DeMartini et al Pressure effects on regurgitation of Pnstipomoides filamentosus 



255 



precision for unbagged fish. Species or collections in 

 which the incidence of partial regurgitation is higher 

 thus may be less precise, depending on whether loss 

 of regurgitated prey is size-selective and in what 

 manner so. High incidence of total regurgitation will 

 be more likely to overwhelm size biases in retention 

 and thereby lower precision. 



Quantity and types of prey 



Our observations for pink snapper document a gen- 

 eral bias toward underestimating the total amount 

 of food ingested by this species if partial or complete 

 regurgitation goes undetected or is ignored. Because 

 the number of different prey detected reflects the 

 amount of material examined, underestimation of 

 quantity should often underestimate prey variety as 

 well, as our data illustrate. The observed underesti- 

 mation for all four major prey types argues that our 

 a priori classification of types, although subjective, 

 was sufficient to detect a general bias resulting from 

 regurgitation. 



Sizes of prey 



Size composition of prey can be biased by size-selec- 

 tive regurgitation. In the case of snapper, larger prey 

 items were selectively regurgitated and lost by 

 unbagged specimens. Large prey items might have 

 been more recently ingested, despite likely crepus- 

 cular or nighttime periods of peak foraging (Haight 

 et al., 1993b). The observed bias against retaining 

 large prey may explain much of the underestima- 

 tion of prey volume for unbagged snapper. Fishes 

 with other feeding mechanism structures (e.g. dif- 

 ferent gill-raker size and spacing) may lack such a 

 size bias, or regurgitation may result in the selec- 

 tive loss of smaller prey. Biases toward retaining 

 larger prey can dampen the effect of regurgitation 

 on underestimation of the total amount of prey 



Implications for diet studies 



Depending on the specific research question, future 

 studies of the food and feeding habits of juvenile pink 

 snapper may benefit from the use of specimens col- 

 lected at depth by divers. If simple comparisons of, 

 say, the relative abundances of major prey types 

 among fish collections are all that is needed, perhaps 

 unbagged fish will suffice. Conversely, if longshore 

 comparisons of large prey — perhaps comprising the 

 bulk of the diet — are necessary, or if field estimates 

 of total food intake or gastric evacuation rates are 

 required, it is obvious that specimens retrieved di- 

 rectly to the sea surface will provide underestimates. 



Clearly, ingestion estimates for unbagged specimens 

 will need to be adjusted for regurgitation loss, but 

 collection date effects that are strongly nonlinear may 

 make this difficult in practice. Further research is 

 needed to develop methods for quantifying effectively 

 the magnitude of regurgitation if collection date ef- 

 fects are nonmonotonic. Comparisons of diet among 

 unbagged snapper captured at different depths will 

 also be biased in terms of composition and size, as 

 well as in terms of quantity of prey, because the 

 amounts, diversity, and sizes of prey are interrelated 

 and because the magnitude of regurgitation is influ- 

 enced by extent of pressure change and hence depth 

 of capture. 



Bowman (1986) noted the progressive effects of 

 depth on extent of regurgitation for other species. 

 Prior explicit consideration of the interrelated biases 

 resulting from regurgitation, on the composition, size, 

 and quantity of prey has been lacking. The current 

 consensus that diet descriptions require compound 

 measures of both occurrence and bulk or mass data 

 (Hyslop, 1980; Bowen, 1983; but see MacDonald and 

 Green, 1983) reinforces the argument for critically 

 evaluating prey data which may have been affected 

 by barotrauma or other causes. 



Acknowledgments 



We thank A. Everson, T Kazama, and K. Landgraf 

 for assistance with diving or fishing operations. We 

 also thank C. Boggs, R. Brodeur, W. Haight, J. 

 Parrish, J. Polovina, M. Seki, and two anonymous 

 reviewers for constructive criticisms of drafts of the 

 manuscript. 



Literature cited 



Bowen, S. H. 



1983. Quantitative description of the diet. In L. A. Nielsen, 

 and D. L. Johnson (eds.). Fisheries techniques, p. 325- 

 336. Am. Fish. Soc, Bethesda, MD. 

 Bowman, R. E. 



1986. Effect of regurgitation on stomach content data of 

 marine fishes. Environ. Biol. Fish. 16:171-181. 

 Bromley, P. J. 



1994. The role of gastric evacuation experiments in quan- 

 tifying the feeding rates of predatory fish. Rev. Fish Biol. 

 Fish. 4:36-66. 

 Daan, N. 



1973. A quantitative analysis of the food intake of North 

 Sea cod, Gadus morhua. Neth. J. Sea Res. 6:479-517. 

 Durbin, E. G., A. G. Durbin, R. W. Langton, and 

 R. E Bowman. 



1983. Stomach contents of silver hake, Merluccius 

 bilineans, and Atlantic cod. Gadus morhua, and estima- 

 tion of their daily rations. Fish. Bull. 81:437-454. 



