Marine Biology

, 163:77 | Cite as

Non-random feeding enhances the contribution of oceanic zooplankton to the diet of the planktivorous coral reef fish Dascyllus flavicaudus

  • Katharine M. Hanson
  • Emilie L. Schnarr
  • James J. Leichter
Original paper

Abstract

Combining gut content analysis and sampling of ambient zooplankton, we examine departures from random feeding in a planktivorous coral reef fish and explore the effects of apparent non-random feeding behavior on the relative contribution of oceanic versus reef-associated zooplankton to fish diet. The planktivorous damselfish Dascyllus flavicaudus appears to exhibit strong positive electivity for oceanic copepods including Candacia spp. and copepods from the families Oncaeidae and Corycaeidae and consistent negative electivity for cyclopoid copepods (Oithonidae). In total, prey taxa categorized as oceanic in origin contributed 10–76 % of total zooplankton biomass in fish guts. The summed contribution of oceanic prey taxa to fish diet was significantly higher than expected under a model of random feeding based on the availability of oceanic versus reef-associated prey as sampled by zooplankton net tows. The feeding behavior of D. flavicaudus appears to be visibility-selective rather than or in addition to size-selective, as electivity across prey taxa could not be explained by differences in prey size alone.

Introduction

For coral reefs surrounding Pacific islands and atolls, the productivity of adjacent oceanic waters has been identified as a primary factor driving the abundance and dynamics of top reef predators including sharks (Nadon et al. 2012), monk seals (Baker et al. 2007), and piscivorous fishes (Williams et al. 2015). These positive relationships between producers at the base of oceanic food webs and higher trophic level consumers on reefs highlight the potential for trophic linkages between offshore, oceanic primary productivity and local secondary productivity in coral reef ecosystems. The feeding behaviors of key trophic guilds, such as planktivorous fishes, may be important factors within the suite of physical and biological processes that influence trophic connectivity between oceanic and coral reef food webs.

Feeding by planktivorous fishes represents a trophic pathway through which oceanic resources may enter reef systems and subsequently become redistributed to multiple levels within the reef food web. Planktivorous fishes are ubiquitous components of coral reef ecosystems, accounting for up to 50 % of total fish biomass on reefs (Williams and Hatcher 1983; DeMartini et al. 2008) and yielding some of the highest rates of biomass productivity among coral reef fishes (Depczynski et al. 2007). Planktivorous fishes contribute to secondary production in reef consumers from multiple trophic levels, and these fishes are often prey for piscivorous fishes and invertebrate predators (Norris and Parrish 1988). Planktivore feces contribute to detrital food webs and are directly consumed by detritivorous reef fishes (Robertson 1982). Nutrients excreted by planktivorous fishes are taken up by sessile organisms including reef corals and anemones (Cleveland et al. 2010). Planktivorous fishes on reefs have been described as a ‘wall of mouths’ that intercepts oceanic zooplankton transported across the reef (Emery 1973; Hamner et al. 1988). The utilization of oceanic resources by coral reef planktivores thus provides a trophic link through which offshore, oceanic primary productivity may contribute to secondary productivity within the reef ecosystem.

The zooplankton assemblage available to coral reef planktivores includes pelagic, oceanic species that are transported into the reef system as well as species that are restricted to or originate from the reef (Hamner et al. 2007). The latter reef-associated plankton includes holopelagic copepods, demersal microcrustaceans, and the eggs and larvae of reef fishes and invertebrates. Early studies concluded that such autochthonous, reef-associated resources are sufficient to support reef consumers, including planktivores, and that reefs are essentially self-supporting communities (Sargent and Austin 1954) that receive no net gain from allochthonous, oceanic plankton (Odum and Odum 1955). The emerging understanding of coral reef ecology highlights the importance of both oceanic (allochthonous) and reef-derived (autochthonous) resources for reef productivity (Hamner et al. 2007; Wyatt et al. 2013). For consumers that link oceanic and reef food webs, such as planktivorous fishes, the relative importance of oceanic versus reef-associated food sources may be influenced both by the availability of those resources and by consumers’ feeding behaviors.

Feeding behaviors both of select species and communities of consumers can have cumulative, ecosystem-scale effects on nutrient cycling and food web dynamics (Schmitz et al. 2010). The ecological implications of selective feeding have been examined for a number of coral reef fish guilds. Selective feeding by corallivorous fishes can influence the distribution and abundance of reef corals (Cole et al. 2008) and may influence the progression of coral diseases (Chong-Seng et al. 2010). Piscivorous groupers have been shown to prey disproportionately on rare species, resulting in shifts in the composition of resident fish assemblages (Stier et al. 2014).

Selective feeding has also been documented for planktivorous larval stages of various coral reef fishes (Sampey et al. 2007; Llopiz and Cowen 2009) and the mechanisms of selective feeding have been examined for the nocturnal planktivorous reef fish Apogon annularis (Holzman and Genin 2005). In contrast to nocturnal planktivores that prey predominantly on reef-associated zooplankton, diurnal planktivorous fishes consume large quantities of oceanic zooplankton (Hamner et al. 1988). Diurnal planktivorous fishes may remove up to 60 % of the plankton transported onto a reef (Glynn 1973), and they represent a trophic link whereby oceanic nutrients enter the reef system (Hobson 1991). Thus, the flux of allochthonous, oceanic plankton into reef food webs is likely to be significantly larger during the day than at night (Hamner et al. 1988).

In the present study we investigate feeding in the diurnal, planktivorous coral reef damselfish Dascyllus flavicaudus sampled from the island of Moorea, French Polynesia. We quantify the taxonomic composition of zooplankton sampled from both fish guts and ambient zooplankton assemblages and combine these data (1) to test the null hypothesis that D. flavicaudus feeds randomly on the available zooplankton prey and (2) to quantify departures from random feeding for specific zooplankton prey types. We categorize zooplankton prey taxa observed in the fish diet as oceanic, reef-associated, or mixed origin on the basis of natural history and prior studies of the distribution of zooplankton taxa in the waters surrounding Moorea. We then explore how the observed fish feeding behavior influences the relative importance of oceanic versus reef-associated resources in the sampled diet of D. flavicaudus.

Materials and methods

Study location and species

Fish and zooplankton were collected from reefs surrounding Moorea, a high volcanic island in the Society Archipelago of French Polynesia (17°32′S, 149°50′W). Three distinct reef habitats (fringing reef, backreef, and forereef) form concentric bands around the island. The yellowtail damselfish D. flavicaudus is common to Moorea’s reefs and can be found in all three habitats. D. flavicaudus exhibits feeding behaviors typical of small-bodied, diurnal planktivorous reef fishes—during the day individuals hover over coral colonies to feed on zooplankton that they detect visually in the water column. D. flavicaudus is a particulate feeder that visually selects and pursues individual zooplankters from the water column (Lazzaro 1987) and then uses suction feeding to ingest prey (Frederich et al. 2008).

Prey use

To identify the zooplankton prey taxa consumed by D. flavicaudus in Moorea, we analyzed the gut contents of fish collected from eight sampling locations, each separated by a minimum of 400 m. We sampled from four locations on the forereef (ca. 10 m depth) and four locations on the shallow backreef (<3 m depth) (Fig. 1). At all locations, paired fish and zooplankton collections were conducted in the late afternoon (15:00–17:00 local time). At each of the eight sampling locations, five fish were collected by SCUBA divers using handheld microspears and then euthanized in chilled seawater (N = 40 total fish). Gut contents were removed from the pouch-like stomach. Gut content samples were rinsed over a 100-μm mesh screen with filtered seawater. The portions ≥100 μm were retained, preserved in 1.8 % buffered formaldehyde in seawater and later enumerated under a dissecting microscope.
Fig. 1

Map of the island of Moorea, French Polynesia (17°32′S, 149°50′W) indicating the location of sampling sites where paired fish and zooplankton collections were made. Black symbols denote forereef sampling sites; backreef sampling sites are shown in gray

Zooplankton prey items were classified into 15 groups of varying taxonomic specificity (Table 1). Prey types that occurred in guts of only a few individual fish (e.g., coral and polychaete larvae) were grouped with unidentifiable organisms and categorized as ‘other’ zooplankton prey. These ‘other’ organisms were included in the calculation of the percent contribution of prey to fish diet based on numbers of prey items as well as log-likelihood statistics and forage ratios, but were not included in biomass calculations (see below).
Table 1

Habitat associations and estimated biomass (μg C animal−1) of zooplankton prey groups comprising the diet of the reef fish D. flavicaudus

Prey group

Association

μg C animal−1

Oncaeidae

Oceanic

2.8

Corycaeidae

Oceanic

2.5

Candacia spp.

Oceanic

12.2

Pelagic Harpacticoida

Oceanic

0.8

Ostracoda

Reef

15.6

Benthic Harpacticoida

Reef

1.7

Cladocera

Reef

15.6

Fusiform fish eggs

Reef

17.5

Decapoda

Reef

17.9

Oithonidae

Mixed origin

0.5

Round fish eggs

Mixed origin

8.8

Other Calanoida

Mixed origin

6.5

Appendicularia

Mixed origin

3.5

Other Copepoda

Mixed origin

4.6

Copepoda nauplii

Mixed origin

0.07

Biomass estimates are based on published data for zooplankton taxa (see “Materials and methods” section)

The entire gut contents were enumerated for an initial subsample of fish (N = 18). Due to the large number of prey items in each fish gut (Fig. 2), the remaining gut content samples were split using a Folsom plankton splitter, and zooplankton were then counted from subsamples. Rare taxa (<100 individuals total) were counted from the entirety of the sample. Numerically dominant taxa (>100 individuals total) were enumerated from one or more aliquots of the gut sample until at least 100 individuals were counted. Subsample counts were multiplied by appropriate aliquot factors to yield an estimate of total abundance per fish gut.
Fig. 2

Histogram of the total number of prey items in gut contents of the reef fish D. flavicaudus (N = 38 fish sampled across 8 study sites)

Prey availability

The ambient zooplankton assemblage was sampled at each study location by a SCUBA diver swimming with a plankton net (153-μm mesh net and cod end, 30 cm diameter) at a height of 1 m above the reef. At each location, a zooplankton tow was conducted immediately following and within 10–20 m of the location of fish collection. The volume of water sampled was measured using a mechanical flow meter outfitted with a low-velocity rotor (model 1030R6 General Oceanics Inc., Miami, FL, USA). The water volume sampled by zooplankton net tows averaged 9.2 m3 ± SD 2.4 (N = 8 total tows). Environmental zooplankton samples were fixed in a 1.8 % buffered formaldehyde and seawater solution and later enumerated under a dissecting microscope. All environmental zooplankton samples were split prior to enumeration (as above).

Examining random feeding

Log-likelihood goodness-of-fit tests (Manly et al. 2002) were used to evaluate two null hypotheses: (1) The occurrence of zooplankton prey taxa in fish gut contents did not vary among individual fish; (2) D. flavicaudus fed randomly on the available zooplankton such that the relative abundance of zooplankton taxa in fish gut contents did not differ from their relative abundance in the environment. Both hypotheses were tested for each of the eight collection locations. The first hypothesis was evaluated by calculating the log-likelihood statistic XL12 as:
$$X_{{{\text{L}}1^{2} }} = 2\sum\limits_{j = 1}^{n} {\sum\limits_{i = 1}^{I} {u_{ij} \ln\left[ {\frac{{u_{ij} }}{{E(u_{ij} )}}} \right]} }$$
for i = 1 to I prey taxa and j = 1 to n fish. Here uij is the observed count of prey group i in the stomach of fish j. E(uij) is the expected number of prey i in the gut of the jth individual assuming that the jth individual feeds in the same manner as the other fish sampled. E(uij) is calculated as:
$$E(u_{ij} ) = \left[ {\frac{{u_{i + } u_{ + j} }}{{u_{ + + } }}} \right]$$
where ui+ is the sum of individuals of prey group i across n fish, u+j is the sum of all prey types for fish j, and u++ is the grand sum of all prey enumerated in n gut contents samples (notation follows Manly et al. 2002). The significance of the resulting test statistic (XL12) was evaluated against the critical value from the Chi-square distribution with (I − 1) (N − 1) degrees of freedom and a Bonferroni-corrected significance level. To evaluate the hypothesis that D. flavicaudus exhibits random feeding on the available prey assemblages, a second log-likelihood statistic (XL22) was calculated, this time including counts from the zooplankton tow sample as if this environmental sample was another individual fish gut content sample (yielding N + 1 total observations). The difference between these two statistics:
$$D = X_{{{\text{L}}2^{2} }} - X_{{{\text{L}}1^{2} }}$$
is distributed as a Chi-square random variable with DFL2 − DFL1 degrees of freedom. A calculated value of D that exceeds the Chi-square critical value indicates non-random feeding or apparent selectivity (Manly et al. 2002).

Electivity as a metric of apparent non-random feeding

Electivity is defined as the proportion of distinct food types in a consumer’s diet relative to the proportion of available food types in the environment. We quantified the electivity exhibited by D. flavicaudus for each zooplankton group by calculating a forage ratio (wi) that compares the proportion of prey group i in fish gut contents to the proportion of group i in the environmental sample. D. flavicaudus often live in large groups on single coral heads. We intentionally collected fish from separate coral heads (distanced by >2 m) in each location to minimize the possibility of feeding interference among individuals collected within a sampling location. The corresponding zooplankton tows are used as estimates of prey availability at each collection location. A forage ratio (wi) was calculated for each zooplankton prey group at each of the eight reef locations as:
$$w_{i} = \left( {\frac{{u_{i + } }}{{u_{ + + } }}} \right)\left( {\pi_{i} } \right)^{ - 1}$$
where ui+ is the sum of counts of prey group i across n fish and u++ is the grand sum of all prey taxa enumerated in n gut content samples (Manly et al. 2002). πi is the proportion of prey type i within the environmental zooplankton sample. Confidence intervals (95 %) for wi were calculated according to Manly et al. (2002) and adjusted using a Bonferroni correction for the total number of prey taxa. The expected value of wi under random feeding is 1, indicating that fish feed on zooplankton prey taxa in the same proportion as their availability in the environment. Zooplankton taxa with forage ratios (wi ± CI) that exceed 1 are associated with positive electivity, while taxa with forage ratios (wi ± CI) < 1 are associated with negative electivity. Forage ratios (wi ± CI) that overlap 1 indicate that feeding on prey group i is indistinguishable from random feeding.

Prey size

To evaluate size-selective feeding in the sampled D. flavicaudus, body sizes of prey from fish gut contents were measured. Gut samples were imaged using the ZooScan digital imaging system (Gorsky et al. 2010). Individual zooplankton (N = 25–100) from each prey group were randomly selected from the resulting images and measured (prosome length or maximum length, μm) using ImageJ (NIH, v. 1.46). Spearman’s rank correlation was used to examine the relationship between median body size and median forage ratio across prey taxa (‘cor.test’ in the statistical computing environment R, version 2.14.1). Body size ranges of Oithonidae and copepod nauplii, prey taxa that were rare in the observed fish diet and absent from the imaged portions of the gut samples, were obtained from the literature, and for these two taxa midpoints of the reported size ranges were used in the place of median measured values. Midpoints were also used in lieu of median values for prey taxa that were rare in the imaged portions of gut samples (Cladocera and Ostracoda, N = 2 total images per taxa).

Influence of feeding electivity on fish diet

To estimate the contribution of oceanic and reef-associated zooplankton to total fish diet, we used prior published studies to distinguish between zooplankton prey taxa that are predominately oceanic in origin and those that are reef-associated (Table 1). The zooplankton taxa classified as oceanic are those known to have a pelagic, oceanic life history or primarily oceanic distribution, namely the pelagic Harpacticoida (here primarily Macrosetella gracilis) (O’Neil 1998) and copepods from the genus Candacia (Grice 1963). Two other copepod taxa were classified as oceanic—the poecilostomatoids Oncaeidae and Corycaeidae. The mean abundance of these two taxa sampled near Moorea’s reefs fell from 53 and 76 animals m−3, respectively, at a sampling station 1 km seaward of the reef tract to <10 animals m−3 at all sampling locations within the reef system (Lefevre 1986; Renon 1989). Similarly, the taxa classified as reef-associated are the eggs and larvae of reef organisms (here fusiform fish eggs and Decapoda) and taxa abundant within but not outside the reef system (Ostracoda, Cladocera, benthic Harpacticoida). The remaining taxa could not definitively be classified as oceanic or reef-associated on the basis of life history or distribution and were considered mixed origin (Table 1).

Count data from zooplankton tow and fish gut samples were transformed to biomass using estimates of carbon content (μg C animal−1) for each prey group (Table 1). Biomass estimates for similar taxa were obtained from the literature and adjusted for discrepancies in the body size of animals sampled from fish guts in Moorea relative to the size of the reference organisms (Yúfera et al. 1999; Satapoomin 1999; Logerwell and Ohman 1999; Baguley et al. 2004; Alldredge and King 2009; Almeda et al. 2010). The environmental zooplankton samples from each collection location were used to calculate expected dietary evenness and percent contribution of oceanic and reef-associated taxa to fish diet under random feeding. These expected values were compared with values calculated using the observed proportions in the gut contents. Prey counts were summed across all fish within a collection location, resulting in one pair of expected and observed measurements for each of the eight sampling locations. Paired t tests were used to evaluate the null hypothesis of no mean difference between observed and expected values for evenness and for the percent contribution of oceanic and reef-associated prey to total prey biomass in fish gut samples. The differences between paired means were examined for departures from normality using Shapiro–Wilk tests. Statistical analyses were performed using the ‘shapiro.test’ and ‘t.test’ functions in R (v.2.14.1).

A modified version (ED) of Simpson’s diversity index (DS) was used to assess the taxonomic breadth of fish diet:
$$D_{\text{S}} = \left( {\sum\limits_{i = 1}^{n} {p_{i}^{2} } } \right)^{ - 1}$$
where pi is the proportion of prey type i in the diet of a consumer. This index (DS) can be expressed as an evenness index (ED) where values of DS for each sample are scaled to the maximum possible value of DS according to:
$$E_{\text{D}} = \frac{{D_{\text{S}} }}{{D_{{{\text{S}}\,\hbox{max} }} }} = \left( {\sum\limits_{i = 1}^{n} {p_{i}^{2} } } \right)^{ - 1} \times S^{ - 1}$$
Here, S is the total number of species or prey types. In our samples, the number of possible prey types remains constant (S = 15) across sampling locations, as defined by the taxonomic resolution of the gut contents analysis. The contribution of oceanic and reef-associated zooplankton to total fish diet was calculated by summing the percent contribution of all oceanic and then all reef-associated taxa to the total zooplankton biomass for both environmental (expected) and fish gut contents (observed) samples. The biomass of ‘mixed-origin’ prey was excluded from these calculations, and thus the combined oceanic and reef contributions are less than the total estimated prey biomass.

Results

Prey use

Dascyllus flavicaudus feeds in the water column during daylight hours. In the present study, fish were collected in the late afternoon (15:00–17:00 local time). The sampled fish guts contained hundreds of zooplankton prey items (median = 871 prey items per fish, Fig. 2).

Four zooplankton prey taxa were numerically dominant in fish gut samples: copepods from the families Oncaeidae and Corycaeidae, appendicularians, and calanoid copepods (Fig. 3). Together these four taxa comprised 50–97 % of total prey items sampled from individual fish. The remaining taxa typically comprised <20 % of gut contents with some taxa contributing <1 % of total prey enumerated. The four dominant zooplankton taxa (Oncaeidae, Corycaeidae, appendicularians and calanoid copepods) were present in all fish gut contents sampled. Some non-dominant prey taxa (fish eggs, pelagic harpacticoid copepods, Oithonidae, Candacia spp., other copepods) were found in over 60 % of all sampled fish guts, while other taxa were both non-dominant and infrequently encountered, occurring in <30 % of gut samples (ostracods, copepod nauplii). Two taxa, benthic harpacticoid copepods and cladocerans, were rare or absent from guts of forereef fish but were found in >90 % of guts sampled from backreef fish.
Fig. 3

The percent contribution of zooplankton groups to total fish gut contents (in black, N = 20 fish per habitat) and to total zooplankton collected in net tows (in gray, N = 4 tows per habitat) for samples collected in the backreef (a) and forereef (b) habitats. Vertical lines indicate median values, boxes extend from 25th to 75th percentile, whiskers to 1.5 times the interquartile range, symbols indicate samples that fall beyond this range. Prey groups are arranged in order of descending median percent contribution to fish gut contents

Prey availability

The composition of zooplankton taxa in fish gut contents did not mirror the proportional contribution of zooplankton taxa in the environmental samples (Fig. 3). Three of the zooplankton taxa that were numerically dominant in gut samples (Oncaeidae, Corycaeidae and appendicularians) each accounted for <20 % of zooplankton found in environmental samples. In contrast, calanoid copepods and Oithonidae dominated the environmental samples. Together these two taxa comprised 35–78 % of the total zooplankton in environmental tows, while their combined contribution to fish gut contents ranged from as low as 1 % to a maximum of 22 % of total prey items. The patterns of abundance among zooplankton taxa in both gut and environmental samples and the discrepancies between gut and environment were parallel in samples collected from the forereef (Fig. 3a) and those collected from backreef habitats (Fig. 3b). Copepods from the families Corycaeidae and Oncaeidae and appendicularians were the most abundant prey items in guts of both forereef and backreef fish, and in both reef habitats these taxa were overrepresented in fish diet relative to their environmental abundance. Calanoid copepods dominated the environmental zooplankton samples collected in both reef habitats, and these taxa were relatively underrepresented in gut samples from both forereef and backreef fish.

Non-random feeding in D. flavicaudus

At each of the eight collection locations, the relative contribution of zooplankton prey types to fish gut contents varied significantly among individual fish, evidenced by log-likelihood statistics (XL12) that were two- to tenfold greater than critical values (Table 2). With the inclusion of the environmental samples, the dissimilarity among samples increased, further inflating the second log-likelihood statistics (XL22). At each reef location, the difference between these two statistics (D) was large, reflecting the disparity between fish gut contents and environmental zooplankton samples and indicating non-random feeding by D. flavicaudus.
Table 2

Goodness-of-fit tests of random feeding in the reef fish D. flavicaudus

Collection location

Gut contents

Gut contents + environmental abundance

D = XL12 − XL22

df

XL12

\(\chi_{\text{crit}}^{2}\)

df

XL22

\(\chi_{\text{crit}}^{2}\)

df

D

\(\chi_{\text{crit}}^{2}\)

Forereef A

56

212

86

70

6934

103

14

6722

31

Forereef B

56

512

86

70

2400

103

14

1887

31

Forereef C

56

180

86

70

4215

103

14

4036

31

Forereef D

56

246

86

70

646

103

14

400

31

Backreef A

56

952

86

70

6270

103

14

5318

31

Backreef B

42

165

69

56

3250

86

14

3085

31

Backreef C

56

396

86

70

4456

103

14

4060

31

Backreef D

42

618

69

56

4463

86

14

3846

31

Statistics are calculated for paired fish and zooplankton collections for each reef location based solely on fish gut contents (XLI2) and including environmental zooplankton abundance (XL22). D is the difference between the models. Critical χ2 values are based on the degrees of freedom (df) for each test (N = 4 or 5 gut content samples, N = 1 environmental sample, and N = 15 prey taxa per location) and α = 0.006 (Bonferroni correction for N = 8 simultaneous tests)

Feeding electivity

Apparent non-random feeding by D. flavicaudus resulted in both over- and underrepresentation of zooplankton taxa in fish gut contents relative to the composition in the ambient zooplankton assemblage (Fig. 4). Fish exhibited strong positive electivity for copepods from the family Oncaeidae. At all eight locations sampled, the forage ratio (wi ± 95 % CI) calculated for this group was greater than 1 (Fig. 4a). Two other prey taxa had significant positive electivity scores in the majority of sampling locations: copepods from the family Corycaeidae (positive scores in 7 of 8 locations) and Candacia spp. (positive scores in 6 locations). Forage ratios calculated for fusiform fish eggs and appendicularians were split between positive electivity scores and neutral scores (Fig. 4b). Forage ratios for Oithonidae, calanoid copepods (excluding Candacia spp.), and ostracods indicated consistent negative electivity for these taxa across all sampled locations (Fig. 4c). Both Oithonidae and calanoid copepods were present in large numbers in each environmental sample, but accounted for small proportions of the total prey items in fish guts (Fig. 3). Evidence for negative electivity was also found for benthic harpacticoid copepods and copepod nauplii (Fig. 4d). The forage ratios for the remaining five prey taxa varied among reef locations with no consistent evidence for positive or negative electivity (Fig. 4e). Within prey taxa, electivity scores did not vary with reef habitat between forereef and backreef sampling locations.
Fig. 4

ae Forage ratios (wi ± 95 % confidence intervals) for fish feeding on 15 zooplankton prey groups. Bars span the 95 % CI for forage ratios calculated at each of eight locations from paired fish gut and zooplankton samples. The expected value of wi under random feeding is 1 (dashed line, note log scale). Blue lines indicate positive electivity where (wi ± CI) > 1; red lines indicate negative electivity where (wi ± CI) < 1. Black lines indicate wi was indistinguishable from random feeding (no electivity). fj Body sizes of prey collected from fish gut samples. Zooplankton groups are as in ae. Vertical lines indicate median values; boxes extend from 25th to 75th percentile; whiskers to 1.5 times the interquartile range

Prey size

There was no correlation between prey body size and forage ratio (Fig. 4f–j; Spearman rank correlation, rs = 0.036, N = 15, P = 0.903). The taxa associated with consistent positive or negative electivity included both small-bodied and large-bodied prey. Fish demonstrated positive electivity for both small (Oncaeidae, Corycaeidae) and large-bodied copepods (Candacia spp.). The calanoid copepods for which D. flavicaudus exhibited negative electivity were some of the largest copepod prey measured (Fig. 5). Both within the copepod taxa and across all prey taxa, there was no evidence that size-based selection alone drove the electivity measured for D. flavicaudus.
Fig. 5

Empirical cumulative distribution functions of prosome lengths measured for four copepod taxa sampled from guts of the reef fish D. flavicaudus (N = 25–100 individuals per taxon in each reef habitat). The electivity for each prey group (+ or −) is indicated. a Backreef, b forereef

Influence of feeding electivity on fish diet

Across sampling locations, environmental samples exhibited evenness values that ranged from 0.113 to 0.375 (max. possible ED = 1) and evenness scores observed in fish gut samples ranged from 0.180 to 0.438. Non-random feeding in D. flavicaudus led to a significant difference between the expected (based on environmental samples) and observed taxonomic evenness of fish diet (Fig. 6a, paired t test, t7 = 2.914, P = 0.022). In all but one sampling location, prey items were more evenly distributed among zooplankton prey taxa in fish gut contents samples than in environmental zooplankton samples. The discrepancy in evenness between fish gut and environmental zooplankton samples was greater at backreef sampling locations than at forereef locations (Fig. 6a).
Fig. 6

a Distribution of prey items among all prey groups (as evenness, maximum possible value = 1) in the diet of the reef fish D. flavicaudus. b, c The percent contribution (as biomass, μg C) of oceanic (b) and reef-associated (c) prey groups to the diet of D. flavicaudus. Gray bars represent expected values under a model of random feeding, calculated using environmental zooplankton tow data. Black bars indicate the observed values from fish gut contents

The oceanic zooplankton contributing to fish diet were copepods from the taxa Oncaeidae, Corycaeidae, Candacia spp., and pelagic Harpacticoida (Table 1). The combined carbon biomass of these oceanic zooplankton taxa comprised a significantly larger proportion of the total carbon biomass in fish gut contents than expected from their abundance in environmental samples (Fig. 6b, paired t test, t7 = 5.669, P < 0.001). The oceanic contribution to ambient zooplankton assemblages was low (mean 10.8 % ± 4.2 SD). In contrast, oceanic taxa contributed up to 76 % of the zooplankton biomass in fish gut contents samples (mean 48.0 % ± 19.1 SD). Non-random feeding by D. flavicaudus resulted in as much as a sevenfold increase in the contribution of oceanic taxa to the total zooplankton biomass in fish diet beyond that expected under random feeding (Fig. 6b). The discrepancy between observed and expected percent contribution of oceanic prey to fish diet was driven largely by the positive electivity for the oceanic copepod taxa Oncaeidae and Corycaeidae (Fig. 4). In contrast, there was no significant difference between the observed and expected contribution of reef-associated zooplankton to fish diet across sampling locations (Fig. 6c, paired t test, t7 = 0.8792, P = 0.409). The observed contribution of reef-associated prey ranged from 2 to 32 %, exceeding expected values at some sampling locations and falling below expected values at others. The differences across sampling locations were driven by interacting negative electivity for Ostracoda and positive electivity for Cladocera and fusiform fish eggs.

Discussion

The diet of D. flavicaudus sampled from the reefs of Moorea, French Polynesia, was dominated by calanoid copepods and poecilostomatoid copepods from the families Oncaeidae and Corycaeidae, and appendicularians. Fish feeding on three copepod taxa (Oncaeidae, Corycaeidae and Candacia spp.) was associated with strong positive electivity scores; by count these taxa comprised proportions of fish gut contents up to 50 times greater than their fractional contributions to environmental samples. Conversely, D. flavicaudus fed on cyclopoid copepods (Oithonidae) and calanoid copepods (excluding Candacia spp.) at much lower proportions than expected from the relative abundance of these taxa in the environmental samples, and thus these two prey taxa were associated with strong negative electivity. The patterns of dominance among prey taxa in fish gut samples were consistent among fish collected from both forereef and backreef habitats. The proportional contributions of prey taxa to environmental samples were also similar in both habitats. It is possible that the diet of D. flavicaudus and patterns of apparent prey selectivity could change under conditions where the composition of available zooplankton prey departs from the conditions we sampled among reef habitats. Such fluctuations in diet with changes in the environmental availability of zooplankton prey have been observed in the planktivorous damselfish Chromis dispilus sampled from temperate reefs in New Zealand (Kingsford and MacDiarmid 1988). In coral reef ecosystems, numerous fish species alter their foraging behaviors and diet to take advantage of the episodic availability of coral spawn and feed preferentially on this lipid- and energy-rich food source (Westneat and Resing 1988; Pratchett et al. 2001).

The patterns of electivity that we document here for D. flavicaudus provide evidence for apparent feeding selectivity by this fish species, but do not imply active feeding preferences by this consumer. Similar patterns of electivity can arise in the absence of active choice as a function of differential encounter rates among available zooplankton prey, driven by the detectability of prey taxa (Holzman and Genin 2005). The vulnerability of prey to a given predator is influenced by both encounter rates and by the capture success of the predator (Pastorok 1981). The abilities of prey, such as the powerful escape jumps exhibited by calanoid copepods (Kiørboe et al. 2010), can influence capture success rates and lead to apparent negative selectivity (Drenner et al. 1978). Prey defenses can also influence apparent prey selectivity. Some tropical cyprinid ostracods produce bioluminescent extracellular secretions (Shimomura et al. 1969). For example, the ostracod Photeros annecohenae emits bursts of blue light as a defense mechanism against predation by nocturnal planktivorous coral reef fishes. It has been observed that diurnal planktivorous fishes reject P. annecohenae after capture, which suggests that these ostracods may also be unpalatable to or chemically defended against fish predators (Rivers and Morin 2012). It is possible that similar dynamics influenced the negative selectivity that D. flavicaudus appeared to exhibit for ostracods in the present study. Conclusive determination of the mechanisms that drive apparent selectivity could be facilitated by controlled field or laboratory experiments in which prey abundance and the distribution of prey characteristics can be manipulated (Hessen 1985; Holzman and Genin 2005); the present field study provides evidence for apparent non-random feeding in the natural environment.

It is also possible that planktivores including D. flavicaudus actively select prey based on the nutritional value of the available zooplankton. The nutritional quality of available foods types has been found to influence the diet of herbivorous and grazing reef fishes (Clements et al. 2009). As noted above, some planktivorous and omnivorous coral reef fishes have been observed to feed preferentially on lipid- and energy-rich coral eggs and larvae during mass spawning events on reefs (Westneat and Resing 1988; Pratchett et al. 2001). In the present study, D. flavicaudus exhibited positive electivity for the fusiform fish eggs characteristic of the parrotfish genera Chlorurus and Scarus. This apparent selectivity may have been driven by the nutritional quality of scarine fish eggs. None of the copepod taxa found in the environmental or the fish gut samples are lipid-rich taxa (Lee et al. 2006). Calanoid copepods from the genus Calanus are known to have relatively higher lipid content among the generally lipid-poor tropical epipelagic copepods (Lee and Hirota 1973). However, in the present study fish exhibited consistent negative electivity for calanoid copepods with the exception of Candacia spp.

The patterns of non-random feeding that we document for D. flavicaudus are based on the assumption that the environmental net tows adequately sampled the zooplankton prey field available to the fish. The most likely biases of the net sampling technique are under sampling of small-bodied zooplankton and of zooplankton that are fast-swimming or otherwise able to escape from the nets. However, predation by planktivorous fish is also limited by prey size and prey escape ability. The smallest bodied prey types encountered in fish gut samples were copepod nauplii and oncaeid copepods (Fig. 4f, i). Copepod nauplii were more abundant in environmental zooplankton samples than in fish gut samples (Fig. 3). Thus, correction for any methodological underestimate of naupliar abundance in the environment would have strengthened the calculated negative electivity that the fish exhibited for copepod nauplii. Underestimates of the environmental abundance of small-bodied oncaeid copepods could have biased the strong positive electivity fish exhibited for this group. However, the minimum prosome length of oncaeid copepods sampled from fish guts was 350 μm (Fig. 5), and the environmental samples were collected with 153-μm mesh nets. Thus it is unlikely that the plankton nets systematically underestimated the availability of copepod prey including the small-bodied Oncaea spp. It is possible that our sampling failed to detect fragile food items such as gelatinous zooplankton or appendicularian houses that may have been destroyed by the net tow collections or digested more rapidly within the fish guts than crustacean prey items.

In planktivorous fish, the reaction distance and the probability of prey detection increase with increasing zooplankton prey size and with increasing zooplankton pigmentation (Utne-Palm 1999). In the present study, prey size alone did not explain the differences in electivity among the copepod taxa that dominated the diet of D. flavicaudus. Fish exhibited strong positive electivity for the relatively small-bodied oncaeid copepods (350–800 μm) and negative electivity for the larger-bodied calanoid copepods (500–1750 μm). The copepod prey taxa for which D. flavicaudus consistently exhibited positive electivity (Candacia spp., Corycaeidae, Oncaeidae) contain multiple species with pigmented bodies or structures (Fig. 7a–c). It is possible that such pigmentation enhanced the visibility of these prey types to D. flavicaudus and that this species exhibits visibility-selective feeding rather than or in addition to size-selective feeding as has been shown for other visual-feeding planktivorous fishes (Zaret and Kerfoot 1975). In addition, field observations by Alldredge (1972) in the Caribbean documented that up to 33 % of discarded appendicularian houses were occupied by the poecilostomatoid copepod Oncaea mediterranea—often 1–5 individuals per house. This type of association among oncaeid copepods and mucous structures could increase the visibility of the associated copepods to particulate feeders like D. flavicaudus.
Fig. 7

Dissecting microscope photographs of individuals corresponding to the four dominant copepod prey groups sampled from guts of the reef fish D. flavicaudus. Copepod taxa and electivity (+ or −) are: a Oncaeidae (+), b Corycaeidae (+), cCandacia spp. (+). and d other Calanoida (−). Individuals imaged were retrieved from fish gut content samples

Swimming in oithonid copepods is characterized by long periods where individuals remain motionless in the water column, punctuated by intermittent leaps (Hwang and Turner 1995). On the basis of swimming behaviors as well as coloration, Paffenhöfer (1993) predicted that the relatively transparent and hydrodynamically quiescent oithonid copepods should be less vulnerable to vertebrate predators than the pigmented oncaeid copepods which exhibit punctuated, jerky swimming motions (Hwang and Turner 1995). In the present study, fish exhibited strong negative electivity for Oithonidae and strong positive electivity for Oncaeidae.

Patterns of negative electivity for oithonid copepods, similar to those we observed in the damselfish D. flavicaudus, have also been documented for invertebrate planktivores on coral reefs. A study of the scleractinian coral Meandrina meandrites on reefs in St. Croix found that Oithona sp. contributed only 18.5 % of the prey items sampled from coral coelentera, while this copepod accounted for 75.1 % of individuals enumerated from zooplankton tows (Johnson and Sebens 1993). A study of Jamaican reefs found negative selectivity for Oithona sp. by the corals Madracis mirabilis and Montastrea cavernosa (Sebens et al. 1996). The apparent non-random feeding evidenced in prior studies for corals and in the present study for the damselfish D. flavicaudus suggests that the importance of zooplankton taxa to reef planktivores cannot be predicted on the basis of environmental abundance alone. Such patterns of prey use (e.g., apparent negative selectivity for Oithonidae) exhibited by multiple types of planktivores (both tactile and visual feeders) could have substantial cumulative influence on the contribution of zooplankton types to the larger reef food web. Studies of the feeding behaviors of key trophic guilds, including planktivorous fishes, contribute to the evolving understanding of the trophic connectivity between oceanic and coral reef food webs.

Notes

Acknowledgments

We thank M. D. Ohman for providing access to ZooScan equipment and for guidance on the design and analysis of the study. We thank A. Stier, L. Bentley, D. Combosch, W. Boudreau, K. Willis, and L. Sala for their contributions to collections, imaging and image analysis. Comments from M. D. Ohman, L. Levin, L. Aluwihare, and I. Abramson improved the manuscript and analyses. This research was supported by awards from the National Science Foundation through the Moorea Coral Reef LTER program and NSF award OCE-0927448 to JJL, a Northeastern University Three Seas Teaching Fellowship to KMH, and an NSF REU fellowship to ELS via the California Current Ecosystem LTER. Fish collection was approved by French Polynesian research permits and the UCSD Institutional Animal Care and Use Committee (Protocol # S06369).

References

  1. Alldredge AL (1972) Abandoned larvacean houses: a unique food source in the pelagic environment. Science 177:885–887. doi:10.1126/science.177.4052.885 CrossRefGoogle Scholar
  2. Alldredge AL, King JM (2009) Near-surface enrichment of zooplankton over a shallow back reef: implications for coral reef food webs. Coral Reefs 28:895–908. doi:10.1007/s00338-009-0534-4 CrossRefGoogle Scholar
  3. Almeda R, Calbet A, Alcaraz M et al (2010) Effects of temperature and food concentration on the survival, development and growth rates of naupliar stages of Oithona davisae (Copepoda, Cyclopoida). Mar Ecol Prog Ser 410:97–109. doi:10.3354/meps08625 CrossRefGoogle Scholar
  4. Baguley JG, Hyde LJ, Montagna PA (2004) A semi-automated digital microphotographic approach to measure meiofaunal biomass. Limnol Oceanogr Methods 2:181–190CrossRefGoogle Scholar
  5. Baker J, Polovina J, Howell E (2007) Effect of variable oceanic productivity on the survival of an upper trophic predator, the Hawaiian monk seal Monachus schauinslandi. Mar Ecol Prog Ser 346:277–283. doi:10.3354/meps06968 CrossRefGoogle Scholar
  6. Chong-Seng KM, Cole AJ, Pratchett MS, Willis BL (2010) Selective feeding by coral reef fishes on coral lesions associated with brown band and black band disease. Coral Reefs 30:473–481. doi:10.1007/s00338-010-0707-1 CrossRefGoogle Scholar
  7. Clements KD, Raubenheimer D, Choat JH (2009) Nutritional ecology of marine herbivorous fishes: ten years on. Funct Ecol 23:79–92. doi:10.1111/j.1365-2435.2008.01524.x CrossRefGoogle Scholar
  8. Cleveland A, Verde EA, Lee RW (2010) Nutritional exchange in a tropical tripartite symbiosis: direct evidence for the transfer of nutrients from anemonefish to host anemone and zooxanthellae. Mar Biol 158:589–602. doi:10.1007/s00227-010-1583-5 CrossRefGoogle Scholar
  9. Cole AJ, Pratchett MS, Jones GP (2008) Diversity and functional importance of coral-feeding fishes on tropical coral reefs. Fish Fish 9:286–307. doi:10.1111/j.1467-2979.2008.00290.x CrossRefGoogle Scholar
  10. DeMartini EE, Friedlander AM, Sandin SA, Sala E (2008) Differences in fish-assemblage structure between fished and unfished atolls in the northern Line Islands, central Pacific. Mar Ecol Prog Ser 365:199–215. doi:10.3354/meps07501 CrossRefGoogle Scholar
  11. Depczynski M, Fulton CJ, Marnane MJ, Bellwood DR (2007) Life history patterns shape energy allocation among fishes on coral reefs. Oecologia 153:111–120. doi:10.1007/s00442-007-0714-2 CrossRefGoogle Scholar
  12. Drenner RW, Strickler JR, O’Brien WJ (1978) Capture probability: the role of zooplankter escape in the selective feeding of planktivorus fish. J Fish Res Board Canada 35:1370–1373CrossRefGoogle Scholar
  13. Emery AR (1973) Comparative ecology and functional oseteology of fourteen species of damselfish (Pisces: Pomacentridae) at Alligator Reef, Florida Keys. Bull Mar Sci 23:649–770Google Scholar
  14. Frederich B, Pilet A, Parmentier E, Vandewalle P (2008) Comparative trophic morphology in eight species of damselfishes (Pomacentridae). J Morphol 269:175–188. doi:10.1002/jmor CrossRefGoogle Scholar
  15. Glynn PW (1973) Ecology of a Caribbean coral reef. The Porites reef-flat biotope: part II. Plankton community with evidence for depletion. Mar Biol 22:1–21. doi:10.1007/BF00388905 CrossRefGoogle Scholar
  16. Gorsky G, Ohman MD, Picheral M et al (2010) Digital zooplankton image analysis using the ZooScan integrated system. J Plankton Res 32:285–303. doi:10.1093/plankt/fbp124 CrossRefGoogle Scholar
  17. Grice G (1963) A revision of the genus Candacia (Copepoda: Calanoida) with an annotated list of the species and a key for their identification. Zool Meded 38:171–194Google Scholar
  18. Hamner WM, Jones MS, Carleton JH et al (1988) Zooplankton, planktivorous fish, and water currents on a windward reef face: Great Barrier Reef, Australia. Bull Mar Sci 42:459–479Google Scholar
  19. Hamner W, Colin P, Hamner P (2007) Export-import dynamics of zooplankton on a coral reef in Palau. Mar Ecol Prog Ser 334:83–92. doi:10.3354/meps334083 CrossRefGoogle Scholar
  20. Hessen DO (1985) Selective zooplankton predation by pre-adult roach (Rutilus rutilus): the size-selective hypothesis versus the visibility-selective hypothesis. Hydrobiologia 124:73–79CrossRefGoogle Scholar
  21. Hobson ES (1991) Trophic relationships of fishes specialized to feed on zooplankters above coral reefs. In: Sale P (ed) The ecology of fishes on coral reefs. Academic Press, San Diego, CA, pp 69–93CrossRefGoogle Scholar
  22. Holzman R, Genin A (2005) Mechanisms of selectivity in a nocturnal fish: a lack of active prey choice. Oecologia 146:329–336. doi:10.1007/s00442-005-0205-2 CrossRefGoogle Scholar
  23. Hwang J, Turner J (1995) Behaviour of cyclopoid, harpacticoid, and calanoid copepods from coastal waters of Taiwan. Mar Ecol 16:207–216CrossRefGoogle Scholar
  24. Johnson AS, Sebens KP (1993) Consequences of a flattened morphology: effects of flow on feeding rates of the scleractinian coral Meandrina meandrites. Mar Ecol Prog Ser 99:99–114CrossRefGoogle Scholar
  25. Kingsford MJ, MacDiarmid AB (1988) Interrelations between planktivorous reef fish and zooplankton in temperate waters. Mar Ecol Prog Ser 48:103–117CrossRefGoogle Scholar
  26. Kiørboe T, Andersen A, Langlois VJ, Jakobsen HH (2010) Unsteady motion: escape jumps in planktonic copepods, their kinematics and energetics. J R Soc Interface 7:1591–1602. doi:10.1098/rsif.2010.0176 CrossRefGoogle Scholar
  27. Lazzaro X (1987) A review of planktivorous fishes—their evolution, feeding, behaviors, selectivities and impacts. Hydrobiologia 146:97–167CrossRefGoogle Scholar
  28. Lee RF, Hirota J (1973) Wax esters in tropical zooplankton and nekton and the geographical distribution of wax esters in marine copepods. Limnol Oceanogr 18:227–239CrossRefGoogle Scholar
  29. Lee RF, Hagen W, Kattner G (2006) Lipid storage in marine zooplankton. Mar Ecol Prog Ser 307:273–306CrossRefGoogle Scholar
  30. Lefevre M (1986) Variations spatio-temporelles du peuplement zooplanctonique du lagon de l’isle de Moorea (Archipel de La Societe, Polyneise Francaise). l’Universite Pierre et Marie Curie, ParisGoogle Scholar
  31. Llopiz J, Cowen R (2009) Variability in the trophic role of coral reef fish larvae in the oceanic plankton. Mar Ecol Prog Ser 381:259–272. doi:10.3354/meps07957 CrossRefGoogle Scholar
  32. Logerwell EA, Ohman MD (1999) Egg-brooding, body size and predation risk in planktonic marine copepods. Oecologia 121:426–431. doi:10.1007/s004420050948 CrossRefGoogle Scholar
  33. Manly BFJ, McDonald LL, Thomas DL et al (2002) Resource selection by animals: statistical design and analysis for field studies, 2nd edn. Springer, BerlinGoogle Scholar
  34. Nadon MO, Baum JK, Williams ID et al (2012) Re-creating missing population baselines for Pacific reef sharks. Conserv Biol 26:493–503. doi:10.1111/j.1523-1739.2012.01835.x CrossRefGoogle Scholar
  35. Norris JE, Parrish JD (1988) Predator-prey relationships among fishes in pristine coral reef communities. In: Proceedings of the 6th international coral reef symposium, vol 2, Australia, pp 107–113Google Scholar
  36. O’Neil JM (1998) The colonial cyanobacterium Trichodesmium as a physical and nutritional substrate for the harpacticoid copepod Macrosetella gracilis. J Plankton Res 20:43–59. doi:10.1093/plankt/20.1.43 CrossRefGoogle Scholar
  37. Odum AJ, Odum EP (1955) Trophic structure and productivity of a windward coral reef community on Eniwitok Atoll. Ecol Monogr 25:291–320CrossRefGoogle Scholar
  38. Paffenhöfer GA (1993) On the ecology of marine cyclopoid copepods (Crustacea, Copepoda). J Plankton Res 15:37–55CrossRefGoogle Scholar
  39. Pastorok RA (1981) Prey vulnerability and size selection by Chaoborus larvae. Ecology 62:1311–1324CrossRefGoogle Scholar
  40. Pratchett M, Gust N, Goby G, Klanten SO (2001) Consumption of coral propagules represents a signicant trophic link between corals and reef fish. Coral Reefs 20:13–17CrossRefGoogle Scholar
  41. Renon J (1989) Le zooplancton des milieux reifo-lagonaires de Polynesie. l'Université d'Orléans, Orléans, FranceGoogle Scholar
  42. Rivers TJ, Morin JG (2012) The relative cost of using luminescence for sex and defense: light budgets in cypridinid ostracods. J Exp Biol 215:2860–2868. doi:10.1242/jeb.072017 CrossRefGoogle Scholar
  43. Robertson D (1982) Fish feces as fish food on a Pacific coral reef. Mar Ecol Prog Ser 7:253–265. doi:10.3354/meps007253 CrossRefGoogle Scholar
  44. Sampey A, McKinnon AD, Meekan MG, McCormick MI (2007) Glimpse into guts: overview of the feeding of larvae of tropical shorefishes. Mar Ecol Prog Ser 339:243–257. doi:10.3354/meps339243 CrossRefGoogle Scholar
  45. Sargent MC, Austin TS (1954) Biologic economy of coral reefs. Bikini and nearby atolls. Part 2. Oceanography (biologic). U.S. Geological Survey Professional Paper 260-B, pp 293–300Google Scholar
  46. Satapoomin S (1999) Carbon content of some common tropical Andaman Sea copepods. J Plankton Res 21:2117–2123. doi:10.1093/plankt/21.11.2117 CrossRefGoogle Scholar
  47. Schmitz OJ, Hawlena D, Trussell GC (2010) Predator control of ecosystem nutrient dynamics. Ecol Lett 13:1199–1209. doi:10.1111/j.1461-0248.2010.01511.x CrossRefGoogle Scholar
  48. Sebens KP, Vandersall KS, Savina LA, Graham KR (1996) Zooplankton capture by two scleractinian corals, Madracis mirabilis and Montastrea cavernosa, in a field enclosure. Mar Biol 127:303–317CrossRefGoogle Scholar
  49. Shimomura O, Johnson FH, Masugi T (1969) Cypridina bioluminescence: light-emitting oxyluciferin–luciferase complex. Science 164(80):1299–1300CrossRefGoogle Scholar
  50. Stier AC, Hanson KM, Holbrook SJ et al (2014) Predation and landscape characteristics independantly affect reef fish community organization. Ecology 95:1294–1307CrossRefGoogle Scholar
  51. Utne-Palm AC (1999) The effect of prey mobility, prey contrast, turbidity and spectral composition on the reaction distance of Gobiusculus flavescens to its planktonic prey. J Fish Biol 54:1244–1258. doi:10.1111/j.1095-8649.1999.tb02052.x CrossRefGoogle Scholar
  52. Westneat MW, Resing JM (1988) Predation on coral spawn by planktivorous fish. Coral Reefs 7:89–92CrossRefGoogle Scholar
  53. Williams DM, Hatcher AI (1983) Structure of fish communities on outer slopes of inshore, mid-shelf and outer shelf reefs of the Great Barrier Reef. Mar Ecol Prog Ser 10:239–250. doi:10.3354/meps010239 CrossRefGoogle Scholar
  54. Williams ID, Baum JK, Heenan A et al (2015) Human, oceanographic and habitat drivers of central and western Pacific coral reef fish assemblages. PLoS One 10:e0120516. doi:10.1371/journal.pone.0120516 CrossRefGoogle Scholar
  55. Wyatt ASJ, Lowe RJ, Humphries S, Waite AM (2013) Particulate nutrient fluxes over a fringing coral reef: source-sink dynamics inferred from carbon to nitrogen ratios and stable isotopes. Limnol Oceanogr 58:409–427. doi:10.4319/lo.2013.58.1.0409 CrossRefGoogle Scholar
  56. Yúfera M, Parra G, Santiago R, Carrascosa M (1999) Growth, carbon, nitrogen and caloric content of Solea senegalensis (Pisces: Soleidae) from egg fertilization to metamorphosis. Mar Biol 134:43–49CrossRefGoogle Scholar
  57. Zaret TM, Kerfoot C (1975) Fish predation on Bosmina longirostris: body-size selection versus visibility selection. Ecology 56:232–237CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  • Katharine M. Hanson
    • 1
    • 2
  • Emilie L. Schnarr
    • 3
    • 4
  • James J. Leichter
    • 1
  1. 1.Scripps Institution of OceanographyLa JollaUSA
  2. 2.Chapel HillUSA
  3. 3.Environmental Systems ProgramUniversity of California, San DiegoLa JollaUSA
  4. 4.Center for International Earth Science Information NetworkEarth Institute at Columbia UniversityPalisadesUSA

Personalised recommendations