453 



Feeding habits of the dwarf weakfish 

 (Cynoscion nannus) off the coasts 

 of Jalisco and Colima, Mexico 



Alma R. Raymundo-Huizar 



Centro Universitano de la Costa, Departamento de Ciencias 



Universidad de Guadalaiara 



Av. Universidad 203 



Puerto Vallarta, Jalisco, CP 48280 Mexico 



Horacio Perez-Espana 



Centro de Ecologia y Pesquerias 

 Universidad Veracruzana 

 Dr, Castelazo s/n. Xalapa 

 Veracruz, CP 91 190 Mexico 



Maite Mascaro 



Laboratono de Ecologia y Conducta 

 Unidad Academica Sisal 

 Universidad Nacional Autonoma de Mexico 

 Sisal, Yucatan, CP 97355 Mexico 



Xavier Chiappa-Carrara 



Unidad de Investigacion en Ecologia Marina, FES-Z 



Mexico, DF, CP 09230 Mexico 



Present address: Unidad Academica Sisal 



Universidad Nacional Autonoma de Mexico 



Sisal, Yucatan, CP 97355 Mexico 

 E-mail address (for X. Chiappa-Carrara, contact author) chiappaig'servidor unam mx 



Sciaenids from the Pacific coast of 

 Mexico are used as a second-class 

 fish species for human consumption 

 (Aguilar-Palomino et al., 1996). The 

 dwarf weakfish (Cynoscion nannus) 

 (Castro-Aguirre and Arvizu-Mar- 

 tinez, 1976) is often caught as bycatch 

 in the shrimp fishery but, because 

 of its small size (<27 cm TL, total 

 length), it is not considered a valuable 

 resource. This species can be found 

 in great numbers in waters between 

 100 and 812 m (Allen and Robert- 

 son, 1994; Fischer et al., 1995) asso- 

 ciated with the soft-bottom regions 

 off the coast of Jalisco and Colima 

 (Gonzalez-Sanson et al., 1997). 



Previous studies of the trophic bi- 

 ology of the Sciaenidae (Chao and 

 Musik, 1977; Campos and Corrales, 

 1986; Chao, 1995; Pelaez-Rodriguez, 

 1996; Cruz-Escalona. 1998; Lucena 

 et al., 2000) have shown that they 



feed on a variety of small fish and 

 benthic invertebrates (Allen and Rob- 

 ertson, 1994). However, there are few 

 studies concerning the feeding habits 

 of C. nannus, and its dietary prefer- 

 ences are not known. Considering its 

 abundance, C. nannus must play an 

 important role in the trophic rela- 

 tionships of soft-bottom ecosystems 

 in this region. 



Most studies describing the feeding 

 habits of fish have used the normal- 

 ized version of the breadth niche in- 

 dex proposed by Levins (1968). This 

 index is based both on the number of 

 food resources and on the proportion 

 of prey used by a species. The appro- 

 priate distribution function for this 

 index ensures sample independence 

 among prey found in any particular 

 stomach. Distribution functions based 

 either on the number or the relative 

 biomass or volume of dietary items do 



not ensure such independence, given 

 that all items found in any particular 

 stomach are statistically associated 

 (Hurlbert, 1984). Therefore, neither 

 the number nor the relative biomass 

 or volume of dietary items should be 

 used to calculate the Levins index. 

 The only distribution function that 

 ensures statistical independence is 

 that which is based on the proportion 

 of stomachs in which a certain food 

 resource is found (Krebs, 1999). 



Considering the ecological impor- 

 tance of studying the feeding habits 

 of this abundant fish species, we ex- 

 amined trophic breadth variations 

 (temporally and ontogenetically) of 

 C. nannus. When attempting to cor- 

 rectly apply the Levins index, we 

 used the distribution function of prey 

 that ensures statistical independence 

 among sampling units. 



Materials and methods 



The sampling area was located in 

 the central region of the continental 

 shelf off the Pacific coast of Mexico, 

 where the mouth of the river Cuitz- 

 mala, in Punta Farallon, Jalisco 

 (19°22'N, 105°01'W), is the northern 

 limit, and Cuyutlan, Colima (18 : 55'N, 

 104°08'W), is the southern limit. Sam- 

 ples of C. nannus were collected on a 

 monthly basis from January to Decem- 

 ber 1996 (except February, August, 

 and September) on the research vessel 

 BIP V, equipped with a trawl net with 

 a pair of codends. Sampling was car- 

 ried out over seven transects perpen- 

 dicular to the coast, each comprising 

 four bathymetric strata: 20, 40, 60, 

 and 80 m mean depth. 



Fish were individually identified, 

 measured (TL, ±1 mm), and the total 

 weight of each fish was recorded to 

 the nearest 0.1 g. The stomachs of 

 individual fish were dissected and 

 preserved in 10% neutralized forma- 

 lin. Stomach contents were analyzed 



Manuscript submitted 16 May 2003 

 to the Scientific Editor's Office. 



Manuscript approved for publication 



20 December 2004 by the Scientific Editor. 



Fish. Bull. 103:453-460 (2005 1. 



