Hayse: Feeding, age, growth, and reproduction of Chaetodipterus faber 



69 



each taxon was determined by displacement, except in 

 the case of small items. Volumes of small items were 

 estimated using a 1 x 1 mm grid. Percent frequency of 

 occurrence (%F), percent of total number (%A^), and 

 percent of total volume {%V) of stomach contents were 

 calculated by higher taxonomic categories for the en- 

 tire data set and for the data separated by length in- 

 tervals and collection strata, and these values were 

 used to calculate an index of relative importance {IRI) 

 (Pinkas et al. 1971): 



IRI = (%N + %V) X %F. 



Shortly after laboratory analysis of stomach contents 

 commenced, it became apparent that most of the prey 

 items observed were either colonial or fragments of 

 soft-bodied organisms that could only be counted as one 

 individual per stomach each time they were encoun- 

 tered. The primary countable organisms were amphi- 

 pods and copepods, which were present in large num- 

 bers, but accounted for only a small portion of the 

 volume. In the opinion of some researchers (Lagler 

 1956, Crow 1982), percent number can lead to a some- 

 what distorted view if small organisms are present in 

 large numbers even though they are of minor impor- 

 tance to the diet of a fish. Consequently, a modified 

 index (MI) of relative importance which did not incor- 

 porate %N into the formula was also calculated for each 

 taxonomic category: 



MI = %F X %V. 



Stomach contents from 27 fish taken by hook-and-line 

 were analyzed separately in the above manner and com- 

 pared with the data from fish collected by nets and 

 spearfishing. 



Age and growth 



Sagittae were chosen as structures for age determina- 

 tion since they were easily accessible, easily stored, and 

 exhibited well-defined growth zones. Other structures 

 were deemed unsuitable as ageing structures for vari- 

 ous reasons. Scales were small, frequently regenerated, 

 and the marks on them were considerably more dif- 

 ficult to interpret than those on the sagittae. Vertebrae 

 were difficult to obtain and time-consuming to prepare 

 for analysis. 



No differences were observed between the left and 

 right sagittae of individual fish and, unless damaged 

 or unavailable, the left sagitta was used in age analysis. 

 Cleaned whole sagittae were examined using both 

 reflected and transmitted light, as necessary, in order 

 to establish the location of presumed annuli. Whole 

 otoliths were used since 100% agreement was obtained 



between counts of opaque zones from a subsample 

 {n =18) of otolith sections and counts from the cor- 

 responding whole otoliths. Sagittae were placed, 

 concave-side-up, in a dish containing cedar oil and view- 

 ed with reflected light against a dark background at 

 25 X using a stereoscope equipped with an ocular micro- 

 meter (1 ocular unit = 0.04 mm). The distance (in ocular 

 units) from the kernel to each opaque mark and from 

 the outermost mark to the edge of the sagitta (marginal 

 increment) was determined from measurements made 

 along the ventral radius of each sagitta (Fig. 2). These 

 measurements were performed on two occasions sep- 

 arated in time by at least 2 weeks, and the two sets 

 of measurements were compared to see if otolith mea- 

 surements were repeatable. 



Both least-squares linear regressions (Sokal and 

 Rohlf 1981) and geometric mean (GM) functional 

 regressions (Ricker 1973, Sokal and Rohlf 1981) were 

 used to describe the relationship of TL to ventral otolith 

 radius (OR), SL to TL, and lengths to weight. Mean 

 back-calculated TL at age was derived by the Fraser- 

 Lee method using the intercept of the OR-TL relation- 

 ship (Poole 1961, Carlander 1982). Mean back-calcu- 

 lated TLs were weighted by the reciprocal of the 

 standard error of the mean squared, and the von Ber- 

 talanffy growth equation (Bertalanffy 1938) was then 

 fitted to mean back-calculated TL at age by using the 

 SAS NLIN procedure while employing Marquardt's 

 algorithm and the SAS NLIN weight statement 

 (Helwig and Council 1979). 



Reproduction 



Reproductive tissues were processed in an Auto- 

 Technicon 2A Tissue Processor, vacuum infiltrated, 

 blocked in paraffin, and sectioned (6-10 i^m thick) on 

 a rotary microtome. Sections were stained with Har- 

 ris hematoxylin and counter-stained with eosin-Y. 

 Mounted sections were then examined with a com- 

 pound microscope, and sex and maturity stages were 

 assigned according to the criteria of Waltz et al. (1979). 

 The ratio of males to females and maturity stages by 

 month were examined to determine sex ratios and 

 spawning period of C. faber off South Carolina. 



Results 



Analysis of stomach contents 



Eighty percent of the stomachs obtained by hook-and- 

 line contained food, and at least seven prey species 



Reference to trade name.s does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



