496 



Fishery Bulletin 104(4) 



Mysidacea was the most important prey item. In this 

 geographic area, the mysids are also the most impor- 

 tant prey item for other rocky bottom fishes (Gonzalez 

 and Oyarzun, 2003). Mysids are found in large and 

 dense patches a few centimeters from the benthos (Silva 

 and Stuardo, 1985). 



The changes in prey types and their differences in 

 IRI were also supported by a correspondence analysis 

 (Fig. 2), which separated the sampling locations into 

 four major areas according to recorded prey types as 

 follows: 1) 23°S-33°S. 2) 37°S-40°S, 3) 45°S. and 4) 

 Golfo Nuevo (43°S), Argentina. These geographic areas 

 agree well with the proposed biogeographic units for 

 this region by Briggs (1995) and by Camus (2001). Ac- 

 cording to Briggs (1995). two faunistic provinces are 

 recognized in the southeastern Pacific: the Peruvian 

 province, which extends from Peru (ca. 10°S) down to 

 the northern Chilean coast (ca. 30°S), and the Magel- 

 lanic province located on the southern Chilean coast 

 (ca. 43°S). Between these provinces a transitional zone 

 is found, in which species typical of the Peruvian prov- 

 ince and the Magellanic province overlap (Brattstrom 

 and Johanssen, 1983; Lancelloti and Vasquez, 2000). 

 Additionally, Camus (2001) recognized five areas of 

 distributional discontinuities along the Chilean coast: 

 (1) 30°S; (2) 33°S; (3) 37°-38° S; (4) 41°-43°S, and (5) 

 52°-53°S. Thus prey, such as R. typus, are distrib- 

 uted along the Peruvian province and transitional zone, 

 whereas prey like mysids are present mainly between 

 latitudes 37°S and 40"S. 



Alternatively, Balech (1954) and Pequeiio (2000) con- 

 sidered all of South America's southern extremity as 

 the same biogeographic province. The records of prey 

 items (i.e., Miinida gregaria and M. subrugosa) in Golfo 

 Nuevo (Argentina coast) support the existence of a 

 connection between the southeastern Pacific and the 

 southwestern Atlantic. These prey species are also 

 distributed along the Chilean coast from Chiloe (42°S) 

 southward. However, these prey items were not found in 

 Aysen fjord, which could be explained by the presence 

 of two zones in the Aysen area with different physical 

 and faunistic conditions: the fjord channels and a ma- 

 rine zone (Arntz et al., 1999). Fish were sampled in the 

 Aysen channels, where species of Miinida are absent 

 (Retamal, 1981). Therefore, abundance and availability 

 of prey may be a factor explaining the differences in 

 the dietary composition of S. capensis between biogeo- 

 graphical regions. 



The present study indicates the existence of geograph- 

 ic variations in the dietary composition of S. capensis, 

 and these variations generally agree with the zoogeo- 

 graphic limits proposed in the literature. However, the 

 interpretation of this result is limited because of the 

 absence of an adequate background on the abundance 

 and biogeography of each prey item. 



Feeding strategy 



According to Ojeda and Farina (1996), S. capensis is a 

 nocturnal predator that feeds selectively on the decapod 



crustaceans R. typus and P. perlatus in the rocky subti- 

 dal shores of the central Chilean coast. Ojeda and Fariiia 

 suggested that these species are consumed preferentially 

 from a large array of potentially abundant invertebrate 

 prey items. Crustaceans are more active at night than 

 during the day, which could explain their large rate 

 of consumption by rockfishes that are active a night. 

 Decapod crustaceans have the highest caloric content 

 (Duarte et al., 1980), therefore Ojeda and Fariiia (1996) 

 proposed that the specialized diet of S. capensis reflects 

 an optimal foraging alimentary strategy based on the 

 maximization of available energy. On the other hand, 

 Moreno and Zamorano (1980) indicated that the feeding 

 strategy of a species has a direct relationship with its 

 morphological and behavioral adaptations that preclude 

 being preyed upon, and these adaptations could restrict 

 the potential prey items that they could capture. In this 

 context, Moreno et al. (1979) and Moreno (1981) studied 

 the bathymetric distribution of six species of fishes from 

 the rocky subtidal shore of the Valdivian coast, including 

 S. capensis. They established that the fishes segregate by 

 habitats in order to take refuge from predators, mainly 

 the sea lion Otaria flavescens; such segregation was not 

 related to the bathymetric distribution of their most 

 important prey items. 



Sebastes capensis shows buccal traits characteristic of 

 carnivorous speciesl it has a good protrusion capacity 

 and its teeth are inclined in a dorsal-caudal orienta- 

 tion that prevents the escape of prey once they have 

 been caught (Alvarado, 1985). Sebastes capensis is a 

 poor swimmer; its anatomy conforms to that of sed- 

 entary species in that its head and mouth are big in 

 comparison to the rest of the body, and it has big eyes 

 that allow an extensive angle of vision facilitating the 

 capture of their prey and the detection of their preda- 

 tors. Therefore, their capacity to catch prey depends 

 mainly on characteristics of their prey. All prey species 

 of S. capensis show similar behavioral traits (i.e., swim- 

 ming organisms that move between rocky intertidal and 

 subtidal zones) and S. capensis waits to capture them. 

 The specialized feeding strategy of S. capensis may be 

 a consequence of the availability of prey species moving 

 in the vicinity of the refuges of this fish. 



The feeding strategy index, as defined by Amund- 

 sen et al. (1996), has been used successfully for active 

 predators like sharks (e.g., Lucifora et al., 2000; Vogler 

 et al. 2003). However, this index does not integrate 

 either predator behavior or the availability of prey in 

 a given environment. Therefore, the criteria used to 

 differentiate specialist from generalist predators must 

 be clarified. According to the feeding strategy index 

 of Amundsen et al. (1996), S. capensis is a specialized 

 predator (preying mainly on crustaceans, although it 

 consumes some prey species, and other rare prey species 

 in almost all locations analyzed). In reality, S. capensis 

 is an ambush predator that does not actively search 

 for its prey. Because this fish waits in rocky refuges 

 for mobile prey, the probability of it encountering prey 

 depends, to a large extent, on prey abundances, prey 

 distributions, and prey behaviors. 



