FISHERY BULLETIN: VOL. 73, NO. 4 



METHODS 



Kelp perch, white seaperch, and senoritas were 

 observed in the field and laboratory. Over a 2-yr 

 period, scuba divers spent more than a total of 125 

 h watching and collecting fish both day and night 

 at depths ranging from 1 to 20 m. Study localities 

 included areas of reef and kelp in the Santa Bar- 

 bara Channel— off the Santa Barbara mainland 

 and off Santa Cruz Island located across the 

 Channel some 42 km to the south. 



Food 



During 27 scuba dives made between February, 

 1971 and March, 1973, we collected a total of 115 

 kelp perch (measuring 43-142 mm, averaging 103 

 mm standard length). 111 white seaperch (74-203, 

 139 mm), and 65 senoritas (110-227, 169 mm). Gut 

 contents were found in and identified from 50 kelp 

 perch, 55 white seaperch, and 53 senoritas. All fish 

 were taken with a small, 15-prong pole spear. 

 Later they were slit ventrally, fixed in 10% For- 

 malin,- washed, and preserved in 45% isopropanol. 

 During the analysis, fish were identified by num- 

 bered tag only so as to minimize any bias that 

 might result from forehand knowledge of time of 

 collection, etc. 



For all three species, the simple, tubular gut, 

 which is "stomachless" in the sense of Chao (1973), 

 was excised, measured, and divided from front to 

 back into three sections of equal length, here ar- 

 bitrarily called the "fore-," "mid-," and "hindgut," 

 respectively. Fullness of each section was scored 

 subjectively from 1 (empty) to 5 (full and distend- 

 ed). Scores were plotted against time of day. 

 Because fish were sampled throughout the year, 

 their times of collection were seasonally adjusted 

 relative to times of sunrise and sunset as listed in 

 solar tables for the particular dates. Adjusted time 

 of collection, rounded to the nearest 2-h interval, 

 was measured on the relative scale with sunrise 

 arbitrarily set at 0600 h and sunset at 1800 h. 



Displayed under a dissecting scope, the contents 

 of the foregut were sorted into broad taxonomic 

 categories of food items, which were segregated in 

 a small, partitioned tray. Then the percent volume 

 of each item in the array was estimated by eye. 

 Estimates were made quickly to the nearest per- 

 cent, and their total over the array often exceeded 



-Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



100% per fish. Yet at the outset, independent es- 

 timates of the same item did not vary substan- 

 tially among successive trials, and series tended to 

 regress toward a mean value (an observer's 

 overestimation of volume on one trial was often 

 countered by his underestimation of the same 

 volume on the next). Item volumes were later 

 standardized to 100%. In computing species 

 means, fish with empty guts were not counted and 

 all others were weighted equally, regardless of fish 

 size or gut fullness (cf . Zaret and Rand 1971). 



The frequency of occurrence of a dietary item 

 was expressed as the percent of fish with non- 

 empty foreguts that contained the item. The rank 

 order of item frequencies was highly correlated 

 with that of volumes. Kendall's tau correlation 

 coefficients for the kelp perch, white seaperch, and 

 senorita were 0.51, 0.85, and 0.70, respectively 

 {P<Om). 



Activity 



In the field, feeding bites made by individuals of 

 each species were counted in six, 2-h intervals 

 between dawn and dusk. Standardized as bites per 

 minute, these counts were necessarily limited to 

 solitary individuals that could be followed by a 

 scuba diver for periods of about 3 to 5 min. It was 

 impossible, for example, to discern the very rapid 

 movements of senoritas feeding in large 

 aggregates, which often formed during the early 

 morning hours. The species value for each 2-h in- 

 terval is the mean of 13-37 observations, each of a 

 different fish. All' counts were made during the 

 first week of October, 1974. 



To supplement field observations, swimming 

 movements of individuals were observed in an 

 outdoor tank. Fish were caught live either by hook 

 and line during the day or by hand net underwater 

 at night and transported in aerated containers to a 

 1-m-deep, 500-gallon, circular concrete tank at the 

 University of California Santa Barbara Marine 

 Laboratory. The tank was located outdoors under 

 a lath roof and so was exposed to a natural, but 

 subdued, 24-h cycle of light and dark. To eliminate 

 visual disturbance, a black opaque plastic shroud 

 perforated with several small peep holes was 

 erected around the sides of the tank. Two, 10- W 

 incandescent red lights, placed 1.5 m above the 

 waterline, provided continuous dim light, 

 especially useful for nocturnal observations. The 

 temperature of the continuously running, filtered 

 seawater, which varied no more than 3°C during 



816 



