86 
Fishery Bulletin 99(1) 
1997 
1998 
Figure 2 
Total density (number/100 m 3 ) of larval fishes by 
egg type and sample date in Penobscot Bay, Maine: 
4-7, 15-16, and 29-30 April; 12-13, 28-29 May; 
9-10 and 24-25 June 1997; and 18 March, 7, 15, 30 
April 1998. 
larvae (>35% of the samples) were rock gunnel ( Pholis 
gunnellus), sea raven ( Hemitripterus americanus), long- 
horn sculpin (Myoxocephalus octodecimspinosus), grubby 
(M. aenaeus), and Ammodytes sp., whereas P virens and 
S. punctatus were the least common (<3%). Myoxocepha- 
lus octodecimspinosus (30% of total number), P. gunnellus 
(27%), and Ammodytes sp. (11%) were the most abundant 
larvae collected. The greatest number of species (12) were 
found during mid-March (981). Densities of all larvae com- 
bined declined from 11.4 larvae per 100 m 3 in mid-March 
(981) to about one larva per 100 m 3 in mid-April (98III). 
The high larval densities in mid-March resulted from the 
high numbers of M. octodecimspinosus and P. gunnellus 
larvae collected at that time. 
Larvae originating from demersal eggs dominated the 
catches in both years (Table 3). In terms of abundance, 
most larvae collected in 1997 came from demersal eggs 
(70%) that dominated the catch at all stations. Larvae 
from demersal eggs composed most of the catch (>90%) 
until mid (79%) and late June (60%) when the number 
of larvae originating from pelagic eggs increased (Fig. 2). 
Overall, five of 23 species (22%) of larvae came from pe- 
lagic eggs in 1997, but by late June (97VII), four of the 
eight species (50%) collected hatched from pelagic eggs. 
Larvae from demersal eggs dominated the catches in 1998 
to an even greater extent than in 1997 (Fig. 2). Two of 
the 16 species (13%) and four larvae (0.04%) collected in 
1998 originated from pelagic eggs (three specimens of G. 
morhua and one of P. virens). Ulvaria suhbifurcata were 
the most abundant larvae in 1997 with densities up to 9.1 
per 100 m 3 at the lower bay stations, whereas in 1998, M. 
octodecimspinosus and P gunnellus were the most abun- 
dant larvae, with station densities up to 12.9 and 6.1 lar- 
vae per 100 m 3 , respectively. Densities of sand lance (Am- 
modytes sp.) were highest (4 per 100 m 3) at the upper bay 
and midbay stations in 1997 and 1998, respectively. Pleu- 
ronectes americanus were more abundant in 1997 with 
densities up to 1.7 larvae per 100 m 3 at the midbay sta- 
tions (B2, B3, B4). Larvae from pelagic eggs were rare in 
both years. 
Several species appeared in plankton samples from most 
sampling dates in 1997 (Table 4). One species, H. platte- 
soides was taken during all seven cruises but was absent 
from the 1998 collections. Ammodytes sp. occurred from 
early April (971) through late May (97V) in 1997 and from 
mid-March (981), early April (9811), and late April (98IV) 
in 1998. Pleuronectes americanus and L. atlanticus were 
present from mid-April (9711) to late June (97VII) and 
H. americanus and P. gunnellus were found from early 
April (971) to mid June (97VI). Four of the five species 
of fish larvae originating from pelagic eggs were taken 
during only one cruise. In 1998, three species, P. gunnel- 
lus, H. americanus and M. aenaeus, were collected on all 
four cruises. Atlantic herring ( Clupea hare/igus) and wry- 
mouth ( Cryptacanthodes maculatus ) occurred from mid- 
March (981) through mid April (98III). Myoxocephalus oc- 
todecimspinosus, M. scorpius, and L. coheni were taken 
only during the first two cruises. 
Size of the most abundant species changed little over the 
survey time period (Table 4). For example, winter floun- 
der larvae were 2-4 mm during mid April (9711) and 3-7 
mm during late June (97VII) indicating successive spawn- 
ing events or no observed growth. The length-frequency 
distributions of most larvae were not significantly differ- 
ent among sample dates (Kolmogorov-Smirnov two sam- 
ple test, P>0.05). However, the proportion of larger larvae 
increased over time for three taxa collected over a two 
month period in 1997 (i.e. Ammodytes sp., P. gunnellus, 
H. plattessoides. Table 4). Few species showed a change in 
length frequency distribution during the shorter 1998 sur- 
vey (Kolmogorov-Smirnov two sample test, P>0.05). 
Temporal changes in the abundance of fish larvae were 
observed in both years. Two peaks in mean density of lar- 
val fish were observed among all locations in 1997, where- 
as only a single peak was seen during the shorter 1998 
survey. In 1997, the mean density was initially about 1.9 
larvae per 100 m 3 in early (971) and mid-April (9711), then 
declined to 0.6 larvae per 100 m 3 by mid May (97IV). Lar- 
val density increased to 1.2 larvae per 100 m 3 by late May 
(97V) and peaked at 3.9 larvae per 100 m 3 in late June 
(97VII). Most of the initial reduction in density was due 
to fewer Ammodytes. sp., C. maculatus, and H. americanus 
in the collections, whereas the June increase was due to 
U. subbiffurcata, and to a lesser extent to E. cimbrius and 
P. americanus. Densities of fish larvae in 1998 declined 
from a peak of 11.4 larvae per 100 m 3 in mid-March (981) 
to about 1 larva per 100 m 3 in mid-April (98III), before 
increasing to 1.4 larvae per 100 m 3 on late April. Fewer 
Ammodytes. sp., C. maculatus . L. coheni, the three Myoxo- 
cephalus sp., and P. gunnellus accounted for the initial re- 
duction. The increase during the latter part of the 1998 
survey was due to greater numbers of L. atlanticus, P. 
americanus, and U. subbiffurcata larvae in collections. 
Relative abundance patterns of 12 of the most abundant 
taxa (by the 5% criterion) in 1997 and 13 taxa in 1998 
