PHOTOPERIODISM IN FISHES 659 



THE GASTEROSTEID SEXUAL CYCLE 



The gasterosteid (stickleback) cycle is the only fish sex cycle other 

 than the cyprinid studied enough to afford comparisons as an annual 

 entity. Although it has interested several workers (Craig-Bennett, 

 1931; Merriman and Schedl, 1941; vanden Eeckhoudt, 1946; Kazan- 

 skii, 1952; Baggerman, 1957), a fully coherent, integrated account of 

 this cycle, with regard both to histophysiologic and behavioral re- 

 sponses to extrinsic factors at all seasons of the year cannot yet be 

 given. Craig-Bennett ( 193 1 ) described the normal seasonal changes in 

 the testicular histology of Gasterosteus aculeatiis and showed that 

 spermatogenesis may be completed at variable intervals from as much 

 as 21/2 months to just in advance of functional maturity, and that 

 breeding color is the concomitant of the latter, not the former. How- 

 ever, his conclusion (from experiments) that the sex cycle is unaffected 

 by day length is contradicted by results obtained by others (vanden 

 Eeckhoudt," 1946; Kazanskii, 1952; Baggerman, 1957). His experi- 

 mental specimens (all males) developed nuptial color whether exposed 

 to the ostensible short or long photoperiods (Craig-Bennett, 1931, pp. 

 238-248), which, if the results of the other experimenters are valid, 

 implies an inadvertent covert source of effective supplementary il- 

 lumination. Since the influence of temperature is meaningless without 

 reference to day length and vice versa, uncertainty about lighting con- 

 ditions makes uncertain the results ascribed to variant temperatures. It 

 is clear, however, that in late summer the extinction of breeding color 

 can be delayed by depressing, and hastened by elevating the tempera- 

 ture. 



The experiments of vanden Eeckhoudt (1946) on Gasterosteus were 

 begun between the middle and end of February, ending within two 

 months; in all of them the temperature was the same (10°C). Males 

 from the wild in both December and February had quiescent testes 

 with lobules bulging with spermatozoa and walls bordered by resting 

 spermatogonia, other cell types lacking, a condhion strictly comparable 

 to that of the autumn-winter testes of Enneacanthiis (Harrington, 

 1956). Under increasing day length, however administered, this condi- 

 tion of the testes persisted, but a constant short day length (8 hr) 

 caused the unexpected resumption of spermatogenetic activity, with 



