FISHERY BULLETIN: VOL. 84, NO. 4 



wasconducted beneath and within either of the two 

 forest types, in water 6 to 22 m deep. Underwater 

 observations were recorded on formated data sheets 

 using plastic paper. 



In conducting the benthic survey, I slowly swam 

 from one end of the transect to the other and iden- 

 tified and enumerated the fish that were observed. 

 A fish was included in the benthic survey if it was 

 observed within 0.5 m of the bottom and was not 

 a member of a school of typically midwater fish 

 located momentarily near the bottom. A fish ob- 

 served swimming through the transect in front of 

 the diver was included. An effort was made to in- 

 spect all crevices, caves, and ledges, and to move 

 aside algae to locate fish. A description of unfamiliar 

 fish was made in the field and its identity later deter- 

 mined in field guides if possible. Small, relatively 

 cryptic species were probably underestimated in the 

 process of these visual surveys (Brock 1982). 



The midwater transect was searched about 3 m 

 above the tape. Repetitive ascents and descents 

 were made at 5 m intervals to detect fish occurring 

 throughout the water column. The sizes of very 

 large schools were estimated. All fish observed 

 within the length of the 50 m tape were recorded. 

 Unidentified species were treated as they were dur- 

 ing the benthic survey. 



After the survey was completed, an index of the 

 bottom profile was recorded by measuring the water 

 depth at each meter mark along the tape. Two 

 methods of determining bottom profile were used: 

 first, an objective, and later, a subjective measure. 

 The objective relief index was the sum of the dif- 

 ferences between each of the 50 consecutive depth 

 measurements along the 50 m transect. During the 

 second half of this study (1983) a subjective relief 

 index was assigned to the general vicinity of each 

 transect; this was determined by the greatest ver- 

 tical relief observed along the transect line: = flat, 

 no relief; 1 = low relief <1 m); 2 = moderate relief 

 (1 to 2 m); 3 = high relief (2 to 4 m); and 4 = ex- 

 treme relief (more than 4 m). 



Two measures of species diversity were used to 

 compare the fish assemblages in Macrocystis and 

 Nereocystis forests: 1) total number of species found 

 on all transects within one canopy type and 2) the 

 Shannon-Weaver index of diversity, H' (Pielou 

 1966). 



Because of heterogeneity between sample vari- 

 ances, fish density distributions were compared with 

 the nonparametric Mann- Whitney test. A minimum 

 acceptable level of significance of 0.05 was assigned. 



RESULTS 



Twenty-seven species of fish were identified 

 within the spatial limits of the transects (Tables 1, 

 2, 3). An additional 8 species were identified within 

 the kelp forest, but outside the transect limits. 

 Juvenile rockfish were considered a single group, 

 and occasionally an unidentified fish was observed. 



In Macrocystis forests, 26 species were identified 

 within the transects and 10 species outside the 

 transects; in Nereocystis forests, the respective 

 totals were 23 and 4 species. Three additional types 

 of fish were observed that could be identified only 

 to the family level (Table 3). Four species observed 

 only in Macrocystis forests were white seaperch, 

 Phanerodon furcatus; rainbow seaperch, Hypsurus 

 caryi; China rockfish, Sebastes nebulosus; and black- 

 eye goby, Coryphopterus nicholsi. One species was 

 observed only in Nereocystis beds, the jacksmelt, 

 Atherinopsis calif orniensis. Species not observed 

 within both transect types were relatively uncom- 

 mon, but were observed in and around both forest 

 types during this study. 



Fishes that could not be identified to species or 

 family level were rare, occurring on only 6 (8%) of 

 the transects (Table 3). 



Table 1.— Summary of presence/absence of fish species 

 encountered [midwater (M) and benthic (B), years pooled] 

 throughout study. 



802 



