Aberdeen, WA. They weighed 20.5 ± 6.3 g and mea- 

 sured 132 ±17 mm in fork length at the completion 

 of these tests. Fish were acclimated to water tem- 

 peratures of 15.0±0.5°C (5°C per week with 2 weeks 

 at 15°C) in acclimation tanks described by Ebel 

 et al. (1971), and Dawley and Ebel (MS). Fish were 

 fed once each day a maintenance ration of Oregon 

 Moist Pellets®. 



Fish were exposed to various nitrogen super- 

 saturation levels in shallow rectangular tanks 

 (Dawley and Ebel, MS) with water 23 cm deep from 

 the same source as used in the acclimation tanks. 

 Tests were performed at 15.0 ± 0.56°C at nitrogen 

 gas (N 2 + Ar) saturations of 103.5, 105.4, 110.1, and 

 116.0 ± 2.0% as calculated from Weiss (1970). These 

 nitrogen saturations corresponded to total gas 

 pressure (TCP) saturations (N 2 + Ar + 2 ) of 99.9, 

 102.0, 105.9 and 110.2%. Mean oxygen saturations 

 ranged between 87.5 and 90.5%. Fish were continued 

 on the same maintenance ration of Oregon Moist 

 Pellets® during these tests. 



Sixty fish were taken concurrently from the 

 acclimation tanks and placed in each of three 

 experimental and three control tanks (102.0 versus 

 99.9; 105.9 versus 99.9; and 110.2 versus 99.9% TGP). 

 Tests at each saturation level were continued until 

 35 days had elapsed (the 35-day test period was 

 chosen to parallel the average exposure time of 

 seaward migrant steelhead passing downstream 

 through the nitrogen supersaturated waters of the 

 lower Snake and Columbia Rivers). At this time, 

 each fish was given an individual voluntary swim- 

 ming performance test (2 to 5 min) in an inclined 

 vee trough against a water current of 1.28 m/sec 

 (15.0 ± 0.5°C) (Schiewe, in press). A variety of 

 experimental limitations dictated that these fish 

 must be used in the swimming performance test 

 before being used in the present blood chemistry 

 work. Therefore, to minimize short-term blood 

 chemistry changes due to the swimming challenge, 

 each experimental and control group was placed 

 in a control tank for 24 hr before blood sampling 

 (Grant and Mehrle, 1972; Wells, 1932; Nakano and 

 Tomlinson, 1967; and Hill and Fromm, 1968). 



Fish were then individually anesthetized with 

 neutralized MS-222 (tricaine methanesulfonate) 

 (Wedemeyer, 1970a) in a concentration sufficient to 

 reach plane three anesthesia (Klontz and Smith, 

 1964), [in about 5 min]. The fork length of each fish 

 was measured. Blood was then collected from the 

 severed caudal peduncle and processed after the 

 method of Miles and Smith (1968) except that plain 

 hematocrit tubes were used to obtain serum. These 

 fish were then weighed and examined for signs of 

 gas bubble disease (Dawley and Ebel, MS). Indi- 

 vidual serum samples were pooled within each ex- 

 perimental or control group and immediately sealed 



and frozen at 0°C until chemical analysis was 

 performed. Pooling was necessary to make up the 

 minimum sample volume needed for the following 

 analysis. 



A Technicon® SMA 12-60 (Sequential Multiple 

 Analyzer) was used to determine serum calcium, 

 phosphate, glucose, urea, uric acid, cholesterol, 

 total protein, total bilirubin, alkaline phosphatase 

 (AP-ase), lactate dehydrogenase (LDH) and glutamic 

 oxalacetic transaminase (SGOT). All samples were 

 run sequentially with a reference synthetic serum 

 (General Diagnostic's Calibrate Automated 

 "Lock-in"), while three control synthetic sera 

 (General Diagnostic's Calibrate) were used to cali- 

 brate the output of each of the 12 channels. Mea- 

 surement error was less than 1.0% with the excep- 

 tion of cholesterol and LDH which were 1.2%. 

 Serum calcium was determined by the method of 

 Kesler and Wolfman (1964) as modified by Techni- 

 con®. Inorganic phosphate was detected by a 

 Technicon® modification of the method of Fiske and 

 Subbarow (1925). Glucose was detected by a 

 modification of the procedures of Brown (1961), 

 and Bittner and McCleary (1963). Urea nitrogen was 

 determined by a modification of the method of 

 Marsh et al. (1965). Uric acid is measured by a 

 Technicon® modification of the cupric-neocuproine 

 procedure originally described by Bittner et al. 

 (1963). A Technicon® modification of the Lieberman- 

 Burchard reagent was used in the direct determina- 

 tion of serum cholesterol. Total protein was 

 quantitated by the Technicon® modification of the 

 biuret reaction. The HABA anionic dye |[(2-(4' 

 - hydroxyazobenzene) benzoic acid)]} used in the 

 Technicon® SMA 12/60 was unusable for salmonid 

 albumin, and a manual procedure using the brom- 

 cresol green method of Doumas et al. (1971) was 

 substituted in these experiments. The method for 

 the estimation of total bilirubin is based on a Tech- 

 nicon® modification of the procedure of Jandrassik 

 and Grof (1938). Alkaline phosphatase was deter- 

 mined by a modification of the King-Armstrong 

 procedure developed by Marsh et al. (1959). LDH 

 was measured by a method based on the procedure 

 of Hochella and Weinhouse (1965). SGOT was 

 quantitated after the procedure of Morgenster.n 

 et al. (1966). 



Serum sodium and potassium were measured 

 on a Corning Model 170 digital flame photometer, 

 while serum chloride was determined by the method 

 of Schales and Schales (1941). Correlation coeffi- 

 cients were calculated for all of the above blood 

 characteristics along with oxygen, nitrogen and 

 TGP saturations, and mean wet weight and fork 

 length. 



Water analysis for nitrogen and oxygen were 

 performed daily in each test tank using the methods 



Changes in Blood Chemistry 97 



