566 



Fishery Bulletin 92(3), 1994 



The scale method appears to underage older weak- 

 fish. Assuming otolith ages were valid, 4 of the 5 fish 

 in this study older than age 6 were underaged by 

 scales. Although 4 out of 155 fish may seem insig- 

 nificant, the importance of correctly ageing these fish 

 cannot be judged only by the number of discrepan- 

 cies. These fish represent the beginning of an asymp- 

 tote in growth and fish in the asymptotic range are 

 often rare in highly exploited stocks. Obtaining and 

 correctly ageing a few weakfish in this range is criti- 

 cal to correctly estimating the parameters of the von 

 Bertalanffy growth curve. 



Annulus formation on weakfish otoliths and scales 

 shows different patterns. The formation of otolith 

 annuli over a discrete time period suggests it may 

 be caused by environmental variables. The most com- 

 monly suggested environmental influences on annu- 

 lus formation are temperature, salinity, food, and 

 light (Simkiss, 1974). Weakfish form annuli on their 

 otoliths in April and May, when they migrate from 

 offshore winter grounds to estuarine feeding and 

 spawning grounds. Thus, annulus formation may be 

 linked to their migration into a different environment. 



Weakfish scales, in contrast, have a more variable 

 time of annulus formation suggesting a cause other 

 than general environmental conditions. Scales may 

 undergo resorption whereas otoliths do not ( Simkiss, 

 1974 ), and spawning has been linked to scale resorp- 

 tion with a consequent scale mark in salmon and 

 trout (Crichton, 1935). Spawning may also be linked 

 to formation of annuli on weakfish scales (Merriner, 

 1973). Weakfish mature at age 1 (Merriner, 1976; 

 Shepherd and Grimes, 1984) and are multiple spawn- 

 ers with a protracted spawning period from May 

 through August (Lowerre-Barbieri"). However, indi- 

 vidual spawning periods are asynchronous and vary 

 greatly, especially in time of termination. Spawning 

 activity and annulus formation may be linked in two 

 ways: 1 ) annuli could form on scales early in the 

 spawning season when resources are shifted towards 

 production of reproductive materials — especially the 

 yolking of oocytes, or 2) annuli might form near the 

 end of the season, owing to the cumulative drain of 

 protracted spawning, causing a cessation in growth 

 and thus an annulus. A connection between scale 

 annulus formation and spawning in weakfish would 

 explain the high level of variation in time of annulus 

 formation and the higher accuracy of ages based on 

 scales taken from females, because females usually 

 invest more energy in reproduction. It might also 

 explain the small growth increment between annuli 



1 and 2 if one-year-old weakfish begin spawning later 



2 Lowerre-Barbieri, S. K. 1993. Reproductive biology of weakfish, 

 Cynosaon regalis, in the Chesapeake Bay region. School of Ma- 

 rine Science, VIMS. College of William and Mary, unpubl. manuscr. 



in the season than older fish, owing to a threshold 

 size necessary to reach maturity. 



Our results indicate both scales and otoliths 

 present problems for back-calculation of weakfish. 

 Although scales showed a strong relationship be- 

 tween body and hardpart size and no seasonal dif- 

 ferences in growth, their long and variable time of 

 annulus formation may cause considerable error 

 (Smith, 1983). It is impossible to determine if a fish 

 formed its annuli at the same time each year. Be- 

 cause annuli can form from April to August, incre- 

 ments may represent 8-16 months of growth rather 

 than approximately one year of growth. Additionally, 

 scale annuli are more difficult to distinguish than 

 otolith annuli, making SAR's difficult to measure and 

 somewhat subjective. However, otoliths show sea- 

 sonal change in the body to hardpart relationship, 

 making a season-specific back-calculation equation, 

 such as we developed, inappropriate for fish collected 

 outside of that season. Additionally, comparisons 

 between back-calculated and observed sizes at age 

 were complicated by the weakfish migrational pat- 

 tern, since weakfish age ranges in the Chesapeake 

 Bay vary seasonally — older fish are present only in 

 spring and only occasionally in fall (Joseph, 1972). 



There was no clear evidence of Lee's phenomenon, 

 as older fish did not consistently show smaller 

 hardpart size at younger ages. The five oldest fish 

 did, however, demonstrate considerably smaller 

 OAR's at age 1 than did their younger counterparts. 

 Nevertheless, these same fish did not demonstrate 

 consistently smaller OAR's at consecutive ages than 

 did younger fish. Thus, the smaller OAR's at age 1, 

 rather than demonstrating Lee's phenomenon, may 

 simply reflect when most fish of those year classes 

 were born, i.e. fish born early in the spawning sea- 

 son would have larger OAR's at age 1 because they 

 had more time to grow before winter, than did fish 

 born later in the season. 



Previous criticism of back-calculation has focused 

 mainly on the body size to hardpart relationship and 

 its calculation (Campana, 1990; Casselman, 1990; 

 Francis, 1990; Ricker, 1992). However, the validity 

 of back-calculation also depends on the constancy, 

 clarity, and pattern of hardpart growth increments. 

 The different growth increment patterns we found 

 between scales and otoliths demonstrate the need to 

 understand hardpart growth better, how it relates 

 to somatic growth and what causes annulus forma- 

 tion on different hardparts. 



Future studies of weakfish age and growth should 

 be based on sectioned otoliths because scales appear 

 inaccurate once growth becomes asymptotic. This 

 common failing of the scale method has been reported 

 for many species (Beamish and McFarlane, 1987 ). It 



