The loss of coastal land and associated infrastructure by marine erosion is a recurrent and widely reported problem in the Western Indian Ocean region (Arthurton, 1992; IOC, 1994; Mwanje, 1997; Kairu and Nyandwi, 2000; UNEP-GPA, 2004). In particular, coastal erosion and consequent shoreline recession poses a continuing threat to investment of tourism-related infrastructure, necessitating expensive engineered coast protection measures and even the abandonment of hotel developments (Nyandwi, 2001).
Studies of erosion “hotspots” have identified or suggested a range of causal factors. These include marine influences related to climate change, with increased wave energy impacting the coast because of increased storminess and sea-level rise, and also tsunamis (UNEP, 2005). Human activities have also been considered influential. These include damming and land-use change in the hinterland (Kitheka et al., 2004; Snoussi et al., 2007) and, notably in Tanzania, sand-mining in streams discharging to the shore causing declines in the natural replenishment of beach sand (Griffiths, 1987; Shaghude et al., 1994; Masalu, 2002). In Kenya, accelerated shoreline recession has been linked to inappropriate adjoining coastal protection structures (Kairu, 1997).
This paper reports on studies that have been carried out and funded under the Marine Science for Management (MASMA) of the Western Indian Ocean Marine Science Association (WIOMSA) – which developed a regionally applicable Manual for Assessment and Design of Mitigation Strategies on Shoreline Change in Tanzania and Kenya. The main goal of the study was to investigate the underlying causes of shoreline changes, their socio-economic impacts and to recommend effective and relevant mitigation and adaptation options. It is based on the analysis of two Indian Ocean mainland coastal sites, one in Kenya, the other in Tanzania. The sites were selected as being representative of the region in terms of climate, shoreline processes and developmental pressures, though not in terms of the types and scales of their catchments.
Materials and methods
The study sites
The Bamburi site extends over 14 km between Shanzu and Nyali. It is characterized by cliffs of Pleistocene limestone and a lagoon platform up to 1,600 m wide with a fringing reef bar flanking the deep ocean. There are extensive- to pocket beaches with carbonate as well as siliciclastic sand, sandy beach plains occupying embayments in the cliffs and local intertidal beach rock. The site is bounded by the deep-water tidal creeks of Mtwapa to the north and Tudor to the south. Only one stream, the seasonal Mtopanga River, discharges to the sea within the study site, its outflow usually blocked by a littoral bar. The entire coastal strip has been developed with private residences and beach hotels. The beaches and lagoons are recreational and fisheries resources, the whole shore lying within a marine reserve administered by the Kenya Wildlife Service.
The Kunduchi site extends for some 18 km between Ras Kiromoni and Msasani Bay. It lies within the Dar es Salaam seascape (Wagner, 2007). There is an extensive beach of predominantly siliciclastic sand and some intertidal outcrops of beach rock. The shore is flanked by a sandy beach plain including a spit bounding Manyema Creek and, around Ras Kiromoni, by cliffs of Pleistocene limestone. Unlike the Bamburi site, there is no fringing reef. Much of the shore has a fringing sandy, inter- to sub-tidal platform, with extensive seagrass meadows and, off Manyema Creek, a tidal delta with migratory sandbars. The platform is flanked by shelf waters, proximal to the western margin of the Zanzibar Channel. These are interrupted by shoals and patch reefs around islands of Pleistocene limestone and designated as marine reserves, administered by Tanzania’s Marine Parks and Reserves Unit. The hinterland consists of Mio-Pliocene clay-bound sands and gravels (Griffiths and Lwiza 1987) drained to the shore by five seasonal rivers. Manyema Creek, partially reclaimed for salt production, has no significant hinterland stream feed. Like Bamburi, the site has been developed for private residences and beach hotels and is an important recreational and fisheries resource.
Both sites are subject to an alternating climate regime of SE monsoon (April-November) and NE monsoon (December-March), with the “long rains” (Masika) usually between March and June and the “short rains” (Vuli) between October and December.
The rate and pattern of shoreline changes and the most important natural and anthropogenic causes of shoreline changes was established by systematic monitoring of 20 beach profiles for over 2 years with reference to existing hydrodynamic, meteorological and geological datasets in combination with other historical datasets (such as aerial photos/ satellite images, meteorological, etc.).
The relationship between shoreline changes and coastal communities’ livelihood patterns including tourism; these include the impacts of human activities on shoreline change and vice versa, was determined through stakeholder interviews and focused group discussions.
An examination of the existing mitigation measures in relation to their effectiveness (function and cost) and efficiency was made, an inventory prepared, and appropriate options recommended.
At both sites, direct observations, transect walks and semi-structured interviews (Bunce et al., 2000) using interview guides/semi-structured questionnaires with open-ended questions were used, as well as community perceptions and attitudes towards sand supply to the shoreline. Respondents were randomly selected and included the fishers, farmers, sand miners and traders.
Beach profile monitoring on the Bamburi shore
Monitoring of the 20 beach profiles surveyed on the Bamburi shoreline revealed marked changes in the distribution of beach sand during the monsoon cycle (Shaghude et al., 2013). The profiles were observed to change according to the prevailing wind-wave forcing of the respective monsoon seasons. During the SE monsoon (April to October), at the southern end of the site between Beach Profile 00 (BP 00) and BP 01 (around Nyali beach), and at the northern end between BP 17 and BP 19 (at Serena beach), there was major northward longshore sand transport, leaving bare rocky shores. During the NE monsoon (November to March) this transport was reversed, and sand was re-accumulated. The reversible trends during the NE and SE Monsoon were clearly illustrated by a rise (accumulation of sand) or fall (depletion of sand) in the relative heights of the profiles at each location. The plots in Fig. 2 illustrate the magnitudes of the changes to the profiles BP 00 – 19 as recorded during three survey campaigns (C) carried out over a 13-month period, C2 (27th August 2007), C4 (8th April 2008) and C7 (30th September 2008). Most noticeable is the increase in relative beach sand height during the NE monsoon, particularly at BP 01, 03 to 07 and 09, where at 20 m and 30 m from the benchmark the increases ranged from 0.3 to 1.0 m. During this season, the overall relative change was an increase in sand height. Conversely, during the SE monsoon, far more significant loss was witnessed, particularly at the 20 m and 30 m distances from the benchmark, of 0.4 to 1.0 m. A complete reversal in the accumulation and depletion of beach sand was also witnessed at BP01, 05, 08, 11, 14, 16 and 18.
Social vulnerability to shoreline changes
Stakeholder interviews and focused group discussions at the Bamburi site revealed that erosion (of the beach plain sand at the backshore) has increased significantly over the last 20–40 years. According to coastal residents, this is most severe during the SE monsoon season. Locally the shore has retreated by 150–200m during the last 20 years. At the Kunduchi site, the reported backshore erosion of the sand spit at the former Africana Hotel since 1967 was estimated to be more than 200 metres.
At Bamburi beach, both fishermen and informal traders such as curio sellers have observed shoreline changes that could affect the shoreline area in which they currently do business. In response to interviews, 30% of respondents stated that the (high-tide) water level had increased and they have been forced to move their businesses; 24% observed shoreline erosion and loss of land adjacent to the shoreline and 17% cited increased construction of sea defenses as the cause of increased erosion. At Nyali, the fishermen have lost part of their fish landing site and have been forced to relocate their temporary banda (shelter) to a narrow space bounded by private land. If displaced by further shoreline regression, they have nowhere else to go. Curio sellers similarly have been impacted by coastal squeeze, losing the open spaces used for the display of their wares. They are now obliged to operate without shelter and moving with the tides. According to the respondents, the lack of shelter leads to pneumonia and related health problems.
When asked about causes of shoreline change, 35% of the respondents cited natural factors, 20% cited increased construction of sea defenses, 8% cited increased human activity at the beach, 6% cited the effects of monsoons, while 28% could cite no reason.
When asked about the social costs of shoreline change, 48% of the respondents stated that they have been displaced after their beach was submerged. 6% cited loss of income, 2% cited the loss of their fish landing site, while 44% did not offer any suggestions about social costs. Respondents identified the main economic costs to be those of replacing infrastructure, and loss of income.
An inventory of protection measures on the Bamburi and Kunduchi shores
At the Bamburi site, a survey of the existing measures (Shaghude et al., 2015) showed that 47% of the 15 km shoreline has some form of protection, mostly masonry walls and revetments. Palisades of coconut tree trunks have also been used. No groynes have been installed on this shore. A similar survey at Kunduchi study site revealed that about 54% of the shoreline is protected in some way, mostly with groynes (Fig. 3). Other common protection methods were masonry seawalls and revetments. Quarried-rock blocks and boulders and concrete blocks have also been used as aprons for protecting fenced-walls, especially in the southern parts of the shore within the Msasani Bay. Quarried-rock boulders were also used as revetments for protecting eroding backshore north of Mdumbwe River.
The costs associated with the protection measures at the Kunduchi site were found to be considerable, extending beyond those of the original capital outlay into long-term maintenance. Measures adopted at the mouth of the Manyema Creek and along its adjoining shorelines are of three types — groynes constructed of quarried limestone blocks, revetments of limestone blocks and seawalls of masonry, some with rock-block aprons. The largest investment, revetments installed in 1999 around the Kunduchi Beach was reported to be valued at US$ 4.5 million. Later, quarried-rock groynes 100–150m long, 2–3m wide and 2–3m high were constructed at a cost of US$ 30,000–50,000. The construction of a 500m seawall was estimated to have cost about US$ 800,000 (equivalent to about US$ 1,600 per metre length).
Beach profile monitoring on the Bamburi shore
The beach profiling results from the Bamburi shoreline show that, while there is considerable intra-annual longshore movement of beach sand resulting in periodic major volume changes, the net transport resulting from the mutually opposing NE and SE monsoons appears to be largely balanced, with no permanent loss of sand from the site. Although there were periodic local erosion events on that shore, especially in the months of January and October, sand volumes were replenished by subsequent beach accretion. This clearly indicates that where the tourism or residential infrastructure are sited appropriately, they will not be adversely affected by loss of beach sand.
Unlike the Bamburi shoreline, the data from the Kunduchi site indicated a long-term overall dominance of forcing from the SE monsoon, with sand “lost” northwards from the Kunduchi shore system around the limestone headland of Ras Kiromoni (Fig. 1) to a beach plain repository at Ununio. These contrasting sand transport regimes may be explained by the different coastal orientations of the respective sites in relation to the monsoonal winds and their consequent wave and current regimes (Shaghude et al., 2013). As a result, any structure inappropriately sited along the shore will be adversely affected by the loss of beach sand and consequent backshore erosion causing shoreline regression.
Social vulnerability to shoreline changes
The shoreline erosion has negatively affected community livelihoods through the phenomenon known as coastal squeeze—a consequence of shoreline regression towards developed coastal land. Increased construction of inappropriate sea defences for hotel and residential developments along the coast and increased human activity at the beach, including sand mining, will exacerbate the problem. Fishermen are the most vulnerable and are therefore concerned that, if the erosion continues at the same pace for another 10 years, they may have to find alternative landing sites elsewhere.
An inventory of protection measures on the Bamburi and Kunduchi shores
While some existing protection structures are well designed and accommodate wave run-up and strong wave-generated currents at their toes, others were observed to be of inappropriate design and thus exacerbating scour, resulting in the undermining and subsequent failure of the seawall, progressive erosion at the backshore and significant beach flattening. Protection of existing investments, e.g. by means of seawalls or groynes, has proved costly. Generally, these protective measures degrade the aesthetic and amenity value of the beach and may even threaten biodiversity, e.g. by impeding turtle nesting. In some instances, expensive protective measures have proved ineffective or unnecessary, or have even exacerbated the hazard (Nyandwi, 2001a, b). There were also no clear regional standards/guidelines for the construction of coastal structures to take account of shoreline changes.
Understanding shoreline change and its drivers should constitute key elements in coastal planning for the successful development of coastal tourism. Beach deposits are an important natural defence against coastal erosion; knowledge of their provenance and fate should be a priority. The potential economic loss due to shoreline change should therefore be considered when siting tourism facilities and determining the protection or mitigation measures to be employed. Adequate policy and legislation must also be put in place to ensure that development along the shoreline is sustainable and in accordance with an enforceable functional zoning plan that considers the conservation of the coastal environment and its resources.
Where appropriate mitigation measures are required to protect key tourist infrastructure, buildings and recreational facilities along the coastline/islands, engineers and planners should be involved in their design and construction. In this regard, there is need for clear standards/guidelines to be used in the assessment and design of management options or mitigation structures to address shoreline change in the region. The guidelines should give strategic mitigation options aimed at reducing the risks associated with shoreline change to coastal tourism development and to coastal communities in general, balancing socio-economic pressures against environmental considerations; and they should be sustainable, at least over the time frame of shoreline change.
The options for strategic management include: “Do nothing” – taking no action; Protecting the shoreline to prevent inundation and erosion; Accommodating erosion and inundation (e.g. through the physical adaptation of structures or the introduction of construction setback regulation); and Retreating from the changing shoreline to a less vulnerable position.