FISHERY BULLETIN: VOL. 79, NO. 1 



diurnal foraging area (Clarke 1970). In contrast, 

 C. punctipinnis is highly gregarious during both 

 day and night, and often individuals crowd noc- 

 turnal shelters among the rocks. When its noctur- 

 nal resting places are far from its diurnal feeding 

 grounds, C punctipinnis migrates between the 

 two locations during twilight in prominent pro- 

 cessions (Hobson and Chess 1976). These migra- 

 tions tend to be better defined in the evening than 

 in the morning. At times evening migrations 

 began up to 30 min or more before sunset, while at 

 other times they were not evident until about sun- 

 set. Generally, however, the migrations peaked 

 from shortly before sunset until about 15 min af- 

 ter, and then continued at greatly reduced levels 

 for another 10 min or so before ending. 



Chromis punctipinnis is among the most 

 numerous species on many nearshore reefs in 

 southern California (Limbaugh 1955; Quastl968), 

 and in some places apparently there is insufficient 

 rocky shelter to accommodate at night the vast 

 numbers that forage in the water column by day. 

 In such places excess individuals sometimes clus- 

 ter after dark in dense numbers on the sand next to 

 the reef. Occasionally C. punctipinnis is in the 

 water column at night, but does not seem to feed at 

 this time. This is attested by the empty stomachs 

 of all 11 specimens examined from predawn collec- 

 tions by Hobson and Chess (1976). 



Labridae: Oxyjulis caltfornica, Halichoeres 

 semicinctus, and Semicossyphus pulch&r 



The Californian labrids are so obviously quies- 

 cent at night that we have no doubt that, like 

 tropical labrids (Hobson 1965, 1974), they do not 

 feed at this time. Also, like their tropical relatives, 

 they follow precise patterns when shifting be- 

 tween diurnal and nocturnal modes (Hobson 1972; 

 Domm and Domm 1973). Oxyjulis californica, the 

 senorita, buries in the sand at nightfall (Herald 

 1961) , a habit also attributed to H. semicinctus , the 

 rock wrasse, by Limbaugh (1955), who noted that 

 this species "sleeps buried with head protruding." 

 Feder et al. (1974) repeated this statement, as did 

 Fitch and Lavenberg (1975), who reported that it 

 also "burrows between or under rocks to escape 

 predators or to sleep." We found that H. 

 semicinctus consistently took nocturnal shelter 

 amid low benthic algae (from which it often was 

 unintentionally flushed by our nocturnal ac- 

 tivities). Both O. californica and H. semicinctus 



were consistent in timing their descent to shelter 

 in the evening, and their rise into the water the 

 next morning. Data on the timing of these events 

 were collected during 1973 and 1974 at Fisher- 

 man's Cove, Santa Catalina Island, in a sand- 

 bottomed habitat dominated by the low brown 

 alga Dictyopteris zonaroides. The last O. califor- 

 nica seen entering the sand on four evenings at 

 this site slipped from view 11-22, x = 16.8, min 

 after sunset, which agrees with Bray and Ebeling 

 (1975), who on three occasions observed this 

 species entering sand and rubble "About 15 min 



after sunset " The first O. californica seen here 



on 11 mornings appeared 11-22, x = 15.8, min be- 

 fore sunrise. During the same period at this site 

 the last H. semicinctus seen taking cover on six 

 evenings disappeared 20-24, x = 22.0, min after 

 sunset, and the first to appear on 10 mornings 

 emerged 18-25, x = 22.8, min before sunrise. It 

 may be significant that H. semicinctus , which 

 grows to a larger size, tended to be active later in 

 the evening and earlier in the morning. Among 

 diurnal fishes on tropical reefs, the larger individu- 

 als tend to retire later and rise earlier (Hobson 

 1972). Comparable data are lacking forS. pulcher, 

 the sheephead (the largest of the three Californian 

 labrids), even though this species was numerous 

 at the observation site during the day. Semicos- 

 syphus pulcher shelters among rocks at night 

 (Hobson 1968b), and probably because the obser- 

 vation site lacked rocks, this species left the area 

 sometime before going under cover. On 4 evenings 

 the last S. pulcher departed the observation site 

 10-21, X = 18.0, min after sunset, and on 11 morn- 

 ings the first to return arrived 11-24, x = 17.0, min 

 before sunrise. On just one evening, in a nearby 

 rocky area where the species found shelter, the last 

 active S. pulcher was seen 30 min after sunset. 



Because S. pulcher rested in distinctive shelters 

 and was visible throughout the night (Figure 8), it 

 offered the best opportunity to investigate consis- 

 tency in resting places among individuals. 

 Reportedly, at least some tropical wrasses return 

 each evening to specific resting sites (Winn and 

 Bardach 1960; Starck and Davis 1966; Hobson 

 1972). So the resting places of nine S. pulcher were 

 located at midnight during November 1973, and 

 then revisited at the same hour once each week for 

 3 wk. On the first return only two of the nine 

 positions were occupied. No. 3 and No. 5 — both by 

 what appeared to be the same individuals that had 

 been there before. On the second return, again 

 only two positions were occupied: No. 5 seemed to 



12 



