Seagrass bed, Seychelles. Seagrasses required bright sunlight and are generally restricted to < 12 m depths, though in exceptional circumstances can be found down to 30 m or more. Thalassodendron ciliatum is one of the dominant species, forming extensive beds on reef and sandy substrates.
© David Obura

Common seagrasses in the WIO are Thalassia hemprichii in the upper photo, and Cymodocea (thick) and Halodule (thin) in the lower photo.
© David Obura (top), © Julien Wickel (below)

Green sea turtles primarily feed on the cylindrical seagrass Syringodium isoetofolium, which is among the rare seagrasses.
© David Obura

Seagrasses

Seagrasses are marine angiosperms widely distributed in both tropical and temperate coastal waters creating one of the most productive aquatic ecosystems on earth. The distribution of seagrasses ranges from high intertidal to subtidal soft and hard bottoms, including sandy bays, mud flats, lagoons, estuaries, coral reef patches and sheltered and exposed reef platforms. They often form extensive mono- and multi-specific meadows in depths < 12 m, and in the WIO tend to be in close association with coral reefs and mangroves. With sufficient water clarity they grow up to a depth of 70 m.
Extensive seagrass beds are found in all countries of the WIO, where they have received limited scientific attention compared to mangroves and coral reefs. Of the 24 seagrass species in the tropical Indo-Pacific, the WIO holds 12 species. The greatest diversity of seagrass species is along the Mozambique, Tanzania and Kenyan coastlines, and decreases eastwards into the islands. The Saya de Malha bank supports the largest contiguous seagrass beds in the world, which thrive on the relatively flat bank in depths shallower than 20 m.
Mixed seagrass beds with a high diversity are common, up to 8 or 10 species at the same locality have been reported for Mozambique, Tanzania and Madagascar. Two of the most common species are Thalassia hemprichii and Thalassodendron ciliatum both forming extensive beds in most parts of the region. T. hemprichii is found in more protected habitats or on intertidal flats, whereas T. ciliatum normally inhabits exposed or semi-exposed habitats, and can anchor on both sandy and rocky substrates. Also common in the region are Halophila ovalis, Cymodocea rotundata, Cymodocea serrulata, Syringodium isoetifolium and Halodule uninervis. Enhalus acoroides, Halophila stipulacea and H. minor are mainly reported from northern Mozambique to Tanzania and in some locations in Kenya. Zostera capensis is a more temperature species, and is only common in southern Mozambique and South Africa where large monospecific stands may occur. But the species has also been recorded in Kenya, Madagascar and Mayotte. Halophila beccarii is only known in Madagascar, Halophila decipiens is a new but relatively common species and widespread.

Significance - Seagrass beds are among the most productive aquatic ecosystems in the biosphere. They are important as nursery grounds, foraging areas for sea turtles, fish and dugong, and predation refuges for numerous fish and invertebrate populations. Seagrass beds provide great benefits for commercial, subsistence and recreational fisheries. Due to the complex architecture of the leaf canopy in combination with the dense network of roots and rhizomes, seagrass beds stabilize bottom sediments and serve as effective hydrodynamic barriers reducing wave energy and current velocity, thereby reducing turbidity and coastal erosion. Further, seagrass beds trap large amounts of nutrients and organic matter in the bottom sediment. Through microbial decomposition, seagrass biomass may enter the marine food web as detritus and thus support productivity through recycling of nutrients and carbon. Due to their high productivity, they are often a food source for animals resident in adjacent ecosystems such as coral reefs, and may increase the biodiversity in these systems.
Due to their high productivity and trapping of carbon in biomass and sediment trapping, seagrass beds are among the most significant shallow marine carbon sinks, storing up to 500 tonnes/ha (or 50 g/m²), of which nearly all of this is trapped in the sediment. This is equivalent to the amount of carbon stored in primary tropical forests. Along with mangroves, seagrasses are therefore of significance in carbon sequestration to reduce greenhouse gas buildup in the atmosphere and oceans.

Threats - Pressure on seagrass beds in the region is increasing due to growing coastal populations and human disturbance from e.g. pollution, eutrophication, sedimentation, fishing activities and collection of invertebrates, though there is little quantitative evidence on specific threats. Reduced water clarity from land-based impacts reduces the depth at which seagrasses can grow.

Management, mitigation and restoration - Seagrass ecosystems in the WIO are valuable resources for fisheries at both local and regional scales, with much of the artisanal fishing in coral reef areas being focused in seagrass habitats on the reefs. Still, seagrass research in the WIO lags behind other ecosystems and other regions, and is mainly focusing on botanic diversity and ecology. Seagrasses have not been the focus of management in the region, though most MPAs focused on coral reefs include large areas of seagrass beds.
From 2000-2005, die-back of seagrass beds (Thalassondendron ciliatum) as a result of sea urchin (Eucidarus thouarsii) predation in Kenya caused some concern, but natural recovery apparently occurred. In Kenya sea urchin removal was found to assist seagrass recovery. In Mayotte, mortality of T. ciliatum on the inner barrier reef flat has been observed. More focused research, site selection and zoning of MPAs with seagrass beds as the prime focus are needed, as well as studies on land-based impacts. Restoration of seagrass beds is common practise in some parts of the world, but has not been trialled in the WIO.

Key References - Bjork et al. (2008); Duarte et al. (in press; Frouin & Bourmaud (2008); Gullström et al. (2002); Hily et al. (2010); Milchakova et al. (2005); Murray et al. (2011); Richmond (2011); Short et al. (2007); Uku et al. (2007). --> References