194 



Fishery Bulletin 105(2) 



^.i°^.<*^^o^vvv"l*vv^-^"■'^»'V^^'^'VVvv' 



Figure 4 



Monthly mean Nereocystis luetkeana density (no. of Nereocys- 

 tis/120 m^ ±standard error). The mean was averaged across the 

 five sites initially containing canopy, n = 248 transects. 



habitats in adjacent, deeper water. Puget Sound 

 rockfishes tend to move to shallower water in 

 summer and deeper water in winter (Moulton 

 and MillerM, possibly avoiding increased storm 

 surge during winter months. Fish communities 

 inhabiting seasonally and annually variable 

 kelp beds in the north Pacific must be capable 

 of enduring a wider variety of environmental 

 variables over the course of a season, year, or 

 lifetime than those occupying the relatively 

 stable, perennial canopy-dominated kelp beds 

 of more temperate zones. The importance and 

 magnitude of seasonal cues vary among kelp 

 and fish species throughout their ranges and 

 the cues include temperature, photoperiod, tur- 

 bidity, increased frequency of storms and surge 

 in winter, and prey and nutrient availability 

 (Ebeling et al., 1980, Dayton, 1985). However, 

 the thresholds of many environmental factors 

 are to some extent temperature-dependent 

 (Dayton, 1985), providing an easily quantified 

 surrogate variable for seasonality in the present 

 study. Kelp beds (and associated fishes) at the 



Miller, B. S., and L. L. Moulton. 1988. Charac- 

 terization of Puget Sound fishes for the EPA Bay 

 Program, p. 77-84. In Proceedings, First Annual 

 Meeting on Puget Sound Research, vol. 1, Seattle, 

 WA, March 18-19, 1988. Puget Sound Action Team, 

 Office of the Governor, P.O. Box 40900, Olympia 

 WA 98504-0900. 



Table 2 



Logistic regression results on the presence of fishes by families and by total abundance. Only families composing greater than 

 207c of the total fish abundance are included. Independent variables are time (n = 503 transects for tests of "month" or n=208 

 transects for test of "year," the comparison of June to September of years 2002 and 2003 ), kelp densities (canopy kelp and annual 

 and perennial understory kelp groups; /! = 473 transects), and physical habitat variables (water temperature, rugosity, substrate 

 size, and verticality). Only significant values (P<0.05) are reported. 



Analysis group 



Independent variable 



Parameter estimate 



Standard error 



Wald's x^ 



P-value 



Hexagrammidae 



Scorpaenidae 

 Gadidae 

 Total fish 



Annual kelp understory 

 Water temperature 

 Rugosity 

 Substrate size 



Year 



Water temperature 



Month 



Water temperature 



Month 



Year 



Canopy 



Annual understory 



Water temperature 



Substrate size 



0.62 



4.26 



-2.48 



1.90 



0.64 

 9.00 



0.27 

 4.28 



0.08 

 0.28 

 0.66 

 0.21 

 5.12 

 0.80 



0.17 

 0.91 

 1.15 

 0.47 



0.29 

 3.00 



0.08 

 1.62 



0.03 

 0.13 

 0.32 

 0.10 

 0.78 

 0.38 



