Biodiversity assessments within seagrass beds were conducted in six sites in Lamu, Kenya; namely, Kiweni, Tauzi, Wange, Ntopate, Manda toto and Ngoi. The objective of the assessment was to collect baseline information of the seagrasses of Lamu County in Kenya. Our findings revealed that nine out of the twelve seagrass species found in Kenya were found in the study sites. The dominant seagrass species Thalassodendron ciliatum was found to occur in deeper subtidal areas, while the pioneering species occurred in intertidal shallower areas. Average shoot densities per site ranged from 2.4 ± 1.7 shoots m-2 of Cymodocea serrulata to 1025.9 ± 139.3 shoots m-2 of Syringodium isoetifolium in Ngoi. Canopy heights ranged from 0.1 ± 0.1 cm of Halodule ovalis in Tauzi to 16.3±1.3.5 cm of Syringodium isoetifolium in Ngoi. Although the seagrass species characteristics were much lower than that found in similar mangrove fringed seagrass beds of Gazi Bay, the majority of the fish found in these seagrass areas were invertivores which indicates that these seagrass sites form a refugia for adult fish populations with nursery grounds being located elsewhere.
This study forms the first comprehensive assessment of the seagrasses of Lamu and it provides important baseline information on seagrass beds. Such biodiversity information provides important support for decision-making for coastal areas that are targeted for infrastructural development such as Lamu. Biodiversity information of such critical habitats form a critical data base for marine spatial planning and can be used to effectively guide the integration of biodiversity and coastal livelihoods in the sustainable development of Kenya’s coastal areas.
Seagrasses are marine flowering plants that form a part of coastal habitats worldwide. They are critical components of coastal and marine environments, providing some of the most economically important ecosystem services of any marine habitat (Costanza et al., 1997; Orth et al., 2006). They often occur in vast meadows which provide nurseries, shelter, and food for a variety of commercially, recreationally and ecologically important marine species. In the tropical and subtropical Indo-Pacific, seagrass meadows are characterized by high species diversity with mixed vegetation (Hemminga and Duarte, 2000). Twelve genera and 60 species have been reported worldwide (Short et al., 2007) with twelve species being found in Kenya. Among the twelve species, Thalassondendron ciliatum is the most dominant one (Ochieng and Erftemeijer, 2003). Seagrass meadows support numerous charismatic faunal species, including turtles, dugongs and seahorses (Hughes et. al., 2009) as well as associated faunal communities that vary with seagrass species (De Troch et al., 2001).
The ecological roles of seagrasses are vast owing to their importance as nutrient, contaminant and sediment filters of estuarine and coastal waters, and their close linkage to other communities such as coral reefs and mangroves (Nybakken, 2001). The relatively high rate of primary production in seagrasses drives detritus-based food chains, which helps to support many organisms (Adam and King, 1995). Seagrasses are also critical in the shoreline protection from erosion (Pantusa et al., 2018). Naturally, seagrasses show clear zonation patterns with water depth, sediment structure and exposure to air and sunlight during low tide. Species that are tolerant to exposure are found higher up on the intertidal, while those that cannot withstand exposure occur submerged in pools of water.
Seagrass decline has been reported from various places around the world (Orth et al., 2006). In Kenya, the major threat has resulted from over-explosion of sea urchins leading to loss of seagrass cover (Eklöf et al., 2008). However, recovery has been evident in several places. Despite their importance, seagrass beds have largely been left out of management regimes compared to other ecologically sensitive habitats such as mangroves and coral reefs. To ensure inclusion of seagrasses in national monitoring programs, they have been included nationally as part of the National Coral Reef and Seagrass Conservation Strategy (2013). The knowledge of seagrass biodiversity in the Lamu is insufficient yet these ecosystems support most of the country’s marine species and this area is a hotspot for biodiversity. To address such impacts and many more, there was need to come up with effective management tools for these ecosystems that ensure that the biodiversity and productivity is well maintained in order to support the ever-increasing demand of local communities. However, prior to the recommendation of appropriate management strategies, there was need for baseline information on the distribution and status of such critical coastal and marine ecosystems.
Monitoring of these critical habitats as part of the implementation of this important national strategy was the main basis for the assessment of the status of seagrass beds of Lamu County in 2014 under the Kenya Coastal Development Project. Lamu is emerging as a hotspot for conservation due to the high level of ongoing development activities under the LAPSSET project which include the development of a port and coal plant as well as plans for a resort city. This was the first ever detailed assessment of the seagrasses in the Lamu and total of six sites were surveyed to gain a broad understanding of the seagrass assemblages in the Lamu Archipelago and to document the baseline on which to understand future changes that may occur in this critical ecosystem.
Materials and methods
Six sites were surveyed; namely Kiweni, Tauzi, Wange, Ntopate, Manda toto and Ngoi. The field survey focused on providing detailed information (distribution and abundance) on high priority intertidal and shallow subtidal seagrass ecosystems in the sites. Seagrass cover was assessed visually as percentage cover based on transects. At each site, two transects were laid perpendicular to the shore within which parameters were recorded in 3 random quadrats after every 20 meters. The distance between transects was approximately 200 m apart to cover intertidal and sub-tidal seagrass habitats. Sampling sites were randomly selected within each site, for assessment but deliberate attempts made to ensure all possible seagrass habitats were assessed.
Seagrass cover, shoot density and canopy height were determined using standard quadrats of 25 cm x 25 cm for each parameter as described in the SeagrassNet protocol (Short and Coles, 2001) in order to capture the zonation patterns of seagrass. For the seagrass cover, the cover by each seagrass species within the quadrat was estimated on a percent cover scale of 0 to 100%. The shoot density was estimated by counting all the shoots of individual species within the quadrat. In accordance to Short and Coles (2001), the canopy height was estimated by taking the canopy average height of the dominant species in the quadrat and omitting the tallest 20% of the leaves. Seagrass identification was made to the lowest taxonomic level according to Waycott et al. (2004).
Within each quadrat, algal percentage cover was also estimated and species observed noted and identified using identification keys from Oliveira et al. (2003). Underwater Visual Census (UVC) through snorkeling was used to estimate the densities and composition of fish as described by English et al. (1997). The fish communities were surveyed using transects of 50 m length and 5 m width (250 m2) marked along lines at each site parallel to the shore. At each site, visual censuses were conducted, by swimming, along at least 3 transects on the shallow lagoons. The distance between the transects at each site was at least 10 to 20 m.
Seagrass habitats in the study sites
Within Lamu seagrass meadows, seagrass cover ranged from 38 – 68% in Tauzi and Manda toto respectively (Fig. 1), with all the locations exhibiting mixed seagrass species with significant differences observed between locations (Kruskal-Wallis, H=47.95 p<0.001) (Fig. 2). The primary seagrass species observed in Kiweni, Wange and Ntopate was Thalassia hemprichii, although Tauzi and Manda toto exhibited a high cover of Cymodocea rotundata while Ngoi showed a high cover of Syringodium isoetifolium (Fig. 2). Additionally, zonation typically exhibited by seagrass species was observed with opportunistic mixed species occurring in the shallow intertidal areas while climax monospecific species occurred in the deeper areas. In terms of algal cover, Kiweni, Ntopate and Ngoi showed a high cover at 11%, 8% and 9%, respectively and significant variation between sites was observed (Kruskal-Wallis, H = 74.41 p < 0.001).
Table 1 shows the mean shoot density of seagrass in the various sampling locations. Mean seagrass shoot density was higher in Wange and Ngoi for Halodule uninervis and Syringodium isoetifolium, respectively with significant differences being observed between locations for all species apart from Halophila ovalis (Kruskal-Wallis p < 0.001). The site with the highest canopy height was Ngoi with Syringodium isoetifolium having the highest shoots (Table 2). As was the case with shoot densities, canopy height also showed significant site differences apart from Enhalus acoroides and Halophila ovalis (Kruskal-Wallis p < 0.05).
Patterns of fish assemblages in the study sites
A total of 48 species belonging to 19 families were recorded in the seagrass beds that were surveyed (see Appendix 1 in the supplementary file to the online version of this article). Overall, the mean fish density and diversity was highest in Thalassia hemprichii dominated meadows of Kiweni and Ntopate followed by Tauzi and Manda toto that were dominated by Cymodocea rotundata (Fig. 3). In terms of species richness, the highest number of taxa (14 species) was observed at Ntopate followed by Kiweni and Ngoi (12 species) while at Tauzi, Manda toto and Wange reported 8, 7, and 6 species, respectively. Similar to fish density, T. hemprichii meadows exhibited higher species richness compared to the other sites. However, in terms of biodiversity, Syringodium isoetifolium dominated meadows of Ngoi showed higher species richness. Significant differences in terms of fish densities were observed between locations (Table 3). Similarly, species richness also exhibited significant differences between the locations surveyed (Kruskal-Wallis H = 15.385, 5df, P < 0.01).
The classification of fish into trophic groups showed that, in terms of percentage of the number of species, most were invertivores followed by herbivores, while there were very few detritivores, planktivores and omnivores (Fig. 4).
This study attempts to document seagrass meadows in the Lamu Archipelago, a world heritage site. A total of 9 out of the possible 12 seagrass species were observed in the shallow intertidal areas. The substrate was mainly covered with multispecies seagrass communities in the shallow intertidal zones while those found in deeper areas consisted mainly of monospecific communities. Continuous seagrass communities were mainly observed although fragmented communities were also common. A comparison of the seagrass densities and cover revealed that that the seagrasses of Lamu were below the higher values that have been documented for Gazi Bay, which has a similar mangrove fringed seagrass community (Githaiga et al., 2017).
Variability of shoot density and canopy height observed in this study was indicative of changes in light availability since perpendicular transects were used starting from intertidal towards subtidal zones. However, we cannot rule out other environmental factors such as nutrient availability that were not determined in the scope of this study. Man-made impacts such as beach seining for fish could have also influenced the seagrasses parameters (Karama et al., 2017) as fishing occurs in these sites.
This study showed variation of fish assemblages spatially with the meadows dominated by Thalassia hemprichii displaying a higher fish density and diversity. These meadows also exhibited comparatively higher algal cover than the other meadows. These results are similar to studies that have compared mixed meadows in relation to associated fauna (Ochieng and Erftemeijer, 2003).
Our study highlights the importance of seagrass habitats for large predatory fish as well as herbivores. The fish structure was mostly dominated by large predators (invertivores) and key herbivores. The large proportion of invertivores also points to the fact that these seagrass beds support an adult population of fish. Adult fishes, have been documented to feed on the infauna of seagrass beds while the diet of juveniles stages is mainly based on seagrass-derived detritus (Ochieng and Erftemeijer, 2003). The absence of detrivores and planktivores, indicates that the juvenile population may be very small in these seagrass beds. This is in contrast to the study by Igulu et al. (2014) that shows the high prevalence of juveniles in Indo-Pacific seagrass beds. As the seagrass beds of Lamu are surrounded by mangrove forests, it is postulated that the mangroves may serve as habitats for the juvenile fish and play a key role as nursery grounds in this region. Additional work is required to characterize the size classes of the fish to confirm our findings. Our results also demonstrate the need to conduct further work in the region to determine the inter-connectedness of the mangroves and seagrass beds. The fact that our study was conducted at the same time of the day, may have hidden fishery patterns that are determined by lunar cycles, seasonal variability as well as daily rhythms of nutrient and water exchange. This presents a gap for future research and offers an opportunity for in depth understanding of the use of the key ecosystems in this region that also include coral reefs.
According to the IUCN Red list status, most of the fish species observed during the survey were classified as species of least concern (LC) or having not yet been evaluated (IUCN Red list, 2014). However, the Chaetodon auriga (Threadfin butterflyfish) belongs to a list of species whose population is decreasing and no species-specific conservation measures have been put into place. This species of fish is mostly harvested by artisanal and aquarium fishes along the Kenyan coast (Mangi and Roberts, 2006). This fish is typically a coral fish and and its occurrence in these seagrass beds provides insights into the habitat connectivity and it means that efforts to preserve seagrass areas would also preserve this population.
The structure of the fish communities also provides insights into human impacts on the environment as reflected in the distinct differences in fish communities between the study sites. At Kiweni, for example, community conservation efforts, in general, assist in maintaining a high species richness and fish abundance. On the other hand, at Ntopate, in as much as there is no community conservation effort, there was also high fish species diversities and abundance. However, fish densities and species diversities were low in the sheltered side at Ngoi and Wange. Ngoi near Kipungani was observed to be a site that once used to be rich in corals and it now has a mixture of both seagrass and scattered small coral head patches. The degradation of the site could have been as a result of human disturbances on land (sand harvesting and mangrove cutting) which appear to have an impact on the marine habitat hence the low fish densities and diversities.
The importance of seagrass ecosystems in supporting small scale fisheries is crucial as highlighted by de la Torre Castro et al., (2014) where they noted that seagrasses are often overlooked as prominence is given to mangrove and coral reef habitats. The study by de la Torre Castro (2014) further demonstrated that the value of seagrass fisheries was comparable to that of mangroves and corals and that catches from seagrass areas were more stable (de la Torre Castro et al., 2014). Although our study was limited to the health of the seagrass beds and presence of fish in the various study sites, our findings provide an opportunity for an in depth interrogation of how the fish species of the seagrass beds contribute to the small scale artisanal fisheries of Lamu.
These seagrass beds are yet to reach the level of destruction of seagrass meadows that has been documented in places like Diani, in Kenya (Eklöf et al., 2008) where the increase in the sea urchin, Tripneustes gratilla has had devastating effects diminishing seagrass cover, distribution and health in several areas along the Kenyan coast.
Overall, our results suggest that the diverse habitats of Lamu are in good condition and have a healthy and resilient seagrass ecosystem that supports a predominantly adult fish community. However, the proposed developments such as the Lamu Port-Southern Sudan-Ethiopia Transport project (LAPSSET) project may have detrimental impacts in these meadows hence the need to ensure that the environmental assessments for these projects provide for ecosystem protection from potential threats such as overfishing, sedimentation and eutrophication from the proposed developments (Björk et al., 2008). Sedimentation and eutrophication have been cited as the major cause of seagrass loss globally and it is recommended that monitoring should be conducted annually so that the full impact of port development (added nutrients from human settlement, and sedimentation) on the critical fishing grounds in seagrass beds is ascertained. The clearance of expansive mangrove areas to support infrastructure, development and increased sedimentation impacts on corals will have implications on the small scale fisheries that are dependent on the linkages between these ecosystems. In the context of the wider landscape, there is also need to understand the role of coastal forests, which extend up to the Tana system, on the health of the ecosystems that form the Lamu Seascape. The decimation of these forests for human settlement and farming present new pressures to the marine areas that are downstream (Hamerlynck et al., 2010). Okafor-Yarwood et al., (2020) have provided a reflection on infrastructural development for the Blue Economy and in their view sustainable development takes place when socio-ecological and environmental considerations are put in the center of the economic dialogue. They use the full spectrum sustainability evaluation tool (FSS) in their evaluation which embraces ecological parameters based on the biodiversity and species assemblages as one of the measures of habitat and ecosystem integrity (Okafor-Yarwood et al., 2020). The information generated in this study provides a first step towards providing some aspects of the health of the seagrass ecosystems in Lamu. Such data is critical as it provides information for future sustainability assessments.
Although this was a one-time study, the information is important in the development Marine Spatial Plans that integrate aspects of landscape protection that are inclusive of mangroves, seagrass and coral reef habitats to foster resilience in the face of continued development in this region.
The work would not have been possible without logistical support from Kenya Wildlife Service (KWS), KMFRI, State Department of Fisheries (SDF) and World Wide Fund for Nature (WWF). We thank the team leader of Component 2 of KCDP, Mr. Stephen Mwangi and team members their support in several aspects of data collection. Local fishermen at the study sites and the managers of community conserved areas are also acknowledged for their support in work within the different study sites.
This work was made possible by the GEF grant to the Kenyan Government and the Kenya Marine and Fisheries Research Institute (KMFRI) through the Kenya Coastal Development Project (KCDP). Support was provided through KMFRI to attend the GLOW 9 conference in Kisumu where this paper was presented.
Supplemental material for this article can be accessed on the publisher’s website.