Access to this full-text is provided by Wiley.
Content available from Ecology and Evolution
This content is subject to copyright. Terms and conditions apply.
13740
|
Ecology and Evolution. 2019;9:13740–13751.
www.ecolevol.org
1 | INTRODUCTION
Top predators are characterized by some of the largest, most enig‐
matic, and threatened species today on Earth (Hammerschlag &
Gallagher, 2017). Often occupying upper trophic tiers, predators
can influence prey directly through consumption and also indirectly
via the perceived risk of predation. These nonconsumptive effects
can drive food‐risk trade‐of fs that alter behavior, physiology, and
foraging strategies in potential prey (Beauchamp, Wahl, & Johnson,
2007; Heithaus, Frid, Wirsing, & Worm, 2008; Rasher, Hoey, & Hay,
2017). In doing so, predators drive important ecosystem processes
that may induce cascading effects throughout entire ecosystems
(Estes et al., 2011). Despite the impor tant roles they play in eco‐
system dynamics, many populations of large predators are declin‐
ing rapidly as a result of overexploitation, and habitat loss, among a
myriad of other threat s (Lennox, Gallagher, Ritchie, & Cooke, 2018).
Received: 22 October 2018
|
Revised: 25 Februar y 2019
|
Accepted: 25 S eptember 2019
DOI: 10.100 2/ece3.5784
ORIGINAL RESEARCH
Evaluating the effects of large marine predators on mobile prey
behavior across subtropical reef ecosystems
Lindsay M. Phenix1,2 | Dana Tricarico1 | Enrique Quintero1 | Mark E. Bond3 |
Simon J. Brandl4 | Austin J. Gallagher1
This is an op en access article under t he terms of the Creat ive Commons Attributio n License, which permits use, dist ribution and reproduc tion in any medium,
provide d the orig inal work is proper ly cited .
© 2019 The Authors. Ecology and Evol ution pub lished by J ohn Wiley & Sons Ltd.
1Beneath the Waves, Herndon, VA, USA
2Three Seas Progr am, Nor theas tern
University, Nahant, MA, USA
3Florida International Universit y, North
Miami, FL , USA
4Depar tment of Biologic al Sciences, Simon
Fraser Universit y, Burnaby, BC , Canada
Correspondence
Lindsay M. Phenix, Beneath the Waves , PO
Box 126, Herndon, VA 20172, USA.
Email: phenix.l@husky.neu.edu
Funding information
Herbert W. Hoover Foundation; C. and
M.Jones
Abstract
The indirect effect of predators on prey behavior, recruitment, and spatial relation‐
ships continues to attract considerable attention. However, top predators like sharks
or large, mobile teleosts, which can have substantial top–down effects in ecosystems,
are often difficult to study due to their large size and mobility. This has created a
knowledge gap in understanding how they affect their prey through nonconsumptive
effects. Here, we investigated how different functional groups of predators affected
potential prey fish populations across various habitats within Biscayne Bay, FL. Using
baited remote underwater videos (BRUVs), we quantified predator abundance and
activity as a rough proxy for predation risk and analyzed key prey behaviors across
coral reef, sea fan, seagrass, and sandy habitats. Both predator abundance and prey
arrival times to the bait were strongly influenced by habitat type, with open homog‐
enous habitats receiving faster arrival times by prey. Other prey behaviors, such as
residency and risk‐associated behaviors, were potentially driven by predator interac‐
tion. Our data suggest that small predators across functional groups do not have large
controlling effects on prey behavior or stress responses over short temporal scales;
however, habitats where predators are more unpredictable in their occurrence (i.e.,
open areas) may trigger risk‐associated behaviors such as avoidance and vigilance.
Our data shed new light on the importance of habitat and context for understanding
how marine predators may influence prey behaviors in marine ecosystems.
KEYWORDS
baited remote under water video stations, predation risk, predator, risk effects, sharks
|
13741
PHENIX Et a l.
While effects of apex predators are relatively well studied in
terrestrial ecosystems (e.g., Suraci, Clinchy, Dill, Roberts, & Zanette,
2016), their roles in marine systems are generally less understood
(e.g., Casey et al., 2017; Sandin et al., 2008). Sharks, for instance,
are traditionally considered the de facto top predator in marine
ecosystems, and their vulnerabilities to fishing (Gallagher, Kyne, &
Hammerschlag, 2012) and general patterns of population decline
(e.g., Ferretti, Worm, Britten, Heithaus, & Lotze, 2010) have rein‐
forced the import ance of understanding the implications of their re‐
movals on ma rine ecosystem s. Often unifor mly characte rized as apex
predators due to their size and trophic position in marine food webs
(Heupel, Knip, Simpfendorfer, & Dulvy, 2014; Hussey et al., 2014),
sharks may exert strong controlling influences on prey through be‐
haviorally‐mediated, nonconsumptive processes (i.e., predation risk)
(Heithaus et al., 2008; Heithaus, Wirsing, Burkholder, Thomson, &
Dill, 2009). However, the degree to which sharks actually influence
the behavior and physiology of prey species remains understudied
and controversial (Casey et al., 2017; Roff et al., 2016; Rupper t,
Travers, Smith, Fortin, & Meekan, 2013). Studies have suggested
that on coral reefs, herbivorous fish reduce their feeding rates when
exposed to a larger, stationary shark decoy (C atano, Barton, Boswell,
& Burkepile, 2017; Madin, Gaines, & Warner, 2010; Rizzari, Frisch,
Hoey, & McCormick, 2014), but it is unknown whether this acute
suppression actually triggers a long‐term reduction in feeding or if it
simply redistributes the prey fish to a different area. Similarly, it re‐
mains unknown how other sympatric marine teleost predators, such
as barracudas (family Sphyraenidae) or morays (family Muraenidae),
compare to sharks with regard to their nonconsumptive effects on
prey. Nonconsumptive effects would be expected to be particularly
prevalent in shallow, open ecosystems where a larger prey item's
opportunity for escape from roving, apex predators are limited
(Heithaus et al., 2009), thus suggesting a potential ef fect of habitat
complexity.
The lack of a generalizable predator effect (consistency in direc‐
tion and strength) may be expected in diverse, three‐dimensional
ecosystems such as coral reefs where water is clear and opportu‐
nities to shelter temporarily are extensive. These habitats provide
increased visibility for and detectability of mobile, roving predators.
Studies have suggested that in coral reef food webs, reef‐associated
sharks and large teleosts occupy similar trophic niches (Bond et al.,
2018; Frisch et al., 2016; Roff et al., 2016), which may allow for the
detection of generalizable effects of predators on prey or may divert
or dilute the nonconsumptive effects of species traditionally consid‐
ered apex predators on larger prey species. Our knowledge of non‐
consumptive effects of marine predators on prey may benefit from
examining predator–prey interactions under varying environmental
conditions.
An increasingly popular technique for noninvasively assessing
the relative abundance and behavior of mobile fish populations,
while removing diver bias, is baited remote underwater video (BRUV)
surveys (Whitmarsh, Fairweather, & Huveneers, 2017). BRUVs con‐
sist of an underwater camera focused on a standardized bait source
positioned in the field of view (FOV), with the unit orient ated down
current from the camera. Individuals attracted to the bait that swim
into the FOV are “captured” on camera (Armstrong, Bagley, & Priede,
1992), providing a permanent record of observations that can be re‐
viewed multiple times. This record improves the accuracy of the data
and allows for detailed analyses such as those required for examin‐
ing animal behavior. They have also been used in studies assessing
predator–prey relationships (e.g., Klages, Broad, Kelaher, & Davis,
2014) and could be readily used to investigate the potential effects
of marine predators on a suite of prey species, across a variety of
habitats and conditions.
Here, we used BRUVs to examine the nonconsumptive effects
of multiple marine predators on various mobile prey species, across
the varying habitats of Biscayne Bay, Florida. We evaluated these
predator–prey interactions in three ways: (a) inferring ambient risk
in each habitat by quantif ying relative predator abundance and for‐
aging activity; (b) assessing habitat‐specific responses of potential
prey species by measuring prey arrival (as a prox y for apprehensive‐
ness); and (c) gauging risk‐associated behaviors of prey as well as
prey residency at the bait stations (Bond et al., 2019). We hypothe‐
sized that (a) predator activity would be greater in complex habitat s
(Bruno, Stachowicz, & Bertness, 2003; Hutchinson, 1957); (b) prey
would take longer to arrive in less complex, more open habitats due
to limited shelter opportunities; (c) prey residencies would increase
and the number of risk‐associated behaviors would decrease in more
complex habitats (Bruno et al., 2003).
2 | METHODS
2.1 | Study site
This study was conducted from January 21 to August 31, 2017 in the
waters of Biscayne Bay, Florida, USA, including within the boundaries
of Biscayne National Park (BNP; 25°45′42.05″N, 80°11′30.44″W;
Figure 1). This area extends from Key Biscayne to Key L argo and
connects to the Florida Reef Tract, the third largest coral reef system
worldwide. The area is defined by a mixture of coral reefs, seagrass
beds, sof t corals, and sand flats. Biscayne Bay is a shallow water la‐
goon in which a variety of habitat s provide important functional, on‐
togenetic , and trophic value for mangrove and reef‐associated fish,
including sharks and rays, as well as sea tur tles and marine mammals
(Serafy, Valle, Faunce, & Luo, 2007).
2.2 | Baited Remove Underwater Video
(BRUV) surveys
Predator–prey interactions among and between mobile elas‐
mobranch and teleost communities were assessed throughout
Biscayne Bay and Biscayne National Park using baited remote un‐
derwater video (BRUV) sur veys. Each BRUV consisted of a 48‐cm
tall metal pyramid frame with the sides converging at a flat, square
platform (Figure 2). Additional weights (two, 0.5 kg dive weights)
were added to each BRUV frame to increase stability. Each BRUV
was equipped with a 100‐cm PVC bait pole, with a mesh bait bag
13742
|
PHENIX Et al.
(150 mm × 200 mm) attached at the end (via zip ties) containing
~450 g of freshly minced Spanish sardines (Sardinella spp.). High‐
definition action cameras (GoPro Hero and Hero+) were secured to
the square platform and positioned to face outward, with the bait
bag within the estimated 160° FOV, all lights and flashing sensors
on the cameras were deactivated. All footage was shot at 1,080p
high‐definition at 30 frames per second.
All BRUVs were deployed from a boat and lowered to the
sea floor via 30‐m ropes att ached to a visible sur face buoy.
Deployment depths ranged from 1.3 m to 12.8 m with an aver‐
age depth of 6.7 m. In‐water free‐divers were occasionally used to
navigate the BRUVs away from living corals and to ensure proper
orientation on the benthic substrate. BRUVs were deployed in
contiguous areas in groups of three to five, spaced ~300–500 m
apart, and were allowed to soak for 60 min. Deployments were
focused in the following habitat types: coral reef (defined by the
presence of coral colonies and structures), sea fan (defined by
the presence of patchy sea fans), seagrass (defined by contiguous
areas dominated by seagrass), and sand (defined by low‐rugose
habitat s with open sandy areas). Deployments occurred during
daylight hours, bet ween 08 00 and 1330 hr. During each round
of BRUV deployments, we measured depth and water tempera‐
ture (°C) using a HANNA handheld probe (Hanna Instruments, HI
98193). Temperature was recorded as a control to account for any
possible anomalies; average water temperature was 24.4°C across
seasons. We characterized the habitat type as coral reef, sand, sea
fan, or seagrass (based on 50% coverage or higher) and whether
the site was inside or outside the boundaries of Biscayne National
FIGURE 1 Map of BRUV sur vey
deployments in Biscayne Bay, FL, USA.
Red dots = dr y season, green dots = wet
season. White line represents boundary of
Biscayne National Park
|
13743
PHENIX Et a l.
Park (a national park with varying fishing regulations, though it is
not a no‐take zone nor a marine reserve).
2.3 | Video analysis and variables considered
Each 60 ‐min vi d e o wa s re v i e w e d an d an a l y zed in re a l ti m e . An al y s i s
be ga n onc e the BRUV was fi rmly planted in t he be nt hos (~15–30 s)
after deployment. Predators were categorized into three trophic
tiers. Upper trophic predators included barracudas (Sphyraena bar-
racuda), as well as large bodied (>2 m) mid‐water feeding sharks.
Large bodied mid‐trophic predators included large benthic feeding
sharks and green moray eels (Gymnothorax funebris), while small
bodied mid‐trophic predators encompassed small bodied sharks
(<2 m) and spotted moray eels (Gymnothorax moringa). Groupings
were determined based on relative size and the presumed corre‐
lating trophic pressures they placed on the ecosystem (Bond et
al., 2018; LaymanWinemiller, Arring ton, & Jepsen, 2005). Seven
common prey families were identified and used to measure habitat
risk and ri sk ef fec ts: filefi sh (famil y Mon ac ant hi dae), gru nt s (family
Haemulidae), jacks (family Carangidae), porgies (family Sparidae),
rays (family Dasyatidae & Urotrygonidae), snappers (family
Lutjanidae), and triggerfish (family Balistidae). These prey families
were chosen due to their observed abundance in the surveyed
habitat s, and since they reflect a range of consumed prey items
for members of the trophic levels listed above. For example, bar‐
racuda are known to be import ant predators of the selected fami‐
lies in our study region (Hansen, 2015). Large shark species found
in Biscayne Bay and Florida Bay, such as blacktip (Carcharhinus
limbatus), bull (Carcharhinus leucas), great hammerhead (Sphyrna
mokarran), and lemon sharks (Negaprion brevirostris), retain higher
trophic positions than many of the prey families and are known
fish predators (Gallagher, Shiffman, Byrnes, Hammerschlag‐Peyer,
& Hammerschlag, 2017; Hammerschlag, Luo, Irschick, & Ault,
2012; Matich, Heithaus, & Layman, 2011; Roemer, Gallagher, &
Hammerschlag, 2016). Bonnetheads (Sphyrna turbo) and Atlantic
sharpnose (Rhizoprionodon terraenovae) sharks may have varying
feeding patterns, but are primarily inshore feeders with diets con‐
si s tin g of tel e ost s, cru sta cea ns, an d ceph a lop ods (Plu mle e & Well s,
2016). Similarly, grunts, jacks, and snapper have been found inside
the stomachs of nurse sharks in Florida (Castro, 20 00). While there
is limited data on moray eel diet in our study area, work from other
Caribbean areas suggest s that they are piscivorous and readily
consume snappers or grunts (Randall, 1967; Young & Winn, 20 03).
The relative risk of each habitat where a BRUV was deployed
was estimated using t wo predator‐focused variables: (a) predator
abundance (maxNb and maxN) and (b) predator foraging activity.
Predator abundance was quantified for each trophic grouping
(maxNb) by tallying the number of distinctly different individuals,
determined by family, sex, size, and markings, observed throughout
the entire video duration (Bond et al., 2012). Additionally, a com‐
bined predator abundance was taken from each BRUV in the form
of maxN, which represents the maximum number of predators,
regardless of grouping, present together at one time (Bond et al.,
2012). We quantified predator foraging activity rates on the bait
bags by recording the number of bites from predators and whether
severe damage occurred to the bag (0 = no damage, 1 = severe
FIGURE 2 (a) The BRUV assembly; base 74 cm × 74 cm, slant height 72 cm, total height 48cm. (b) Still image captured from BRUV
deployment with a bonnethead shark (Sphyrna tiburo) in frame. (c) Still image captured from BRUV deployment with schooling yellow
snappers (Ocyurus chrysurus) and a southern stingray (Hypanus americanus) in frame
13744
|
PHENIX Et al.
damage). Bait bags were categorized as “severe damage” if the bag
had major lacerations or rips, or if the bag was totally removed from
the pole. Nonpredatory fish also have the potential to inflict dam‐
age to the bags (i.e., triggerfish), so any instances of damage to the
bags from nonpredatory fishes (ascer tained via video validation)
that could have confounded the detectability of our bait were not
included in these analyses.
Potential responses of prey species to ambient predation risk
were estimated using arrival times for each prey family (as a proxy
for apprehensiveness), as well as evaluating three prey‐focused be‐
haviors (burst swimming, schooling, and bait residency). Arrival time
(s) was measured by recording the total elapsed time until the first
individual from each prey family arrived on camera. Burst swimming
events (defined as a short, rapid swimming behavior away from the
frame; Gallagher, Brandl, & Stier, 2016; Gallagher, Lawrence, Jain‐
Schlaepfer, Wilson, & Cooke, 2016) and schooling events (defined as
instances where groups of five or more conspecific individuals were
present; Viscido, Parrish, & Grünbaum, 2005) were recorded for the
previously defined prey groups. Bait residency (sec) was evaluated
for each replicate as follows: the first fish, regardless of species, to
make contact with the bait was monitored until it had moved an esti‐
mated three or more body lengths distance from the bait bag (Bond
et al., 2019).
2.4 | Statistical analyses
Because data violated assumptions of normality and homogeneity
of variance (confirmed using Shapiro–Wilk's and Levene's tests),
we performed a zero‐inflated generalized linear model (GLM)
with a negative binomial error distribution and a log‐link func‐
tion to assess the ambient risk of each habitat , with habitat type
and its interaction with predator functional groups specified as
the independent variables and predator maxNb as the response
variable. Similarly, we per formed a GLM with a negative binomial
error distribution on prey arrival times, with the response vari‐
able being the arrival time of prey species and the independent
var iables bein g habitat type, preda tor maxN, and their interactio n.
Instances where an individual from a prey family did not appear
on the BRUV footage (i.e., not arriving) were excluded from the
model. Because this resulted in low replicates for some prey fish
species (e.g., rays), we did not specify prey species as an independ‐
ent variable and assumed that effects of predators are general‐
ized across all prey species. For both GLMs, we used the obtained
parameters for predictions and then plotted the predicted values
against the raw data to visualize both the obtained patterns and
the model fit.
Predator foraging activity and prey behaviors were then visual‐
ized using a nonmetric multidimensional scaling ordination (nMDS)
based on a Man hat tan distance. Further more, a PERM ANOVA was
run on the same distance matrix in order to determine if habitat
type, predator maximum abundance, or their interaction affected
prey behavior. Finally, we analyzed correlations bet ween predator
foraging and prey risk‐associated behaviors for each habitat using
a set of Spearman rank correlation analyses . All st atistical analyses
were performed using R Studio (R Core Team).
3 | RESULTS
A total of 194 deployments were made, within a total survey area
of ~15 km2. Of these, 37 deployments were discarded due to the
BRUV tipping over in heavy current or poor visibility, leaving a
total of 157 videos (n = 157) that were used in analyses ( Table 1).
A total of 184 predators were recorded by the BRUVs through‐
out the sampling period (Table 2). Of those predators, 80 indi‐
vidual elasmobranchs from eight species (7 shark species, 1 ray
species) were recorded, in addition to 88 barracuda and 16 moray
eels. There were limited seasonal differences in maximum preda‐
tor abundances (maxN) and prey arrival times across habitats, ex‐
cept for seagrass beds, where maximum predator abundance was
substantially higher in the wet season (0.690 ± 0.0.123 individuals,
mean ± SE) than in the dr y season (0.091 ± 0.063). In fact, no barra‐
cudas or large bodied mid‐trophic predators were observed in sea‐
grass habitats during the dry seasons. However, prey arrival times
in seagrass beds did not differ between the two seasons.
Predator abundances (maxNb) were significantly different
among habitat types, with coral reefs having the highest average
maximum number of predators per deployment (2.21 ± 2.04), fol‐
lowed by sea fan habitats, sand, and seagrass habitats (Table 3,
Table 4). Predictions from the GLM further suggest an interaction
effect between trophic level grouping and habitat. Coral reefs had
the greatest mean abundance of upper trophic and large bodied
mid‐trophic predators, whereas sea fan habitats had the greatest
mean abundances of small bodied mid‐trophic predators (Figure 3).
Prey arrival times were significantly influenced by the interactive
effects of habitat and the cumulative maximum number of preda‐
tors (maxN) (Table 5). Grunts, porgies, and snappers arrived com‐
paratively early at the BRUV deployments, while stingrays arrived
substantially later. The GLM revealed that the effect of maximum
predator numbers in sand, sea fan, and seagrass habitats are neg‐
ative and significantly different from effects of predators on coral
reefs, where cumulative predator maximum number and prey ar‐
rival time were positively correlated. This is further supported by
the predictions from the model, which show a steep negative re‐
lationship in sand and seagrass habitat s, a nearly flat but slightly
TABLE 1 BRUV deployments by season and habitat type
Habitat
Season
Dry (January–April)
Wet
(May–December)
Coral reef 4 15
Sea fan 934
Seagrass 22 30
Sand 16 27
|
13745
PHENIX Et a l.
negative relationship in sea fan habitats and a positive relationship
for coral reefs (Figure 4).
The nMDS ordination of bo t h pr e d a t o r for a ging ac t i v i t y (i.e., num ‐
ber of bites) and prey behavior in response to habitat type showed
little variation among habitats (Figure 5). Generalized predator for‐
aging activity was not significantly influenced by any habit at type,
although BRUVs deployed on coral reefs experienced the highest
average number of predatory bites (2.211 ± 3.441 bites, mean ± SE)
and instances of severe damage to the bait bag (0.263 ± 0.452 in‐
stances, mean ± SE). Prey burst swimming (4.579 ± 7.932 events)
and schooling events (6.053 ± 4.801 events) also had the highest
average occurrences on coral reefs when compared to sand, sea
fans, and seagrass habitats (Table 6). Average prey residency at the
bait was the greatest in sea fan habitats (32.211 ± 32.527 s). The
PERMANOVA to test the explanator y power of habitat, predator
maximum number, and their interac tion on different behaviors, al‐
beit revealing a significant habitat effect (p = .001), only explained
~10% of the variation in the data and no effect of predator maximum
number or its interaction with habitat was observed. The Spearman
rank correlation test showed significant correlations between pred‐
ator and prey behaviors in sand, seagrass, and sea fan habitats, but
not on coral reefs (Figure 6). Schooling behavior was the only one to
show a positive relationship with predator maximum numbers across
sand, seagrass, and sea fan habitats.
4 | DISCUSSION
Predator–prey interactions can structure marine habitats by ac‐
tively changing habitat use, foraging behaviors, and food‐web
dynamics (Morosinotto, Thomson, & Korpimäki, 2010). We pre‐
dicted that prey fishes would be more apprehensive and thus
arrive later in the field of view of the BRUV in habitats with in‐
creased predator abundance and vice versa in those with fewer
TABLE 2 Summary of predatory species observed on BRUVs in the present study
Upper trophic Large mid‐trophic Small mid‐trophic
Barracuda (Sphyraena sp.) 88 Green Moray (Gymnothorax
funebris)
4Atlantic Sharpnose (Rhizoprionodon
terraenovae)
14
Blacktip (Carcharhinus limbatus) 3 Nurse (Ginglymostoma cirratum)22 Blacknose (Carcharhinus acronotus) 3
Bull (Carcharhinus leucas) 2 Sawfish (Pristis pectinata) 1 Bonnethead (Sphyrna turbo)34
Great Hammerhead (Sphyrna
mokarran)
1 Spotted Moray (Gymnothorax moringa)12
Tot al 94 27 63
Upper trophic Large mid‐trophic Small mid‐trophic Max Nb
Coral reef 1.16 (±0.384) 0.474 (±0.140) 0.579 (±0.318) 2.21 (±2.04)
Sand 0.674 (±0.169) 0.093 (±0.045) 0.140 (±0.053) 0.907 (±1.231)
Sea fan 0.581 (±0.245) 0.256 (±0.067) 0.721 (±0.206) 1.56 (±1.94)
Seagrass 0.346 (±0.095) 0.058 (±0.033) 0.288 (±0.092) 0.692 (±1.15)
TABLE 3 Mean predator abundance
per BRUV deployment across the four
habitat t ypes (coral reef, sand, sea fan, and
seagrass), decomposed into the different
trophic levels and their combined
abundance (MaxNb)
Coefficients Estimate SE Z value Pr (>|z|)
Intercept (coral reef:large
mid‐trophic)
−9. 041 0.43 −21. 03 ***
Sand −1. 5 98 0.685 −2.3 3 *
Sea fan −0. 659 0.555 −1. 19 ns
Seagrass −2. 137 0.739 −2.8 9 **
CR Upper trophic 0.886 0. 551 1 .61 ns
Small mid‐trophic 0.2 52 0.591 0.43 ns
SD Upper trophic 1.993 0 .594 3.35 ***
Small mid‐trophic 0.435 0. 697 0.62 ns
SF Upper trophic 0.82 0.443 1.85 .
Small mid‐trophic 1.016 0.434 2.34 *
SG Upper trophic 1.766 0.667 2.65 **
Small mid‐trophic 1.594 0 .675 2.36 *
Note: Significant codes: ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1.
TABLE 4 Summary results from a zero‐
inflated negative binomial generalized
linear model used to test the effects
of habitat t ype on predator abundance
(maxNb) by trophic level
13746
|
PHENIX Et al.
predators. Our results suggest that this pattern held true only for
coral reefs, where predator numbers appeared to have a negative
effect on prey arrival, while in all other habitat s, the two vari‐
ables were positively correlated. While coral reefs offer increased
structural complexit y and refuge for prey, they also increase po‐
tential predation risk by obscuring prey fish's field of view (Bond
et al., 2019). These components of the habitat may provide preda‐
tors with a functional advantage when hunting, thereby creating
a more dangerous environment and increasing prey vigilance in
these areas. Thus, the interaction between habitat features and
the probability of predator detection and successful escape can
result in altered prey risk‐associated behaviors and vigilance
(Heithaus et al., 2009). It has been recently argued that predators
may exact greater influences on prey behavior where predation
risk is predictable (Creel, 2018). While we did not measure pre‐
dictability of predation risk in our study, abundance of predators
in certain habitats, a potential prox y for exposure, may have re‐
sulted in a pro‐active response of apprehensiveness toward the
bait, although this remains speculative.
Pre dators are know n to ma tch p rey dis tr i b u t ions on small sc ales
when prey is abundant (Heithaus & Dill, 2006), and, as observed
in the present study, coral reefs generally contain high numbers of
piscivores (Hixon & Beets, 1993), which can inversely affect prey
abundance on reefs (Beukers‐Stewart, Beukers‐Stewart, & Jones,
2011). On average, grunts and snappers arrived on coral reefs and
in sea fans long before any predators. Whether predation risk is
“predictable” or chronic on coral reefs remains unknown, but our
findings offer an interesting potential link to the predicted food‐
risk effects as described in the “control of risk” hypothesis (Creel,
2018).
Animals often express their antipredator‐behaviors in high
risk situations that are brief and infrequent (Lima & Bednekoff,
1999). These acute “reactive” responses are linked to areas of
FIGURE 3 Mean predicted predator abundance (±95% confidence inter vals) from a zero‐inflated negative binomial GLM across four
habitat t ypes: coral reef (CR), sand (SD), sea fan (SF), and seagrass (SG). Predicted predator abundance values, as well as mean predicted
abundance by habitat (dashed lines) are overlaid on top of raw observational data
Upper TrophicLarge Mid-TrophicSmall Mid-Trophic
TABLE 5 Summary results of a negative binomial generalized
linear model of the effects of habitat type on maximum combined
predator abundance (maxN)
Coefficients Estimate SE Z value Pr (>|z|)
Intercept (coral
reef)
5.865 0.156 37.6 ***
Sand 1.331 0 .172 7. 73 ***
Sea fan 0.463 0.174 2.66 **
Seagrass 1.035 0.168 6.18 ***
maxN 0.167 0.10 0 1. 67 ns
Sand: maxN −0.601 0.141 −4.26 ***
Sea fan: maxN −0.247 0 .114 −2 .17 *
Seagrass: maxN −0.366 0.130 −2. 8 2 **
|
13747
PHENIX Et a l.
FIGURE 4 Predicted mean prey arrival time (y‐axis) as a function of maximum combined predator abundance (x‐axis) across four habitat
types based on a negative binomial GLM. Predicted fits (±95% confidence intervals) are overlaid on top of raw observational data of seven
prey families across four habitat types. CR, coral reef; SD, sand; SF, sea fan; SG, seagrass
0
1,000
2,000
3,000
4,000
02468
Predicted arrival time of prey (in seconds elapsed 95% CI)
Filefish Jack SnapperGrunt Tr iggerfish
Porgy Ray
Coral reef
Sand
Seafan
Seagrass
FIGURE 5 Multidimensional ordination
of predator foraging activit y (predator
bites and bait damage) and prey risk‐
associated behaviors (burst swimming,
schooling, and prey residency) across four
habitat t ypes. Colors match the previously
used habitat‐specific colors
−1.5 −1.0 −0.5 0.00.5 1.
01
.5
−1.0 −0.5 0.00.5
NMDS1
NMDS2
pred.bites
prey.res
schooling
burst.swm
13748
|
PHENIX Et al.
unpredictable predation risk (Creel, 2018). We hypothesized that
potential behavioral risk effects might be highest in open areas.
Interestingly, we obser ved faster prey arrival times in more open,
homogenous habitats such as sea fans, sandy areas, and seagrass.
The lack of resources in these open, plain habitats may have ren‐
dered our BRUVs a more attractive source of food, resulting in
both prey and predators arriving sooner; in our study, we found
that grunts and snappers were much quicker to arrive to a habi‐
tat where predators were more abundant (Nagelkerken & Velde,
2004). It is also possible that these open and homogenous habitats
provide increased escape routes to prey if needed, thus making
the m worth the “risk.” Addition ally, since predators are often tran‐
sient in these habitats (Hammerschlag, Morgan, & Serafy, 2010),
attacks may be less predictable. Therefore, our obser ved patterns
for behavioral effects in these habitats may stem from a combi‐
nation of resource provisioning and unpredictability of predation
risk.
In general, juvenile and small bodied sharks (i.e., small mid‐tro‐
phic predators) can be found in shallow waters to minimize their own
preda tion risk (Gu tt ri dge et al., 2012; He up el et al., 2014). Mor e th an
half of the sharks captured on the BRUVs were species that reach
maximum sizes of <2 m. While it stands to reason that smaller pred‐
ators induce a weaker response in prey than larger conspecifics or
species (due to gape limitations), smaller mesopredators (hawkfish,
Parrachirrhites arcatus) have been found to have equal nonconsump‐
tive effects compared to larger conspecifics (Gallagher, Brandl, et al.,
2016; Gallagher, Lawrence, et al., 2016). Most predators (regardless
of trophic grouping) in our videos did not stay for prolonged periods
of time and, as such, they represent an acute, but relatively inconsis‐
tent, pulsed source of predation risk. Finally, some small species (e.g.,
bonnetheads) may also have limited ef fects on prey because both ju‐
veniles and adults primarily feed on crabs, lobsters, and cephalopods
(Bethea et al., 2007).
The extrapolation of our results beyond our study design is hin‐
dered by several caveat s. Firstly, we do not know whether arrival
times are truly a consequence of perceived predation risk or if they
are a function of var ying densities of individuals which could not
be controlled. We also did not measure water currents at each of
TABLE 6 Mean predator foraging activity (bites and severe damage) and prey response behavior (burst swimming, schooling, and
residency) across four habitat types
Predator bites Severe damage Burst swimming Schooling Prey residency
Coral reef 2.211 (±3.441) 0.263 (±0.452) 4.579 (±7.324) 6.052 (±4.8 01) 24.316 (±18.973)
Sand 0.791 (±1.684) 0.070 (±0.259) 0.698 (±3.377) 1.395 (±2.555) 8.814 (±17.14)
Sea fan 1.558 (±4.078) 0.136 (±0.351) 1.605 (±3.13) 4. 628 (±4.232) 32.211 (±32.527)
Seagrass 0.865 (±2 .360) 0.096 (±0.298) 0.745 (±1.741) 2 .980 (±3.906) 20.192 (±27.652)
FIGURE 6 Correlation plot of prey risk
behaviors (burst swimming, schooling,
and prey residency) compared to predator
foraging activity (bites and damage) across
four habit at types
(c) Sea fan (n = 43)
(a)
Coral reef (n = 19)(b) Sand (n = 43)
(d) Seagrass (n = 52)
|
13749
PHENIX Et a l.
our BRUV stations, which could have affected the bait dispersal at
different rates, thus changing detection potential by prey species.
Furthermore, our statistical power was weakened by poor visibility
(resulting in the exclusion of 37 replicates) and a category 5 hur‐
ricane, which ended data collection a bit early and thus prevented
extended sampling. In future studies, dusk or night time deploy‐
ments should be added to observe predator–prey interactions after
dark, which may be especially important for sharks on coral reefs
(Hammerschlag et al., 2017).
The role of “apex”‐predators on reefs has been brought into
question in recent years (see Roff et al., 2016). While we caution
overextending the results of this study to other regions, our dat a
suggest that predators regardless of their trophic position do
not significantly control mobile prey behavior on short temporal
scales, across habitats. Instead, a habitat‐specific response to a
consistent signal of mobile predators on reefs may result in pro‐
active prey vigilance and subtle food‐risk trade‐offs. Specifically,
less complex habitats where predators are known to patrol yet re‐
main temporally unpredictable in their occurrence due to limited
numbers and potentially wider activity areas may induce different
reactive behavioral effects such as schooling and burst swimming,
which, when extended over larger time scales, could have meta‐
bolic and fitness‐level impacts on prey. Taken together, these re‐
sults suggest that context is important when trying to disentangle
the effects of top predators on prey in costal marine habitats, and
future studies should examine the interactions between mobile
predators and habit at in order to link predation risk theory to
observations.
ACKNOWLEDGMENTS
This work was supported by funding to Beneath the Waves from the
Herber t W. Hoover Foundation, as well as from C. and M. Jones. We
are grateful to M. Riera, E. Pritchard, C. Perry, R. Tricarico, Shake‐A‐
Leg, and the International Seakeepers Society for their assistance
with this project. This work was conducted under a Biscayne
National Park permit BISC‐0 0076 to AJG.
CONFLICT OF INTEREST
The authors declare no competing interests.
AUTHOR CONTRIBUTIONS
A. J. G., M. E. B., and S. J. B. conceived and designed the study. L. M.
P., D. T., E. Q., and A. J. G. conduced the field work. L. M. P. and S.
J. B. performed the analyses. All authors contributed to writing the
manuscript and gave approval.
ORCID
Lindsay M. Phenix https://orcid.org/0000‐0001‐7462‐878X
Simon J. Brandl https://orcid.org/0000‐0002‐6649‐2496
Austin J. Gallagher https://orcid.org/0000‐0003‐1515‐3440
DATA AVAIL ABI LIT Y S TATEM ENT
All data are deposited and available in Dryad at the following address
https ://doi.org/10.5061/dryad.wm37p vmh5.
REFERENCES
Armstrong, J. D., Bagley, P. M., & Priede, I. G. (1992). Photographic and
acoustic tracking observations of the behaviour of the grenadier
Coryphaenoides (Nematonurus) armatus the eel Synaphobranchus
bathybius, and other abyssal demer sal fish in the North Atlantic
Ocean. Marine Biology, 112(4), 535–544. ht tps ://doi.org /10.1007/
BF00 3 4 6170
Beauchamp, D. A., Wahl, D., & Johnson, B. M. (2007). Predator‐prey
interac tions. In C. S. Guy, & M. J. Brown (Eds.), Analysis and inter-
pretation of freshwater fisheries data (pp. 765–842). Bethesda, MD:
American Fisheries Society.
Bethea, D. M., Hale, L., Carlson, J. K., Cortés, E., Manire, C. A., &
Gelsleichter, J. (2007). Geogr aphic and ontogenetic variation in the
diet and daily r ation of the bonnethead shark, Sphyrna tiburo, from
the eastern Gulf of Mexico. Marine Biology, 152(5), 1009–1020. https
://doi.org/10.1007/s00227‐007‐0728‐7
Beukers‐Stewart, B. D., Beukers‐Stewart, J. S., & Jones, G. P. (2011).
Behavioural and developmental responses of predatory coral
reef fish to variation in the abundance of prey. Coral Reefs, 30(3),
855–864.
Bond, M. E., Babcock, E. A., Pikitch, E. K., Abercrombie, D. L., Lamb, N.
F., & Chapman, D. D. (2012). Reef shar ks exhibit site‐fidelity and
higher relative abundance in marine reserves on the Mesoamerican
Barrier Reef. PLoS ONE, 7(3), e32983. ht tps ://doi.org /10.1371/journ
al.pone.0032983
Bond, M. E., Valentin‐Albanese, J., Babcock, E. A., Heithaus, M. R.,
Grubbs, R. D., Cerrato, R., … Chapman, D. D. (2019). Top pred‐
ators induce habit at shifts in prey within marine protected
areas. Oecologia, 190(2), 375–385. ht tps ://doi.or g/10.1007/
s00442‐019‐04421‐0
Bond, M. E., Valentin‐Albanese, J., Babcock, E. A., Hussey, N. E., Heithaus,
M. R., & Chapman, D. D. (2018). The trophic ecology of Caribbean
reef sharks (Carcharhinus perezi) relative to other large teleost pred‐
ators on an isolated coral atoll. Marine Biology, 165(4), 67. https ://doi.
org /10.1007/s00227‐018‐3322‐2
Bruno, J. F., Stachowicz, J. J., & Bertness, M. D. (2003). Inclusion of fa‐
cilitation into ecological theor y. Trends in Ecology & Evolution, 18(3),
119–125. https ://doi.or g/10.1016/S0169‐53 47(02)000 45‐9
Casey, J. M., Baird, A . H., Brandl, S. J., Hoogenboom, M. O., Rizzari, J. R.,
Frisch, A. J., … Connolly, S. R. (2017). A test of trophic casc ade the‐
ory: Fish and benthic assemblages across a predator density gradient
on coral re efs. Oecologia, 183 (1), 161–175. ht tps ://doi.or g/10.1007/
s0 04 42‐ 016 ‐375 3‐8
Castro, J. I. (2000). The biolog y of the nur se shark, Ginglymostoma cirra‐
tum, off the Florida east coast and the Bahama Islands. Environmental
Biology of Fishes, 58 (1), 1–22.
Catano, L. B., Barton, M. B., Boswell, K. M., & Burkepile, D. E. (2017).
Predator identity and time of day interact to shape the risk–reward
trade‐off for herbivorous coral reef fishe s. Oecologia, 183 (3), 763–
773 . http s ://doi.or g/10.1007/s 00 4 42‐ 016 ‐37 94‐z
Creel, S. (2018). The control of risk hypothesis: Reactive vs. proactive
antipredator responses and stress‐mediated vs. food‐mediated costs
13750
|
PHENIX Et al.
of response. Ecology Letters, 21, 947–956 . ht tp s ://doi. org/10.1111/
ele.12975
Estes, J. A., Terborgh, J., Brashares, J. S., Power, M. E., Berger, J., Bond,
W. J., … Marquis, R. J. (2011). Trophic downgrading of planet earth.
Science, 333(6040), 301–306.
Ferretti, F., Worm, B ., Britten, G . L ., Heithaus, M. R., & Lotze, H.
K. (2010). Patterns and ecosystem consequences of shark de‐
clines in the ocean. Ecology Letters, 13(8), 1055–1071. https ://doi.
org /10.1111/j.1461‐0 248 .2 010.014 89.x
Frisch, A . J., Ireland, M ., Rizzari, J. R ., Lönnstedt, O. M., Magnenat, K. A.,
Mirbach, C. E., & Hobbs, J. P. A. (2016). Reassessing the trophic role
of reef shar ks as apex predators on coral reefs. Coral Reefs, 35(2),
459–472. htt ps ://doi.org /10.10 07/s00338‐016 ‐1415‐2
Gallagher, A. J., Br andl, S. J., & Stier, A. C. (2016). Intraspecific variation
in body size does not alter the effects of mesopredators on prey.
Royal Societ y Open Science, 3(12), 160 414. https ://doi.or g/10.1098/
rsos.160414
Gallagher, A. J., Kyne, P. M., & Hammerschlag, N. (2012). Ecological risk
assessment and its application to elasmobranch conservation and
management. Journal of Fish Biology, 80(5), 1727–1748.
Gallagher, A. J., Lawrence, M. J., Jain‐Schlaepfer, S. M., Wilson, A. D., &
Cooke, S. J. (2016). Avian predators transmit fear along the air–water
interf ace influencing prey and their parent al care. Canadian Journal
of Zoology, 94(12), 863–870. https ://doi.org/10.1139/cjz‐2016‐0164
Gallagher, A . J., Shiffman, D. S., Byrnes, E. E., Hammerschlag‐Peyer, C.
M., & Hammerschlag, N. (2017). Pat terns of resource use and isoto‐
pic niche overlap among three species of sharks occurring within a
protected subtropical estuar y. Aquatic Ecology, 51(3), 435–448.
Guttridge, T. L., Gruber, S. H., Franks, B. R., Kessel, S. T., Gledhill, K. S.,
Uphill, J., … Sims, D. W. (2012). Deep danger: Intra‐specific preda‐
tion risk influences habitat use and ag gregation formation of juvenile
lemon sharks Negaprion brevirostris. Marine Ecology Progress Series,
445, 279–291. https ://doi.org/10.3354/meps0 9423
Hammerschlag, N., & Gallagher, A. J. (2017). Extinction risk and con‐
servation of the earth's national animal symbols. BioScience, 67(8),
744–749. https ://doi.org/10.1093/biosc i/bix054
Hammerschlag, N., Luo, J., Irschick, D. J., & Ault, J. S. (2012). A compar‐
ison of spatial and movement patterns between sympatric preda‐
tors: Bull sharks (Carcharhinus leucas) and Atlantic tarpon (Megalops
atlanticus). PLoS ONE, 7(9), e45958. https ://doi.org/10.1371/journ
al.pone.0045958
Hammerschlag, N., Morgan, A. B., & Serafy, J. E. (2010). Relative pre‐
dation risk for fishes along a subtropical mangrove–seagrass ec‐
otone. Marine Ecology Progress Series, 401, 259–267. https ://doi.
org/10.3354/meps0 8449
Hammerschlag, N., Skubel, R . A., Calich, H., Nelson, E. R., Shiffman, D.
S., Wester, J., … Gallagher, A. J. (2017). Nocturnal and crepuscular
behavior in elasmobranchs: A review of movement, habitat use, for‐
aging, and reproduction in the dark. Bulletin of Marine Science, 93(2),
355 –374 .
Hansen, N. R. (2015). Feeding ecology and habitat utilization of the Great
Barracuda Sphyraena barracuda (Edwards 1771) in Southeast Florida.
Nova Southeastern Univer sity MS Thesis.
Heithaus, M. R., Dill, L. M. (2006). Does tiger shark predation
risk influence foraging habitat use by bottlenose dolphins at
multiple spatial scales? Oikos, 114(2), 257–264. https ://doi.
org /10.1111/j.200 6.00 30 ‐1299.14 443. x
Heithaus, M. R., Frid, A., Wirsing, A. J., & Worm, B. (20 08). Predicting
ecological consequences of marine top predator declines. Trend s
in Ecology & Evolution, 23(4), 2 02–210. ht tps ://doi.or g/10.1016/j.
tree.2008.01.003
Heithaus, M. R., Wirsing, A. J., Burkholder, D., Thomson, J., &
Dill, L. M. (2009). Towards a predictive framework for preda‐
tor risk ef fect s: The interaction of landscape features and prey
escape tactics. Journal of Animal Ecology, 78 (3), 556–562. https ://doi.
org /10.1111/j.1365 ‐2656 .2 00 8.01512.x
Heupel, M. R., Knip, D. M., Simpfendorfer, C . A., & Dulvy, N. K. (2014).
Sizing up the ecological role of sharks as predators. Marine Ecology
Progress Series, 495, 291–298. https ://doi.org/10.3354/meps1 0597
Hixon, M. A., & Be et s, J. P. (1993). Pred ation , prey refuges, and the struc‐
ture of coral‐reef fish assemblages. Ecological Monographs, 63(1),
77–10 1 .
Hussey, N. E., MacNeil, M. A ., McMeans, B. C., Olin, J. A., Dudley, S. F.,
Cliff, G., … Fisk, A. T. (2014). Rescaling the trophic struc ture of marine
food webs. Ecology Letters, 17(2), 239–250. ht tps ://doi.o rg /10.1111/
ele.12226
Hutchinson, G. E. (1957). Concluding remark s. Cold Spring Harbor
Symposia on Quantitative Biology, 22, 415–427.
Klages, J., Broad, A., Kelaher, B. P., & Davis, A. R. (2014). The influence of
gummy sharks, Mustelus antarcticus, on observed fish assemblage
structure. Environmental Biology of Fishes, 97(2), 215–222. https ://
doi.org /10.10 07/s10641‐013‐013 8‐2
Layman, C. A., Winemiller, K. O., Arring ton, D. A., & Jepsen, D. B. (20 05).
Body size and trophic position in a diverse tropic al food web. Ecology,
86(9), 2530–2535. https ://doi.org/10.1890/04‐1098
Lennox, R. J., Gallagher, A. J., Ritchie, E. G., & Cooke, S. J. (2018).
Evaluating the eff icacy of predator removal in a conflict‐prone world.
Biological Conservation, 2 24, 27 7–28 9. h t tps ://d oi . or g /1 0. 10 16/j .
biocon.2018.05.003
Lima, S. L., & Bednekoff, P. A. (1999). Temporal variation in danger
drives antipredator behavior: The predation risk allocation hy‐
pothesis. The American Naturalist, 153(6), 649–659. https ://doi.
org /10.108 6/30 3202
Madin, E. M. , Ga in es , S. D., & Warner, R. R. (2 010 ). Fie ld evi de nce for per‐
vasive indirect ef fect s of fishing on prey foraging behavior. Ecolog y,
91(12), 3563–3571 . ht tps ://doi.org /10.18 90/0 9‐2174.1
Matich, P., Heithaus, M. R ., & Layman, C. A . (2011). Contrasting patterns
of individual specialization and trophic coupling in two marine apex
predators. Journal of Animal Ecology, 80(1), 294–305.
Morosinotto, C., Thomson, R. L., & Korpimäki, E. (2010). Habit at selec‐
tion as an antipredator behaviour in a multi‐predator landscape: All
enemies are not equal. Journal of Animal Ecology, 79(2), 327–333.
https ://doi. org/10.1111/j .1365‐2656.2009.01638.x
Nagelkerken, I., & Van der Velde, G. (2004). Relative importance of inter‐
linked mangroves and seagrass beds as feeding habitats for juvenile
reef fish on a Caribbean island. Marine Ecology Progress Series, 274,
153–159. https ://doi.org/10.3354/meps2 74153
Plumlee, J. D., & Wells, R. D. (2016). Feeding ecolog y of three coastal
shark species in the nor thwest Gulf of Mexico. Marine Ecolog y
Progress Series, 550, 163–174. https ://doi.org/10.3354/meps1 1723
Randall, J. E. (1967). Food habits of reef fishes of the West Indies. Studies
of Tropical Oceans, 5, 665–847.
Rasher, D. B., Hoey, A. S., & Hay, M. E. (2017). Cascading predator effec ts
in a Fijian coral reef ecosystem. Scientific Repor ts, 7, 15684.
Rizzari, J. R., Frisch, A. J., Hoey, A . S., & McCormick, M. I. (2014). Not
worth the risk: apex predators suppress herbivor y on coral reefs.
Oikos, 123(7), 829–836.
Roemer, R. P., Gallagher, A. J., & Hammer schlag, N. (2016). Shallow water
tidal flat use and associated specialized foraging behavior of the
great hammerhead shark (Sphyrna mokarran). Marine and Freshwater
Behaviour and Physiology, 49(4), 2 35 –24 9.
Roff, G., Doropoulos, C., Rogers, A., Bozec, Y. M., Krueck, N. C., Aurellado,
E., … Mumby, P. J. (2016). The ecological role of sharks on coral reefs.
Trends in Ecology & Evolution, 31(5), 395–407.
Ruppert, J. L., Travers, M. J., Smith, L. L., Fortin, M. J., & Meekan, M. G.
(2013). Caught in the middle: Combined impacts of shark removal
and coral loss on the f ish communities of coral reef s. PLoS ONE, 8(9),
e74648. https ://doi.org/10.1371/journ al.pone.0074648
|
13751
PHENIX Et a l.
Sandin, S. A., Smith, J. E., DeMartini, E. E., Dinsdale, E. A., Donner, S. D.,
Friedlander, A. M., … Sala, E. (2008). Baselines and degradation of
coral reefs in the Northern Line Islands. PLoS ONE, 3(2), e1548. https
://doi.org/10.1371/journ al.pone.0001548
Serafy, J. E., Valle, M ., Faunce, C. H., & Luo, J. (2007). Species‐specific
patterns of fish abundance and size along a subtropical mangrove
shoreline: An application of the delta approach. Bulletin of Marine
Science, 80(3), 60 9–624.
Suraci, J. P., Clinchy, M., Dill, L. M., Roberts, D., & Zanette, L. Y. (2016). Fear
of large carnivores causes a trophic c ascade. Nature Communications,
7, 10698. http s ://doi.org/10.1038/ncomm s10698
Vis cido, S. V., Pa rr ish, J. K., & Grü nbaum , D. (2 00 5). The effect of po pula‐
tion size and number of influential neighbors on the emergent prop‐
erties of fish schools. Ecological Modelling, 183 (2–3), 347–363. https
://doi.org/10.1016/j.ecolm odel.2004.08.019
Whitmarsh, S. K., Fairweather, P. G., & Huveneers, C. (2017). What
is Big BRUVver up to? Methods and uses of baited under water
video. Reviews in Fish Biology and Fisheries, 27(1), 53–73. https ://doi.
org/10.1007/s11160‐ 016‐945 0‐1
Young, R. F., & Winn, H. E. (2003). Activity patterns, diet, and shelter
site use for t wo species of moray eels, Gymnothorax moringa and
Gymnothorax vicinus, in Belize. Copeia, 1, 4 4–55. https ://doi.org/10.
1643/0045‐8511(2003)003[0044:APDAS S]2.0.CO;2
How to cite this article: Phenix LM, Tricarico D, Quintero E,
Bond ME, Brandl SJ, Gallagher AJ. Evaluating the effects of
large marine predators on mobile prey behavior across
subtropical reef ecosystems. Ecol Evol. 2019;9:13740–13751.
https ://doi.org/10.1002/ece3.5784
Available via license: CC BY
Content may be subject to copyright.
