the last quarter (Figure 2). In April 1974 the col- 

 lector obtained several thousand eggs over a 3-day 

 period about 12 days before the last quarter; al- 

 though the eggs were of the same diameter as P. 

 sexfilis eggs, the larvae were not reared for posi- 

 tive identification. In view of the consistency of 

 subsequent spawning data, and the low ovarian 

 weights which two previous studies found in P. 

 sexfilis in April (Lowell 1971; Morris*'), we believe 

 these were eggs of another species. 



Some of the eggs found in the collector were 

 transparent, buoyant, and developing normally, 

 while others sank and were opaque and obviously 

 not viable. Among the spawnings in which 1,000 

 or more eggs were collected, an average of 52*%^ of 

 the eggs were viable (range, 0-989^). The airlift 

 collector apparently damaged the eggs to some 

 extent. For example, on 13 June 1974, when 75^^ 

 of the eggs from the collector were viable, eggs 

 obtained by dip net at the time of spawning 

 showed over 90^'r viability. It is not known how 



^Morris, D.E. 1964. Life history study of the moi, Po/vrfac- 

 tylus sexfilis. Job Completion Rep., Project No. F-5-R-11, Div. 

 Fish Game, State of Hawaii, 15 p. Division of Fish and Game, 

 Honolulu, Hawaii. 



Figure 2. — Duration of spawning cS. Poly dactyl us sexfilis rela- 

 tive to the time of the last quarter ((J) for the six spawning 

 months in 1974, 1975, and 1977. Data are given only for the last 

 four spawning months of 1976. 



many of the fish in the net participated in each 

 spawning or whether the same fish spawned each 

 month, although Kanayama'^ believed that indi- 

 vidual P. sexfilis spawned more than once in a 

 season. 



Time of Spawning 



The developmental stage of eggs found in the 

 collector in the morning indicated that spawning 

 had occurred shortly after sunset. Visual monitor- 

 ing of water sampled continuously from the net on 

 34 spawning nights showed that with few excep- 

 tions the fish spawned between 2030 and 2130 h 

 (Figure 3, Table 2). The times recorded were those 

 when eggs were first observed in the collector; it is 

 possible that additional spawnings took place 

 slightly later on the same night, and behavioral 

 observations (see below) indicated that this may 

 have been true on at least some nights. The time of 

 spawning did not vary with the time of sunset 

 (Figure 3) and appeared unrelated to the time of 

 moonrise (which occurred generally between 2300 

 and 0400 h during the spawning season). 



''Kanayama, R. 1967. Life history study of the moi, 

 Polydactylus sexfilis. Job Completion Rep., FVoject No. F-5-R- 

 15, Div. Fish Game, State of Hawaii, 9 p. Division of Fish and 

 Game, Honolulu, Hawaii. 



_l l_ 



I II 111 IV V VI 



SPAWNING MONTHS 



Figure 3. — Times of .spawning oi^ Polydactylus sexfilis during 

 the six spawning months iroman numerals) of 1974 in relation to 

 the tidal cycle. Dots indicate observed times of spa wning in 1974, 

 and the horizontal lines delineate the time of spawning as indi- 

 cated by data from 1974, 1975, and 1977 (see Table 2). Dotted 

 horizontal lines show the times of sunset in 1974. Vertical lines 

 for each spawning night indicate time between the evening high 

 and low tides, i.e., the duration of the outgoing tide, as measured 

 by a tide gage at Coconut Island in 1974. 



902 



