Johnson et al.: Reproductive success of Clupea pallasi after the Exxon Valdez oil spill 
757 
PWS SE AK 
Age (years) 
Figure 6 
Mean (+SE) percent live larvae of Pacific herring by female parent age, site, 
and region in Alaska, 1995. Sample sizes are indicated in Figure 5. Overall P- 
value from ANOVA is listed above each graph. Significant differences were 
Rocky Bay, ages 3 and 4 < age 5, 6, and 7 (P<0.047). 
stages of herring are more suscep- 
tible to the effects of oil according to 
laboratory (Linden, 1978; Pearson et 
al., 1985; Carls, 1987; Kocan et al., 
1987; Rice et al., 1987) and field stud- 
ies (Brown et al., 1996a; Norcross et 
al., 1996). Abnormal larvae have poor 
survival potential (Kocan et al., 
1996a), and thus the exposure of eggs 
and larvae to oil in PWS may have 
resulted in increased mortality. Fur- 
thermore, the same oil concentra- 
tions that caused significant genetic 
damage also caused significant 
physical damage in developing em- 
bryos (Carls et al. 5 ); thus early death 
would likely preclude recruitment of 
genetically damaged individuals to 
spawning populations. 
Although genetic damage was de- 
tected in larvae collected in the oil- 
contaminated areas of PWS in 1989 
(Hose et al., 1996; Brown et al., 
1996a), we did not inspect larvae for 
genetic damage. Concomitant labo- 
ratory measurements of larvae that 
had been artificially contaminated 
indicated that genetic response was 
not a more sensitive measure of oil 
exposure than the parameters we 
examined (Carls et al. 5 ). In addition, 
artificial exposure of prespawning 
adults to relatively high oil concen- 
trations (58 ppb, initial PAH) did not 
cause genetic defects in artificially 
spawned progeny (Carls et al. 14 ). 
Other defects observed in larvae 
from PWS in 1989 included physical 
damage, assessed by scored indices 
(Hose et al., 1996). Carls et al. 5 ob- 
served that two of these indices, peri- 
cardial edema and finfold condition, 
were more sensitive to oil damage 
than was the genetic response. Be- 
cause we did not detect significant 
pericardial abnormalities in larvae 
from PWS six years after the spill, it 
is likely that the genetic condition of these larvae 
has not been adversely affected. 
The failure of the 1989 year class of herring in PWS 
to recruit to the spawning population may have been 
partly attributable to the spill, but it is impossible 
to separate oil effects from other natural factors. At 
the sites we sampled in PWS, the 1989 year class 
usually represented <4.0% of the spawning popula- 
tion (ADF&G 15 ). Larval survival in PWS was reduced 
an estimated 52% in 1989 as a result of the spill 
(Brown et al., 1996a); such loss supports inferences 
of poor survival that are based on laboratory obser- 
15 ADF&G (Alaska Department of Fish and Game). 1995. Her- 
ring test fishery data. Commercial Fisheries Management 
and Development Division, PO Box 669, Cordova, AK 99574. 
