Sewall and Rodgveller: Changes in body composition and fatty acid profile during embryogenesis of Sebastes maliger 
217 
bryos developing inside the mothers’ bodies are supplied 
with maternal nutrients (viviparous), or rely entirely on 
their yolk sacs (ovoviviparous). Quillback rockfish have 
been described as viviparous (MacFarlane and Norton, 
1999), and ovoviviparous (Matala et al., 2004). Previ- 
ous research using radiocarbon-labeled amino acids 
found that embryos of black rockfish (S. melanops) took 
up nutrients from intraovarian fluid, but only at very 
late stages of development — presumably after they had 
hatched and their mouths and digestive systems were 
sufficiently functional (Yoklavich and Boehlert, 1991). 
MacFarlane and Bowers (1995) also found evidence of 
matrotrophy (postfertilization maternal nutrient pro- 
visioning) occurring in yellowtail rockfish because a 
radio-labeled phospholipid was transferred from moth- 
ers to embryos before their mouths opened, and the 
amount increased as they developed. Reviews of these 
and other studies have thus supported viviparity in 
rockfish (e.g., Parker et al., 2000). The reduction in dry 
tissue mass seen among the quillback rockfish embryos 
in our study was lower than the 25% to 55% range of 
dry mass losses typically seen in strict lecithotropes 
(MacFarlane and Bowers, 1995), which rely entirely on 
nutrients provided to the egg before fertilization, sug- 
gesting that quillback rockfish are also partly matrotro- 
phic. The degree to which nutrition is obtained from the 
yolk rather than from maternal intraovarian fluids is 
unclear for quillback rockfish; therefore it is important 
to view data regarding mass loss and energy use given 
here as minimums. 
It is likely that maternal traits (e.g., the size and 
age of the female parent) influence the biochemical 
compositions of rockfish embryos and larvae (Sogard et 
al., 2008). This introduces the possibility of maternal 
effects confounding the relationship between develop- 
mental stage and body composition (e.g., if our samples 
were biased towards larger females yielding the earlier 
stages of embryos). However, it is likely that develop- 
mental processes accounted for most of the differences 
that we found between early-stage embryos and hatched 
larvae. Developmental stage showed a much stronger 
relationship to lipid concentration (r 2 = 0.87) than did 
maternal length (r 2 =0.39). Maternal length was only 
weakly correlated with developmental stage (r 2 =0.26), 
and this correlation was largely driven by the pres- 
ence of one large fish with stage 10 larvae. Removing 
this fish and its larvae resulted in virtually no rela- 
tionship between maternal length and developmental 
stage (r 2 = .15). This highlights one of the conclusions 
that can be drawn from our data: developmental stage 
should be accounted for when investigating maternal 
effects among wild caught fish with progeny at various 
stages. 
If quillback rockfish preferentially use lipid as an 
energy source over protein, it would be useful to inves- 
tigate how various maternal traits influence the relative 
rates of lipid and protein loss in embryos. For example, 
do embryos from older, larger, or fatter parents have 
greater lipid reserves, and do they exhibit lower rates 
of protein loss? 
Why not simply use size or total energy content as 
indicators of viability? Such an approach is indicated 
by our finding that changes in total energy content per 
larva largely reflected changes in dry mass from early 
to late stages, rather than changes in the proportions 
of lipid and protein. In addition, there is great vari- 
ability among species in the size and energy content of 
eggs and embryos — the early stage quillback rockfish 
embryos in our study were on average more than 2.6 
times heavier on a dry mass basis than yellowtail rock- 
fish embryos (Eldridge et al., 2002). Greater larval size 
may also confer advantages through reduced predation 
and increased range of feeding opportunities, and was 
probably the force driving the uptake of water during 
early development that we observed. However, various 
studies have found no relationship between egg size and 
offspring viability (reviewed in Kamler, 1992). Straight- 
forward interpretation of the relationship of egg or em- 
bryo size and total energy content to larval viability 
is confounded by findings suggesting that larvae from 
smaller eggs often use yolk energy for growth more ef- 
ficiently than those from larger eggs, and may undergo 
compensatory growth in later development (reviewed in 
Kamler, 1992). Even under conditions of food scarcity, 
where larger larvae may be expected to be at an advan- 
tage, results have been inconsistent; for example, larval 
length did not appear related to starvation resistance of 
black rockfish larvae (Berkeley et al., 2004). 
Oil globule volume 
Given the importance of lipid as an energy source for 
developing quillback rockfish embryos, the strong cor- 
relation of OGV with total lipid we found suggests that 
OGV may serve as an indicator of energetic status. Some 
maternal trait, such as age (e.g., Berkeley et al., 2004), 
may strongly influence OGV and be responsible for the 
variability. Investigating changes in the lipid class com- 
ponents (e.g., TAG and polar lipids) of the oil globules, 
as well as whole embryos, could provide information 
useful for better understanding the relationship of the 
oil globules to condition. The strength of the relationship 
between OGV and larval survival should also be inves- 
tigated experimentally with quillback rockfish larvae. 
Using OGV as an indicator of energetic status represents 
a potentially large savings in resources required, com- 
pared with analytical chemistry techniques. 
Fatty acid profile 
The major FA components of the lipids in quillback 
rockfish embryos and larvae were generally similar 
to those reported elsewhere for many species of adult 
fish (reviewed in Tocher, 2003): predominantly the n-3 
PUFAs 22:6n-3 and 20:5n-3; 20:4n-6 as the main n-6 
PUFA; large quantities of the MUFA 18:ln-9; and 16:0 
and 18:0 as the main SFAs. Previous researchers have 
also reported high levels of n-3 PUFAs in marine fish 
eggs (e.g., Tocher & Sargent, 1984); however, there can 
be marked interspecific differences in the precise order 
