Anweiler et al.: Effects of temperature and hypoxia on the metabolic performance of Morore saxatilis 
341 
before transfer to a 90-L swim tunnel respirometer 
(Loligo Systems, Tjele, Denmark) containing water at a 
salinity of 1 and the treatment temperature. Blinders 
were placed over the front, working portion of the inner 
chamber to create a darkened environment for the fish 
and prevent accidental startling from personnel mon¬ 
itoring the swim chamber. Approximately 30 min was 
allowed for recovery from the effects of anesthesia, and 
recovery was defined as the return of fish to resting on 
the bottom of the respirometer in a normal posture and 
with normal patterns of resting gill ventilation. 
After recovery, flow in the tunnel was set to 10 cm/s, 
and fish were left overnight (-1600-0900). The overnight 
period allowed fish to recover from any stress due to han¬ 
dling or effects of anesthesia because the results of prelim¬ 
inary overnight trials indicate that oxygen consumption 
returned to baseline approximately 5 h after transfer. The 
overnight period also allowed a fish time to orient itself in 
the swim tunnel against a low-velocity flow. The darkened 
front portion of the chamber aided fish in using the work¬ 
ing portion of the chamber instead of resting against the 
back grid. During the overnight recovery period, the swim 
tunnel respirometer was continually flushed with water 
from the outer bath. An air stone was placed in the outer 
bath to achieve a constant ambient DO level of over 90%. 
On the morning of each trial, a DO galvanic probe 
(MINI-DO, Loligo Systems; with accuracy of ±1% air sat¬ 
uration) was calibrated with a 2-point calibration method 
and secured in the inner chamber of the flume through a 
port. A bright light was placed over the back portion of the 
swim chamber to further encourage fish to use the dark¬ 
ened portion of the swim chamber. Water in the outer bath 
was deoxygenated by bubbling nitrogen gas through a high- 
pressure air stone. A solenoid valve, linked to the DO gal¬ 
vanic probe, controlled the flow of nitrogen gas in the outer 
chamber. The oxygen saturation in the inner chamber was 
lowered from the ambient level by continuously flushing 
in deoxygenated water from the outer bath. The treatment 
DO level was typically reached within 1 h. Once the treat¬ 
ment DO concentration was reached, fish were held at this 
concentration for 30 min to allow for acclimation. 
After acclimation, flow in the chamber was increased 
and intermittent respirometry was used to repeatedly 
measure oxygen consumption. Oxygen consumption mea¬ 
surements were recorded in milligrams per kilogram per 
hour in 7-min intervals that consisted of a 2-min flush, fol¬ 
lowed by a 1-min equilibration period then a 4-min mea¬ 
surement period in which oxygen saturation was recorded 
once every second. The flush cycle served, among other 
things, to restore oxygen that had been depleted during 
each equilibration and measurement period. The speed of 
water flow in the chamber was increased by 15 cm/s every 
42-56 min, resulting in 6-8 oxygen consumption measure¬ 
ments at each speed. At least 3 of those measurements 
were more than 20 min after the speed change; these mea¬ 
surements were required for SMR calculations (see the 
next section). 
Trials ended when the fish was exhausted, which was 
defined as failure to swim against the current and use 
of the tail to maintain position against the back screen 
for at least 15 s. A recent review of methods for eliciting 
MMR found that the swim-tunnel method is acceptable 
for species, such as the striped bass, that are active swim¬ 
mers (Norin and Clark, 2016). The timing of exhaustion 
was calculated by subtracting the time that we increased 
speed for the first speed increment of the trial from the 
time at which the fish became exhausted. After exhaus¬ 
tion, flow in the chamber was decreased to 10 cm/s, water 
was brought to a DO level of 100%, and fish were left 
to rest for 15 min before being transferred to the recov¬ 
ery tank. After the fish was removed, the chamber was 
closed again and 3 measurements of background oxy¬ 
gen consumption were recorded by using the previously 
described 7-min cycle. 
Metabolic calculations 
Mean background bacterial oxygen consumption was 
appreciable at each temperature (20°C: 12.1 mg/h [SD 2.8]; 
25°C: 28.9 mg/h [SD 8.9]; and 32°C: 32.6 mg/h [SD 12.3]). 
Note that these levels represent a substantial portion of 
the measured oxygen consumption values (39% [SD 12] 
when measuring SMR and 13% [SD 6] when measuring 
MMR, averaged across all temperature and DO-level treat¬ 
ments). Although these values are relatively high, they 
likely occurred because 1) fish were acclimated in the respi¬ 
rometry system for -12-16 h prior to measurements being 
recorded and 2) the dechlorinated tap water used to fill the 
flume was held in 200-L bins for 2-3 d prior to testing to 
allow it to reach the testing temperature (also see Svendsen 
et al., 2016). Therefore, the average background bacterial 
oxygen consumption (computed from the 3 measurements 
made for that trial) was subtracted from the total recorded 
oxygen consumption to calculate the oxygen consumption 
of the fish (Svendsen et al., 2016). 
The SMR for each individual was calculated by fitting 
a linear function to the log of oxygen consumption on 
swimming speed (Fig. 2) and then by using this function 
to extrapolate the SMR at a speed of zero (Brett, 1964; Fry, 
1971). Only oxygen consumption values measured 20 min 
after each speed change were used in the calculations for 
SMR. The MMR was the single maximum oxygen con¬ 
sumption rate achieved by the fish during the entire trial. 
This reading occurred approximately 50% of the time at 
the maximum speed step and the other 50% of the time 
at the speed step just before the maximum. The AMS for 
each fish was calculated as follows: 
AMS = MMR - SMR, (1) 
where AMS = aerobic metabolic scope (mg-kg _1 -h _1 ); 
MMR = maximum metabolic rate (mg-kg _1 -h _1 ); and 
SMR - standard metabolic rate (mg-kg _1 -h _1 ). 
Statistical analysis 
All analyses were completed in JMP 13 (SAS Institute Inc., 
Cary, NC). Differences in weight and relative condition fac¬ 
tor (Kn; Le Cren, 1951) between treatments were analyzed 
