I

June -/August

June -October

June - June -August August

May -i October

¡Spring & Summer

Marine Waters

Late Non- Female Male Juvenile Spawning Spawners Spawners "Atlantic Ocean Adults

Larval

Early Juvenile

Figure 13.1. Seasonal movements of Atlantic sturgeon in the Hudson River by life stage, gender, and reproductive condition (developed by M. Bain, Cornell University).

VanEenennaametal. (1996) couldcollectspawning specimens only at two historically important up-river fishing sites (Hyde Park and Catskill) and they argued thatreproductionnear the saltfrontwas unlikely given the physiological requirements of the species' early life stages. Sonic tagging by Bain et al. (1998; 2000) showed that spawning occurs over a period of weeks at several sites between km 113 and 184. They also found that many subadults and adults (some believed to be post-spawners) aggregate midsummer through fall just upstream of the Bear Mountain Bridge, part of a complex seasonal phenology that varies by life stage, gender, and reproductive condition (Fig. 13.1).

Haley (1998) developed a gastric lavage technique to study sturgeon food habits without sacrificing them. When applied to twenty-third Atlantic sturgeon captured in the river between June and September 1996, major prey items were found to include polychaetes, isopods, and amphipods.

In October 1994, 4,929 hatchery-reared, six-month-old Atlantic sturgeon were stocked into the Hudson near km 90. Because these fish were marked, they were distinguishable from wild specimens from the same year class. Peterson, Bain, and Haley (2000) used this difference to perform a Peterson estimate of abundance of wild age-1 Atlantic sturgeon, which generated a value of 4,313 individuals, with a 95 percent confidence interval of 1,917-10,474. This estimate was about 80 percent lower than those obtained by Dovel and Berggren (1983) from the mid 1970s.

Atlantic sturgeon had been fished in the Hudson since prehistoric times (Brumbach, 1986) and in the late nineteenth and early twentieth centuries, they were skinned, sliced, and then sold under the name "Albany beef" (Greeley, 1937). However, the Hudson River's population appeared not to experience the magnitude of overfishing that occurred in the Delaware River during the caviar craze of the late nineteenth century (for example, Secor and Waldman, 1999).

But there was no evidence of a significant population change in Atlantic sturgeon in the

Hudson until the mid 1980s when numbers of juveniles observed in commercial fishing bycatch and utilities-sponsored monitoring programs decreased substantially (Bain et al., 1998). Moreover, as alternative species declined, some commercial fishermen inside and outside of the Hudson River began to focus on Atlantic sturgeon. Landings in NewJersey's coastal waters increased from 5,900 kg in 1988 to approximately 100,000 kg in 1990. And NewYork's coastal harvest rose from about 7,700 kg in 1993 to almost 16,000 kg in 1994 (Waldman, Hart, andWirgin, 1996a). Mixed-stock analysis using mitochondrial DNA (mtDNA) markers indicated that 99 percent of the sturgeon in the New York Bight fisheries were of Hudson River origin (Waldman et al., 1996a). In response to this increase and to the overall scarcity of Atlantic sturgeon in other river systems (Waldman andWirgin, 1998), a moratorium on commercial harvest in U.S. waters was enacted in 1998 that may extend for as much as 40 years. (Limburg et al., this volume).

Acipenser brevirostrum shortnose sturgeon. Short-nose sturgeon is a small, federally-endangered acipenserid which occurs in about twenty estuaries from New Brunswick to Florida (color plate 6). Many of its populations have been reduced and number only in the hundreds to thousands (ASMFC, 1998).

Shortnose sturgeon are long lived, reaching almost seventy years. Hudson River males don't mature until age three and females not until age six, or later (Bain, 1997). Early growth is rapid: yearlings may exceed 30 cm by their second summer (Dovel, Pekovitch, and Berggren, 1992). However, the rate of growth declines substantially as subadults and adults (Bain, 1997). Indeed, it may halt, based on the results of Bain et al. (1998) who recaptured nineteen individuals between 1993 and 1995 that were tagged by Dovel et al. (1992) in 1979 and 1980. Of these recaptures, seventeen were adults when tagged and on average had grown only 86 mm and some showed no increase in length.

There is limited information on the feeding habits of shortnose sturgeon in the Hudson River. Carlson and Simpson (1987) examined forty-two young-of-the-year and ten yearling specimens killed on the cooling water intake screens at a power plant at km 228. Predominant prey varied seasonally but included midge larvae, amphipods, and isopods. Although Carlson and Simpson found little reliance on molluscs for young age classes, Bain et al. (1998) determined that shortnose sturgeon also eat the non-native zebra mussel Dreis-sena polymorpha. Haley (1998) found principal food items to include soft-bodied invertebrates (e.g., amphipods, chironomids, and isopods) and shelled organisms such as snails and zebra mussels.

Data syntheses (Geoghegan, Mattson, and Keppel, 1992) from three utilities-sponsored monitoring programs between 1983 and 1988 (excluding winter) showed that shortnose sturgeon from 10 to 120 cm TL were most common between Kingston and Croton Point. Dovel et al. (1992) and Bain (1997) demonstrated a complex seasonal pattern of movements for reproductively-mature adults that is dependent on whether individuals are spawning in a given year. The frequency of spawning for the St. John River population was estimated at between every third to fifth year for females, and every second year for males (Dadswell, 1979). However, these values may not apply to the Hudson River populationinasmuchasDoveletal. (1992)reported the occurrence ofsome tagged shortnose sturgeon on the spawning grounds in successive years.

Prespawning adults overwinter aggregated at a bend in the river near Esopus Meadows (Bain, 1997). They then reaggregate and spawn far up river from about km 190 to immediately downstream of the Federal Dam. Later, postspawners disperse widely, mixing with nonspawners, primarily in channels of the fresh- and brackish-water reaches. As winter approaches adults again segregate according to whether they will spawn the following year, with nonspawners wintering farther down river than spawners, between km 54 and 61 (Bain, 1997).

Recent work demonstrated stock differences between shortnose sturgeon from the Hudson and other populations. These included some mor-phometric features in comparison with specimens from the Kennebec and Androscoggin Rivers (Walsh et al., 2001) and mtDNA control region sequences in comparison with populations across the species' range (Grunwald et al., 2002).

There was no concerted commercial fishery for shortnose sturgeon in the Hudson River (where they were also known as "roundnosers"; Dovel et al.,

1992), although some individuals were harvested (Boyle, 1969; Dovel et al., 1992), particularly in the central river, where Greeley (1937) stated it was of "some commercial importance." Bain (personal communication, Cornell University) estimated the abundance of adult shortnose sturgeon in the Hudson during the late 1990s at about 61,000, a more than four-fold increase since the early 1980s (based on Dovel et al., 1992). Bain believes the Hudson's population may exceed the total of the approximately twenty other populations of the species. It may be that the shortnose sturgeon population of the Hudson River is recovered and is a candidate for de-listing under the 1973 U.S. Endangered Species Act. If de-listing should occur, it would be the first instance for a population segment of a fish species.

anguillidae

Anguilla rostrata American eel. This catadromous species spawns in the Sargasso Sea; afterwards its developing larvae (leptocephali) are advected toward North America (ASMFC, 2000). Leptocephali metamorphose into glass eels (transparent juveniles) in coastal habitats and are subsequently transported into estuaries. They become pig-mented as they move farther inland. They are then called "yellow" eels, residing for upwards of 30 years before metamorphosing into "silver" eels and returning to the Sargasso Sea to spawn and die. Because of their random recruitment to river systems, American eels form one genetically panmictic population.

Eels have been little studied within the Hudson River estuary. They were known to move upstream past the Federal Dam as far as Saratoga Lake (Brumbach, 1986) and into several Adirondack lakes (Greeley and Bishop, 1933). Greeley (1937) found eels in numerous streams and lakes in the drainage of the lower Hudson River.

Peak immigration of elvers into the Hudson estuary occurs from mid-March through April (Mattes, 1989). Stomach analysis by Mattes (1989) of 468 eels trawled between Manhattan and Troy revealed a generalized feeding pattern with forty-eight prey species found. Important fooditems included crustaceans, molluscs, and fish.

The commercial fishery for yellow eels in the Hudson River has been closed since 1976 because of contamination with polychlorinated biphenyls (PCBs); specimens were found to have PCB levels of 50 to 75 ppm (Blake, 1982). U.S. landings of American eel together with estimates of eels passed at dams and taken in surveys indicate a substantial decline since 1979 (ASMFC, 2000). Although a significant glass eel fishery never developed in New York, in 1995 the state has imposed a minimum length limit for harvest in marine waters of 15 cm (ASMFC, 2000).

clupeidae

Alosa sapidissima American shad. American shad is the largest North American alosine, reaching weights of 10 pounds or more. Populations occur in rivers from the St. Lawrence to the St. Johns. American shad has long been fished with nets for both its flesh and roe throughout the Hudson River and the New York bays (Fig. 13.2). Because its PCB body burdens meet federal standards, American shad is the only fin fish still legally harvestable in the Hudson by commercial fishermen.

American shad spawn in spring at water temperatures between 12° and 21°C, mostly between dusk and midnight. In the Hudson, American shad are main channel spawners (although I am aware of anecdotal reports of adult American shad occurring in the Croton and Rondout Rivers). Eggs of American shad have been collected from the Albany area to about km 83, but with highest densities in northern sections (Schmidt et al., 1988). Limburg (1996) performed a fine-scale study of growth and seaward migration of the 1990 year class of American shad in the Hudson River. She found this migration to be both size- and age-related, with upriver fish consistently smaller and younger than downriver specimens until late in the season. Otolith analysis showed that recruitment of juveniles from the 1990 year class was not spread proportionately across birth dates. Although most American shad spawning activity occurred in early to mid May, the year class was established mainly by individuals hatched in June; a discrepancy attributed to probable weather-related effects on food availability to larvae.

Juvenile American shad in the Hudson River prey primarily on epiphytic rather than benthic invertebrates, particularly chironomid larvae and formi-cid hymenopterans (Grabe, 1996). They often feed at the surface during afternoon and evening, consuming terrestrial insects.

Figure 13.2. Shad fishing in New York Bay using stake nets, circa 1867 (courtesy of the New York Public Library).

American shad range coastally from Labrador to Florida. They are long-distance but mainly inshore migrators; during an average life span of five years at sea, an individual may travel 20,000 km (Dadswell et al., 1987). Dadswell et al. (1987) analyzed 1,837 returns from 11,579 American shad tagged in the Hudson River and New York Bight. Although more than 85 percent of recaptures occurred in the Hudson, recaptures from outside the river rangedfrom just south of Cape Hatteras to the upper reaches of the Bay of Fundy and to Halifax on Nova Scotia's east coast.

Leggett and Carscadden (1978) demonstrated that American shad show latitudinal variation in reproductive characteristics, including whether all individuals in a population died after first spawning (semelparity) or if repeat spawning occurred (iteroparity). Repeat spawning of American shad in the Hudson River appears to occur at a rate of more than 40 percent (ASMFC, 1998).

American shad population size in the Hudson declined from about 2.3 million in 1980 to 404,000 in 1996 (ASMFC, 1998). Inriver commercial landings fell from 2.6 million pounds in 1980 to less than 250,000 pounds in 1996. Although inriver fishing mortality rates fell over that period, coastal fishing mortality rates ("intercept fishery") increased, and total fishing mortality rates have remained stable and independent of stock decline from 1980 to 1996. Moreover, recruitment has been high and relatively stable, yet adult stock size continued to decline to historic low levels. It may be that abiotic or biotic factors (possibly striped bass predation) rather than overfishing have caused the recent decrease in the Hudson's shad stock (ASMFC, 1998).

A number of genetic studies have shown significant differences among American shad stocks, including the Hudson River's (for example, Nolan, Grossfield, and Wirgin, 1991; Waldman et al., 1996b). Brown (1996) used genetic data to estimate the relative contributions of American shad stocks to coastal intercept fisheries. In New York Bight waters proximal to the Hudson River, the Hudson River stock was estimated to compose 21 percent of the 1994 catch and 22 percent of the 1995 catch.

Within-population variation of American shad has received little attention. Interesting differences in external characteristics were observed by commercial fishermen early in the 1900s (Greeley, 1937). Greeley reported that two types were seen: a form called the "blueback" and the much less common "yellowback." These types differed subtly in ground color, pigmentation pattern, and snout dimensions. Yellowbacks already were rare at the time of Greeley's survey.

Alosa mediocris hickory shad. Hickory shad is a medium-sized alosine (max length ~60 cm) that ranges from the Bay of Fundy to the St. Johns River, Florida. Its life historyremains largely undescribed. Adults often occur in inshore waters in the New York Bight and Long Island Sound, sometimes in high abundance, particularly near river mouths. Unlike the other North American shads, hickory shad is highly piscivorous, but it also consumes invertebrates.

Little is known of hickory shad within the Hudson River. Beebe and Savidge (1988) noted that it has been recorded from the Yonkers to Indian Point regions of the Hudson River. Smith (1985) reported that anglers occasionally take them in the Hudson. Bean (1903) stated that it was commonly caught at Gravesend Bay (at the mouth of the Hudson estuary) between September and November. Bean (1903) reported that the ascent of hickory shad precedes that of American shad in rivers where they co-occur. But although hickory shad have been caught in the Hudson River, there does not seem to be any conclusive evidence that they reproduce there.

I believe hickory shad abundance in the coastal waters of the New York Bight and Long Island Sound has increased substantially in the 1990s based on anecdotal reports in the angling literature and my personal observations, a trend that also has been noted in Connecticut and the Chesapeake Bay (ASMFC, 1999). If so, it is not clear whether this increase represents increased reproduction in the Hudson River or, more likely, immigration from other sources.

Alosa pseudoharengus alewife. Alewife is a small alosine that occurs in rivers and coastal waters from Newfoundland to South Carolina. It differs from the other "river herring," the blueback, in several characteristics, most notably its greater body depth and pale, instead of dark, peritoneum.

Alewives co-occur with blueback herring in many drainages between New Brunswick and South Carolina. A maximum total length of 38 cm has been attributed to both species (Loesch, 1987). Unlike blueback herring, the alewife landlocks readily, often in dwarf form. These populations are referred to as "sawbellies" and are widely used as bait for salmonids. Landlocked alewives exist in numerous reservoirs in the Hudson River drainage, including in the New York City Reservoir Supply System (e.g., Kensico Reservoir; Greeley, 1937).

Alewives begin spawning at between 5 and 10°C and prefer lakes and ponds or slow-flowing sections of streams. Fecundity is variable but is related to size and age, reaching as many as 467,000 eggs. First spawning of both river herrings occurs from ages three to six, but is dominated by age-four fish (Loesch, 1987). Fecundity appears to peak at age-six and declines afterward. Alewives return to sea after spawning but little is known of their marine migrations.

There has been little research on either river herring in the Hudson drainage, but some system-specific information has emerged. Schmidt, Klauda, and Bartels (1988) analyzed data from early life history surveys performed between 1976 and 1979. Juveniles appeared in seine collections from late June to early July, about a month later than American shad young. In the upper and middle estuary, seine catches declined in early July while they increased in offshore bottom trawls, signifying a movement from the shallows and mouths of streams to deeper waters. Summer catches showed amovement downriver, with numbers declining after August, presumably because of emigration. Juveniles were found downriver through mid December, but some may remain in the estuary all winter. Indeed, elemental composition analysis of small but spawning adult alewives from Coxsackie Creek indicated that some may remain within the Hudson River estuary for their entire life cycle (R. Schmidt, personal communication). Hudson juveniles feed primarily on chironomids and amphipods, with no diel differences in feeding activity (Grabe, 1996).

Lake and Schmidt (1997) surveyed fish usage during spring of a small Hudson River tributary, Quassaic Creek, located near Newburgh. In 1996, the first alewife specimens were seen on April 3, with only males caught until April 14. Peak spawning appeared to occur on April 19 at a water temperature of 9.4°C, with a secondary peak in mid-May, but spawning continued until early

June. The total adult spawning run in Quassaic Creek for 1996 was estimated at about 5,600 individuals. Lake and Schmidt (1998) investigated the alewife spawning run in Quassaic Creek more intensively in 1997. They found fecundity to range from 15,000 to 135,000 eggs andfemale totallengths of between 232 to 318 mm. Spawning activity was similarly bimodal to 1996 and the total number of eggs deposited in this single system was estimated at 162 million. In more extensive surveys in 1998 and 1999, Schmidt and Lake (2000) documented river herring (mainly alewife) runs in twenty-eight of the thirty-eight Hudson River tributaries they sampled. Although river herring are rarely sought for food from the Hudson River, they are harvested for bait (Vargo, 1995).

Alosa aestivalis blueback herring. Blueback herring range from Nova Scotia to Florida. Where they co-occur in spawning rivers with alewives, bluebacks achieve some niche separation by running later in the season (peak spawning lagged by two to three weeks) and by choosing relatively swiftly running sites to spawn in (Loesch, 1987).

Because they spawnlater, recruitment of juvenile bluebackherringin the Hudsonsystem occurs after that of American shad and alewife (Schmidt et al., 1988). They also found that juvenile blueback herring numbers dwarfed other alosines in summer. Blueback herring are the most planktivorous of juvenile alosines and are most active during daylight (Schmidt etal.,1988;Grabe, 1996). Primary prey includes copepods, chironomids, and cladocerans.

In recent decades, blueback herring have become more abundant in the Hudson River system, but it appears that this increase is due mainly to an expansion of range. Daniels (1995) noted that Greeley (1937) found blueback herring at only 4 of 112 sites sampled; whereas in Daniels' own 1990 survey and in monitoring conducted in the 1970s and 1980s, blueback herring dominated summer catches at inshore sites. Surveys have shown that blueback herring are either absent or uncommon in most of the tributaries examined that enter the Hudson below the dam at Troy. In their 1998 and 1999 surveys, Schmidt and Lake (2000) found adult blueback herring in only 7 of the 28 tributaries in which river herring were encountered. Bluebacks were seen in the four northernmost tributaries, which ranged to the Poesten Kill (km 241). The reason for the apparent increase in the population size of bluebacks in the Hudson system is that large numbers of spawners move up the mainstem tidal Hudson River to the Federal Dam and then pass through the hydroelectric turbines or the navigationlock. MacNeill (1998) believes they spawnnear Rome, New York, at the highest elevation of the Mohawk corridor.

osmeridae

Osmerus mordax rainbow smelt. Rainbow smelt is a slender, elongate fish that ranges from the Hudson River northward to the Arctic. Maximum size is about 30 cm. Little is known of their marine movements. Spawning occurs after individuals have passed two winters at sea. Rainbow smelt reproduce in early spring; egg fecundity ranges between 103-104 and eggs hatch in ten to thirty days. In salt waters smelt eat crustaceans, nereid worms, and fish. In fresh water they consume shrimp, amphipods, oligochaetes, and insect larvae.

In his 1936 survey, Greeley (1937) encountered young smelt at seven stations between Rhinecliff and Port Ewen. Rainbow smelt once ran into many Hudson River tributaries in early spring when the mainstem reached about 4°C (Boyle, 1969). These runs were actively fished in locations such as the Rondout River and Saugerties Creek (Greeley, 1937). They have also been observed in the Croton River (Boyle, 1969; Rose, 1993), Wappingers Creek, Columbiaville Creek, Roeliff-Jansen Kill, Black Creek, Crum Elbow Creek, Fishkill Creek, and Esopus Creek (Rose, 1993). The last significant tributary runs appear to have occurred in 1979 (Rose, 1993) and recreational net fisheries have since disappeared.

All indications are that rainbow smelt in the Hudson declined during the late 1900s and became extinct shortly before 2000. Between 1974 and 1980, annual catches in the Hudson River utilities near-shore survey ranged from 108 to 1,880 individuals (Daniels and Lawrence, 1991). But in 1981, the survey caught only forty-six, and just a single specimen was sampled between 1982 and 1989. In a 1988 survey of Hudson River tributaries, Schmidt and Limburg (1989) caught several smelt larvae in Catskill Creek but not in fifteen others (located between km 42 and 220). Schmidt and Lake (2000)

caught a single smelt larva in the Moordener Kill in 1998.

Rose (1993) analyzed the Hudson River utilities survey data to describe spawning locations and abundance trends in the mainstem between 1974 and 1990. Although anecdotal accounts indicated that Hudson River tributary runs had dwindled to almost undetectable levels by the early 1980s, the Long River Survey did not show a decline in the Hudson River over that period, that is, high average densities of post-yolk sac larvae were seen in 1986, 1988, and 1990. However, Rose found that the center of abundance of these larvae had shifted since the 1970s from the lower river to the Catskill to Albany regions. Since 1995, rainbow smelt essentially disappeared from the Hudson River ichthy-oplankton (Daniels et al., in press). Despite comparable levels of effort with earlier periods during which 103-104 post-yolk sac larvae were collected annually in the Long River Survey, between 1996 and 2000 only four individuals (during 1998) were found.

gadidae

Microgadus tomcod Atlantic tomcod. The Hudson River is the southern population limit of this small, boreal gadoid. In the Hudson, spawning occurs in early- to mid-winter, primarily between the Tappan Zee and Poughkeepsie reaches, and centered near West Point (Klauda, Moos, and Schmidt, 1988). Egg incubationlasts sixty-one to seventy days (Dew and Hecht, 1994). Larvae become distributed throughout the Hudson, but at highest densities downriver. Between April and November juveniles are most abundant in the Tappan Zee and West Point regions and south to Manhattan (Dew and Hecht, 1994). However, summer distributions appear to advance upriver with the salt front, with high monitoring catches being made in the Indian Point region (Klauda et al., 1988).

Juvenile Atlantic tomcod in the Hudson River were found to feed mainly on invertebrates, par-ticularlycalanoidcopepods (McLaren etal., 1988). Adults displayed greater piscivory, but invertebrates still predominated, especially Gammarus, Neomysis, Crangon, and chironomid larvae. McLaren et al. (1988) identified three phases of first-year growth for Hudson River tomcod: a summer phase of little or no growth (because of high temperature stress) separated rapid growth phases during spring and fall. Dew and Hecht (1994) estimated a decrease in weight gain from 2.9 percent d to 1.3 percent d as temperatures rose above 13°C in late May.

The Hudson River population has an unusual maturation schedule and age-structure. Both sexes begin to mature reproductively at nine months and are capable of spawning at eleven months. In early winter, these newly-mature individuals move up-river along with older tomcod to spawn. However, unlike in northern estuaries where they may live for several more years, only three age classes are known from the Hudson River, with individuals older than age-1 being rare. McLaren et al. (1988) found that among spawners collected during the winters of 1975-76 through 1979-80, age-1 fish comprised 92-99 percent of the annual totals, with age-2 individuals making up most of the remainder. Age-3 fish were extremely rare (<0.1 percent of total).

Because the tomcod population is composed almost completely of one age class, its annual abundance is unusually responsive to environmental conditions that influence recruitment and subsequent survival; thus, its abundances fluctuate widely. For example, winter spawning population estimates were 12.5 million in 1982-83, and 6.7 million, 2.1 million, 3.5 million, and 5.9 million in succeeding years (Mattson, Geoghegan, and Dunning, 1992).

Most Hudson River tomcod remain within the estuary for life. However, a few individuals tagged in the Hudson in the late 1970s were recaptured in lower New York Bay, the East River, and western Long Island Sound (Klauda et al., 1988), suggesting broader limits for this population.

Because it is bottom dwelling, dependent on benthic prey, has a lipid-rich liver, and tends to remain within its natal estuary, tomcod from highly polluted systems tend to accumulate unusually high concentrations of contaminants. In the Hudson River, these include PCBs and other chlorinated hydrocarbons (Courtenay et al., 1999). Tomcod from the Hudson exhibit biochemical responses and molecular damage not observed in those from cleaner estuaries (Wirgin et al., 1994) and a remarkably high prevalence of liver tumors (Deyet al., 1993).

moronidae

Moronesaxatilis striped bass. This species appears to be the most popular sport fish in the Hudson River. It also was harvested commercially in the river until the fishery was closed in 1976 because of PCB contamination. The striped bass population of the Hudson River is the northernmost of the three main migratory stocks (together with Chesapeake Bay and Delaware River) that support the fishery along the northeast U.S. coastline.

StripedbassspawnintheHudsonfrom early May through early June. Between 1974 and 1979, maximum egg densities occurred at water temperatures of 12°-22°C and a spatial peak occurred between km 54 and 98 each year, although early life stages were encountered in all regions between Albany and Yonkers (Boreman and Klauda, 1988). However, the downriver limit of egg deposition is usually associated with the position of the salt front. Sample densities of juveniles peaked between km 38 and km 74. But juveniles may occur anywhere within the estuary and even into western Long Island Sound (LoBue and McKown, 1998). A high proportion of juveniles and yearlings winter in the lower river, particularly off Manhattan.

Several studies have described early life history processes of striped bass in the Hudson. Pace et al. (1993) examined relationships among abundances of life stages and temperature and flow conditions between 1974 and 1990. They concluded that temperature and river flow were not related to interan-nual variation in recruitment and, that there were no statistically significant relationships between the abundances of yolk-sac larvae, post yolk-sac larvae, and juveniles. They believe the absence of relationships among early life stages means that differences in mortality are less important than variability in egg numbers in determining larval abundance.

Limburg et al. (1997) tracked the abundance and food consumption of larval striped bass during 1994. The cladoceran Bosmina longirostris and large copepodites and adult copepods composed 97.4 percent of the diet. They found that larval cohorts extant before the zooplankton bloom had the least available food but also the lowest respiration costs. Postbloom cohorts had both high consumption rates and respiration rates due to increased temperatures. Cohorts coincident with the bloom had moderately high consumption rates and lower metabolic costs relative to late cohorts. The investigators concluded that larval cohorts coincident with the bloom possess an energetic advantage relative to early cohorts butnotrelative to late cohorts.

Hurst and Conover (1998) explored the possible effects of winter mortality stemming from bioenergetic stresses on striped bass recruitment in the Hudson. They found that age-1 abundance was negatively correlated with the severity of winter; however, numbers of age-0 fish were not correlated with abundance at age 1 - leading to the conclusion that winter mortality greatly modified year class strength. A progressive increase in the mean length of young-of-the-year fish, coupled with a decrease in the coefficient of variation in length, occurred during some winters. This, combined with laboratory experiments that showed that growth in length requires temperatures in excess of 10°C suggest that these changes result from a bias toward greater mortality of smaller specimens. Dunning etal. (1997) provided evidence that subadult striped bass in the lower Hudson River feed during winter at temperatures wellbelow10°C (although consumption was not compared with growth).

As part of the Hudson River Cooling Tower Settlement Agreement, the electric utilities were to operate a hatchery that would plant 600,000 young-of-the-year striped bass in the Hudson annually (Dunning etal., 1992). Hatchery-reared individuals were marked with coded magnetic-wire tags before stocking. In the course of trawl sampling for these individuals to estimate their contributions to the Hudson River stock, many wild specimens were caught and these were tagged with external tags to learn more about their movements and to monitor year class abundance. The goal of 600,000 hatchery-reared specimens per year never was met because of disease problems and the resultant contributions to thestockofthe 1.3 millionfish stocked from 1983 through 1987 were small (e.g., 0.1 percent to 1984cohort,3percentto 1985 cohort). Moreover, the hatchery-reared striped bass showed more limited movements inside the Hudson River than did wild fish (Wells et al., 1991). For these reasons, the hatchery ceased operations in 1995.

Striped bass stockings in the Hudson River appear to have made little contribution to coastal waters. Waldman and Vecchio (1996) examined hatchery-reared specimens from Hudson River and Chesapeake Bay sources among more than 1,500 striped bass caught in haul seines in eastern Long Island in 1991 and 1992. Hatchery-reared specimens comprised 3.5 percent of the total catch in 1991 and 2.5 percent in 1992. Although only about twice as many marked individuals were stocked in Chesapeake Bay as in the Hudson River, the recapture ratio of Chesapeake Bay to Hudson specimens was 62:1 in 1991 and 37:2 in 1992 - a difference attributed to differences in relative survival or vagility. Hatchery-reared specimens from both stocks also showed a pronounced increase in the incidence of broken striping patterns, a phenomenon that remains unexplained (Waldman and Vecchio, 1996).

The increase in the abundance of the Hudson River striped bass stock that occurred since the late 1970s appears to have affectedits coastal migratory range. Tagging studies performed between 1948 and 1952 concluded that the Hudson River stock limited its movements outside the river to western Long Island (Raney, Woolcott, and Mehring, 1954). McLaren et al. (1981), based on tagging done in 1976 and 1977 reported broader coastal movements, but only as far northward as Newbury-port, Massachusetts, and an absence of a relationship between fish length and distance from the river. But tagging conducted in the Hudson River between 1984 and 1988 showed an expansion of range, with recoveries made northward as far as the Annapolis River, Nova Scotia (Waldman et al., 1990a). They also found a strong relationship between fish length and distance from the river, and interpreted the absence of such a relationship in the results of McLaren et al. (1981) as an artifact of size-dependent tag retention (Waldman, Dunning, andMattson, 1990b).

Although the Hudson's striped bass stock makes coastal migrations, it was suspected that not all individuals participated and that some were residential in the freshwaters of the river (Raney et al., 1954) or did not move past the lower estuary (Clark, 1968). Secor (1999) used elemental analysis of otoliths to reconstruct the salinity history of individuals in relation to age. His work supported the notion of "contingents," finding evidence for resident Hudson River, New York Harbor, western Long Island Sound, and coastal Atlantic components to the total Hudson River stock. Individuals of the Hudsonresident contingent (primarily males) showed elevated PCB levels consistent with greater lifetime exposure to PCB sources (Zlokovitz and Secor, 1999).

The relative contributions of the migratory striped bass stocks to coastal waters vary with their spawning success and subsequent conservation. Berggren and Lieberman (1978) analyzed striped bass caught in coastal waters in 1975 using morphological stock identification approaches and estimated that the Hudson population comprised only 7 percent of the stockmixture, with 90 percent originating from the Chesapeake Bay. VanWinkle, Kumar, and Vaughan (1988) reanalyzed these data on a year-class basis and found that the contribution of the Hudson stock approached 50 percent for some cohorts. Wirgin et al. (1993) applied a mtDNA-based approach to this question with a collection of specimens (N = 112) from eastern Long Island made in 1989, and estimated a Hudson component of 73 percent. A similar study (Wirgin et al., 1997) performed on a larger sample (N = 362) collected in 1991 and including nuclear DNA markers also showed a Hudson River stock contribution of about 52 percent. These later estimates of large contributions of the Hudson River population to coastal waters reflect both its increased abundance and the decline of the Chesapeake stock over this period.

Was this article helpful?

0 0

Post a comment