FISHERY BULLETIN: VOL. 80, NO. 3 



that 5. caprinus reaches 90-123 mm TL at age I. 

 Back-calculated lengths of S. chrysops are 120- 

 155 mm TL at age I and 182-213 mm at age II 

 (Finkelstein 1969a; Hamer footnote 9). Sizes at 

 age I appear to be similar in these species, but S. 

 chrysops is much larger at age II. Our growth 

 data and the constant size noted in later gonad 

 maturity stages (Fig. 6) indicate that S. caprinus 

 markedly diverts energy from somatic growth to 

 gonadal development as it matures. The differ- 

 ent sizes of these congeners at age II probably 

 reflect this drastic diversion of energy in S. ca- 

 prinus. 



MAXIMUM SIZE AND LIFESPAN 



Results 



The maximum size typically reached by S. 

 caprinus is about 200 mm TL. The largest fish 

 collected off Texas (n = 22,924) was 182 mm TL, 

 and the largest specimen collected in the north 

 central Gulf (n — 1,576) aboard the Oregon II was 

 193 mm TL. 



The maximum lifespan of S. caprinus typically 

 appears to be 2.5-3 yr. In the period October 

 1977-March 1980, 99% of the specimens captured 

 off Texas were <146 mm TL, and 99.5% were 

 <149mm(Fig. 11). Many age II fish captured off 

 Texas were as large as 155 mm TL. The largest 

 fish collected was age II when captured in March 

 1979, or age III if it was a member of the poorly 

 defined 1976 year class (Fig. 2). These data indi- 

 cate that a value of ti = 2.5-3 yr is reasonable for 

 this Beverton-Holt model parameter (Gulland 

 1969). This estimate is supported by data from 

 the north central Gulf (Fig. 7) in which 99% of the 

 fish were <172 mm TL and 99.5% were <176 

 mm. Fish of these sizes were age II or III in the 

 north central Gulf (Fig. 7). 



Discussion 



The maximum sizes reported herein are 

 slightly larger than the maximum size reported 

 by Hildebrand (1954), Caldwell (1955), and Chit- 

 tenden and McEachran (1976). The only pub- 

 lished record of S. caprinus much larger than 

 200 mm TL is that of Franks et al. (1972) who 

 captured a specimen 256 mm TL in the north 

 central Gulf. The apparent larger size of individ- 

 uals at given percentages of the catch in the 

 north central Gulf might reflect growth differ- 

 ences between areas, or the nonrandom sub- 



58CN 



80 120 160 



TOTAL LENGTH (mm) 



200 



Figure 11.— Length frequency (moving averages of three) and 

 cumulative percent frequency of all Stenotomus caprinus col- 

 lected off Freeport, Tex., October 1977-March 1980. 



sampling from the Oregon //catch which would 

 probably select larger fish. Values of ti would 

 vary from year to year due to variation in post- 

 spawning survival noted later. 



MORTALITY AND POSTSPAWNING 

 SURVIVAL 



Results 



Stenotomus caprinus has a total annual mor- 

 tality rate of about 83-99% on a time-specific 

 basis. Time-specific total annual mortality rates 

 (1 — S) were calculated from the expression S = 

 N ( /No where S = rate of survival, and N, and No 

 are the numbers of fish collected at age each 

 cruise in depths of 18-100 m. The pooled esti- 

 mate, using Heincke's procedure (Ricker 1975) 

 was 98.95% comparing the 1977 and 1978 year 

 classes and individual rates generally exceeded 

 98%. These values probably overestimate 1 — S, 

 because the 1978 year class was stronger than 

 the 1977 year class (Fig. 10). The pooled estimate 

 comparing the apparent 1976 and 1977 year 

 classes was 99.79%. The pooled estimate compar- 

 ing the 1978 and 1979 year classes was 84.94% 

 which may be fortuitous because it included one 

 data set (October 1979) for which an exception- 

 ally large number of fish from the 1979 year class 

 were captured. A minimum pooled mortality 

 estimate for the 1978 and 1979 year classes is 

 40.73%, if the exceptional October 1979 data set 

 is excluded. This is probably a large underesti- 

 mate because the 1979 year class was so much 

 weaker than the 1978 year class. Realistic esti- 

 mates could not be calculated in most instances 

 comparing the 1978 and 1979 year classes be- 

 cause N, exceeded iV , which largely reflects the 

 much greater strength of the 1978 year class 

 (Fig. 10). Time-specific mortality estimates 

 made from Chittenden and McEachran's (1976) 



536 



