Materials and Methods

Animals—Eleven adult yellow perch (Perca flavescens [2 males and 9 females]), 5 adult walleye pike (Sander vitreus [2 males and 3 females]), and 8 adult koi (Cyprinus carpio [4 males and 4 females]) were included in the study. All fish were housed at the Milwaukee County Zoo in 94.0 × 91.4 × 457.2-cm (37 × 36 × 180-inch), 3,558.3-L (940-gallon) cement holding tanks on a direct flow-through system with a water temperature of 17°C (62.6°F). All fish had been housed at the zoo for a minimum of 4 months and appeared in good health at the time of the study as determined by complete veterinary examinations. Food was withheld from all fish for 96 hours prior to surgery, which was 1 feeding cycle because the fish were normally fed every 3 to 4 days. This study was approved by the University of Wisconsin-Madison School of Veterinary Medicine Institutional Animal Care and Use Committee.

Experimental procedures—Each fish was captured manually with a nylon net. Immediately after capture, with the fish still in the net, a mixed venous-arterial blood sample (0.1 mL) was collected from the caudal tail vein of the conscious fish into a 1-mL heparinized syringe^a with a 25-gauge needle. Because fish arteries and veins are thin-walled, collection of peripheral blood samples may typically result in a mixture of arterial and venous blood. Blood samples were immediately analyzed by use of a handheld clinical blood analyzer^b and cartridge^c for determination of base excess of the extracellular fluid, blood lactate concentration, pH, Pco₂, and Po₂. Blood gas values were measured at analyzer temperature (37°C) and corrected to water temperature (17°C), with the assumption that ambient water temperature and individual fish body temperatures were equivalent.

After the initial blood sample was collected, each fish was placed into a 37.9-L (10-gallon) anesthetic induction tank containing tricaine methanesulfonate^d (150 mg/L) buffered with sodium bicarbonate powder (75 mg/L).^e An air stone attached to an air pump was maintained in the anesthetic bath that aerated the water and ensured a relatively high oxygen content in the tank. Once each fish had lost its righting reflex, it was considered anesthetized and removed from the induction tank, weighed, and transferred to a surgical table where a second blood sample was collected and analyzed. A surgical plane of anesthesia was maintained by use of a recirculating anesthesia unit,⁹ with tricaine methanesulfonate concentrations between 100 and 125 mg/L. Although the flow rate of water into the mouth of each fish was not measured, the same anesthesia unit was used for every fish so that water flow remained constant throughout the study.

Each anesthetized fish was placed in dorsal recumbency, the ventral midline was cleaned by making 1 swipe with a sterile saline (0.9% NaCl) solution–soaked gauze sponge, and the surgical site was draped. The skin and coelomic membrane were incised by use of a No. 15 scalpel blade to expose the coelomic cavity. Female perch and walleye were ovariectomized, with males left sexually intact, whereas all koi were gonadectomized. Because the gonads extend from the level of the caudal opercula to just cranial to the cloaca in koi, a long midline incision and splitting of the cartilaginous pectoral girdle were required in this species. During all gonadectomy surgeries, blunt dissection from a caudal to cranial direction was used to exteriorize the gonad and bipolar radiosurgical forceps^f were used to simultaneously cut and coagulate the ovarian artery and vein. If the ovarian vessels were very large, a single hemostatic clip^g was applied for ligation.

On completion of surgery, incisions were closed in a single, continuous layer by use of 3-0 polyglyconate^h suture in a Ford interlocking pattern. Prior to completion of suture closure, sterile saline solution was infused into the coelomic cavity to maintain neutral buoyancy. Povidone-iodine ointmentⁱ was applied topically to the incision to help prevent subsequent infection. Immediately following closure, a final blood sample was collected and analyzed. Each fish then received IM injections of long-acting oxytetracycline (20 mg/kg [9.1 mg/lb]) and ketoprofen (2 mg/kg [0.9 mg/lb]) that were administered after the final blood sample was obtained to prevent any alteration in blood lactate concentrations as a result of the injections. Each fish was then placed in a 37.9-L recovery tank containing water from the cement holding tank and monitored visually for operculations and return of the righting reflex during recovery. Once fully recovered, fish were returned to their original large cement holding tanks. A 2.27-kg (5-lb) aquarium salt block was placed into the holding tank to provide approximately 0.7% salinity to decrease the osmotic gradient and aid in recovery. The increased salinity in the water decreases the osmotic flow of water into the fish, especially with a large surgical wound present as a defect in the natural epidermal barrier. The increased salinity also decreases the environmental microorganisms and increases the production of mucus on the skin, which is a part of the passive immune system of fish, leading to decreased risk of postoperative microbial infection. Fish were monitored with daily visual inspections for 2 weeks after surgery.

Because of technical problems associated with the analyzer cartridges, 1 perch only had blood values available from the preanesthesia period, so data from this fish were removed from all analyses. Additionally, postoperative results for 1 walleye were delayed for approximately 5 minutes because the analyzer cartridge malfunctioned and this fish therefore had to have new a new blood sample obtained. Because of the small sample size of the walleye group in this study, the data from this fish were included in the analyses.

Statistical analysis—Data are presented as mean ± SD for each physiologic variable, in each species, and at each time point. When values for Po₂, Pco₂, and lactate were below the analyzer detection level, a value just below the detection range (4.9 mm Hg for Po₂ and Pco₂ and 0.29 mmol/L for lactate) was assigned to calculate the mean and SD. Additionally, a Friedman nonparametric test was run for these 3 variables to confirm that the values assigned for Po₂, Pco₂, and lactate did not affect the results.

Mean values for each physiologic variable for each species for the 3 time points were compared by use of a 1-way repeated-measures ANOVA with a compound symmetry correlation structure.^j Residuals were checked for normality and constant variance across groups. If results of the F test were significant, post hoc pairwise comparisons that were corrected with the Tukey-Kramer method were performed.

For temperature dependent variables (pH, Po₂, and Pco₂), analyses were performed on data measured at 37°C because those data were measured by the analyzer (vs corrected for patient temperature by use of an algorithm). Base excess was calculated by the analyzer with the following formula:

View Expanded

where BE is the base excess and HCO₃ represents the blood bicarbonate concentration. For perch, mean blood lactate concentrations immediately after surgery for survivors and nonsurvivors were compared by use of a Welch 2-sample t test. No values for blood lactate concentration were below the detection level immediately after surgery. No comparisons were made for walleye or koi because all fish survived.

Middle 95% reference ranges, defined as the values between the 2.5th and 97.5th percentiles, for the blood variables at each time point were calculated for perch and koi. Actual ranges rather than reference ranges were calculated for walleye because of the small sample size (n = 5). The 2 perch that died were excluded from these calculations. Statistical analyses were performed by use of a commercial software package,^j and significance was set at values of P < 0.05.

Results

During the 2-week postoperative monitoring period, all walleye and koi survived but 2 female perch died. The first perch died within 10 minutes after surgery, and a hepatic sarcoma was found on necropsy examination. The second perch died 10 days postoperatively and was found to have a tail infarction on necropsy.

Mean blood pH values for perch decreased significantly (P = 0.005) from before anesthesia (blood samples obtained while the fish were conscious) to during anesthesia. Mean blood pH values for perch increased significantly (P < 0.001) from during anesthesia to immediately after surgery so that there was no significant (P = 0.830) difference between the mean pH between before anesthesia and immediately after surgery in this species. Mean blood pH values for walleye and koi did not change significantly throughout the study (Table 1).

Mean blood Pco₂ decreased immediately after surgery in perch so that it was significantly different from values before anesthesia (P = 0.019) and during anesthesia (P = 0.003). Mean blood Pco₂ decreased significantly (P = 0.039) in koi from before to during anesthesia.

Mean blood Po₂ increased significantly in both perch (P = 0.021) and koi (P = 0.004) from before to during anesthesia. Furthermore, in koi, mean Po₂ increased significantly (P = 0.035) from before anesthesia to immediately after surgery.

The mean base excess significantly decreased in all 3 species from before anesthesia to during anesthesia (Table 1). Mean base excess increased significantly (P = 0.002) in perch from during anesthesia to immediately after surgery, such that the mean base excess before anesthesia was not significantly (P = 0.281) different from the mean base excess immediately after surgery. In walleye, the increase in mean base excess from during anesthesia to immediately after surgery was not significant (P = 0.434). The mean base excess increased in koi from during anesthesia to immediately after surgery, but this increase was not significant (P = 0.386); there was a significant (P = 0.001) difference in the mean base excess values of koi between the before anesthesia and immediately after surgery time points.

Mean blood lactate concentrations increased significantly in all 3 species from before anesthesia to immediately after surgery (Table 1). Mean blood lactate concentrations significantly increased in perch and koi from before to during anesthesia, but there were no significant changes in the blood lactate concentrations of walleye at these same time points. Immediately after surgery, koi had mean ± SD blood lactate concentrations of 5.46 ± 1.56 mmol/L, whereas walleye had mean blood lactate concentrations of 7.25 ± 3.26 mmol/L. Surviving perch had a mean blood lactate concentration immediately after surgery of 6.06 ± 1.47 mmol/L, which was significantly (P < 0.001) lower than the blood lactate concentrations at the same time point in the 2 perch that died (ie, 10.58 and 10.72 mmol/L).

Middle 95% reference ranges for blood values (pH, Pco₂, Po₂, and base excess) at each time point for perch and koi and a range of values for walleye were determined (Table 2).

Discussion

Results of the study reported here suggest that elevated blood lactate concentrations immediately after surgery (≥ 10 mmol/L) may be predictive of a poor short-term survival rate after surgery in yellow perch and perhaps other fish species. In the present study, all fish recovered from surgery and anesthesia uneventfully, except for the perch that died shortly after surgery and had a hepatic sarcoma. During the postoperative monitoring period, 1 additional perch died and histologic examination indicated that it had a tail infarction. In perch, caudal tail vein venipuncture was more difficult than in the other fish species. Captive perch are maintained at a low water temperature (17°C), so peripheral vasoconstriction may have contributed to the difficulties encountered in venipuncture. Tail infarction in the affected perch may have been the result of vascular damage and thrombosis secondary to multiple venipuncture attempts. Although the number of attempts was not recorded, this fish did require a higher number of attempts than some other perch.

In the present study, mean blood Po₂ increased in both perch and koi from before to during anesthesia and also in koi from before anesthesia to immediately after surgery. This was unexpected because, in a previous study,⁵ a profound decrease in Po₂ with an increased anesthetic duration was reported for pacu (Piaractus brachypomus). In the present study, fish were operculating (breathing) spontaneously throughout the anesthesia period, and the water in the recirculating anesthesia unit was well oxygenated, which may explain the different results. Alternatively, because the blood samples in our study were considered mixed venous-arterial blood samples, it is possible that the variability in Po₂ negated any significant changes in Po₂ over time. A third possibility is that the variation could be the result of the differing water temperatures (17°C [62.6°F] in this study and a mean of 21.5°C [70.7°F] in the previous study⁵) because oxygen solubility is higher in colder water.

The blood Pco₂ decreased in both perch and koi in the present study, which was unexpected. This decrease in blood Pco₂ was immediate in koi and occurred from before to during anesthesia, and in perch, the significant decrease in blood Pco₂ did not occur until immediately after surgery. During anesthetic procedures in fish, opercular movement (respiration) typically decreases in rate and depth, compared with that of conscious fish. Therefore, the Pco₂ was expected to increase with increasing duration of anesthesia. In red pacu, anesthesia with either tricaine methanesulfonate or eugenol contributed to a consistent increase in the mean Pco₂ of mixed venous-arterial blood.⁵ However, blood Pco₂ initially increased, then decreased with time in anesthetized carp (C carpio).¹⁰ These carp¹⁰ were anesthetized with carbon dioxide, which may explain the initial increase in Pco₂ but does not explain the subsequent decline. There may also be anatomic or physiologic differences between fish species that account for the variation in results between the different studies. Differences in water carbonate alkalinity and other water-quality variables may also explain the discrepancy in results, as alkalinity affects the blood Pco₂ of fish.¹¹ Furthermore, it is possible that the fish in the present study also had an increase and decrease in Pco₂ which were not detected because of the timing of blood sample collection.

In the present study, all 3 fish species had significant decreases in mean base excess from before to during anesthesia; however, only perch had a significant decrease in mean blood pH over the same time frame and koi actually had a decrease in Pco₂ during the same time frame. It is uncertain whether perch are unable to buffer their blood as well as walleye or koi, if perch struggled more at capture leading to respiratory or metabolic acidosis or both, or if perch naturally maintain a lower blood pH than the other 2 species. The observation that the base excess values of perch returned to baseline (preanesthetic) values immediately after surgery suggests that capture and exertion may have led to the observed changes in pH. However, base excess values of koi did not return to baseline values immediately after surgery. One possible explanation for this is that the surgical procedure took longer in koi as a result of the larger incision and pectoral splitting necessary to access the gonads. Additionally, in brook trout, it has been reported that¹² plasma protein concentration contributed substantially to blood buffering capacity, compared with the capacity of plasma protein buffering in humans. Mean base excess in the walleye did not significantly (P = 0.067) change between before anesthesia and immediately after surgery, but the sample size was small. Perhaps with a larger number of fish, the changes would have been significant, as they were in koi.

As expected with restraint, anesthesia, and surgery, in the present study, all 3 species had a significant increase in blood lactate concentrations from before anesthesia to immediately after surgery. Interestingly, there was no significant increase in blood lactate concentrations from during anesthesia to immediately after surgery in any of the 3 species, suggesting that the surgery itself played a minor role in these increased blood lactate concentrations. Instead, restraint and anesthesia alone were sufficient to contribute to the increased lactate concentrations observed over time. In both perch and koi, there was a significant increase in lactate concentration from before to during anesthesia. Although blood lactate concentrations of walleye increased from before to during anesthesia, the change was not significant (P = 0.066) and may have been a function of the small sample size. Our results are consistent with those of a previous study¹³ of splake (Salvelinus fontinalis × Salvelinus namaycush), a species closely related to walleye, in which anesthesia contributed to increased blood lactate concentrations that peaked 1 hour after anesthetic recovery. Additionally, a study¹⁴ of bone fish (Alula vulpis) revealed that blood lactate concentrations took > 4 hours to return to baseline values after only 4 minutes of exercise, and prolonged hypoxia contributed to an increase in blood lactate concentrations in sculpin (Monocephalus scorpius).⁴ However, the fish in the present study were not hypoxemic, the water was well oxygenated, and they were not manipulated for > 1 hour. The prior study⁴ of sculpin revealed that 4 hours of hypoxemic conditions were needed to induce the increase in blood lactate concentrations. Other studies^15,16 have found a significant effect of water temperature and time of day on the magnitude of increased blood lactate concentrations after air-emersion or exhaustive exercise. These effects were minimized in the present study as all fish were maintained at the same water temperature and with the same controlled artificial 12-hour light and 12-hour dark cycle.

Although the temperature dependent variables (pH, Po₂, and Pco₂) in the present study were collected at both 37° and 17°C, the decision was made to calculate and report the data at the analyzer temperature (ie, 37°C). Prior studies¹⁷ have indicated that cold-blooded vertebrates including fish maintain an acid-base balance despite changes in body temperature. Furthermore, the data at 37°C are actually measured by the analyzer, whereas the patient temperature data are calculated by algorithms designed for human patients that may not be valid for fish. Analyzing and presenting the data at 37°C will enable other researchers and clinicians to use the reference ranges, regardless of patient temperature.

A recent study³ evaluated the use of a handheld clinical analyzer in 2 species of rockfish (Sebastes melanos and Sebastes mystinus) and identified failures of the individual cartridges and a number of limitations in the use of the analyzer. However, the prior investigators did not measure blood lactate concentrations or the effects of surgery.³ A previous study¹⁸ in chickens (Gallus gallus) revealed that the handheld system was reliable, was easy to use, and had an acceptable accuracy, compared with that of traditional analyzers. In the present study, the handheld analyzer was chosen because of portability and the ability to rapidly identify trends in serial blood gas values and blood lactate concentrations in fish. However, because use of mammalian temperature-correction factors for determination of fish blood gas tensions and acid-base status yields values that are significantly different from those measured directly at fish body temperature,¹⁹ further studies of healthy and diseased fish are indicated. Additional investigation of other fish species will be necessary to determine whether high blood lactate concentrations are predictive of impending death.