EBELING and BRAY: ACTIVITY OF REEF FISHES 



the four zones. These shifts are measured in Table 

 4 by values of A/;: 



Ih 



= 2 liPi 



1 = 1*- 



'iday)(l) - iPi 



night) (^)J , 



where p,day is the proportion of individuals of a 

 species observed during the day in zone / (i = 1,2, 

 3, or 4 for the shelter through mid-water zones, 

 respectively) and p, nj^ht is the proportion observed 

 at night. The Ih's range from +3.0, when all 

 observed individuals of a species undergo a max- 

 imum shift downward from the mid-water zone 

 during the day to the shelter zone at night, to -3.0, 

 when all individuals undergo the reverse max- 

 imum shift upward. A Ih = 0.0 indicates little or 

 no shift, in that the species' proportional distribu- 

 tion among zones does not change from day to 

 night. 



Fish species varied considerably in the degree to 

 which they changed zones between day and night, 

 although the patterns of shifting upward or 

 downward were similar within families (Table 4). 

 Some species changed their vertical position little 

 if at all: several species of rockfishes; the cabezon, 

 Scorpaenichthys marmoratus; Rhacochilus tox- 

 otes; Damalichthys vacca; and blackeye goby, 

 Coryphopterus nicholsii. Other species changed 

 their vertical position markedly between day and 

 night. Individuals of Chromis punctipinnis and 

 Pimelometopon pulchrum, which had near-max- 

 imum positive values of Ih, move about in the 

 water column during the day and shelter in holes 

 at night. No individuals of Oxyjulis californica 

 were seen at night (recall that they descend from 

 mid-water to bury themselves in sand or gravel 

 patches). Assuming that burying individuals are 

 in the "shelter zone!' Ih for Oxyjulis = 2.91. No 

 species had a large negative value of Ih, i.e., no 

 species mostly contained individuals that rose 

 from the bottom to mid-water at night. Hobson 

 and Chess (1976) noted that during the day most 

 Sebastes atrovirens were "seated on rocky strata" 

 whereas at night they "hovered in mid-water." In 

 the present study, the A/i of S. atrovirens was 

 small but positive (Table 4); however, this species 

 was relatively rare in our transects. 



Several lines of evidence indicate that many of 

 the kelp-bed fishes observed become less active 

 and do not regularly feed at night. The levels of 

 activity often could be inferred from direct obser- 

 vations. Many species that swam about and fed on 

 or above the reef during the day were found deep 

 in holes and crevices at night and would flee from 



their shelter only when vigorously disturbed. 

 These species included Hypsypops rubicundus, C. 

 punctipinnis, P. pulchrum, and juvenile S. mys- 

 tinus. Some individuals of P. pulchrum reportedly 

 secrete a mucous envelope about themselves 

 (Wiley 1974), and we often found this fish wedged 

 deep in crevices in an apparent state of torpor at 

 night. Individuals of Girella nigricans were also 

 found in holes or on the bottom but were more 

 responsive to our presence. Previous diel analyses 

 of gut contents substantiate our present impres- 

 sions that the following species are strictly day- 

 time feeders: H. rubicundus (by Clarke 1970), 0. 

 californica (by Bray and Ebeling 1975), juvenile S. 

 mystinus (by Thomas Bailey unpubl. data), and C. 

 punctipinnis (by Hobson and Chess 1976; Bray 

 unpubl. data). 



Our analyses of fish-gut emptiness revealed that 

 even many of the kelp-bed fishes not undergoing 

 such obvious diel changes in vertical position may 

 stop feeding at dusk (Figure 1, Table 5). Although 

 all five demersal surf perches {Embiotoca jacksoni, 

 E. lateralis, Hypsurus caryi, Damalichthys vacca, 

 and Rhacochilus toxotes) generally remain in the 

 suprabenthic and bottom zones both day and 

 night, their diel patterns of gut emptiness indicate 

 that all but R. toxotes do not feed at night. Median 

 scores of gut fullness for E. jacksoni reached 

 maximum values in the afternoon and declined 

 after sunset; at dawn, all guts examined were 

 empty (Figure la). Fully 88% of the fishes speared 

 during daylight hours contained food in their 

 foreguts (Table 5). Although 39% of the fishes 

 collected at night contained food, 89% of these 

 were collected before midnight. Thus it is likely 

 that the food contained in the foreguts of these 

 individuals was eaten before nightfall and had not 

 yet passed into the second quarter of their guts. 

 Foreguts of E. lateralis, H. caryi, and D. vacca 

 show the same pattern (Figure Ib-d). In fact, all 

 four species had significantly less food in their 

 guts during the night than during the day (Table 

 5). Gnose (1968) also observed that individuals of 

 E. lateralis collected from off" Oregon had empty 

 guts at dawn. Additionally, two other kelp-bed 

 surfperches that commonly occur in mid-water, 

 Phanerodon furcatus and Brachyistius frenatus, 

 which is rare at Naples Reef, feed mainly during 

 the day (Bray and Ebeling 1975; Hobson and Chess 

 1976). 



In contrast, median scores for fullness of R. 

 toxotes reached maximum values at night, and 

 many foreguts were empty during the day (Figure 



709 



