84 
Fishery Bulletin 99(1) 
Table 2 
Correlations among environmental variables and factor scores from varimax-rotated PCA of the log-transformed environmental 
data. Underlined values significant at P = 0.05, adjusted for multiple comparisons such that P < 0.01 for a significant correlation. 
Depth 
Pvol 
Temp 
Salinity 
Factor 1 
Factor 2 
1997 
Depth 
1.00 
0.08 
0.98 
Pvol 
0.07 
1.00 
0.84 
-0.06 
Temperature 
-0.13 
-0.35 
1.00 
-0.68 
-0.05 
Salinity 
0.20 
0,44 
-0.24 
1.00 
0.71 
0.25 
1998 
Depth 
1.00 
0.08 
0.83 
Pvol 
-0.08 
1.00 
-0.90 
0.08 
Temperature 
0.18 
-0.71 
1.00 
0.90 
-0.09 
Salinity 
-0.16 
-0.29 
0.21 
1.00 
0.42 
-0.68 
Station mean water temperatures for the top 20 m of 
the water column in 1997 were variable and ranged from 
2.3°C on 4 April to 10.8°C on 25 June. Station mean salin- 
ity ranged between 27.8 ppt on 30 April and 31.2 ppt on 
16 April. Plankton volume ranged from 0.8 to 44.9 mL per 
100 m 3 and generally was greatest during early April (971) 
with a secondary peak in volume observed in late June 
(97VII). Greatest plankton volumes occurred at the lower 
bay stations. In 1998, mean water temperatures ranged 
from 2.4°C on 18 March to 5.3°C on 30 April. Mean salin- 
ity ranged between 29.6 ppt on 18 March and 30.6 ppt on 
30 April. Station plankton volume in 1998 ranged from 6.1 
to 280 mL per 100 m 3 and was greatest during late April 
(98IV) at the lower bay stations. 
We compared environmental variables among months 
to aid in the characterization of Penobscot Bay. PCA of 
log-transformed variables identified two factors that ex- 
plained 69% of the total variance in 1997 and 76% in 
1998 (Table 2). Factor 1 was related to temperature, plank- 
ton volume, and salinity in 1997 and in 1998. Factor 2 
was related to depth in both years. MAN OVA of mean 
factor scores indicated significant differences by month 
in 1997 (Wilk’s lambda=0.792, P<0.001) and 1998 (Wilk’s 
lambda=0.778, PcO.001). Temperature was negatively cor- 
related and plankton volume and salinity were positively 
correlated with PCA factor 1 in 1997 (Table 2). Post hoc 
comparisons of monthly scores for factor 1 indicated that 
the mean scores for April were different from those for 
May and June. In 1998, temperature was positively cor- 
related and plankton volume negatively correlated with 
PCA factor 1 (Table 2) and post hoc comparisons indicated 
that the mean scores for March were different from those 
for April. Salinity was correlated with both PCA factor 1 
(positively) and factor 2 (negatively) in 1998. Depth was 
positively correlated with factor 2 in both 1997 and 1998. 
Post hoc comparisons of monthly scores for PCA factor 2 
indicated that the mean scores for April were different 
from those for May and June in 1997; the mean score for 
March was different from that for April in 1998. 
Fish larvae 
Differences existed between the two years in the kinds of 
fish larvae found in Penobscot Bay. At least, 26 species (23 
individuals [1.5%] not identified to species) belonging to 
15 families were identified (Table 3). A total of 23 species 
(15 families) were found in 1997, and 16 species (9 fami- 
lies) in 1998; 13 species were common to both years. Dif- 
ferences in taxonomic composition were mainly due to 
the collection of rare specimens during the year, such as 
inquiline snailfish (Liparis inquilinus) in 1997 and pol- 
lock ( Pollachius virens ) in 1998. In addition, several spe- 
cies that were present in 1997, including American eel 
(Anguilla rostrata), capelin (Mallotus villosus), fourbeard 
rockling ( Enchelyopus cimbrius), Atlantic tomcod (Micro- 
gadus tomcod ), cunner ( Tautogolabrus adspersus), Atlantic 
mackerel (Scomber scombrus), alligatorfish ( Aspidophoroi - 
des monopterygius), windowpane flounder (Scophthalmus 
aquosus ), and American plaice (Hippoglossoides plattes- 
soides), did not occur in the 1998 collections. In contrast, 
pollock, snake blenny (Lumpenus lampraetiformis) and 
Arctic shanny (Stichaeus punctatus) were collected only in 
1998. The assemblage of species was slightly more diversi- 
fied in 1997 than in 1998 with Shannon-Wiener indices of 
H'=3.1 and 2.8, respectively. The three most abundant spe- 
cies in 1997 (U. subbifurcata, Ammodytes sp., P america- 
nus) constituted about 59% of all captures compared with 
68% in 1998 (M. octodecimspinosus , P gunnellus, Ammo- 
dytes sp.). 
A total of 779 larvae of 23 species of fishes were collect- 
ed during the seven cruises in 1997 (Table 3). The most 
commonly occurring larvae (>35% of the samples) were At- 
lantic seasnail (Liparis atlanticus), winter flounder (Pleu- 
ronectes americanus), radiated shanny (Ulvaria subbifur- 
cata ), sand lance (Ammodytes sp.), and H. plattessoides. 
Rare larvae (<1% of the samples) included Atlantic cod 
( Gadus morhua), Gulf seasnail ( Liparis coheni ), L. inqui- 
linus, shorthorn sculpin (Myoxocephalus scorpius), and S. 
aquosus. The highest number of species (14) occurred dur- 
