MILLER and SUMIDA: DEVELOPMENT OF CARANX MATE 



larvae in a tank are selected for preservation. 

 This effect is apparent from the larger (than 

 predicted from the curve) mean size of the five 

 fish on day 36, which were the last specimens 

 in the tank when it was emptied. 



One source of error in relating early growth 

 rates of larvae to those of larger larvae is their 

 shrinkage upon preservation. Five groups of ten 

 live larvae (one to five days old), ranging in mean 

 standard length from 2.46 to 2.85 mm, ranged in 

 length from 2.13 to 2.55 mm 24 h after preserva- 

 tion. Shrinkage was also observed to begin within 

 seconds after death when larvae died while being 

 observed microscopically. This latter observa- 

 tion suggests that shrinkage was not entirely due 

 to the effects of formaldehyde on body proteins. 

 The percent shrinkage was not correlated in a 

 simple way with size. Although, presumably, 

 this percentage decreases with increased size of 

 larvae, the shrinkage values, which ranged from 

 8 to 22'7f in the five groups, can introduce 

 significant error into estimates of early larval 

 growth. Newly hatched larvae shrank as much as 

 30% when they died. 



Farris (1959) described two growth stanzas of 

 the jack mackerel, viz. (A) from hatch to day 3 

 and (B) from day 3 to 7. If these corres- 

 pond to our first two segments, then the second 

 segment (day 3-7) of growth in omaka is twice 

 that of jack mackerel (0.195 mm/day compared to 

 0.10 mm/day). Alternatively, comparing growth 

 from hatching to yolk absorption — day 3 in 

 omaka, day 6 in jack mackerel — yields a similar 

 difference, 0.48 and 0.26 mm/day, respectively. 

 The comparisons suggest that effects of starva- 

 tion may occur prior to complete yolk absorption. 

 Farris' growth rates for segment B (on starved 

 fish) may be underestimates. Lasker, Feder, 

 Theilacker, and May (1970) found that larvae 

 may begin to feed before complete yolk absorp- 

 tion. Comparisons need to be made between 

 starved and fed yolk-sac larvae of the same 

 species reared in the same physical environment 

 before a definitive answer can be reached. 



BODY PROPORTIONS 



As Marr (1955) pointed out, expression of 

 relationships between body dimensions as ratios 

 contributes nothing more than plots of the 

 original measurements, so the latter were used. 

 Relationships between standard length and 



1) head length, 2) eye diameter, 3) snout to anus 

 length, and 4) body depth at pectoral fin were 

 all adequately described by an equation of the 

 form: Y = a + 6(SL). All of the data used in 

 the regressions are from one series of reared 

 omaka larvae (Table 1). The ratios all adequately 

 describe specimens captured in the field. 



In the following discussions of these relation- 

 ships, comparisons are made between the omaka 

 and jack mackerel (T. symmetricus), described 

 by Ahlstrom and Ball (1954). The latter is the 

 only carangid larva for which these kinds of data 

 are published. With similar data for other 

 carangid species, these may prove useful in a 

 key to carangid larvae. 



Head Length 



Head length was related to standard length 

 according to the equation: HL = -0.2796 + 0.3477 

 (SL in mm) (Figure 5). Unlike T. symmetricus 

 (Ahlstrom and Ball, 1954), there was no inflection 

 in the curve at ca. 4 mm. The slope of the regres- 

 sion line for omaka (0.3477) is not very different 

 than that for the jack mackerel (0.378), so this 

 ratio would not be very useful by itself in dis- 

 tinguishing the two species. Ahlstrom and Ball 

 ( 1954) did find a different slope (0.556) in the jack 

 mackerel larvae smaller than 4.2 mm, but several 

 of our smaller larvae would fit either regression. 



Eye Diameter 



The relationship between eye diameter and 



STANDARD LENGTH (MM) 



Figure 5. — Relationship between standard length and head 

 length ofCaranx mate larvae. 



511 



