174 



Fishery Bulletin 91(1). 1993 



structure within the estuary, the frequency distributions of available 

 salinities and temperatures were different for all fish tracks (log- 

 likelihood test, Zar 1984; p<0.001). Hence, it was impossible to com- 

 pare the distributions of temperature and salinity experienced by 

 individual fish. No analysis was made on the depth data transformed 

 to salinity and temperature for portions of fish tracks below depths 

 sampled by the CTD, because the available frequency distributions 

 of salinity and temperature could not be calculated for these depths 

 and the distance from the fish to the bottom could not be accurately 

 determined. Due to these problems, an average of 83.2% (range 50.9- 

 100%) of the depth observations for individual tracks were converted 

 to experienced salinity and temperature. Fish 3 was not analyzed for 

 temperature and salinity preference because it spent nearly all its 

 time below depths sampled with the CTD. However, the frequency 

 distributions of temperatures and salinities occupied by fish showed 

 modes between 14° and 16°C for five of seven fish, and 17 and 19%c 

 for four of seven fish (e.g., Fish 4 and 5, Fig. 4). The log-likelihood 

 test indicated that all fish occupied different distributions of tem- 

 perature and salinity than they would have experienced by random 

 vertical movements in their environments (p<0.001). 



the bottom. This resulted from record- 

 ing water depth under the boat, which 

 generally followed a short distance be- 

 hind the fish rather than directly above 

 it. Although this discrepancy makes it 

 impossible to accurately determine the 

 distance of the fish from the bottom, Fish 

 1 and Fish 3 spent the majority of their 

 time close to or on the bottom, and Fish 

 4 spent approximately 35% of its time 

 near the bottom. 



Fish encountered a wide range of sa- 

 linities ( 7. 8-33. 69c c) and temperatures 

 (8.9-22.9°C; Table 3). Due to the dynam- 

 ics of tidal currents and vertical water 



Discussion 



In general, the tracked fish moved with tidal currents, milled during 

 periods of low current velocity, and reversed their direction of move- 

 ment with the change of tides. The results suggest that tidal cur- 

 rents are a major component to horizontal fish movements in the 

 Columbia River estuary. Chinook salmon had higher mean ground 

 speeds during ebbing tides than during flooding tides, presumably 

 because tidal and riverine flows are additive during ebbing tides and 

 antagonistic during flooding tides. 



These findings tend to agree with other estuarine tracking studies 

 (Groot et al. 1975, Fujioka 1970) of Pacific salmon. Fujioka (1970) 

 found that the position of chinook salmon tracked in the Duwamish 

 River estuary was dependent on the tidal stage, with fish generally 



