214 
Fishery Bulletin 107(2) 
Table 4 
Simple linear regression parameters relating body composition (response variables) to oil globule volume for quillback rockfish 
( Sebastes maliger ) embryos and preparturition larvae. Each sample was a composite of hundreds of larvae from the same parent 
(n=ll maternal females). Only data from samples for which protein and lipid analyses were both completed were included. 
Response variable 
Slope 
Intercept 
r 2 
ANOVA F 
P 
Lipid mass 
1.13 
0.0133 
0.672 
F J9 =18.45 
0.002 
Lipid (% wet mass) 
247 
-0.101 
0.943 
F 7 9 =150.14 
<0.001 
Lipid (% dry mass) 
403 
20.3 
0.701 
F ;9 =99.41 
0.001 
Protein mass 
1.40 
0.0537 
0.432 
F ;9 = 6. 85 
0.028 
the PUFA 20:4n-6 (arachidonic acid). No groups of FAs 
based on degree of saturation were apparently depleted 
more rapidly than others, as the percentage mass losses 
of SFAs, MUFAs, and PUFAs were approximately equiv- 
alent to the percentage of total FA mass loss (Fig. 5). 
Some FAs showed relatively little contribution to total 
FA mass loss despite having large initial masses, indi- 
cating that they were conserved, particularly the SFA 
18:0; and the PUFA 20:4n-6 (Eq. 3, Table 5). Mean- 
while, the largest absolute mass losses were found for 
the SFA 16:0 (palmitic acid); the MUFA 18:ln-9 (oleic 
acid); and the n-3 PUFAs 22:6n-3 (docosahexaenoic 
acid, DHA) and 20:5n-3 (eicosapentaenoic acid, EPA), 
which together accounted for 71% of the total loss in 
FAs. Thus, there were clear differences in the contribu- 
tions of different FAs to the overall lipid use. 
| 
Discussion 
We found that while both lipid and protein mass are 
consumed by quillback rockfish embryos during develop- 
ment, lipid is used more rapidly and contributes a larger 
portion of total energy than protein. This is consistent 
with results from other studies of rockfish, and affirms 
the importance of measuring lipid levels when assess- 
ing larval condition. However, we also found dif- 
ferences in the specific rates of use of protein and 
lipid compared to other rockfish, which illustrates 
the diversity of patterns of energy use and changes 
in body composition among species. 
In our study, OGV was highly correlated with 
lipid content. This relationship could be important 
for future studies researching the energetic status 
of rockfish embryos and preparturition larvae. 
Using OGV as an indicator of energy reserves at 
any stage of development, and knowing the rela- 
tionship between OGV and developmental stage, 
may allow for interpreting the energetic health 
of embryos at any developmental stage. This is 
a considerable advantage for field-based studies, 
given the difficulty of capturing significant num- 
bers of gravid females with embryos or larvae at 
the same developmental stage, and the risks of 
introducing experimental effects when parents 
are held until larvae are released. Our results 
also illustrate that indicators of condition applied 
to different species should be interpreted with 
differences in their biochemistries in mind (e.g., 
in quillback rockfish OGV is strongly related to 
total lipid, whereas in black rockfish the two are 
unrelated) (Berkeley et al., 2004). 
Our study represents the first attempt to char- 
acterize FA use during embryogenesis for a rock- 
fish species. Although aquaculture studies have 
investigated FA requirements for rockfish, these 
have typically involved manipulating the diets of 
adults and juveniles (e.g., Lee, 2001) and likely 
Figure 4 
Multidimensional scaling plot of quillback rockfish ( Sebastes 
maliger ) early-stage embryos (•) and hatched, preparturition 
larvae (O) according to their fatty acid compositions based on 
an Aitchison distance matrix. Developmental stage for each 
sample is given in parentheses. Each sample was a composite 
of hundreds of larvae from the same parent (n = 8 maternal 
females). Comparisons only included those samples for which 
lipid data were available. 
