Reef sharks may influence the foraging behaviour of mesopredatory teleosts on coral reefs via both risk effects and competitive exclusion. We used a “natural experiment” to test the hypothesis that the loss of sharks on coral reefs can influence the diet and body condition of mesopredatory fishes by comparing two remote, atoll-like reef systems, the Rowley Shoals and the Scott Reefs, in northwestern Australia. The Rowley Shoals are a marine reserve where sharks are abundant, whereas at the Scott Reefs numbers of sharks have been reduced by centuries of targeted fishing. On reefs where sharks were rare, the gut contents of five species of mesopredatory teleosts largely contained fish while on reefs with abundant sharks, the same mesopredatory species consumed a larger proportion of benthic invertebrates. These measures of diet were correlated with changes in body condition, such that the condition of mesopredatory teleosts was significantly poorer on reefs with higher shark abundance. Condition was defined as body weight, height and width for a given length and also estimated via several indices of condition. Due to the nature of natural experiments, alternative explanations cannot be discounted. However, the results were consistent with the hypothesis that loss of sharks may influence the diet and condition of mesopredators and by association, their fecundity and trophic role. Regardless of the mechanism (risk effects, competitive release, or other), our findings suggest that overfishing of sharks has the potential to trigger trophic cascades on coral reefs and that further declines in shark populations globally should be prevented to protect ecosystem health.
Citation: Barley SC, Meekan MG, Meeuwig JJ (2017) Diet and condition of mesopredators on coral reefs in relation to shark abundance. PLoS ONE 12(4): e0165113. https://doi.org/10.1371/journal.pone.0165113
Editor: Timothy Darren Clark, University of Tasmania, AUSTRALIA
Received: April 23, 2016; Accepted: January 11, 2017; Published: April 19, 2017
Copyright: © 2017 Barley et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: All funding came from the University of Western Australia, an IPRS scholarship and the Australian Institute of Marine Science. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
There is strong evidence that apex predators can structure food webs in ecosystems by consuming prey, also known as “lethal effects” . Less well understood are the costly, sub-lethal “risk” effects that predators generate, causing potential prey and competitors to alter their behaviour in order to reduce their risk of being attacked or consumed [2,3]. To avoid predation, prey may shift to safer, less profitable habitats [4,5], forage less actively or efficiently [6,7], alter diet selectivity , eat less food or smaller food items [9,10] and become more vigilant . Ultimately, such shifts can impact prey growth rates and reproductive fitness [12,13]. In addition, risk effects can generate niche “partitions” between competitors that reduce conflict over prey items or habitats [12,14]. Removal of a predator can lead to dissolution of niche partitions, allowing competitors to access formerly rare, shared prey and/or expand their diets to include novel prey items [15,16]. Competitive release following loss of a predator may also occur in the absence of risk effects simply due to an increase in the abundance of shared prey items .
Sharks are typically considered to be the apex predators on coral reefs , yet evidence that they play an important ecological role in these systems is equivocal . Most of the evidence that they impose “top-down” effects on reef fishes has come from relatively small-scale experiments focused not on sharks but on the broader piscivorous guild [20,21], but see . At larger scales (whole reefs), the evidence for or against the importance of sharks comes from large-scale, “natural experiments” where large, carnivorous fishes have been removed by the pervasive and chronic human activity of fishing. Some of these studies have failed to find clear evidence for piscivore-driven trophic cascades on reefs [19,23–30], but a natural experiment conducted at the Northern Line islands, where fishing has created a 5 to 7-fold difference in predator abundance across reefs , demonstrated that sites with higher predator abundance were characterised by reductions in several prey attributes including abundance and foraging activity , size and longevity , condition  and size-at-sex change . The role of reef sharks therefore remains ambiguous, with growing evidence that, in certain contexts, they may act as functionally redundant mesopredators that influence reef systems primarily via competition rather than predation [19,34–36].
Comparisons between reefs where fishing has reduced numbers of sharks and reefs where shark assemblages remain intact constitute a natural experiment that can be used to examine the ecological role of these predators . While this approach cannot fully control for alternative factors, it offers one of the few ways to gather evidence at scales relevant to the range of movement of sharks within tropical marine ecosystems (entire reefs and groups of reefs) . However, in most cases, fishing removes not just sharks but also prey and competitor species, such that a reduction in shark-induced effects is compounded by the imposition of a new type of mortality and a loss of select species within the food web .
Two reef systems located over 300 km from the coast of northwestern Australia offer a unique opportunity to examine how the loss of sharks may influence reef fish assemblages in a situation that is largely free of the secondary effects of fishing [40,41] (Fig 1). The Scott Reefs and the Rowley Shoals have been the subject of the Australian Institute of Marine Science Long Term Monitoring Program for over two decades and are characterised by similar habitat structures, benthic communities, areas (~180 km2) and productivity levels, in addition to having well-characterised disturbance histories . Importantly, however, the two locations are sufficiently distant from each other (~400 km) to preclude intermixing of shark and fish populations [43,44]. The Rowley Shoals are a marine protected area that was established in 1990 and has relatively low levels of charter and illegal fishing due to its remoteness from both the Australian mainland and Indonesia . In contrast, sharks have been selectively targeted at the Scott Reefs for centuries by Indonesian fishers, a practice legalised by the Australian government in 1974 . As a result, shark abundance at the Scott Reefs is now 4–17 times lower than at the Rowley Shoals, with some species that are the preferred targets of fishing (e.g. silvertips Carcharhinus albimarginatus and scalloped hammerheads Sphyrna lewini) almost entirely absent at the former location [46,47]. In contrast, populations of teleosts are still relatively intact at the Scott Reefs as fishers primarily target them for subsistence and do not have the capacity to transport large quantities of teleosts . The contrast in shark abundance between the Scott Reefs and the Rowley Shoals offers the opportunity to examine the importance of these predators within the context of a large-scale (hundreds of km, multiple reefs) natural experiment.
The map depicts the Scott Reefs, where sharks are targeted by Indonesian fishermen, and the Rowley Shoals, where sharks are protected.
A study by Ruppert et al. (2013) found that the shark-depleted fish assemblage at the Scott Reefs had significantly greater numbers of mid-sized carnivores than the Rowley Shoals, suggestive of mesopredator release . The aim of this analysis is to explore whether mesopredators have also undergone changes in diet and condition at the Scott Reefs relative to the Rowley Shoals due to the removal of sharks. In the absence of risk effects, prey have been shown to upregulate activity and preferentially forage in riskier but more profitable habitats relative to prey in predator-rich habitats [20,48,49]. Importantly, risk effects can influence the behaviour of species even when predation rates are low or non-existent .
We therefore hypothesized that mesopredators at the Scott Reefs would consume more fish and squid, and fewer benthic invertebrates, than conspecifics at the Rowley Shoals. The pursuit of fish and other prey items typically requires mesopredators to leave the relative safety of the coral reef and venture into the water column, where the risk of attack by predators and/or competitors is higher . In contrast, slow-moving benthic invertebrates can be consumed within the relative safety of the reef and have shorter handling times than fishes . We also hypothesized that loss of competitive effects would lead to a shift in diet from benthic to water column-based prey in mesopredatory teleosts due to breakdown in niche partitioning. Prey differ in calorific content, therefore long-term, consistent prey switching should ultimately result in measurable differences in condition between populations of the same species. We hypothesized that the condition of mesopredators, defined as body weight, height and width for a given length, should improve in response to loss of predation and/or competition due to increased consumption of energy-rich fishes and squids . The evidence for these hypotheses is examined and, within the context of this natural experiment, alternative processes that might also create the observed patterns in diet and condition between these reef systems are considered.
This research was conducted under UWA Ethics Approvals: RA3/100/1279, RA3/100/1172 and relevant permits were obtained from the Department of Fisheries (Exemptions 2148 and 2316), Department of Parks and Wildlife (SF009530, SF0010280 and SF009530) and the Department of Sustainability, Environment, Water, Population and Communities (AU-COM2012-170).
Ruppert et al. (2013) estimated that reef sharks (primarily silvertip Carcharhinus albimarginatus and grey reef Carcharhinus amblyrhynchos) were ~3 times more abundant at the Rowley Shoals than at the Scott Reefs, based on baited remote underwater video systems (BRUVS). The mean of the maximum number of sharks seen together in a single frame of BRUVS footage was 0.4 per hour at the Scott Reefs vs ~1 per hour at the Rowley Shoals . Whereas the Scott Reefs shark assemblage is characterised primarily by relatively small, “mesopredatory” shark species such as grey reef C. amblyrhynchos, zebra Stegostoma fasciatum and white tip reef Triaenodon obesus, the Rowley Shoals assemblage additionally contains larger, species such as silvertip C. albimarginatus, tiger Galeocerda cuvier, scalloped hammerhead Sphyrna lewini and great hammerhead Sphyrna mokarran .
Data on fish abundance for the five focal species (see below) and an additional eleven other mesopredators (Table A in S1 File) were extracted from the Long Term Monitoring Program (LTMP) database of the Australian Institute of Marine Science (AIMS), which annually surveyed three sites per location (the Scott Reefs and the Rowley Shoals) between 1994 and 2008. Each site was comprised of five, 50 m-long transects marked with star pickets every 10 m at a depth of 6–9 m. Transects were 10 m apart. For each survey, a trained scuba diver counted large, mobile species at a distance of up to 2.5 m on either side of the transect line while swimming at a speed of 10 m/min. An assistant swam 10 m behind the surveyor, laying a tape measure on the sea floor. Once the five transects were complete, the surveyor returned along the tape measure and counted small, site-attached species.
Fishes of five focal species were collected at the Scott Reefs in November 2012 and February 2014 and at the Rowley Shoals in April 2013 and November 2013. Up to 90 individuals of Lutjanus gibbus, Lutjanus decussatus and Lutjanus bohar and 50 individuals of Lutjanus kasmira and the lethrinid Monotaxis grandoculis were sampled by free divers using spear guns at each location (Table 1). Fishes were collected at depths of up to 20 m, euthanised using the iki jime method and stored in an ice slurry for a maximum of three hours. Body dimensions in mm (standard, fork and total length) were measured using a purpose-built fish board with inset tape measure. Digital calipers were used to measure body height at the anterior edge of the dorsal fin and body width at the first ray of the dorsal fin, while total wet weight (g) was recorded using a digital balance to an accuracy of 0.01 g. Sex and sexual stage based on  were also recorded following examination of the gonads. Guts were removed and preserved in 10% formalin for a month, before being washed in water and the formalin replaced with 70% ethanol.
Also shown are mean (range) values for length of five mesopredatory fishes at the Scott Reefs and the Rowley Shoals. The p-values for results of a 2 x 2 contingency table with no fixed margins (% females) and a two-sample t-test assuming unequal (L. decussatus, L. kasmira and L. bohar) and equal (L. gibbus, M. grandoculis) variances (total length) are shown.
We chose to study L. gibbus, L. decussatus, L. kasmira, M. grandoculis and L. bohar because they consume a range of benthic invertebrates and teleosts and therefore have the potential to display dietary shifts in response to changes in risk effects. In addition, the five species occupy relatively high trophic positions within reef food webs and therefore have the potential to act as both the prey and competitors of reef sharks. Moreover, the species differ in life history (e.g. maximum size, size at reproduction, longevity, foraging behaviour) allowing for the generality of the findings to be examined. For example, L. gibbus reaches a maximum length of 50 cm, while L. decussatus, L. kasmira, M. grandoculis and L. bohar attain lengths of 35 cm, 40 cm, 60 cm and 90 cm, respectively . Lutjanids and lethrinids consume crabs, bivalves, gastropods, echinoderms, worms, cephalopods, shrimp and small teleosts . Although the diet of sharks is poorly understood, there is evidence that reef-associated sharks consume both lutjanids and a diverse array of herbivorous fishes including acanthurids and scarids [34,55–57]. In addition, both reef sharks and lutjanids, including L. bohar, L. gibbus and L. decussatus, are primarily nocturnal feeders, although evidence is limited [58,59]. As a result, the latter are likely to experience stronger risk effects than diurnally active fishes, although risk effects can occur without direct consumption [50,60,61].
Comparability of collections and locations.
The LTMP dataset was filtered for the 16 most abundant mesopredators and years common to both locations. Data from 1994, 1999, 2003 and 2004 were excluded because only one location was sampled in these years. Mean abundances per site and location (the Scott Reefs and the Rowley Shoals) were calculated for each species (Table A in S1 File). Mean abundance was compared between locations using a paired two sample for means t-test. Mean abundance data for each species in each location was then analysed using least squares regression. We used a 2 x 2 Chi-Square contingency table with no fixed margin to determine whether sex ratios of sampled fishes differed between locations. Total lengths of each species were compared between locations using a t-test based on the log10 transform of individual values. Data series of temperature for the same time period and depths were available for single sites on the reef slopes of the Scott Reefs (6 m) and the Rowley Shoals (Mermaid Reef, 7 m) between June 2007 and August 2012 from the AIMS LTMP temperature database . Mean monthly average temperature at each location was compared using a paired t-test.
Gut content analysis.
The volume and wet mass (formalin-fixed and ethanol-preserved) of intact guts from L. gibbus, L. decussatus and L. bohar were measured and prey items extracted. The presence of scales, jaws and other items were used to identify fishes, while eyes, carapaces and chelipeds were used to identify crustaceans. The length and morphology of beaks were used to distinguish between squid and octopi . Prey items were identified to species where possible using compound and dissection microscopes and photographed. Individuals that had empty stomachs were excluded from the analysis. Prey items were pooled into ten categories (Table 2) and two broader groups: “water column” and “benthic”. The percentage of water column-derived prey items found in the guts was calculated for each individual to account for the fact that some mesopredators contained more than one prey item. To assess whether the percentage of prey items in the guts that originated in the water column was significantly higher at the Scott Reefs than at the Rowley Shoals, a 2×2 Chi-Square contingency table with no fixed margin was used.
We calculated length–weight relationships, where condition was estimated from the residuals of an ordinary least squares regression of log-transformed body mass against length, following . Lengths (L; mm) and weights (W; g) for each fish were first log10 transformed to stabilize variance. A dummy variable (DV = 1 for the Scott Reefs, DV = 0 for the Rowley Shoals) was included to test for a main effect (DV) and the presence of an interaction DV × log L. The general model was log W = b0 + b1 × log L + [b2 × DV] + [b3 × log L × DV]. Predicted weight at the minimum, median and maximum lengths was back-calculated (10log W) for each species at each location and the difference between the two estimates calculated as the percentage difference in weight:
This approach was repeated for body height (%ΔH) and width (%ΔWi) on length using non-transformed data. Two independent indices of condition, including Fulton’s condition factor (K) and relative condition factor Kn were also calculated (Methodology in S1 File).
Comparability of collections and locations
The fish assemblage at the Scott Reefs was characterized by greater numbers of mesopredators than the Rowley Shoals (paired two-sample for means, one-tailed t-test: p = 0.004, t16 = 1.75). The slope of the regression of mesopredator abundance at the Scott Reefs vs. abundance at the Rowley Shoals was significantly different from 1 (b1 = 1.79 ± 0.16; Fig 2). At the Scott Reefs, L. gibbus, L. decussatus, L. kasmira and M. grandoculis were 2.8, 1.7, 1.9 and 1.6 times more abundant than at the Rowley Shoals, respectively. In contrast, L. bohar was 0.5 times less abundant at the Scott Reefs than at the Rowley Shoals.
Abundance data were sourced from the Australian Institute of Marine Science Long-Term Monitoring Program. The plot includes the five study species (black) and eleven non-focal species (white). The dashed line shows a 1:1 relationship between abundances at each location. Abundance was averaged across 3 sites at each location (Clerke, Imperieuse and Mermaid at the Rowley Shoals and North Scott, South Scott and Seringapatam at the Scott Reefs). The inset figure shows logged values for mean fish abundance per site-location. See Table A in S1 File and the AIMS website for details:http://www.aims.gov.au/docs/research/monitoring/monitoring.htm.
A total of 707 fish of the five focal species were collected from the Scott Reefs and the Rowley Shoals (Table 1). The sex of 70% of these fishes was determined, with sex ratios consistent between locations for all species except L. bohar (2×2 contingency table with no fixed margins: χ2 = 4.9, 0.025<p<0.05). Mean length was determined to be similar at the Scott Reefs and the Rowley Shoals for M. grandoculis (two-sample, two-tailed t-test assuming equal variances: p = 0.5, t100 = 0.68) and L. bohar (p = 0.05, t176 = 2.0), but differed for L. gibbus (p = 0.005, t174 = 2.8), L. decussatus (p = 3.7E-05, t165 = 4.24), and L. kasmira (p = 0.0001, t95 = 4.0; Fig A in S1 File). There was no significant difference in paired monthly temperatures between the two locations (paired two-sample, two-tailed t-test for means: p = 0.45, t82 = 0.76).
Gut content analysis
A total of 401 prey items were extracted from the stomachs of 75 L. gibbus, 71 L. bohar and 65 L. decussatus (Table 2). Guts were dominated by fish and crabs, although other benthic prey items such as gastropods were also abundant, particularly in L. gibbus. For all mesopredators, counts of water column-derived prey items were significantly higher at the Scott Reefs than at the Rowley Shoals (2×2 contingency table with no fixed margins; L. gibbus, 40.1±6.6% vs 25.5±6.1%, χ2 = 10.7, 0.001<p<0.005; L. decussatus, 78.4±6.0% vs 49.4±8.9%, χ2 = 11.9, p<0.001; L. bohar, 85.6±4.6% vs 63.8±7.3%, χ2 = 30.9, p<0.001; Fig 3). Among the prey found in the guts of L. bohar were fishes belonging to Scaridae, Sygnathidae (Corythoichthys spp), Lethrinidae (M. grandoculis), Serranidae (possibly Epinephelus fasciatus), Acanthuridae, Balistidae and Chaetodontidae. Scarids were also found in the guts of L. decussatus while echinoderms were only found in the guts of L. gibbus at the Rowley Shoals.
Plot of the mean percentage of prey items from the water column (±SE) in the guts at the Scott Reefs (pale grey) and the Rowley Shoals (dark grey) for three species of mesopredator (2x2 Chi-Square contingency table with no fixed margin: L. bohar, χ2 = 30.9, p<0.001, L. decussatus, χ2 = 11.9, p<0.001 and L. gibbus, χ2 = 10.7, 0.001<p<0.005).
All linear regressions on log-transformed length and weight data were significant with a clear effect of location for L. bohar, L. decussatus and L. kasmira (Fig 4; Table C in S1 File). All species were heavier for a given length at the Scott Reefs relative to the Rowley Shoals including L. gibbus (+7.6%), L. decussatus (+14.6%), L. kasmira (+28.2%), M. grandoculis (+7.5%) and L. bohar (+8.8%; Table B in S1 File). Lutjanus bohar was the only species that had a significant interaction between length and location, with an inflection point at 601 mm. Below this length, L. bohar were heavier for a given length at the Scott Reefs than at the Rowley Shoals. At the observed median length of 468 mm, fish were 8.8% heavier at the Scott Reefs than the Rowley Shoals, while at the maximum length of 715 mm, animals at the Scott Reefs were 2.4% lighter (see Fig B in S1 File).
Differences in the regression lines between species might be obscured due to variation in the range of the axes. See Table C in S1 File for statistics.
Linear regressions on height-length data were significant for all species, with an effect of location for all species except L. gibbus and no interactions (Table C in S1 File). Respectively, L. decussatus, L. kasmira, M. grandoculis and L. bohar were 4.1%, 5.1%, 3.5% and 8.6% deeper bodied for a given length at the Scott Reefs than at the Rowley Shoals (Table B in S1 File).
Linear regressions on width-length data were significant for all species except L. gibbus, while L. bohar had a significant interaction between location and the width-length relationship. Lutjanus decussatus, L. kasmira and M. grandoculis were 5.9%, 9.2% and 7% wider, respectively, for a given length at the Scott Reefs than the Rowley Shoals. The significant interaction between length and location for L. bohar (linear regression, DVxL, p = 0.021, n = 78) meant that while individuals were 5.1% wider at the median length at the Scott Reefs than at the Rowley Shoals, the smallest animals were no different in width between the two locations (Lmin: 0.2%) yet substantially wider at the Scott Reefs at the largest sizes (10.3% wider at 702 mm).
Our results are consistent with the hypothesis that reef sharks influence the diet and condition of mesopredators on coral reefs. On shark-depleted reefs, mesopredatory fishes were relatively abundant and fed on a diet that was largely composed of fishes and squid, whereas at the Rowley Shoals, where higher densities of sharks were present, abundances of mesopredators were lower and the diets of the same species were dominated by benthic invertebrates. These differences in diet were accompanied by variation in morphology and condition, such that fishes at the Scott Reefs tended to be heavier, deeper-bodied and wider for a given body length than fishes of the same species at the Rowley Shoals. The results of this study suggest that sharks may impose significant top-down effects on reef fish assemblages. However, it is unclear whether the mechanism by which these predators structure ecosystems is based on risk effects, competition, or a mixture of both.
Consistent with the hypothesis that reef sharks are important predators that impose risk effects on potential prey, mesopredators consumed a greater proportion of low-risk, benthic food items at the Rowley Shoals relative to conspecifics at the Scott Reefs. As a prey source in a predator-rich environment, benthic invertebrates living within the interstices of the reef represent a less risky prey option than fishes or squid, which are typically located in the water column, where risk of predation is higher . Fishes are also faster-moving, rove over larger areas, require more activity and energy to capture and have longer handling times than benthic invertebrates [66,67]. Additionally, the act of predation on reef fishes generates alarm cues, alerting other prey to the presence of the predator and initiating learning about predator behaviours that makes future prey harder to catch [68,69]. Together, these factors make fish and squid far riskier prey items than benthic invertebrates for mesopredators in situations where they, in turn, are subject to the threat of being consumed. Our results are consistent with previous studies showing that dugongs, bottlenose dolphins and turtles manage their risk of being attacked by tiger sharks by foregoing energetically profitable food in the centre of sea grass meadows, where escape options are limited, in favour of lower quality food on the edge of meadows [4,5,70].
It is entirely possible that the shift in diet recorded in mesopredators at the Scott Reefs relative to the Rowley Shoals reflects competitive rather than predatory release. There is growing evidence that sharks and large piscivores such as lutjanids and serranids share food resources and occupy overlapping trophic niches [19,34]. A reduction in the number of sharks at the Scott Reefs may not only have increased the availability of formerly rare, shared prey species to mesopredators, but also weakened niche partitions, allowing mesopredators to consume novel teleost prey that were previously off-limits. This would be consistent with previous studies in a range of taxa that have demonstrated niche expansion following the loss of a competitor [17,71,72]. Declines in the blue shark Prionace glauca, for example, led to changes in habitat use by the salmon shark Lamna ditropis, potentially causing the latter to incorporate new cephalopod prey species into their diets . The finding that competition may be an important mechanism by which sharks structure reef fish assemblages would be consistent with recent studies that have concluded that reef sharks may operate as functionally redundant mesopredators rather than apex predators . However, it is likely that sharks regulate fish assemblages via a complex mixture of both competitive and predatory effects that is continuously evolving to reflect the relative sizes of the reef sharks and teleosts within the population.
A diet of fish (5.4–6.1 kcal/g) and squid (5.5 kcal/g) is far more energetically rewarding than one composed of relatively low energy prey such as crustaceans (3.4–3.7 kcal/g) and gastropods (1.1 kcal/g), in addition to being less energetically costly to digest and quicker to assimilate . Diet is one of the key determinants of condition . To test if the observed shift in mesopredatory diet from invertebrates towards higher-quality prey from the water column was correlated with improved condition, we compared the body dimensions of mesopredators at the Scott Reefs and the Rowley Shoals. All five mesopredators were significantly heavier for a given length at the Scott Reefs compared to the Rowley Shoals. The regression of log-transformed body mass against length for L. bohar was characterized by an inflection point at 61 cm, above which there was effectively no difference in condition between individuals at Scott Reefs and the Rowley Shoals. Given a mean maximum length of the two dominant species of reef shark at the two locations of 1520 mm as measured on stereo-Baited Remote Underwater Video Systems , L. bohar may enter into a size refuge at a length of ~55 cm as reef sharks typically do not consume prey >36% of their length . Moreover, predators tend to feed on prey between one and three orders of magnitude smaller in mass than themselves . Given a maximum published mass for C. amblyrhynchos of ~34 kg, prey over 3.4 kg are therefore unlikely to be consumed (but see ). Length-weight regressions suggest that an L. bohar of this weight is ~59 cm, corroborating size refugia as a possible explanation for the inflection point. Footage from the stereo-BRUVS also support the premise that large L. bohar escape predation at larger sizes, as they depict large individuals feeding aggressively on bait in the presence of sharks .
In addition to being heavier for a given length at the Scott Reefs relative to the Rowley Shoals, mesopredatory fishes at the former location were also significantly deeper bodied for a given length at the shark-depleted location. This finding is consistent with studies of body shape in the Bahamas mosquitofish Gambusia hubbsi, which found that low-predation environments selected for fishes with deeper bodied anterior regions than high-predation environments, potentially due to a reduced need to engage in fast-starts . Larger body depth may also have been selected for at the Scott Reefs due to increased competition between mesopredatory fishes following a reduction in shark numbers. Fishes with greater body depths can swim for more prolonged periods of time and may therefore be better able to secure prey, access mates and conserve energy . The lack of evidence for a significant increase in height for L. gibbus at the Scott Reefs compared to the Rowley Shoals was surprising, however L. gibbus may be anatomically constrained in body height, as it is deeper-bodied than other lutjanids .
Mesopredatory fishes were also significantly wider for a given length at the Scott Reefs than at the Rowley Shoals. Lutjanus decussatus were 5.9% wider at the median length at the Scott Reefs than at the Rowley Shoals, while L. kasmira, M. grandoculis and L. bohar were 9.2%, 7.0% and 5.1% wider, respectively. Increased width acts as a good measure of condition because it is indicative of better developed fat reserves and propulsive muscles in the caudal region . Indices such as Fulton’s K factor also confirmed that all mesopredatory fishes were in significantly better condition at the Scott Reefs compared to the Rowley Shoals. Improved condition among mesopredatory fishes at the Scott Reefs may be a consequence of the observed shift in their diet towards fish prey relative to conspecifics at the Rowley Shoals. However, a reduction in the abundance of sharks at the Scott Reefs may also have promoted the condition of mesopredatory fishes by allowing them to forage more effectively and for longer periods [7,83,84], waste less energy on costly escapes from attacks [49,85,86] and experience fewer physiological stress responses, which can trigger muscular protein catabolism via increased plasma corticosterone levels [87–89].
Paired reef systems constitute a useful window into the debate on the role of sharks on coral reefs. However our results must be interpreted cautiously in light of the fact that natural experiments are not replicated, randomized and controlled trials . Indeed, one possible alternative explanation for the observed results is that the temperature regime at the two locations differed. However, we could find no consistent differences in average water temperatures between locations. In addition, it is possible that the results were influenced by other differences between the locations and/or the sampling design. However, Ruppert et al. (2013) argued that the two reef systems were similar in terms of productivity, habitat structure, the composition and average cover of reef size, benthic communities and the frequency, intensity and history of disturbance. Moreover, it is impossible to exclude the possibility that the results were an artefact of intra- and inter-annual variation. As it was logistically unfeasible to collect all samples at the same time at both locations, fishes were collected at the Scott Reefs in February 2014 and November 2012 and at the Rowley Shoals in April 2013 and November 2013. Temporal fluctuations in environmental conditions may therefore have influenced the results.
It is also possible that the recorded differences in the diet and condition of mesopredatory fishes between locations reflected differences in the relative availability of their prey rather than predation risk. However, herbivores and other potential prey of mesopredatory fishes were significantly more abundant at the Rowley Shoals compared to the Scott Reefs, which suggests that the distribution of prey items is not a factor . It is also plausible that the observed patterns in the data were driven by differences in the abundance of benthic invertebrates between the two locations. Assessing the abundance of benthic prey was beyond the scope of this study, since invertebrates hide in the interstices of the reef during the day and are difficult to survey accurately. As gonads were not weighed, it is possible that variation in gonad size reflecting reproductive status influenced our results. However, the sex ratios and spawning stages of collected fishes were similar across locations such that additional weight associated with, for instance, egg production is an unlikely factor. Furthermore, the study species are thought to reproduce throughout the year, thus small differences in the timing of sampling would be unlikely to produce disproportionate numbers of reproducing animals between locations. Parasites can also affect condition , but no obvious differences in parasite loads were observed between locations when analysing guts.
An alternative explanation for the observed trends is overfishing of teleosts at the Scott Reefs. For instance, it is possible that fishers have removed significant quantities of mesopredatory teleosts at this location, driving competitive release in the remaining piscivores. However, teleost-targeted fishing at the Scott Reefs is thought to be limited in nature, as Indonesian fishers focus predominantly on sharks, sea cucumbers and other high-value items and have limited abilities to transport large quantities of fishes . Teleost fishing at the Scott Reefs is also an unlikely explanation as many of the mesopredatory fish species supposedly targeted by fishers for subsistence are more abundant at the Scott Reefs relative to the Rowley Shoals .
Irrespective of their cause, high indices of condition are likely to have major effects on the reproductive success of mesopredatory fishes. Fish in good condition are more fecund, spawn earlier and produce larger eggs that give rise to more viable larvae than those produced by fish in poorer condition [90–92]. Given that genetic and demographic studies show that the fish and coral communities of these oceanic coral reefs are largely self-seeding , improved condition of mesopredatory fishes may directly contribute to the greater abundances of these species at the Scott Reefs compared to the Rowley Shoals. This does not require strong stock-recruitment relationships, even though there is evidence for their existence on coral reefs , because maternal contributions play a major role in survivorship of new recruits on coral reefs via mortality bottlenecks . If better condition of adults is indeed aiding reproductive output and survivorship of young, as a number of studies suggest is likely, this could provide a means for differences in the abundance of mesopredatory fishes between these isolated reef systems to “snowball” through time.
Variation in the diet of mesopredatory fishes between locations is likely to have important implications for energy flows, community composition and top-down control of invertebrate populations on coral reefs. Although reef sharks may not be apex predators, our results suggest that their loss may still have the capacity to alter the foraging behaviour, diet and fitness of mesopredatory fishes through direct predation, risk effects or competition. As most mesopredatory teleosts have different life histories to reef sharks (e.g. shorter generation times) and smaller gape sizes [95,96], such changes are unlikely to “compensate” for the loss of sharks. Juvenile lutjanids and lethrinids are important predators of newly-recruited reef fishes in coral reef ecosystems and greater abundances of these medium-sized predators are likely to influence rates of survival and foraging behaviour in other young fishes, including herbivores and invertivores . Such displacement of lethal and risk effects within the food web has the potential to influence the integrity of the reef system by altering the quantity and type of resources that are consumed. In addition, dietary shifts in mesopredatory fishes may cause benthic invertebrates to undergo lower rates of mortality at the Scott Reefs relative to the Rowley Shoals, assuming that other fishes do not change patterns of feeding to exploit this resource.
Although natural experiments lack the manipulation and replication that have traditionally been considered the “gold-standard” of the scientific method, there is growing consensus that these are one of the few techniques available to test hypotheses at large, ecologically relevant scales . The results of this study show that mesopredatory fishes on two comparable reef systems have very different diets and condition indices that are consistent with variation in the abundance of reef sharks and associated differences in risk and/or competitive effects. Our findings have important implications for reef systems worldwide, because sharks are being removed by fishing from coral reef systems at an unprecedented rate [18,99]. This analysis suggests that continued exploitation of sharks may lead to significant shifts in the diet, condition and therefore reproductive fitness of mesopredatory fishes in these environments that have the potential to propagate through coral reef food webs. Further research is required to ascertain the relative contribution of competition and predation-induced risk effects to the observed outcome, particularly given that lutjanids do not seem to comprise a significant proportion of the diets of reef sharks . In addition, acoustic/archival tagging  and/or behavioural observations would help to clarify whether removal of sharks leads to shifts in habitat use and diet among prey and competitors. However, our results strongly suggest that populations of sharks and other predators in marine systems should be protected in order to maintain the health of the broader ecosystem.
Fig A in S1 File. Frequency distributions by size class (mm) of L. gibbus, L. decussatus, L. kasmira, M. grandoculis and L. bohar at the Scott Reefs (grey) and the Rowley Shoals (black), where the size class represents the upper limit of the values. Fig B in S1 File. Plot of the length (mm) of L. bohar vs percentage difference in weight (green), height (orange) and width (blue) at the Scott Reefs compared to the Rowley Shoals, based on regression relationships for each of the measured parameters. Fig C in S1 File. Condition indices (K and Kn) plotted against Log Length (Log L) on the x-axis for focal species at the Scott Reefs (grey) and the Rowley Shoals (black) for, from top to bottom: L. gibbus, L. decussatus, L. kasmira, M. grandoculis and L. bohar, where and . W was defined as total body weight of a fish (g), Wobserved as the recorded weight of an individual (g) and Wstandard as the weight (g) predicted by the formula Wstandard = a × Lb. Species-specific coefficients a and b were sourced from Fishbase. Table A in S1 File. Mean fish abundances per site at each location of 16 mesopredators at the Scott Reefs and the Rowley Shoals, with focal species in bold. The values were calculated using the Long Term Monitoring Program (LTMP) database of the Australian Institute of Marine Science (AIMS; (http://www.aims.gov.au/docs/research/monitoring/monitoring.html). Standard error estimates are presented in parentheses. Table B in S1 File. Percentage difference in weight (W), height (H) and width (Wi) of each of five mesopredatory fishes at the Scott Reefs (S) relative to the Rowley Shoals (R), where %ΔW = 100×(WS-WR)/WR. H and Wi were substituted into this equation to calculate differences in these variables. Differences were calculated for the median value of length for each species. All median differences were significant (regression analysis), except for comparisons of the H and Wi of L. gibbus. Species characterised by significant interactions between location and length are marked with an asterisk; parentheses contain % differences at the minimum and maximum lengths. Table C in S1 File. Linear regression of (A) Log Weight (Log W) and Log Length (Log L), (B) Height (H) and L and (C) Width (Wi) and L. Log W = b0 + b1× Log L + [b2×DV] + [b3×Log(L)*DV], where a dummy variable (DV = 1 for the Scott Reefs, DV = 0 for the Rowley Shoals) was included to test for a main effect (DV) and the presence of an interaction DV × log L. In the formula, bo was the coefficient of the intercept, b1 was the coefficient of log L, b2 was the coefficient of the DV, b3 was the coefficient of DV×L, DV was (0,1) for the Rowley Shoals (R) and Scott Reefs (S), respectively, and DV×L was only included if the coefficient was significant. For (b) and (c), H and Wi were substituted for W in the same equation, but with no logarithmic transformations. Standard errors are indicated in parentheses. Asterisks indicate significance (* = p<0.05;** = p<0.001, and *** = p<0.0001).
This research was funded in part by an International Postgraduate Research Scholarship, the University of Western Australia, Perth and the Australian Institute of Marine Science (AIMS). We would like to thank AIMS for providing UVC-based fish abundance data via the AIMS Long Term Monitoring Program and temperature data, the collection of which was funded by Woodside Energy Ltd. We also thank the Department of Fisheries, the Department of Parks and Wildlife (DPaW), the Department of Sustainability, Environment, Water, Population and Communities for arranging permits to conduct research at the Scott Reefs and the Rowley Shoals and the crew on board the RV Solander.
- Conceptualization: SB JM MM.
- Data curation: SB JM.
- Formal analysis: SB JM.
- Funding acquisition: SB JM MM.
- Investigation: SB.
- Methodology: SB JM MM.
- Project administration: SB JM MM.
- Resources: SB JM MM.
- Software: SB JM MM.
- Supervision: SB JM.
- Validation: SB JM MM.
- Visualization: SB.
- Writing – original draft: SB JM MM.
- Writing – review & editing: SB JM MM.
- 1. Ritchie EG, Johnson CN. Predator interactions, mesopredator release and biodiversity conservation. Ecol Lett. 2009;12: 982–998. pmid:19614756
- 2. Wirsing AJ, Heithaus MR, Frid A, Dill LM. Seascapes of fear: Evaluating sublethal predator effects experienced and generated by marine mammals. Mar Mammal Sci. 2008;24: 1–15.
- 3. Preisser EL, Bolnick DI, Benard MF. Scared to death? The effects of intimidation and consumption in predator-prey interactions. Ecology. 2005;86: 501–509.
- 4. Wirsing AJ, Heithaus MR, Dill LM. Living on the edge: dugongs prefer to forage in microhabitats that allow escape from rather than avoidance of predators. Anim Behav. 2007;74: 93–101.
- 5. Heithaus MR, Frid A, Wirsing AJ, Dill LM, Fourqurean JW, Burkholder D, et al. State-dependent risk-taking by green sea turtles mediates top-down effects of tiger shark intimidation in a marine ecosystem. J Anim Ecol. 2007;76: 837–844. pmid:17714261
- 6. Madin EMP, Gaines SD, Warner RR. Field evidence for pervasive indirect effects of fishing on prey foraging behavior. Ecology. 2010;91: 3563–3571. pmid:21302828
- 7. Metcalfe NB, Huntingford FA, Thorpe JE. Predation risk impairs diet selection in juvenile salmon. Anim Behav. 1987;35: 931–933.
- 8. Houtman R, Dill LM. The influence of predation risk on diet selectivity: A theoretical analysis. Evol Ecol. 1998;12: 251–261.
- 9. Ibrahim AA, Huntingford FA. Laboratory and field studies of the effect of predation risk on foraging in three-spined sticklebacks (Gasterosteus aculeatus). Behaviour. 1989;109: 46–57.
- 10. Catano LB, Rojas MC, Malossi RJ, Peters JR, Heithaus MR, Fourqurean JW, et al. Reefscapes of fear: Predation risk and reef hetero-geneity interact to shape herbivore foraging behaviour. J Anim Ecol. 2015;1: 146–56.
- 11. Brown JS. Vigilance, patch use and habitat selection: Foraging under predation risk. Evol Ecol Res. 1999;1: 49–71.
- 12. Werner EE, Peacor SD. A review of trait-mediated indirect interactions in ecological communities. Ecology. 2003;84: 1083–1100.
- 13. Clinchy M, Sheriff MJ, Zanette LY. Predator-induced stress and the Ecology of Fear. Funct Ecol. 2013;27: 56–65.
- 14. Brook LA, Johnson CN, Ritchie EG. Effects of predator control on behaviour of an apex predator and indirect consequences for mesopredator suppression. Dickman C, editor. J Appl Ecol. 2012;49: 1278–1286.
- 15. Palomares F, Caro TM. Interspecific Killing among Mammalian Carnivores. Am Nat. 1999;153: 492–508.
- 16. Schuette P, Wagner AP, Wagner ME, Creel S. Occupancy patterns and niche partitioning within a diverse carnivore community exposed to anthropogenic pressures. Biol Conserv. 2013;158: 301–312.
- 17. Moreno RS, Kays RW, Samudio R. Competitive Release in Diets of Ocelot (Leopardus Pardalis) and Puma (Puma Concolor) After Jaguar (Panthera Onca) Decline. Journal of Mammalogy. 2006. pp. 808–816.
- 18. Robbins WD, Hisano M, Connolly SR, Choat JH. Ongoing Collapse of Coral-Reef Shark Populations. Curr Biol. 2006;16: 2314–2319. pmid:17141612
- 19. Roff G, Doropoulos C, Rogers A, Bozec YM, Krueck NC, Aurellado E, et al. The Ecological Role of Sharks on Coral Reefs. Trends Ecol Evol. 2016;31: 395–407. pmid:26975420
- 20. Stallings CD. Indirect effects of an exploited predator on recruitment of coral-reef fishes. Ecology. 2008;89: 2090–2095. pmid:18724719
- 21. Gagliano M, McCormick MI. Maternal condition influences phenotypic selection on offspring. J Anim Ecol. 2007;76: 174–182. pmid:17184366
- 22. Rizzari JR, Frisch AJ, Hoey AS, McCormick MI. Not worth the risk: Apex predators suppress herbivory on coral reefs. Oikos. 2014;123: 829–836.
- 23. Friedlander A, DeMartini E. Contrasts in density, size, and biomass of reef fishes between the northwestern and the main Hawaiian islands: the effects of fishing down apex predators. Mar Ecol Prog Ser. 2002;230: 253–264.
- 24. Jennings S, Grandcourt EM, Polunin NVC. The effects of fishing on the diversity, biomass and trophic structure of Seychelles’ reef fish communities. Coral Reefs. 1995;14: 225–235.
- 25. Jennings S, Polunin NVC. Impacts of predator depletion by fishing on the biomass and diversity of non-target reef fish communities. 1997; 71–82.
- 26. Russ GR, Alcala AC. Marine reserves: Rates and patterns of recovery and decline of large predatory fish. Ecol Appl. 1996;6: 947–961.
- 27. Mumby PJ, Dahlgren CP, Harborne AR, Kappel C V., Micheli F, Brumbaugh DR, et al. Fishing, trophic cascades, and the process of grazing on coral reefs. Science. 2006;311: 98–101. pmid:16400152
- 28. Mumby PJ, Steneck RS, Edwards AJ, Ferrari R, Coleman R, Harborne AR, et al. Fishing down a Caribbean food web relaxes trophic cascades. Mar Ecol Prog Ser. 2012;445: 13–24.
- 29. Sandin SA, Smith JE, DeMartini EE, Dinsdale EA, Donner SD, Friedlander AM, et al. Baselines and degradation of coral reefs in the Northern Line Islands. PLoS One. 2008;3.
- 30. Rizzari JR, Bergseth BJ, Frisch AJ, Brock JB, Frisch AJ, Bergseth BJ, et al. Impact of Conservation Areas on Trophic Interactions between Apex Predators and Herbivores on Coral Reefs. Conserv Biol. 2014;29: 418–429. pmid:25185522
- 31. Ruttenberg BI, Hamilton SL, Walsh SM, Donovan MK, Friedlander A, DeMartini E, et al. Predator-induced demographic shifts in coral reef fish assemblages. PLoS One. Public Library of Science; 2011;6: e21062.
- 32. Walsh SM, Hamilton SL, Ruttenberg BI, Donovan MK, Sandin SA. Fishing top predators indirectly affects condition and reproduction in a reef-fish community. J Fish Biol. 2012;80: 519–37. pmid:22380551
- 33. DeMartini EE, Friedlander AM, Holzwarth SR. Size at sex change in protogynous labroids, prey body size distributions, and apex predator densities at NW Hawaiian atolls. Mar Ecol Prog Ser. 2005;297: 259–271.
- 34. Frisch AJ, Ireland M, Rizzari JR, Lonnstedt OM, Magnenat KA, Mirbach CE, et al. Reassessing the trophic role of reef sharks as apex predators on coral reefs. Coral Reefs. 2016;
- 35. Robbins WD, Renaud P. Foraging mode of the grey reef shark, Carcharhinus amblyrhynchos, under two different scenarios. Coral Reefs. 2015;35: 253–260.
- 36. Heupel MR, Knip DM, Simpfendorfer CA, Dulvy NK. Sizing up the ecological role of sharks as predators. Marine Ecology Progress Series. 2014. pp. 291–298.
- 37. Barley SC, Meeuwig JJ. The Power and the Pitfalls of Large-scale, Unreplicated Experiments. Ecosystems. 2016;
- 38. Baum JK, Worm B. Cascading top-down effects of changing oceanic predator abundances. J Anim Ecol. 2009;78: 699–714. pmid:19298616
- 39. Stevenson C, Katz LS, Micheli F, Block B, Heiman KW, Perle C, et al. High apex predator biomass on remote Pacific islands. Coral Reefs. 2006;26: 47–51.
- 40. Russell BC, Vail LL. Report on Traditional Indonesian Fishing Activities at Ashmore Reef Nature Reserve. 1988.
- 41. Nowara GB, Newman SJ. A history of foreign fishing activities and fishery-independent surveys of the demersal finfish resources in the Kimberley region of Western Australia. Perth; 2001.
- 42. Ruppert JLW, Travers MJ, Smith LL, Fortin M-J, Meekan MG. Caught in the middle: combined impacts of shark removal and coral loss on the fish communities of coral reefs. PLoS One. 2013;8: e74648. pmid:24058618
- 43. Underwood JN, Smith LD, Van Oppen MJH, Gilmour JP. Ecologically relevant dispersal of corals on isolated reefs: Implications for managing resilience. Ecol Appl. 2009;19: 18–29. pmid:19323171
- 44. Field IC, Meekan MG, Speed CW, White W, Bradshaw CJA. Quantifying movement patterns for shark conservation at remote coral atolls in the Indian Ocean. Coral Reefs. 2010;30: 61–71.
- 45. Fox J. Legal and Illegal Indonesian Fishing in Australian Waters. In: Cribb R., Ford M, editor. Indonesia beyond the Waters Edge: Managing an Archipelagic State. Institute of Southeast Asian Studies; 2009. pp. 0–248.
- 46. Meekan M, Cappo M. Non-destructive Techniques for Rapid Assessment of Shark Abundance in Northern Australia. Perth; 2004.
- 47. Meekan M, Cappo M, Carleton J, Marriott R. Surveys of Shark and Fin-fish abundance on reefs within the MOU74 Box and Rowley Shoals using Baited Remote Underwater Video Systems. 2006;
- 48. del Mar Palacios M, Warren DT, Mccormick MI. Sensory cues of a top-predator indirectly control a reef fish mesopredator. Oikos. 2015.
- 49. Lima SL, Dill LM. Behavioral decisions made under the risk of predation: a review and prospectus. Can J Zool. NRC Research Press Ottawa, Canada; 1990;68: 619–640. Available: http://www.nrcresearchpress.com/doi/abs/10.1139/z90-092#.U9tyPPmSx8E
- 50. Creel S, Christianson D. Relationships between direct predation and risk effects. Trends Ecol Evol. Elsevier; 2008;23: 194–201.
- 51. Galarowicz TL, Wahl DH. Foraging by a young-of-the-year piscivore: the role of predator size, prey type, and density. Can J Fish Aquat Sci. 2005;62: 2330–2342.
- 52. Farrell ED, Hssy K, Coad JO, Clausen LW, Clarke MW. Oocyte development and maturity classification of boarfish (Capros aper) in the Northeast Atlantic. ICES J Mar Sci. 2012;69: 498–507.
- 53. Froese R, Pauly D. Fishbase. In: www.fishbase.org. 2016.
- 54. Kulbicki M, Bozec Y-M, Labrosse P, Letourneur Y, Mou-Tham G, Wantiez L. Diet composition of carnivorous fishes from coral reef lagoons of New Caledonia. Aquatic Living Resources. 2005. pp. 231–250.
- 55. Papastamatiou YP, Wetherbee BM, Lowe CG, Crow GL. Distribution and diet of four species of carcharhinid shark in the Hawaiian Islands: Evidence for resource partitioning and competitive exclusion. Mar Ecol Prog Ser. 2006;320: 239–251.
- 56. Wetherbee BM, Crow GL, Lowe CG. Distribution, reproduction and diet of the gray reef shark Carcharhinus amblyrhynchos in Hawaii. Mar Ecol Prog Ser. 1997;151: 181–189.
- 57. Papastamatiou YP, Friedlander AM, Caselle JE, Lowe CG. Long-term movement patterns and trophic ecology of blacktip reef sharks (Carcharhinus melanopterus) at Palmyra Atoll. J Exp Mar Bio Ecol. 2010;386: 94–102.
- 58. Nagelkerken I, Dorenbosch M, Verberk WCEP, Cocheret De La Morinière E, Van Der Velde G. Day-night shifts of fishes between shallow-water biotopes of a Caribbean bay, with emphasis on the nocturnal feeding of Haemulidae and Lutjanidae. Mar Ecol Prog Ser. 2000;194: 55–64.
- 59. Gillibrand C, Harris A, Mara E. Inventory and spatial assemblage study of reef fish in the area of Andavadoaka, South-West Madagascar (Western Indian Ocean). West Indian Ocean J Mar Sci. 2007;6: 183–197.
- 60. Burkholder DA, Heithaus MR, Fourqurean JW, Wirsing A, Dill LM. Patterns of top-down control in a seagrass ecosystem: Could a roving apex predator induce a behaviour-mediated trophic cascade? J Anim Ecol. 2013;82: 1192–1202. pmid:23730871
- 61. Heithaus MR, Wirsing AJ, Thomson J a., Burkholder D a. A Review of lethal and non-lethal effects of predators on adult marine turtles. J Exp Mar Bio Ecol. 2008;356: 43–51.
- 62. Long Term Monitoring Program Temperature Database. In: Australian Institute of Marine Science. 2015.
- 63. Pinkas L, Oliphant MS, Iverson ILK. Food Habits of Albacore, Bluefin Tuna, and Bonito in California waters. Fish Bulletin. 1971.
- 64. Le Cren ED. The length-weight relationship and seasonal cycle in gonad weight and condition in the perch (Perca fluviatilis). J Anim Ecol. 1951;20: 201–219.
- 65. Renones O, Polunin NVC, Goni R. Size related dietary shifts of Epinephelus marginatus in a western Mediterranean littoral ecosystem: an isotope and stomach content analysis. J Fish Biol. 2002;61: 122–137.
- 66. Nilsson PA, Brönmark C. Prey vulnerability to a gape-size limited predator : behavioural and morphological impacts on northern pike piscivory. Oikos. 2000;88: 539–546.
- 67. Harper DG, Blake RW. Energetics of Piscivorous Predator-Prey Interactions. J Theor Biol. 1988;134: 59–76.
- 68. Mitchell MD, McCormick MI, Ferrari MCO, Chivers DP. Friend or foe? The role of latent inhibition in predator and non-predator labelling by coral reef fishes. Anim Cogn. 2011;14: 707–714. pmid:21519901
- 69. Kelley JL, Magurran AE. Learned predator recognition and antipredator responses in fishes. Fish Fish. 2003;4: 216–226.
- 70. Heithaus MR, Dill LM. Does tiger shark predation risk influence foraging habitat use by bottlenose dolphins at multiple spatial scales? Oikos. 2006;114: 257–264.
- 71. LaPoint SD, Belant JL, Kays RW. Mesopredator release facilitates range expansion in fisher. Anim Conserv. 2014; n/a-n/a.
- 72. Bolnick DI, Ingram T, Stutz WE, Snowberg LK, Lau OL, Paull JS. Ecological release from interspecific competition leads to decoupled changes in population and individual niche width. Proc R Soc B Biol Sci. 2010;277: 1789–1797.
- 73. Okey TA, Wright BA, Brubaker MY. Salmon shark connections: North Pacific climate change, indirect fisheries effects, or just variability? Fish and Fisheries. 2007. pp. 359–366.
- 74. Abitia-Cardenas LA, Galvan-Magana F, Rodriguez-Romero J. Food habits and energy values of prey of striped marlin, Tetrapturus audax, off the coast of Mexico. Fish Bull. 1996;95: 360–368.
- 75. Lloret J., Shulman G., Love RM. Condition and Health Indicators of Exploited Marine Fishes. Wiley-Blackwell; 2014.
- 76. Barley SC, Meekan MG, Meeuwig JJ. Species diversity, abundance, biomass, size and trophic structure of fish assemblages on coral reefs in relation to shark abundance. Mar Ecol Prog Ser. 2017;
- 77. Bethea DM, Buckel JA, Carlson JK. Foraging ecology of the early life stages of four sympatric shark species. Mar Ecol Prog Ser. 2004;268: 245–264.
- 78. Woodward G, Ebenman B, Emmerson M, Montoya JM, Olesen JM, Valido A, et al. Body size in ecological networks. Trends Ecol Evol. 2005;20: 402–409. pmid:16701403
- 79. Mourier J, Maynard J, Ballesta L, Parravicini Valeriano, Clua E, Domeier ML, Planes S. Extreme Inverted Trophic Pyramid of Reef Sharks Supported by Spawning Groupers. Curr Biol. 2016;26: 1–6.
- 80. Langerhans RB, Gifford ME, Joseph EO. Ecological speciation in Gambusia fishes. Evolution (N Y). 2007;61: 2056–2074.
- 81. Nanami A, Shimose T. Interspecific differences in prey items in relation to morphological characteristics among four lutjanid species (Lutjanus decussatus, L. fulviflamma, L. fulvus and L. gibbus). Environ Biol Fishes. 2012;96: 591–602.
- 82. Fuiman LA. Growth gradients in fish larvae. J Fish Biol. 1983;23: 117–123. Available: http://deepblue.lib.umich.edu/bitstream/handle/2027.42/72250/j.1095-8649.1983.tb02886.x.pdf?sequence=1
- 83. Amo L, López P, Martín J. Nature-based tourism as a form of predation risk affects body condition and health state of Podarcis muralis lizards. Biol Conserv. 2006;131: 402–409.
- 84. Brown JS, Kotler BP, Smith RJ, WO W II. The effects of owl predation on the foraging behavior of heteromyid rodents. Oecologia. Springer-Verlag; 1988;76: 408–415.
- 85. Béchet A, Giroux JF, Gauthier G. The effects of disturbance on behaviour, habitat use and energy of spring staging snow geese. J Appl Ecol. 2004;41: 689–700.
- 86. Carrascal L, Polo V. Coal tits, Parus ater, lose weight in response to chases by predators. Anim Behav. 1999;58: 281–285. pmid:10458879
- 87. Wingfield JC, Silverin B. Effects of corticosterone on territorial behavior of free-living male song sparrows Melospiza melodia. Horm Behav. 1986;20: 405–417. pmid:3793022
- 88. Munck A, Guyre PM, Holbrook NJ. Physiological functions of glucocorticoids in stress and their relation to pharmacological actions. Endocrine Reviews. 1984. pp. 25–44. pmid:6368214
- 89. Scheuerlein A, Van’t Hof TJ, Gwinner E. Predators as stressors? Physiological and reproductive consequences of predation risk in tropical stonechats (Saxicola torquata axillaris). Proc Biol Sci. 2001;268: 1575–82. pmid:11487404
- 90. Marteinsdottir G, Steinarsson A. Maternal influence on the size and viability of Iceland cod Gadus morhua eggs and larvae. J Fish Biol. 1998;52: 1241–1258.
- 91. Ma Y, Kjesbu OS, Jørgensen T. Effects of ration on the maturation and fecundity in captive Atlantic herring (Clupea harengus). Can J Fish Aquat Sci. 1998;55: 900–908.
- 92. Trippel EA, Neil SRE. Maternal and seasonal differences in egg sizes and spawning activity of northwest Atlantic haddock (Melanogrammus aeglefinus) in relation to body size and condition. Can J Fish Aquat Sci. 2004;61: 2097–2110.
- 93. Underwood JN, Travers MJ, Gilmour JP. Subtle genetic structure reveals restricted connectivity among populations of a coral reef fish inhabiting remote atolls. Ecol Evol. 2012;2: 666–79. pmid:22822442
- 94. Vigliola L, Meekan MG. Size at hatching and planktonic growth determine post-settlement survivorship of a coral reef fish. Oecologia. 2002;131: 89–93.
- 95. Hixon MA. Predation as a process structuring coral reef fish communities. In: Sale PF, editor. The Ecology of Fishes on Coral Reefs. San Diego: Academic Press; 1991. pp. 475–508.
- 96. Urban MC. The growth-predation risk trade-off under a growing gape-limited predation threat. Ecology. 2007;88: 2587–2597. pmid:18027761
- 97. McCormick MI, Hoey AS. Larval growth history determines juvenile growth and survival in a tropical marine fish. Oikos. 2004;106: 225–242.
- 98. Miao S, Carstenn S. A new direction for large-scale experimental design and analysis. Front Ecol Environ. 2006;4: 227–227.
- 99. Dulvy NK, Fowler SL, Musick JA, Cavanagh RD, Kyne M, Harrison LR, et al. Extinction risk and conservation of the world’s sharks and rays. Elife. 2014;3: 1–35.
- 100. Clark TD, Taylor BD, Seymour RS, Ellis D, Buchanan J, Fitzgibbon QP, et al. Moving with the beat: heart rate and visceral temperature of free-swimming and feeding bluefin tuna. Proc Biol Sci. 2008;275: 2841–2850. pmid:18755679