490 



Fishery Bulletin 104(4) 



IC'S - 



20°S 



30 S 



40 S 



paraiso (32°54'S; 71°44'W), Talcahuano (36°44'S; 

 73°07'W), Valdivia (39°45'S; 73"08'W), and Aysen 

 Fjord (45°34'S; 72°05'W— and in the Golfo Nuevo 

 (42°45'S; 65°01'W) on the Argentinian coast. 



Fish were captured between April and No- 

 vember of 2003. Details for each sampling pe- 

 riod in each location are given in Table 1. The 

 depths of capture varied from 15 to 50 m. Fish 

 were captured 1) with hand lines from artisanal 

 boats, 2) by submarine divers, or 3) they were 

 acquired from local fishermen. Overall, a total 

 of 613 specimens were analyzed, of which 375 

 were males and 238 were females. 



Laboratory analysis 



Immediately upon being captured, the fish were 

 transferred from the sampling location to the 

 laboratory, where they were frozen until analy- 

 sis. The total length (TL) of each fish was mea- 

 sured to the nearest 1 cm of precision, total 

 weight was recorded to the nearest 0.5 g preci- 

 sion), and sex was determined by macroscopic 

 observation of the gonads. 



The prey items found in the stomachs were 

 sorted, counted, weighed, and identified to the 

 lowest possible taxonomic level, according to 

 Holthuis,^ Garth,'* Holmquist,^ and Retamal 

 (1981) for crustaceans, and according to Moreno 

 and Castilla-' for fishes. 



Dietary analysis 



Traditional methods of dietary analysis (counts, 

 frequencies of occurrence, and weights of indi- 

 vidual prey [Hyslop, 1980]) were used. These 

 methods can either underestimate or overesti- 

 mate the results. For this reason, a compound 

 method of analysis was used. One of the most 

 widely used compound indices in fish diet studies is the 

 index of relative importance, IRI (Pinkas et al., 1971), 

 which is expressed as a percentage, % IRI (Cortes, 

 1997). This index was calculated at the species and 

 phylum level. 



50 S 



2 Holthuis, L. 1952. The Crustacea Decapoda Macrura of 

 Chile. Reports of the Lund University Expedition 1948-49. 

 Lunds Universitets Arsskrift 247(10:1-110. Lunds Uni- 

 versitets, Lund, Sweden. 



^ Garth, J. 19.57. The Crustacea Decapoda Brachyura of 

 Chile. Reports of the Lund University Expedition 1948- 

 49. Lunds Universitets Arsskrift. number 7, 125 p. Lunds 

 Universitets, Lund, Sweden. 



■• Holmquist, C. 1957. Mysidacea of Chile. Reports of the 

 Lund University Expedition 1948-49. Lunds Universitets 

 Arsskrift. number 6, 52 p. Lunds Universitets, Lund, 

 Sweden. 



5 Moreno, C, and J. Castilla. 1978. Guide for the recognition 

 and observation of fish from Chile. Expedition of Chile, 

 120 p. Universidad Austral de Chile, Casilla 567-Isla Teja, 

 Valdivia, Chile. [In Spanish.] 



gcw 75°w eo-w 45°w acw 



80 W 70 W 



Figure 1 



Sampling locations for red rockfish iScba^tcs capensis) 

 along the Chilean and Argentinian coasts. l=Antofagasta, 

 2 = Coquimbo, 3=Valparaiso, 4=Talcahuano, 5=Valdivia, 

 6=Aysen, and 7 = Golfo Nuevo (Argentina). 



Feeding strategy 



The analysis of dietary composition proposed by Amund- 

 sen et al. (1996) was used. 



This is based on a graphic representation; prey-spe- 

 cific abundance is plotted against the prey's frequency 

 of occurrence on a two-dimensional space: 



Pi 



1^ 



100, 



where Pi = the prey-specific abundance of prey ;'; 



Si = the stomach contents (by weight) comprising 



prey /; and 

 S,, = the total stomach contents (by weight) in 



only those predators with prey / in their 



stomachs. 



This index gives information about prey importance 

 (dominant versus rare) and the feeding strategy (gen- 



