The relationships between fish health, metabolic rate, swimming performance and recovery in return-run sockeye salmon...

Transcription

1 See discussions, stats, and author profiles for this publication at: The relationships between fish health, metabolic rate, swimming performance and recovery in return-run sockeye salmon... Article in Journal of Fish Diseases December 2004 DOI: /j x Source: PubMed CITATIONS 62 READS authors, including: Keith B Tierney University of Alberta 60 PUBLICATIONS 1,057 CITATIONS SEE PROFILE All content following this page was uploaded by Keith B Tierney on 29 August The user has requested enhancement of the downloaded file.

2 Journal of Fish Diseases 2004, 27, The relationships between fish health, metabolic rate, swimming performance and recovery in return-run sockeye salmon, Oncorhynchus nerka (Walbaum) K B Tierney and A P Farrell Department of Biological Sciences, Simon Fraser University, Burnaby, BC, Canada Abstract The repeat swimming ability and oxygen uptake (Mo 2 ) of adult sockeye salmon, Oncorhynchus nerka (Walbaum), were assessed at ambient water temperatures at three field locations along their migration route. Following these measurements, internal and external fish condition was evaluated according to United States Environmental Protection Agency guidelines. Here we report on the physiological characteristics of fish having either moderate or severe levels of disease and injury. Routine oxygen uptake (Mo 2 ) did not differ between healthy fish and those with indices of ill health. In contrast, fish classified as sick, which included conditions of damaged internal organs, an Ichthyophonus spp. heart infection, a Saprolegnia spp. gill infection, and skin wounds, had a lower post-exercise Mo 2 and were unable to repeat their critical swim speed (U crit ) on the second swim test. Moderate levels of disease or injury did not significantly affect either U crit or post-exercise Mo 2. We conclude that the ability of adult salmon to recover quickly from exercise may be a useful indicator of sublethal pathologies. Keywords: autopsy-based assessment, Ichthyophonus, Mo 2, Saprolegnia, sockeye salmon, U crit. Introduction Fish health can be evaluated through autopsy-based assessment. Guidelines developed for detecting Correspondence K B Tierney, Department of Biological Sciences, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada ( ktierney@sfu.ca) Present address: A P Farrell, Centre for Aquaculture and the Environment, University of British Columbia, Vancouver, BC, Canada macroscopic internal and external abnormalities (Goede & Barton 1990) are regularly used by the United States Environmental Protection Agency (Klemm, Stober & Lazorchak 1993) and Environment Canada to evaluate fish health. While successful, any autopsy-based assessment fails to appraise the functional capabilities of a fish and, by definition, requires that fish be killed, an assessment requirement that could compromise any populations under pressure from other stressors. Reduction in swimming performance has been used as a non-lethal means of evaluating the functional capabilities of a fish. Critical swimming speed (U crit ) decreases with toxicant exposure (Farrell, Gamperl & Birtwell 1998; Jain, Birtwell & Farrell 1998), and with loadings of parasites (Butler & Milleman 1971; Wagner, McKinley, Bjørn & Finstad 2003), fungus (Jain et al. 1998), and bacteria (Swanson, Baxa, Young, Cech & Hendrick 2002). Moreover, some challenges cause a decrease in the ability of fish to recover U crit and re-perform. For example, both seawater-shock (Brauner, Iwama & Randall 1994) and fungal infections (Jain et al. 1998) can impair repeat swimming ability, which means the quotient of the second and first U crit tests, the recovery ratio (RR; U crit2 /U crit1 ), falls below unity. If swimming ability is also related to autopsy-based health indices, then the ability of fish to recover quickly and repeat a U crit test may be useful as a non-lethal indicator of health, as proposed by Jain et al. (1998). Such an assessment approach may be particularly valuable for mature adult salmon, whose swimming ability is critical to successful upstream migration. With repeat U crit testing, the expectation is that healthy fish would be able to swim to the same U crit on each test whereas unhealthy fish would not. This 663

3 suggestion has not been tested by comparing standard indices of health from autopsy tests with swimming performance. A recent study examined the swimming performance and oxygen uptake (Mo 2 ) of healthy Alberni Inlet sockeye salmon, Oncorhynchus nerka (Walbaum) (Somas River stock, BC, Canada), as they moved from the marine environment to their spawning site upriver (Farrell, Lee, Tierney, Hodaly, Clutterham, Healey, Hinch & Lotto 2003). A number of fish were excluded from the analysis because post-mortem examination revealed that they had a mixture of internal and external abnormalities and injuries. Here, we use these data to test the hypothesis that autopsy-based indices of ill health are associated with changes in metabolic rate, swimming performance and recovery. Materials and methods Fish and locations Details of fish sources and experimental locations were previously reported in detail by Farrell et al. (2003), where results for healthy Somas sockeye salmon were compared with other stocks of adult salmon. Here, the results for the fish that showed abnormalities or injury are presented and compared with those published earlier for healthy fish of the same stock. Returning Somas sockeye salmon were tested in 1998, while they were either in (a) salt water, at Bamfield Marine Station (BMS; Bamfield, BC), or in (b) fresh water, at Pacifica Papers (PA; Port Alberni, BC), shortly after making the transition from sea water, or (c) at the terminus of their journey, Robertson Creek Fish Hatchery (RCH; Port Alberni, BC), after being in fresh water for at least a week. In 1999, sockeye salmon were tested again at the Pacifica Papers location. All tests were performed under ambient water conditions. A total of 66 sockeye salmon were tested in 1998 and 1999, all of which were captured using seine nets and transferred to holding tanks using knotless nylon nets. Thirty-two fish were classified as unhealthy and of these fish, 25 completed the repeat swim test. Two fish did not swim a second time and five with severe fungal infections died overnight in the respirometer. Repeat swim tests and respirometry U crit and oxygen uptake (Mo 2 ) were measured using a 215-L Brett-type respirometer (see Farrell et al. 2003). Briefly, fish were anaesthetized with 0.2 mg L )1 MS222 (tricaine methanesulphonate; Syndell Laboratories, Vancouver, BC, Canada) buffered 1:1 with NaHCO 3, measured for mass and length, and transferred into the respirometer. Fish recovered for 45 min at a low water speed (0.3 body lengths per second; BL s )1 ), after which they were given a practice swim where water speed was increased in increments of 0.15 BL s )1 every 2 min until the fish was unwilling to swim faster. The practice swim was used to set the ramp velocities and minimize any training effect. Following overnight acclimatization (12 14 h), fish were given a ramp-u crit test as described by Jain, Hamilton & Farrell (1997). A ramp-u crit protocol allowed for two successive U crit tests to be performed daily on each fish. Fish were given a 45 min recovery period between the first and second U crit measurements (U crit1 and U crit2, respectively). Recovery ratio (RR) was calculated as the quotient U crit2 /U crit1. Mo 2 was measured in 1999 using an OxyGuard Mark III oxygen electrode (Point Four Systems, Port Moody, BC, USA). Mo 2 was calculated as: ðdo 2 concentrationþðrespirometerwatervolumeþ= ðfishmassþ=ðtimeþ with O 2 concentration measured in mg L )1 ; respirometer water volume (215 L) equal to total volume less fish volume (assuming 1 kg ¼ 1 L); and time in minutes. Mo 2 was measured four times during each swimming trial: routine Mo 2 (measured prior to U crit1 ); post-exercise-1 Mo 2 (measured immediately following U crit1 ); recovery Mo 2 (measured 40 min after U crit1 ); and post-exercise-2 Mo 2 (measured immediately following U crit2 ). Routine and recovery Mo 2 values were the average Mo 2 taken over a 5 min period. Post-exercise Mo 2 was plotted against time, exponential decay curves were fitted to the data (Sigma Plot; SPSS Scientific, Chicago, IL, USA), and a 2-min post-exercise value was interpolated from the equation. Autopsy health evaluation After completing the second U crit test, fish were killed by cervical dislocation and a blood sample was taken into a 3 ml sodium-heparin Vacutainer (VWR Canlab, Mississauga, Ont., Canada) by either cardiac or caudal puncture. Two 40 ll heparinized capillary tubes (VWR) were then filled from the Vacutainer, capped with Critoseal (VWR) 664

4 and centrifuged at g for 3 min. Haematocrit (Hct) was measured for all fish within 20 min of blood collection to limit temporal changes (Soivio, Nyholm & Westman 1973). Leucocrit (Lct) was also measured for the 1999 fish, as outlined by McLeay & Gordon (1977). The average of both capillary tubes was used to determine Hct and Lct for individual fish. Each fish was visually inspected internally and externally according to United States Environmental Protection Agency (EPA) guidelines for qualitatively assessing fish health (Klemm et al. 1993). Fish were sexed and liver, spleen, kidney, gonad, mesenteric (visceral) fat, gill, pseudobranch and fin condition were categorized and recorded for each fish (Table 1). In addition to EPA criteria, the heart was inspected for infection and observations of external fungus load and wounds were recorded. When present, fungus load was considered moderate if patches were <1 cm 2 and severe if >1 cm 2. Similarly, any wounds were considered moderate if <1 cm 2 and severe if >1 cm 2. The relative masses of the ventricle (RVM), spleen (SSI), liver (HSI) and gonads (GSI) were calculated as: (organ mass/animal mass) 100%. The condition factor (K FL ) of fish was calculated using fork length (in cm) and fish mass (in kg) as: (animal mass 10 5 )/ fork length 3. Analysis Abnormal conditions and their severity were used to establish a ternary system of grading fish (Table 2). Fish were classified as sick when at least one serious abnormality was noted either internally or externally, i.e. severe fin erosion, severe fungal infection, severe hindgut inflammation, infected heart, injured spleen, haemorrhaging liver, or severe skin wound. Fish were classified as moderately sick when a less obvious abnormality was present, i.e. mild active fin erosion, minor exterior wound and fungus, or abnormally (green) coloured liver. The healthy fish from the Somas stock, reported in Farrell et al. (2003), had no abnormalities or wounds, and data on swimming performance, Mo 2, and unreported somatic indices and haematology information, are used for statistical comparisons. Studies have consistently shown that healthy adult Pacific salmon can repeat their U crit performance with a min recovery period between tests (Jain et al. 1998; Farrell et al. 2003; Lee, Devlin & Farrell 2003), and thus have an RR of 1.0. We defined impaired repeat swimming ability as an RR < It is reasonable to assume that U crit was resolved to 0.5% as the U crit test method had speed increments of approximately 10% U crit, each with a 20 min duration. Nevertheless, fish experiencing difficulty with a velocity transition could cause a minor underestimate of U crit if U crit bordered on that speed increment. Therefore, an RR < 0.95 seemed a reasonably conservative threshold. As a second measure of impaired repeat swimming ability, and as ratios present difficulties for parametric analyses, we regressed U crit2 on U crit1 in cm s )1 (using temperature and length as covariates) to test whether the health groups differed in their ability to recover and to highlight values outside the 95% CI of the regression model for healthy fish [for healthy fish U crit2 ¼ (1.04) (U crit1 ) ) (1.49) (temp); P < , R 2 ¼ 0.77]. While temperature was a significant covariate (P ¼ 0.005), length was not (P ¼ 0.555). Mo 2 and U crit comparisons among health groups and locations were made using analysis of covariance (ANCOVA) followed by a Tukey HSD post hoc test. Specifically, Mo 2 differences were tested using absolute O 2 uptake (mg O 2 min )1 ) with temperature and mass as covariates, and U crit was analysed in cm s )1 using temperature and length as covariates. To test whether specific abnormalities affected swimming ability, fish with similar abnormalities were grouped and compared with healthy fish. Differences in the proportions of RR < 0.95 between health groups were compared using the chi-square test. Differences in organ size were tested using organ mass with body mass as a covariate. Allometric differences in mass length relationships for health groups were tested by regressing (log)mass on (log)length, using sex and year as covariates. Data analyses were conducted with a fiducial limit of significance at 5% using JMP 5.0 (SAS, Toronto, Ont., Canada). Sigma Plot 8.0 (SPSS Scientific) was used for graphing. All values are presented as mean SEM except where stated otherwise. Results Fish health status Variation existed within all visual assessment categories except kidney condition. Across locations, liver colour went from a fat-rich light brown at 665

5 Table 1 Visual inspection categories used to assess sockeye salmon. Guidelines were taken from the EPA (Fish Field and Laboratory Methods for Evaluating the Biological Integrity of Surface Waters; Klemm et al. 1993), except for fungus, wound and heart categories, which were added External examination Internal examination Eyes (1 ¼ L, 2 ¼ R) Mesenteric fat (caeca coverage) N Normal 0 None E1, E2 Exophthalmia 1 Slight, less than 50% H1, H2 Haemorrhagic 2 50% coverage B1, B2 Blind 3 More than 50% coverage M1, M2 Missing 4 100% fat covered OT Other Spleen Pseudobranchs B Very dark red N Normal R Red S Swollen G Granular, rough L Lithic NO Nodular I Inflamed E Enlarged OT Other OT Other Gills Hindgut inflammation N Normal 0 None P Pale 1 Mild or slight OT Other 2 Severe Fins Kidney 0 No active erosion N Normal 1 Mild active erosion S Swollen 2 Severe active erosion, haemorrhaging M Mottled Opercula G Granular 0 No shortening U Urolithiasis 1 Shortening, some gill exposed OT Other 2 Severe shortening, gill exposed Liver Fungus A Red 0 None B Lighter or less red 1 Mild, <1 cm 2 C Fatty, light tan coloured 2 Severe, >1 cm 2 D Nodules in liver Wound E Focal discolouration, marbling 0 None OT Other 1 Mild, <1 cm 2 Bile 2 Severe, >1 cm 2 0 Yellow, not full 1 Yellow, full 2 Light green, green 3 Dark green, blue green Sex M F Heart N Normal I Infected Bamfield to a marbled red at Robertson Creek. Physically damaged internal organs were found in two fish (liver, BMS-09; spleen, RCH-10) and infected hearts (evident as obvious, small <1 mm 2 circular white focal lesions, later identified as Ichthyophonus spp.) were found in four fish (BMS- 06, PA98-04, PA98-09, PA99-20), one of which (BMS-06) was excluded from the main analysis as it did not perform a second U crit. Histology revealed that the cardiac Ichthyophonus infection was not accompanied by myocardial degeneration. Two fish with severe hindgut inflammation were also noted (PA99-05, PA99-22). Among the external abnormalities, physical gill damage, probably caused by a seine net, was found in one sockeye (PA99-17; scored as moderately sick), and minor gill tip loss was found in three other sockeye (PA99-03, PA99-04, PA99-05). External fungal infections were present in 15 of 25 fish, four of which were considered severe (Table 2). As fungal infections in migrating adult sockeye are known to initiate fin erosion (Collins, Elling & Gauley 1962), fin deterioration generally paralleled fungus burden. Severe external wounds were also noted on three fish (from a gillnet on BMS-03, and <1 cm deep wounds on PA98-01, PA99-03). Variations in the 666

6 Table 2 Repeat critical swimming abilities, abnormalities and other parameters of moderately sick (a), sick (b) and RR ¼ 0 (c) sockeye salmon sampled at three locations [Bamfield Marine Station (BMS); Port Alberni (PA), Robertson Creek Fish Hatchery (RCH)] Conditional assessment New additions Overall Fish Temp ( C) Sex Mass (kg) Ucrit1 (BL s )1 ) Ucrit2 (BL s )1 ) RR Hct (%) Lct (%) KFL Eyes Gills Fins Gut Spleen Liver Heart Fungus Wound Ext. Int. (a) BMS M N N 0 0 R C N PA M N N 0 0 R E N PA F N N 1 0 R C N PA F N N 1 0 R C N PA F N N 1 0 R C N PA M N N 1 0 R C N PA F N N 1 0 R C N PA M N N 1 0 R C N PA F N N 1 1 B C N PA F N N 1 0 B C N PA M B N 1 0 B C N PA F H N 1 0 B C N PA M N OT 0 0 B C N RCH F N N 1 0 R E N Average SEM n (b) BMS M N OT 0 0 R C N BMS M N N 0 0 R OT N PA F N N 2 0 R C N PA F N N 0 0 R C I PA F N N 0 0 R C I PA F OT OT 2 0 B C N PA F H OT 2 0 B C N PA M N OT 2 2 B C N PA M N N 1 0 B C I PA M N N 2 2 B C N RCH M N N 0 0 OT E N Average SEM n (c) BMS M N N 0 0 R C I PA F N N 0 0 OT C N Only those abnormalities considered health concerns are reported. The scoring key is given in Table 1, except for Overall categories, where abnormalities are scored as 0 ¼ nominal, 1 ¼ moderate, 2 ¼ severe. 667

7 relative amount of mesenteric fat and the colourations of the spleen and bile were not considered negative health factors. Swimming performance and recovery The majority (12 of 14) of fish classed as moderately sick had an RR > 0.95 (Table 2) and U crit2 values within the 95% CI for healthy fish (Fig. 1). In fact, U crit1 and U crit2 were identical to those previously reported for healthy fish (Fig. 2a). One moderately sick fish (PA98-10) with an RR of 0.80 had an exterior fungal infection, which in time may have caused severe gill damage and mortality (see Lethally sick fish). Unlike moderately sick fish, the majority (seven of 11) of fish classed as sick had recovery ratios less than 0.95 (Table 2), and (six of 11) had U crit2 values below the 95% CI of healthy fish (Fig. 1). The exceptions included a sick fish with an infected heart (PA99-20) that had an acceptable RR (0.98) and U crit2, unlike other fish where this condition indicated poor recovery (three of four fish had RR < 0.95; BMS-06 RR ¼ 0, PA98-04 RR ¼ 0.81, PA98-09 RR ¼ 0.88). In addition, a fish with a 1 cm 2 haematoma on the liver (BMS-09) recovered and Figure 1 Actual vs. predicted U crit2 values for Somas River sockeye salmon, based on a regression model for condition-free (healthy) fish from Farrell et al. (2003) (see Materials and methods; model 95% CI shown on graph; P < , R 2 ¼ 0.77). Fish performing poorer than expected appear below lower confidence limit. Figure 2 (a) Repeat critical swimming performance (U crit1 first bar, U crit2 second bar), recovery ratio (RR ¼ U crit2 /U crit1 ), and (b) post-exercise oxygen uptake (Mo 2 )at 2 min after U crit of sockeye salmon in three states of health (see Table 2). The 2-min post-exercise Mo 2 was interpolated using curve fit equations from each fish. n is given in brackets below bars; values with same letters are not significantly different (Mo 2 and U crit, ANCOVA and Tukey HSD; RR, chi-square). Dotted lines show mean 95 CI for (a) U crit1 and (b) post-exercise Mo 2 at 2 min for healthy fish. 668

8 swam well, as did one with both severe fin rot and a large (11 cm long by 2 cm deep) external wound (PA99-03). The two fish with severe hindgut inflammation (PA99-05 and PA99-22) had contrasting results. PA99-22 performed poorly on the first test and improved on the second (1.80 and 2.03 BL s )1 ;RR¼ 1.13), whereas PA99-05 performed well on the first test but was unable to recover (2.20 and 1.90 BL s )1 ;RR¼ 0.86). PA99-05 differed from PA99-22 by additional loss of gill tips. To test which abnormalities were statistically related to reduced performance, the data for fish with similar conditions were grouped and compared with data for healthy fish. Mild fungal infection was not related to a decrease in either RR or U crit, whereas severe fungal infection significantly reduced RR (P ¼ 0.026) and both U crit s(u crit1, P ¼ 0.029; U crit2, P ¼ 0.010). Severe fungal infection was characterized by gill tip loss and severe fin rot. Infected hearts impaired U crit2 (P ¼ 0.004) and RR (P ¼ 0.024), but not U crit1. Hindgut inflammation, the relative amount of mesenteric fat, and the colourations of liver, spleen and bile were not related to U crit. Interestingly, the presence of damaged internal organs (spleen and liver) reduced U crit in both swim tests (U crit1, P ¼ 0.053; U crit2, P ¼ 0.008) but did not affect RR. Respirometry and fish health Neither routine Mo 2 ( , and mg O 2 kg )1 min )1 for healthy, moderately sick and sick groups, respectively) nor recovery Mo 2 ( , and mg O 2 kg )1 min )1 for healthy, moderately sick and sick groups, respectively) differed significantly between health groups. However, sick fish had a lower post-exercise Mo 2 for U crit2 compared with moderately sick fish (P ¼ 0.023) (Fig. 2b), which corresponded to a lower U crit2 (P ¼ ). Both temperature and mass were significant covariates of post-exercise Mo 2, but not of routine or recovery Mo 2 [post U crit1 Mo 2 ¼ 34.6 ) (1.05) (temp) ) (1.48) (mass); post U crit2 Mo 2 ¼ 49.6 ) (2.56) (temp) + (10.5) (mass)]. Lethally sick fish A sockeye salmon with a hook wound that had penetrated the body cavity (and was subsequently found to have damaged the liver) swam poorly (U crit1 ¼ 1.66 BL s )1 ) and died during the 45 min recovery period following U crit1. At temperatures of 18 C or greater, Saprolegnia spp. infections of the gills (loss of 1 10% of the gilltips) were associated with the overnight death of five fish in the respirometer although each fish had performed acceptably during the practice swim. With these fish, the severe fungal infection appeared to have removed the external mucous layer from the body in addition to causing gill tip loss. Haematology and somatic indices Haematocrit was found to be independent of health ranking, U crit, RR and Mo 2. For moderately sick fish, Hct was %, and for sick fish, Hct was %. Healthy fish had a significantly larger Lct ( %) than moderately sick fish ( %), but not sick fish ( %). The lack of significant difference between sick and healthy fish was due to one sockeye with an infected heart (PA99-20) and an Lct (1.55%) within the range of healthy fish ( %), as excluding this fish from the analysis causes the Lct of sick fish ( %) to be significantly lower than healthy fish. None of the somatic indices were related to health. However, liver mass decreased and gonad mass increased as fish matured and reached the hatchery (Fig. 3) [data includes healthy fish (previously unreported) from Farrell et al. 2003; n ¼ 59]. The mass of the ventricle and spleen did not vary significantly across location, group or sex (data not shown). There were significant differences in allometries between years. Specifically, 1999 fish tended to be leaner, especially when greater than 55 cm [for (log) mass (y) and (log) length (x): 1998, y ¼ (1.53) (x) ) 2.42; 1999, y ¼ (1.74) (x) ) 2.73]. Discussion This study is the first to directly relate abnormalities revealed by the US EPA fish health assessment guidelines with Mo 2, swimming performance and recovery in fish. The major conclusion was that a moderate level of sickness did not significantly affect Mo 2, U crit and RR. In contrast, fish with more severe indices of ill health swam well on a U crit test, but had trouble recovering and consequently their U crit2,rrand post-u crit2 Mo 2 were all significantly reduced. Furthermore, two additional sick fish did not 669

9 Figure 3 (a) Liver (hepatosomatic index, HSI%) and (b, c) gonad (gonadosomatic index, GSI%) size of return-run sockeye salmon at three points along inland migration route [data shown for present study and Farrell et al. (2003); n ¼ 59]. Actual values are given below bars; values with same symbols are not significantly different (ANCOVA, Tukey HSD). swim a second time (i.e. RR ¼ 0) and five fish with severe fungal infections died overnight following a practice swim. Thus, the ability of fish to recover quickly (i.e. 45 min) from a swim test and perform again seems to be a potentially useful assessment tool for identifying weakened functional capacity (in this case associated with abnormalities and wounds). This finding has particular relevance to fish, such as salmon, that migrate long distances and face hydraulic challenges that must be negotiated with repeated swimming efforts. Our finding that minor abnormalities did not adversely affect repeat swimming performance is consistent with earlier work by Collins et al. (1962). They noted that the swimming ability of adult sockeye continuously ascending a fishway became impaired when abrasions and minor fungal infections on the head and fins developed into more serious infections around day 4. Overall, some abnormalities related specifically to impaired U crit2 and RR, while others affected both U crit values equally. Prior knowledge of absolute swimming performance of the fish being assessed would increase the power of the swimming assessments and points to the need for tests of fish swimming performance on reference populations, if swimming performance is to be adopted as a useful non-lethal assessment tool. Lct was lower for fish with abnormalities, which was not surprising given that Lct is known to decrease with stress (McLeay & Gordon 1977) and acute and chronic infection (Wedemeyer, Gould & Yasutake 1983). As a result, fish health assessments could be further strengthened with Lct measurements (e.g. Adams, Brown & Goede 1993), provided blood sampling could be performed without harming fish to be released. Curiously, 1999 fish ranked as sick appeared heavier than healthy fish, which may mean that healthy sockeye salmon were trim marathon swimmers. Hct fell within the normocythaemic range and therefore any negative effects we observed on either Mo 2max or U crit were not related to anaemia (Gallaugher, Axelsson & Farrell 1992). In summary, observations of serious pathologies were correlated with reductions in either recovery ratio or swimming speed. Conversely, a moderate amount of disease or injury did not appear to hinder swimming ability. In the future, and with further refinement, swimming assessments hold promise as a reliable assessment of fish health without autopsy. Although the significance of the observed ailments remains unknown, future studies could monitor fish health after an assessment of swimming performance had been made. 670

10 Acknowledgements Several people and organizations facilitated this project. Sincere thanks to Larry Cross from Pacifica Papers (now Norske Skog), Ray Volk and Glenn Rasmussen from the Robertson Creek fish hatchery, all 1998 Bamfield Marine Sciences Centre staff, Jim Mitchell from the Canadian Department of Fisheries and Oceans, James Shoults of SFU Science Technical and Margo Moore of Simon Fraser. Special thanks to Joanne Constantine of BC Ministry of Agriculture and Food, and the Hupaçasath First Nations, especially Chief Judith Sayers, Susan Lauder, Steven Tatoosh and the entire Tatoosh family. This study was funded by a Forest Renewal British Columbia (FRBC) grant to APF. References Adams S.M., Brown A.M. & Goede R.W. (1993) A quantitative health assessment index for rapid evaluation of fish condition in the field. Transactions of the American Fisheries Society 122, Brauner C.J., Iwama G.K. & Randall D.J. (1994) The effect of short-duration seawater exposure on the swimming performance of wild and hatchery-reared juvenile coho salmon (Oncorhynchus kisutch) during smolting. Canadian Journal of Fisheries and Aquatic Sciences 51, Butler J.A. & Milleman R.E. (1971) Effect of the salmon poisoning trematode, Nanophyetus salmincola, on the swimming ability of juvenile salmonid fishes. Journal of Parasitology 57, Collins G.B., Elling C.H. & Gauley J.R. (1962) Ability of salmonids to ascend high fishways. Transactions of the American Fisheries Society 91, 1 7. Farrell A.P., Gamperl A.K. & Birtwell I.K. (1998) Prolonged swimming, recovery and repeat swimming performance of mature sockeye salmon Oncorhynchus nerka exposed to moderate hypoxia and pentachlorophenol. Journal of Experimental Biology 201, Farrell A.P., Lee C.G., Tierney K., Hodaly A., Clutterham S., Healey M., Hinch S. & Lotto A. (2003) Field-based measurements of oxygen uptake and swimming performance with adult Pacific salmon using a mobile respirometer swim tunnel. Journal of Fish Biology 62, Gallaugher P., Axelsson M. & Farrell A.P. (1992) Swimming performance and haematological variables in splenectomized rainbow trout, Oncorhynchus mykiss. Journal of Experimental Biology 171, Goede R.W. & Barton B.A. (1990) Organismic indices and an autopsy based assessment as indicators of health and condition of fish. American Fisheries Society Symposium 8, Jain K.E., Hamilton J.C. & Farrell A.P. (1997) Use of a ramp velocity test to measure critical swimming speed in rainbow trout (Oncorhynchus mykiss). Comparative Biochemistry and Physiology 117A, Jain K.E., Birtwell I.K. & Farrell A.P. (1998) Repeat swimming performance of mature sockeye salmon following a brief recovery period: a sensitive measure of fish health and water quality. Canadian Journal of Zoology 76, Klemm D.J., Stober Q.J. & Lazorchak J.M. (1993) Fish Field and Laboratory Methods for Evaluating the Biological Integrity of Surface Waters. U.S. Environmental Protection Agency (EPA/ 600/R-92/111), Cincinnati, OH. Lee C.G., Devlin R.H. & Farrell A.P. (2003) Swimming performance, oxygen consumption and excess post-exercise oxygen consumption in adult transgenic and ocean-ranched coho salmon. Journal of Fish Biology 62, McLeay D.J. & Gordon M.R. (1977) Leucocrit: a simple hematological technique for measuring acute stress in salmonid fish, including stressful concentrations of pulpmill effluent. Journal of the Fisheries Research Board of Canada 34, Soivio A.S., Nyholm K. & Westman K. (1973) Notes on haematocrit determinations on rainbow trout, Salmo gairdneri. Aquaculture 2, Swanson C., Baxa D.V., Young P.S., Cech J.J. Jr & Hendrick R.P. (2002) Reduced swimming performance in delta smelt infected with Mycobacterium spp. Journal of Fish Biology 61, Wagner G., McKinley R.S., Bjørn P.A. & Finstad B. (2003) Physiological impact of sea lice on swimming performance of Atlantic salmon. Journal of Fish Biology 62, Wedemeyer G.A., Gould R.W. & Yasutake W.T. (1983) Some potentials and limits of the leucocrit test as a fish health assessment method. Journal of Fish Biology 23, Received: 15 March 2004 Revision received: 13 July 2004 Accepted: 20 September View publication stats