384 
Fishery Bulletin 115(3) 
appear to be uniform, nor did they follow any particular 
pattern (Suppl. Fig. 2) (online only). Shark AOCS3, for 
instance, exhibited all 3 types of behavior almost in 
the same proportion (Fig. 4), and these behavior types 
alternated frequently throughout its time series. Long 
sequences of the same behavior type were rare for this 
individual, the maximum being 9 consecutive days for 
type I. In fact, long sequences of one behavior type were 
seen only when a particular type was also dominant 
for an individual, such as type I for shark IOCS1 and 
type III for shark AOCS5. For all sharks, one isolated 
day of any behavior type occurred more frequently than j 
any sequence of one type of behavior. Only for one in- > 
dividual, shark AOCS4, was there a clear shift in be- 
havior type with time observed. During its first 50 days 
of monitoring, this shark almost exclusively exhibited 
type-I behavior. After this period, shark AOCS4 started 
to alternate type I with the other behavior types. At 
the same time, the number of gaps present in the first 
50 days of the time series for this shark weakened the 
veracity of this otherwise constant pattern. 
Three sharks (AOCS5, AOCS6, and AOCS7) were 
