Powles and Warlen Recruitment season, size, and age of Angutlla rostrata entering an estuary near Beaufort 



301 



zone, and from the "elver" check (transition 

 mark) to the edge of the otolith (see Cieri 

 and McCleave, 2000). Total ages were esti- 

 mated as the age from the hatching ring to 

 the elver mark. Glass-eel gi'owth zone ages 

 (increments from the outer boundary of met- 

 amorphic zone to elver check) were read as 

 in Wang and Tzeng ( 1998). 



Age studies 



The mean ages in the glass-eel growth zones 

 of eels taken at Pivers Island in February 

 were compared with mean ages in glass-eel 

 growth zones of eels taken at an upper estu- 

 ary site, the Black Creek Mill Pond entrance 

 (also in February). The difference in mean 

 age presumably reflected the time required 

 to travel the distance upriver to the collec- 

 tion site. 



The total ages and the ages found in the 

 glass-eel growth zone were compared with 

 those for small and larger glass eels arriving during the 

 recruitment season. Glass eels were divided into length 

 groups (from 46.1-48.0 TL to 58.1-60.0) and correspond- 

 ing age frequencies were assigned to each length group. 

 A linear regression was then fitted to the data to assess 

 whether or not smaller fish were older or younger than the 

 longer members of their recruitment class. 



Statistical analysis 



Seasonal trends in the weekly length distributions were 

 examined by linear regression (SAS Institute, 1996) for 

 each sampling (recruitment) season, and for the whole 

 10-year period ( 1985-86 to 1994-95). 



The differences in otolith increment widths in the early 

 (proximal) glass-eel growth zone (increments 10-15) and 

 later ( mid-distal ) glass-eel growth zone ( increments 30-35 ) 

 from the New Brunswick and North Carolina samples 

 were examined by analysis of variance (ANOVA). Age dis- 

 tributions of glass eels captured by month were compared 

 by using the F-test. Statistical significance was accepted 

 at P < 0.05. 



again (F.j=4.88; slope 2.04, 7-2=0.78, P=0.113), approaching 

 the previous mean level. 



Seasonal recruitment 



Although glass eel recruitment sometimes occurred over 

 a 5 1/2 month period (mid-November to early May), most 

 recruitment of young eels to the estuary occurred from 

 December through April (Fig. 2). Although there was some 

 variation among years within the recruitment period, the 

 largest catches usually occurred in February and March. 

 Peak catches varied from early in the season (1985-86) to 

 late in the season (1987-88). In 1993-94, densities were 

 highest in mid-January and mid-March. In (1990-91), the 

 year with the lowest catches, almost all of the catch occurred 

 in December and January, with virtually none afterwards. 

 Wlien seasonal changes in length of glass eels among re- 

 cruitment years were analyzed (Table 1), the slope of the 

 regression for all 10 years (Fig. 3) was positive, but not sig- 

 nificant (Fj=19.50, P=0.001). Four seasons had negative 

 regression slopes and slopes for the other years were posi- 

 tive, three of which were significant. 



Results 



Annual recruitment 



The total density of glass eels in the Beaufort estuary 

 varied considerably between 1985 and 1995, but there was 

 no significant trend (Fj,= 1.605, adj r'=0.123, slope=-0.031, 

 P=0.246) among the 10 consecutive recruitment years in 

 the Beaufort estuary (Fig. 1). Densities were highest in 

 1988-89 and 1993-94 ( 13.5-14.0 eels/100 m') and lowest 

 in 1990-91 (1.5 eels/100 m^i, a ninefold difference. Apart 

 from the low of 1990-91, all other years had total densi- 

 ties of more than seven eels/100 m*. After the 1990-91 low, 

 the total annual density of elvers at Beaufort increased 



Age studies 



The mean age observed in the glass-eel growth zones 

 of eels collected in February from the Pivers Island and 

 Black Creek sites differed significantly, by eight days (F^= 

 6.65 l,P=0.003i. The Pivers Island eels averaged 36.2 d (52"9 

 ±0.9 mm TL) for their as yet uncompleted glass-eel phase, 

 and the Black Creek elvers averaged 44.6 d (55.9 ±0.9 mm 

 TL). The distance between the two sites was about 9.5 km. 



In respect to otolith growth, the mean widths of the first 

 10-15 daily growth increments in the glass-eel growth 

 zone (Table 2) were not significantly different (F2=1.363, 

 P=0.185) between the northern eels (New Brunswick) and 

 the southern eels (North Carolina). However, both sam- 

 ples of otoliths from North Carolina indicated that mean 



