346 



Fishery Bulletin 101(2) 



Figure 2 



Photomicrograph of daily growth increments from the otolith of a striped mullet at 

 lOOOx magnification 14 days after OTC marking. The OTC mark is indicated and 

 each subsequent daily growth increment is marked. 



We also estimated the accuracy of the assigned ages 

 of the fish. At the end of the study, 69 previously aged 

 specimens were chosen at random and re-aged without 

 knowledge of the prior assigned age. The percent agree- 

 ment between the two sets of age readings was examined, 

 and an age-bias plot was constructed for comparisons of the 

 coefficients of variation between the two sets of ages for the 

 same specimens (Campana et al., 1995). 



The growth of juvenile striped mullet and the relation- 

 ship of size at age once they recruit into the estuarine 

 nursery habitats were described by using a least squares 

 regression; 



Y = aX^± £, 



where Y = total length; 

 X = age; 



a = the y-intercept; 



b = the regression coefficient (slope); and 

 f = the error term. 



This would give a basic growth model from the juvenile 

 aging study. 



The distribution of the birthdates for the juveniles aged 

 was backealculated by subtracting the daily age of the fish 

 from the date of capture. For the historical-sui-vey fish (cap- 

 tured with rotenone), age was predicted from the application 

 of the growth equation (least squares regression of size at 

 age) to the observed lengths. We assumed in using the pre- 

 dicted age to estimate birthdates for the historical-survey 



fish that both groups (i.e. historical and current survey) 

 had equal variances and similar growth patterns. 



The historic data analyzed in our study were obtained 

 from rotenone and plankton-net surveys in South Carolina 

 from 1986 to 1991. The sampling technique for these surveys 

 has been described (see Wenner et al.'). These data included 

 length and hydrographic data similar to the data collected 

 during our juvenile aging study and represented an addition- 

 al 5498 fishes. The historical data were collected from three 

 different salinity zones (oligohaline, mesohaline, and polyha- 

 line) primarily within the Charleston Harbor estuary. The 

 growth of juvenile striped mullet for the historical data was 

 calculated from changes in mean total length per month by 

 using a least squares regression. For fishes collected with ro- 

 tenone, month of capture was substituted for age in the model. 

 Mean size per month for each year of the rotenone survey 

 (1986 to 1991) was determined separately. The growth from 

 the juvenile aging study and from the historical data were 

 then compared by using an analysis of covariance ( ANCOVA) 

 to compare the differences in the slopes and intercepts. 



The different salinity regimes within the estuaries were 

 defined by using the Venice scale of estuarine salinities 

 (Anonymous, 1959). This scale provides five distinct salin- 

 ity regimes within an estuary: freshwater (0 ppt), oligohaline 

 (0.1 to 5 ppt), mesohaline (5.1 to 18 ppt), polyhaline (18.1 

 to 30 ppt), and euhaline (>30.1 ppt). Growth was compared 

 between the different salinity regimes and significance was 

 tested by comparing the changes in mean size per month 

 within each salinity regime against the others by using an 

 ANCOVA. 



