Abstract: This study examines the distribution and diversity of six common families of reef fishes (Pomacentridae, Chaetodontidae, Pomacanthidae, Acanthuridae, Labridae and Scaridae) in two reef zones at 20 sites along the Egyptian coast of the Red Sea. A total of 91 fish species were recorded, representing 41 genera. The Sharm El-Sheikh region has the highest number of species (75), whereas the northern part of the Gulf of Aqaba has the lowest number (40). The highest average abundance was recorded at Ras Mohammed (2921 fish/600 m3). Most individuals belonged to the families Pomacentridae (52.1%), followed by the Acanthuridae (19.5%), Labridae (14.7%), Chaetodontidae (7.1%), Scaridae (5.2%) and Pomacanthidae (1.3%). Average species richness ranged from 6.2 (northern Gulf of Aqaba) to 10.2 (Sharm El-Sheikh). In general, the number of species of all selected families was higher on the reef slope than on the reef flat. There was no clear zonation (restriction to reef flat versus reef slope) in the distribution of Pomacentridae, Acanthuridae, Labridae and Scaridae along the Egyptian reefs, although preferences were evident. On the other hand, certain species of Chaetodontidae and Pomacanthidae clearly preferred the reef slope. This study reveals clear geographical differences in the fish communities of the Egyptian Red Sea reefs, differences that should be taken into consideration for crucial future conservation efforts here.
INTRODUCTION
Fishes are the most visible and important mobile component in the coral reef ecosystem (Letourneur et al., 2000). The high diversity of coral reef fish communities includes a large within-habitat component (Goldman and Talbot, 1976), wherein large numbers of species may co-occur in a very small space. Most coral reefs show a clear zonation and, within a reef, numbers and types of organisms may vary in the different zones.
Fishes are also influenced by this reef zonation, which is reflected in great spatial heterogeneity in terms of substrate composition and structural complexity (Done, 1982; Glynn et al., 1996; Rajasuriya et al., 1998). Accordingly, certain fish species or assemblages are characteristic for certain zones (Bell and Galzin, 1984; Harmelin-Vivien, 1989; Alwany, 1997; McClanahan and Arthur, 2001; Garpe and Öhman, 2003). They may be selective or non-selective, obligate, facultative or opportunistic in relation to their habitat (Bergman et al., 2000). Many reef fishes associate with particular microhabitats within the above-mentioned zones (Sale, 1991), although the importance of such associations in determining larger-scale patterns of distribution and abundance appears to vary widely among species (Munday, 2000).
While the Red Sea fish fauna is taxonomically quite well known compared with other parts of the tropical Indo-Pacific Ocean, the structure of coastal fish communities has been less well investigated (Khalaf and Kochzius, 2002). The present study investigates community structure at local (reef flat versus slope) and at latitudinal (23-29° N) spatial scales. Information about the distribution and species diversity of reef fish communities along the Egyptian Red Sea coast is a prerequisite for future conservation and management measures.
MATERIALS AND METHODS
Study Area
Twenty reefs were studied in the summer of 1999 along the Egyptian coast
of the Red Sea, spanning a distance of 1200 km (Fig. 1 and
Table 1) from 23° N to 29° N. The reef type varied
between coastal, barrier and island reefs along the Egyptian Red Sea coast.
The study sites cover five different geographical regions (Gulf of Aqaba, Sharm
El-Sheikh, Hurghada, Hamata and Shalateen) of the Egyptian coast of the Red
Sea (Table 1).
| Fig. 1: | The northern Red Sea showing the Egyptian coast and the study sites. For site names, compare numbers in Table 1 |
| Table 1: | List of sites and regions showing their locations, Reef Type (RT), site codes and brief description of Reef Flat (RF) and Reef Slope (RS) along the Egyptian Red Sea coast. (C = Coastal, I = Island, B = Barrier, WA = Wave Action, TW = Turbid Water, CU = Current) |
Families Investigated
The species composition of the Egyptian Red Sea reefs is examined here based
on six common families, representing a wide range of species of different sizes
and trophic affiliations: damselfishes (Pomacentridae), butterflyfishes (Chaetodontidae),
angelfishes (Pomacanthidae), surgeonfishes (Acanthuridae), wrasses (Labridae)
and parrotfishes (Scaridae). These fishes are highly visible and represented
about 88% of total fish population in the northern Red Sea (Alwany et al.,
unpublished data) on shallower reefs. The goal was not to provide a taxonomically
exhaustive list: Some species were present in the study area but were absent
inside our investigated transects. Fish nomenclature follows that in FishBase
(http://www.fishbase.com).
Field Methods
Underwater visual census techniques have been used to record fish densities
on reefs for fifty years (Brock, 1954) and form the basis for population ecology
studies and management decisions (Harmelin-Vivien et al., 1985). Furthermore,
they provide rapid estimates of the relative abundance and distribution of reef
fishes (Samoilys and Carlos, 2000). Here, members of the six families were counted
using this approach along transects (100 χ6 χ1 = 600 m3)
on the reef flat (RF, depth: 0.5-1 m) and reef slope (RS, depth: 1-8 m). On
the reef flat, fishes were observed using snorkeling, on the reef slope using
SCUBA during day-time from 1100 to 1400.
Data Analysis
The data were analysed statistically using the software packages PRIMER
(V 5.0) and SPSS (V 11.5). The species present in the two zones were compared
using the Bray-Curtis similarity index. Species richness was expressed by considering
the number of species (D) and species diversity and homogeneity were determined
using the Shannon-Wiener diversity index (H') and the evenness index (J') (Pielou,
1966). These parameters were calculated for each site by pooling data from the
sample replicates. When necessary, abundance data were square root transformed
to produce normality and homogeneity of variance.
RESULTS
General Distribution and Abundance
The species recorded at different reef types at twenty sites along the Egyptian
Red Sea coast are listed in Table 2. In the six selected families,
a total of 91 fish species representing 41 genera were counted. The Sharm El-Sheikh
region had the highest number of species (site 10-RM, 75 species), while the
northern part of Gulf of Aqaba had the lowest number (site 3-NE, 40 species).
The highest average abundance was also recorded at site 10-RM (2921 fish/600
m3), with the lowest value at sites 1-BS and 12-SI (222 fish/600
m3). Table 2 shows that most individuals belonged
to the Pomacentridae (52.1%, 29 species), follow by the Acanthuridae (19.5%,
8 species), Labridae (14.7%, 26 species), Chaetodontidae (7.1%, 11 species),
Scaridae (5.2%, 12 species) and Pomacanthidae (1.3%, 5 species). In general,
the number of species of all families was higher on the reef slope than reef
flat. The results of one-way ANOVA, based on diversity indices and showing the
effect of factors site and zone, are given in Table 3. The
number of species varied highly significantly between the two zones (p<0.001),
while the influence of sites was not significant (p = 0.632). The number of
individuals also differed significantly between zones (p = 0.007), but not among
sites (p = 0.233). Average species richness ranged from 6.2 at site 3-NE to
10.2 at site 7-RN (p = 0.874 among sites and p < 0.001 among zones). The
highest evenness index (J�) was recorded at site 17-MS (0.9), while site 4-CA
yielded the lowest value (0.6) (p = 0.151 among sites and p = 0.446 among zones).
Average Shannon-Wiener diversity (H�) varied between 3.8 at site 15-MI and 2.4
at site 4-CA (p = 0.188 among sites and p = 0.057 among zones).
The Bray-Curtis similarity index cluster analysis of the reef flat according to fish abundance is shown in Fig. 2. This dendrogram revealed two main groups: 1. Gulf of Aqaba, Sharm El-Sheikh, Hurghada and Hamata regions (11 sites) and 2. Shalateen region (6 sites + 2 sites from the Sharm El-Sheikh region). One site (12-SI) was not assigned to either group identified above. The reef flat at this site had many rocks and was not suitable for fishes. According to this grouping, the reef flat of the Egyptian Red Sea coast is therefore divided into north (29°-24° N) and south geographical regions (24-23° N). On the other hand, the cluster analysis based on fish abundance on the reef slope revealed four main groups (Fig. 3): 1. Sharm El-Sheikh and Hurghada regions (5 sites), 2. Shalateen region (6 sites), 3. Hamata region (3 sites) and 4. Gulf of Aqaba (6 sites).
| Table 2: | Presence (+) and absence (-) of reef fish species, includes the number of species, number of individuals, species richness (D), evenness (J�), Shannon-Wiener (H�) and percentage of families along the Egyptian coast of the Red Sea (for abbreviations see Table 1) |
| Table 3: | Results of one-way ANOVA performed on diversity indices showing the influence of sites and zones (**p<0.01, ***p<0.001) |
| Fig. 2: | Cluster analysis dendrogram (Bray-Curtis similarity, square root transformation, group average) showing associations of 20 sites on the reef flat along the Egyptian Red Sea coast |
| Fig. 3: | Cluster analysis dendrogram (Bray-Curtis similarity, square root transformation, group average) showing associations of 20 sites on the reef slope along the Egyptian Red Sea coast |
This analysis clearly differentiates the geographical distribution of fish assemblages on the Egyptian reef slope. Accordingly, the reef slope (from reef edge to 10 m depth) of the Egyptian Red Sea coast can be divided into four geographical regions based on fish assemblages: 1. Gulf of Aqaba region (27-29° N), 2. Sharm El-Sheikh and Hurghada region (25-27° N), 3. Hamata region (24-25° N) and 4. Shalateen region (23-24° N). Thus, the reef flat and reef slope yielded very similar results, although the latter provided a somewhat more differentiated picture.
Pomacentridae (Damselfishes)
Pomacentridae had by far the highest number of species and individuals in
both reef zones along the Egyptian Red Sea coast (Fig. 4).
In general, Amblyglyphidodon leucogaster and Pomacentrus sulfureus
were the most frequent species here (represented at all sites). The least common
species were ascyllus trimaculatus, Neopomacentrus xanthurus and
P. leptus. This family is not restricted to a particular zone but does
show preferences. Two groups were distinguished: The first preferred the reef
flat (Plectroglyphidodon leucozona and P. lacrymatus), while the
second preferred the reef slope (D. aruanus, D. marginatus,
Chromis dimidiata, C. weberi, C. ternatensis, P. sulfureus
and A. leucogaster). C. dimidiata was the most abundant damselfish
species at all investigated sites (142.2 fish/600 m3); it clearly
prefers the reef slope (95.4%) over the reef flat (4.6%) due to its feeding
activities (planktonic feeder). Overall, this family shows no clear geographical
trend along the Egyptian reefs, although Abudefduf sordidus, for example,
apparently preferred the southern region (Table 2). The one-way
ANOVA showed an influence of species (p<0.001), sites (p = 0.002) and zones
(p<0.001, Table 4) on the abundance.
Chaetodontidae (Butterflyfishes)
The average number of species and individuals was higher on the reef slope
than on the reef flat (Fig. 4). Chaetodon austriacus
was the most abundant butterflyfish species (11.7 fish/600 m3),
followed by C. paucifasciatus (7.3 fish/600 m3). Chaetodon
auriga and C. fasciatus were recorded at all sites. A zonation preference
(i.e., more abundant on reef slope) was very clear for some species, such as
C. semilarvatus and Heniochus intermedius, even though the latter
two were hardly restricted to the reef slope at all sites. On the other hand,
a preferred geographical distribution was recorded for C. larvatus and
C. mesoleucos, the former being restricted to southern Egyptian reefs
(Hurghada region and south from 23 to 27° N), the latter to the Shalateen
region (23 to 24° N). Neither species was recorded in the Gulf of Aqaba.
Furthermore, C. semilarvatus and C. melannotus were not observed
at any Gulf of Aqaba sites (north) and were rare in the Shalateen region (south).
One-way ANOVA indicated that the butterflyfish abundances varied highly significantly
among the species, sites and zones (p<0.001).
| Fig. 4: | Average number of species and individuals for six families along the Egyptian Red Sea coast on the Reef Flat (RF) and Reef Slope (RS) |
| Table 4: | Results of one-way ANOVA performed on fish abundance showing the influence of fish species, sites and reef zones (*p<0.05, **p<0.01, ***p<0.001) |
Pomacanthidae (Angelfishes)
This family had the lowest number of species and individuals, with values
being higher on the reef slope (Fig. 4). Centropyge multispinis
was the most abundant representative (3.8 fish/600 m3) and was clearly
restricted to the reef slope of all sites. Pomacanthus imperator was
the rarest angelfish, present in only 5 of 20 sites. Pygoplites diacanthus
was the most frequent representative on the Egyptian reefs (represented
at all sites) and was equally present on both reef zones. Geographically, the
Shalateen region (in the south) was richer in angelfishes (5 species). P.
asfur and P. maculosus were found frequently at the southern sites
and the former was very rare in the northern sites. One-way ANOVA shows no significant
difference in angelfish abundance among sites (p = 0.954), but the difference
was highly significant among species and zones.
Acanthuridae (Surgeonfishes)
The surgeonfishes were the most regularly distributed fish group along the
Egyptian reefs, except in the northern part of the Gulf of Aqaba. The number
of species was higher on the reef slope, but the number of individuals was nearly
the same in both zones (Fig. 4). Acanthurus nigrofuscus
was the most abundant surgeonfish (36.7 fish/600 m3). Some acanthurids
preferred the reef flat (A. nigrofuscus, A. sohal and Ctenochaetus
striatus), while others preferred the reef slope (A. gahhm, Naso
lituratus, N. unicornis, Zebrasoma desjardinii and Z.
xanthurum). No geographical trend was observed in the distribution of acanthurids,
except that A. gahhm was not recorded in the Gulf of Aqaba (27°
to 29° N). One-way ANOVA indicated that the surgeonfish abundances varied
highly significantly among species. In contrast, the effect of sites and zones
was not significant (p = 0.053 and 0.795, respectively).
Labridae (Wrasses)
Wrasses had the second highest number of species on the reef slope after
damselfishes. The average number of species was higher on the reef slope than
on the reef flat, but individual numbers were very similar in both zones (Fig.
4). Thalassoma rueppellii was the most abundant species (63.7 fish/600
m3), followed by the cleaner wrasse Labroides dimidiatus (12.6
fish/600 m3). Over all sites, three species (Gomphosus coeruleus,
L. dimidiatus and T. rueppellii) were recorded frequently and regularly.
There was no clear zonation preference in the distribution of wrasses: Some
species (e.g., T. rueppellii) preferred the reef flat, others the reef
slope (e.g., G. coeruleus and L. quadrilineatus). The remaining
species roved frequently in all zones. No clear geographical pattern of wrasse
distributions along the Egyptian reefs emerged. One-way ANOVA showed that the
influence of species is highly significant (p<0.001), but the influence of
sites and zones was not significant (p = 0.284 and p = 0.642, respectively).
Scaridae (Parrotfishes)
The average number of scarid species and individuals was somewhat higher
on the reef slope than on the reef flat (Fig. 4). Hipposcarus
harid was the most abundant parrotfish (10.2 fish/600 m3).
On the other hand, Chlorurus sordidus was the most frequent species
(represented in all sites). No clear zonation preference was evident, except
for Cetoscarus bicolor, which preferred the reef slope (75%) over the
reef flat (25%). Most other species had very similar abundances in both zones.
No clear geographical pattern of parrotfish distributions emerged. One-way ANOVA
showed a highly significant difference in parrotfish abundance among species
and sites (p<0.001) and a significant difference among zones (p = 0.022).
DISCUSSION
Local populations of marine reef fishes often show great spatial variation in abundance (Holbrook et al., 2000). This variation results from a combination of many physical and biological factors that affect fish distribution and diversity. One example is the different distribution of fish groups, whereby herbivores are generally much more abundant in the shallow than in deeper reef zones (Bouchen-Navaro and Harmelin-Vivien, 1981). This probably reflects the richness of algae in this zone. In contrast, carnivorous fishes are usually more abundant on the reef slope.
Jennings et al. (1996) reported that the differences in the diversity and abundance of reef fish communities may also be attributed to spatial and temporal variations in recruitment (e.g., Doherty, 1991), habitat effects (e.g., Williams, 1991) and other factors. Habitat availability (Caley et al., 1996) and habitat preferences (Caley, 1995; Tolimieri, 1998b) also play a role. A variety of other factors, such as reef types or environmental conditions, can also interact with habitat associations to influence patterns of distribution and abundance (Wellington, 1992). Patterns of habitat use might also change from reef to reef depending on the intensity of interactions with other species (Werner et al., 1983). Finally, settlement and post-settlement processes influence abundance and distribution patterns (Gutiérrez, 1998).
General Distribution
Pomacentridae and Labridae dominated the fish fauna in terms of species
richness along the Egyptian Red Sea reefs and Pomacentridae were most abundant.
This result echoes the situation on the Great Barrier Reef and in New Caledonia,
where Pomacentridae and Labridae are the dominant fishes (pomacentrids contribute
the highest percentage of species, followed by labrids; Williams and Hatcher,
1983; Letourneur et al., 1997). The situation is also very similar in
the Indian Ocean and central Pacific, except that the ranking of the first two
groups is reversed: labrids contribute the highest percentage of species, followed
by pomacentrids (Letourneur, 1996; Gladfelter et al., 1980). Closer to
the study area, Khalaf and Kochzius (2002) demonstrated the same situation (as
in the Indian Ocean) on a Jordanian reef (Gulf of Aqaba, Red Sea). In contrast,
Caribbean reefs have a different composition (fewer Labridae, more Serranidae;
Gladfelter et al., 1980; Pattengill et al., 1997). In the present
study, pomacentrids also had the highest number of individuals. In agreement
with our results, Khalaf and Kochzius (2002) demonstrated higher abundances
at the reef slope versus shallow reefs (reef flat), due to schooling planktivorous
fishes. On the Great Barrier Reef, Russ (1984b) demonstrated that the assemblages
of herbivorous fishes (most of the species we investigated are herbivorous)
on the reef flat tend to have relatively low numbers of species and individuals.
Our results clearly support this relationship on the Egyptian Red Sea reefs.
Habitat Preferences
The specialization of some species on reef flat habitats and of others on
reef slope habitats shows a certain degree of habitat segregation in the investigated
Red Sea reefs. In principle, these species may either be occupying their preferred
habitat (e.g., for feeding, nesting, sheltering, etc.) or be excluded from other
habitats by competition. On the other hand, Russ (1984b) listed a few species
that are almost equally abundant in all zones (e.g., Chlorurus sordidus,
Naso unicornis and C. gibbus) on the Great Barrier Reef. He also
reported certain species as being most abundant in the shallow (reef flat) zones
(Ctenochaetus striatus and Scarus frenatus) or deep (reef slope)
zones (S. niger). Our results agree with those of Russ for some species
and disagree for others. Thus, in the present study, C. sordidus and
C. gibbus were equally abundant in all zones in agreement with Russ, but
N. unicornis preferred reef slope habitats (86.9%). Agreement was also
achieved for C. striatus (reef flat), but S. frenatus was equally
abundant in both zones. Our results for S. niger also agree with the
findings of Russ (1984b).
Our results contrast with those of Ormond et al. (1996), who reported that damselfish are restricted to certain parts of the reef along the southern coast of Sinai (Egyptian Red Sea). In the present study, however, some pomacentrids did prefer certain zones, but were also recorded in other zones (although in low abundances). For instance, Chromis dimidiata preferred the reef slope (95.4%) over the reef flat (4.6%) due to its feeding activities (planktivorous, Masuda and Allen, 1993). Ormond et al. (1996) stated that many damselfishes were more abundant in some reef zones than others, some being restricted to a single zone, e.g., Pomacentrus sulfureus, which was only recorded on the reef edge. The present study, however, indicated that P. sulfureus was not restricted only to the reef edge, but also found on the reef flat (6.1 fish/600 m3, 20.1%). On the other hand, the zonation preferences of butterflyfishes were very clear for some species. For example, Chaetodon semilarvatus and Heniochus intermedius were exclusively restricted to the reef slope of all investigated sites (100%) and were never recorded on the reef flat.
Geographical Distribution
Sheppard et al. (1992) examined the species richness of the families
Pomacentridae, Chaetodontidae, Acanthuridae, Labridae, Scaridae and Serranidae
(i.e., 5 of the 6 families investigated here) in the different regions of the
northern Red Sea, Gulf of Suez and Gulf of Aqaba. They noted a decreasing number
of species, from 100 at the southern coast of Egypt to 75 at the northern Gulf
of Aqaba. Our results yielded virtually the same percentage between the two
regions. In the present study, a total of 76 species (belonging to the studied
families) were recorded in the southern region of the Egyptian Red Sea, while
a total of 54 species were recorded in the northern Gulf of Aqaba (Sheppard
et al., 1992, 75.0%; our value: 71.1%). This small difference between
the two percentages probably reflects the number of species of Serranidae (not
considered in the present paper) and Pomacanthidae (not considered by Sheppard
et al., 1992) in the Red Sea.
The structure of both fish and coral communities differs between the northern and southern Red Sea reefs (Sheppard and Sheppard, 1991), with the northern part showing higher diversity. Rich coral communities are also present in the central part of the Red Sea (Roberts et al., 1992). Our results revealed a more detailed geographical differentiation in terms of fish communities for the northern part. Based on reef flat assemblages, we can further divide this part of the Egyptian Red Sea coast into a north (24-29° N) and south regions (23-24° N). On the other hand, the reef slope yields four regions: Gulf of Aqaba (27-29° N), Sharm El-Sheikh and Hurghada (25-27° N), Hamata (24-25° N) and Shalateen (23-24° N, Fig. 3).
Shackley (1999) considers this coast to be one of the fastest growing resort areas in the world. Hawkins and Roberts (1996) predicted a 13-fold expansion of coastal tourism in Egypt and estimated that this would be associated with a rise from 19 to >26% of Egypt’ s reefs being affected by tourism. In fact, the number of beds increased here from 64,500 to 145,200 between 1998 and 2002 alone and 84.5% of beds now under construction are in the Red Sea and South Sinai area (Ibrahim and Ibrahim, 2006). This trend will no doubt entail many critical environmental problems in the future for the coral habitats here. Such projected ecosystem degradation will require a huge scientific effort to provide information and support conservation and management measures, with one of the priority tasks being a better understanding of fish distribution and abundance. This study has grasped the opportunity to provide such baseline data as long as these reefs and their fish communities are still in a near-natural state.
CONCLUSIONS
The Red Sea is known for its high diversity and high degree of endemism, of coral reef fishes. At the same time, tourism along the Egyptian Red Sea coast is expanding at an unprecedented pace. Nowhere worldwide may it be more important to determine the pre-exploitation status of coral reef fish communities. This contribution meets that challenge by examining six common families of coral reef fishes-which together make up a significant percentage of the overall fish population; provide most of the indicator fish species for coral reef health- along a 1200 km stretch of this coast that will face the greatest impact. We show that the distribution of these fishes is not uniform, either geographically (north-south) or habitat-wise (reef flat versus reef slope). Such information is critical in helping to evaluate tourism expansion, to control the protected status of these reefs and to determine future management and conservation efforts.
ACKNOWLEDGMENTS
This study was supported by a grant of the ÖAD, Austria. We thank Prof. M.H. Hanafy, Prof. E. Thaler, Prof. S. El-Etreby and Dr. M.A. Kotb for their suggestions and encouragement in planning this study. This work would not have been possible without the kind assistance of the Department of Marine Science (Suez Canal University, Egypt) and the Institute of Zoology and Limnology (University of Innsbruck, Austria). The authors wish to thank Prof. Joerg Ott, Department of Marine Biology at the University of Vienna in Austria, for providing working space.