CLARKE: ECOLOGY OF STOMIATOID FISHES 



Gonostoma ebelingi (306; 12-158 mm) 



Gonostoma ebelingi occurred at 520-700 m dur- 

 ing the day and at 125-300 m at night. The size 

 composition of the catches changed with depth for 

 both day and night series. The largest taken shal- 

 lower than 150 m at night was 45 mm, and most 

 fish taken below 200 m exceeded 60 mm. During 

 the day, few fish over 100 mm were caught above 

 600 m, and none less than 75 mm were caught 

 deeper. 



Female G. ebelingi matured at about 120 mm, 

 but males, as far as could be told without histologi- 

 cal studies, appeared to mature at about 100 mm. 

 There was, however, no evidence to suggest that 

 this species is a protandrous hermaphrodite as 

 observed in G. gracile by Kawaguchi and Marumo 

 ( 1967). There were no obvious seasonal differences 

 in the percentages of ripe females among mature 

 females, but differences in size composition of the 

 catches of different series suggested some season- 

 ality in spawning. 



All G. ebelingi caught in December were over 70 

 mm. In March, all fish were either less than 50 mm 

 or over 70 mm. In June, AQF/c of the catch was 20-60 

 mm and the rest about evenly distributed between 

 61 and 150 mm. In September, the catch was again 

 bimodal with all fish either smaller than 70 mm or 

 over 100 mm. Recently transformed juveniles 

 were taken only in March and September and 

 were most abundant in March. More data would, 

 of course, be helpful, but it seems that G. ebelingi 

 spawns principally in early spring and early fall. 



Gonostoma elongation (1,346; 10-218 mm) 



In all series but December 1970, G. elongatum 

 occurred at 560-725 m during the day and moved 

 to 60-265 m at night. In December, no individuals 

 over 115 mm long were caught in the upper layers 

 at night, but larger fish were taken in two night 

 tows within the day depth range (of 22 fish, 9 were 

 117-200 mm. Figure 3). These deep night catches 

 were not large relative to those expected from 

 contamination, and consequently may have re- 

 sulted from encountering patches in the shallow- 

 layers. However, the difference in size suggests 

 that the large individuals did not migrate. 



Size-frequency curves from night samples at dif- 

 ferent depths during the same series were fre- 

 quently significantly different and consistently 

 indicated that small fish occurred shallower (Fig- 



100 



50 75 100 125 150 



STANDARD LENGTH(mm) 



200 



Figure 3. — Cumulative size-frequency curves for Gonostoma 

 elongatum taken at 170 m (A, 25 individuals), 190-200 m (B, two 

 tows, 11), 265 m (C, 13), and 750 m (D, two tows, 22l at night 

 during December 1970. The curves were all significantly differ- 

 ent (P<0.05) from each other. 



ure 3). Those caught above 100-125 m were less 

 than 35-40 mm, most caught between 175 and 200 

 m were 60-80 mm, and larger fish were taken 

 mostly in tows below 200 m. It was not clear 

 whether a similar pattern existed at depth during 

 the day. 



There was considerable sexual difference in 

 size at maturity. Males appeared to reach matur- 

 ity at about 120 mm; the largest male was 161 

 mm. The smallest mature female was 193 mm. 

 Some small, clearly immature females (120-140 

 mm) were found, but unfortunately, no fish be- 

 tween 161 and 193 mm was collected. Although 

 histological studies of specimens of all sizes are 

 obviously necessary, the above data suggest that 

 some G. elongatum mature directly as females, 

 while others are protandrous hermaphrodites. 

 Kawaguchi and Marumo (1967) have shown that a 

 congener, G. gracile is a protandrous hermaphro- 

 dite. Butler ( 1964) has shown that in some species 

 of pandalid shrimps, a group within which pro- 

 tandrous hermaphroditism frequently occurs, the 

 degree of hermaphroditism varies throughout the 

 species' ranges. Butler suggests that this is a re- 

 sult of varying ecological factors. Varying degrees 

 of protandry may similarly occur among the 

 Gonostoma spp. 



Too few mature females were collected to assess 

 any seasonal trends in gonad ripeness, but the 

 pooled size composition data for each series indi- 



341 



