374 



Fishery Bulletin 102(2) 



snapper were observed over open habitat by a SCUBA 

 visual survey despite our observations that red snapper 

 are attracted to SCUBA divers. Thus we suggest that a 

 shift in habitat was more likely the cause of this absence 

 than trawl avoidance. 



The distinct diet shift as red snapper changed habitats 

 was independent of increasing size and suggested that 

 different benthic habitats play a critical role in the early 

 life history of this species. This separation was completely 

 independent of "a priori" knowledge of sample location and 

 fish size. For example, the MDS analysis showed almost 

 complete separation based on habitat rather then fish size 

 (Fig. 5). These differences between open and reef habitat 

 were readily apparent when prey taxa were separated 

 into lower taxonomic categories. For example, fishes such 

 as Halichoeres spp., Serranus spp., and Centr-op/istis spp., 

 were found only in the diets of reef-collected red snapper. 

 These species are closely tied to reef structure (Nelson 

 and Bortone, 1996). Prey shrimp also showed distinct 

 differences in red snapper diets between habitats. Over 

 open habitat, Mysidacea, Penaeidae, and Sergestidae were 

 important components. After the shift to reef habitat, 

 Mysidacea were absent and Penaeidae and Sergestidae 

 were greatly reduced, and Sicyoninae, Hippolytidae, and 

 Alpheidae became the dominant shrimp components. 

 The latter are all families typically associated with reef 

 habitats (Chance, 1970; Pequegnat and Heard, 1979). 

 One exception was the increased feeding on Squillidae, an 

 open habitat crustacean, at the largest size classes of this 

 study (220-250 mm SL; Fig. 9). For crabs, the separation 

 was not as clear, because of the dominance of Portunidae, 

 which can be assigned to both open and reef habitats. 

 However, increases in reef crabs were still apparent with 

 habitat shift, i.e., Diogeninae, Porcellanidae, and Leucosi- 

 idae can all be considered reef prey. Although Bradley and 

 Bryan (1975) pooled "juvenile" red snapper over open and 

 reef habitats, they did show a marked increase in fish prey 

 above 175 mm FL. This increase was almost exclusively 

 due to Batrachoididae or toadfishes, which are typically 

 found in reef habitat. We did not observe any toadfish prey 

 in our juvenile red snapper collections, but its presence in 

 this earlier study is consistent with present findings show- 

 ing a shift to feeding on reef-habitat prey. 



Red snapper diet shifted to greater percentages of reef- 

 prey with movement to reef habitat, but with this shift 

 they also continued feeding on other prey. This flexibility 

 in feeding habits allows red snapper to take advantage 

 of prey from wide-ranging habitats. Similar diet shifts 

 related to habitat shifts have been shown in schoolmaster 

 snapper, (L. apodus) (Rooker, 1995). The schoolmaster 

 snapper shifted from nearshore mangroves to coral reef 

 habitats near 70 mm SL; diets offish s70.0 mm SL were 

 dominated by crustaceans, particularly amphipods and 

 crabs. Fish >70.0 mm SL fed on fishes and to a lesser ex- 

 tent crabs, shrimps, and stomatopods. Similar diet shifts 

 were also shown for several fish species of Puget Sound. 

 For example in pile perch {Rhacochilus vacca), striped 

 seaperch iEmbiotoca lateralis), and quillback rockfish 

 (Sebastes maliger), the smallest juveniles preyed on open- 

 habitat plankton and benthic fauna, and medium-size 



and larger fish (>121 mm) of all three species shifted their 

 diets to include reef-associated prey ( Hueckel and Stayton, 

 1982). However, at larger sizes these three species were 

 not totally dependent on reef-associated prey. 



We have examined red snapper diets based on specific 

 volume of food. Although many other studies have used 

 an index of relative importance (IRI; Pinkas et al., 1971: 

 Cortes, 1997), we were specifically interested in the nutri- 

 tional value of particular prey, and prey separation into 

 open-habitat or reef-habitat. With IRIs these separations 

 would be more difficult to define, e.g., pelagic prey with 

 high numbers might be considered more important, but 

 actually provide little nutritional value to red snapper 

 diets (Macdonald and Green, 1983). Future studies on the 

 effects of red snapper predation on prey distributions may 

 be better suited for using IRIs. 



In summary, red snapper diets from open habitat 

 showed prey taxa associated with open sand-mud sub- 

 strate and the planktonic environment. Open-habitat prey, 

 such as chaetognaths, are known to be pelagic as well as 

 benthic, as are sergestid shrimp, calanoid copepods, my- 

 sids, and stomatopods (Williams, 1968; Manning, 1969; 

 Gosner, 1978; Stuck et al., 1979; Alldredge and King, 

 1985; Lindquist et al., 1994 ). Red snapper shifted diets to 

 reef-associated prey with their habitat shift, and this diet 

 shift was independent offish size. These diet shifts were 

 clearly apparent for both fish and shrimp prey but less 

 so for crab prey. As shown with marine fish species from 

 Puget sound, red snapper diets from reef habitat were not 

 restricted to reef-associated prey. For example, squids 

 were an important prey over both open and reef habitats 

 in the present study and our findings agree with those 

 of Bradley and Bryan (1975). The squids Loligo sp., and 

 Lolliguncula sp. are both abundant in nearshore coastal 

 waters and are not typically associated with reef structure 

 (Gosner 1978; Laughlin and Livingston, 1982; Hopkins 

 et al., 1987). Availability and ease in capture could be a 

 key as to why squid are important for red snapper over 

 size ranges of 40 to 240 mm SL. This flexibility in feeding 

 habits allows red snapper to take advantage of prey from 

 wide-ranging habitats, but clear shifts to additional reef 

 prey supports the hypothesis that reef structure provides 

 new prey resources. 



Acknowledgments 



We thank Joseph Conti, Kori M. Heaps, and Frank S. 

 Rikard for help in field collections and invertebrate 

 identification. This study was funded by NOAA, NMFS. 

 MARFIN grant number USDC-NA47FF0018-0. This is a 

 contribution of the Alabama Agricultural Station. 



Literature cited 



Alldredge, A. L., and J. M. King. 



1985. The distance demersal zooplankton migrate above 

 the benthos: implications for predation. Mar. Biol. 

 84:253-260. 



