The nesting and feeding grounds of the green turtle, Chelonia mydas, in Oman are of regional importance (Baldwin and Al-Kiyumi 1999). The present study focuses on the diet composition of stranded green turtles in the area of Ra's Al Hadd, a Nature Reserve internationally important as feeding and nesting grounds for this species.

The green turtle is the most commonly encountered marine reptile in Oman (Ross and Barwani 1982). Over 275 beaches along the 1700-km stretch of coastline are used for nesting, including those on offshore islands from the northern region of Musandam to the Oman-Yemen border in the south (Salm 1991). There are large concentrations of turtle nesting at Ra's Al Hadd (ca. 6000 females per year; Ross and Barwani 1982). Ra's Al Hadd is a protected area for nesting sea turtles, covering a stretch of coast of approximately 70 km from Khawr Al Jaramah in the north to Ra's Al Ru'ays in the south, and an area of 120 km² including beaches, coastal lands, and offshore areas (Al-Kiyumi 1997).

The patterns of movements of green turtles between Oman nesting grounds and local or distant feeding grounds are unknown but appear to be rather complex. Extensive feeding grounds are in the Arabian Gulf, on the coasts of Oman and Yemen, and in the Red Sea (Ross and Barwani 1982). Many of Oman's green turtles are also known to feed in waters of other countries in the Arabian Gulf and Gulf of Aden (Salm and Salm 2001). Specifically in Oman, there are extensive and regionally important feeding grounds that occur along much of the coast, with particularly dense aggregations of feeding turtles to be found in the Barr al Hickman—Masirah Island region, along the Sahil al Jazir coastline and between Salalah and Hasik in Dhofar and scattered along the Batinah coast (Ross and Barwani 1982; Salm 1991). Although the feeding area along Arabian Sea coasts is extensive (around 800 km), development of dense beds of seagrasses is limited, especially during the summer season, due to the southwest monsoon known locally as khareef. The strong southwesterly winds drive currents parallel to Arabian Sea coasts from June to September with cold (21°C) and low salinity (35.5‰–35.7‰) water upwelled onto the continental shelf from depths of about 150 m. The high levels of nutrients in the sea combined with low temperatures account for the annual development of dense beds of macroalgae along Arabian Sea coasts. These conditions, however, are not well suited for the growth of seagrass leaves (Jupp et al. 1996).

The first dietary studies on Chelonia mydas in Oman were carried out by Ross (1985) who analyzed the stomach contents of 9 green turtles captured on the east side of the Masirah channel, and data showed that 1 turtle had eaten only algae, 3 others had eaten only seagrass, 2 had eaten seagrass and a trace of algae, while 3 others had eaten bulk amounts of both seagrass and algae. Quantitative data on food items were not obtained by species. The most important plant food items (forming > 40% by volume of stomach contents) included the seagrasses Halodule uninervis and Halophila ovalis, the green algae Chaetomorpha aerea and Ulva lactuca, and the brown macroalgae Sargassum illicifolium.

Green turtles in Oman are threatened by an assortment of human activity as well as by natural processes and events (Baldwin and Al-Kiyumi 1999). Of the human pressures, fishery-related threats are particularly severe, especially near feeding grounds or nesting beaches (Hare 1991). Most of the dead green turtles found in Ra's al Hadd have been incidentally caught in fishing gear. Other human activities, such as recreation and general coastal development, increasingly threaten turtle populations (Salm 1989). The first surveys of turtles in Oman were undertaken in 1977, focusing on Masirah Island and Ra's Al Hadd (the two Nature Reserves), starting with a turtle-tagging project. More than 23,000 turtles were tagged during the first 10 years of the project (1977–1986), including a few on the feeding grounds of Masirah Island. Subsequently, additional turtles have been tagged, mostly at Ra's Al Hadd, increasing the total number of tagged individuals to more than 50,000 (Salm and Salm 2001).

Current management strategies in Oman are based on the protection of nesting beaches, largely from recreation-related threats and primarily within these two Nature Reserves. Protection of turtles at other stages of their life history, including feeding and migrating turtles, is vital for their survival (Salm and Salm 2001). Therefore, the knowledge of the diet and feeding grounds of these green turtle populations is essential for planning management actions for animals and their selected food resources.

Methods

In July and August 2001, during the nesting season, 15 stranded green turtles (7 adult females, 7 adult males, 1 juvenile) were found on the beach in the Ra's Al Hadd area. Some of the females had shelled eggs, indicating active breeding. The probable cause of death was determined to be from artisanal fisheries activities (Ferreira 2002). The carapace length and width, the plastron length, the sex and site of recovery of all the turtles were recorded, and the digestive-tract contents were removed. From the 15 digestive tracts, only 11 contained food items (5 adult males, 5 adult females, 1 juvenile) that were used in this study. But it was not possible to quantify the diet contents of the juvenile sample due to the state of the food items, so only 10 samples were considered to quantify the green turtles diet composition.

The diet samples were collected and preserved in 4% buffered formalin/seawater solution in clear plastic vials stored in darkness to reduce color fading. After a week, the samples were transferred to 70% alcohol (Forbes 1999). Food items were identified to the lowest taxonomic level possible. The wet and dry weights of the items were obtained by drying them at 60°C until a constant mass was reached.

In the turtles with empty stomachs, the esophagus generally contained a bolus (also used for analyses), probably representing the last few bites taken before death. Some of the food found in the esophagus and stomach was in a remarkably fresh state, apparently unchanged by digestion, but in other cases the food was already digested.

Quantification of the diet composition allows the description of the food items found in the digestive tract contents, as well as their proportions. Two methods were applied to quantify the diet composition: the numeric method and the gravimetric method (Windell and Bowen 1978).

1. With the numeric method, the frequency of occurrence was determined. The frequency of occurrence (%FO) of a food item is given by the percentage of turtles that contain the same item in their digestive tracts: %FO = (ni / N) × 100 where ni is the number of tracts that contain the item i, and N is the total number of tracts with prey (Windell and Bowen 1978). Following Hyslop (1980), the turtles with an empty digestive tract were not considered for the calculation.

2. In the gravimetric method or weight contribution, the food items are separated according to their taxonomic categories and are weighed individually. The weight of each food item can be expressed as the percentage of the total weight of digestive tract contents found in all examined turtles (Hynes 1950): %W = (wi / Wt) × 100 where wi is the total dry weight of a particular prey group or taxon, and Wt is the total weight of contents in all the digestive tracts.

The Index of Relative Importance (IRI) was adapted from Hyslop (1980) calculating each diet category i by the equation:

where FO is the frequency of occurrence, W the weight, and n is the number of diet categories.

The importance of diet items with a high ash content and therefore high relative weight (e.g., calcareous algae, sponges, spicules, and exoskeletons) will be overestimated in a gravimetric analysis whereas diet items with low ash content will be underestimated. The IRI provides a more reliable measure for ranking the relative importance of diet categories because frequency of occurrence and mass are integrated.

Results and Discussion

The size of the adult turtles ranged between 98 and 111 cm CCL (curved carapace length) for the females and between 91.5 and 98 cm for the males; the juvenile had a CCL of 43 cm. Five adult males, 5 adult females, and the juvenile were considered for statistical analyses. It was possible to identify 42 different dietary items (Table 1); in decreasing order of importance (dry weight), these consisted of 19 taxa of algae (49%), 22 taxa of animal matter (26%), 2 seagrasses (15%), and substrate particles (3%; sand, pebbles, and shell fragments). Recovered items of litter (7%) included plastic bags, nylon chord, mesh bags, and tarp fragments. The digestive tract contents ranged from 3 to 16 g of dry weight. Weights of the taxa, litter, and substrate categories are presented, with FO and IRI data, in Table 2.

Table 1. Taxonomic groups found in the digestive tracts of green turtles, Chelonia mydas, at Ra's Al Hadd, Oman.

View Fullscreen

Table 2. Quantitative description of the diet of green turtles (Chelonia mydas) at Ra's Al Hadd, Oman, and Index of Relative Importance (IRI).^a

View Fullscreen

The digestive tract contents showed a strong tendency toward herbivory with indications of the turtles having foraged primarily on marine algae, although it was observed that they had also eaten seagrasses.

The brown algae Nizamudinnia zanardinii was the most commonly ingested food item, with 70% frequency of occurrence and an IRI of 24. The prevalence of N. zanardinii among the diet samples reflects its great abundance along Arabian Sea coasts. Every year thousands of tons of this seaweed appear along Arabian Sea coasts of Oman in the cold, nutrient-rich upwelled waters of the annual summer southwest monsoon. The Oman Seaweed Project (OSP 1999) carried out a comprehensive survey of seaweeds in Oman including data on taxonomy, the distribution, cover, and biomass of seaweeds found in various sites; maximum biomass data for N. zanardinii ranged from 0.60 to 2.33 kg DW/m².

The green alga Cladophoropsis javanica was another common species in the diet samples. This species appears to be rare, and there are only two records of its existence: at the southern tip of Masirah Island (Wynne and Jupp 1998) and at Sur (B.P. Jupp, unpubl. data, 1999). However, dense beds of this species could have been overlooked in surveys to date.

Other commonly found species or that were a significant component by weight in the samples included Ulva rigida, Gracilaria canaliculata, Hormophysa triquetra, Stoechospermun marginatum, Ahnfeltia plicata, Codium dwarkense, and Grateloupia sp. These were also found in surveys in the Oman Seaweed Project (OSP 1999).

The red algae were only sparsely represented in these diet samples, in contrast to other studies where green turtles were found to forage primarily on Rhodophyta classes in Brazil and Mexico (Ferreira 1968; Seminoff et al. 1998).

The major components of the seagrass meadows in Oman are Halophila and Halodule species, which are hardy pioneer species well suited to the conditions along Arabian Sea coasts (Jupp et al. 1996). The seagrasses species Halophila ovalis and Halodule uninervis were abundant in the diet samples, both being preferred by the green turtles, similar to that reported in other green turtle diet studies (Oman: Ross 1985; India and Yemen: Bjorndal 1997). Halodule uninervis was found more commonly, occurring in 50% of the tracts but Halophila ovalis formed a higher weight percentage and had an IRI of 13. These species are also found along the north coast of Oman, limited to sandy/salty areas such as Ra's Suwadi, Bandar Jissah, and Sur Harbour (Jupp et al. 1996). Further detailed studies on seagrasses in Oman, including biomass data, are given in Jupp et al. (1996).

Preference feeding tests with C. mydas in Oman (Al-Ajzoon 1993) showed a preference for macroalgae: Nizamuddinia > Sargassum > Ulva, and a preference for seagrasses: Halophila > Syringodium > Halodule. It is of interest that the same two preferred taxa, Nizamuddinia and Halophila, from these tests form the highest percentage by weight of plant components (13 and 12%, respectively, Table 2) in the digestive tract contents. This implies some selectivity on the part of the turtles for these macrophyte species. Ross (1985) attributed the differential abundance of algal species in the habitat and in the diet of green turtles in Oman to selective feeding. It is of interest that Ross (1985) reported a different suite of algae (Chaetomorpha, Ulva, and Sargassum) in the stomach contents he examined in the Masirah Channel, which may reflect different availabilities of algae in feeding grounds visited by the turtles. However, the number of samples from both studies is too small to conclude selective feeding. The extent to which the diet of green turtles is determined by selective feeding or by the relative abundance of different diet species has been addressed in several studies (Bjorndal 1997).

Analyzing males and females separately it was observed that the brown algae Nizamudinnia zanardinii was the most commonly ingested species by males, and the green algae Cladophoropsis javanica the most commonly ingested by females.

Data on any selective feeding and the distribution and biomass of both macroalgae and seagrasses are important in assessing the status of feeding grounds for C. mydas. To better facilitate nutrient assimilation, green turtles appear to take small bites of algae during consumption. In the examined diet samples, algae were commonly fragmented to small size. The resulting small food particle size is functionally beneficial to nutrient assimilation providing opportunity for efficient digestion (Bjorndal 1997).

Green turtles in the Arabian Sea supplement their diet with animal matter. The possibility that such food resources make a major contribution of vitamins, trace minerals, or essential amino acids for herbivorous green turtles has been discussed (Bjorndal 1985) and may explain why, even with abundant algae and seagrass resources, green turtles feed on animal matter. Small invertebrates, including polychaete worms and tubes, small gastropod and bivalve shells, hydrozoans, and bryozoans, were found in almost all samples. The Gastropods and Cephalopods were the most abundant, with both occurring in 70% of the tracts, however, both show differences for IRI index, 41 and 4, respectively. This apparent higher use of nonalgal resources may reflect high prevalence of invertebrates in the shoreline habitats near Ra's Al Hadd. It is important to note that the animal matter species could also be associated with the algae and seagrasses. Little is known on the distribution of these fauna in feeding grounds.

The relative importance of animal matter, especially the gastropods, is believed to be exaggerated due to the use of dry weight as an index of measurement. Rigid invertebrate parts have overestimated weights compared with algal and seagrass specimens and soft-bodied invertebrates, like cephalopods, since they are completely digested and leave no identifiable remains behind. With the exception of the gastropods, the species richness and abundance of nonalgal species recovered in this study are likely to be underestimated.

The frequent occurrence of substrate particles (sand, pebbles, and shell fragments) in diet samples from all sites suggests that feeding turtles may be ingesting this material incidentally as they closely crop seagrasses and algae. Substrate may also be ingested as turtles feed on Sabellid worms and other benthic organisms. Green turtles also consumed plastic bags, nylon cord, mesh bags, and tarpaulin fragments. Again, the relative importance of these materials is exaggerated by the use of dry weight as an index of measurement, as with the animal matter. There is no evidence to suggest this as a cause of death, especially because they represent a small gravimetric percentage of the total diet. However, the amount of debris and other plastic materials ingested is usually small, and so it is important to put these apparently small percentages in perspective, even though mortality resulting from ingestion of debris is extremely difficult to estimate (National Research Council 1990).

The results from this study must be interpreted with care. Small sample sizes, possible misidentifications of prey species, the fact that data were obtained during the reproductive season, and analysis of dead turtles may contribute to erroneous conclusions.

Much research is needed to elucidate the relationships between the foraging ecology of sea turtles and their role in the marine ecosystems and between the nutrition of sea turtles and their productivity. Several gaps remain in our knowledge of feeding habits of green turtles in Oman, but understanding diet selection is critical for assessing the significance of feeding ground habitats, an undertaking that is ever more critical as wildlife managers are expected to make difficult decisions about which habitats to protect. Further quantitative studies are needed that address relative abundance of potential food items in the environment to the food items that are ingested.