Short communication Prey regurgitation in the grouper Cephalopholis argus By J. Dierking* and A. L. Meyer   Department of Zoology, University of Hawaii, Honolulu, HI, USA Introduction Groupers (Serranidae: Epinephelinae) are among the most important predatory reef fishes in the tropics and subtropics (Morris et al., 2000). Members of this family commonly regurgitate prey (i.e. expel prey from their stomachs by contraction of the esophagic muscle) during capture both by spearfishing (Kingsford, 1992) and by hook-and-line (Beukers- Stewart and Jones, 2004), presumably as a physiological stress reaction. However, no quantitative estimate of regurgitation has been published to date for a grouper, which poses problems for studies of grouper dietary ecology (Parrish, 1987). In particular, while stomach content analysis of groupers commonly shows high stomach vacuity (i.e. preva- lence of empty stomachs), it is usually unclear whether this is a characteristic of feeding ecology, or an artifact of regurgitation (see e.g. Renones et al., 2002). Furthermore, lack to account for regurgitation can strongly negatively bias prey consump- tion estimates (Bowman, 1986). Concern about ecosystem level effects of grouper declines due to overfishing (Morris et al., 2000) has led to recent efforts to better describe their feeding ecology. Clearly, lack of data on regurgitation in groupers stands in the way of this effort. The aim of this study was therefore to determine regurgitation rates for the grouper Cephalopholis argus in Hawaii obtained by spearfishing, one of the main capture techniques applied in studies of grouper feeding (Parrish, 1987). Specific goals were to interpret the ecological significance of the large stomach vacuity commonly observed in this species, and to use the observed regurgitation rate to develop an approximate cor- rection factor to account for regurgitation in prey consump- tion estimates of C. argus, which may also be applicable to other groupers. Materials and methods Cephalopholis argus has one of the widest natural distributions of any grouper, ranging from the Red Sea to the central Pacific Ocean. In Hawaii, naturally depauperate of groupers, the species was introduced in 1956 to create a new fishery. In July 2003 a total of 285 C. argus specimens from 17 sites in Hawaii were obtained by spearfishing on SCUBA. Following the approach by DeMartini et al. (1996) for the estimate of regurgitation, speared fish were sealed in plastic bags imme- diately after capture (i.e. underwater). In the few cases in which regurgitation occurred before sealing, regurgitated items were collected from the water column and included in the specimenÕs bag. Upon return to the laboratory, each specimen was then analyzed for occurrence of regurgitation (i.e. prey found in the mouth cavity or between the gill rakers, or completely expelled from the mouth cavity and found in the bags). The M of prey items was measured to the nearest milligram, and the degree of digestion classified as ÔLittleÕ, ÔMediumÕ, or ÔStrongÕ based on a scale of external character- istics (not shown). In addition, the standard length (SL) of C. argus specimens was determined to the nearest mm, and mass (M) to the nearest 5 g. Regurgitation rate was calculated as the number of regur- gitation incidents per number of specimens that had consumed prey (i.e. specimens with full stomachs and with empty stomachs that were artifacts of regurgitation). In addition, the fraction of the total M of all prey that was regurgitated was calculated. Stomach vacuity (the proportion of individuals with empty stomachs) and mean stomach fullness (M of prey per M of predator) of C. argus were then calculated twice: a first time while accounting for regurgitation (i.e. counting empty stomachs that were artifacts of regurgitation as full, and including regurgitated prey M in the calculation), and a second time without accounting for regurgitation. Finally, regurgitation rates for C. argus of different SL (<24 cm, 24–30 cm, >30 cm), and containing prey of different digestion status (Little, Medium, Strong), and prey of different M (£ 2.5 g, 2.5–7.5 g, >7.5 g), were compared using chi-square tests (Minitab version 14). Results Overall, regurgitation occurred in 25.7% (41 ⁄ 159) of C. argus specimens with full stomachs, and led to the loss of 33.2% (473.2 g ⁄ 1424.5 g) of the total M of stomach contents. Detailed results of the regurgitation analysis are shown in Table 1. Regurgitation was generally complete (i.e. all stomach contents expelled), which is not surprising as most stomachs contained only a single prey item. The ÔtrueÕ vacuity rate (i.e. accounting for regurgitation) in the analyzed sample was 44.9%, compared to 58.6% when not accounting for regurgi- tation. This means that almost half of the stomachs in the sample were naturally empty, compared to only 13.7% empty due to regurgitation. Regarding mean stomach fullness, the ÔtrueÕ value was 0.68% of own body M, compared to 0.46% when not accounting for regurgitation. Regurgitation thus Current addresses: *Centre dÕOce´anologie de Marseille, UMR CNRS 6540, Universite´ de la Me´diterrane´e, Station Marine dÕEndoume, rue de la Batterie des Lions, F-13007 Marseille, France.   Division of Aquatic Resources, Department of Land and Natural Resources, Anuenue Fisheries Research Center, Honolulu, HI, USA. J. Appl. Ichthyol. 25 (2009), 600–602 Ó 2009 The Authors Journal compilation Ó 2009 Blackwell Verlag, GmbH ISSN 0175–8659 Received: December 20, 2008 Accepted: February 10, 2009 doi: 10.1111/j.1439-0426.2009.01275.x U.S. Copyright Clearance Centre Code Statement: 0175–8659/2009/2505–0600$15.00/0 Applied Ichthyology Journal of