FISHERY BULLETIN: VOL. 70, NO. 4 



from central California south to Magdalena Bay, 

 Baja California, and the Gulf of California south 

 to Mulege on the west coast, and south to Agia- 

 bampo Bay on the east coast (Barlow, 1963). 

 The typical habitat of this species is the inter- 

 tidal region of coastal sloughs. Barlow (1961, 

 1963) discussed the systematics and some as- 

 pects of the ecology of G. mirahilis. The popu- 

 lation of Gillichthys used in these studies occurs 

 in the Alviso salt ponds located at the southern 

 end of San Francisco Bay, Calif. Carpelan 

 (1957) described seasonal changes in the hy- 

 drobiology of these ponds. 



De Vlaming (1972b) described the reproduc- 

 tive cycle of G. mirahilis and suggested that sea- 

 sonal temperature changes may be involved in 

 regulating sexual cycling in this species. The 

 spawning period is protracted, extending from 

 December to June. Gonadal regression occurs 

 in July; the gonads remain regressed during 

 August and September. Gonadal recrudescence 

 begins in late September, reaching completion 

 by early December. 



The aim of the present study was to determine 

 the effects of various light and constant temper- 

 ature regimes on gonadal function in G. miror 

 bills, with the hope of evaluating the influence 

 of these factors in regulation of the annual sex- 

 ual cycle. The phenological data on reproductive 

 cycling in this species presented by de Vlaming 

 (1972b) was used as a basis for these studies. 

 Some of the previous studies with teleosts have 

 shown that the effect of the environmental 

 synchronizer (s) varies with the stage of gonadal 

 maturity. Consequently, the effects of photo- 

 period and constant temperature treatments 

 were examined during different phases of the 

 gonadal cycle. 



MATERIALS AND METHODS 



Samples of G. mirabilis were captured in the 

 Alviso habitat at several different times during 

 the year and thus in different phases of gameto- 

 genesis. Since males were more abundant in 

 these samples, a greater number of experiments 

 were conducted with this sex. Several fish from 

 each sample were sacrificed, and the gonads ex- 

 amined at the time of capture; these fish served 



as a reference for the experiments that followed. 

 In the following discussion the fish sacrificed 

 from the samples from nature will be referred to 

 as initial controls. In many of the experiments, 

 samples of fish from the natural habitat were 

 collected and sacrificed upon termination of the 

 experiment; these fish will be referred to as 

 terminal controls. To facilitate quantification 

 of gonadal response, animals of approximately 

 equal size were utilized in these experiments. 



Experimental fish were maintained in 56- (no 

 more than 10 fish per tank) or 132-liter (no more 

 than 18 fish per tank) tanks. Recirculating 

 filtered seawater was used in all of these exper- 

 iments. The bottom of each tank was covered 

 with fine gravel. The experimental tanks 

 were housed in constant temperature rooms 

 (± 1.5°C). Temperatures selected for these 

 experiments are within the range normally ex- 

 perienced by this species during the year. 



Various photoperiods were also employed in 

 these experiments. Light was provided by 20-w 

 warm-white fluorescent bulbs suspended above 

 the tanks. Salinity was maintained at a con- 

 stant 35 %c, and pH between 8.0 and 9.5 (these 

 pH's are consistent with those experienced by 

 the fish in the Alviso ponds). The fish in these 

 experiments were provided with a varied diet 

 consisting of brine shrimp, chopped fish, boiled 

 egg white, and beef kidney and liver; all fish 

 ate voraciously. 



Upon termination of each experiment, the 

 weight and standard length of each fish were re- 

 corded. Gonads were weighed and prepared for 

 histological examination in the same manner as 

 previously reported (de Vlaming, 1972b) . Grav- 

 imetric data are expressed in absolute weights 

 since it was shown (de Vlaming, 1972b) that 

 gonadal weight is independent of body weight 

 (and length) in the size range used. Spermato- 

 genesis and oogenesis were divided into six and 

 five recognizable phases (Tables 1 and 2) to fa- 

 cilitate quantitative evaluation of gametogenetic 

 activity. 



Statistical comparisons of gonadal weights be- 

 tween experimental groups were made by using 

 the Mann- Whitney U test (Siegel, 1956, p. 184- 

 193). This nonparametric test is suitable for 

 small sample sizes and can be used to determine 



1138 



