McBnde and McKown: Consequences of dispersal to temperate estuaries for Caranx hippos 



531 



some individuals had reached approximately 20 cm FL by 

 October and November, suggesting that absolute growth 

 could exceed 1 mm/d. The prolonged appearance of small 

 fish in southern estuaries, in particular, and the lack of 

 age-specific data precluded more detailed comparisons of 

 growth rates. 



Estuarine residency 



In the Hudson River system, C. hippos occurred annually 

 but its density varied by an order of magnitude among 

 years (Table 1). This species occurred as far upstream as 

 rkm 102 (i.e. lower part of stratum 7) but was most abun- 

 dant in the upper portion of Haverstraw Bay (Fig. 4A). Sea- 

 sonally, C. hippos resided in the Hudson River system from 

 July to November and was most common from late July 

 to early October (Figs. 4B, 5A). The median temperature 

 for collections of C. hippos was 26°C in Haverstraw Bay 

 and 24°C in Jamaica Bay (Fig. 6). Near Haverstraw Bay, 

 where salinities were much lower than those of Jamaica 

 Bay, C. hippos was capable of invading nearly freshwater 

 regions of the Hudson River (Fig. 6). Abundance measured 

 with a 61-m net, set in a comparable manner in both bays, 

 showed that the abundance of C. hippos was much higher 

 in Jamaica Bay than in Haverstraw Bay. Abundance in 

 Great South Bay, New York, was also consistently higher 

 than in Haverstraw Bay (Fig. 5A), even though a smaller, 

 30-m net was used. 



Fork length offish collected during 1986-1993 in the Hud- 

 son River estuary ranged from 2.9 to 17.6 cm. Growth rate 

 was modeled by using data for New York and New Jersey 

 estuaries (Fig. 5B; ;)=439), and the resulting equation, 



FL= 136.6exp(-exp{-0.0315[DOy-211]}), (r2=0.75), 



predicted a peak instantaneous growth rate for July 30, 

 after which growth slowed and fish reached a mean asymp- 

 totic length of 13.7 cm. The distribution of residuals for 

 this growth model suggested that the asymptotic length 

 was biased (i.e. lowered) by the presence of small, presum- 

 ably young, juveniles in August-September; therefore we 

 consider these growth estimates to be preliminary. 



Continental shelf distributions 



A total of 657 C. hippos were collected at 134 stations 

 between 33°47'N and 41°00'N and at depths from 6 to 38 m 

 (Fig. 7A). Although present in all but four years during 

 1972-1996, C. hippos were collected only in the months 

 of July ( 1 of 134 stations=0.8'7r ), September (84.3'7, ), Octo- 

 ber ( 14.2'7( ), and November (0.8''* ) during this period. Fish 

 ranged in size from 3 to 29 cm FL, and modal size ( 17 cm I 

 was the same for fish collected both north and south of 

 36°00'N(Fig. 7B). 



Discussion 



Densities of C hippos in Haverstraw Bay, where our study 

 was focused, were generally low compared with other New 



Fork length (mm) 



Figure 2 



Monthly length frequencies of young-of-the-year Caranx 

 hippos in estuaries from North Carolina to Florida (open 

 bars) and New York or New Jersey (filled bars). Data for 

 .southern states are taken from Berry 1 1959; Table 23. 

 Addenda, p: 534-5351, and data for northern states are 

 from material archived at the New York State Museum, 

 Academy of Natural Sciences in Philadelphia, and Amer- 

 ican Museum of Natural History (listed in McBrideM. 

 n = number offish measured. 



York estuaries. We did observe considerable variation in 

 densities of C. hippos between months, years, and estuar- 

 ies, which we believe is at least partially the result of this 

 species' relatively low density and shoaling (e.g. aggre- 

 gating) behavior (Kwei, 1978). Larval dispersal processes 

 or habitat may also affect juvenile C. hippos densities 

 between temperate estuaries. This variability appears to 

 have led to contradictory reports of C. hippos abundance 

 at temperate latitudes: some investigators concluded that 

 this species is "rare" or "scarce" (e.g. Bean, 1900; Greeley, 

 1937), whereas others have considered it to be "common" 

 or "abundant" ( DeKay, 1842; Smith, 1985). 

 . Our review of seining data across a latitudinal range 

 from 27°N to 41°N demonstrated that densities of C hippos 

 propagules transported to subtropical estuaries were not 



