Deree: Age and growth, dietary habits, and parasitism of Enchelyopus ambrius 



41 



magnitude of the variance was stabilized at 30 fish 

 per age class, yielding approximately 270 specimens 

 to be aged. A length-based stratification was devel- 

 oped to avoid underrepresentation of the least abun- 

 dant fish located at the extremes (Fournier, 1983). 

 After removing the 77 lower (95-129 mm TL) and 54 

 upper (250-328 mm TL) extremes, 150 rocklings be- 

 tween 130 and 249 mm TL were randomly selected, 

 then added to the extremes, yielding a total 

 subsample size of 281 fish. 



Otoliths from 281 fourbeard rocklings were etched 

 and stained, then aged with a compound microscope 

 at a magnification of 40 by using transmitted light. 

 After three complete trials, an ANOVA was run on 

 the assigned age classes. 



Assigned age classes were compared by using the 

 Fraser-Lee back-calculation formula which assumes 

 that the relationship between individual fish length 

 and otolith length is maintained proportionally 

 throughout the back-calculation (Carlander, 1981) 



It is defined as 



L =a + 0D (L -alOR), 



Where L^ = estimated total length (TL) when otolith 



has formed annulus i\ 



a = intercept of the (TL):otolith length (OL) 



regression determined from the sample; 



OD^ - otolith distance from the nucleus to 



annulus /; 

 OR = otolith radius measured form the 

 nucleus to the edge of the anterior ros- 

 trum; and 

 L^ - TL at capture. 



Lengths at assigned ages were compared to back- 

 calculated lengths at ages by means of correlation 

 analysis. Total length back-calculation measure- 

 ments were used to construct a length-frequency dis- 

 tribution to further corroborate ages assigned to 

 length classes by otolith analysis. A Kolmogorov- 

 Smirnoff test was used to test for differences between 

 the observed and back-calculated length-frequency 

 distributions. 



Specimens from the GM 94-12 sur\'ey were labeled 

 according to station, cut lengthwise along the abdo- 

 men, and preserved in 10% formalin. Stomachs were 

 removed and their contents, after identification, sum- 

 marized by using both numerical and frequency of 

 occurrence indices. These analyses were further di- 

 vided into two length categories: 1-year-old (105-121 

 mm TL) and 2-7 year old (119-271 mm TL) fish, on 

 the basis of assigned age-class information. The size 

 range of the first year class was calculated by using 

 ±1 standard deviation from the mean. The second 



category, 2-7 year old fish, was calculated by using 

 the -1 deviation of the mean size of the age-2 year 

 class and the +1 deviation of the age-7 year class. 

 Fish (« = 1) that fitted within the size limits of these 

 two age groups were placed in the category with the 

 smallest difference between mean age-class length 

 and length of the fish. 



In addition to the 36 stomach samples from the 

 GM 94-12 survey, a sample of 112 frozen fish from 

 the GM 93-20 survey was obtained to inspect for oc- 

 currence of parasitism. Each gastrointestinal tract 

 was removed and cut into its component parts, then 

 sliced lengthwise to allow for inspection with a dis- 

 secting microscope. Nematodes were labeled accord- 

 ing to the location of infection, preserved in 70% etha- 

 nol, then transferred to 70% ethanol and 5% glycerine. 



Tr3rpanorhynch blastocysts and digenetic trema- 

 todes were preserved and stored in 10% buffered for- 

 malin. Trypanorhynchs were removed from their 

 blastocysts, transferred to 70% ethanol, and stained 

 overnight in Mayer's paracarmine diluted 1:9 in 70% 

 ethanol. Specimens were destained in 0.1 HCl in 70% 

 ethanol until internal organs were visible, neutral- 

 ized with 70% sodium bicarbonate for 20 minutes, 

 then dehydrated in a graded ethanol series prior to 

 clearing in methyl salicylate. Digenetic trematodes 

 were similarly prepared. Each specimen was 

 mounted on a glass slide in Canada balsam with 

 slides stored flat until the mounting medium had 

 hardened, when they were viewed under a dissect- 

 ing scope for identification. 



Helminth data were summarized by using preva- 

 lence, abundance, and intensity indicies. Prevalence 

 (%P) is defined as the number of infected individu- 

 als divided by the number of individuals examined 

 multiplied by 100, abundance (A) as the mean num- 

 ber of parasite species per host examined, and in- 

 tensity range (I range) as the number of parasite 

 species in each infected host (Margolis et al., 1982). 

 Helminth data were further divided into nine length- 

 class categories with ±1 standard deviation from the 

 mean of each age class. Again, fish («=3) that fitted 

 within the limits of two age classes were treated in the 

 same manner as in the food habits analysis. Splitting 

 the data into length-class categories provided a better 

 indication of the representative parasites in a respec- 

 tive year class than did dividing the data arbitrarily. 



Results 



Age and growth 



The length-frequency distribution of E. cimbrius 

 {n-727) sampled during the GM 93-20 survey ranged 



