RADTKK and DEAN: INCRKMKNT FORMATION OF OTOLITHS OF MUMMIOIKx; 



drigh and Bruce (1957), who showed that a light 

 stimulus synchronized emergence in fruit flies. 

 More study is necessary to determine the timing 

 of light needed for increment formation as well 

 as the quantity and quality of light necessary. 

 Whether the control of increment formation is an 

 endogenous or exogenous rhythm (Harker 1957) 

 is beyond the scope of these experiments. But the 

 experiment on increment initiation in the dark 

 group with 1 min of light exposure on day 10 in- 

 dicated that light can act as a synchronizing 

 stimulus, similar to that observed by Pittendrigh 

 and Bruce (1957). Mugiya et al. (1980) found that 

 D formation was initiated when light inter- 

 rupted a photo period of 12L:12D or longer light 

 period, but they did not determine the minimum 

 dark period necessary for formation of the Dor 

 the free running period for the D and I. 



When F. heteroclitus larvae were hatched in 

 L12:D12 and then placed in light regimes other 

 than a 24-h photoperiod, the increment forma- 

 tion became aphasic in each group and incre- 

 ment formation occurred at a slower rate. The 

 "biological clock" of this group seemed to be out 

 of phase under photoperiods other than those 

 with a 24-h periodicity. A great deal of very 

 exciting work is necessary to resolve these funda- 

 mental questions on increment control. 



Effects of Temperature and 



Body Growth on Otolith 



Formation in Larvae 



Under the various experimental conditions 

 employed in this study, daily otolith increments 

 formed regardless of body growth or otolith 

 growth rate (Fig. 6a, b, c), so it was possible to 

 determine age and daily growth rates of individ- 

 ual larvae which lived under different environ- 

 mental conditions. Although F. heteroclitus lar- 

 vae grew faster at 30°C than at24°C, the number 

 of increments was still directly related to chrono- 

 logical age. This documents the reliability of oto- 

 lith increments for the age estimation of mum- 

 michog larvae. It has been demonstrated that 

 daily increments exist in several other species of 

 fish (Pannella 1971, 1974; Brothers et al. 1976; 

 Struhsaker and Uchiyama 1976; Ralston 1976; 

 Taubert and Coble 1977) and the relationship be- 

 tween increment counts and fish and otolith size 

 was shown for the Atlantic silversides(Barkman 

 1978). In this study, otolith diameter increased 

 with increased body length and increments 

 formed on a daily basis with wider increments 



found in younger fish than older fish. This is con- 

 sistent with the fact that younger fish are grow- 

 ing faster, and although the relationship is non- 

 linear, it is predictable and these results are 

 consistent with those of Methot and Kramer 

 (1979). 



Estimation of Age of 

 Wild Fish 



Daily increments observed in field samples 

 were easier to interpret than increments found 

 in laboratory-reared larvae. We were not able to 

 make age estimations of field collections of mum- 

 michogs from length-frequency histograms, but 

 it was possible to determine the age and growth 

 rate of individual larvae from increment counts. 



Ralston (1976) and Struhsaker and Uchiyama 

 (1976) determined growth rates of the millet- 

 seed butterfly fish, Chaetodon miliaris, and the 

 nehu, Stolephorus purpureus, respectively, and 

 found that the growth, as represented in incre- 

 mental units in the otolith, was nearly linear. 

 Similar results were obtained by Barkman 

 (1978) for Atlantic silversides and Methot and 

 Kramer (1979). Our results are consistent with 

 theirs: that increment formation is independent 

 of growth rate but is age dependent; thus growth 

 rates can be estimated for individual larval fish. 



Analysis of the age structure of samples of wild 

 larval mummichogs showed that larvae hatched 

 on or near the time of full and new moons. This is 

 corroborated by observations on the reproduc- 

 tive biology of F. heteroclitus by Taylor et al. 

 (1977, 1979) and DiMichele and Taylor (1978), 

 New Zealand white bait, Galaxias maculatus, by 

 McDowell (1968), and Atlantic silversides by 

 Middaugh (1981). Eggs of the California grun- 

 ion, an intertidal spawner, have been found 

 to hatch during spring tides (Clark 1925) and 

 have otolith increments at hatching (Brothers et 

 al. 1976). An analysis of age structure of wild 

 populations of mummichog larvae, as deter- 

 mined from their otoliths showed that South 

 Carolina mummichogs spawn from March to 

 mid-August and have a lunar spawning perio- 

 dicity during that season. Analysis of otolith in- 

 crements enabled us to differentiate individual 

 fish in the wild population of the same size but of 

 different ages. 



Photoperiod is a critical factor in increment 

 formation, but other factors such as diurnal 

 migratory behavior, rhythmic feeding, tempera- 

 ture, respiration, and tidal rhythms might also 



213 



