CLARKE: FECUNDITY AND SPAWNING FREQUENCY OF NEHU 



be unequivocally discriminated under a dissecting 

 microscope on the basis of size or opacity, the 

 ovaries were teased apart and all ova in the ad- 

 vanced mode counted. This technique eliminated any 

 error in fecundity determination due to subsampling 

 of the ovaries, but meant that very few determina- 

 tions were based on specimens with oocytes smaller 

 than ca. 0.65 mm. In most of the latter cases, even 

 if an advanced mode was clearly evident from the 

 size-frequency determinations, it could not be 

 unequivocally discriminated for total counts under 

 the dissecting scope. Females with hydrated ova 

 free from the follicles and segregated from the 

 smaller oocytes were not used for fecundity deter- 

 minations. 



After the oocytes in the largest mode were 

 counted, the entire ovaries were rinsed with distilled 

 water from the slide into a preweighed aluminum 

 pan. The stomach contents were removed from the 

 fish and the body cavity was examined for parasites, 

 specifically the presence of ca. 5 mm long nematodes 

 around the liver and pyloric caeca. The fish was 

 placed in a preweighed pan, and any tissue remain- 

 ing on the slide was rinsed with distilled water into 

 the same pan. The fish and gonads were dried at 

 60 °C for 24 h after which the pans were reweighed 

 to the nearest 0.1 mg, and dry weights of the fish 

 and gonads determined by subtraction. 



In all cases, fecundity and relative fecundity refer 

 strictly to batch fecundity. Relative fecundity will 

 be given as eggs per gram ovary-free dry weight, 

 and gonad to somatic weight ratio (G/S) will be given 

 as percent of dry weight values. Dry weights were 

 used because of the difficulty in making consistent 

 wet weight determinations on such small fish and 

 even smaller ovaries. Careful wet-dry weight deter- 

 minations on 10 females and gonads indicated that 

 preserved nehu without gonads are about 73% water 

 and that ovaries with yolked, but unhydrated, 

 oocytes are about 60% water. To compare nehu 

 fecundity data with those from other studies which 

 had used wet weights, individual nehu dry weights 

 were divided by 0.27, and relative fecundity and 

 fecundity weight relationships were recalculated. 

 This procedure admittedly ignored any variability 

 in the wet-dry weight relationship. The G/S values 

 given here can be multiphed by 0.675 (0.27/0.40) to 

 make them roughly comparable to values based on 

 wet weight from other studies. 



Unless otherwise noted, all regressions given 

 below are Model II (or "functional"), GM regressions 

 (Ricker 1973). Results of regressions using natural 

 logarithms are expressed as power curves (antilog 

 form). The 95% confidence limits for slopes of linear 



regressions and exponents of power curves ( = 

 slopes of In-ln regressions) were calculated from for- 

 mulae in Ricker (1975). For any previous studies 

 which had given results from Model I regressions, 

 original fecundity and weight data were used to 

 calculate functional regressions. 



RESULTS 



Maturity and Oocyte Development 



The smallest mature females were 35 mm SL, the 

 same minimum size reported by Leary et al. (1975), 

 but in many of the samples most of the fish <40 mm 

 SL were immature. Among the fish from the 36 

 samples from which more than cursory examina- 

 tions were made, 30% of the 134 specimens <40 mm 

 SL were immature. Only 8% of the 227 between 41 

 and 45 mm SL and <2% of the 284 over 45 mm were 

 immature. 



Nehu oocytes begin to elongate at about 0.3 mm 

 in length but remain relatively translucent with little 

 visual evidence of vitellogenesis until about 0.4 mm 

 long. Oocytes longer than 0.5 mm were almost 

 always opaque, and those over about 0.6 mm were 

 densely opaque and yellow to yellow-brown in color. 

 The first signs of hydration appeared in oocytes 

 about 0.75 mm long. The yolk became more trans- 

 lucent and globular rather than granular in apparent 

 texture, and the perivitelline space was evident at 

 one or both ends. All ova longer than 0.8 mm were 

 white in appearance and had an evident perivitelline 

 space. At about this size or slightly larger, ova had 

 left the follicles and begun moving to the main ovi- 

 duct. Comparisons of fish from closely spaced purse 

 seine samples taken just before and during spawn- 

 ing indicated that migration of hydrated ova from 

 the follicles to the oviduct occurred in <0.5 h. Only 

 in a few fish with the ova segregated or partially 

 spawned were one or two hydrated ova left in the 

 follicles. Apparently once hydration begins, all ova 

 in a batch are normally ovulated and spawned at one 

 time. 



Separate batches of maturing oocytes become 

 distinct from the numerous small oocytes between 

 0.45 and ca. 0.60 mm. In fish with LMX <0.45-0.50 

 mm there was little or no evidence of a separating 

 size-frequency mode of oocytes. Variably separated 

 modes with LMD at 0.45-0.55 mm were present in 

 fish with the LMX at 0.55-0.65 mm. Usually modes 

 centered at 0.60 mm or larger and with LMX over 

 0.65 mm were clearly separated from smaller and 

 less opaque oocytes. There was no evidence from 

 size-frequency data that, once oocytes reached ca. 



129 



