Bertram et al.: Growth and development during the early life stages of Pleuronectes americanus 
7 
The relationship between larval growth rate and 
larval-period duration, however, may not be straight- 
forward. Although larval growth rate is the primary 
factor influencing larval-period duration, its effects 
appear to be modified by the duration of the latency 
period. Larvae that grow rapidly tend to reach maxi- 
mum larval length at an early age. However, indi- 
viduals that reach maximum larval length at an early 
age have a longer latency period than those larvae 
that reach maximum length late in the larval pe- 
riod. This finding suggests that rapid growth is as- 
sociated with a long latency period. Slower-growing 
larvae, in comparison, may reach metamorphosis at 
a later age but have a considerably shorter latency 
period. Moreover, this suggests that metamorphosis 
may require a minimum duration, independent of 
size. These findings are consistent with Ricklefs’ 
(1973, 1979) hypothesized tradeoff between growth 
rate and the acquisition of mature tissue function in 
birds. In this connection, it is noteworthy that a 
tradeoff between growth rate and the rate of protein 
turnover has been documented for the mussel Mytilus 
edulis (Hawkins et al., 1986). Our findings are also 
consistent with Balon’s (1990) suggestion that 
through epigenetic interactions, individuals within 
a clutch may form distinct developmental groups — 
some being more altricial and others more precocial. 
Growth rate estimates will be biased if mortality 
is size dependent. Biased growth-rate estimates will, 
in turn, reduce estimates of variation in growth rate. 
However, the extent of variability in larval growth 
rates reported here are not due to size-dependent 
mortality. Our analysis could not detect size-depen- 
dent mortality, and there was no reduced variability 
in growth until the end of week 4. Survival to meta- 
morphosis was relatively high (26 out of 53 [49%] 
from family 1) for individuals replaced on day 22. 
High survival from day 1 to metamorphosis (175 out 
of 400 [44%]) was also observed for group-reared lar- 
vae in one of the rearing aquaria for family 3. Impor- 
tantly, the CV for size and age at metamorphosis was 
similar for the individual and group-reared larvae. The 
CV for age at metamorphosis was 0.14 and 0.17 for 
individually reared and group reared larvae ( n =175), 
respectively. The CV for size at metamorphosis was 
0.045 and 0.046 for individually reared and group- 
reared larvae (n=175), respectively. (Note that the CVs 
for age and size at metamorphosis for the full data set 
of group reared larvae [n=205] were indistinguishable 
from those reported above for the reduced data set 
[n = 175]). The similarity of CVs for age and size at 
metamorphosis implies similar scope for variation in 
growth rate despite differences in the rearing protocol. 
We used a small number of female broodstock. This 
small number of fish, however, did not preclude in- 
sight into the potential for variation in larval growth 
and development at the population and species level. 
Previous research on early life history traits in win- 
ter flounder has shown that most of the total varia- 
tion in metamorphic traits (age and size at meta- 
morphosis) occurred within rather than among ma- 
ternal families (Chambers and Leggett, 1992). Dif- 
ferences between families in the relationship between 
size and age at metamorphosis (Chambers and 
Leggett, 1987; see below) and length at metamor- 
phosis (Chambers and Leggett, 1992, Bertram et al., 
1993), although detectable, appear small in compari- 
son with the similarities between families for varia- 
tion in age at metamorphosis. Indeed, Chambers and 
Leggett (1992) reported that most variation in age 
at metamorphosis resided within each rearing 
aquarium. The CV’s for age and size at metamor- 
phosis reported here are similar to those reported by 
Chambers and Leggett (1987) despite differences in 
rearing temperatures and origins of broodstock em- 
ployed in the two studies. Chambers and Leggett 
( 1992) developed several qualitative expectations for 
parental influences on larval flatfishes. They sug- 
gested that parentage is likely to influence larval 
traits but that its contribution to the total pheno- 
typic variation in larvae is expected to diminish dur- 
ing the larval period. In addition, the degree of pa- 
rental influence is likely to be trait-specific. The ab- 
sence of parental effects on important traits such as 
larval-period duration (age at metamorphosis) sup- 
ports the potential generality of our results on early 
life history traits based on few parents. Moreover, in 
the absence of field data, laboratory-based research 
such as this represents the only basis for character- 
izing and predicting the dynamics of patterns of in- 
dividual larval growth and development. 
The survival consequences of individual variation 
in larval growth and development reported here are 
presently unknown. We do not know whether indi- 
viduals that grow rapidly and metamorphose at an 
early age have a survival advantage over those that 
grow more slowly and metamorphose at an older age. 
Despite the limited supporting evidence, there has 
been widespread acceptance of the hypothesis that 
rapid larval growth conveys a survival advantage 
because those individuals are large at age and often 
have a reduced larval-period duration (Bertram, 
1993; Leggett and Deblois, 1994). D ’Amours (1992) 
tested directly the hypothesis that rapid larval 
growth increases survival by using wild 0-group ( 17— 
47 d) Atlantic mackerel, Scomber scombrus. Compar- 
ing the otolith microstructure of larvae from one co- 
hort captured at two different intervals in time, he 
found no evidence of higher survivorship among 
faster-growing larvae. In addition, two studies have 
