Starr et a\ Movements of Sebasles paucispinis and S. chloiostictus in Monterey submarine canyon 



331 



Eight of'the 16 tags transmitted information about 

 depth. We recorded signals from five of the eight tags 

 for only short time periods after the fish were re- 

 leased. Two of these five fish (tag-3 and tag-12 fish) 

 moved vertically to within 15 m of the surface 9-12 

 hours after tagging, then returned below a depth of 

 90 m for a 12-15 day period before signals were lost 

 (see Fig. 11 for an example of tag-3 fish movements). 

 A third fish (tag-4 fish) fluctuated ±10 m around a 

 depth of 90 m for a week, moved vertically to a depth 

 of 14 m, and returned to a depth of about 90 m for 

 another three weeks before signals ceased. A fourth 

 tag (tag 13) was recorded for 3 days and exhibited 

 frequent fluctuations of 20 m in depth before signals 

 ceased. 



Data from the remaining three fish containing 

 depth transmitters were recorded throughout the 

 study. Depth transmissions from tag 14 varied less 

 than 3 m, whereas depth transmissions from tag 7 

 (Fig. 12) and tag 9 (Fig. 13) indicated cyclical ver- 

 tical movements of ±10-20 m, and occasional deep- 

 er dives. The greatest variation in depth was exhib- 

 ited from tag 7. That fish made several dives of 100 

 m; once it moved from 100 m to 220 m and back in 

 a 20-h period. Tag-7 and tag-9 fish also demonstrat- 

 ed a diurnal periodicity in vertical movements. Fish 

 were more active and higher during the day and less 

 active and deeper at night (Fig. 14). The AN OVA and 

 Scheffes F test statistics indicated significant differ- 

 ences in both depth and change in depth between 

 dawn, day, dusk, and night hours (Table 6). 



Environmental parameters fluctuated but did not 

 appear to be related to fish movements. Salinity 

 and temperature data from the S-4 current meter 

 at 100-m depth fluctuated within 24-h periods, but 

 there were no obvious trends within or between 

 years. Salinity averaged 33 ppt in 1997 and 34.5 ppt 

 in 1998. Water temperature fluctuated from 10 to 

 15°C in 1997, whereas in 1998 it was more consis- 

 tent, ranging from 9 to 11°C. Current speed and direction 

 fluctuated in what appeared to be a tidal basis (Shea and 

 Broenkow, 1982), but since there was no obvious relation- 

 ship to recorded movements, this relationship was not ex- 

 plored in more detail. 



Discussion 



Fish movements 



The in situ tagging procedures we developed alleviated many 

 problems associated with surface tagging and provided 

 means for tracking deeper-water rockfishes. We expected 

 gi'eenspotted rockfish to move only small distances because 

 of their affinity to seafloor habitats such as overhangs and 

 crevices (Stein et al.. 1992; Yoklavich et al., 2000). Our work 

 confirmed that gi-eenspotted rockfish are relatively seden- 

 tary. The greenspotted rockfish we tagged with depth trans- 

 mitters moved less than ±3 m vertically during the study, 

 except for a few occasions when a fish swam down 20-40 



26 



24 



13 • 



25 



20 



17 



21 

 12 

 10 



4 



3 

 27 

 18 

 14 



9 



7 



dlDOtlKJCD 



o o 



17 Aug 



17 Sep 



1 



17 Oct 



1998 



17 Nov 



17 Dec 



Figure 8 



Dates of signal receptions in study area from sonic transmitters 

 implanted in bocaccio in 1998. Circles represent individual occur- 

 rences of recorded signals; solid lines represent almost continu- 

 ous occurrences of recorded signals. Dashed lines on the graph 

 are shown to separate the groups of tagged fish discussed in the 

 text. 



m, only to return to its original depth within 2 hours. This 

 small variation in depth displayed by the tagged greenspot- 

 ted rockfish indicated that these fish most likely do not leave 

 canyon wall habitats to feed. Love (1996) indicated that 

 they eat mainly small invertebrates, but also cephalopods 

 and fishes. It is possible that prey in the water column is 

 advected toward them along the canyon wall (Isaacs and 

 Schwartzlose, 1965; Genin et al., 1988). They do not appear 

 to migrate vertically to feed on scattering layer organisms 

 as do other rockfishes such as yellowtail rockfish (Sebastes 

 flavidus) off Oregon (Perewa et al., 1969). 



Horizontal movements of the tagged gi'eenspotted rock- 

 fish were on the order of hundreds of meters to a few ki- 

 lometers. Receivers 2 and 3 each recorded cross-talk from 

 signals of the same two tags (tags 2, 6) at several different 

 times, indicating that these fish alternately swam close 

 to one receiver, then at a later time swam close to a sec- 

 ond receiver that was more than 300 m away (Starr et 

 al., 2000). Although most movements were contained well 

 within the 3-km long study area, half of the fish made so- 

 journs out of the study area for short time periods. We 

 think these small, short-term movements represent forag- 

 ing activity along the canyon ledge. 



