance costs associated with the higher water tem-perature. As a result, the overall cost associatedwith N gain in base-f-AT fish (N-cost index = 9.94moles of 62 consumed per mole of N stored) wassimilar to that of the base temperature group(9.66).

Effects of Elevated AmmoniaThe physiological effects of low environmental

ammonia are not well documented. The vast ma-jority of studies concerning the chronic effects ofammonia on freshwater fish were conductedat levels considerably higher than the 70 fxmolammonia/L we used. Effects of sublethal ammoniatoxicity on fish may include reduced growth (Rob-inette 1976; Rice and Bailey 1980; Beamish andTandier 1990), reduced feeding (Beamish and Tan-

03 dark hatched = base+Am• solid = baseD open = base+ATCZ3 light hatched = base+AT+Am

E

COCO

200

150

100

50

0Initial 30 60 90

Time (days)FIGURE 6.—Effects of +2°C (AT) and 70 p,mol

ammonia/L (Am) on (A) total ammonia (N = 10-20)and (B) Na* (N = 5-10) concentrations in blood plasmaof juvenile rainbow trout fed to satiation twice daily.Data are means + SE. Columns with the same letter onthe same sampling day are not significantly different (P> 0.05). The initial plasma ammonia value may be ar-tificially high because the fish were small and samplingwas invasive (caudal severance).

dler 1990), gill damage (Smart 1976), and plasmaion disturbance (Buckley et al. 1979; Thurston etal. 1984; Twitchen and Eddy 1994). With respectto ion disturbance, fish exposed to ammonia in thepresent study not only exhibited significantly moreplasma ammonia than base and base+AT fish byday 90, but significantly more plasma Na* as well(Figure 6). Although the mechanisms of branchialammonia excretion in freshwater teleosts remaincontroversial (Randall and Wright 1987), twomechanisms are thought to be responsible for themajority of the ammonia excreted under most en-vironmental conditions: (1) passive NH3 diffusionalong the partial pressure (PNH^) gradient fromblood to water and (2) electroneutral exchange of

METABOLIC COST OF HIGH TEMPERATURE AND AMMONIA 269

NH4" for Na* (Wood 1993). Because the PNH3gradients calculated here for the ambient controland ammonia group at 14°C were 262 and 305fxtorr, respectively (based on measured total am-monia in plasma and water, measured water pH,arterial blood pH values from Randall and Cam-eron 1973, water and plasma pK values from Cam-eron and Heisler 1983, and equations from Wrightand Wood 1985), and because a significant in-crease in plasma Na^ concentration was observedover time in these fish, it is likely that both mech-anisms contributed to the excretion of ammonia towater.

One effect that has not been reported previously,however, is the stimulation of fish growth by ex-posure to such a low level of ammonia. In thepresent study, juvenile rainbow trout exposed oversummer to 70 jxmol ammonia/L at base water tem-perature (base+Am) had significantly greaterweight gain (Figure 2A), better N and energy con-version (Tables 3, 4), and higher N retention ef-ficiency at a lower metabolic cost (Figure 3A) thanambient controls. The N-cost index for this groupwas only 6.58 moles of Ch consumed per mole ofN stored. Thus more N, in the form of protein,was being retained and put towards growth (Tables2, 3). These results contrast with those of Beamishand Tandler (1990), who reported no change ineither carcass protein or growth of juvenile laketrout Salvelinus namaycush exposed for 60 d toapproximately 330 |xmol total ammonia/L at 11°C.The increase in growth can not be explained byincreased N consumption because ammonia-ex-posed fish ate no more food than controls (Figure2B). Instead, the ammonia appears to have stim-ulated protein turnover (unpublished data from ourlaboratory), resulting in better N conversion effi-ciency. In just 30 d, hepatosomatic indices (liverweights as proportions of body weights) were el-evated; by day 60, liver protein synthesis, degra-dation, and accretion rates were significantly in-creased. This increase in protein turnover mayhave been an indirect result of ammonia detoxi-fication mechanisms acting in response to rela-tively higher levels of plasma ammonia (Figure6A). For example, ammonia is incorporated intoglutamine (Levi et al. 1974; Walton and Cowey1977) and other amino acids (Iwata et al. 1981;Dabrowska and Wlasow 1986), which may in turnbe used as substrates for protein synthesis, ulti-mately improving growth. Indeed, concentrationsof glutamine in the livers of the ammonia-exposedfish appeared to be elevated in comparison to thebase temperature group (Figure 5B).

The above arguments also hold true for thosefish exposed to the ammonia at +2°C(base-f AT+Am) up to day 30; beyond this period(above 16°C), the stimulatory effect of ammoniawas lost and the combination became deleterious.Nitrogen retention and absorption efficienciesdropped sharply (Table 3) and the amount of en-ergy lost as heat increased (/?, Table 4) betweendays 30 and 90, which undoubtedly contributed tothe high N-cost index (11.06 moles of O2 con-sumed per mole of N stored) calculated for the 90d of exposure. From these results, it is apparentthat increased metabolic costs associated with bothelevated temperature and ammonia detoxificationwere manifest in the high N-cost index estimatedfor these fish. We conclude that juvenile rainbowtrout fed to satiation and exposed over summer toa simulated warming scenario of +2°C can makethe necessary metabolic adjustments necessary tomaintain growth. However, in the presence of sub-lethal ammonia, the cost of growth will increaseand growth may be compromised.

AcknowledgmentsThis study was sponsored by a Natural Sciences

and Engineering Research Council Strategic Grantawarded to C. M. Wood and D. G. McDonald. Wesincerely thank R. W. Wilson for contributionsnecessary to start the project and I. J. Morgan forassistance with the manuscript preparation. Wethank Jacqueline Dockray, Teresa Banka, KristenYoung, and Mathew Norton for their splendid tech-nical assistance.

ReferencesAdams, S. M., and J. E. Breck. 1990. Bioenergetics.

Pages 389-415 in C. B. Schreck and P. B. Moyle,editors. Methods for fish biology. American Fish-eries Society, Bethesda, Maryland.

Agradi, E., L. Bonomi, E. Rigamonti, M. Liguori andP. Bronzi. 1995. The effect of dietary lipids ontissue lipids and ammonia excretion in Europeaneels (Anguilla anguilla). Comparative Biochemistryand Physiology A 111: 445-451.

Atherton, W. D., and A. Aitken. 1970. Growth, nitrogenmetabolism and fat metabolism in Salmo gairdneri,Rich. Comparative Biochemistry and Physiology36:719-747.

Atkins, P. W. 1987. Molecules. Freeman, New York.Beamish, F. W. H., and A. Tandler. 1990. Ambient am-

monia, diet and growth in lake trout. Aquatic Tox-icology 17:155-166.

Bergmeyer, H. U. 1985. Methods of enzymatic analysis.Academic Press, New York.

Bidgood, B. E, and A. H. Berst. 1969. Lethal temper-atures for Great Lakes rainbow trout. Journal of theFisheries Research Board of Canada 26:456-459.

270 LINTON ET AL.

Birkett, L. 1969. The nitrogen balance in plaice, soleand perch. Journal of Experimental Biology 50:375-386.

Boutilier, R. G., T. A. Heming, and G. K. Iwama. 1984.Appendix: physicochemical parameters for use infish respiratory physiology. Pages 403-430 in W.S. Hoar, and D. J. Randall, editors. Fish physiology,volume 10, part A. Academic Press, New York.

Braefield, A. E., and M. J. Llewellyn. 1982. Animalenergetics. Blackie and Son, Glasgow, UK.

Brett, J. R., and T. D. D. Groves. 1979. Physiologicalenergetics. Pages 280-344 in W. S. Hoar, D. J.Randall, and J. R. Brett, editors. Fish physiology,volume 8. Academic Press, New York.

Buckley, J. A., C. M. Whitmore, and B. D. Liming.1979. Effects of prolonged exposure to ammoniaon the blood and liver glycogen of coho salmon(Oncorhynchus kisuich). Comparative Biochemistryand Physiology C 63:297-303.

Cameron, J. N., and N. Heisler. 1983. Studies of am-monia in the rainbow trout: physico-chemical pa-rameters, acid-base behavior, and respiratory clear-ance. Journal of Experimental Biology 105:107-125.

Cho,C. Y. 1990. Fish nutrition, feeds, and feeding: withspecial emphasis on salmonid aquaculture. FoodReviews International 6:333-357.

Cho, C. Y, S. J. Slinger, and H. S. Bayley. 1982. Bio-energetics of salmonid fishes: energy intake, ex-penditure and productivity. Comparative Biochem-istry and Physiology B 73:25-41.

Cho, C. Y, and S. J. Kaushik. 1990. Nutritional ener-getics in fish: energy and protein utilization in rain-bow trout (Salmo gairdneri). World Review of Nu-trition and Dietetics 61:131-172.

Crocker, C. L. 1967. Rapid determination of urea ni-trogen in serum or plasma without deproteinization.American Journal of Medical Technology 33:361-365.

Dabrowska, H., and T. Wlasow. 1986. Sublethal effectof ammonia on certain biochemical and hematolog-ical indicators in common carp (Cyprinus carpio L.).Comparative Biochemistry and Physiology C 83:179-184.

Dawson, W. R. 1992. Physiological responses of ani-mals to higher temperatures. Pages 158-168 in R.L. Peters, and T. E. Lovejoy, editors. Global warm-ing and biological diversity. Yale University Press,New Haven, Connecticut.

Elliott, J. M. 1982. The effects of temperature and rationsize on the growth and energetics of salmonids incaptivity. Comparative Biochemistry and Physiol-ogy B 73:81-91.

Elliott, J. M., and W. Davison. 1975. Energy equivalentsof oxygen consumption in animal energetics. Oec-ologia 191:95-201.

EPA (U.S. Environmental Protection Agency). 1985.Ambient water quality criteria for ammonia. EPA,Report 440/5-85-001, Washington, D.C.

Erickson, R. J. 1985. An evaluation of mathematicalmodels for the effects of pH and temperature on

ammonia toxicity to aquatic organisms. Water Re-search 19:1047-1058.

Folch, J., M. Lees, and G. H. Sloane-Stanley. 1957. Asimple method for the isolation and purification oftotal lipids in animal tissues. Journal of BiologicalChemistry 226:497-509.

Fromm, P. O. 1963. Studies on renal and extra-renalexcretion in a freshwater teleost, Salmo gairdneri.Comparative Biochemistry and Physiology 10:121-128.

Fry, F. E. J. 1967. Responses of vertebrate poikilo-therms to temperature. Pages 375-409 in A. H.Rose, editor. Thermobiology. Academic Press, NewYork.

Hansen, J., and six coauthors. 1988. Global climatechanges as forecast by Goddard Institute for spacestudies three-dimensional model. Journal of Geo-physical Research 93:9341-9364.

Iwata, K., I. Kakuta, M. Ikeda, S. Kimoto, and N. Wada.1981. Nitrogen metabolism in the mudskipper, Per-iophthalmus cantonensis: a role of free amino acidsin detoxification of ammonia produced during itsterrestrial life. Comparative Biochemistry andPhysiology A 68:589-596.

Jayaram, M., and F. W. H. Beamish. 1992. Influence ofdietary protein and lipid on nitrogen and energylosses in lake trout, Salvelinus namaycush. CanadianJournal of Fisheries and Aquatic Sciences 49:2267-2272.

Jobling, M. 1980. Effects of starvation on proximatechemical composition and energy utilization ofplaice, Pleuronectes platessa L. Journal of Fish Bi-ology 17:325-334.

Jobling, M. 1993. Bioenergetics: feed intake and energypartitioning. Pages 1-41 in J. C. Rankin, and F. B.Jensen, editors. Fish ecophysiology. Chapman andHall, London.

Kaya, C. M. 1978. Thermal resistance of rainbow troutfrom a permanently heated stream, and of twohatchery strains. Progressive Fish-Culturist40:138-142.

Kun, E., and E. B. Kearney. 1971. Ammonia. Pages1802-1806 in H. U. Bergmeyer, editor. Methods ofEnzymatic Analysis, volume 4. Academic Press,New York.

Kutty, M. N. 1972. Respiratory quotient and ammoniaexcretion in Tilapia mossambica. Marine Biology16:126-133.

Lauff. R. F, and C. M. Wood. 1996. Respiratory gasexchange, nitrogenous waste excretion, and fuel us-age during starvation in juvenile rainbow trout.Journal of Comparative Physiology B 105:542-551.

Larmoyeux, J. D., and R. G. Piper. 1973. Effects ofwater reuse on rainbow trout in hatcheries. Pro-gressive Fish-Culturist 35:2-8.

Levi, G., G. Morisi, A. Coletti, and R. Catanzaro. 1974.Free amino acids in fish brain: normal levels andchanges upon exposure to high concentrations invivo, and upon incubation of brain slices. Compar-ative Biochemistry and Physiology A 49:623-636.

Magnuson, J. J., L. B. Crowder, and P. A. Medvick.

METABOLIC COST OF HIGH TEMPERATURE AND AMMONIA 271

1979. Temperature as an ecological resource.American Zoologist 19:331-343.

Mayer, F. L., and six coauthors. 1992. Physiologicaland nonspecific biomarkers. Pages 5-85 in R. B.Huggelt, R. A. Kimerle, P. M. Mehrle, Jr., and H.L. Bergman, editors. Biomarkers: biochemical,physiological, and histological markers of anthro-pogenic stress. Lewis Publishers, Chelsea, Michi-gan.

Mehner. T., and W. Wieser. 1994. Effects of temperatureon allocation of metabolic energy in perch (Percafluviatilis) fed submaximal rations. Journal of FishBiology 45:1079-1086.

Miller. G. L. 1959. Protein determination on larger sam-ple sizes. Analytical Chemistry 31:964.

Mount, D. L. and W. A. Brungs. 1967. A simplifieddosing apparatus for fish toxicology studies. WaterResearch 1:21-29.

Olson. K. R., and P. O. Fromm. 1971. Excretion of ureaby two teleosts exposed to different concentrationsof ambient ammonia. Comparative Biochemistryand Physiology A 40:999-1007.

Randall, D. J., and J. N. Cameron. 1973. Respiratorycontrol of arterial pH as temperature changes inrainbow trout Salmo gairdneri. American Journal ofPhysiology 225:997-1002.

Randall, D. J., and P. A. Wright. 1987. Ammonia dis-tribution and excretion in fish. Fish Physiology andBiochemistry 3:107-120.

Reid, S. D., J. J. Dockray, T. K. Linton, D. G. McDonald,and C. M. Wood. 1995. Effects of a summer watertemperature regime representative of a global warm-ing scenario on growth and protein synthesis inhard- and softwater- acclimated juvenile rainbowtrout (Oncorhynchus mykiss). Journal of ThermalBiology 20:231-244.

Rice, S. D., and J. E. Bailey. 1980. Survival, size, andemergence of pink salmon, Oncorhynchus gorbus-cha, alevins after short- and long-term exposures toammonia. U.S. National Marine Fisheries ServiceFisheries Bulletin 78:641-647.

Robinette, H. R. 1976. Effect of selected sublethal lev-els of ammonia on the growth of channel catfish(Ictalurus punctatus). Progressive Fish-Culturist38:26-29.

Rubin, E. S., and six coauthors. 1992. Realistic miti-gation options for global warming. Science 257:148-149.

Russo, R. C. 1985. Ammonia, nitrite, and nitrate. Pages455-471 in G. M. Rand and S. R. Petrocelli. editors.Fundamentals of aquatic toxicology and chemistry.Hemisphere. Washington. D.C.

Schindler, D. W., and eight coauthors. 1990. Effects ofclimatic warming on lakes of the central boreal for-est. Science 250:967-970.

Skoog. D. A., and D. M. West. 1982. Fundamentals ofanalytical chemistry. Saunders College Publishing,New York.

Smart, G. 1976. The effect of ammonia on gill structureof the rainbow trout (Salmo gairdneri). Journal ofFish Biology 8:471-475.

Soderberg, R. W. 1995. Flowing water fish culture. CRCPress, Boca Raton, Florida.

Soofiani, N. M., and A. D. Hawkins. 1982. Energeticcost at different levels of feeding in juvenile cod,Gadus morhua L. Journal of Fish Biology 21:577-592.

Sprague, J. B. 1985. Factors that modify toxicity. Pages124-163 in G. M. Rand, and S. R. Petrocelli, edi-tors. Fundamentals of aquatic toxicology and chem-istry. Hemisphere, Washington. D.C.

Steffens, W. 1981. Protein util ization by rainbow trout(Salmo gairdneri) and carp (Cyprinus carpio): abrief review. Aquaculture 23:337-345,

Swift, D. R. 1955. Seasonal variation in the growth rale,thyroid gland activity and food reserves of browntrout (Salmo trutta L.). Journal of Experimental Bi-ology 32:751-764.

Thomas. R. E., and five coauthors. 1986. Effects offluctuating temperature on mortality, stress, and en-ergy reserves of juvenile coho salmon. Transactionsof the American Fisheries Society 115:52-59.

Thurston. R. V.. and R. C. Russo. 1983. Acute toxicityof ammonia to rainbow trout. Transactions of theAmerican Fisheries Society 112:696-704.

Thurston, R. V., and seven coauthors. 1984. Chronictoxicity of ammonia to rainbow trout. Transactionsof the American Fisheries Society 113:56-73.

Trippel, E. A., R. Eckmann, and J. Hartman. 1991. Po-tential effects of global warming on whitefish inLake Constance, Germany. Ambio 20:226-231.

Twitchen, I. D., and F. B. Eddy. 1994. Effects of am-monia on sodium balance in juvenile rainbow troutOncorhynchus mykiss Walbaum. Aquatic Toxicolo-gy 30:27-45.

Verdouw, H., C. J. A. van Echted, and E. M. J. Dekkers.1978. Ammonia determination based on indophen-ol formation with sodium sal icy late. Water Research12:399-402.

Walton, M. J., and C. B. Cowey. 1977. Aspects of am-moniogenesis in rainbow trout. Salmo gairdneri.Comparative Biochemistry and Physiology B 57:143-149.

Wang, Y., M. P. Wilkie, G. J. F. Heigenhauser, and C.M. Wood. 1994. The analysis of metabolites inrainbow trout white muscle: a comparison of dif-ferent sampling and processing methods. Journal ofFish Biology 45:855-873.

Wilson, R. W. H. L. Bergman, and C. M. Wood. 1994a.Metabolic costs and physiological consequences ofacclimation to aluminum in juvenile rainbow trout(Oncorhynchus mykiss). I : Acclimation specificity,resting physiology, feeding, and growth. CanadianJournal of Fisheries and Aquatic Sciences 51:527-535.

Wilson, R. W, H. L. Bergman. and C. M. Wood. 1994b.Metabolic costs and physiological consequences ofacclimation to aluminum in juvenile rainbow trout(Oncorhynchus mykiss). 2: Gil l morphology, swim-ming performance and aerobic scope. CanadianJournal of Fisheries and Aquatic Sciences 51:536-544.

Wood, C. M. 1993. Ammonia and urea metabolism and

272 LINTON ET AL.

excretion. Pages 379-424 in D. H. Evans, editor.The physiology of fishes. CRC Press, Boca Raton,Florida.

Wood, C. M. 1995. Excretion. Pages 379-438 in C.Groot, L. Margolis, and W. C. Clarke, editors. Phys-iological ecology of Pacific salmon. UBC Press,Vancouver.

Wright, P. A., and C. M. Wood. 1985. An analysis ofbranchial ammonia excretion in fresh water rainbowtrout: effects of environmental pH change and so-

dium uptake blockade. Journal of Experimental Bi-ology 114:329-353.

Wurtsbaugh, W. A., and G. E. Davis. 1977. Effects offish size and ration level on the growth and foodconversion efficiency of rainbow trout, Salmogaird-neri Richardson. Journal of Fish Biology 11:99-104.

Received April 9. 1996Accepted October 20. 1996