FISHERY BULLETIN: VOL. 83, NO. 3 



tionately with fish volume, since most slopes of In 

 (fat volume) on In (length) were near 3. Slopes 

 smaller than 3 occurred in those seasons when fish of 

 all sizes had little fat. Some slopes during seasons of 

 high fat volume were significantly greater than 3, in- 

 dicating that larger fish had disproportionately 

 greater fat volume than smaller fish. This was parti- 

 cularly true of, but not limited to, the species and 

 sexes with high-magnitude fat cycles (males and 

 females of 5. entomelas and females ofS.flavidus). It 

 would be interesting to discover whether larger 

 females of 5. pinniger have high-amplitude fat 

 cycles, since the slopes of the fat regressions for 

 female S. pinniger were relatively high. 



The dependence of fat volume on fish size and 

 season of collection, while applicable on average, 

 does not apply to every individual. The scatter about 

 the regressions of fat volume on length was large. 

 Some of the scatter was probably due to errors in 

 measurement, but a large part was real. We have 

 observed large differences in the amount of visceral 

 fat among fish of the same size and sex in the same 

 collection. The reasons for such variation may be dif- 

 ficult to resolve, since it is difficult to know the 

 history of individuals collected in the field. 



However, our data were adequate in most species 

 to detect seasonal changes on average. We wanted 

 to compare the timing of fat cycles and gonad cycles, 

 and thus learn something of the function of fat 

 storage in rockfish. Our initial hypothesis was that 

 stored fat would be used in reproduction, based on 

 Roberts' (1979) observations. 



The data for males do not support this hypothesis. 

 Fat and gonad cycles were nearly coincident, with 

 peaks in both cycles occurring in fall and/or summer 

 and lows in winter/spring. Thus the energy for 

 gonadal growth was probably derived from current 

 ingestion, not from material previously stored as fat. 

 Maximum somatic growth also coincides with the 

 summertime growth of testes and fat deposits (e.g.. 

 Love and Westphal 1981), so all three processes may 

 be competing for energy consumed during this time. 



The situation is more complicated in females. 

 Some fat was deposited between spring and summer 

 in females of S. entomelas, S. goodei, Sind S. Jlavidus, 

 and females of all species gained fat between sum- 

 mer and fall. Some gonadal growth took place be- 

 tween spring and summer in females of all species 

 except S. paucispinis. In all but 5. pinniger, 

 however, ovarian volume increased steadily between 

 summer and winter. The growth of ovaries through 

 fall was due largely to vitellogenesis, while the 

 greater growth of ovaries into winter was due 

 primarily to embryogenesis and hydration of ova 



(Moser 1967). Additional vitellogenesis probably 

 occurred during winter in the multiple spawners, 5. 

 paucispinis and S. goodei (Moser 1967; MacGregor 

 1970). 



In females, then, fat deposition usually began in 

 summer, perhaps slightly before the initiation of 

 vitellogenesis, but continued into the main period of 

 vitellogenesis between summer and fall. Thus, like 

 males, females of these species deposited fat more or 

 less concurrently with gonadal maturation and 

 somatic growth. The depletion of visceral fats oc- 

 curred between fall and spring in females of most 

 species, during and after gestation. It is possible that 

 fat reserves are used for the maturation of additional 

 ova in multiple spawners or are involved in the nutri- 

 tion of embryos (Boehlert and Yoklavich 1984), but 

 as in males, reserves are not used in the initial devel- 

 opment of gonads. 



There were two main differences in the fat cycles 

 of males and females: females usually had larger fat 

 cycles than males, and the peak fat volume of 

 females occurred in fall, while fat volume in males 

 usually reached a plateau that spread through sum- 

 mer and fall. It is possible that courtship activity in 

 males (Helvey 1982) draws energy from fat deposi- 

 tion between summer and fall, or that this activity 

 draws time from feeding. As a result, females may 

 continue to fatten after fat deposition ceases in 

 males. It is also possible that females require more 

 reserves in winter and they somehow are able to ac- 

 quire these extra reserves. 



The synchronous depletion of reserves in males 

 and females, however, indicates a common function 

 for such reserves. We suggest two possible func- 

 tions: 



First, some rockfish may migrate during the 

 period of fat depletion. Love (1981) presented 

 evidence for seasonal movements in SebaMes 

 paucispinis and S. entomelas off southern California, 

 but had no data on the extent or direction of move- 

 ment. Females of S. alutus undertake seasonal 

 migrations covering as much as 300 m of depth (sum- 

 marized in Gunderson 1977). Lenarz (pers. obs.) 

 found evidence of seasonal movement in female S. 

 jordani. Several species of rockfish undertake little 

 or no seasonal migration, including ^S. ./7a rvV/M.s from 

 shallow waters off Alaska (Carlson and Barr 1977). 

 Since seasonal migrations in rockfish may not be 

 long and seem to occur primarily in females, we sug- 

 gest that other possibilities be ruled out before ac- 

 cepting migration as a major use of fat reserves. 



Second, we suggest that fat reserves are used for 

 maintenance during wintertime periods of reduced 

 food availability. Fattening, as well as growth and 



308 



