You are on page 1of 16

Environmental Management (2009) 43:1–16

DOI 10.1007/s00267-008-9198-z

PROFILE

Managing Dive Tourism for the Sustainable Use of Coral Reefs:


Validating Diver Perceptions of Attractive Site Features
Maria C. Uyarra Æ Andrew R. Watkinson Æ
Isabelle M. Côté

Received: 11 October 2007 / Accepted: 5 August 2008 / Published online: 23 September 2008
Ó Springer Science+Business Media, LLC 2008

Abstract It has been argued that strategies to manage ability to perceive/recall some of the attributes. Perceived
natural areas important for tourism and recreation should differences in environmental attributes, such as surface
integrate an understanding of tourist preferences for spe- conditions, underwater current, and the likelihood of
cific natural features. However, the accuracy of tourist encountering rare fish and sea turtles, were not empirically
recalled perceptions of environmental attributes, which are validated. The fact that divers perceive correctly differ-
usually derived from post hoc surveys and used to establish ences in the condition of some of the key biological
management priorities, is currently unmeasured. We tested attributes that affect dive enjoyment reinforces the need to
the validity of the relationship between tourist-stated maintain overall reef condition at satisfactory levels.
preferences and actual condition of coral reefs around the However, variation in accuracy of perceptions owing to
Caribbean island of Bonaire. Using standardized ques- demographic factors and attribute type suggests the need
tionnaires, we asked 200 divers to select their most and for caution when using public perceptions to develop
least favorite dive sites and the attributes that contributed environmental management strategies, particularly for
to that selection. We also carried out ecological surveys at coral reefs.
76 of the 81 dives sites around the island to assess the
actual conditions of the attributes indicated as important for Keywords Diving tourism  Preferences  Perceptions 
site selection. Fish- and coral-related attributes were key Coral reefs  Management  Caribbean
features affecting dive enjoyment. In general, divers
appeared to be able to perceive differences between sites in
the true condition of biological attributes such as fish The rapid expansion of the SCUBA diving industry has led
species richness, total number of fish schools, live coral to increasing popularity of holiday destinations harboring
cover, coral species richness, and reef structural complex- coral reefs (Cope 2003). However, the high vulnerability of
ity, although men and women divers differed in their corals to disturbances and the great variety of natural and
anthropogenic stressors affecting reefs make the sustain-
able management of this diverse ecosystem both
M. C. Uyarra (&)
complicated and necessary (Roberts 1997; Hoegh-Guld-
School of Biological Sciences, University of East Anglia,
Norwich NR4 7TJ, UK berg 1999; Gardner and others 2003). Divers often collide
e-mail: m.uyarra@uea.ac.uk with the reef, and these actions can lead to short- and
longer-term damage to corals (Harriott and others 1997;
A. R. Watkinson
Hawkins and others 1999; Tratalos and Austin 2001;
School of Environmental Sciences, University of East Anglia,
Norwich NR4 7TJ, UK Walters and Samways 2001). The extent of this damage is
e-mail: a.watkinson@uea.ac.uk often directly related to the intensity of site use by divers
(Dixon and others 1993; Hawkins and others 1999, 2005;
I. M. Côté
Barker and Roberts 2004). As a result, to balance the
Department of Biological Sciences, Simon Fraser University,
Burnaby, BC, Canada V5A 1S6 benefits arising from the recreational use of coral reefs with
e-mail: imcote@sfu.ca the conservation of this sensitive ecosystem, management

123
2 Environmental Management (2009) 43:1–16

strategies designed to regulate how divers distribute area of influence on large mammals. In this instance, for
themselves across reefs should be underpinned by a good example, basing recommendations about the number of
understanding of why divers dive where they do. This, in trails in a habitat or minimum recommended distances to
turn, requires knowledge of diver preferences and percep- observe wildlife solely on visitor perceptions would lead to
tions of coral reef habitats. great underestimates of the real extent of human
Visitors’ perceptions of impacts and of acceptable disturbance.
standards for environmental conditions can provide We attempted to establish the nature of preferences–
essential information for the sustainable management of perceptions–true conditions relationships for recreational
tourist destinations, especially in protected areas (Daily diving tourists visiting the reefs on the southeastern
2000). To this end, visitor surveys are usually administered Caribbean island of Bonaire. Bonaire is a popular diving
to identify attributes and conditions considered to be most destination, which is famous for its shore-diving. There are
important in determining visitors’ quality of experience. 60 dive sites around the main island, with 26 more around
Standards for attributes and conditions derived from such the satellite island of Klein Bonaire (Van’t Hof 1997).
surveys, based on perceptions of visitors, are then pro- These numerous and easily accessible sites give divers the
posed. In this way, standards for facilities (e.g., Moore and opportunity to experience many dive sites over a short time
Polley 2007), visitor impacts (Moore and Polley 2007), and period. In addition, dive sites in Bonaire offer high stan-
crowding levels (Dixon and others 1993) have been sug- dards in terms of biological features (e.g., coral conditions,
gested or implemented in various terrestrial and marine fish communities, presence of turtles) (Hawkins and others
natural areas. However, the efficacy of visitor-derived 1999) compared to most other Caribbean destinations.
standards as a basis for natural area management depends These characteristics make Bonaire an ideal place to
largely on the accuracy of visitor perceptions of actual examine the preferences of divers, without the usual risk of
environmental conditions. answers being biased by lack of experience of some of the
Perceptions can be readily altered if, for example, visi- attributes by respondents (Shafer and Inglis 2000), and to
tors have partaken in a recreational activity in the natural compare the perceptions of divers to the real condition of
environment of interest (Dearden and others 2007). Per- the reefs.
ceptions also depend heavily on demographic factors such The aims of this study were threefold. First, we sought
as nationality (Leujak and Ormond 2007), gender (Priskin to determine which site features divers associated with
2003), experience (Fakeye and Crompton 1991), and edu- positive and negative dive experiences. Second, we
cation (Lothian 2002; Priskin 2003). These sources of examined whether the specific recalled perceptions of
variability in perceptions suggest that, on average, visitor attributes contributing to the choice of most and least
perception may bear little relationship to actual environ- favorite dive sites related to the actual condition of dive
mental conditions, which poses problems if perceptions are sites. To do so, we combined the results of postholiday
to contribute to the design of recreational management questionnaires on dive experiences of tourists with exten-
planning (Noe and others 1997; Newsome and others sive biological surveys of dive sites around Bonaire.
2004). Finally, we explored the effect of gender as one potential
There have been relatively few attempts to compare source of variability in perceptions. Our results are dis-
visitor perceptions of any environment to actual conditions. cussed in the light of the usefulness and limitations of
Most of these studies have focused on the perceptions of perception surveys for managing dive tourism.
campsite conditions, and they generally show a lack of
agreement between visitor and expert assessment of envi-
ronmental impacts at campsites (Martin and others 1989) Methods
or between camper rankings of impacted sites and objec-
tively measured conditions of environmental attributes Study Area
(Farrell and others 2001; Hillery and others 2001).
In other tourism contexts, the results are mixed. Visitor The study was carried out between February and June 2006
perceptions matched actual conditions of riparian areas on the island of Bonaire (12°100 N, 68°150 W), Netherland
(Hoover and others 1985), natural hazards (Méheux and Antilles (Fig. 1). This small island (180 km2) lies approx-
Parkerb 2006), and bathing water suitability (Smith and imately 80 km north of Venezuela (BNMP 2007;
others 1995). In contrast, perceptions of visitors of damage www.bmp.org). The near-pristine condition of the coral
on coral reefs in South Sinai (Red Sea) did not correlate reef surrounding the island and the warm, arid climate are
with actual measurements of damage (Leujak and Ormond two key attractive features of the island for tourism (Uyarra
2007). Similarly, Taylor and Knight (2003) found that and others 2005). Bonaire has been ranked as one of the top
backcountry trail users significantly underestimated their 10 diving destinations in the world for 5 consecutive years

123
Environmental Management (2009) 43:1–16 3

approximately 59,000 tourists per annum, of which 45%


are divers. By comparison, only *15% of visitors to the
Red Sea are divers (Leujak and Ormond 2007). Bonaire
diving tourism attracts a relatively upmarket clientele and
more experienced divers than many other diving destina-
tions (Medio and others 1997; Rouphael and Inglis 2001;
Uyarra and others 2005; Leujak and Ormond 2007).

Diver Surveys

Questionnaire surveys were carried out from February to


June 2006, which encompassed both the low and the high
tourist seasons (TCB 2006). For the purpose of this study,
we required information from tourists who had completed
their holiday to obtain a full list of dive sites visited by
each individual. Thus, we surveyed the dive experiences of
tourists using standardized questionnaires at the departure
lounge of Bonaire Flamingo Airport, before the interna-
tional departures of the following airlines: KLM, American
Airlines, Continental Airlines, and Air Jamaica. These
airlines were the only four international carriers with
scheduled flights to Bonaire. The departure times of these
companies ranged from 0200 (KLM) to 1500 (also KLM).
Because most visitors to Bonaire came on 1- or 2-week
package holidays that, in most cases, ended over the
Fig. 1 Location of dive sites selected by divers exclusively as most weekend, we carried out surveys once every 2 weeks over
favorite (black circles), exclusively as least favorite (white circles), a 24-h period between Saturday night and Sunday night.
and as both most and least favorite (stippled circles) around Bonaire All departures of all four airlines were captured in each of
and Klein Bonaire. Dive sites that were not selected either as most or
the seven 24-h survey periods. During each survey period,
as least favorite dive sites are indicated by a crossed circle
we approached only respondents that were sitting. We
(Diving 2007) and it offers some of the best coral and fish started with the person sitting closest to the entrance and
communities remaining in the Caribbean (Hawkins and proceeded on a next-to-pass basis along the rows of seats.
others 1999). All corals around the island and its satellite Only tourists who had been diving during their holidays
island, Klein Bonaire, are protected down to a depth of were asked to fill the questionnaire. If people were in
60 m as part of the Bonaire National Marine Park (BNMP), couples or families, we only asked the most active diver to
which was designated in 1979 (Van’t Hof 1997). Boat participate in the survey.
anchoring, spearfishing, and collection of any kind are We initially tested the questionnaire by distributing it to
forbidden within the park, although fishing using tradi- two sets of 20 respondents across the four airlines and then
tional methods is permitted. Sizes and species of the catch adjusted the questions for greater clarity to create the final
are regulated. Upon purchase of a park tag, divers can dive standardized questionnaire. Time and budgetary constraints
everywhere, except in two small reserves where only sci- limited our sample size to 200 respondents.
entific diving is allowed (BNMP 2007; www.bmp.org). The questionnaire had three parts. The first section was
The economy of Bonaire depends heavily on the diving aimed at characterizing the diving clientele of Bonaire.
industry, which accounts for *35% of the country’s GDP Respondents were asked to provide general information,
(Department of Economics and Labour Affairs of Bonaire including gender, age, nationality, and number of logged
2006; www.bonaireeconomy.org). This importance of dives to date and during their holiday in Bonaire.
diving is largely due to a lack of alternative resources and In the second section, respondents were asked to indi-
tourist attractions, such as sandy beaches or cultural cate which dive sites, of all those visited during the current
monuments existing at other diving destinations (e.g., Red holiday, they ranked as their most and least favorite. For
Sea). Based on figures from the Bonaire tourism board each site, we provided a set of 20 attributes, including a
(TCB 2006) and on the number of marine park fees sold range of biological (fish abundance, rare fish, large fish,
(BNMP, personal communication, 2007), we estimated that coral condition, sponges, color, and turtles), environmental
over the last 10 years, Bonaire has received an average of (surface conditions, entry point conditions, underwater

123
4 Environmental Management (2009) 43:1–16

current, level of difficulty [easy or difficult], visibility, and with yellow painted stones at roadside entry points. In our
weather), type of dive (night, wreck, drift), and other quantitative assessment of dive site attributes, we consid-
external attributes (dive master guidance, dive operator ered all 21 Klein Bonaire sites, and 55 of the 60 sites
service, and number of divers at the site). We presented around Bonaire. We excluded the Town Pier and Salt Pier,
each attribute in opposite fashion for most and least as they offer almost exclusively night dives, and Bise
favorite dive site. For example, we listed the attributes Morto, Boca Slaagbaai, and Willemstoren Lighthouse, due
offering ‘‘positive’’ options (e.g., ‘‘good coral condition’’) to the special logistics required to access them. For each
in relation to the most favorite dive site, whereas for least dive site we assigned scores to or measured the values of
favorite dive sites we presented the negative side of the 14 of the 20 attributes offered in the questionnaire as
same attribute (e.g., ‘‘poor coral condition’’). Both ‘‘easy potential causes of enjoyment. The attributes ‘‘drift dive,’’
dive’’ and ‘‘difficult dive’’ attributes were offered as such ‘‘guidance,’’ and ‘‘service quality’’ depended on the dive
under both most and least favorite dive sites since, operators rather than on the dive sites, hence they were not
depending on the diver, each can make a positive (or measured. We deemed the attributes ‘‘weather’’ and
negative) contribution to the enjoyment of a dive. ‘‘underwater visibility’’ to be very good and relatively
Respondents ticked boxes next to all of the attributes that constant across the island throughout the time of the sur-
had contributed to the selection of their most and least veys (average air temperature of 29°C [unpublished data;
favorite site, and we interpreted this selection as repre- www.wunderground.com 2007] and 30-m underwater vis-
senting the respondent’s recalled conditions of these ibility [Uyarra, personal obsservation, 2006]). Colorfulness
attributes. The 20 attributes were derived from Shafer and of the reef could not be measured objectively, hence it was
Inglis (2000) and Uyarra and others (2005), the former omitted from comparisons of perceived and actual condi-
focusing on the experience of tourists on the Great Barrier tions (but retained in analyses of diver enjoyment). We
Reef and the latter on the selection of holiday destinations measured several indexes of the attributes ‘‘fish abun-
in the Caribbean. Respondents were given the option to add dance’’ and ‘‘coral condition’’ (see below). In the end, each
any attribute contributing positively or negatively to the site was described by 20 scores which matched closely the
enjoyment of the dive if it was not listed in the attributes offered to divers on the questionnaire. To collect
questionnaire. these data we used a variety of methods, which we describe
To obtain a quantitative measure of the level of enjoy- below and summarize in Table 1.
ment provided by most and least favorite dive sites,
respondents were asked to indicate their enjoyment of each Fish Abundance
of their selected two dive sites with a tick on a continuous
line with orientative markers at 0, 25, 50, 75, and 100. The Fish surveys were carried out at 76 dive sites using line
distance of each tick from the zero mark allowed the per- transects (English and others 1997). In order to minimize
centage of maximum enjoyment (100%) to be measured for possible day variation, all fish surveys were carried out
most and least favorite dive sites for each respondent. between 09.00 and 13.00. To characterize the fish com-
Finally, in the third section we provided a map of the munities that divers encounter during a dive, the transects
BNMP which included a list of all dive sites in Bonaire and were run at two depths, 10 and 20 m, at which divers spend
Klein Bonaire, their location, and the access type (boat or most of their diving time (Uyarra, personal observation,
shore). Respondents indicated on this map the sites they 2006). Each transect was 30 m long, which is approxi-
had visited during the current holiday, how many times mately one-third the average distance covered by a
they had visited each site, and whether they had accessed recreational diver from the entry point during a swimming
each site by boat or from shore. Following the protocol (i.e., nondrift) dive in Bonaire (Uyarra, personal observa-
described above, we collected an additional 200 question- tion, 2006), and placed parallel to the shore. Each transect
naires, for a total of 400, with this section only. was surveyed twice by a diver swimming at a constant
During the interviews, we explained the purpose of the speed (3 m min-1). During the first pass we recorded the
study and asked for the consent of the respondents to use species and numbers of all swimming fish within 2.5 m of
the information provided anonymously. In all cases consent the transect line. On the second pass we recorded the same
was given. information on benthic fish species (excluding blennies and
gobies). We estimated numbers when encountering groups
Measurements of Dive Site Attributes of more than 200 fish, which we refer to as ‘‘schools,’’ and
recorded the number of such schools. These data provided
In 2006, the official BNMP map indicated 60 active dive two indexes of ‘‘fish abundance’’: total number of fish and
sites around Bonaire and 21 around Klein Bonaire, all of number of fish schools. Data from the two depths were
which were marked either with permanent moorings or combined by adding them to generate a single value for

123
Table 1 Summary of dive site attributes, as stated in the survey questionnaire, quantitative and qualitative indexes used to represent these attributes, values or scores, and sources of data for
each index
Attribute Index Value/score Source of data

Fish abundance Total number of fish Total number of fish counted on transects Biological surveys—fish surveys
Number of fish schools Counts of fish schools (groups [ 200 fish) at each site Biological surveys—fish surveys
Fish species richness Number of species recorded at each site Biological surveys—fish surveys
Rare fish 1, unlikely; 2, likely; 3, very likely Dive masters observations
Environmental Management (2009) 43:1–16

Large fish Average number of large fish recorded at each site Biological surveys and recreational dives
Coral condition Percentage cover of live coral Average % cover of live coral Biological surveys—photo quadrats
Percentage cover of dead coral Average % cover of dead coral Biological surveys—photo quadrats
Percentage cover of rubble Average % cover of rubble Biological surveys—photo quadrats
Coral species richness Number of species recorded at each dive site Biological surveys—photo quadrats
Structural complexity 1, relatively flat; 2, some structural relief; Personal observations
3, many good structural formations;
4, outstanding formations (e.g., walls)
Sponges Percentage cover of sponges Average % cover sponges Biological surveys—photo quadrats
Turtles Number of individuals captured and recaptured/site (2006) Bonaire Turtle Conservation
Surface conditions 1, calm; 2, choppy; 3, rough Personal quantification
Entry point 1, easy; 2, intermediate; 3, advanced Personal quantification
Underwater current 1, mild; 2, moderate; 3, strong Personal quantification
Easy and difficult dive 1, very easy; 2, easy; 3, intermediate; 4, advanced Personal quantification
Visibility No score assigned
Weather No score assigned
Popularity of dive sites Number of logged dives at each dive site (400 individuals) Questionnaires
Wreck dive 0, absence of wreck; 1,: presence of wreck Personal observation
Night dive 0, day dive; 1, night dive Personal observation
Drift dive No score assigned
Guidance No score assigned
Quality of dive operator No score assigned
5

123
6 Environmental Management (2009) 43:1–16

each of these two indexes for each dive site. We used the formations, such as walls, etc.). We did not measure other
species identification data to provide the third index of signs of coral health, such as breakage and scarring,
‘‘fish abundance,’’ namely fish species richness (i.e., total because these occur at an extremely low frequency away
number of fish species recorded at each site). from stationary points of interest (such as seahorses and
frogfish) (Uyarra and others 2005).
Rare Fish There were therefore five indexes of ‘‘coral condi-
tion’’—percentage cover of live coral, percentage cover of
We asked 10 dive masters working with four different dive dead coral, percentage cover of rubble, coral species
operators to indicate known locations of seahorses, pipefish richness, and structural complexity—and one index for the
and frogfish. These charismatic species are cryptic and rel- ‘‘sponges’’ attribute, percentage cover of sponges.
atively rare, and sought after by divers and divemasters
(Uyarra and Côté 2007). This information was used to assign Sea Turtles
a score to each site reflecting the likelihood of encountering
‘‘rare fish’’ (1 = unlikely encounter, 2 = likely encounter, A score was assigned to each dive site for the ‘‘turtle’’
3 = very likely encounter). attribute based on surveys carried out in 2006 by Sea Turtle
Conservation Bonaire (STCB). Researchers and volunteers
Large Fish from this nongovernmental organization capture and tag all
turtles encountered during methodical surveys of the near-
We compiled our personal sightings of large fish at each shore waters off Bonaire and Klein Bonaire. We used
dive site, taken during the course of the fish and coral capture frequencies at each site as an index of the site-
surveys (inside and outside the transect areas) as well as specific likelihood of turtle sightings.
during recreational dives (n = 2–7 dives per site). ‘Large
fish’ included sharks, rays, barracudas, tarpons, and the Dive Features
three largest Caribbean parrotfish (midnight Scarus coel-
estinus, rainbow S. guacamaia, and blue S. coeruleus). The We assessed the general dive site conditions in terms of
average number of sightings per dive was used as an index ‘‘surface conditions,’’ ‘‘entry point,’’ ‘‘underwater current,’’
of large fish abundance at each diving site. and ‘‘dive difficulty.’’
Repetitive boat/car trips along the coasts of Bonaire and
Coral Condition and Sponges Klein Bonaire while accessing dive sites (n = 4–90 per
site) allowed us to score site-specific general surface con-
The attributes ‘‘coral condition’’ and ‘‘sponges’’ were ditions on a 3-point scale (1 = calm, 2 = choppy,
assessed through surveys of benthic habitat composition. 3 = rough waters).
At each site, we took 40 photoquadrats (i.e., 20 photo- ‘‘Entry point difficulty’’ was scored on a 3-point scale.
quadrats at each depth), spaced evenly above and below A site was rated as having an easy entry (score = 1) when
each of the two transect lines laid out for fish surveys using (a) there was a smooth substratum, such as sand or small
a digital camera in an underwater housing. Each photo- rubble, to access the water; (b) there were no breaking
quadrat covered an area of 1 9 0.5 m, giving a total waves; and (c) there were no shallow reefs, fire coral, or
surveyed area of 10 m2 of benthos per site at each depth. sea urchins which could harm the diver while entering
Percentage cover of live and dead coral, rubble, and the water. The entry point was rated as intermediate diffi-
sponges was measured by digitally overlaying a grid, with culty (score = 2) when the entry site did not fulfill one
10 9 10-cm cells, on each photograph and counting the of the previously mentioned requirements, and difficult
number of cells dominated by each benthic type. Percent- (score = 3) when the entry to the dive site did not fulfill
age cover of each benthic type was averaged across two or more of the requirements. All dive sites which could
photoquadrats separately at each depth and then averaged be accessed only by boat were rated as easy (score = 1) for
between depths. this attribute.
Corals were identified to species, whenever possible. ‘‘Underwater current’’ was also scored on a 3-point
The total number of coral species recorded was used as an scale, reflecting how easy the initial (up-current) part of
index of coral species richness at site. In addition, we each dive is (1 = mild or no current, i.e., diver can easily
classified the structural complexity of each site on the kick against the current; 2 = moderate current, i.e., diver
basis of in situ observations, assigning a score ranging from exerts some kicking force to overcome current; 3 = strong
1 to 4 (1 = relatively flat, 2 = some structural relief, current, i.e., diver must fin energetically to achieve forward
3 = many remarkable coral formations, 4 = outstanding movement and breathing rate increases).

123
Environmental Management (2009) 43:1–16 7

Finally, dive difficulty was assessed based on the scores visitors from each nationality stated in this source and that
assigned to the difficulty of the ‘‘surface conditions, ‘‘entry in our own questionnaire-derived data. To describe the
point,’’ and ‘‘underwater current’’ attributes and an addi- diving clientele of Bonaire, we reduced the impact of
tional feature: the length of swim from shore to reef crest, extremely experienced divers, who had logged more than
beyond which the dives take place. Distance from shore to 1000 dives to date, by transforming the total number of
the reef crest was measured using Google Earth and was logged dives into a categorical variable with six levels: (1)
considered short (score = 1) when 0–50 m, intermediate beginner, 1–10 dives, (2) somewhat experienced, 11–25
(score = 2) when 51–100 m, and long (score = 3) dives, (3) moderately experienced, 26–100 dives, (4)
when [ 100 m. Once all dive sites had scores for the four moderately to highly experienced, 101–250 dives, (5)
attributes contributing to dive difficulty, dives were highly experienced, 256–1000 dives, and (6) extremely
deemed very easy (score = 1) when none of these attri- experienced, 1000 ? dives. We applied chi-square tests to
butes had a high score (i.e., score = 3), easy (score = 2) categorical variables included in the general information
when one of the attributes had a high score, intermediate section of the questionnaire to examine the distribution of
(score = 3) when two of the attributes had a high score, respondents across categories.
and advanced (score = 4) when three or more of the To assess the relative importance of the 20 attributes
attributes had a high score. The dive difficulty scores were presented in the questionnaire for the enjoyment of the
used for both the ‘‘easy dive’’ and the ‘‘difficult dive’’ most and least favorite dives, we counted the number of
attributes considered in the questionnaire. times each attribute was selected. To check whether similar
Significant correlations (all p values \0.01) between our attributes (in opposite directions) affected the selection of
quantitative assessments of the most variable attributes, most and least favorite dive sites, we carried out a Pearson
i.e., ‘‘underwater current’’ and ‘‘dive site difficulty,’’ and correlation between the number of times each attribute was
those reported in diving guides written by diving experts selected when identifying most favorite dive sites and that
who are long-term residents of Bonaire (Van’t Hof 1997; for least favorite dive sites.
Daryanani 2006; Porter 2007) supported the validity of our We examined whether average condition or high vari-
assessments. ability in condition explained why some dive sites were
chosen as both most and least favorite. To do so, we
Popularity of Dive Sites compared the conditions (means) and variability (measured
as coefficients of variation) of all continuously measured
Diver use of each site was derived from the information attributes among sites chosen solely as most favorite, solely
provided in the third section of the questionnaire, which as least favorite, and as both, by applying Kruskal Wallis
was answered by 400 respondents. The total number of tests.
dives made at each site by all people surveyed during their The extent to which most and least favorite dive sites
holiday was used as an index of the ‘‘number of people’’ were enjoyed by divers was inspected by carrying out a
attribute. Although this score probably does not reflect the pairwise comparison between the percentage of maximum
exact number of divers encountered underwater, it should enjoyment provided by the individually selected most
nevertheless reflect the likelihood of encountering other favorite and that for least favorite dive sites. The percent-
people at a given site. ages of maximum enjoyment were not normally distributed
and could not be transformed to meet the assumption of
Other Miscellaneous Attributes parametric testing. Therefore, we used a nonparametric
Wilcoxon test.
The three sites with a wreck, i.e., Hilma Hooker, Front To examine the relationship between diver perception of
Porch, and La Machaca, were assigned a score of 1 for the dive site characteristics and actual site conditions, we
‘‘wreck’’ attribute (no wreck = 0). The Town Pier and Salt compared empirical measurements of those attributes that
Pier were deemed primarily ‘‘night’’ dive sites (day were indicated by respondents as affecting the selection of
dive = 0; night dive = 1). most and least favorite dive sites (e.g., high fish abundance
stated for most favorite dive site or low fish abundance
Statistical Analyses stated for least favorite dive site). It was unfortunately not
possible to use a paired design in this case (e.g., comparing
To check the representativeness of our sample of respon- fish abundances at the sites selected as most and least
dents, we obtained information on the nationalities of favorite by a given respondent) because this unduly
tourists visiting Bonaire during our period of study from a restricted sample sizes. Instead, the comparisons contrasted
report produced by the Bonaire tourism board (TCB 2006). all real values of, for example, fish abundance at dive sites
We carried out a correlation between the proportion of selected because of high fish abundance to all values of fish

123
8 Environmental Management (2009) 43:1–16

abundance at sites selected because of low fish abundance. population of tourists in terms of nation of origin. The
Samples sizes in these comparisons therefore represent the average age (±SD) of respondents was 43 (±13) years
number of respondents that indicated the specific attribute (n = 200), and the sex ratio was evenly divided between
as an affecting factor. None of the attributes was normally men (55%) and women (45%; v21 = 1.64, p = 0.20). Half
distributed and could be transformed to meet the assump- of the respondents were returning visitors to the island.
tion of parametric testing; we therefore used nonparametric Thirty-nine percent of respondents were moderately
Mann–Whitney U tests in all comparisons. We excluded experienced (26–100 dives) divers. Twenty-two percent
the wreck and night dive attributes from this analysis, as were moderately to highly experienced divers and 20%
presence/absence of a wreck and day/night dives were not a were highly experienced. On average (±SD), divers had
matter of perception. To minimize the rate of type I error logged 16 (±8) dives during their holiday in Bonaire, or 2
arising from multiple comparisons, we adjusted p values (±1) dives per day.
throughout using the false discovery rate method (FDR), a
correction deemed to be more appropriate than the often- Diver Enjoyment of Dive Sites and Dive Site
applied Bonferroni method, which has been criticized for Characteristics
being too conservative (Benjamini and Hochberg 1995). To
apply this method, original p values are first ranked from Eighty-four percent of respondents were able to identify
smallest to largest. Adjusted p values are then obtained by the single site they had preferred during their holiday and
multiplying the original p values by the total number of gave reasons for this choice. The identification of least
attributes and divided by the rank of the specific attribute. preferred sites was, however, more difficult: 16% of
Adjusted p values are then deemed to be significant if they respondents declared not to have a least favorite dive site
are lower than the standard significance value (p = 0.05). and 25% did not provide an answer to this question.
We explored the effect of gender on perceptions. We Sixty-four of the 81 available dive sites around Bonaire
used Mann–Whitney U tests, as our data did not meet the and Klein Bonaire were visited by one or more respon-
requirements of MANOVA testing, and applied FDR to the dents. Each dive site was visited by an average of 39 ± 33
results. Other demographic variables, such as nationality of our respondents. Hilma Hooker was the most popular
and experience, could not be considered because of sample site, with 171 visits, followed by Thousand Steps (137), Oil
size limitations. Slick Leap (109), and Angel City (105). Of the 64 sites that
were visited, 21 sites were exclusively chosen as most
favorite and 19 sites were chosen only as least favorite
Results (Fig. 1). The top five favorite dive sites, based on the
number of respondents choosing them, were Angel City
Dive Tourist Characteristics (n = 14 respondents), Something Special (13), Invisibles
(9), Hilma Hooker (9), and Karpata (8)/Small Wall (8), all
We approached 356 tourists at the airport. Of those, 143 located around Bonaire. The five least favorite dive sites
were nondivers and were therefore unsuitable respondents. were Hilma Hooker (8), Town Pier (8), Rappel (6), Bari’s
Thirteen divers declined to answer our survey, resulting in Reef (6), and Front Porch (6). Twenty-four sites were
a 94% response rate for suitable respondents. Of the final chosen by different people as both most and least favorite
200 respondents, 27 individuals did not complete the sec- sites (Fig. 1). The average condition of continuously
tion relating to most favorite dive sites and 50 omitted the measured attributes at these dive sites did not differ sig-
least favorite dive site sections, leading to variable sample nificantly from that at sites selected solely as most or least
sizes in the analyses presented here. Of the 400 divers that favorite (Kruskal-Wallis tests: all p values [ 0.05). In
were asked to complete section 3 of the questionnaire, 295 addition, the variability in condition (as measured by
respondents indicated how many times they visited each coefficients of variation) across all attributes was not sig-
dive site during their holiday. nificantly different at these sites than at those chosen solely
Respondents were from 12 different nationalities, orig- as most and least favorite (Kruskal Wallis test: v22 = 0.06,
inating predominantly from the United States (74%), the n = 51, p = 0.97).
Netherlands (14%), and the United Kingdom (5%). Even The enjoyment experienced by divers when diving at
though the Bonaire Tourism Board recorded 10 additional their favorite site was significantly higher than when diving
nationalities among arrivals in the first half of 2006, the at their least favorite site (median [interquartile range]:
distribution of countries of origin of our respondents was favorite site, 90% [80.89–100.00], versus least favorite,
significantly correlated with that reported by the TCB 48% [25.00–54.72]; Wilcoxon Z = -9.11, n = 111,
(Pearson correlation, r = 0.90, n = 22, p \ 0.001), indi- p \ 0.001). All 20 attributes offered in the questionnaire
cating that our sample was representative of the total received votes from divers as factors affecting dive site

123
Environmental Management (2009) 43:1–16 9

Table 2 Proportion of respondents indicating each attribute as a attributes when selecting their least favorite site. Respon-
factor affecting dive enjoyment, positively for most favorite dive sites dents offered additional attributes contributing to most/
and negatively for least favorite dive sites
least favorite dive site selection. In particular, three divers
Attribute (as stated Proportion of respondents (%) for indicated that site structural complexity increased dive site
in questionnaire) enjoyment, whereas overgrowing algae (two divers), litter
Most favorite Least favorite
dive sites dive sites (three divers), and lack of lobsters (one diver) negatively
affected the enjoyment of a dive.
Fish abundance 80.3 34.0
Individual divers usually selected different attributes as
Color 75.1 25.3
reasons for the choice of most and least favorite dive sites.
Coral condition 74.6 22.0
However, the frequencies at which each attribute was
Visibility 64.7 18.7 selected in most and least favorite dive sites questions were
Weather 43.3 3.3 significantly correlated (Pearson correlation, r = 0.81,
Dive site popularity 40.5 11.3 n = 20, p \ 0.001; Fig. 2). Three factors appeared to be
Turtles 38.1 22.6 particularly important determinants of dive site enjoyment:
Entry point 37.0 12.0 fish abundance, coral condition, and reef color (Fig. 2).
Sponges 36.8 4.6 Thus the high level of enjoyment experienced at favorite
Large fish 35.3 20.0 dive sites was primarily attributed to high fish abundance,
Surface conditions 32.4 10.0 vivid reef colors, and good coral condition, whereas the
Easy dive 30.6 2.0 low level of enjoyment of least favorite dive sites was
Underwater current 28.9 8.7 attributed to low fish abundance, dull reef colors, and poor
Rare fish 25.4 17.3 coral condition. Enjoyment of favorite dive sites was also
Dive master guidance 25.4 2.6 attributed to good underwater visibility, while the lack of
Wreck dive 11.0 2.0 enjoyment of least favorite dive sites was further ascribed
Quality dive operator 11.0 1.3 to the absence of sea turtles (Fig. 2).
Night dive 6.9 3.3
Difficult dive 4.0 4.0 Do Diver Perceptions Match Dive Site Characteristics?
Drift dive 3.5 2.0
Note: Most respondents selected more than one attribute in each case,
A comparison of the conditions of attributes selected as
hence totals exceed 100% affecting the enjoyment of most and least favorite dive sites
revealed that diver perceptions matched actual condition
enjoyment, and these attributes received a significantly for most biological attributes (Table 3).
higher proportion of votes when highlighted as factors Fish abundance, which was the most often cited crite-
affecting the selection of most favorite than least favorite rion selected by respondents as affecting dive site
dive sites (paired t-test t20 = 5.53, p \ 0.001; Table 2). enjoyment (both positively and negatively) (Table 2), was
Respondents highlighted, on average, 7 ± 3 (SD) attributes significantly higher at sites perceived as having more fish
when selecting their most favorite dive site but only 3 ± 2 than at sites perceived as having fewer fish according to

Fig. 2 Relationship between


the number of times each of 20
attributes was cited as
contributing to the enjoyment of
most favorite dive sites and the
number of times the opposite
version of the attributes was
cited as contributing to the lack
of enjoyment of least favorite
dive sites

123
10 Environmental Management (2009) 43:1–16

Table 3 Differences between most and least favorite dive sites in attributes indicated by respondents as affecting dive site enjoyment using
Mann–Whitney U tests
Attribute/index; measured Most favorite dive site Least favorite dive site U p Significance
variable after FDR
Median [Interquartiles] n (no. sites) Median [Interquartiles] n (no. sites)

Fish abundance
Total number of fish 1518 [1161, 2158] 127 (45) 1153 [878, 2022] 46 (20) 2396 0.07
Number of fish schools 1 [0, 3] 127 (45) 0 [0, 1] 46 (20) 1609 \0.001 ***
Fish species richness 44 [41, 47] 127 (45) 41 [38, 44] 46 (20) 1929 0.001 **
Rare fish 3 [1, 3] 43 (20) 3 [2, 3] 24 (17) 495 0.75
Large fish 1 [0, 3] 60 (27) 0 [0, 1] 27 (18) 514 \0.02
Coral condition
% live coral 39.66 [30.52, 43.03] 120 (47) 25.17 [21.44, 38.39] 30 (18) 964 \0.001 ***
% dead coral 33.85 [31.52, 40.46] 120 (47) 33.38 [29.01, 38.22] 30 (18) 1510 0.17
% rubble 5.40 [2.70, 7.14] 120 (47) 13.61 [3.63, 21.79] 30 (18) 1095 0.001 **
Coral species richness 18 [17, 19] 120 (47) 16 [11, 19] 30 (18) 1212 0.005 *
Structural complexity 3 [3, 2] 120 (47) 2 [1, 3] 30 (18) 1121 0.002 **
Sponge % cover 1.62 [0.66, 2.70] 60 (28) 1.43 [0.40, 4.85] 7 (18) 171 0.42
Turtles 0.5 [0, 5] 64 (29) 0 [0, 2.25] 30 (21) 879 0.48
Surface conditions 1 [1, 2] 55 (29) 2 [1, 3] 15 (11) 274 0.03
Entry point 1 [1, 2] 63 (27) 1 [1, 2] 15(12) 443 0.67
Underwater current 2 [1, 2] 47 (27) 2 [1, 2] 13 (9) 171 0.05
Difficult dive 2 [1, 3] 7 (5) 1 [1, 1.5] 6 (4) 9.5 0.14
Easy dive 1 [1, 1] 51 (24) 1 [1, 1] 3 (3) 46 0.78
Popularity of dive site 12 [3, 37] 67 (31) 13 [2, 37.5] 17 (10) 556 0.88
Note: n represents the number of respondents selecting a particular attribute as characterizing their most and least favorite dive site. The number
of dive sites chosen as most and least favorite on the basis of each attribute is given in parentheses. Significant results after applying false
discovery rate (FDR) corrections at the *0.05, **0.01, and ***0.001 significance levels, respectively

two of the three indexes of fish abundance, namely the Perceptions of dive site conditions were affected by
number of fish schools and fish species richness (Table 3). gender. When ‘fish abundance’ was important for site
Several indexes of coral condition differed between dive selection, both men and women divers perceived differ-
sites selected because of coral condition. Most favorite dive ences between most and least favorite dive sites in the
sites chosen on the basis of good coral condition had a number of fish schools (Fig. 3). However, only sites
significantly higher cover of live coral, lower cover of coral selected by male divers differed significantly in the like-
rubble, higher coral species richness, and more complex lihood of sighting large fish (Fig. 3). Moreover, when dive
reef structure (Table 3) than least favorite dive sites chosen sites were selected as most and least favorite on the basis of
on the basis of their poor coral condition. The percentage ‘‘coral condition,’’ only selections made by males differed
cover of dead coral, however, did not differ significantly significantly in actual conditions of percentage cover of
between sites considered to have good coral condition and live coral, percentage cover of rubble, structural com-
those perceived as having poor coral condition (Table 3). plexity, and coral species richness (Fig. 3).
Encounters with large fish initially were significantly
more likely at most favorite dive sites selected by divers
because of the presence of large fish than at least favorite Discussion
sites perceived to lack such fish, although this difference
fell just short of significance after FDR correction. Our study shows that the level of enjoyment experienced
All other attributes which contributed to the choice of by divers depends highly on dive site conditions, and that
most or least favorite dive sites (Table 2), including rare the accurate recall by divers of these conditions depends on
fish, turtles, percentage cover of sponges, popularity, sur- attribute type and diver gender. Our results confirm the
face conditions, entry point, current, easy dives, and known importance of fish and coral condition as key
difficult dives, did not differ significantly between most attributes affecting diving experience, but we identify
and least favorite sites (Table 3). specific aspects of these attributes that are particularly

123
Environmental Management (2009) 43:1–16 11

Fig. 3 Actual condition of


attributes which contributed to
the selection of most favorite
(MF) and least favorite (LF)
dive by male and female divers.
Medians and interquartile
ranges are shown. Significant
results after applying FDR
corrections at the *0.05, **0.01,
and ***0.001 level,
respectively. White circles
represent outliers

appreciated by divers. For fish abundance, these include condition, important indexes are live coral cover, coral
number of fish schools, fish species diversity, and, possibly, species richness, and reef structural complexity. The facts
the likelihood of encountering large fish. For coral that many reef features contribute to diving enjoyment and

123
12 Environmental Management (2009) 43:1–16

that divers are able to appreciate differences in some of These species were not those considered in this study to be
these key biological conditions should provide strong rare, charismatic species (e.g., frogfish, seahorses), which
incentive to maintain overall reef quality. However, the are known to be sought after by divers (Uyarra and others
variable accuracy of visitor perceptions also signals the 2005), since the likelihood of encountering these species
need for complementing this source of information with was similar at most and least favorite sites. Further
real measurements (or expert opinions) when developing investigation of the fish community composition at most
coral reef management strategies. and least preferred sites may shed light on this finding.
‘‘Coral condition’’ was another key attribute used in the
Attributes Contributing to Dive Enjoyment: Perceptions selection of most and least favorite dive sites. Leujak and
Versus Reality Ormond (2007) suggested that divers in the Red Sea per-
ceive the percentage cover of live coral as a sign of reef
Fish- and coral-related attributes are the features that health and that cover in the range of 25–30% is needed in
tourists interacting with the marine environment value the order for a reef to be considered healthy by the majority of
most (Shafer and Inglis 2000; Williams and Polunin 2000; visitors. Our study shows that most and least favorite dive
Uyarra and others 2005). The responses of divers in this sites differed in live coral cover. Interestingly, most
study supported this general consensus by indicating fish favorite dive sites selected because of their ‘‘good coral
abundance and coral condition as two of the main factors condition’’ had a median live coral cover of nearly 40%,
influencing the selection of most and least favorite dive compared to *25% at least favorite sites perceived as
sites. However, a closer examination of the conditions of having ‘‘poor coral condition.’’ These values are higher
these attributes at dive sites revealed that divers perceive than those suggested by Leujak and Ormond (2007) as a
only specific aspects of these attributes well. representation of healthy reefs. This difference could be
‘‘Fish abundance’’ was the attribute that most often explained by differences in general knowledge of coral
affected the selection of most and least favorite dive sites. reefs, nationalities, and/or education between Bonaire and
We had expected to find that diver perception of high fish Red Sea divers (Petrosillo and others 2006; Leujak and
abundance would be supported by actual differences in Ormond 2007) or perhaps by the expectation of local coral
total fish numbers between most and least favorite dive conditions formed by diving at different sites within each
sites. The fish abundance at most favorite dive sites was area. Thus, 25–30% live coral cover may be at the high end
24% higher, on average, than at least favorite dive sites, but of the coral cover spectrum on Red Sea reefs but at the
this difference was not statistically significant. This may be lower end of the range for reefs on Bonaire. Moreover, our
explained if our necessarily limited surveys did not capture study shows that sites perceived as having good coral
natural variability in daily fish abundance or if perceived condition also had overall more coral species, greater reef
fish abundance depends on more than just numbers. There complexity, and less coral rubble. However, we did not find
is evidence for the latter since the perception of high fish any evidence that these sites had a lower cover of dead
abundance appeared to be influenced by the number of fish coral. This is interesting, as Leujak and Ormond (2007) did
schools at a dive site. Fish schools in our study were groups not find an association among the extent of breakage and
containing more than *200 individuals. Fish density abrasion, other signs of reef damage, and the perception of
within schools is usually high compared to that of non- dive sites as being unhealthy. It is possible that breakage,
schooling fish, thus the outcome of encountering such large abrasion, and what we measured as dead coral are signs of
groups may cause an overall impression of high fish damage that are more difficult for divers to identify than
abundance. Moreover, those sites selected as most favorite rubble cover.
because of high fish abundance had more fish species Other biological attributes, such as the likelihood of
present than those selected as least favorite sites because of encountering large fish, rare fish, and, particularly, sea
the lack of fish. Fish diversity was already known to affect turtles, were also frequently mentioned by respondents as
the experience of divers on Bonaire (Uyarra and others affecting the enjoyment of dives (see also Shafer and Inglis
2005), and we now suggest that encounters with different 2000; Williams and Polunin 2000; Uyarra and others
fish species during a dive appear to contribute to the per- 2005). However, the actual likelihood of encountering rare
ception by divers of enhanced fish abundance. Note that fish, large fish, or sea turtles did not differ between sites
most favorite dive sites had, on average, only three more selected as most and those selected as least favorite dive
species than sites selected as least favorite based on fish sites because of the presence (or absence) of these species.
abundance. This small difference suggests either that This could be due to inaccurate estimates of encounter
respondents were remarkably aware of the fish diversity likelihoods derived from our information sources. Alter-
they encountered or that some particularly noticeable natively, fortuitous encounters with such fauna may not
species were present at sites deemed to be most favorite. depend only on their actual abundance. For example,

123
Environmental Management (2009) 43:1–16 13

sightings of rare fish may largely depend on the ability of in these attributes at both sites was not supported by
dive masters (or divers themselves) to find these species empirical comparisons. Therefore, while respondents
(Uyarra and Côté 2007), on diving at the right time of day showed a good aptitude to perceive most biological dif-
(or night), or on sheer good luck. Ensuring a high likeli- ferences between sites, we found a mismatch between the
hood of encounters with large fauna has, in some cases, perception of physical features and their actual conditions.
been sufficient to attract thousands of divers to specific This disparity could be explained by temporal variation in
diving locations. Hamelin Bay in Australia (Newsome and those attributes. The scores we assigned to surface condi-
others 2004) and Stingray City in the Cayman Islands tions, for example, described the average condition of this
(Shackley 1998), for example, attract thousands of divers attribute at each dive site, but individual divers could easily
and snorkelers every year to observe stingrays. have experienced atypical conditions during their visit to a
Environmental attributes such as weather and clear given site. The ‘personal average’ experience of each diver
waters are among some of the most important attributes for for each site, which is often based on a single site visit
choosing tropical holiday destinations and contribute (Uyarra, unpublished results), may therefore differ sub-
greatly to the enjoyment of such holidays (Mercado and stantially from the site average considered in our
Lassoie 2002; Sheller 2004; Uyarra and others 2005). In comparisons.
our study, good weather and good underwater visibility
were the environmental attributes that were most fre- Effect of Gender on Perceptions of Dive Site
quently indicated as positive factors. This is in contrast Conditions
with the small number of respondents indicating poor
weather and low visibility as factors in their choice of least Gender biases in perceptions of human-caused damage in
favorite dive sites. This suggests that the present quality of natural habitats have been noted previously. Priskin (2003),
both attributes is high in Bonaire. Smith and others (1995) for example, found that women perceived most nature-
examined specifically tourist preferences for water color based recreational activities undertaken on the Western
and clarity levels in freshwater recreational environments Australian coast as more damaging than men did. By
and suggested that blue, rather than yellow, waters and contrast, there were no gender differences in the percep-
clarity of 2.2? m are perceived as suitable for bathing. tions of tourism development impacts on an Italian marine
Uyarra and others (2005) also found that clear waters were protected area (Petrosillo and others 2006).
important in the selection of holiday destinations and Our results suggest that gender affects the ability to
contributed to holiday enjoyment. Nevertheless, ours is the perceive differences in reef conditions, especially for coral-
first study to include an ‘‘underwater visibility’’ attribute in and fish-related attributes. Whereas selections made by
a questionnaire aimed at divers and we found that high both male and female divers for most and least favorite
visibility is associated with a positive enjoyment of dives. sites on the basis of fish abundance matched existing dif-
At present, the underwater visibility in Bonaire fairly ferences in actual numbers of fish schools, only male divers
consistently exceeds 30 m (Uyarra, personal observation), perceived differences in coral conditions. Additionally,
which is considered to be excellent, but may change with only male divers perceived and/or recalled accurately the
plans to develop the coastline in the near future (Nemeth abundance of large fish. Such a result could be explained
and Nowlis 2001; Burke and Maidens 2004). by differences in the respective interests of male and
The enjoyment of difficult dives suggested by Ewert and female divers. Male divers are known to be more venturous
Hollenhorst (1997) applied to very few divers. A much and enjoy diving deeper and exploring cavities (Rouphael
higher proportion of divers enjoyed easy diving, for which and Inglis 2001). As a result men might form a more
Bonaire is well known. However, wreck and night dives, representative picture of coral condition and increase their
considered as more adventurous dives, were more fre- chances of encountering large fish.
quently selected as reasons for selecting most favorite than
least favorite dive sites. This may be because these types of Implications for Managing Diving Tourism and Coral
dives offer more unusual, and hence memorable, experi- Reef Conservation
ences in terms of underwater features than dives deemed to
be technically difficult. It is clear that the state of many biological attributes of dive
Most other attributes relating to the dive itself, such as sites affects diving enjoyment. Therefore, to maintain the
current, entry characteristics, and surface conditions, attractiveness of Bonaire, or any other location, as a diving
appeared to be only moderately important in the selection destination, efforts should be targeted at protecting and
of most and least favorite dive sites. Interestingly, when maintaining the condition of all reef attributes at a level
these dive attributes were the stated reasons for choosing that makes dives highly enjoyable and satisfies expecta-
most and least favorite sites, the perception of a difference tions of visitors (Mercado and Lassoie 2002).

123
14 Environmental Management (2009) 43:1–16

Without being specifically informed by results such as satisfaction levels/standards may also differ geographi-
ours, the management actions implemented by the BNMP, cally (e.g., Leujak and Ormond 2007 vs our study) as well
as well as local legislation, have so far been largely as over time, as tourism markets adapt to the changing
effective at satisfying the expectations of divers of reef quality of the natural resources on which they depend
conditions (Uyarra and others 2005) by preventing boats (Jobbins 2006). Moreover, if perceptions of environmental
from anchoring, reducing illegal spear-fishing, and stop- conditions are formed on the basis of personal thresholds,
ping divers from using gloves (BNMP 2007; www.bmp.org it may be too late to implement preventative management
). Similarly, other marine protected areas across the world by the time tourists perceive negative impacts (Leujak
have also been effective at enhancing and protecting some and Ormond 2007). Finally, it is often easier to recall
of the attributes that divers prefer (Alcala 1988; McCl- obvious impacts (e.g., debris on reefs) than less noticeable
anahan and others 1999; Williams and Polunin 2000). Such but equally important ones (e.g., water pollution) (Chin
effectiveness is attributable to the ecosystem-based man- and others 2000). For these reasons, if management
agement approach that marine protected areas represent, strategies are to be based on questionnaires that establish
which is focused on preserving not single attributes but the visitor preferences and perceptions, the relationship
overall complexity of reef ecosystems (Wilkinson 2006). between perceptions and real conditions, as well as the
Beyond overall protection, which lessens stressors and effect of respondent characteristics, will therefore need to
disturbances, more specific measures can be put in place to be established a priori to evaluate the usefulness of
benefit specific reef attributes of interest to divers (Wil- responses. Visitor surveys may be useful for identifying
kinson 2006). For example, educational programs (e.g., via arrays of management strategy options. However, bio-
compulsory dive briefings [Medio and others 1997]) to logical surveys will still be required if the aim is to
prevent reef trampling, stricter fishing regulations or limits develop specific management strategies and assess reef
on catch size (Williams and Polunin 2000) to increase fish condition.
abundance and size, and stricter control over coastal In conclusion, the links among preferences, perceptions,
development to maintain high coral cover and underwater and true conditions of habitat features uncovered here have
visibility (Mora 2008) would be effective management three important implications. They highlight the need to
strategies targeted at enhancing the condition of some of maintain or enhance overall reef attractiveness. They
the reef attributes most valued by divers. confirm the usefulness of questionnaires to establish the
While visitor surveys are clearly useful for establishing preferences of different sectors of the tourism market and
tourist preferences, their role as a basis for the formula- to outline arrays of potentially suitable management strat-
tion of environmental standards or to assess the egies. Finally, they suggest that although questionnaires
effectiveness of management strategies is more ambigu- may be useful sources of information to measure indirectly
ous. Direct comparisons between the recalled experiences some of the more highly visible indexes of environmental
of divers and quantitative data describing dive site con- quality, they do have considerable limitations that can be
dition allow us to conclude that the perceptions of divers overcome with direct habitat observations.
of features deemed to be particularly attractive and
enjoyable are in many cases accurate. However, this may Acknowledgments The authors thank Bonaire National Marine
Park, Wanna Dive, Toucan Diving, Larry’s Wild Side Diving Bo-
not be a general conclusion. Bonaire is unique in terms of naire, Habitat, Bonaire Flamingo Airport, and Chat ‘n’ Browse for
its diving clientele. Divers visiting Bonaire are generally their logistical support. Special thanks go to Elsmarie Beukenboom,
quite experienced compared to those at most other diving Edwin Domacassé, Ramón de León, Fernando Simal, and Simone
destinations (e.g., Red Sea) (Leujak and Ormond 2007). Wackenhut for their substantial help in setting up the project. We also
thank Jokin del Amo, Denice Boelen, Gregory la Croes, Daniel Dı́az,
The accuracy of diver perceptions of true reef condition Aleks Maljkovic, Rowan Martin, Allison Perry, Karel Rosario, John
observed in this study was almost certainly affected by Tayleur, Sixto Trenidad, Elly Versteeg, Marco Visbeen, Johan Visser,
this extensive experience (Priskin 2003). In addition, Sam Williams, Sherman Winklaar, Matthijs Winters, and, especially,
although perceptions were often accurate, this accuracy Jorge W. Ferrón, Fred Fisher, Luis Gorrı́n, Ramón de León, Duvan
Rı́os, and Julio Abraham for field assistance. Susan Porter and Mabel
varied among attributes and among respondents. Biolog- Nava (Sea Turtle Conservation) provided valuable help. We are also
ical attributes were perceived more accurately than grateful to Julie-Ann Frans, Kerenza Rannou, and Crisanta Martha
physical attributes, and gender was a significant source of and to all the tourists who gave up their holiday time to complete our
variability of perceptions. Other demographic factors such survey. We thank Jennifer A. Gill for her statistical and intellectual
support. Funding was provided by Bonaire National Marine Park,
as education and nationality, which could not be exam- Simon Fraser University, PADI AWARE, Reef Conservation UK, and
ined here, may also affect the accuracy of perceptions the Philip Rickett Foundation. M.C.U. was funded by a scholarship
(Priskin 2003; Petrosillo and others 2006; Leujak and from the Gobierno de Navarra and now by Fundación Caja Madrid.
Ormond 2007), and this effect may be more acute in I.M.C. is supported by the Natural Sciences and Engineering
Research Council of Canada.
locations where diver experience is more limited. Diver

123
Environmental Management (2009) 43:1–16 15

References reconciling conflicting uses. USDA Forest Service General


Technical Report RM-120. USDA, Fort Collins, CO
Alcala AC (1988) Effects of marine reserves on coral fish abundances Jobbins G (2006) Tourism and coral-reefs-based conservation: Can
and yields of Philippine coral reefs. Ambio 17:194–199 they coexist? In: Côté IM, Reynolds JD (eds) Coral reef
Barker NHL, Roberts CM (2004) Scuba diver behaviour and the conservation. Cambridge University Press, Cambridge, UK
management of diving impacts on coral reefs. Biological Leujak W, Ormond RFG (2007) Visitor perceptions and the shifting
Conservation 120:481–489 social carrying capacity of south sinai’s coral reefs. Environ-
Benjamini Y, Hochberg Y (1995) Controlling the false discovery mental Management 39:472–489
rate—a practical and powerful approach to multiple testing. Lothian A (2002) Australian attitudes towards environment. Austra-
Journal of the Royal Statistical Society Series B—Methodolog- lian Journal of Environmental Management 9:45–61
ical 57:289–300 Martin SR, McCooL SF, Lucas RC (1989) Wilderness campsite
Burke L, Maidens J (2004) Reefs at risk in the Caribbean. World impacts: Do managers and visitors see them the same?
Resource Institute, Washington, DC Environmental Management 13:623–629
Chin CLM, Moore SA, Wallington TJ, Dowling RK (2000) McClanahan TR, Muthiga NA, Kamukuru AT, Machano H, Kiambo
Ecotourism in Bako National Park, Borneo: visitors’ perspec- RW (1999) The effects of marine parks and fishing on coral reefs
tives on environmental impacts and their management. Journal of northern Tanzania. Biological Conservation 89:161–182
of Sustainable Tourism 8:20–35 Medio D, Ormond RFG, Pearson M (1997) Effect of briefings on rates
Cope R (2003) The international diving market. Travel & Tourism of damage to corals by scuba divers. Biological Conservation
Analyst 6:1–39 79:91–95
Daily GC (2000) Management objectives for the protection of Méheux K, Parkerb E (2006) Tourist sector perceptions of natural
ecosystem services. Environmental Science & Policy 3:333–339 hazards in Vanuatu and the implications for a small island
Daryanani DE (2006) Bonaire dive guide. Sales Deepak Daryanani, developing state. Tourism Management 27:69–85
Kralendijk, Bonaire Mercado L, Lassoie JP (2002) Assessing tourists’ preferences for
Dearden P, Bennett M, Rollins R (2007) Perceptions of diving recreational and environmental management programs central to
impacts and implications for reef conservation. Coastal Man- the sustainable development of a tourism area in the Dominican
agement 35:305–317 Republic. Environmental Development and Sustainability
Diving S (2007) Top 100 readers’ choice Awards. SCUBA Diving 4:253–278
Dixon JA, Scura LF, Vanthof T (1993) Meeting ecological Moore SA, Polley A (2007) Defining indicators and standards for
and economic goals—marine parks in the Caribbean. Ambio 22: tourism impacts in protected areas: Cape Range National Park,
117–125 Australia. Environmental Management 39:291–300
English S, Wilkinson C, Baker V (1997) Survey manual for tropical Mora C (2008) A clear human footprint in the coral reefs of the
marine sciences, 2nd edn. Australian Institute of Marine Science, Caribbean. Proceedings of the Royal Society 275:767–773
Townsville, Australia Nemeth RS, Nowlis JS (2001) Monitoring the effects of land
Ewert AW, Hollenhorst SJ (1997) Adventure recreation and its development on the near-shore reef environment of St. Thomas,
implication for wilderness. International Journal of Wilderness USVI. Bulletin of Marine Science 69:759–775
3:21–26 Newsome D, Lewis A, Moncrieff D (2004) Impacts and risks
Fakeye PC, Crompton JL (1991) Image differences between pro- associated with developing, but unsupervised, stingray tourism at
spective, first-time, and repeat visitors to the Lower Rio Grande Hamelin Bay, Western Australia. International Journal of
Valley. Journal of Travel Research 30:10–16 Tourism Research 6:305–323
Farrell T, Hall TE, White DD (2001) Wilderness campers’ perception Noe FP, Hammitt WE, Bixler RD (1997) Park user perceptions of
and evaluation of campsite impacts. Journal of Leisure Research resource and use impacts under varied situations in three national
33:229–250 parks. Journal of Environmental Management 49:323–336
Gardner TA, Côté IM, Gill JA, Grant A, Watkinson AR (2003) Long- Petrosillo I, Zurlini G, Corlianò ME, Zaccarelli N, Dadamo M (2006)
term region-wide declines in Caribbean corals. Science 301: Tourist perception of recreational environment and management
958–960 in a marine protected area. Landscape and Urban Planning
Harriott V, Davis D, Banks SA (1997) Recreational diving and its 79:29–37
impact in marine protected areas in Eastern Australia. Ambio Porter S (2007) Bonaire shore diving made easy. Kralendijk, Bonaire
26:173–179 Priskin J (2003) Tourist perceptions of degradation caused by coastal
Hawkins JP, Roberts CM, Van’t Hof T, de Meyer K, Tratalos J, nature-based recreation. Environmental Management 32:
Aldam C (1999) Effects of recreational scuba diving on 189–204
Caribbean coral and fish communities. Conservation Biology Roberts CM (1997) Connectivity and management of Caribbean coral
13:888–897 reefs. Science 278:1454–1457
Hawkins JP, Roberts CM, Kooistra D, Buchan K, White S (2005) Rouphael AB, Inglis GJ (2001) ‘‘Take only photographs and leave
Sustainability of scuba diving tourism on coral reefs of Saba. only footprints’’? An experimental study of the impacts of
Coastal Management 33:373–387 underwater photographers on coral reef dive sites. Biological
Hillery M, Nancarrow B, Griffin G, Syme G (2001) Tourist Conservation 100:281–287
perception of environmental impact. Annals of Tourism Shackley M (1998) ‘Stingray City’—managing the impact of
Research 28:853–867 underwater tourism in the Cayman Islands. Journal of Sustain-
Hoegh-Guldberg O (1999) Climate change, coral bleaching and the able Tourism 6:328–338
future of the world’s coral reefs. Marine and Freshwater Shafer CS, Inglis GJ (2000) Influence of social, biophysical, and
Research 50:839–866 managerial conditions on tourism experiences within the Great
Hoover SL, King DA, Matter WJ (1985) A wilderness riparian Barrier Reef World Heritage Area. Environmental Management
environment: visitor satisfaction, perceptions, reality, and man- 26:73–87
agement. In: Johnson RR, Ziebell CD, Patten DR, Ffolliiot PF, Sheller M (2004) Natural hedonism: the invention of Caribbean
Hamre RH (eds) Riparian ecosystems and their management: islands as tropical playgrounds. In: Duval DT (ed) Tourism in

123
16 Environmental Management (2009) 43:1–16

the Caribbean: trends, development, prospects. Routledge, Uyarra MC, Côté IM, Gill JA, Tinch RRT, Viner D, Watkinson AR
London (2005) Island-specific preferences of tourists for environmental
Smith DG, Croker GF, Mcfarlane K (1995) Human perception of features: implications of climate change for tourism-dependent
water appearance. 1. Clarity and color for bathing and aesthetics. states. Environmental Conservation 32:11–19
New Zealand Journal of Marine and Freshwater Research 29: Van’t Hof T (1997) New guide to the Bonaire Marine Park. Harbour
29–43 Village Beach Resort, Kralendijk, Bonaire
Taylor AR, Knight RL (2003) Wildlife responses to recreation and Walters RDM, Samways MJ (2001) Sustainable dive ecotourism on a
associated visitor perceptions. Ecological Applications 13: South African coral reef. Biodiversity and Conservation
951–963 10:2167–2179
TCB (2006) Bonaire tourism. Half-year statistics report 2006. Wilkinson C (2006) Status of coral reefs of the world: summary of
Kralendijk, Bonaire threats and remedial action. In: Côté IM, Reynolds JD (eds)
Tratalos JA, Austin TJ (2001) Impacts of recreational SCUBA diving Coral reef conservation. Cambridge University Press, Cam-
on coral communities of the Caribbean island of Grand Cayman. bridge, UK
Biological Conservation 102:67–75 Williams ID, Polunin NVC (2000) Differences between protected and
Uyarra MC, Côté IM (2007) The quest for cryptic creatures: Impacts unprotected reefs of the western Caribbean in attributes preferred
of species-focused recreational diving on corals. Biological by dive tourists. Environmental Conservation 27:382–391
Conservation 136:77–84

123

You might also like