WALTERS: ECOLOGY OF HAWAIIAN SERGESTID SHRIMPS 



aflanticus appear to stop migrating altogether. 

 During cruise 70-12 part of tlie populations of 

 Sergestes orientalis and S. consohrinus migrated, 

 while the rest of the populations remained at their 

 daytime ranges; during Echo IV both species 

 appeared to migrate normally. Other species, such 

 as Sergestes vigilax and Sergia scintillans, have 

 shown no indications of nonmigratory behavior. 

 Species showing the best evidence of nonmigra- 

 tion are all members of the shallow migratory 

 group, but sampling was inadequate to determine 

 definitely whether any species in the deep group 

 are also nonmigrators. The data from cruise 70-12 

 and the June 1973 Teuthis cruise further suggest 

 that when a species is not migrating its daytime 

 depth can also be abnormal. Sergia gardineri, 

 normally found between 650 and 775 m during the 

 daytime, was taken as deep as 1,200 m on these 

 cruises, and Sergestes atlanticus, normally found 

 between 600 and 725 m, was taken down to 800 m. 



The nonmigration of some sergestids around 

 full moon is a separate behavior from the moon- 

 light depression below 150 m. Nonmigration is not 

 a direct effect of increased light levels. During 

 Echo IV the moon was often heavily obscured by 

 cloud, yet the nonmigratory species remained 

 deep. During Cruise 70-12, nonmigratory species 

 remained deep until the next-to-last night, when 

 normal migration resumed, although light inten- 

 sity in the surface waters could not have been 

 radically different than on the previous night. 

 Nonmigratory behavior has been observed in 

 December and June, suggesting that it occurs 

 during most seasons of the year. 



Studies of seasonal variation in vertical migra- 

 tion can be complicated by moon effects. For 

 example, Donaldson (1973) found abnormally deep 

 distributions day and night for Sergia splendens 

 during a February 1972 cruise. He also cited data 

 from the same cruise showing that sergestid 

 numbers were strongly influenced by moonlight in 

 the upper 100 m, both at the quarter and at full. 

 Knowing only that the moon was in various phases 

 during the February 1972 cruise, it is impossible to 

 separate seasonal effects from moonlight effects 

 for 5. splendens. Other mid-water groups show 

 nonmigratory behavior not tied to lunar phase. 

 Riggs (pers. commun.) found that species of the 

 penaeid shrimp Gennadas did not migrate during 

 our November 1972 cruise, which sampled near 

 new moon when sergestids appeared to be mi- 

 grating normally, and concluded that a seasonal 

 factor was involved. In summary, the depth struc- 



ture of the mesopelagic community changes in a 

 bewilderingly complex manner under the 

 influence of ambient light, lunar phase, season, 

 and probably other undiscovered effects. 



Feeding Chronology and Diet 



In studies of the diets of mesopelagic animals, 

 the time of day when feeding takes place is as 

 interesting a datum as the kinds of prey eaten. 

 The most widely accepted theory of the function of 

 vertical migration holds that mesopelagic animals 

 move into the food-rich shallow water at night to 

 feed in the dark and retreat into deeper water at 

 sunrise to escape the efficient visual predators of 

 the epipelagic zone (Marshall 1954). If this theory 

 is correct, an examination of the feeding chron- 

 ology of vertical migrators should reveal that the 

 majority, at least, of feeding occurs at night. 



Table 4 compares the stomach contents of day- 

 caught with night-caught sergestids from the 

 DSB III cruise of February 1973. The night sam- 

 ples as a whole had a lower percentage of empty 

 stomachs, a greater amount of food in the 

 stomachs, and a lesser degree of digestion than the 

 day samples, indicating that most feeding oc- 

 curred at night. Unfortunately, only two species 

 were abundant both day and night. Sergestes 

 armatus fed more at night than during the day- 

 time, although most specimens had empty 

 stomachs regardless of time of day. Sergestes 

 erectus actually had a lower percentage of empty 

 stomachs during the daytime than at night, but 

 the night specimens on the average were fuller 

 than the day specimens. Other studies of feeding 

 chronology in sergestids, notably those of Omori 

 (1969) on Sergia lucens, Judkins and Fleminger 



(1972) on Sergestes similis, and Foxton and Roe 

 (1974) on a variety of Atlantic species, also in- 

 dicated that most feeding occurs at night. How- 

 ever, the DSB III day samples contained a number 

 of individuals with appreciable amounts of food in 

 their stomachs, showing that a certain amount of 

 feeding occurs during the daytime. Donaldson 



(1973) found that Sergestes sargassi, S. pectinatus, 

 and Sergia japonica appeared to feed around the 

 clock. The first two species also live in Hawaiian 

 waters; unfortunately, they only occurred in the 

 night samples of DSB III, so this study could not 

 test his observations. 



If Hawaiian sergestid species have specialized 

 by dietary preference, they might be expected to 

 exhibit specialized structures for catching prey. 



829 



