1. Juvenile American shad collections, from 10 Atlantic coast rivers, examined in meristic 2. Frequencies of meristic counts from juvenile American shad in samples from the Con- necticut River, Conn., and Mass 3. Frequencies of meristic counts from juvenile American shad in samples from the 4. Analysis of variance on meristic counts for differences between locations and between years within rivers, and differences between rivers, for samples of juvenile American shad from the Connecticut and Hudson Rivers. 5. Frequencies of meristic counts from juvenile American shad in samples from the Analysis of variance on meristic counts for differences between locations and between years within rivers, and differences between rivers, for samples of juvenile American 14. Analysis of variance on meristic counts for differences between locations and between years within rivers, and differences between rivers, for samples of juvenile American 8 Comparative Study of Juvenile American Shad By PAUL R. NICHOLS, Fishery Biologist (Research) Bureau of Commercial Fisheries Biological Laboratory ABSTRACT Forty-five juvenile American shad, Alosa sapidissima (Wilson), collections, from 10 major shad producing rivers along the Atlantic coast of North America, were examined to see if differences in meristic counts suggested evidence of discrete river populations. Four meristic characters-- pectoral, dorsal, and anal fin rays and scutes--were used. The difference in the counts between locations and between years within rivers was small compared to that between rivers. The differences in counts between rivers indicated that discrete populations of juvenile shad occurred in rivers. INTRODUCTION In studies to discover causes of the decline in yield of American shad, Alosa sapidissima (Wilson), and to determine factors favoring recovery of the fishery, it was essential that the number and distribution of populations be known. The shad is widely distributed along the Atlantic coast from the St. Lawrence River, Canada, to the St. Johns River, Fla. This species is anadromous, spending most of its life in the ocean, but ascending coastal rivers to spawn. The spawning migrations into the rivers begin earliest in the southern part of the range (November in St. Johns River, Fla.) and are progressively later northward (June in St. John River, Canada). A female spawns about 250,000 eggs, and hatching occurs in 6 to 8 days at a water temperature of 17° C. The young shad stay in the rivers until autumn, attaining a length ranging from 75 to 145mm., and then migrate to sea. After spending from 2 to 6 years in the ocean, shad return to the rivers to spawn. Those spawning in rivers south of Cape Hatteras, N.C., normally die after spawning, while north of Cape Hatteras the proportion of fish spawning for the second time or more progressively increases from about 15 to 25 percent in Chesapeake Bay tributaries to about 45 to 55 percent in the Connecticut River. Several workers have reported evidence of different shad populations along the Atlantic coast. Differences between shad from different areas based on meristic counts (Fischler, 1959; Hildebrand and Schroeder, 1928; Hill, 1 1959; Vladykov and Wallace, 1938), growth rates (Hammer, 1942; Hildebrand and Schroeder, 1928), and fecundity (Davis, 1957; Lehman, 1953) indicated the occurrence of different populations. Recapture on the spawning ground of shad tagged in prior seasons indicated that they returned to their native streams to spawn (Hollis, 1948; Nichols, 1960). Also, the fact that the runs in the northern rivers were self-perpetuating and fluctuated independently (Talbot and Sykes, 1958) suggested different populations. The purpose of this study was to determine if discrete populations of shad could be identified on the basis of consistent differences in counts of meristic characters in juveniles from 10 rivers. As used in this report, a "population" is a group of fish having similar meristic characteristics, of which the nature of origin, genotypic and/or phenotypic, of the characteristics has not been determined. MATERIALS AND METHODS For this study, the Atlantic coast was divided into three geographical areas: North Atlantic (Maine to Virginia); Chesapeake Bay (Maryland and Virginia); and South Atlantic (North Carolina to Florida). In 10 major shad producing rivers within these areas, 45 collections of juvenile shad, from 43 to 146 mm. fork length, were taken near spawning and The homing instinct of the Chesapeake Bay shad, Alosa sapidissima (Wilson), as revealed by a study of their scales. Thesis (typewritten), 1942, University of Maryland, 45 p. nursery areas (table 1). Fifty specimens were sampled at random from each collection, covering the size range in each, for the analysis. Table 1.--Juvenile American shad collections, from 10 Atlantic coast rivers, examined in meristic studies Size range fork length Area and river Collect- Loca- Collec- Speciing gear tions tions mens Number Number Number Mm. 23 1222 North Atlantic: Connecticut... Seine Hudson........ Seine Chesapeake Bay: Susquehanna... Seine Rappahannock.. Trawl 2 South Atlantic: Neuse........ Seine Edisto... Trawl Ogeechee... St. Johns... Trawl 2123 200 200 250 Total... 45 2,250 In the North Atlantic area, collections of juvenile shad were taken from the Connecticut and Hudson Rivers. From the Connecticut, collections were taken above South Hadley Falls Dam at Holyoke, Mass., about 85 miles from the river mouth, and at Enfield, Conn., about 15 miles downstream of South Hadley Falls Dam, in the fall of 1954, 1957, and 1958. The collections taken above South Hadley Falls Dam were considered as an introduced population, because the dam blocked upstream migrating fish from this area for more than 100 years until a fish-passage facility was installed in 1952. In the Hudson, collections were taken at Piermont, N. Y., in the brackish water section about 30 miles from the river mouth; at Kingston Point, N. Y., in the freshwater section about 50 miles upstream from Piermont; and at Catskill, N. Y., about 25 miles upstream from Kingston Point, in the autumn of 1950 and 1951. Additional collections were taken at Kingston Point in 1954, 1957, and 1958. In the Chesapeake Bay area, collections of juvenile shad were taken from the Susquehanna, Rappahannock, York, and James Rivers. From the Susquehanna River, collections were taken below the Conowingo Dam in 1958; from the Rappahannock River, at Long Point in 1954 and in Batchelors Bay in 1958; from the York River, at the Pamunkey Indian Reservation in 1953, 1954, 1956, and 1958 and at the Mattaponi Indian Reservation in 1954 and 1958; and from the James River, at Walcot Wharf, Va., in 1954 and at Claremont Beach, Va., in 1954 and 1958. In the South Atlantic area, collections of juvenile shad were taken from the Neuse, Edisto, Ogeechee, and St. Johns Rivers. From the Neuse River, collections were taken at Bridgeton, N.C., in 1950, 1954, 1957, and 1958 and at Streets Ferry, N.C., in 1954; collections from the Edisto River at Crosby Landing, S.C., were available for 1938 and 1939 and were taken in 1957 and 1958; collections from the Ogeechee River were available for 1938 and 1939 from Kings Ferry, Ga., and were taken in 1957 and 1958 at the State Park, near Richmond Hill, Ga.; and collections from the St. Johns River were taken at Mandarin, Fla., in 1954, in Lake Harney in 1954, and at Palatka, Fla., in 1954, 1957, and 1958. Using a binocular microscope, counts were made of left pectoral, dorsal, and anal fin rays and total scutes. Fin ray counts included all rudiments, and the last elements in anal and dorsal fins, originating from the same base, were counted as one ray. The dorsal fin origin often required dissection to expose embedded rays. Scales occasionally had to be removed to expose enveloped scutes and anal fin rays. No attempt was made to separate scutes into anterior and posterior counts. Not a single abnormal fin or scute was encountered out of the 2,250 specimens examined. Analysis of variance (Snedecor, 1956; Steel and Torrie, 1960) was used to test if meristic count means of specimens were statistically different at the 1 percent level (indicated by two asterisks in the tables) between rivers, locations within rivers, and years within rivers. Before comparing the means, group variances were tested for homogeneity. ANALYSES OF MERISTIC COUNTS Differences and similarities in meristic counts for samples of shad from within individual rivers and between rivers are discussed in the following sections by geographical area. North Atlantic Area Meristic counts we re made of juvenile shad taken at each location in the Connecticut and Hudson Rivers (tables 2 and 3). Connecticut River. -- Mean meristic counts for the Holyoke samples (above Hadley Falls Dam) in most instances were slightly higher than those for Enfield samples (below Hadley Falls Dam). The difference in pectoral fin ray counts was significant between locations (table 4). The difference in each mean meristic count was nonsignificant between years for the Enfield samples. No analysis was made for differences between years in the meristic counts from above Hadley Falls Dam. Hudson River. -- Differences in meristic counts were not significant between locations (Piermont-Kingston-Catskill) for the years 1950 and 1951. Since Kingston Point was the |