312 
Fishery Bulletin 1 14(3) 
the highly successful recruitment of juvenile copper 
rockfish observed in Puget Sound in 2006 (LeClair et 
ah, 2007; Palsson et ah, 2012). Although the numbers 
of juveniles did not influence the overall pattern of 
change in seasonal abundance for either species, juve- 
nile recruits during one or more years may have been 
plentiful enough to account for the overall statistically 
significant greater proportion of small fish (< 20 cm) 
observed in the summer. Other potential explanations 
for the disproportion include seasonal mortality, preda- 
tion, and movement. 
Washington et al.^ and Gowan (1983) reported fe- 
male first-spawning lengths for both brown and cop- 
per rockfish in Puget Sound to be in excess of 20 cm 
(age 3-4 years) and our observations at PHAR are 
in general agreement, although a single late-stage 
gravid brown rockfish less than 20 cm was encoun- 
tered in our study and we have observed late-stage 
gravid brown rockfish elsewhere in Puget Sound as 
small as 18 cm (determined by cannulation). All of the 
late-stage gravid brown rockfish sampled by Hess at 
al. (2012) were in excess of 20 cm. In our study, the 
percentages of late-stage gravid rockfish in relation 
to the number of potential female spawners are con- 
servative because some fish >10 cm would not have 
reached maturity. The higher numbers of late-stage 
gravid brown rockfish encountered by Hess et al. 
(2012) than the numbers we observed reflect an inten- 
tional sampling bias toward gravid rockfish in their 
study. 
We observed more late-stage gravid brown rockfish 
in the spring {N=35) than in the summer (N=22). The 
larger length classes comprised a statistically signifi- 
cant greater proportion of the spring observations, 
and this is consistent with Bobko and Berkeley (2004) 
and Love et al. (1990), who found that parturition oc- 
curs earlier for older black iS. melanops) and yellow- 
tail rockfish. Cooper (2004) also found the same to be 
true for copper rockfish in Puget Sound. Because of 
sampling bias, we did not examine length-by-season 
for the late-stage gravid brown rockfish sampled by 
Hess et al. (2012). Nevertheless, the mean length of 
the late-stage gravid fish sampled in that study falls 
near the center of the length class that accounted for 
most of our observations of late-stage gravid brown 
rockfish. In Puget Sound, peak parturition is known 
to occur earlier in the year for copper rockfish than 
for brown rockfish (DeLacy et al.^°; Washington et 
al.^). Curiously, we did not observe any late-stage 
gravid copper rockfish in the spring at PHAR. Of the 
13 late-stage gravid copper rockfish observed, 10 were 
encountered during the final 2 years of the study and 
only 2 of those were in the largest length class. On 
the basis of length-frequency changes over time and a 
substantial increase in copper rockfish abundance over 
*°DeLacy, A. C., C. R. Hitz, and R. L. Dryfoos. 1964. Mat- 
uration, gestation, and birth of rockfish (Sebastodes) from 
Washington and adjacent waters. Wash. Dep. Fish. Fish. 
Res. Pap. 2:51-67. 
the study period after 2006, we surmise that most 
of the observed late-stage gravid copper rockfish be- 
longed to the strong 2006 year class noted above. If 
so, they were just reaching maturity during the final 
years of the study and may have been spawning later 
in the season, as has been noted for black, yellowtail, 
and copper rockfish (see above). 
Consistent with many habitat selection models, the 
results of Matthews (1990a, 1990b) indicated that the 
apparent homing ability of some rockfish species may 
enable them to embark on periodic exploratory excur- 
sions in response to unfavorable changes in habitat, 
allowing them to assess other environments but re- 
turn to their point of departure if more suitable sur- 
roundings are not encountered. Matthews (1990c) fur- 
ther noted that the winter disappearance of canopy- 
forming bull kelp, with the structure and associated 
prey it provided, may have explained the seasonal 
exodus of brown and copper rockfish she observed on 
naturally formed low-relief reefs. Although bull kelp 
does not occur at PHAR, the seasonal presence of non- 
floating seaweeds may provide similar levels of refuge 
and prey. The study areas of Matthews (1990c) includ- 
ed a high-relief artificial reef (Boeing Creek Artificial 
Reef) also located in the main basin of Puget Sound. 
The reef is comparable in age, size, depth, and con- 
struction to PHAR; is subject to similar wave energy, 
current, and temperature regimes; and supports a 
similar ichthyofauna and flora that is devoid of cano- 
py-forming kelp. However, the highest densities of >20 
cm brown and copper rockfish recorded by Matthews 
at that site occurred during the fall and winter, not 
during the summer as observed in our study. 
Resiliency to temperature fluctuations is not known 
for copper rockfish. However, Wilson et al. (1974) stud- 
ied metabolic compensation in response to temperature 
in brown rockfish, and vermilion rockfish {S. miniatus). 
This latter species resides below the thermocline and 
is therefore not exposed to the same seasonal tempera- 
ture fluctuations experienced by brown rockfish resid- 
ing above the thermocline. Wilson et al. concluded that 
there are metabolic differences between the 2 species 
that correlate with differences in depth distribution, 
and that brown rockfish have a higher capacity to accli- 
mate over a wider range of temperatures. Both brown 
and copper rockfish have a similar biological range, oc- 
curring from the subtropics to the subarctic (Horn et 
al., 2006) and are found in warmer inland seas, as well 
as colder oceanic waters (Love et al., 2002). The mean 
daily recorded temperature changes at PHAR were 
highest during the spring when both species began 
appearing on the reef, and during the summer when 
they appeared on the reef in their greatest abundance. 
Although we do not have temperature data beyond 
the offshore perimeter of the reef, water column data 
(available at website) in East Passage conducted by the 
Washington State Department of Ecology indicate that 
diurnal and seasonal temperature regimes become less 
labile with increasing depth. If, as noted by Neill and 
Gallaway 1989, fish move in response to the totality of 
