ORIGINAL PAPER Spatio-temporal variations in the diet and stable isotope composition of the Argentine hake Merluccius hubbsi Marini, 1933 of the continental shelf of southeastern Brazil Elizabeti Yuriko Muto • Lucy Satiko Hashimoto Soares Received: 6 July 2009 / Accepted: 14 March 2011 / Published online: 31 March 2011 Ó Springer-Verlag 2011 Abstract The feeding ecology of Merluccius hubbsi was investigated in 2 regions of SE Brazil. The major food sources for the hakes were fish, crustaceans, and squid. In the upwelling system of Cabo Frio, the diet was very similar in the summers of 2001/2002 and spring 2002; fish were the most important prey followed by crustaceans. In Ubatuba, euphausiids were an important prey during the winter 2001 (100 m), while in the summer 2002, fish and amphipods predominated in the diet in the shallower site (40 m) and squid in the deeper site (100 m). The hakes showed temporal differences in stable isotope signatures in both regions, while C:N ratios varied only in Cabo Frio. d 15 N and d 13 C (bulk and corrected for lipid content) increased with fish length, which seems to be related to the increasing importance of fish and decreasing importance of euphausiids and amphipods in the diet of larger hakes. The mean trophic level of 3.7 for M. hubbsi was estimated using d 15 N of bivalves as baseline and the fractionation of 3.4% between trophic levels. Introduction Stable isotopes (SI) have been applied in ecological studies since the distribution of SI in natural systems reflects the integrated history of physical and metabolic processes within ecosystems (Peterson and Fry 1987). The isotopic composition of d 13 C in animals reflects their diets with an enrichment factor of 0.8 ± 1.1% due to preferential loss of 12 CO 2 during respiration, preferential assimilation of 13 C during digestion, and metabolic fractionation during syn- thesis of different tissues (DeNiro and Epstein 1978; Rau et al. 1983). The 13 C/ 12 C ratios are useful in tracing food webs where primary producers substantially differ in iso- tope composition (Peterson and Fry 1987). With regard to d 15 N, light isotopes are preferentially eliminated through nitrogenous excretion (DeNiro and Epstein 1981), and a mean enrichment of 3.4 ± 1.1% is expected between two consecutive trophic levels (Minagawa and Wada 1984). The differences between SI signatures of consumers and their diets, and the transfer of these signatures through successive feeding processes allow identifying the position of organisms in the food chain. The advantage of using d 15 N is that this isotope records the main flow of energy in the lower trophic levels, represents a time-integrated measure of trophic position, and detects trophic interac- tions not observed through stomach content analyses (Vander Zanden et al. 1997). The low primary production characteristic of the SE-S Brazilian waters is supported by regenerated nitrogen (NH 4 ? ); however, the supply of new nitrate (NO 3 - ) by the advection of the SACW (South Atlantic Central Water) on the continental shelf enhances the new production (Aidar et al. 1993; Gonzalez-Rodriguez et al. 1992). The food webs of Cabo Frio (CF) and Ubatuba (UB) were investi- gated using carbon and nitrogen SI (project DEPROAS/ PRONEX) to evaluate the influence of the coastal upwelling of CF and the seasonal eutrophication promoted by the subsurface advection of SACW on the continental shelf of UB on the trophodynamics of the demersal com- munity (Corbisier 2006). Because different forms of nitrogen bear distinct SI signatures, this tool has been Communicated by M. A. Peck. E. Y. Muto (&) Á L. S. H. Soares Laborato ´rio de Ecologia Tro ´fica, Departamento de Oceanografia Biolo ´gica, Instituto Oceanogra ´fico, Universidade de Sa ˜o Paulo, Sa ˜o Paulo, SP 05508-120, Brazil e-mail: mutokika@usp.br 123 Mar Biol (2011) 158:1619–1630 DOI 10.1007/s00227-011-1674-y