A baseline survey of macrobenthic fauna inhabiting shallow subtidal habitats was carried out around Huwar islands during two different cruises in May 2003. A total of 144 quadrats (0.25 m2), 48 subsurface seawater samples and replicate samples of sediments from 16 sites resulted in a collection of 119 species belonging to macrobenthos. Physicochemical parameters of the seawater namely depth (0.2 – 1.5 m), water temperature (28–32°C), pH (7.5–8.0), and dissolved oxygen (5.7–8.5 mg l−1) did not vary significantly between eastern and western shores. However, marked variation in salinity (46.5–54.5 ‰) and chlorophyll-a content of seawater (0.58–6.3 μ g l−1) were found. There was some indication that the eastern coast is more biologically diverse. Sediment organic content was slightly higher along the eastern coast. The substrate was typical of other areas around Bahrain and composed of sand, mud and harder encrustations. Grain size analysis indicated the dominance of fine to very fine sand fractions. Microbial indicators of seawater contamination were found in five of the sixteen sites and Enterococci were found at almost all sites with higher numbers recorded along the western coast. Similar results were found for Salmonella–Shigella counts. The total coliform count was over 100 Colony Forming Units with the exception of a few sites, and a considerable total vibrio count was also recorded. The island's coastal water was nevertheless considered relatively clean from pathogenic bacteria as compared to other areas around Bahrain. A monitoring of macrobenthos and pathogenic bacteria should be conducted in order to endure good water quality around Huwar. A survey of this sort requires dedicated time and resources in order to establish a monitoring program because of the fast rate in urban developments within this region.
The Archipelago of Huwar [NB: Huwar is spelled Hawar in some other references] is located 26 km away southeast from the main island of the Kingdom of Bahrain 25° 50.5″ N and 50° 39.5″ E, comprising six large and over thirty smaller islands, covering 51.5 km2 representing around 20% of the total land area of Bahrain. The coastal area extends to about 17 km and the main island is three kilometres in width (Figure 1). These limestone islands have cliffs at maximum height of around 13 m, and they are at or near sea level with shallow surrounding waters at maximum depths of around six meters (Aspinall et al., 2003). The salinity is normally around 42–58% (Vousden, 1985) and the tide is semidiurnal, typical of Bahrain and the Arabian Gulf. The spring tide does not exceed 2.5 m (Aspinall et al., 2003). The Huwar islands are considered to be one of the least spoilt natural terrestrial and marine ecosystems in the Arabian Gulf (Hill, 2005; King, 1999). The islands enjoy relatively pristine marine and coastal habitats and encompass some endangered wildlife due to remoteness (as well as military restriction in the past).
Different species of fishes, rays and jellyfishes were observed, including dwarf mussels, solitary and colonial ascidians on the surface of rocks, tube worms, barnacles and rock oysters. Four common species of seagrasses Halodule uninervis, Syringodium isoetifolium, Halophela ovalis and H. stipulacea flourish around these islands. These seagrasses are relatively associated with fine-grained sediment types (Aspinall et al., 2003). The seagrass beds accommodate some of the most rare and endangered animals such as sea dugongs and green turtles. The rocky habitats along the coastline are an important homeland for a variety of rare marine birds such as Sooty Falcons and Ospreys.
Huwar and adjacent areas have been recently declared as protected areas (Decree number 16, 1996, Article 1 of the State of Bahrain) issued by the council of Ministers. This area has also been nominated for inclusion on the Natural World Heritage List and listed as a RAMSAR site in February, 1997.
Data is generally lacking on the macrobenthic communities which could be used as a tool for bio-monitoring studies. The main objective of this survey was therefore to fill this gap with special reference to macrobenthos spatial diversity, physicochemical parameters and microbial water quality and organic content of sediment and grain size distribution.
Materials and methods
Sixteen sites were selected, at two nautical mile intervals, starting from the north moving along the east coast southward and continuing along the western coast northwards of the main island (Figure 1). Samples were collected over a two day period. Tidal range is small, around 0.3 m, based on the tide table for Huwar. The selection of stations was based on visual inspection of apparent change in habitat type for example, embayments or more exposed locations.
Water depth was estimated using Secchi disc and by walking into the shallower water. The temperature of water (°C), pH (pH units), oxygen concentration (mg l− 1) and salinity were measured using portable meters (Radiometer, pH 82, Oxygen meter Eil7130 and refractometer, AtagoF/ mill 8901).
Aliquots of dried sediment were used to calculate the percentages of different grain size fractions and characterized using the Wenthworth class size analysis method. Ash-free dry weight was determined according to MOOPAM (1999) prior to the determination of organic matter. Ashing was completed within 12 hours at 500°C and the percent organic matter was calculated from the differences in the weights of oven dried and ash free samples relative to the oven dried weights (Brower, 1998).
Chlorophyll-a concentration (μ g l− 1) was determined spectrophotometrically following the method outlined in Parsons et al. (1984). Water samples of 500 ml each were filtered whilst on board the research vessel. Each filtrate was then placed on a sterile Petri dish in a cool dark place during transportation back to the laboratory.
Samples were obtained using standard quadrats (0.25 m2), the area was scooped using a hand shovel to a depth of 10 centimetres, transferred into a polythene bag and stored on ice while on board. Peterson grabs (equivalent to 0.0675 m2) were also used wherever the tide and the current prohibited the use of the quadrat sampling technique. Nine replicates were obtained at each site. Sieving was carried out on site using 0.5 mm mesh size.
The data set from each site was analysed using the unweighted pair group method, average clustering and non-metric Multivariate Dimensional Scaling (MDS) ordination adopting Bray Curtis similarity index (ANOSIM) in order to detect any possible pattern in the animal assemblage among stations of different sites. The observed pattern is further analysed in relation to different environmental factors (BIOENV). All multivariate analysis was carried out using PRIMER software version 5.
Sixteen subsurface seawater samples were collected at 0.5–1 m depth for bacteriological analysis, from the 16 sites during two different cruises, around the coastline using shallow water sampler in sterile glass bottles. A composite water sample was freshly prepared by mixing 250 ml of water collected from five different stations at each site. Aliquots of 100 ml of seawater samples were diluted in sterile seawater to give a ratio of 1:10. Procedures for total and faecal coliforms, faecal streptococci and vibrios were those of the generally accepted methods of American Public Health Association (APHA, 1989). Samples were either directly filtered using pre-sterilized 0.45 μ m Millipore filter sets, or injected by means of sterile syringes into broth media.
The following pathogenic and potentially pathogenic bacteria were considered: Shigella and Salmonella (isolated on SS agar medium), Pathogenic Vibrios (isolated on TCBS agar), Faecal coliforms (isolated on m-FC agar), Enterococci (isolated on m-Entero agar), Total coliforms (isolated on lactose broth for MPN test). For confirmation of coliforms, m-Endo agar was used. Identification of all gram-negative catalaze positive bacteria was completed using API 20E identification kits.
Comparing to the western coast topographical feature, which is characteristically smooth, the eastern coast of Huwar has a varied and sheltered topography whereas the western coast is more exposed. The sea has a variety of distinguished habitats such as sea grass beds and both hard and soft bottoms, soft corals, extensive algal beds. Table 1 shows the coordinates for each selected site. A mixture of soft mud, sand and rocks was a characteristic feature of these areas. Occasionally anoxic muddy substrates were also found but areas were completely covered with sand as a result of current urban developments on these islands.
Chlorophyll-a was higher along the east coast (Mann-Whitney U test = 8, n1 = 8, n2 = 8, p = 0.012) whereas salinity was higher along the western coast (U = 11.5, n1 = 8, n2 = 8, p = 0.029). Chlorophyll- a concentration varied significantly, 0.58–6.3 μ g l− 1, with higher concentrations recorded at site number 7 (sheltered area in a bay near to animal reserve on the east coast) and site 14 (near municipality chalets on the west coast). The shallow water was clear (visibility reached to the bottom of the sea) except at some locations where seagrass reached a few meters.
Northern areas were characterised by medium (0.25 mm) to fine (0.125 mm) sand whereas southern and western sites had coarser sand granules (2 mm). Very fine sand was more common along the east coast. The organic content ranged from 5–15 % (Table 1). The sediment's organic content was slightly higher along the eastern coast (Mann-Whitney U test = 11, n1 = 7, n2 = 8, p = 0.05).
Species abundance (N) and total number of species (S) encountered at each site are listed in Table 2. Species area curve (not shown), using the cumulative number of samples, indicated that as the number of sampling units increases as more new species were encountered until the curve reached stability. Relative abundance, known as the species importance curve, was also investigated using both density and frequency of occurrence. Table 2 also shows various computed diversity indices. Table 3 shows the total number of major groups of macrobenthic invertebrates examined and the total number of species in each group. Polyplachophora were regarded as a rare species in this study as only two individuals were found. Occasionally, a single species dominates the whole of the sampling units indicating the lack of biodiversity. Polychaetes were abundant whereas crustaceans were more diverse. Syllus cornuta and Mitrella blanda were commonly found species. The lowest species diversity was recorded at site thirteen. Species diversity did not vary significantly between east-west coasts (Mann-Whitney U test = 18.5 n1 = 8, n2 = 7, p = 0.271).
Results of cluster analysis (ANOSIM) and MDS for macrobenthic faunal assemblages are shown in Figure 2 and Figure 3. Both of the above plots identically indicated four cluster groups. Site number 1 and 5 showed 20% similarity. This is also applied to sites 13 and 14. The remaining site groupings namely, (8 and 16); (6, 9 and 10) and (11 and 12) showed similarity at 40%. The similarity could be attributed to the sharing of 40% of species composition. The relationship between the biotic and abiotic variation was determined using Spearman Rank correlation (Table 4). The analysis revealed that the best correlation is between variables 2 (i.e. pH), 4 (i.e. salinity) and 5 (i.e. chlorophyll-a). The combined action of these factors was found to affect around 32% of the observed biological pattern. However, this result was statistically insignificant (R = 0.14) at p > 0.05.
Bacterial counts were recorded on selective media, followed by morphological characterization. Variations of the total count of indicator and pathogenic bacteria are shown in Table 5. Total coliforms showed high concentrations on the following sites: 1, 2, 12, and 13, with low or neglected occurrences from all other sites. Pathogenic Enterococci, were absent at sites 1–9 (eastern coast), while fluctuated occurrences were recorded at sites 10–16 (western coast). No significant difference between eastern and western sites was found for most of the pathogens and indicator bacteria. However, total Enteroccoci and Salmonella/Shigella were statistically higher along the west coast (Mann-Whitney U test = 14, n1 = 8, n2 = 8, p = 0.03) and (Mann-Whitney U test = 4, n1 = 8, n2 = 8, p = 0.001).
Standard ecological methods were used to obtain samples of macrobenthos and the quality of seawater was assessed in terms of both physicochemical and microbial parameters. The lowest spring tides in the Arabian Gulf occur during the night or early morning (Jones, 1986). Sampling macrobenthos would ideally coincide with or just before the daily low tides. Unfortunately difficulties arose from the fact that the intertidal zone around Huwar Islands at low tide is characteristically small. Therefore, sampling subtidally was a better option.
Temperature, pH and dissolved oxygen did not vary significantly between eastern and western coasts. Salinity is generally high around Bahrain islands (43–57‰), United Nation Environment Programme (1985). Tidal range is small (0.3 m), as such salinity was not expected to fluctuate significantly.
High salinity has been generally blamed for lower species diversity (Vousden, 1985; UNEP, 1985). The west coast of Huwar was dominated by fewer species (i.e. lower species richness) as reported by Vousden (1985). Higher salinity along the western coast of Bahrain has been correlated with lower chlorophyll-a concentrations (UNEP, 1985). Lower salinity along the eastern coast of Huwar has been attributed to the water current channelling effects of the topographical features.
Based on the multivariate analysis, we can differentiate four prominent clustering of macrobenthic assemblage. Site 15 clearly stands out. This was attributed to the proximity of this site to some development on the island of Huwar (Chalets) with a modified sandy beach. Lower salinity was also recorded at this site. Both sites number one and five were very similar and this was attributed to their resemblance in the substrate type being fine sand. Likewise, sites 13 and 14 aggregated together and this was possibly because both are exposed to waves and their substrate was of the coarser sand overlaying harder substrate. In addition, the biological similarity between other sites was also related to the similarity of the substrates being mainly fine sediment. However, the Spearman correlation revealed inconclusive interaction between biotic and environmental conditions. The type of substrate therefore, could be an important factor affecting the observed biological pattern.
Biological diversity was assessed using several diversity indices and a different multivariate analysis indicated a homogenous pattern in species diversity. It was therefore suggested that sampling efforts could be reduced without affecting the accuracy of parameter estimation, when a more rapid assessment of biomonitoring is required (Resh and Jackson, 1993). No P-values are generated in the cluster analysis because the test looks only for patterns in the data and clustering is about grouping the variables that are highly correlated (Townend, 2002).
Polychaetes, crustaceans and molluscans are an important part of the diet of commercial fishes. The knowledge of the existence and population density of the important benthic taxa would be essential in any ecological and environmental monitoring. Polyplacophora was noticeably rare; however, the lack of data on the existence of this macrofauna prohibited any comparison to be made. Obstacles may arise in the identification of species where similar species or sub-species possibly co-exist. In addition, juveniles could appear different from adults in the same species.
The most useful measures of species diversity (an expression of the community structure) include both species richness and evenness. The Margalef index (Margalef, 1958), Gleason index (Gleason, 1922) and Menhenick index (Menhinick, 1964) are similar. They consider species number (diversity) and the number of individuals in a species (abundance). In addition to the number of species and individuals, the Simpson index (Simpson, 1949) considers the proportion of the total that occurs in each species. Also, the number of equally abundant species could provide information on the level of the diversity in the community. Evenness, which is referred to as the distribution of individuals among different species, may also be expressed by considering how close a set of observed species abundances are to those from an assemblage of species having the maximum possible diversity. Such evenness measures are called relative diversity (Brower et al., 1998).
A relatively high species diversity was recorded for the crustacean group which would indicate that the environment at the selected sites was a well suited habitat for the presence and abundance of this particular faunal group. To some extent, crustaceans do tend to represent a hardier group and adapt faster to environmental stress. A dominant species could be regarded as a hardy species (or a key species) surviving harsh environmental conditions. Relative abundance/ species importance curves revealed some key species. Ranking of species abundance and frequency of occurrences revealed that Sylus cornuta and Mitrella blanda were the most common species. On the other hand, a lower number of individuals (abundance) for a species are attributed to a stressor at the community level.
The Shannon index indicated a range of values between < 0.5–1 (Shannon, 1948). Other studies carried out in the North Sea found a diversity index in the range of 0.5–6 (Clark, 2001). These indices are useful if, for example, assessing the impact of an oil spill. In recent years, the environmental stress witnessed by much of the coastal areas around Bahrain has been dramatic affecting biological diversity. Chronic pollution effect is evident along the coastal areas and unless revised, a larger number of species would certainly be at greater risk. The Diversity index was lower than those reported around the coastal areas of Bahrain (Vousden, 1985; Linden and Larsson, 2002). The latter authors have reported Shannon index values of around 0.3–4.5 and 0.66–4.3 respectively.
Organic content of sediment was slightly higher in samples obtained from the east coast. Although chemical procedures for estimating organic carbon are available, the method used was based on simple approximation of using ash free dry weight. Particle size of sediment has been related to organism distribution and abundance which would affect the water retention and burrowing ability of the benthic species (Nybakken, 2001). Higher wave actions at some western sites, such as site 14, affect the grain size distribution.
Chlorophyll-a concentration could be used as an indicator of the level of primary production and hence the density of the primary producers (phytoplankton). Higher concentrations of chlorophyll-a at site 7, which was located near animal reserve, and site 14, which was located near municipality chalet were attributed to the anthropogenic activities in proximity to these sites and to direction water currents. The calculated values were within the range (0.2–0.86 mg m−3) quoted in Sheppard et al. (1992) for the Arabian Gulf. Linden and Larsson, (2002) have reported values for chlorophyll-a of around 0.1–5.1 mg m− 3 and higher values (0.9–36.2 mg m− 3) were normally obtained in Mangrove areas (Al-Sayed et al., 2005). These values were also compared to values for Kuwait waters (0.2–9 mg m− 3) and the ocean (0.01–2 mg m− 3) (Al-Yamani et al., 2004).
Faecal contamination is a major concern in Bahrain aquatic environments (Al-Sayed et al., 2005; Ghanem et al., 1995). This contamination can originate from human and non-human sources. Escherichia coli are one of several faecal coliform bacteria that inhabit the intestines of many warm-blooded animals that sometimes contaminate water courses. It is therefore, necessary to differentiate contamination sources in order to accurately assess human health risks.
The variation of bacterial presence with special reference to coliforms was attributed to human activities and raw sewage dispersion. The distribution of pathogenic Enterococci was used as an index of water pollution since most of the members of this group grow normally on saline media. Statistically significant difference existed in microbial water content between the east-west coastal areas.
Coastal marine ecosystems in many parts of the world including Bahrain are under unrelenting stress caused by urban development; habitat destruction; toxic substances; pathogens; introduction of exotic species and natural toxins. Biomonitoring could play a vital role in governmental and industrial strategies to identify, control and reduce these problems and to monitor marine ecosystem health.
The environmental parameters studied during this survey provided a baseline for the effect of many influencing and rapidly changing environmental stresses in the Gulf. Tolerance and adaptability of the benthic organisms to extremely high salinity, high temperature regimes, distracted microhabitats and desiccation should be some of the future topics for research.
Recognising the importance of Huwar as a nature reserve both locally and internationally in sustaining a healthy environment in the Arabian Gulf, considerable time was devoted to the collection of information about macrobenthic animals, microbial diversity and water quality around the islands. Physicochemical parameters of the seawater were within the normal levels for the Gulf waters. Macrobenthic faunal diversity (moderate to low) around Huwar was typical of the Gulf's naturally stressed environment. Nevertheless it was healthier than many other parts around the coastline of the main Island of Bahrain. Cluster analysis indicated distinct groupings. The eastern coast appeared to be more biologically diverse. This was attributed to topographically varied eastern coastal habitats. The microbial indicator of seawater contamination was evident along the western coasts and attributed to human activity sources. Salinity was higher along the west coast. Chlorophyll-a concentration was variable but significantly higher along the east coast and/but generally low, indicative of low primary productivity. Organic matter was slightly higher along the east coast and the bottom substratum was dominated by fine sand. Biomonitoring using similar indicators should be carried out on a regular basis around the coastal waters of Huwar and other small islands within the Arabian Gulf.
This survey was supported by the Deanship of Scientific Research, University of Bahrain. Special thanks and appreciation go to Miss Tarneem Al Mousawi for her duly assistance with figures, typing and linguistics.