Laidig et al.: Growth dynamics in early life history of Sebastes jordani 



617 



(0 



"H oo 



w 

 K -5.0-1 



10 



-15.0 



.,111:11:-" 



■■■:;::■ -. .. 



»:'!•! I:!.:-:- 



urn** i'::iliiii! m. 



IliiPte 



20 40 60 80 100 120 140 160 



Age (d) 



Figure 9 



Relationship between back-calculated standard length and age 

 for shortbelly rockfish. Solid line in the upper panel represents 

 the composite function of relationships shown in Figures 6 

 and 7. Filled circles are means of the observed data at 10-day 

 intervals, bracketed by ±2 standard errors. Dashed lines 

 enclose the range of data. Residuals (observed-predicted) of 

 the fit are shown below. 



Discussion 



Our goal was to develop a model that could be used to 

 provide accurate estimates of back-calculated growth 

 from the otolith microstructure of larval and juvenile 

 shortbelly rockfish. A crucial assumption of the study 

 was that the increments were formed daily. Many 

 authors have stressed the importance of validating the 

 daily periodicity of otolith growth increments (Jones 

 1986, Geffen 1987). Without validation, growth rates 

 may be biased and their use in models of juvenile 

 dynamics and recruitment could lead to serious errors. 

 For validation purposes, we followed a cohort of fish 

 through the May-June time period, measuring their 

 progressive change in average length. We found that 

 the rate of change in length of this cohort agreed with 

 the estimated growth rate based on a regression of 

 length against number of growth increments enumer- 

 ated for aged fish. From this result, we conclude that 

 the increments we counted were formed daily. It is 

 noted, however, that the daily periodicity of these in- 

 crements is validated only for the size ranges and dates 



observed during the extended May- June cruise, but we 

 have no reason to assume that the increments are not 

 daily at earlier stages of growth. 



Other evidence exists that daily increments form in 

 otoliths of Sebastes spp. Yoklavich and Boehlert (1987) 

 demonstrated the daily periodicity of increment forma- 

 tion in otoliths of black rockfish S. melanops by mark- 

 ing the otoliths with oxytetracycline (OTC) and by 

 autoradiography. Also, Laidig (unpubl. data) found no 

 difference between increment counts and the number 

 of days following a fluorescent OTC mark, verifying 

 that the increments of brown rockfish S. auriculatus 

 were interpreted correctly as daily. 



Our segmented model of fish size versus otolith size, 

 for use in back-calculating fish length, may describe 

 significant events during the early life of S. jordani. 

 The first major events following extrusion, in the early 

 life of rockfish, are first feeding, flexion, and transfor- 

 mation from larva to juvenile. Moser et al. (1977) noted 

 that shortbelly rockfish larvae undergo flexion at 

 8.0-10.0 mm SL. This range essentially corresponds to 

 the size of fish in segment II. The slope of this segment 

 was much reduced, suggesting that although fish 

 length increases little during flexion, the otolith con- 

 tinues to grow. Likewise, Moser et al. (1977) found that 

 juvenile transformation began after 27 mm SL, and we 

 observed that the intersection of segments III and IV 

 occurred at a length of about 30 mm SL. It is possible, 

 of course, that the segments v/e have described are 

 peculiar to the time and place that we collected the fish. 

 For example, the slope of the fish size to otolith size 

 relationship could have changed at specific points in 

 time in response to altering oceanographic conditions 

 (see also the discussion on "buffering" below). Before 

 this type of alternative explanation can be discarded, 

 similar SL-OR regressions need to be established in 

 other years. 



Relative to the entire organism, the sagittal otolith 

 is a simple structure, especially during early life 

 history. During the first 160 days of life, the sagitta 

 develops from a spherical primordium into an oval- 

 saucer shape. During this time a single Gompertz curve 

 (Fig. 6) adequately described growth along one of its 

 dimensions (i.e., the postrostral axis). 



In contrast, the whole organism is morphologically 

 and developmentally complex. Although a simple Gom- 

 pertz curve, when fitted to back-calculated length 

 (SLjj) against age, resulted in a high r 2 value, an un- 

 acceptable pattern was evident in a plot of the resid- 

 uals. Other authors also have found that smooth models 

 (e.g., Gompertz and von Bertalanffy curves) did not ac- 

 curately fit SL-at-age data in young-of-the-year fish 

 (Uchiyama and Struhsaker 1981, Bailey 1982, Rosen- 

 berg and Laroche 1982, Campana 1984, Boehlert and 

 Yoklavich 1985). Although these simplified models 



