648 
Fishery Bulletin 95(4), 1997 
the initiation and the end of a poor transmitter sig- 
nal) was recorded. On the basis of these data, the 
portion of the diurnal activity period spent inactive 
was calculated and expressed as a percentage. 
In describing the activity-inactivity patterns of 
female cunners, five parameters were used: 1) onset 
of activity, 2) cessation of activity, 3) duration of di- 
urnal activity, 4) length of inactivity bouts, and 5) 
percent of diurnal activity period spent inactive. 
Analysis of data 
Although a fish was handled for only 1-2 min when 
a transmitter was inserted into its tag holder and 
although field observations did not detect any 
changes in the behavior of fish following this proce- 
dure, nonparametric paired t-tests were used to ex- 
amine whether female cunners showed similar ac- 
tivity on the first complete day of tracking (day 2) 
compared with the following day of tracking (day 3). 
All five activity parameters were tested. 
A nonparametric analysis (Wilcoxon matched-pairs 
signed-ranks test) was used to compare inter- 
individual differences in activity between the two 
pairs of subjects tracked during the same periods 
(Table 1). Comparisons were made between fish A 
and fish B for a total of 11 days (i.e. June 18, 19, 20, 
21, 22, 24, 25, 27, 28, 29, and 30) and between fish C 
and fish D for a total of 9 days (i.e. July 13-21 inclu- 
sive) for all five activity parameters. 
A least-squares multiple regression analysis was 
used to determine whether date, time of day (i.e. 
morning vs. afternoon), or environmental variables 
(water temperature, cloud cover, and sea state) had 
a significant effect on activity. All activity param- 
eters were tested. Because the activity data were 
normally distributed, no transformations were carried 
out. Although there were 141 tracking days, some days 
could not be used for certain activity parameters. For 
example, there were only 72 tracking days during which 
both the time of onset and cessation of activity were 
known for any given fish, both of which are required to 
calculate the duration of diurnal activity. 
The tidal cycle was divided into four phases: low- 
tide, flood-tide, high-tide, and ebb-tide as described 
by Bradbury et al. (1995). For fish A, B, C, and D, 
mean activity parameters (i.e. percentage of time 
spent inactive and length of inactivity bouts) were 
determined for each tidal phase for the duration of 
the tracking period. An analysis of variance with 
three factors was used to test for intra- and inter- 
individual differences in activity during the tidal 
phases. We included only the fish by tide interaction 
term in our analyses because we did not expect any 
temporal variation in tide (tide x date) or activity 
Table 2 
Mean number of minutes before sunrise and after sunset 
when tracked female cunners began and ceased their diur- 
nal activity. Because the data for onset of activity for fish E 
consisted of a single point, no standard deviation is given. 
Onset Cessation 
Fish (minutes before sunrise) (minutes after sunset) 
identi- 
fication 
Mean 
SD 
n 
Mean 
SD 
n 
A 
57.6 
10.02 
9 
57.6 
26.70 
9 
B 
40.2 
23.95 
12 
63.6 
18.10 
12 
C 
48.7 
10.93 
18 
74.3 
19.64 
22 
D 
31.7 
18.48 
6 
43.0 
43.59 
4 
E 
45.0 
— 
1 
50.3 
34.46 
4 
F 
52.0 
21.28 
3 
36.5 
41.16 
4 
G 
49.4 
32.69 
14 
54.1 
20.33 
17 
H 
72.4 
52.54 
28 
18.9 
39.12 
27 
All 
54.7 
36.63 
91 
49.1 
36.15 
99 
(fish x date), given the relatively short time (i.e. 11 
and 9 days) over which observations were made. 
Paired comparison £-tests were used to examine 
whether the one female cunner tracked during both 
the prespawning and spawning period had the same 
activity patterns during both periods. All activity 
parameters were tested. 
Statistical analyses were performed with Minitab 
(Minitab, Inc., 1992) or SPSSX (SPSS Inc., 1990) sta- 
tistical software packages. 
Results 
There were no significant differences (P>0.05, non- 
parametric paired £-test) in activity parameters be- 
tween the first complete day of tracking (day 2) and 
the following day. 
All tagged fish were active during the day, inac- 
tive at night. Activity commenced, on average, 55 
minutes (SD=36.6) before sunrise and ceased 49 min- 
utes (SD=36.2) after sunset; however, there was 
considerable daily variation among individuals 
(Table 2). Throughout the day, activity was inter- 
rupted by periods of inactivity, usually lasting 5-15 
minutes. Among those fish tracked on the same day, 
there were no significant differences between sub- 
jects for any of the activity parameters (Table 3). 
When water temperature was below 5°C, cunners 
were inactive. On 23 and 24 June, for example, strong 
northwesterly winds forced cold water into the study 
area causing the water temperature to drop from 6°C 
to 3°C and the cunners to be inactive for two days. 
On the morning of 26 June the water temperature 
