CLARKE: SPAWNING, EGG SIZE, AND DEVELOPMENT OF HAWAIIAN ANCHOVY 



of 5-10 larvae each from September and six sam- 

 ples of 10 each from March. 



RESULTS 



Spawning 



The dusk-to-midnight samples from all four 

 seasons clearly indicated that spawning was con- 

 centrated within a ca. 1 h period shortly after 

 sunset. Newly spawned eggs, with no evidence 

 of embryonic development under low magnifica- 

 tion, were absent from initial samples and ap- 

 peared in large numbers in samples 1.5-3 hours 

 after sunset. The eggs in samples taken ca. 1 

 hour after first appearance of new eggs were 

 already either mostly or totally in early cleavage 

 stages. In the June series, many of the new eggs 

 in the first sample containing new eggs were 

 already in early cleaveage stages. Newly 

 spawned eggs were very rare in samples taken 

 ca. 1 hour after fu-st appearance and were absent 

 from later samples. Thus spawning appears to be 

 concentrated within a period of 1 hour or less. 

 The estimated midpoints of spawning time 

 (Table 1) indicate that spawning occurs about 3 

 hours earlier in winter than in summer. The dif- 

 ference is not entirely accounted for by the 

 earlier sunset in the winter months; the delay 



after sunset is less than in summer. These esti- 

 mates of seasonal changes in spawning time 

 agree with trends evident from mature females 

 (Clarke 1987). 



Incubation 



In June and September, the incubation period 

 was less than a day, and two day classes of eggs 

 never occurred in the same sample. By early 

 morning, embryos already extended about half- 

 way around the remaining yolk. By late after- 

 noon they extended over three-fourths of the 

 way around the yolk, and the tail had flexed 

 sideways. In both months the eggs apparently 

 hatched about 22 hours after spawning time. 

 During a 1 h period the "old " eggs disappeared, 

 and newly hatched larvae appeared. 



Incubation time was considerably longer in 

 December and March. Eggs taken in the morn- 

 ing and afternoon were noticeably less developed 

 than those from the same times in June or Sep- 

 tember. Both "old" eggs and recently spawned 

 eggs were present between spawning time and 

 midnight in December. No "old" eggs were pre- 

 sent in any morning samples, and the embryos in 

 the latest (2400 h) night samples appeared al- 

 most ready to hatch, judged from comparable 

 stages in the June and September samples. Thus 



Table 1 . — Summary of data on spawning time, duration of egg and prefeeding 

 larval stages, and egg size of the Hawaiian anchovy, Encrasicholina purpurea, in 

 September and December 1984 and March and June 1985. Times of sunrise, 

 sunset, spawning, and hatching are Hawaiian Standard Time. Temperature values 

 are the means and ranges of hourly surface temperatures for a 2 wk period 

 bracketing the sampling dates (see text). Values for egg and yolk volume are 

 means and standard deviations of volumes calculated from length and width 

 measurements. 



595 



