Healthy coral reefs are biologically diverse and provide vital ecosystem services. However, decreasing water quality and global warming are key contributors to coral reef decline, which poses substantial environmental threats. In response to this degradation, an innovative coral reef restoration technology, called Biorock, utilizes weak direct current electric fields to cause limestone deposition on conductive materials, inevitably inducing prolific coral reef growth. Although expediting coral growth, research on how the associated electric fields may impact the behavioural patterns of teleosts and/or organisms (i.e. elasmobranchs) possessing electroreception capabilities is lacking. Therefore, we studied the behavioural responses of two shark species, the bull shark (Carcharhinus leucas) and the Caribbean reef shark (Carcharhinus perezi) and multiple teleost species towards weak direct current electric fields in Bimini, Bahamas. Generalized linear mixed model analyses based on 90 trials illustrate that both the feeding and avoidance behaviors of C. leucas and C. perezi were significantly associated with treatment type, with the weak experimental electrode treatments resulting in the greatest quantity of avoidances and fewest feedings for both species. However, data analyses illustrate that teleost feeding behavior was not observably impacted by experimental treatments. Although the Biorock technology exhibits promise in coral reef restoration, the findings from this study illustrate a need for future large-scale studies assessing shark behavioral patterns around these devices, since the deterrence of apex predators may impact ecosystem balance.
Although coral reefs are among the most threatened ecosystems in the world, they are among the most diverse, second only to tropical forests (Connell, 1978; Bryant et al., 2000). Coral reefs occupy an area of approximately 225 000 to 284 000 km2, which corresponds to only 0.09% to 0.17% of the area of the oceans (Spalding, 1997, 2000; Spalding & Grenfell, 1997; Burke et al., 2001). This biodiverse ecosystem is characterized by exceptionally high productivity (Birkeland, 1997; Bryant et al., 2000; Souter & Linden, 2000) in contrast to the oligotrophic open ocean environment (Bryant et al., 2000). Despite their economic importance from a global perspective, coral reefs have deteriorated in the last 50 years, largely from anthropogenic impacts such as: global warming, overfishing, industrial and domestic pollution, and increased sedimentation (Goreau & Hayes, 1994; Downs et al., 2005; Leão & Kikuchi, 2005; Bruno & Valdivia, 2016).
One coral reef restoration measure is known as Biorock reef restoration technology (Goreau, 2012, 2014). This technology utilizes electric fields ranging from 1.2 and 12.0 V/m to facilitate limestone deposition on conductive materials. The minerals, mainly aragonite (CaCO3) and brucite (Mg(OH)2), are similar in chemical and physical properties to natural coral skeleton limestone, which makes up the bulk of coral reefs. These structures have been demonstrated to have six times higher abundance, higher diversity, and greater evenness of fishes than nearby control reefs (Bakti et al., 2012). However, although teleost-based surveys were conducted, the responses of elasmobranchs (i.e. sharks, skates, and rays) to the technology have not been documented.
Elasmobranch (sharks, skates, and rays) possess a unique electro-sensory system known as the ampullae of Lorenzini (e.g. Kalmijn, 1971, 1982). This sensory system is composed of minute gel-filled pores and has been demonstrated to be used to detect the minute bioelectric fields associated with prey (e.g. Kajiura & Holland, 2002). The mechanism of bioelectric field detection is based around voltage gradients. Each ampullae surface pore leads to a subcutaneous canal filled with conductive jelly (Kalmijn, 1966, 1971, 1974, 2000; Bastian, 1994), which was recently demonstrated to be one of the most effective biological proton conductors (Josberger et al., 2016). Upon encountering a weak electrical stimulus, such as that produced by prey, a voltage gradient is produced between the external pore surface and the internal pore. This differential elicits a neurological impulse, which is sent to the brain where the stimulus is perceived (Kalmijn, 1974, 1982, 1984). Weak electrical stimuli (e.g. 5 nV/cm) have been demonstrated to elicit foraging behavior (Kajiura & Fitzgerald, 2009), whereas stronger electrical stimuli ranging (e.g. 3-7 V/m) have been demonstrated to elicit deterrent responses (Smith, 1966, 1973, 1974; Gilbert & Gilbert, 1973; Smit & Peddemors, 2003; Marcotte & Lowe, 2008; Huveneers et al., 2012). Although stronger electrical stimuli have been observed to elicit deterrent responses, some studies suggest that deterrent-mediated behaviors may be species-specific (Marcotte & Lowe, 2008) and may also vary on a context-specific basis (Huveneers et al., 2012).
Due to their sensitivity to weak (e.g. Kajiura & Fitzgerald, 2009) and strong localized (e.g. Marcotte & Lowe, 2008; Huveneers et al., 2012) electric fields, the electric fields (i.e. 1.2-12.0 V/m) associated with Biorock may elicit deterrent responses in various shark species. Therefore, we aimed to assess how bull shark (Carcharhinus leucas) and Caribbean reef shark (Carcharhinus perezi) behaviors may or may not be affected by exposure to strong and localized electrical fields. In addition, this study aimed to determine if teleost behavior was observably impacted. Based on previous research (Marcotte & Lowe, 2008; Huveneers et al., 2012), we hypothesized that the Biorock-associated electric fields may be sufficiently strong to elicit deterrent responses (e.g. cessation of feeding and increased avoidance behavior) in sharks; however, due to their lack of ampullary pores, teleosts should exhibit no significant behavioral variation in response to electrical fields.
Material and methods
This study was conducted from February to March 2015 in Bimini, Bahamas. Research was conducted at two locations, Triangle Rocks: 25°38′48.23″N, 79°18′44.68″W and Alicetown Channel: 25°43′32.58″N, 79°17′45.90″W. Study sites ranged from 2-8 m deep and were characterized as a region containing sandy substrate (Alicetown Channel) or a coral reef ecosystem (Triangle Rocks). Study sites were selected based on the reliable presence and feasibility of working with the target species. More specifically, Caribbean reef shark (Carcharhinus perezi) behavior was assessed at Triangle Rocks, whereas bull shark (Carcharhinus leucas) behavior was studied in Alicetown Channel. All research was conducted in accordance with the rules and regulations of the assigned Bahamas Department of Marine Resources permit (MAF/FIS/17).
To assess the effects of Biorock-associated electric fields, three experimental treatments were constructed: control, procedural control, and experimental electrode. The control treatment was an experimental baseline trial that lacked any apparatus. The procedural control treatment contained electrodes that were not connected to a power source, and was deployed to assess if the visual stimuli associated with the electrode apparatus were sufficient to elicit attractant or deterrent behavior. Lastly, the experimental electrode treatment contained electrically charged electrodes to assess if Biorock-associated electric fields influenced teleost and elasmobranch behavior.
For the experimental electrode treatment, the electrode apparatus was constructed out of 160 cm (length) by 5 cm (diameter) polyvinyl chloride (PVC) pipe with two electrodes spaced by 1 m, the cathode and anode. The cathode (negative dipole), a 60 cm long, 9 mm diameter steel reinforcing bar connected to a black electrical cable, was securely attached to the PVC pipe. The anode (positive dipole), a 20 cm wire mesh, was connected to a red electrical cable via a small PVC pipe filled with epoxy resin (fig. 1). The long PVC pipe was connected to a piece of wood (65 cm × 10 cm) to maximize structural integrity of the experimental apparatus and make the electrodes float below the water surface. A nylon rope attached to the center of the apparatus allowed it to be lowered from the dock or boat for deployment and retrieval during the trials associated with the procedural control and experimental electrode treatments. The electrodes were powered by a 6 V DC power source that was in-series with a Power Probe CAT IV Multimeter DMM101ES, and a clamp ammeter to measure generated voltage and current.
At each study site, treatments were deployed independently. Using a random number generator through Microsoft Excel, the temporal sequence of treatments (i.e. control, procedural control, and experimental electrode treatment) were randomly selected. The randomization was essential to eliminate the possibility of order effects (Hurlbert, 1984). To commence experimentation, an HD 1080p GoPro camera was placed directly above the treatments to document behavior and determine conspecific density. In addition, the olfactory stimuli (i.e. fine mesh bag filled with minced fish) and experimental apparatus were positioned, and subsequently, bait deployment commenced. Throughout each 6 minute and 15 second trial associated with each individual treatment, one 0.028 kg piece of bait was thrown in the water every 15 seconds approximately one meter from the experimental apparatus. A total of 25 pieces of bait were deployed for each trial, or 75 pieces of bait per replicate (e.g. all three treatments). In addition, there was a three-minute inter-trial duration, with all experimental materials being removed from the water at the start of this duration and re-deployed for the start of the next trial. Every trial contained equal quantities of bait from a particular species to eliminate any preference-based behavior. The species used for bait were: great barracuda (Sphyraena barracuda), wahoo (Acanthocybium solandri), striped grunt (Haemulon striatum), French grunt (Haemulon flavolineatum), mangrove snapper (Lutjanus griseus), crevalle jack (Caranx hippos), king mackerel (Scomberomorus cavalla), and Atlantic sardines (Clupea pilchardus).
During each trial, the maximum number of sharks within view of the apparatus and within a given trial was quantified to assess conspecific density, which was then categorized as low (1-3 sharks; categorized as ‘1’), medium (4-7 sharks; categorized as ‘2’), or high (8+ sharks; categorized as ‘3’). However, to determine overall sample sizes, distinctive characteristics (e.g. presence/absence scars, tags, pigmentation patterns, and dorsal fin notch characteristics) were used to estimate total shark quantity and whether an individual returned over the experimental timeframe. Since teleost quantity varied throughout each trial and individual identification was difficult, the maximum quantity that could be observed within one video frame was used for each teleost species to create an estimate of sample size. In addition to conspecific density, water visibility was also recorded as either low (i.e. could not see the seafloor; categorized as ‘1’), medium (i.e. could see the seafloor but visibility was reduced due to moderate turbidity; categorized as ‘2’), and high (i.e. could see the seafloor with minimal to no turbidity; categorized as ‘3’).
For each trial, the following behaviors were recorded for both sharks and teleosts: feeding and avoidance. A feeding was recorded when an animal was observed to eat the bait. An avoidance was recorded when an abrupt change in swimming pattern, such as an acceleration away or a 45°, 90° or 180° turn away was observed. Each behavior was then aggregated for each species where possible and for each trial.
Data collected throughout the experiment was in the form of frequencies (i.e. counts) for both shark species. However, since the collected data were multi-dimensional, where the main effects of several variables and interaction terms between these variables were of interest, a Poisson generalized linear mixed effect model was used for each behavioural type (i.e. avoidances and feedings). Furthermore, since trials were conducted in a similar location between replicates, C. perezi and C. leucas behaviors were not considered independent. This potential non-independence violates the assumption in generalized linear models; thus, replicates were treated as a random effect whereas the other variables (i.e. treatment type and water visibility) were treated as fixed effects.
The mathematical form of the implemented generalized linear mixed effect model was:
Y represents the column vector of the response variable (counts of shark responses), X is the design matrix of explanatory variables, including all possible interaction terms, β is the column vector of coefficients that correspond to explanatory variables, R is the vector of individual replicates, which is a random effect, and ε represents the vector of errors, which are assumed to follow a normal (Gaussian) distribution whose mean is zero and whose variance is constant. The fixed effects (i.e., X) were treatment type (discrete), water visibility (discrete; only implemented in C. leucas as visibility did not vary sufficiently during the C. perezi trials), and conspecific density (discrete).
The generalized linear mixed effect model (Eq. (1)) was implemented using the ‘lme4’ package of R (Bates et al., 2012; Hyun et al., 2014; R 3.3.0 Statistical Program). Starting with the null model, forward selection was used to determine the best fit model for the data. Subsequent models were created by adding one or several explanatory variables to determine their effect on avoidance frequency (i.e. total avoidances/25) and feeding frequency (i.e. total feedings/25). Typically, visits (i.e. fish swimming within one body length of an experimental apparatus) were used as the denominator and used to create behavioral frequencies within a given trial (e.g. O’Connell et al., 2014). However, for this study, the quantity of baits deployed (i.e. 25 baits) within a trial were used to create the frequencies because the quantity was both standardized within a trial and deemed more accurate, as a standardized region to record a visit within the three-dimensional experimental setting was inherently challenging. We tested the contribution of an explanatory variable, examining the difference in the log-likelihood (Faraway, 2006; Hyun et al., 2014):
is the difference in the log-likelihood between nested and non-nested models in the forward selection process, and is the difference in the number of free parameters between two models. Model selection criteria included: Akaike Information Criteria (AIC), and behavior of model residuals using a quantile-quantile (Q-Q) plot, and associated P-values. Lastly, no C. leucas avoidance behaviors were observed in relation to the control treatment, whereas a substantial quantity of avoidances were observed towards the procedural control and experimental electrode treatments. Therefore, C. leucas avoidance data were transformed by adding one behavioral count to each treatment region (i.e. one avoidance towards the control, procedural control, and experimental electrode treatments).
Unlike the data associated with C. leucas and C. perezi, the data associated with the teleosts were too insufficient to subject to the aforementioned Poisson generalized linear mixed effect model. Therefore, data were separated on a per-species basis and subsequently transformed into behavioral frequencies (e.g. avoidance frequency = total trial avoidances per total pieces of bait) for each behavioral type. However, due to difficulties associated with identifying the jacks (C. ruber and S. rivoliana) and snappers (L. griseus and L. apodus) to the species level during the Alicetown Channel experiment, data were aggregated within the respective fish group and subjected to analysis. After transformation, data were first subjected to a Shapiro-Wilk’s test for normality. Since the data did not meet the statistical assumptions of normality and homogeneity, a non-parametric Kruskal-Wallis test was implemented to assess if any variation in behavioral frequency existed with treatment type.
Thirty replicates (20 at Alicetown Channel and 10 at Triangle Rocks; 90 total trials) were conducted over 16 days during February and March 2015. Throughout experimentation, the mean voltage was 4.7 V for the experimental electrode treatment trials. The mean electrical current during the experimental electrode treatment trials was 0.785 A and therefore, maximum power output was 3.73 W.
Triangle rocks: Caribbean reef shark (Carcharhinus perezi)
Throughout experimentation, the per trial quantity of C. perezi ranged from 7 to 13. For avoidance frequency, the best fit model (A2) included the main effects of treatment type (T) and contained an AIC of 90.80 (table 1). The coefficient and associated P-values with the selected model demonstrate that the experimental electrode () and procedural control () treatments resulted in a significantly higher avoidance frequency when compared to the control (table 2; fig. 2).
Results from the mixed effect models pertaining to Caribbean reef shark (Carcharhinus perezi) and bull shark (Carcharchinus leucas) behavior. For C. leucas avoidance frequency, data were transformed to “total avoidances + 1” for each treatment type to improve the interpretability of the data, as no avoidances occurred towards the control treatment throughout the entire experiment. Trial (R) is treated as a random effect and the others are treated as fixed effects. These fixed variables were T (treatment), Den (conspecific density), and Vis (water visbility). Selected models for avoidance and feeding frequencies were A2, B2, C2, and D2 respectively, based on a combination of Akaike Information Criteria (AIC), and behavior of the residuals of a model using a quantile-quantile (Q-Q) plot, and associated P-values. Significant models () are in bold.
Coefficients, standard errors, t statistic and P-values of explanatory variables for best models A2, B2, C2, and D2 for avoidance and feeding frequencies associated with the Caribbean reef shark (Carcharhinus perezi) and the bull shark (Carcharhinus leucas), respectively, to the treatment. For C. leucas avoidance frequency, data were transformed to “total avoidances + 1” for each treatment type to improve the interpretability of the data, as no avoidances occurred towards the control treatment throughout the entire experiment. Significant models for main effects () are in bold.
For feeding frequency, the best fit model (B2) included the main effects of treatment type (T) and contained an AIC of 194.01 (table 1). The coefficient and associated P-values with the selected model demonstrate that although the experimental electrode and procedural control treatments resulted in fewer feedings, only the experimental electrode treatment () had a significant effect on C. perezi feeding frequency (table 2, fig. 2).
Triangle rocks: teleosts
Throughout experimentation, two teleost species, the bar jack (Caranx ruber; n = 26) and the Bermuda chub (Kyphosus sectatrix; n = 31), were observed. Since no avoidances were observed, a Kruskal-Wallis test was only conducted on the associated feeding frequencies of these two species. However, the feeding frequencies (fig. 3) between treatments were not significantly different for both C. ruber () and K. sectatrix () (table 3).
Kruskal-Wallis tests assessing if any associations with treatment type exist pertaining to the feeding frequencies of two teleost species, bar jack (Caranx ruber) and Bermuda chub (Kyphosus sectatrix), at Triangle Rocks and numerous teleost species, remora (Remora remora), bar jack (C. ruber) and almaco jack (Seriola rivoliana), mangrove snapper (Lutjanus griseus) and schoolmaster snapper (L. apodus), and Bermuda chub (K. sectatrix), at the Alicetown Channel.
Alicetown channel: bull sharks (Carcharhinus leucas)
Throughout experimentation, the per-trial quantity of C. leucas ranged from 1 to 10. For avoidance frequency, the best fit model (C2) included the main effects of treatment type (T) and contained an AIC of 88.69 (table 1). The coefficient and associated P-values with the selected model demonstrate that only the experimental electrode treatment resulted in a significant increase in C. leucas avoidance frequency (; table 2; fig. 4).
For feeding frequency, the best fit model (D2) included the main effects of treatment type (T) and contained an AIC of 365.49 (table 1). The coefficient and associated P-values with the selected model demonstrate that both the experimental electrode () and the procedural control () treatments had a significant effect on C. leucas feeding frequency (table 2); however, the experimental electrode treatment resulted in fewer feedings than did the procedural control treatment (table 2; fig. 4).
Alicetown channel: teleosts
Throughout experimentation, six teleost species were observed: bar and almaco jack (Caranx ruber and Seriola rivoliana, respectively; n = 18) Bermuda chub (Kyphosus sectatrix; n = 12), remora (Remora remora; n = 8), and mangrove and schoolmaster snapper (Lutjanus griseus and Lutjanus apodus, respectively; n = 66). No avoidances were observed, and the feeding frequencies (fig. 3) between treatments were not significantly different for all teleost species: C. ruber and S. rivoliana (), K. sectatrix (), R. remora (), and Lutjanus spp. () (table 3).
This study demonstrated that the Biorock-associated electric fields significantly alter shark behavior, reducing feedings and increasing avoidances (table 2; figs. 2, 4); whereas the behavior of interacting teleosts was not observably influenced (table 3; fig. 3). Interestingly, sharks approached the bait yet completely missed and exhibited signs of disorientation during the experimental electrode treatment. These findings are similar to those studying the effects of various deterrent materials (e.g. electropositive metals and permanent magnets) on spiny dogfish (Squalus acanthias), where observations revealed bait localization difficulties when subjected to deterrents (Stoner & Kaimmer, 2008). Since electroreception has been demonstrated to be important in the final phases of prey capture (e.g. Kajiura & Holland, 2002; Kajiura & Fitzgerald, 2009; Jordan et al., 2013), the electric fields associated with the Biorock technology may mask the unique cues associated with bait thus negatively influencing feeding capabilities.
Elasmobranchs: treatment type
This experiment demonstrated that the behavioral frequencies (i.e. avoidance, feeding) of both species varied significantly between treatments. Mixed effect model analyses illustrate that the experimental electrode treatment contributed substantially to these significant differences (i.e. table 2), which is suggestive that the Biorock-associated electric fields (i.e. 4.7 V) were sufficient to elicit deterrent behavior. The present findings are consistent with previous studies that demonstrate the elasmobranch deterrent capabilities of electro-sensory stimuli (Huveneers et al., 2012; SANSA, 2012), and thus suggest that strong electro-sensory stimuli exhibit the capability to override the desire to feed in certain contexts. In addition to the experimental electrode treatment, the procedural control treatment had a similar effect for C. perezi avoidance frequency and C. leucas feeding frequency. More specifically, avoidance frequencies significantly increased () and feeding frequencies significantly decreased () with respect to the procedural control treatment. With binocular vision and depth perception as a direct result of eye positioning, eye rotation, and head yaw (e.g. McComb et al., 2009), it is possible that the associated visual stimuli of both the procedural control and experimental electrode treatments were visually detected and sufficient to elicit these behavioral responses. However, although the procedural control treatment did impact C. perezi avoidance frequency and C. leucas feeding frequency, the results pertaining to the avoidance and feeding frequencies (figs. 2, 4) illustrate the heightened effect of the experimental electrode treatment, and thus the sole effect of visual stimuli is not sufficient to explain the findings. Therefore, the potential utilization of Biorock-associated electric fields for both coral restoration and shark deterrence warrants further investigation.
For feeding frequency, it was determined that C. perezi and C. leucas were more likely to successfully feed from the procedural control and control treatments than the experimental electrode treatment (4.7 V; table 2; figs. 2, 4). This finding therefore suggests that an electro-sensory threshold exists near or at 4.7 V and this threshold may contribute to temporary feeding inhibition. Although this study aimed to assess the influence of Biorock-associated electric fields on fish behavior, the feeding inhibition threshold has increasing relevance. Due to overexploitation and lack of proper management, sharks now represent the largest group of threatened marine species from a global perspective according to the International Union for the Conservation of Nature (IUCN) Red List (e.g. Manire & Gruber, 1990; Musick et al., 2000; Stevens et al., 2000; Baum & Myers, 2004; Lucifora et al., 2011). Furthermore, sharks are K-selected (i.e. slow growth rate, late maturity, and low fecundity); therefore the likelihood of population rebound in response to their unsustainable harvest is minimal if the present mortality rates continue (Stevens et al., 2000; Dulvy et al., 2014). Since many shark species play a top-down predatory role within their respective ecosystems, population declines at such high levels may have substantial ecological consequences (Myers et al., 2007; Heithaus et al., 2008). Therefore, with both commercial and recreational fisheries continuously contributing to substantial shark population decline (Stevens et al., 2000; Dulvy et al., 2014), these C. perezi findings suggest that deterrents containing higher associated voltages may have an increased elasmobranch-specific deterrent capability.
Elasmobranchs: conspecific density
Organisms occupying a similar ecological niche often compete between (i.e. interspecific competition) and within (i.e. intraspecific competition) species (Nelson & Johnson, 1980; Schoener, 1983), which has been demonstrated to alter shark behavior (Crombie, 1947; Polis, 1981; Stiling et al., 1984; Munday et al., 2001). In previous shark-related studies, conspecific density was determined to negatively influence the effectiveness of deterrent-associated baits/hooks (Brill et al., 2009; Jordan et al., 2011; Robbins et al., 2011). However, in contrast to these findings, conspecific density was not observed to significantly influence the deterrent capability of the experimental electrode treatment. Therefore, this finding may be explained by the fact that density-induced behavioral responses to electro-sensory stimuli are species-specific and the electrodes were so effective that even though a competitive mentality may have been induced, behavioral manipulation was not compromised.
Elasmobranchs: water visibility
Water visibility has been previously observed to alter fish behaviour (Ranåker et al., 2012; O’Connell et al., 2013). With declining water visibility parameters, a heightened reliance on another sense has been observed, known as context-dependent switching (Leahy et al., 2011; Ranåker et al., 2012). For example, in a previous study and when compared to sighted sharks, visually deprived lemon sharks (Negaprion brevirostris) exhibited significant behavioral changes towards permanent magnetic stimuli (O’Connell et al., 2013). All trials in the present study were conducted during daylight hours and water visibility conditions were greater than or equal to 3 m. Thus, the visual range may not have been sufficiently restricted to elicit these previously observed behavioral changes (e.g. O’Connell et al., 2013), and therefore, may explain why water visibility did not significantly impact shark behavior.
Experimental treatments did not observably influence the feeding behavior of teleosts at either study site (table 3; fig. 3). One explanation for these findings may be the apparent lack of a well-developed electro-sensory system in the teleosts targeted (Montgomery et al., 1995; Bell, 2001). However, continued research is needed as strong electric fields (300-400 V) have been demonstrated to have negative consequences on a variety of marine teleosts (e.g. Snyder, 2003) and thus, although not observed in the present study, interspecific variation in teleost responses to Biorock-associated electric fields may occur.
Although not the main focus of the present study, the results illustrate that the electric fields generated by the experimental electrode treatment elicit deterrent responses in the elasmobranchs in this study, but not in the teleosts. Therefore, with elasmobranch bycatch being a substantial concern to the overall health of current elasmobranch populations (Molina & Cooke, 2012), the application of the electric fields used in the present study may also be a great approach to future conservation engineering techniques to prevent shark bycatch by long lines and drift nets. Whether individual electric field devices (e.g. battery operated) are applied directly to hooks or electrodes are continuously spaced along a commercial longline, both warrant further investigation.
Coral reef restoration has become a rapidly expanding discipline that encompasses a variety of approaches and scientific fields, including ecology, geology, physics, economics, and engineering. Within this discipline, a variety of restoration techniques exist, including (1) structural restoration (e.g. creating artificial reefs through activities such as the relocation of rocks/dead coral heads or the sinking of vessels; Carr & Hixon, 1997), (2) biological restoration (e.g. collecting, rehabilitating, and/or transplanting naturally broken coral fragments, culturing coral larvae, or propagating coral colonies; Ellis & Ellis, 2001), and (3) physical restoration (e.g. assessing the environmental parameters where coral growth is occurring to improve health, growth rates, or fecundity, which includes mid-water coral nurseries or the Biorock technology; Shafir & Rinkevich, 2010). In the present study, the effects of the weak direct current electric fields associated with the Biorock reef restoration technology were assessed. Although the electricity makes Biorock reefs grow faster and healthier than other corals, this study found that Biorock-associated electric fields significantly alter C. leucas and C. perezi behavior, but have little observable impact on teleosts. Despite the Biorock technology exhibiting promise in coral reef restoration, the possibility of the technology negatively influencing the “natural behavior” of interacting elasmobranchs is concerning. If long-range inhibition of feeding occurs around these Biorock structures, general and localized top-town ecosystem dynamics could be threatened and therefore, future investigations assessing the distance of influence of Biorock structures on shark behavioral patterns are warranted. In contrast, if Biorock structures elicit close-range deterrent responses, it is possible that these structures may provide localized ecosystem benefits while concurrently serving as a short-range shark deterrent system in relevant areas. Due to the infancy of fish behavioral studies surrounding Biorock structures and the potential positive and/or negative implications of their presence, future research on more species is needed.
TJG and MPU built the experimental apparatus to ensure it appropriately mimicked the functionality of the Biorock Technology. MPU and CPO designed, conducted, and statistically analyzed all fieldwork. It is important to note that the fieldwork and statistical analysis was conducted independently of TJG as a means to eliminate any potential experimental bias due to TJG’s vested interest in the technology. MPU, CPO, and TJG collectively wrote the paper.
We thank the Global Coral Reef Alliance, the Wave Foundation, and O’Seas Conservation Foundation for providing partial funding, equipment, and training to make this project possible. We thank the Bahamian Government for granting us permits for this scientific research. We thank the Bimini Big Game Resort and Marina for permitting us to use their facilities. With are thankful for the comments provided to us by several anonymous reviewers, as the comments largely benefitted this manuscript. We thank Guido Leurs, Lauren Smith Diack, Ryan Reed, Sarah Reed, Rachel Ariel Jacobson, Caroline Collatos, Nicole O’Connell, Debra Friedman, Mary Fraga and many other volunteers who assisted us. Lastly, we thank Dr. Saang-Yoon Hyun who provided the initial framework and guidance needed to understand and carry out the statistical approaches used in this study.
Bakti L.A., Virgota A., Damayanti L.P., Radiman T.H., Retnowulan A., Hernawati, Sabil A., Robbe D. (2012) Biorock reef restoration in Gili Trawangan, North Lombok, Indonesia. In: Goreau T.J., Trench R.K. (Eds) Innovative Methods of Marine Ecosystem Restoration, pp. 59-80. CRC Press, Florida, USA.
Brill R., Bushnell P., Smith L., Speaks C., Sundaram R., Stroud E., Wang J. (2009) The repulsive and feeding-deterrent effects of electropositive metals on juvenile sandbar sharks (Carcharhinus plumbeus). Fish. Bull., 107, 298-307.
Bryant D., Burke L., Mcmanus J., Spalding M. (2000) Reef at Risk: a Map-Based Indicator of Threats to the World’s Coral Reefs. 1st Edition. World Resources Institute, Washington, D.C.
Burke L., Kura Y., Kassem K., Revenga C., Spalding M., Mcallister D. (2001) Pilot Analysis of Global Ecosystems: Coastal Ecosystems. 1st Edition. World Resources Institute, Washington, D.C.
Downs C.A., Woodley C.M., Richmond R.H., Lanning L.L., Owen R. (2005) Shifting the paradigm of coral-reef ‘health’ assessment. Marine Poll. Bull., 51, 486-494.
Dulvy N.K., Fowler S.L., Musick J.A., Cavanagh R.D., Kyne P.M., Harrison L.R., Carlson J.K., Davidson L.N., Fordham S.V., Francis M.P., Pollock C.M. (2014) Extinction risk and conservation of the world’s sharks and rays. Elife, 3, e00590.
Ellis S., Ellis E. (2001) Recent advances in lagoon-based farming practices for eight species of commercially valuable hard and soft corals. Technical Report. CTSA, 147, 63 pp.
Gibbs M.A. (2004) Lateral line receptors: where do they come from developmentally and where is our research going? Brain Behav. Evol., 64, 163-181.
Goreau T.J. (2012) Marine electrolysis for building materials and environmental restoration. In: Kleperis J., Linkov V. (Eds) Electrolysis, pp. 273-290. InTech Publishing, Rijeka, Croatia.
Goreau T.J. (2014) Electrical stimulation greatly increases settlement, growth, survival, and stress resistance of marine organisms. Nat. Resour. J., 5, 527-537.
Heithaus M.R., Frid A., Wirsing A.J., Worm B. (2008) Predicting ecological consequences of marine top predator declines. Trends Ecol. Evolut., 23, 202-210.
Huveneers C., Rogers P.J., Semmens J. (2012) Effects of the Shark Shield™ electric deterrent on the behaviour of white sharks (Carcharodon carcharias). SARDI Publication No F2012/000123-1, Research Report Series No. 632, South Australian Research and Development Institute.
Hyun S.-Y., Cadrin S.X., Roman S. (2014) Fixed and mixed effect models for fishery data on depth distribution of Georges Bank yellowtail flounder. Fish. Res., 157, 180-186.
Jordan L.J., Mandelman J.W., Kajiura S.M. (2011) Behavioral responses to weak electric fields and a lanthanide metal in two shark species. J. Exp. Mar. Biol. Ecol., 409, 345-350.
Jordan L.K., Mandelman J.W., Mccomb D.M., Fordham S.V., Carlson J.K., Werner T.B. (2013) Linking sensory biology and fisheries bycatch reduction in elasmobranch fishes: a review with new directions for research. Conserv. Physiol., 1, cot002.
Josberger E.E., Hassanzadeh P., Deng Y., Sohn J., Rego M.J., Amemiya C.T., Rolandi M. (2016) Proton conductivity in ampullae of Lorenzini jelly. Sci. Adv., 2, e1600112.
Kalmijn A.J. (1974) The detection of electric fields from inanimate and animate sources other than electric organs. In: Fessard A. (Ed.) Handbook of Sensory Physiology, pp. 148-200. Springer-Verlag, Berlin.
Kalmijn A. (1984) Theory of electromagnetic orientation: a further analysis. In: Bolis L., Keynes R.D., Maddrell S.H.P. (Eds) Comparative Physiology of Sensory Systems, pp. 525-560. Cambridge University Press, Cambridge, UK.
Kalmijn A. (2000) Detection and processing of electromagnetic and near-field acoustic signals in elasmobranch fishes. Philos. Trans. R. Soc. London [Biol.], 355, 1135-1141.
Kempster R.M., McCarthy I.D., Collin S.P. (2012) Phylogenetic and ecological factors influencing the number and distribution of electroreceptors in elasmobranchs. J. Fish Biol., 80, 2055-2088.
Leahy S.M., McCormick M.I., Mitchell M.D., Ferrari M.C. (2011) To fear or to feed: the effects of turbidity on perception of risk by a marine fish. Biol. Lett., 7, 811-813.
Leão Z., Kikuchi R. (2005) A relic coral fauna threatened by global changes and human activities, eastern Brazil. Marine Poll. Bull., 51, 599-611.
Marcotte M., Lowe C. (2008) Behavioral response of two species of sharks to pulsed, direct current electrical fields: testing a potential shark deterrent. Mar. Technol. Soc. J., 42, 53-61.
Montgomery J.C., Coombs S., Conley R.A., Bodznick D. (1995) Hindbrain sensory processing in lateral line, electrosensory, and auditory systems: a comparative overview of anatomical and functional similarities. Aud. Neurosci., 1, 207-231.
Munday P.L., Jones G.P., Caley M.J. (2001) Interspecific competition and coexistence in a guild of coral-dwelling fishes. Ecology, 82, 2177-2189.
Musick J.A., Burgess G., Cailliet G., Camhi M., Fordham S. (2000) Management of sharks and their relatives (Elasmobranchii). Fisheries, 25, 9-13.
Myers R.A., Baum J.K., Shepherd T.D., Powers S.P., Peterson C.H. (2007) Cascading effects of the loss of apex predatory sharks from a coastal ocean. Science, 315, 1846-1850.
O’Connell C.P., Guttridge T.L., Brooks J., Finger J.S., Gruber S.H., He P. (2013) Behavioral modification of visually deprived lemon sharks (Negaprion brevirostris) towards magnetic fields. J. Exp. Mar. Biol. Ecol., 453, 131-137.
O’Connell C.P., Hyun S.-Y., Rillahan C.B., He P. (2014) Bull shark (Carcharhinus leucas) exclusion properties of the Sharksafe barrier and behavioral validation using the ARIS sonar technology. Glob. Ecol. Cons., 2, 300-314.
Ranåker L., Nilsson P.A., Brönmark C. (2012) Effects of degraded optical conditions on behavioural responses to alarm cues in a freshwater fish. PLoS One, 7(6), e38411.
Rivera-Vicente A.C., Sewell J., Tricas T.C. (2011) Electrosensitive spatial vectors in elasmobranch fishes: implications for source localization. PLoS One, 6, e16008.
Robbins W.D., Peddemors V.M., Kennelly S.J. (2011) Assessment of permanent magnets and electropositive metals to reduce the line-based capture of Galapagos sharks, Carcharhinus galapagensis. Fish. Res., 109, 100-106.
Shafir S., Rinkevich B. (2010) Integrated long-term mid-water coral nurseries: a management instrument evolving into a floating ecosystem. U. Maur. Res. J., 16, 365-386.
Smit C.F., Peddemors V.M. (2003) Estimating the probability of a shark attack when using an electric repellent. S. Afr. Stat. J., 37, 59-78.
Smith E.D. (1974) Electro-physiology of the electrical shark-repellant. Tran. In. Elect. Eng., 166-181.
Snyder D.E. (2003) Invited overview: conclusions from a review of electrofishing and its harmful effects on fish. Rev. Fish Biol. Fish., 13, 445-453.
South African National Space Agency (SANSA) (2012) Interim report on measurement and analysis of the electric fields produced by the SharkPOD and Shark Shield. Doc No: 6035-0006-722-A1.
Spalding M.D. (2000) Spatial variation in coral reef fish biodiversity at intermediate scales around oceanic islands. In: Proceeding of 9th International Coral Reef Symposium, vol. 1, pp. 23-27.
Stevens J.D., Bonfil R., Dulvy N.K., Walker P.A. (2000) The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J. Mar. Sci., 57, 476-494.
Stiling P.D., Brodbeck B.V., Strong D.R. (1984) Intraspecific competition in Hydrellia valida (Diptera: Ephydridae), a leaf miner of Spartina alterniflora. Ecology, 65(2), 660-662.
Stoner A.W., Kaimmer S.M. (2008) Reducing elasmobranch bycatch: laboratory investigation of rare Earth metal and magnetic deterrents with spiny dogfish and Pacific halibut. Fish. Res., 92, 162-168.
Tricas T.C. (2001) The neuroscience of the elasmobranch electrosensory world: why peripheral morphology shapes behavior. Environ. Biol. Fishes, 60, 77-92.