FISHERY BULLETIN: VOL. 76, NO. 2 



from the Hudson River is the major contributor to 

 the coastal fishery from New Jersey to Mas- 

 sachusetts because the number of striped bass 

 tagged in Chesapeake Bay and recaptured outside 

 the Bay was too low to indicate a large contribu- 

 tion of Chesapeake stock to that fishery. 



Because of the controversy of which stock pre- 

 dominates, we conducted a study to obtain quan- 

 titative estimates of relative percentage of the 

 major stocks in the coastal fishery. A previous 

 study (Grove et al. 1976) demonstrated the feasi- 

 bility of using discriminant analysis on mor- 

 phological characters (counts and morphometric 

 ratios) to distinguish among Hudson, Chesapeake, 

 and Roanoke spawning stocks of striped bass. That 

 study showed that adequate segregation of spawn- 

 ing stocks within the Chesapeake Bay system was 

 not possible. Quantitative estimates of stock com- 

 position based on morphological characters and 

 discriminant analysis have been obtained for 

 sockeye salmon (Fukuhara et al. 1962; Anas and 

 Murai 1969), pink salmon (Amos et al. 1963), and 

 Atlantic herring (Messieh 1975). The present 

 study establishes discriminant functions based on 

 collections of spawning-stock specimens to classify 

 striped bass collected in the Atlantic coastal 

 fishery from southern Maine to Cape Hatteras. 

 The percentage of specimens collected that were 

 classified into each stock was used to estimate 

 the relative contribution of that stock to the 

 fishery. 



METHODS AND MATERIALS 



Collection of Spawning-Stock Specimens 



During the spawning season of 1975, mature 

 striped bass were collected from the natal rivers of 

 major stocks along the Atlantic coast. These fish 

 were assumed to have originated from the rivers 

 (i.e., that striped bass, like salmon and other 

 anadromous fishes, home to their natal stream to 

 spawn). This assumption was supported by tag- 

 ging studies in which striped bass tagged on 

 spawTiing grounds were recaptured on the same 

 spawning grounds in successive years (Mansueti 

 1961; Nichols and Miller 1967). Collections were 

 composed of 232 mature striped bass from the 

 Chesapeake Bay tributaries (70 from the Rap- 

 pahannock River, 53 from the Potomac River, 52 

 from the Choptank River, and 57 from the Elk 

 River and Chesapeake and Delaware Canal), 168 

 from the Hudson River, and 99 from the Roanoke 



River. Only 19 sexually ripe striped bass were 

 collected from the Delaware River above the en- 

 trance to the Chesapeake and Delaware Canal, 

 which confirms findings by Chittenden ( 1971) that 

 spawning in the Delaware River is not substan- 

 tial. Therefore specimens from the Delaware 

 River were omitted from subsequent analyses. 

 Collections were made primarily during April in 

 the Chesapeake Bay tributaries, Delaware and 

 Roanoke Rivers, and during May in the Hudson 

 River. Most specimens were obtained fresh from 

 commercial fishermen using pound nets, haul 

 seines, and gill nets. Some were netted by study 

 personnel. 



To assure an adequate representation of the 

 sexes and multiple year classes in spawning-stock 

 collections, sampling was designed to obtain 

 nearly equal numbers of male and female striped 

 bass and a minimum of 10 individuals in each of 

 the following length categories: ^399, 400-549, 

 550-699, 700-849, and &850 mm. Discriminant 

 functions based on male and female specimens 

 from multiple year classes are needed to analyze 

 an oceanic population which consists of a different 

 sex ratio and broader age structure than that of 

 the spawning stocks. 



Processing of Spawning-Stock Specimens 



Scale samples, counts, measurements, sex, and 

 state of maturity were obtained from each speci- 

 men while in fresh condition. Scale samples from 

 above the lateral line between the first and second 

 dorsal fins were pressed on acetate cards. Ages 

 were determined by the scale annulus method 

 (Mansueti 1961). Measurements from the focus to 

 the first and second annuli were made on mag- 

 nified scale images. The following counts and 

 measurements were taken: number of lateral line 

 scales, left pectoral rays, right pectoral rays, sec- 

 ond dorsal rays, anal rays, upper-arm gill rakers, 

 fork length, snout length, head length, and inter- 

 nostril width. Methods used were those discussed 

 by Hubbs and Lagler (1958) and Grove et al. 

 (1976). 



Counts, measurements, and age determinations 

 were replicated by a second observer and a set of 

 tolerances was established to reduce observation 

 error. When differences between replicated obser- 

 vations exceeded tolerances, the observations 

 were retaken. Means of the replicated counts and 

 means of ratios of the replicated measurements 

 were used in subsequent analysis. 



336 



