Starr et a\: Movements of Sebastes paucispinis and S ch/orostictus in Monterey submarine canyon 



335 



that subadults are more mobile than adults, and as the 

 fish increase in size, they become more sedentary. From tag 

 transmissions and submersible sui-veys, we observed both 

 pelagic and refugial behavior (about 25% of the tagged ani- 

 mals exhibited refugial behavior). The two largest tagged 

 bocaccio were among the five fish that moved the least. The 

 narrow size range of most of the tagged fish, however, pre- 

 cluded an analysis of movements by fish length. 



Tagged bocaccio made frequent small vertical move- 

 ments that were associated with time of day. The magni- 

 tude of the vertical movements matched the vertical relief 

 of the habitats used by tagged fish. During submersible 

 operations, we tracked tagged fish in small schools as they 

 moved along rock scarps that were 10-20 m high and 

 100-200 m long (Starr et al., 2000). These scarps were of- 

 ten at, or just below, the rim of the submarine canyon. The 

 depth transmissions from tags indicated that fish were at 

 the top or just above the rock habitats during the day, and 

 lower and more sedentary at night. We attribute the in- 

 creased activity during the day to the fact that bocaccio 

 are visual predators (Love, 1996) and are thus more apt to 

 forage during the day. 



Four of the eight fish containing depth trans- 

 mitters made rapid vertical movements. Three of 

 the bocaccio rose vertically to spend a short time 

 near the surface, then returned to depths from 

 which they came. These fish all remained only a 

 few weeks in the study area. A fourth fish made a 

 deep dive to 220 m and back to 100 m in less than 

 a day (Fig. 12). The purposes of these dives are un- 

 known, but such dives reinforce our observations 

 during tagging operations that bocaccio can modi- 

 fy the volume of air in their swim bladder. Pearcy 

 ( 1992 ) obsei-ved that yellowtail rockfish also are ca- 

 pable of discharging air from their swim bladders 

 in a relatively short time. 



For some tags, there were short time intei-vals 

 when signals were not recorded. This may have 

 been caused by the canyon topogi-aphy or by fish 

 behavior. The complex topography of Soquel Can- 

 yon and the behavior of the rockfishes may have 

 prevented the receivers from receiving all possible 

 transmissions from tagged fish. The hard, steep 

 walls that undulate along the side of the canyon 

 (Yoklavich et al., 2000) occasionally cause echoes 

 and an original transmission to coincide, thus preventing 

 the receivers from recording a valid signal. Other signal 

 lapses may also have been due to the tendency of rockfish 

 to take shelter under rocks or ledges. From the surface, we 

 occasionally heard weak signals that were unusually high- 

 pitched and that sounded "tinny." We have experienced the 

 same type of reception when tracking shallow-water rock- 

 fishes that took shelter under a rock or ledge. Signals re- 

 turn to full strength and timbre when the fish leaves the 

 crevice. 



Implications for marine reserves 



Tag-recapture studies of shallow water (<100 m) demersal 

 rockfishes typically have indicated very little movement; 



90 

 95- 

 100 



130 



o Tag 7 depths 

 / \ Dayligtit curves 



\ \ 1 \ \ 1 



22 Oct 23 Oct 24 Oct 25 Oct 26 Oct 27 Oct 28 Oct 



1998 



Figure 14 



Depth transmissions recorded for tag-7 bocaccio plotted with day- 

 light curve for the period 22-28 October 1998. 



only one species in ten studies exhibited long-term move- 

 ments greater than 3 km (Stanley et al. 1994; Lea et al., 

 1999). Results from these studies support the idea that 

 small harvest refugia may effectively protect nearshore 

 rockfishes. This would be fortuitous if almost all near- 

 shore rockfishes have small home ranges, because most of 

 the marine reserves in the eastern Pacific Ocean are very 

 small and have been created without regard to typical 

 movements of species (McArdle, 1997; Starr, 1998; Yoklav- 

 ich, 1998). 



Our results indicate a considerable short-term varia- 

 tion in the movements of individual greenspotted rockfish, 

 which may be masked by the long-term nature of tag-re- 

 capture studies. Although almost all fish may remain in a 

 small area over the course of several years, an individual 



