Kramer: Growth and mortality rates of juvenile Paralichthys cahfornicus 



197 



weighting coefficient. Variance of the weighted mean 

 was underestimated because the covariance terms were 

 not included. Resampling techniques to estimate vari- 

 ance (e.g., bootstrap) were impractical because of the 

 large size of the database. 



Age validation and determination 



Laboratory-reared halibut larvae of known age were 

 measured to standard length in mm and their sagittae 

 excised and mounted in resin (Eukitt, 0. Kindler, West 

 Germany) on a microscope slide. Age was estimated 

 using the methods of Methot (1981) and Butler (1987). 

 A microcomputer interfaced to an electronic digitizer 

 was used to measure and count increments on a pro- 

 jected image of the otolith from a high-resolution video 

 camera mounted on a compound microscope. Incre- 

 ment counts of 45 larvae (3.1-9.1 mm SL) that were 

 reared at 16-20°C in the laboratory were regressed 

 against the known age of the larvae to establish a rela- 

 tionship between estimated and known age. Incre- 

 ments were formed daily: the slope of the relationship 

 (0.969) did not differ significantly from unity (P>0.05). 

 The regression of the number of increments on age of 

 halibut larvae (5-29 days) was 



Age (days) = 3.496 + 0.969 x (no. increments) 



where r 2 = 0.981, SE constant = 1.055, SE slope = 

 0.018, and range of increment counts = 1-26) (Fig. 2). 

 Daily formation of rings has also been found in juveniles 

 30-70 mm SL (Kicklighter 1990). The first increment 

 is deposited about 3.5 days after hatching, coinciding 

 with the day of first feeding (Gadomski and Peterson 

 1988). I added 3.5 to the number of increments counted 

 on the otolith so that age was equivalent to the number 

 of days from hatching. 



Ageing of field-caught halibut 



Juvenile halibut from field collections were measured 

 alive and either frozen or preserved in 80% ethanol. 

 Sagittae were dissected and increments counted using 

 the techniques described above. Sagittae from juveniles 

 >20mm SL were polished with 400- and 600-grit wet 

 sandpaper before counting. 



A total of 120 field-caught halibut were aged: 50 

 from Mission Bay, 19 from Agua Hedionda Lagoon, 

 and 51 from the open coast. Larval sagittae are sym- 

 metrical and nearly circular (Fig. 3A), but after meta- 

 morphosis additional foci develop and the sagittae 

 became asymmetrical, with maxiumum deposition 

 along the rostral axis (Karakiri et al. 1989) (Fig. 3B). 

 This shift in the axis of sagittal growth produces areas 



KNOWN AGE (days) 



Figure 2 



Age validation of California halibut sagittae. Number of in- 

 crements counted on the sagittae are compared with the 

 known age of laboratory-reared larval halibut (n 45). Straight 

 line represents a one-to-one relationship of increment number 

 and known age. 



that are difficult to interpret (Fig. 3). These areas 

 correspond to a period of about 7 days after metamor- 

 phosis. I estimated the number of increments in regions 

 of transition between foci by counting the number of 

 increments that occurred in an adjacent area on a dif- 

 ferent axis (Fig. 3). The relationship between standard 

 length (mm) and otolith radius (^m) was linear for 

 halibut MOmrn SL (Fig. 4). 



Mortality estimates 



I did not use data from the 1987 survey for estimating 

 mortality because nearly all of the 1987 year-class oc- 

 curred in bays and comparisons of mortality between 

 bay and coast habitats were an essential step in the 

 analysis. The relationship between abundance and age 

 (estimated from the length-at-age relationship) of the 

 1988 year-class was used to estimate age-specific mor- 

 tality rates. 



I used seven different models to estimate age-specific 

 instantaneous mortality rates. Three of the models 

 were estimates based on the following assumptions 

 regarding the relationship between survival rates and 

 age (Barlow 1982): (1) Age-specific survival rates in- 

 crease linearly with age; (2) age-specific survival rates 

 increase exponentially with age; and (3) age-specific 

 survival rates approach an asymptote with age. The 

 two daily production models estimated age-specific in- 

 stantaneous mortality rates based on the relationship 

 between daily production (abundance of length class/ 

 duration of length class) and age (Lo 1985). The last 



