FISHERY BULLETIN: VOL. 76, NO. 3 



an indication of dusk feeding activity in two 

 species of A rgy rope lee us; however, the data for A. 

 hemigymnus seem to me more consistent with in- 

 creasing stomach fullness throughout the day and 

 a nighttime decline. Except for high dawn values 

 (based on only three specimens from two tows), A. 

 aculeatus shows a similar trend. 



DeWitt and Cailliet ( 1972) found no diel trend in 

 feeding ofCyclothone signata, but, based on fewer 

 empty stomachs in fish caught in the upper part of 

 the depth range, proposed that this species, al- 

 though it does not undertake diel vertical migra- 

 tions, may ascend irregularly to levels of higher 

 prey concentration to feed. Their data also indi- 

 cated that a deeper living species C. acclinidens, 

 had a higher percentage of empty stomachs by 

 day; as noted by the authors, the latter seems to 

 defy any reasonable explanation. 



Legand et al. ( 1972) considered feeding chronol- 

 ogy of 14 species of mesopelagic fishes from the 

 South Pacific. Though trends in stomach fullness 

 of some species are similar to those noted here, 

 e.g., that for Triphoturus microchir (which almost 

 certainly = T. nigrescens) is very similar to that 

 for T. nigrescens near Hawaii, a number of species 

 show patterns quite different from those reported 

 by either the present or other studies. Interpreta- 

 tion of the validity of such "exceptions" is difficult 

 owing to the sparse presentation of Legand et al. 

 Though total numbers of specimens are fairly 

 high, it is not clear that they were equitably dis- 

 tributed among diel periods, from the same area, 

 or from the same season, etc. The percent fullness 

 values are obviously based on wet weights — an 

 imprecise measurement, particularly for stomach 

 contents — and it is not clear whether all fish and 

 stomach contents were weighed or some sort of 

 averaging or regression procedure was employed. 



The feeding patterns shown by previous studies 

 cannot be compared in detail with those presented 

 here; however, there is general agreement in data 

 on the two dominant groups of mesopelagic fishes. 

 Myctophids feed mostly at night, while stom- 

 iatoids tend to feed by day. My interpretations 

 indicate that near Hawaii, the differences are at 

 least partially related to different diel relation- 

 ships of the fishes to vertical distributions of their 

 prey. Other interpretations are obviously possible, 

 e.g., the feeding patterns may prove to be charac- 

 teristic of the two taxa regardless of relationship 

 to prey distribution. It would be of particular in- 

 terest to investigate myctophids with vertical dis- 



tribution patterns similar to those of the 

 stomiatoids, i.e., with shallow day depth ranges at 

 or near high daytime concentrations of zoo- 

 plankton. (Certain Myctophym and Diciphus spp. 

 from Hawaii meet this criterion [Clarke 1973], but 

 were not captured in sufficient numbers to be in- 

 cluded in this study.) 



The diel feeding patterns of mesopelagic fishes 

 could well be related to light rather than (or in 

 addition to) temperature and prey concentration. 

 No data on diel light changes near Hawaii are 

 available; however, data of Kampa ( 1970) from a 

 similar area of clear oceanic water in the North 

 Atlantic show that during full moon the diel 

 change in depths of relevant isolumes is of the 

 order of 300-350 m. Even allowing for considerable 

 differences in extinction coefficients between 

 Hawaii and Kampa's study area, the diel change 

 in isolumes at new moon (when the present sam- 

 ples were taken) off Hawaii is probably at least 

 300-350 m and could be as great as 500 m. The 

 absolute diel change in depth for most of the myc- 

 tophids is over 500 m while that for the 4 day- 

 feeding stomiatoids is ca. 400 m or less (Table 3). 

 Thus it is possible that feeding in both groups 

 occurs when higher light levels are encoun- 

 tered — at night for the myctophids and by day for 

 the stomiatoids. 



Estimation of Rates 



As mentioned previously, neither feeding rate 

 nor stomach evacuation rate can be considered 

 quantitatively without an independent estimate 

 of the other. Because of the difficulty in keeping 

 mesopelagic fishes alive for grazing or evacuation 

 experiments, it will likely be a long time before 

 independent estimates are available. For a few 

 species considered here it is, however, possible to 

 derive "quasi-independent" estimates of evacua- 

 tion rate given certain plausible assumptions. 

 These allow, with further assumptions, rough es- 

 timates of feeding rate and daily ration. 



For any period where feeding rate is zero, 

 changes in stomach fullness are due to evacuation 

 alone, and, if temperature, pressure, etc., remain 

 essentially constant during that period, the rate of 

 evacuation can be assumed to be proportional to 

 the amount of food in the stomach (Kjelson and 

 Johnson 1976; Eggers 1977). The change in 

 stomach fullness would then be described by: 



dSldt = -kS orS, = So^*' 



(1) 



510 



