Site Fidelity and Behavior of Sharks at a Long-Term Shark Feeding Dive Site

Material Information

Site Fidelity and Behavior of Sharks at a Long-Term Shark Feeding Dive Site
Scharfer, Alissa
Burgess, George ( Mentor )
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:


serial ( sobekcm )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.


This item has the following downloads:

Full Text

Journ31l of in.nderr.3adu.3- Re--search

.Oluine 5, issuue 3 - :ecem-,er 20:0:3

Site Fidelity and Behavior of Sharks at a Long-Term Shark Feeding Dive Site

Alissa Scharfer


To encourage tourism, commercial dive boat operators often offer shark-feeding excursions and provide

video footage of the dives to participating SCUBA divers. In a study of Caribbean reef shark feeding dives occurring

at a single location off West End, The Bahamas, videotapes of 36 dives recorded over approximately nine years

were analyzed. Data was collected on many behavioral and environmental variables. Site fidelity of individual

sharks was investigated by documenting, when possible, markings and scars on each shark. It was found that

the number sharks per dive, as well as the parasite load, increased temporally. Parasites were most frequently

found near gill openings. Bumping of the camera, presumably an agonistic act, was observed more often in

females than in males. Males were found to be more aggressive than females during the summer and

overall aggression from both sexes decreased over the duration of the study. The frequency of shark

feeding attempts was correlated with the technique of the human feeder and showed complex temporal trends.

Nine individual females and eight males were identified as appearing in more than a single dive. Although one

shark had a hiatus of 4.5 years between sightings, most of the reappearing individuals were repetitively observed

on site over periods of several months. Such site fidelity may be attributable to feeding-induced conditioning of

the sharks, with possible consequences including ecological disruption and increased danger to humans.


Carcharinus perezi, the Caribbean reef shark, is found throughout tropical western Atlantic waters. It is the

most common shark on or near coral reefs throughout much of the Bahamas. Like many other sharks, the

Caribbean reef shark primarily eats bony fishes (Compagno, 1984). C. perezi also has been observed lying

motionless on the sea floor (Castro, 1983), a behavior termed "sleeping."

An increasingly popular recreational activity has developed over the past two decades. To increase paying

clientele, many dive-boat operations offer interactive shark-feeding dives to recreational SCUBA (and

occasionally skin) divers at some of the region's most popular dive sites (Fig. 1). Although new regulations

prohibit such feeding off the coast of Florida and the Cayman Islands, no such restrictions have been placed

on operations in the Bahamas.

Figure 1. Shark feeding by hand at West End, Bahamas.

The desirability of such shark-feeding activities is the subject of considerable debate. In addition to

concerns expressed about ecological disruption, some have suggested that feeding might lead to increased danger

to humans (Burgess, 1998). A shark attack upon a submerged human often is behaviorally similar to an attack

upon a natural prey item (Nelson, 1977). In other cases, attacks are thought to be the result of agonistic behavior.

In such events the shark may exhibit agonistic displays, indicating possible forthcoming aggression. Such

agonistic displays have been reported as occurring prior to C. perezi attacks on humans (G.H. Burgess, pers.

comm.)1 and are well documented in the gray reef shark (C. amblyrhynchos) (Johnson and Nelson, 1973). In

the latter species, agonistic displays include:

* head swings well out of the normal path, or upward, yielding a weaving or spiraling pattern in the shark's path

* sharp, quick movements or turns increase in number

* back arches like a cat

*one, or both, pectoral fins are lowered

Another possible agonistic display observed in C. perezi is characterized by a brief, but distinctive opening and

closing of the mouth, called "yawning" (G.H. Burgess, pers. com.)1.


A total of 36 shark-feeding dives captured on videotape over the period of 26 November 1992 - 9 September 2001

at a single site off West End, The Bahamas, were analyzed. Many variables concerning the sharks, humans,

and environment were recorded. Variables included water visibility; human activity factors; number of sharks

and humans per dive; shark parasite load; gender of shark and male maturity; and frequency of feeding

attempts, acts of aggression, and "yawning." Human activity factors included the total number of submerged

human divers as well as the feeding technique employed by the human feeder, feeder aggression towards sharks

(as measured by spear jabs), and identity of the individual feeder. Only agonistic displays by sharks occurring prior

to feeding were recorded. These acts included a circling or "zig-zag" swimming pattern, back arching, dropping of

the pectoral fins, and "bumping." Bumping is characterized by the shark's head making contact with the

video camera lens (Fig. 2). Site fidelity of individual sharks, the tendency for a shark to return to the same

location, was investigated by documenting, whenever possible, markings and scars on each shark. Scars on

males and females, the latter resulting from bites from males during the mating process, were particularly useful

for identification. The presence of claspers, a reproductive organ in males (Fig. 3), served to differentiate the

sexes and clasper size was used to estimate male maturity. In tandem, size, sex and scarring patterns served as

the statistical basis for establishing a minimum number of sharks. Parasite abundance and location on the body

were recorded.

Figure 2. Bumping, an antogonistic act.


... . , ,.. .. *

Figure 3. Male claspers.


The number of sharks per dive was found to increase significantly over time (Fig. 4) (male p=0.0025,

female p<0.0001, overall p<0.0001). The sex ratio of observed sharks was 145 males: 133 females. Most

(93.8%) male sharks were mature.

.C 2


M c -

E "
3 0 .5

1992 1994 1996 1998 2000 2002
Figure 4. Number of sharks increasing over time.

The frequency of feeding attempts made by sharks remained constant. However, because the number of
sharks participating on dives increased over time, the frequency of feeding attempts per shark per dive
decreased temporally (p=0.0036). The number of agonistic acts per feeding attempt decreased over time
(p<0.0001) (Fig. 5). Among agonistic acts, lowering of the pectoral fins was observed most frequently, followed
by zig-zag swimming and circling at similar frequencies. Back arching was only observed once. During the
summer, males were significantly more aggressive than females (Fig. 6) (p=0.0424). However, females bumped
the camera more frequently (p=0.0304), with females and males averaging at 0.0685 and 0.0460 bumps per
minute, respectively. Yawning was recorded on five occasions. Frequency of aggression was dependent on
the feeding technique used (p<0.0001). Feeding the sharks by placing the food on the end of a long spear
yielded the most aggression, whereas placing the food under a reef ledge produced the least aggression (Fig. 7).

051. -

c I - I i I I I I

Figure 5. Antagonistic attacks, over time.

Male and Female Seasonal Aggression

Summer Spring Fall Winter
Figure 6. Male and female seasonal aggression.



2 -
6 _

* Female
I Male

(note: winter sample size statistically insignificant)

1-fish on long spear
2=fish on long spear + hand feed
3--fish on short spear
-4fish on short spear + hand feed
5-hand feed
6-fish under reef
7--ish on long spear + under reef

1 2 3 4 5 6 7
Feeding Technique
Figure 7. Frequency of agonistic acts depends on feeding techniques.

There was no correlation between frequency of agonistic acts and individual human feeders. Furthermore, the
feeder did not change his frequency of jabbing the sharks with a spear, therefore the aggression towards sharks
by the feeder remained constant. Water visibility and human aggression towards sharks did not affect
shark behaviors. The relationship between number of human divers participating on a dive and frequency of
shark feeding attempts was curvilinear (p=0.0587) (Fig. 8), with frequency of feeding attempts peaking at
fourteen divers.



o 010 ______1.. --2
111 s 0-
.... " ""

U. 0 - * � :

4 6 8 10 12 14 16 18 20
Minimum Number of Divers
Figure 8. Curvilinear relationship between number of divers and feeding attempts peaks at 14

divers submerged.

Parasite load increased temporally (p=0.0002) (Fig. 9). Of the total number of parasites recorded on sharks,

83.59% were found near gill openings, 6.70% were located on the first dorsal fin, and 9.71% were seen on

other parts of the sharks. Females hosted a heavier parasite load than males (p=0.0653), with the mean number

of parasites on males and females being 0.0282 and 0.0372, respectively. Four male Caribbean reef sharks were

also observed with bulging "lumps" on their left flank in approximately the same location.

.n-1 -

~d. ~

- -

Figure 9. Parasites per shark increasing temporally.

Sharks showed a notable degree of site fidelity (Fig. 10). Nine individual females and eight males were identified

as present on more than a single dive. Although one individual had a hiatus of 4.5 years between sightings, most

of the reappearing individuals were repetitively observed on site over periods of several months. A single

Caribbean reef shark of unidentified sex was observed sleeping on the ocean floor, concealed underneath a

natural reef arch (Fig. 11).

1998 -

Individual Shark
Figure 10. Dates of individual shark sitings.

Figure 11. A "sleeping" Caribbean reef shark.
Figure 11. A "sleeping" Caribbean reef shark.

In addition to Caribbean reef sharks, nurse sharks (Ginglymostoma cirratum), two great hammerhead
sharks (Sphyrna mokarran), and one lemon shark (Negaprion brevirostris) were observed.


The results suggest that sharks were conditioned to feeding at this site. Since the number of sharks observed
per dive increased temporally, it appears that sharks recognized the area as a feeding-site and oriented to
that locale. The site fidelity of numerous individuals reinforces this conclusion. The decrease in aggression
further suggests that the sharks are becoming more familiar with the feeding procedures and their
surroundings. Since the human feeder did not significantly change his frequency of aggression towards sharks,
more confidence can be placed on these interpretations. Although sharks are less aggressive during the dives as
a result of conditioning, if feeding is denied, danger may be posed to unknowing submerged humans at the
feeding site or in adjacent areas frequented by the sharks.

1 n . Z 1'.
I .. ,1 a

The increase in parasite load may be a result of increased shark density with parasites more readily

transferred between hosts. Such a marked change in parasite load underscores concern about ecological

disruption. By inducing the congregation of many Caribbean reef sharks through feeding, shark-feeding

operators may be changing the natural biology of the sharks and promoting parasite transfer.

It was found that a relatively constant number of feeding attempts occurred during the period of an individual

feeding dive, regardless of the number of sharks present. Since the number of sharks increased over the course

of the study, the individual feeding rate (number of feeding attempts per shark) decreased. Thus, the amount of

food each shark consumed during a feeding event decreased over time. In addition, if a social hierarchy develops

in which certain individuals are excluded from access to provided food, then the potential exists for an

increased number of hungry, aggressive sharks in an area, potentially leading to negative interactions with

humans. Feeding also raises concern about ecological disruption. The diet the sharks were consuming over the

course of the study was inconsistent and may potentially result in nutrient deprivation.

A hierarchy of agonistic displays may be postulated from the relative percentages of each act. If the most

frequently observed display is interpreted as the initial indication of aggression, then dropping the pectoral fins is

the initial display. Zig-zag (Fig. 12) or circle patterns of swimming would follow in the hierarchy, followed by

arching of the back. Yawning displays were rare and not seen prior to feeding, suggesting that this behavioral

pattern may follow the above displays hierarchically. Further investigation of agonistic display is required to

confirm this suggested hierarchy.

Figure 12. "Zig-zag" swimming pattern.

Females may react towards threats more actively than males, as they were observed to bump the camera

more frequently. Female Caribbean reef sharks are generally larger than males, possibly leading to this increase

in aggression. Although much remains unknown about the reproductive behavior of C. perezi, its mating season

can be postulated to occur in the summer, since males were found to be more aggressive during this season.

The results show that average shark aggression can be predicted by the type of feeding technique used, and that it

is not the feeder himself to which these sharks are responding. Placing food on the end of a long spear yielded

the most aggression whereas placing the food under a reef ledge yielded the least aggression. This predictive

power may be useful to other shark-feeding dive operations as to prevent shark aggression towards divers.

Finally, the observed "sleeping" shark was one of the few documented sleeping C. perezi in the waters of

the Bahamas, further discrediting the myth that sharks must be in constant motion. The lumps observed on the

left side of male sharks could be a result of some physiological condition or an endoparasite.

The research was a retroactive design, thus only a minimal level of statistical constraint on variables was

possible. Data gathering was limited by the quality and content of existing video tapes, which posed

various limitations. For instance, as technology improved over time, visual acuity improved and more

individual sharks could be identified. Although statistical procedures can help eliminate such confounds, the

project would have yielded more confident results if the research plan had included scientific field

videography protocols.


George H. Burgess, Florida Program for Shark Research, University of Florida, Gainesville, FL


George H. Burgess, Florida Program for Shark Research, University of Florida, Gainesville, FL

Burgess, G.H. 1998. Diving with elasmobranchs: a call for restraint. IUCN Shark Specialist Group, Shark News 11:1-4.

Castro, J. (1983). The Sharks of North American Waters. Texas A&M University Press: 140-141.

Johnson, R., and Nelson, D. (1973). Agonistic Display in the Gray Reef Shark, Carcharhinus menisorrah, and

its relationship to attacks on man. Copeia, NO. 1: 77-83.

Nelson, D. (1977). On the Field Study of Shark Behavior. American Zoology, NO. 17: 501-507

Compagno, J.E.V. 1984. FAO Fisheries Synopsis: FAO Species Catalog, Sharks of the World, NO. 125, Vol. 4:



Back to the Journal of Undergraduate Research

College of Liberal Arts and Sciences I University Scholars Program I University of Florida |

� University of Florida, Gainesville, FL 32611; (352) 846-2032.

Tle S'm s tslw r Sh Gyor Nahou