Relationships between butterflyfish (Chaetodontidae) feeding rates and coral consumption on the Great Barrier Reef

Transcription

1 Coral Reefs (2008) 27: DOI /s REPORT Relationships between butterflyfish (Chaetodontidae) feeding rates and coral consumption on the Great Barrier Reef M. A. Gregson Æ M. S. Pratchett Æ M. L. Berumen Æ B. A. Goodman Received: 4 April 2007 / Accepted: 25 February 2008 / Published online: 28 March 2008 Ó Springer-Verlag 2008 Abstract This study explored differences in the feeding rate among 20 species of coral reef butterflyfishes (Chaetodontidae) from Lizard Island, Great Barrier Reef. Feeding rate, measured as bites per minute (b.p.m.), varied between 2.98 ± 0.65 and ± 0.27 (mean ± SE) according to species and was positively related to the proportional consumption of coral (r 2 = 0.40, n = 20, P \ 0.01), independent of phylogeny (standardised independent contrasts r 2 = 0.29, n = 19, P \ 0.05). All species fed actively throughout the day, with obligate corallivores having a higher feeding rate at all times than either facultative corallivores or non-corallivores. The feeding rate of the obligate corallivores was also highest during the middle of the day. For eight of the species for which data was available, there was a positive correlation between bite rate and competitive dominance (r = 0.71, Communicated by Ecology Editor Professor Peter Mumby. M. A. Gregson (&) Department of Environmental Sciences, Institute for Water and Environmental Resource Management, University of Technology, Sydney, Broadway, NSW 2007, Australia Marcus.A.Gregson@uts.edu.au M. S. Pratchett M. L. Berumen ARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville, QLD 4811, Australia M. L. Berumen Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA B. A. Goodman School of Marine and Tropical Biology, James Cook University, Townsville, QLD 4811, Australia P \ 0.05). Chaetodon ephippium was the only species for which the feeding rate of pairs was higher than for solitary individuals. Keywords Competition Corallivore Diurnal variation Feeding guild Group behaviour Prey quality Introduction Energy acquisition is fundamental to the biology and ecology of all living organisms. Accordingly, choices regarding the type and relative quantities of prey consumed, as well as when and where to feed, greatly influence distributions, abundances, and fitness of most species (Krebs 1978; Hughes 1980). Prey capture is a major component of the time budget for most species, though actual feeding rates (the average number of bites per unit of time) vary greatly. Potential sources of variation include ambient temperature, which can directly affect activity levels of poikilotherms (Meeuwig et al. 2004), the size/maturity of the animal, where ontogenetic changes may include dietary shifts (Choat and Clements 1993), and/ or time of day (Ikeda 1977). An animal s feeding rate also tends to reflect its individual energy requirement relative to the energy obtained from selected food items (Sterner and Hessen 1994; Zharikov and Skilleter 2004). Given equal energetic requirements, organisms feeding on poor-quality prey must feed more frequently and/or ingest a greater volume of food (Birkeland and Neudecker 1981; Beamish and Medland 1986; Lares and McClintock 1991; but see Lawrence et al. 1989). Where energy intake is limited by the time available for feeding, animals must make increasingly stringent decisions to optimise feeding behaviour (Lucas 1983).

2 584 Coral Reefs (2008) 27: Aside from energetic return, feeding rates of animals are influenced by the distribution and abundance of prey (relating to interfeeding search times), and time required to capture and consume individual prey (handling times). For predatory animals, search and handling times are viewed as major limits on feeding rates (Holling 1959). Feeding behaviour may be further constrained by biological interactions with competitors and/or predators, which limit access to certain areas or prey types (Krause and Godin 1994) and reduce overall time available for feeding. Behavioural adaptations, such as group foraging, may counteract the constraints imposed by competition and predation, but there are still likely to be significant energetic costs associated with behaviours to minimise competition and predation (e.g., Lima and Dill 1990; Vahl et al. 2005). In aquatic environments, many teleost fishes have adopted group foraging behaviours, presumably because it reduces predation risk (Godin 1986) and/or increases feeding efficiency (Pitcher et al. 1982). The purpose of this study was to compare and contrast feeding behaviour among sympatric butterflyfishes (family Chaetodontidae), at Lizard Island, in the northern Great Barrier Reef of Australia. Butterflyfishes are a conspicuous and relatively common family of coral reef fishes, which exhibit considerable diversity in feeding habits. Butterflyfishes may feed on corals (hard and/or soft), algae, benthic invertebrates (motile or sedentary, including polychaete worms and crustaceans), and/or plankton (Reese 1977; Anderson et al. 1981; Sano 1989). In general, quantifying and comparing feeding rates among different species are often complicated by differences in feeding mode and gross morphology. Butterflyfishes, however, represent a monophyletic group of fishes with highly conserved jaw morphology (Ferry-Graham et al. 2001). Aside from a few species with elongate jaws, butterflyfishes have short robust jaws, adapted for biting corals and other attached prey (Motta 1988). Feeding mostly involves grabbing and tearing small pieces of soft tissue from benthic invertebrates, while some species supplement their diet by taking small discrete prey items, such as crustaceans. The various specialised morphologies of these fishes do not readily indicate what they feed on (Motta 1988). Butterflyfishes are, therefore, an ideal group to compare feeding rates on different prey types, independent of differences in morphology and foraging mode. This study tested several predictions using data on feeding rates obtained for 20 sympatric species of butterflyfishes at Lizard Island: (1) Do feeding rates of butterflyfishes correlate positively with the proportion of bites taken on hard coral? If coral tissue is a poor-quality food source (sensu Tricas 1989a), it would be expected that obligate corallivores compensate by feeding at much faster rates compared to facultative corallivores and non-corallivores. Conversely, obligate corallivores might compensate for poor-quality prey by feeding for a greater proportion on each day (Zekeria et al. 2002); (2) Do feeding rates vary through the day? It was predicted that there would be significant diurnal variation in bite rates of butterflyfishes, as shown for several other coral reef fishes (Taborsky and Limberger 1980), relating to variation in prey quality; (3) Do species-specific differences in social structure (i.e., solitary, paired, or grouped) or degree of competitiveness influence feeding rate? If feeding rates of butterflyfishes are limited by predation, solitary fishes might be expected to feed much more slowly compared to fishes feeding in pairs or larger groups (Magurran and Pitcher 1983). The theory of competitiveness and its relationship with feeding rate is more complicated. It was expected that competitively dominant species would be less constrained and/or able to monopolise high-quality prey items, and thus feed more rapidly in the presence of con-specifics (Milinski and Parker 1991). However, it could also be argued that competitively dominant fish spend more time defending territories than their subordinates, and hence have a depressed feeding rate (Elliott 2002). Materials and methods Feeding rates of butterflyfishes were obtained from feeding observations conducted between January 1995 and February 2002, at Lizard Island ( S, E), in the northern section of the Great Barrier Reef, Australia. Feeding observations were conducted for a total of 20 different species, including species from three distinct feeding guilds: obligate corallivores, facultative corallivores, and non-corallivores, categorised following Pratchett (2005). Feeding observations were conducted for at least 23, and up to 314, haphazardly selected individuals of each species (Table 1). During feeding observations, individual butterflyfish were followed for 3 min, recording the number of bites taken from each species of scleractinian (hard coral), alcyonarian (soft coral), or any other macro-invertebrate (e.g., Tridacnid clams). The number of bites taken on unidentified prey items from consolidated reef pavement, coral rubble, or sandy substrate was also recorded. These species were most likely feeding on small invertebrates (Zekeria et al. 2002), but no attempt was made to identify the specific source of prey for butterflyfishes feeding on these reef substrates. Observers moved through the study sites methodically to avoid resampling individual fishes. More detailed descriptions of sampling methodology are provided in Berumen and Pratchett (2006) and Pratchett (2005). Bite rates of butterflyfishes (specifically, the mean number of bites taken in each 3-min observation) were compared among species and among feeding guilds. Three distinct feeding guilds were recognised based on proportional

3 Coral Reefs (2008) 27: Table 1 Mean bites per minute (b.p.m., ±SE ) of all Chaetodontid species Species Subgenus Guild Proportional bites on hard corals Total mean b.p.m. (n), ±SE* Mean b.p.m. by time of day (n), ±SE Mean b.p.m. by social unit (n), ±SE h h h Alone Paired C. baronessa Gonochaetodon OC (314), (69), (134), (111), (111), (200), 0.33 C. trifascialis Megaprotodon OC (71), (25), (24), (22), (44), (27), 1.04 C. lunulatus Corallochaetodon OC (302), (60), 0.51 a (142), 0.46 b (100), 0.53 b 8.72 (68), (231), 0.36 C. aureofasciatus Discochaetodon OC (41), (10), (19), (12), (26), (14), 2.03 C. plebius Tetrachaetodon OC (105), (30), (46), (29), (54), (50), 0.53 C. rainfordi Discochaetodon OC (40), (15), (25), (31), (9), 1.01 C. citrinellus Exornator FC (114), (32), (48), (34), (51), (60), 0.52 C. rafflesi Rabdophorus FC (35), (6), (16), (13), (10), (24), 2.28 C. kleinii Lepidochaetodon FC (94), (33), (30), (31), (48), (46), 0.52 C. speculum Tetrachaetodon FC (25), (4), (14), (7), (10), (15), 1.12 C. unimaculatus Lepidochaetodon FC (32), (4), (22), (6), (13), (17), 0.71 C. melannotus Rabdophorus FC (46), (2), (28), (16), (32), (12), 0.97 C. ulietensis Rabdophorus FC (30), (2), (17), (11), (19), (11), 0.83 C. lunula Rabdophorus FC (28), (4), (20), (4), (11), (17), 0.39 C. ephippium Rabdophorus NC (48), (3), (39), (6), (15), 0.76 b 9.05 (32), 0.73 a C. vagabundus Rabdophorus NC (110), (30), (47), (33), (53), (57), 0.33 C. semion Rabdophorus NC (23), (9), (14), 0.95 C. auriga Rabdophorus NC (105), (31), (36), (38), (58), (47), 0.44 Ch. rostratus Chelmon NC (32), (2), 2.17 b 5.69 (14), 0.53 a 3.19 (16), 0.38 b 4.55 (11), (21), 0.48 C. lineolatus Rabdophorus NC (27), (12), 1.27 a 1.69 (15), 0.36 b 2.19 (14), (7), 0.18 Total b.p.m., b.p.m. arranged by time of day, and b.p.m. arranged by social unit are reported. The proportional number of bites on hard corals is also reported. Subgenera are from Smith et al. (2003); feeding guilds (OC = obligate corallivores, FC = facultative corallivores, and NC = non-corallivores) are from Pratchett (2005) * Data set with significantly different values (see Table 2 for post-hoc results) a,b Statistically similar values within each comparison (a = 0.05)

4 586 Coral Reefs (2008) 27: Table 2 Tukey post-hoc analysis results for multiple contrasts ANOVA comparing bite rates of Great Barrier Reef butterflyfish (Chaetodontidae) species C. baronessa C. trifascialis C. lunulatus C. aureofasciatus C. plebius C. rainfordi C. citrinellus C. rafflesi C. kleinii C. speculum C. unimaculatus C. melannotus C. ulietensis C. lunula C. ephippium C. vagabundus C. semion C. auriga Ch. rostratus C. lineolatus C. baronessa ns * * * * * * * * * * * * * * * * * * C. trifascialis ns ns * * * * * * * * * * * * * * * * * C. lunulatus * ns * * * ns * * * * * * * * * * * * * C. aureofasciatus * * * ns ns ns ns ns ns ns ns ns * ns ns ns ns ns * C. plebius * * * ns ns ns ns ns ns ns ns ns * ns ns ns * ns * C. rainfordi * * * ns ns * ns ns ns ns ns ns ns ns ns ns ns ns * C. citrinellus * * ns ns ns * ns ns * * * * * ns ns ns * * * C. rafflesi * * * ns ns ns ns ns ns ns ns ns * ns ns ns ns ns * C. kleinii * * * ns ns ns ns ns ns * ns ns * ns ns ns * * * C. speculum * * * ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns C. unimaculatus * * * ns ns ns * ns * ns ns ns ns ns ns ns ns ns ns C. melannotus * * * ns ns ns * ns ns ns ns ns ns ns ns ns ns ns * C. ulietensis * * * ns ns ns * ns ns ns ns ns ns ns ns ns ns ns * C. lunula * * * * * ns * * * ns ns ns ns * * ns ns ns ns C. ephippium * * * ns ns ns ns ns ns ns ns ns ns * ns ns ns * * C. vagabundus * * * ns ns ns ns ns ns ns ns ns ns * ns ns ns ns * C. semion * * * ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * C. auriga * * * ns * ns * ns * ns ns ns ns ns ns ns ns ns * Ch. rostratus * * * ns ns ns * ns * ns ns ns ns ns * ns ns ns ns C. lineolatus * * * * * * * * * ns ns * * ns * * * * ns * Significant difference in bite rate, ns indicates no significant difference (a = 0.05) consumption of hard (scleractinian) corals: (1) obligate corallivores, which take [80% of bites from corals, (2) facultative corallivores, which take between 20 and 70% of bites from hard corals, and (3) non-corallivores, which take \5% of bites from hard corals. Assignment of species to feeding guilds follows Pratchett (2005). Interspecific variation in the bite rates of butterflyfishes were analysed using a two-way nested ANOVA with species nested within feeding guild to test for variation both within and among feeding guilds. However, related species do not represent statistically independent data, such that comparisons involving conventional statistics may be invalid (Felsenstein 1985; Harvey and Pagel 1991). For example, the obligate corallivores in this study belonged to a single monophyletic group (Fessler and Westneat 2007). To circumvent this problem, phylogenetic analyses were undertaken using a pruned phylogeny of a recent molecular phylogeny of the butterflyfishes (Fessler and Westneat 2007) that included 18/20 species examined in this study. As Chaetodon aureofasciatus and Chaetodon lunulatus were absent from the phylogeny these species were placed with their closest relative/sibling species (i.e., Chaetodon rainfordi and Chaetodon trifasciatus, respectively) (Kuiter 1995). To examine the evolution of mean bite rate in response to the % hard coral consumed by 20 species of butterflyfishes, standardised phylogenetic independent contrasts were calculated using PDTREE (Garland et al. 1992). Traits were standardised by dividing the independent contrast of each trait by the standard deviation of the branch length (square root of the corrected branch lengths) of that trait (Garland et al. 1992). As there were no significant linear or non-linear trends in the data, all branch lengths were deemed adequately standardised as required under a Brownian motion model of evolution. Percentages were converted to proportions and arcsine square-root transformed prior to analyses to achieve normality Quinn and Keough (2002). The independent contrasts of mean bite rate and % hard coral consumed were compared using regression analysis. Independent ANOVA was then used to test for variation in bite rates within species, both between solitary versus paired individuals and at different times through the day. Bonferroni-corrected alpha-levels were used to assess significance across the 20 separate analyses. Feeding rates of solitary versus paired individuals were analysed separately for each species, using one-way ANOVA. To further test the influence of competition on feeding rates of butterflyfishes, interspecific variation in feeding rates of butterflyfishes was related to a dominance hierarchy derived by Berumen and Pratchett (2006), where competitive dominance was

5 Coral Reefs (2008) 27: Fig. 1 Mean number of bites per minute of butterflyfishes (±SE) in the genera Chaetodon and Chelmon. Colour coding represents feeding guilds (OC = obligate corallivores, black bars; FC = facultative corallivores, grey bars; NC = non-corallivores, white bars) Mean Bites per Minute C. baronessa C. trifascialis C. lunulatus C. aureofasciatus C. plebius C. rainfordi C. citrinellus C. rafflesi C. kleinii C. speculum C. melannotus C. unimaculatus C. ulietensis C. lunula C. ephippium C. vagabundus C. semion C. auriga Ch. rostratus C. lineolatus quantified based on the proportion of aggressive interactions between individual fish. Rank correlation was used to test whether feeding rates of butterflyfishes were related to competitive dominance across 8 different species for which competitive interactions have been directly quantified. Feeding observations were conducted throughout the day, with approximately equal sampling within each of three pre-designated time periods: morning ( h), mid-day ( h), and afternoon (1400 h 1800 h). No nocturnal foraging was ever witnessed, or has ever been recorded, for the 20 species considered in this study (cf. Zekeria et al. 2002). To test for diurnal variation in feeding rates of butterflyfishes, average feeding rates of each species were compared among the three different time periods using a series of one-way ANOVAs. Data collected for each species were essentially independent, but Bonferronicorrected alpha-levels were still used to test for significance across the 20 separate analyses. All data were square-root transformed prior to analyses to satisfy assumptions of normality and univariate homogeneity. Results Mean bite rates varied significantly among species (Fig. 1) (F 19, 1602 = 23.4, P \ 0.01), ranging from 2.98 (±0.65 SE) bites per minute (b.p.m.) for Chaetodon lineolatus, up to (±0.27 SE) b.p.m. for Chaetodon baronessa (Tables 1 and 2). Much of the variation in feeding rates was ostensibly related to differences in major dietary items. Notably, mean bite rates varied significantly among feeding guilds (F 2, 1619 = 6.56, P \ 0.001), where obligate corallivores had the highest rate of feeding, taking a mean of (±0.18 SE) b.p.m., compared to 6.53 (±0.23 SE) b.p.m. for facultative corallivores, and 5.7 (±0.18 SE) b.p.m. for non-corallivores (Table 3). There was no significant difference in the feeding rates between facultative corallivores and non-corallivores. Obligate coral-feeders (C. aureofasciatus, C. baronessa, C. lunulatus, Chaetodon plebius, C. rainfordi, and C. trifascialis) also represent a single monophyletic group, distinct from facultative and non-coral feeders, so independent contrasts were necessary to separate effects of ecology and phylogeny. Bite rates of butterflyfishes were significantly and positively related to the proportion of bites taken on hard corals by each of the 20 species (r 2 = 0.40, n = 20, P \ 0.01, Fig. 2), and the results from the phylogenetically corrected data were congruent with the uncorrected data. The relationship between standardised independent contrasts for mean bite rates versus standardised independent contrasts for % coral consumption was weaker compared to the uncorrected data, but nonetheless significant (r 2 = 0.29, n = 19, P \ 0.05, Fig. 3). All butterflyfishes were active during daylight hours and spent virtually all their time feeding. Only C. lunulatus, Table 3 Mean bites per minute (b.p.m., ±SE) and mean b.p.m. arranged by time of day for three Chaetodontid feeding guilds Feeding guild Total mean b.p.m. (n), ±SE Mean b.p.m. by time of day (n), ±SE h h h Obligate Corallivores (873), 0.18 a 9.79 (194), 0.35 b (380), 0.28 a 9.73 (299), 0.30 b Facultative Corallivores 6.53 (404), 0.23 b 7.31 (87), 0.43 b 5.77 (195), a 7.18 (122), 0.52 b Non-corallivores 5.7 (345), 0.18 b 5.4 (66), (157), (122), 0.32 a,b Statistically similar values within each comparison (a = 0.05)

6 588 Coral Reefs (2008) 27: Fig. 2 Regression of mean number of bites per minute against proportional bites on scleractinian corals (r 2 = 0.40, P \ 0.01). Each data point represents the mean bite rate of all individuals studied for each of 20 species. Colour coding represents feeding guilds (OC = obligate corallivores, black dots; FC = facultative corallivores, grey dots; NC = non-corallivores, white dots) Mean Bites per Minute Proportional Number of Bites on Scleractinian Corals Fig. 3 Regression of mean number of bites per minute (standardised for independent contrasts) against proportional bites on scleractinian corals (standardised for independent contrasts) (r 2 = 0.29, P \ 0.05) Standardised independent contrasts of mean bite rates Standardised independent contrasts of % coral consumption Chelmon rostratus and C. lineolatus displayed significant diurnal variation in feeding rates. For C. lunulatus, mean bite rates were much higher at mid-day (11.32 ± 0.46 SE b.p.m.) and during the afternoon (10.04 ± 0.53 SE b.p.m.), than in the morning (7.92 ± 0.51 SE b.p.m.) (F 2, 302 = 10.82, P \ 0.01) (Table 1). Mean b.p.m. for Ch. rostratus and C. lineolatus were significantly higher at mid-day (5.69 ± 0.53 SE, 4.58 ± 1.27 SE, respectively), than morning (3.83 ± 2.17 SE, no morning data for C. lineolatus) or afternoon (3.19 ± 0.38 SE, 1.69 ± 0.36 SE, respectively) (F 2,.32 = 6.99, P \ 0.01; F 2, 27 = 5.8, P \ 0.05, respectively) (Table 1). The diurnal feeding behaviour of each feeding guild was also explored. Both facultative and obligate corallivores displayed diurnal variation in bite rate: obligate corallivores had a higher bite rate during the mid-day time period (F 2, 870 = 7.36, P \ 0.01), while facultative corallivores had a significantly lower bite rate during the same mid-day period (F 2, 401 = 6.54, P \ 0.01) (Table 3). In both guilds, the morning and afternoon bite rates did not differ significantly from each other (Table 3). In addition to differences among feeding guilds, there was also significant variation in feeding rates of butterflyfishes within each feeding guild (F 17, 1602 = 23.40, P \ 0.01), accounting for 22% of overall variation. Interspecific variation in bite rates of butterflyfishes was partly attributable to differences in the proportional targeting of corals. The influence of other butterflyfish on the feeding rate of a focal fish was investigated in two ways during this study. First, competitive behaviour towards another butterflyfish was related to feeding rate. There was a positive correlation (Pearson s coefficient = 0.71, P \ 0.05) between mean bite rates and competitive dominance (measured as aggressive interactions) for eight of the species of butterflyfish examined in this study for which competitive dominance has been previously quantified (Chaetodon auriga, C baronessa, Chaetodon citrinellus, C. plebius, C.

7 Coral Reefs (2008) 27: trifascialis, C. lunulatus, Chaetodon vagabundus, Chaetodon kleinii) (Berumen and Pratchett 2006). Second, the influence of pairing on feeding rate was examined. Within species, significant variation in feeding rates among individuals related to their social arrangement was expected. However, only Chaetodon ephippium exhibited significant variation in feeding rates between solitary and paired individuals. Mean bite rates of paired C. ephippium (9.05 ± 0.73 SE b.p.m., n = 32), were approximately twice that of solitary individuals (4.89 ± 0.76 b.p.m., n = 15) (F 2, 47 = 15.24, P \ 0.01). Discussion Bite rates of butterflyfishes varied considerably among species and among trophic groups (Fig. 1), with obligate corallivores having a higher bite rate compared to facultative corallivores and non-corallivorous butterflyfishes (Table 3). Moreover, there was a significant positive correlation between standardised independent contrasts for mean bite rates versus standardised independent contrasts for % coral consumption across all the 20 species (Fig. 3). One possible interpretation of this is that corals are poorquality prey, as has been reported previously (Tricas 1985, 1989a). Tricas (1989a) estimated that the energetic intake for butterflyfishes feeding on corals was J per bite. In contrast, the energetic intake for butterflyfishes feeding on motile invertebrates is between 0.08 and 3.25 J per bite, depending on whether they target smaller ( lm) or larger ([2 mm) prey items (Botrell and Robins 1984; Fleeger and Palmer 1982). Assuming that non-corallivorous butterflyfishes target larger prey items, then energetic intake from feeding on motile invertebrates may be as much as 10 times the energetic intake derived from coral feeding. In this case, corallivores would indeed need to have a higher bite rate to obtain a similar energy intake to non-corallivores, though considerable research is required to identify specific sources of prey for non-coral feeding butterflyfishes and quantify the energetic content of these prey items (Pratchett 2007). Higher feeding rates among corallivores compared to non-corallivores may also be related to differences in the time required to find suitable prey. When feeding on corals, some butterflyfishes exhibit intense feeding bouts (up to 20 bites in quick succession), after which they tend to move to a new patch of coral tissue (on either an entirely new colony or on different parts of larger colonies) and begin searching (Tricas 1989a), possibly looking for polyp clusters with greatest expansion (Gochfeld 2004). Coralfeeding butterflyfishes typically spend very limited time searching for prey (Tricas 1985). In contrast, butterflyfishes feeding on unidentified prey items on non-coral substrates undertake lengthy inspections (up to 1 min) of the substrate before nearly every bite. The time required to identify small discrete prey items will certainly constrain the maximum possible bite rates for non-coral feeders, though butterflyfishes that bite at exposed tentacles of discrete coral polyps must carefully choose feeding locations to maximise the amount of material consumed with each bite (Gochfeld 2004). Moreover, the morphology of corals limits the amount of tissue consumed with each bite (Tricas 1989a). Few butterflyfishes have sufficiently robust jaws to bite into the carbonate skeleton of scleractinian corals and so bite size is generally limited by tissue depth (Motta 1988). Irrespective of prey quality, corallivorous butterflyfishes may need to feed at a faster rate or for a longer time period to ingest similar quantities of food compared to species feeding on discrete prey items (Birkeland and Neudecker 1981; Tricas 1985, 1989b). Many organisms, especially herbivores, exhibit distinct diurnal patterns of feeding activity (Choat and Clements 1993), tending to maximise feeding when prey is most valuable (Taborsky and Limberger 1980; Zoufal and Taborsky 1991; but see Zemke-White et al. 2002). This study indicated that very few butterflyfishes varied their feeding behaviour through the day; however, obligate corallivores as a group had a higher bite rate during the middle of the day than other time periods (Table 3). This pattern is consistent with studies showing that the nutritional quality of coral and algal species is highest at mid-day (Crossland et al. 1980; Zoufal and Taborsky 1991). Similarly, Harmelin-Vivien and Bouchon-Navaro (1983) and Zekeria et al. (2002) reported that feeding rates of butterflyfish species vary through the day. Although not explicitly tested, this suggests that corallivores may be exploiting their prey when most nutritionally beneficial. In addition to significant interspecific variation in feeding rates among species, there was also significant variability amongst species within the same feeding guild. The above-mentioned trade-off between feeding rate and search time may be responsible for this. Specialised corallivores may take 80 to [99% of their total bites from hard corals. One species of butterflyfish has been recorded taking 88% of its total number of bites from just one coral species (Pratchett 2005). Extreme dietary specialisation may facilitate increased feeding rates as fishes inspect only a few closely positioned coral colonies, reducing search time (Tricas 1985). However, even among the obligate corallivores, all of which took [90% of bites from scleractinian corals, there was a greater than two-fold difference in feeding rate between the fastest and slowest feeders (Fig. 1). Highest feeding rates were recorded for C. baronessa and C. trifascialis, which are the most specialised corallivorous butterflyfishes (Pratchett 2007) and are also competitively dominant (Berumen and Pratchett 2006).

8 590 Coral Reefs (2008) 27: Competitive dominance is strongly present among butterflyfish species, whereby most corallivorous species maintain small, non-overlapping feeding territories and attempt to exclude con-generic corallivores (Berumen and Pratchett 2006). For this study, competitiveness was determined from a dominance hierarchy (Berumen and Pratchett 2006) that details aggressive interactions between individuals. The results indicate that butterflyfishes feeding at the highest rates are also highly competitive obligate corallivores, although this was only restricted to eight of the species studied. In theory, maintenance of feeding territories will reduce time available for feeding, but it is interesting in this study that the most aggressive fishes still manage to have the highest feeding rates (i.e., obligate corallivores). Alternatively, if competitive dominant obligate corallivores have higher-quality food within their territories, then perhaps they obtain more energy per bite. They would thus be expected to feed less than subordinate competitors forced to consume poorer-quality resources with potential impacts on their health (Pratchett et al. 2004; Berumen et al. 2005). Competitive dominants could also have exclusive access to areas with greatest prey density, allowing them to feed more efficiently inside these territories, further reducing the searching time required. For whatever reason, though, maintenance of feeding territories seems to be a worthwhile energy investment for these species. Pairing is the predominant social strategy employed by butterflyfishes (Roberts and Ormond 1992) and is largely related to monogamous breeding (Pratchett et al. 2006). However, pairing or grouping may also confer other benefits, such as increased foraging efficiency (e.g., Gregson and Booth 2005), increased efficiency in defending feeding territories (Roberts and Ormond 1992), and/or increased vigilance against predators (Wilson 1980). Importantly, mutual vigilance in con-specific aggregations should increase the probability of predator detection (Clark and Mangel 1986), consequently increasing the time each individual can spend on other activities, such as feeding (Wilson 1980; Werner and Mittlebach 1981). The results of the present study, however, suggest that pairing had no effect on feeding rate for these butterflyfish species (except for the non-corallivore, C. ephippium, and also see Bonaldo et al. 2005). For corallivorous butterflyfishes, it appears that feeding rates are maximised irrespective of social organisation, and pairing may or may not increase survivorship due to foraging gains (Pratchett et al. 2006). This study has revealed significant variation in feeding rates of sympatric butterflyfish species and guilds at Lizard Island, northern Great Barrier Reef. Social strategy and time of day influenced the feeding rates of few individual species, though two of the three feeding guilds exhibited distinct patterns of feeding through the day. For eight of the study species, there was a positive correlation between bite rate and aggressive interactions, suggesting that corallivorous species may be a more competitive guild. Apparent variation in feeding rates between corallivorous and noncorallivorous butterflyfishes may be best explained in terms of the time required to find their respective prey items, although the nutritional quality of corals relative to other potential prey items remains unclear and warrants further investigation. Acknowledgements Comments from two anonymous reviewers greatly improved this manuscript. This research was funded by a Merit Research Grant from James Cook University awarded to MSP and a Graduate Research Fellowship from the National Science Foundation (USA) to MLB. Field assistance was provided by A.H. Baird, R. Thomas, and S.L. Watson. The authors are grateful to staff at Lizard Island Research Station for ongoing logistical support. References Anderson GRV, Ehrlich AH, Ehrlich PR, Roughgarden JD, Russell BC, Talbot FH (1981) The community structure of coral reef fishes. Am Nat 117: Beamish FWH, Medland TE (1986) Protein sparing effects in large rainbow trout, Salmo gairdneri. Aquaculture 55:35 42 Berumen ML, Pratchett MS (2006) Effects of resource availability on the competitive behaviour of butterflyfishes (Chaetodontidae). Proc 10th Int Coral Reef Symp 1: Berumen ML, Pratchett MS, McCormick MI (2005) Within-reef differences in diet and body condition of coral-feeding butterflyfishes (Chaetodontidae). Mar Ecol Prog Ser 287: Birkeland C, Neudecker S (1981) Foraging behaviour of two Carribean Chaetodontids: Chaetodon capistratus and C. aculeatus. Copeia 1: Bonaldo RM, Krajewski JP, Sazima I (2005) Meals for two: foraging activity of the butterflyfish Chaetodon striatus (Perciformes) in southeast Brazil. Braz J Biol 65: Bottrell HH, Robins DB (1984) Seasonal variations in length, dry weight, carbon and nitrogen of Calanus helgolandicus from the Celtic Sea. Mar Ecol Prog Ser 14: Choat JH, Clements KD (1993) Daily feeding rates in herbivorous labroid fishes. Mar Biol 117: Clark CW, Mangel M (1986) The evolutionary advantage of group foraging. Theor Popul Biol 30:45 75 Crossland CJ, Barnes DJ, Borowitzka MA (1980) Diurnal lipid and mucus production in the staghorn coral Acropora acuminata. Mar Biol 60:81 90 Elliott JM (2002) Shadow competition in wild juvenile sea-trout. J Fish Biol 61: Felsenstein J (1985) Phylogenies and the comparative method. Am Nat 126:1 25 Ferry-Graham LA, Wainwright PC, Bellwood DR (2001) Prey capture in long-jawed butterflyfishes (Chaetodontidae): the functional basis of novel feeding habits. J Exp Mar Biol Ecol 256: Fessler JL, Westneat MW (2007) Molecular phylogenetics of the butterflyfishes (Chaetodontidae): Taxonomy and biogeography of a global coral reef fish family. Mol Phyl Evol 45:50 68 Fleeger JW, Palmer MA (1982) Secondary production of the estuarine, meiobenthic copepod Microarthridion littorale. Mar Ecol Prog Ser 7:

9 Coral Reefs (2008) 27: Garland TJ, Harvey PH, Ives AR (1992) Procedures for the analysis of comparative data using phylogenetic independent contrasts. Syst Biol 41:18 32 Gochfeld DJ (2004) Predation-induced morphological and behavioral defenses in a hard coral: Implications for foraging behavior of coral-feeding butterflyfishes. Mar Ecol Prog Ser 267: Godin J-G (1986) Antipredator function of shoaling in teleost fishes: a selective review. Nat Can 113: Gregson MA, Booth DJ (2005) Zooplankton patchiness and the associated shoaling response of the temperate reef fish Trachinops taeniatus. Mar Ecol Prog Ser 299: Harmelin-Vivien ML, Bouchon-Navaro Y (1983) Feeding diets and significance of coral feeding among Chaetodontid fishes in Moorea (French Polynesia). Coral Reefs 2: Harvey PH, Pagel MD (1991) The comparative method in evolutionary biology. Oxford University Press, Oxford Holling CS (1959) The components of predation as revealed by a study of small mammal predation of the European pine sawfly. Can Entomol 91: Hughes RN (1980) Optimal foraging theory in the marine context. Oceanogr Mar Biol Ann Rev 18: Ikeda T (1977) Feeding rates of planktonic copepods from a tropical sea. J Exp Mar Biol Ecol 29: Krause J, Godin J-G (1994) Shoal choice in the banded killifish (Fundulus diaphanus, Teleostei, Cyprinodontidae): Effects of predation risk, fish size, species composition and size of shoals. Ethology 98: Krebs JR (1978) Optimal foraging: decision rules for predators. In: Krebs JR, Davies NB (eds) Behavioural Ecology: An evolutionary approach. Blackwell Science, Oxford, pp Kuiter RH (1995) Chaetodon lunulatus, a sibling species of C. trifasciatus, with observations on other sibling species of butterfly fish (Chaetodontidae). Revue francaise d Aquariologie 21: Lares MT, McClintock JB (1991) The effects of food quality and temperature on the nutrition of the carnivorous sea urchin Eucidaris tribuloides (Lamarck). J Exp Mar Biol Ecol 149: Lawrence J, Regis M-B, Delmas P, Gras G, Klinger T (1989) The effect of quality of food on feeding and digestion in Paracentrotus lividus (Lamarck) (Echinodermata: Echinoidea). Mar Behav Physiol 15: Lima SL, Dill LM (1990) Behavioral decisions made under the risk of predation: A review and prospectus. Can J Zool 68: Lucas JR (1983) The role of foraging time constraints and variable prey encounter in optimal diet choice. Am Nat 122: Magurran AE, Pitcher TJ (1983) Foraging, timidity and shoal size in minnows and goldfish. Behav Ecol Sociobiol 12: Meeuwig MH, Dunham JB, Hayes JP, Vinyard GL (2004) Effects of constant and cyclical thermal regimes on growth and feeding of juvenile cutthroat trout of variable sizes. Ecol Freshw Fish 13: Milinski M, Parker GA (1991) Competition for resources. In Krebs JR, Davies NB (eds) Behavioural ecology, 3rd ed. (Krebs JR, Davies NB, eds) Blackwell, Oxford, pp Motta PJ (1988) Functional morphology of the feeding apparatus of ten species of Pacific butterflyfishes (Perciformes, Chaetodontidae): An ecomorphological approach. Environ Biol Fish 22:39 67 Pitcher TJ, Wyche CJ, Magurran AE (1982) Evidence for position preferences in schooling mackerel. Anim Behav 30: Pratchett MS (2005) Dietary overlap among coral-feeding butterflyfishes (Chaetodontidae) at Lizard Island, northern Great Barrier Reef. Mar Biol 148: Pratchett MS (2007) Dietary selection by coral-feeding butterflyfishes (Chaetodontidae) on the Great Barrier Reef, Australia. Raffles Bull Zool 2007 Suppl: Pratchett MS, Wilson SK, Berumen ML, McCormick MI (2004) Sublethal effects of coral bleaching on an obligate coral feeding butterflyfish. Coral Reefs 23: Pratchett MS, Pradjakusuma OA, Jones GP (2006) Is there a reproductive basis to solitary living versus pair-formation in coral reef fishes? Coral Reefs 25:85 92 Quinn GP, Keough MJ (2002) Experimental design and data analysis for biologists. Cambridge University Press, Cambridge Reese ES (1977) Coevolution of corals and coral feeding fishes of the family Chaetodontidae. Proc 3rd Int Coral Reef Symp 1: Roberts CM, Ormond RFG (1992) Butterflyfish social behaviour, with special reference to the incidence of territoriality: a review. Environ Biol Fish 34:79 93 Sano M (1989) Feeding habits of Japanese butterflyfishes (Chaetodontidae). Env Biol Fish 25: Smith WL, Webb JF, Blum SD (2003) The evolution of the laterophysic connection with a revised phylogeny and taxonomy of butterflyfishes (Teleostei: Chaetodontidae). Cladistics 19: Sterner RW, Hessen DO (1994) Algal nutrient limitation and the nutrition of aquatic herbivores. Annu Rev Ecol Syst 25:1 29 Taborsky M, Limberger D (1980) The activity rhythm of Blennius sanguinolentus Pallus: an adaptation to its food source? PSZNI Mar Ecol 1: Tricas TC (1985) The economics of foraging in coral-feeding butterflyfishes of Hawaii. Proc 5th Int Coral Reef Congr 5: Tricas TC (1989a) Prey selection by coral-feeding butterflyfishes: strategies to maximize the profit. Environ Biol Fish 25: Tricas TC (1989b) Determinants of feeding territory size in the corallivorous butterflyfish, Chaetodon multicinctus. Anim Behav 37: Werner EE, Mittlebach GG (1981) Optimal foraging: field tests of diet choice and habitat switching. Am Zool 21: Vahl WK, Van der Meer J, Weissing FJ, Van Dullemen D, Piersma T (2005) The mechanisms of interference competition: two experiments on foraging waders. Behav Ecol 16: Wilson EO (1980) Sociobiology: The abridged edition. Harvard University Press, Cambridge Zekeria ZA, Dawit Y, Ghebremedhin S, Naser M, Videler JJ (2002) Resource partitioning among four butterflyfish species in the Red Sea. Mar Freshw Res 53: Zemke-White WL, Choat JH, Clements KD (2002) A re-evaluation of the diet feeding hypothesis for marine herbivorous fishes. Mar Biol 141: Zharikov Y, Skilleter GA (2004) Why do eastern curlews Numenius madagascariensis feed on prey that lowers intake rate before migration? J Avian Biol 35: Zoufal R, Taborsky M (1991) Fish foraging periodicity correlates with daily changes of diet quality. Mar Biol 108: