L .L. Coiro et al. J. Exp. Mar. Biol. Ecol. 247 2000 243 –255
249
where FG is the fractional growth of the constant exposure which corresponds to
max D.O.
the high D.O. concentration of the fluctuation, FG is the fractional growth of the
min D.O.
constant exposure which corresponds to the low-D.O. concentration of the fluctuation, and exposure time is the portion of the whole cycle duration spent at each condition.
In this study, the cyclic exposure time was 50:50 which results in an estimate equivalent to 50 of the adverse effects observed in the continuous low-D.O exposure.
3. Results
In each of the five tests with larval P . vulgaris, each low-D.O.treatment, whether
fluctuating or constant, reduced growth significantly relative to the saturated-D.O. control P ,0.003 – Duncan’s Multiple Range Test Table 1. At the lowest D.O.
exposures in both the constant and fluctuating treatments 1.6 and 1.4 mg l, growth rates were only 22 and 28 of the controls, respectively. Even at the highest hypoxic
D.O. levels tested 3.4 and 3.2 mg l, the growth rates were only 79–85 of the saturated controls.
The results of the five independent sets of exposures were considered jointly to provide a general evaluation. Analysis of covariance showed that there was a significant
difference P ,0.02 in growth impairment between cyclic exposures and constant low-D.O. exposures. Cyclic exposures usually resulted in better growth than corre-
sponding constant exposures at D.O. concentrations equal to the minimum of the cycle. In comparison with the constant exposures, mean growth effects in the cyclic exposures
ranged from 53 to 93 of the corresponding reduced growth of the constant low-D.O. exposure for paired data Table 3. Considering all pairs of constant and cyclic
exposures with equal minima, the average growth impairment for the cyclic exposure was 72.6 of that caused by constant exposure. A similar pattern of response, with
growth impairment effects exceeding those estimated by time-weighted averaging, was also seen in tests conducted with other species and life stages Tables 2 and 4.
Fig. 1 shows larval P . vulgaris fractional growth data for all five sets of exposures
Table 3 Mean difference from control growth for paired cycling and constant exposures P
. vulgaris larvae Test
Low-D.O. concentration Cycling response as percent
[ 60.1 mg l
of constant minimum D.O. response
1 1.6
75.7 1.9
53.4 2
1.9 66.5
2.3 78.0
3 2.3
69.0 5
1.8 92.9
] Average fluctuating
72.6 response as percent of
constant response
250 L
.L. Coiro et al. J. Exp. Mar. Biol. Ecol. 247 2000 243 –255 Table 4
Mean difference from control growth of tested species for paired fluctuating and constant exposures, using pooled test results. Exposures considered paired if concentration is 60.2 mg l. D.O. of constant exposure
60.1 mg l D.O.for P. vulgaris larvae
Species Life stage
Cyclic response Number of
as percent tests
constant response
Palaemonetes vulgaris larval
72.6 5
Palaemonetes vulgaris juvenile
63.0 1
Dyspanopeous sayi larval
83.3 1
Paralichthys dentatus juvenile
61.0 2
plotted against D.O. concentration. The fluctuating and constant exposure data have been
2 2
marked with best fit logarithmic trend lines R 50.765 and R 50.945, respectively, the fluctuating exposure represented by the medium weighted line and the constant exposure
by the heaviest weighted line. The third lightly weighted trend line represents the time-weighted-average response curve based on 50 of the constant exposure response,
calculated using Eq. 3. Fig. 2 also represents the actual fluctuating and constant exposure data, and their respective trend lines as well as the estimated TWA response
curve, but with the TWA adjusted to represent 72.6 of the constant exposure response,
Fig. 1. Actual fluctuating h and constant d growth response and estimated TWA response based on 50 of the constant, as calculated using Eq. 3.
L .L. Coiro et al. J. Exp. Mar. Biol. Ecol. 247 2000 243 –255
251
Fig. 2. Actual fluctuating h and constant d growth response and adjusted estimated TWA response based on 72.6 of the constant, as calculated using Eq. 3.
based on the results from these tests, which more closely represents the actual fluctuating response.
4. Discussion