Maritime and Shipwreck Archaeology in the Western Indian Ocean and Southern Red Sea: An Overview of Past and Current Research
- First Online:
- Cite this article as:
- Lane, P.J. J Mari Arch (2012) 7: 9. doi:10.1007/s11457-012-9102-0
This paper provides an overview of previous archaeological research on shipwrecks and the maritime archaeology of the Western Indian Ocean and southern Red Sea. It highlights the early pioneering research on the Santo António de Tanná, wreck off Mombasa Island, before discussing more recent discoveries, surveys and excavations. Attention is drawn to the important distinction between ‘shipwreck’ and ‘maritime’ archaeology, and the need to develop integrated programmes aimed at investigating the diversity of the maritime heritage of these two regions. Particular attention is also drawn to the increased activity of treasure hunters and other threats to the underwater heritage of these regions, the need to sustain ongoing training and capacity building in maritime archaeology and for strengthening existing legislation.
KeywordsMaritime archaeologyIndian OceanRed SeaShipwrecksMaritime landscapes
Maritime and shipwreck archaeology are comparatively under-developed in the Western Indian Ocean (hereafter WIO) region and adjacent southern Red Sea zone (hereafter RSZ) relative to many other parts of the world, despite some early pioneering studies in these areas. Over the last decade, partly as a result of various British and Irish research initiatives (see below) this situation has begun to change. This period has also witnessed a revival of interest in the region on the part of treasure diving companies, with at least one mounting sustained campaigns. There is also growing awareness of the range of other threats to the underwater heritage of these regions (e.g., Pollard 2012, see also Pollard et al. this volume), and the need to revise existing national archaeological legislation to take into account the changing nature of these threats and the opportunities for collaboration provided under the terms of the 2001 UNESCO Convention on the Protection of the Underwater Cultural Heritage (hereafter CPUCH) (see Lane 2007). A particularly encouraging trend has been the steady rise in interest in underwater archaeology among the various national archaeological agencies and the beginnings of investment in the training of local archaeologists as divers and in maritime and underwater archaeology. However, although the potential for future research is great—especially where a maritime rather than an exclusively underwater perspective is adopted (e.g., Breen and Lane 2003; Lane 2005; Pollard 2008a; Mallinson et al. 2009; Rhodes 2010; Christie 2007, 2011), sustaining these developments will be challenging for most WIO and RSZ archaeological services, not least because of the limited number of trained personnel and scarcity of internal funding for both research and underwater heritage protection.
Accordingly, collaborative projects involving partnerships between WIO and/or RSZ archaeologists and colleagues from countries with more established programmes and expertise in underwater and maritime archaeology are likely to offer the most viable means of sustaining the current impetus, at least for the foreseeable future. As discussed below, several recent and ongoing initiatives have been relatively successful at stimulating both interest in and demand for this type of archaeological expertise. Moreover, most of these projects have been carried out with relatively modest levels of funding, the main key to their success being the emphasis placed on training and local capacity building as part of the overall research process. While all, thus far, have been driven by an external and better resourced partner from the UK, Ireland, France, the USA, the Netherlands or, most recently, China, rather than being developed by local archaeologists, the latter are increasingly beginning to establish their own research agendas and priorities. There is also considerable scope for broader collaboration with partners from others sectors of the Indian Ocean, and from colleagues based in Egypt and South Africa where there already exists considerable expertise in maritime and underwater archaeology (Gribble 2002; Morcos et al. 2003; Chirikure et al. 2010; Khalil 2008; Khalil and Mustafa 2002; see also South African Heritage Resources Agency, n.d. and Alexandria Centre for Maritime Archaeology (CMA) and Underwater Cultural Heritage, n.d.). Possibilities exist also for developing collaboration with colleagues based in other Indian Ocean or Red Sea states, especially where these have well established underwater and maritime archaeology divisions—such as the Marine Archaeology Centre at the National Institute of Oceanography (NIO) at Dona Paula, Goa in India. In an attempt to encourage wider international interest in the underwater heritage of both areas, this article provides a summary of previous underwater shipwreck investigations in the WIO and RSZ and recently completed maritime archaeology projects involving the use of marine geophysics and inter-tidal and/or underwater surveys. The paper concludes with some suggestions regarding possible future research topics and priorities.
Early Shipwreck Archaeology: The ‘Mombasa Wreck’
The wreck was protected beneath a deposit of fine silt and coral fragments that was in places up to 2 m deep. This had helped preserve a large section of the hull (approximately 30 m) from the stern to the bow, and to the bottom of the first deck on the port side. At the start of the excavations a 2 × 2 m grid constructed from scaffolding was laid across the centre of the hull section, following the base-line established during the 1970 test-excavation. Three areas were excavated during the initial season, leading to the exposure of the keelson from the stern knee to just forward of the mast-step and the port-side framing (Piercy 1977). Over the following two seasons the rest of the surviving hull was excavated, along with two trial trenches on the port side aimed at investigating debris that had spread from the vessel as it decayed (Piercy 1978, 1979). Recording of the shape of the hull was undertaken by Jeremy Green from the Western Australia Maritime Museum, who used a combination of photogrammetry and recording lateral sections at one metre intervals along the surviving length of the hull and at right angles to the keelson (Green 1978). The excavations revealed that the ship was built entirely of teak (Tectona grandis), with planks secured by iron fasteners. During the final season in 1980, a 7 × 3 m area was excavated down slope of the stern (i.e., off the port side of the vessel as it lies) so as to continue investigation of the dense concentration of finds on this side (Piercy 1981).
Over 6,000 finds were recovered during the course of the excavations. These included items associated with the sailing and navigation of the vessel, such as the remains of sails, rope, lantern fragments, sail maker’s palms and two compasses, along with a range of fittings; armaments including a bronze breech loading swivel gun, iron cannon, iron shot, grenades, and wooden powder flasks; domestic items used by the ship’s crew and their personal possessions, including Portuguese faience and a variety of other ceramics, wooden bowls and bucket. Goods, including mangrove poles (Powell 1996, 1999), believed to have been destined for commercial disposal by members of the crew were also found on board. Detailed study of these finds and their conservation has been ongoing for a number of years (Mwadime 1991; Piercy 1992), and several reports are available, including analyses of the ceramics (Sassoon 1981, 1983, 1991; Willoughby 1991), armaments (Darroch 1986, 1991), rigging (Thompson 1985, 1988, 1991) and the ship’s structure (Fraga 2007; Jordan 2001), as well as various other artefact types (Hall 1991; Oertling 1991; Richardson 1991). A full report on the excavations and the finds is in preparation (Robin Piercy pers. comm. 2006).
The wreck site has been re-surveyed twice over the past decade or so. In 2001, as part of a broader study of Mombasa’s maritime seascapes (Breen and Lane 2003; McConkey and McErlean 2007) and training exercise, a marine geophysical survey involving bathymetric profiling, side-scan sonar and magnetometry was carried out in Tudor Creek on the north side of Mombasa Island, concentrating on the area of the Old Port, including the section of the seabed where Santo António de Tanná lies (Quinn et al. 2007). This highlighted 50 suspected cultural anomalies on the seabed in the vicinity of the Old Port, many of which were subsequently ‘ground-truthed’ by divers. This included inspection of the wreck site, during which it was noted that exposed timbers and frames were both ‘soft to touch’ and showed signs of active degradation by marine borers (Quinn et al. 2007:1458). The side-scan sonar survey of the wreck site also documented the likely presence of a V-shaped anomaly on the steeply sloping eastern flank of the excavated portion, which may represent a further depositional zone associated with the wreck (Quinn et al. 2007:1458). Three of the other anomalies were the wrecks of nineteenth and twentieth century ‘coastal traders’; others were associated with the remains of a former twentieth century bridge across the creek and other modern activity (Forsythe et al. 2003; Quinn et al. 2007). In addition, another anomaly, designated M2, survives on the seabed as a low mound with locally exposed timbers and artefacts below a veneer of sand. This probably represents the remains of a further historic wreck of seventeenth or eighteenth century date (Quinn et al. 2007:1457–1458).
Two other Portuguese vessels are known to have sunk or at least been destroyed during the siege of Mombasa. On Christmas Day 1696 a manchua (a form of single-masted galiot, with a lateen sail and 12 pairs of oars) ran aground below the site of the early second millennium AD settlement of Kongowea—which lies in the general vicinity of the current Coast Province General Hospital (Sassoon 1980; McConkey and McErlean 2007). This may well have been broken up or burnt where it ran aground. An earlier incident happened on 15th September 1697, when another manchua sank shortly after the Portuguese squadron had entered harbour and was in the process of anchoring opposite Fort Jesus. Between 2005 and 2007, Hans-Martin Sommer, who at the time was the NMK’s maritime archaeologist, made a series of dives in the vicinity of Santo António de Tanná and located timbers ca. 25 m distant from that wreck. These were duly photographed and drawn (these records are lodged at Fort Jesus Museum). On the basis of the position of this material, and the recovery of what might be the remains of a rowlock, Sommer has proposed this could represent the remains of the manchua that sank in 1697 (Sommer 2007a). Further investigations are clearly needed to substantiate this, however.
Past Research: Other Shipwrecks
Elsewhere in the region a number of other shipwrecks (see Fig. 1) have been investigated archaeologically. In several cases investigations have entailed non-intrusive mapping of the vessel remains and their associated contents with minimal recovery of any artefacts. In others, the position of the site has simply been noted with no further investigation as yet. As outlined below, however, a few underwater wreck sites have been subject to controlled excavation, while others have been excavated but possibly to rather questionable standards. These different projects are summarised below.
Black Assarca Island, Eritrea
The remains of a probable seventh century AD wreck were found in 1995 by sports divers close to Black Assarca Island, situated in the middle of the Massawa Channel off Eritrea. The island is a small coral outcrop rising only a few metres above sea level and is known as a shipping hazard today. An initial survey of the site revealed a mound of amphorae and scattered pottery covering an area of about 7 × 15 m. Limited excavations and further mapping of the site were conducted in 1997, directed by Ralph K. Pedersen from the INA under the auspices of the Ministry of Marine Resources of Eritrea, with additional financial support from the British Institute in Eastern Africa (BIEA) (Pedersen 2000, 2008). This later work located other clusters of ceramics of Mediterranean and Near Eastern origin on the seabed. The bulk of the finds were amphorae, most probably used for transporting wine. ‘Ayla-Axum’ type amphorae were the most common form represented in the assemblage. These have a characteristic long, conical ‘carrot-shaped’ form with external rilling (or corrugations). They are so named because examples are known from Aksum in Ethiopia, the capital of the Aksumite Kingdom (Phillipson 2003), and also occur around Ayla/Aila (the ancient name for Aqaba) (Pedersen 2008). Other dated examples have been recovered from Adulis on the Eritrean coast (Munro-Hay 1982; Peacock and Blue 2007), which served as Aksum’s main seaport, and from the Ptolomeic harbour of Berenike in Egypt (Sidebotham and Wendrich 1998). Another amphorae wreck is known from further north in the Red Sea around Fury Shoal, Zabargad Island in Egyptian waters, about 40 km north of the Sudan border, where Dressel 1 and 2 type amphorae dated to the first century BC—first century AD have been observed (Purpura 2004), and more recently mapped (Blue et al. 2010:95–96). A scatter of third or fourth century AD ceramics has also been noted on the seabed near a small island beyond the Dahlak Kebir archipelago in Eritrean waters, and there are also reports of a deep water wreck lying between Black Assarca and Massawa (Pedersen 2008:91); neither have been investigated archaeologically.
Manda and Ungwana, Kenya
Over the years, there have been periodic reports of finds of Chinese and other imported ceramics being recovered from the seabed between Manda and Pate Islands, in the Lamu archipelago, northern Kenya, and examples of some of the material (which is mostly eighteenth century or later in date—Mark Horton personal communation 27/7/2010) recovered is on display in Lamu Museum. The long history of trans-oceanic trade between East African coastal communities and India, China, the Persian Gulf and the Mediterranean world is well documented as a result of archaeological research on coastal sites such as Shanga in northern Kenya and Kilwa in southern Tanzania (see Kusimba 1999; Horton and Middleton 2000; Sinclair 2007 for summaries). Consequently, it is highly likely that traces of wrecks associated with this trade and dating from at least the last few centuries BC onwards may occur all the way along the East African coast. In this sense, these reports of underwater finds of Chinese porcelain are in no way surprising.
More recently, however, it has been suggested that the remains of a Chinese ocean-going junk that formed part of the fleet led by the Ming Dynasty Admiral Zheng He during the fifteenth century (Finlay 2008) may have survived. Since 2006, the NMK and the Chinese government have conducted limited diver surveys aimed at locating this wreck and a more substantial campaign was initiated in 2010, following signing of a joint agreement between the governments of China and Kenya. Initial underwater investigations were conducted in 2006 and 2007, in the channel where fishermen have long reported the presence of pottery, and also further south in the inter-tidal zone and offshore from the site of Ungwana on the northern side of the Tana River delta (Sommer 2007b). This located the remains of three wrecks in the inter-tidal zone around Kipini (all twentieth century in date) and noted the severity of erosion affecting coastal sites here (which, among other impacts had exposed four human burials), but failed to locate any submarine remains. Further underwater survey resumed in late 2010 (Murphy 2010) apparently resulting in the location of 35 wrecks between Lamu in the north and Watamu in the south (Gari 2012), although no significant results have been reported yet. A second phase of research began in 2012 (Xinhua 2012), which may include detailed investigations and excavations of some of the newly located wrecks, which include a 150–200 year old vessel off Mambrui, just north of Mombasa. The possibility of developing five of these wrecks into underwater museums is also under consideration (Gari 2012).
Mafia Archipelago, Tanzania
Ten brass canons, three anchors, fragments of canon balls and a bronze bell were located by divers in shallow water close to the southern shore off Juani Island, Mafia, in 1977. The bell was recovered and bore the following inscription: ‘H … &B H Co. 180_’ (the last figure might be 9) (Antiquities Division, Government of Tanzania 1980:9). The position of the wreck was relocated by Frontier Tanzania in 1990 during preliminary marine biology surveys prior to the establishment of Mafia Marine Park (pers. obsv. 1990). They have yet to be fully documented, however (Lane 2005).
Boudeuse is a low-lying and treacherous cay off Admiral Island, Seychelles. The remains of a wreck of a Portuguese nau were located here by local fisherman in the 1970s, and the site was surveyed by Warren Blake and Jeremy Green from Western Australia Maritime Museum in 1976. The site had been extensively damaged by previous efforts at salvage, and at least 30 cannon are known to have been removed along with other artefacts. Most are reportedly in private collections, although a few are held by the Carnegie Museum in Victoria, Mahe Island. At the time of the 1976 survey, parts of the frame and bottom planking were preserved, covering an area of about 50 × 10 m. It is believed that the vessel represents the remains of Santo António that sank in 1589 (Blake and Green 1986). This should not to be confused with the Santo António de Tanná that sank in Mombasa Harbour, or the other Portuguese ships called Santo António that are known to have sunk elsewhere in the Indian Ocean.
Ilha de Moçambique and Nampula Province, Mozambique
Mozambique Island is a small coral island, roughly 3 km long and 500 m wide, situated at the mouth of Mossuril Bay in the Mozambique Channel about a kilometre from the mainland. Today, it forms part of Nampula Province in northern Mozambique. The island has no fresh water to speak of, and it is not known exactly when the island was first settled. However, it is likely to have been occupied well before the emergence of a Swahili settlement there in the early part of the second millennium AD (Duarte 1993). When Vasco de Gama, who led the first post-medieval European expedition to explore the WIO, landed on the island in 1498 and claimed it for Portugal, he found a thriving port community with Arab and other shipping in its deep, sheltered anchorage between the island and the mainland. Although gold, ivory and timber were among some of the exports, Newitt (2004) suggests that the primary reason for the town’s existence was as a boat-building centre. The skills of the local inhabitants were certainly quickly recognised and utilised by the Portuguese settlers following Vasco de Gama’s return in 1502 (Newitt 2004). Over the next few decades, over a hundred Portuguese nau and other ships visited while on the carreira da India, as the route from Lisbon to Goa was known. By 1540, Mozambique had become the second most important naval base, after Goa, in the Indian Ocean and the town later became one of the most important captaincies of the Estado da India (Newitt 2004). The town remained the capital of Portuguese East Africa until the late nineteenth century when this role was transferred to Lorenço Marques (now Maputo). In view of its long and varied history and the wealth of surviving architectural remains, the island was designated a World Heritage Site in 1991, and an international campaign to restore and conserve this heritage was launched by UNESCO in 1997.
Given the significance of Ilha de Moçambique as a ship-building location, trading port and naval base during its history, the area in the vicinity of the island is known to have numerous wrecks, and surveys along the Mozambique coast show that the waters around the Island “are by far the richest spot along the Mozambican coast with regard to the underwater archaeological heritage” (Macamo 2001:6; see also Duarte, this volume). Preliminary surveys were undertaken during the colonial era and subsequently extended between 1996 and 1999 by archaeologists from Eduardo Mondlane University, with support from Brown University and the US National Park Service (Duarte, this volume). This resulted in the location of several important wreck sites and other archaeological material. However, further, systematic archaeological research has been hampered since the award in 1999 of an exclusive concession by the Mozambique Ministry of Culture to Arqueonautas Worldwide (established in 1995 as a private share holding company in Madeira, [Portugal Região Autónoma da Maderia 2003]) to conduct surveys and excavations of wrecks around Ilha de Moçambique. In April 2003, the company was granted an extension until October 2009 to cover 700 km of coastline along the coast of Nampula Province (Arqueonautas Worldwide 2010, ‘Projects and historical research’), which has since been extended (Duarte, this volume). The company’s partners are Patrimonio Internacional SARL. The company uses a combination of archival research, marine remote sensing (typically magnetometer survey), diver reconnaissance of magnetometer anomalies, and excavation, and a statement of its methods appears on the company website (Arqueonautas Worldwide 2010, ‘Archaeology’). According to the company’s website, thus far some “79 historical shipwrecks have been confirmed by historical research of which 22 are of outstanding cultural and historical value”, although as recently as May 2012 the discovery of a newly located wreck site, possibly the remains of São Miguel e Almas which sank en route from Portugal to India in 1771, was reported on the company website (Arqueonautas Worldwide 2010, ‘Current News’).
While the majority of located wreck sites comprise little more than scatters of ballast and some associated artefacts, several of the better preserved sites have been mapped and partially excavated (for details see Duarte, this volume). The excavated wrecks include the wreck of a Portuguese ship Nossa Senhora de Consolação (1608) designated IDM-003 (Mirabal 2001, 2004a, 2006, 2007); wreck IDM-007, comprising two ballast piles and a scatter of artefacts (which included gold artefacts and amber trade beads), many of which had accumulated in the numerous ‘blow holes’ in the sea bed (Mirabal 2008); wreck IDM-010 which is thought to be that of a mid-nineteenth century English trader, from which several hundred bottles reportedly filled with gin, wine and champagne have been recovered (Mirabal 2007); Nossa Senhora da Madre de Deus e S. José, a Portuguese Indiaman (designated wreck MOG-001) which sank on 2nd September 1802, and comprised a scatter of ca. 100 cannon, some associated cannon balls and artefactual material including several later eighteenth century gold coins and a number of copper sword hilts (Mirabal 2008, 2009); San Jose, which sank in 1622 (designated MOG-003), from which four bronze cannon and ca. 24,000 silver coins have been recovered (Mirabal 2005, 2006, 2007, 2008); and wreck ANG-005, believed to be the remains of Bredenhof, a Dutch East India Company (VOC) vessel, first located in 1986 (see below), from which Arqueonautas Worldwide recovered over 3,700 copper duit coins (Mirabal 2001:8).
Full reports on these excavations are promised, although at present only interim reports, posted on the company website are available. The company has also contributed to the redesign of the Museo da Marinha (Museum of the Navy) on Ilha de Moçambique, which re-opened in August 2007. This contains displays of many of the finds generated by their activities. The company also aspires to be considered a serious archaeological research body. For instance, founder and chief executive of Arqueonautas, Count Nikolaus Sandizell, in an interview for the New York Times in April 2001 reportedly stated that “We don’t consider ourselves treasure hunters. We consider ourselves as more a maritime archaeology group. Our image is a little bit more serious” (Kinetz 2001). However, the company is equally concerned with ‘salvage’. For instance, it sold over six tons of ‘low alpha lead’ from N.S. de Consolação to ‘Global LAL Inc., USA’ in March 2007 (Lane 2007), and in December 2000, Sotheby Holdings Inc. auctioned finds from nine wrecks the company had investigated in the Cap Verde Islands (Sotheby 2000). More recently, Ming ceramics and gold from IDM-002 were sold by Christies in 2004, and further pieces of gold scrap from this wreck were sold by Sedwick Coins in May 2007 (Sedwick 2007). Artefacts, described as mostly ‘repetitive’ (with the exception of the coins, seals and parts of ship structure), can be purchased directly from the company via its ‘online shop’ webpage (Arqueonautas Worldwide 2010—‘ARQ’s On-Line Shop’).
This East India Company ship sank around Bassas d’India in the Mozambique Channel in 1738. The remains of the wreck were located in 1987, and investigated by a team from the French Department of Marine Archaeological Research (DRASM) (Bousquet et al. 1990; L’Hour et al. 1991). The ship was carrying a cargo of Chinese porcelain when it sank, some fragments of which were recovered from the seabed along with some glass fragments. Traces of an unidentified nineteenth century wreck were also located on the reef itself, while a cast iron cannon associated with two large anchors found near the southern tip may represent the remains of Santiago—a Portuguese carrack that was lost in 1585 (Bousquet et al. 1990; see also Duarte, this volume).
This British 5th rate frigate (i.e., a frigate mounting 32–44 guns with the main battery on a single deck) was built in 1797, and was wrecked off Mauritius in 1810. The site was discovered in 1964, and remains lay scattered across the bottom mud at depths extending from 6 to 24 m; the bow was reported to be still in situ embedded in the mud. Parts of the vessel were damaged during salvage operations in 1968, and limited excavation of the site has been conducted by a team from Australia (von Arnim 1998).
This ship, which belonged to the French Compagnies des Indes Orientales, sank on 31st July 1761 off ‘Sandy Island’ (now known as Tromelin Island) while en route to Île de France (Mauritius) from Madagascar after being blown off course by a storm. It was carrying a cargo of Madagascan slaves that had been acquired illegally. One hundred and twenty members of the crew and around 60 slaves were able to reach the comparative safety of the island, and supplies were salvaged from the wreck. The island is a sandy cay little more than a square kilometre in extent with a maximum elevation of ca. 8.5 m (Marriner et al. 2010). It is situated in an area prone to cyclones, and the shores are often swept by storm surges, making settlement hazardous. The French crew built a makeshift boat and sailed away, leaving the slaves behind with food for about 3 months. Despite promising to return, on reaching Madagascar the crew were prevented from returning to help rescue the slaves. Fifteen years later, in November 1776, seven women and an eight-month old baby were rescued by the crew of La Dauphine (Guérout and Romon 2008). The fate of the other slaves is unknown, although they are presumed to have died on the island or perhaps while trying to make their own escape.
Archaeological and related studies were conducted on Tromelin in 2006, 2008 and 2010, coordinated by a French not-for-profit organisation the Groupe de Recherche en Archéologie Navale (GRAN), as part of a broader programme of research on the history of slavery in the WIO coordinated by UNESCO (Guérout 2007; Guérout and Romon 2007). An underwater survey was conducted around the island in 2006, supplemented by geoarchaeological survey of the island aimed at learning more about how the abandoned slave population adapted to this hostile environment. Divers located remains of the ship, including iron anchors and cannon, scattered over a wide area. Each item was recorded in situ and their location on the seabed was plotted. Concurrent with this, test-excavations were undertaken on land in an effort to locate the camp used by the Madagascan and French survivors. Eventually, the remains of a small house built from coral and salvaged material was located, as well as the remains of a small oven which was used for baking biscuits. Further excavations in 2008 exposed two further buildings, and the whole complex appears to represent the remains of a dwelling, workshop and kitchen (Marriner et al. 2010:1303). The reports by Guérout (2007; Guérout and Romon 2007) provide further details, while some of the finds and plans of underwater remains can also be seen on the project website: http://www.archeonavale.org/Tromelin/.
Four cannon and an anchor were located in shallow water near Belitsaky by Michael Parker-Pearson and his team while conducting research on the archaeology of the Antandroy Kingdom. These possibly derive from Degrave, which was wrecked on the southern tip of Madagascar in 1703 while en route back to England from Bengal, forcing members of the crew to abandon ship (Parker-Pearson and Golden 2002; see also van den Boogaerde 2009:89–96). The survivors from Degrave were duly captured and enslaved by the local Antandroy king, Andriankirindra. One of them, Robert Drury, kept a diary of his time on the island and after 16 years on Madagascar, mostly in captivity, he managed to return to England, where his diaries were later published (Parker-Pearson and Golden 2002).
This British warship was captured by John Paul Jones in September 1779, during the American War of Independence following a naval battle in the North Sea off Flamborough Head. After capturing the vessel, John Paul Jones sailed to Texel on the Dutch coast with the rest of his fleet, where Serapis and the rest of Jones’ fleet (except the Alliance) were placed under the French flag. In 1781, Serapis was sent to the French fort on Île Sainte Marie, off the northern coast of Madagascar, under the command of a Captain Roche. While ashore, a fire broke out on board and despite the best efforts of the crew, the ship sank (van den Boogaerde 2009:205–207). The wreck of Serapis was relocated in late 1999 by Richard Swete, an American ‘nautical archaeologist’, and his associate Michael Tuttle, following several years of magnetometer survey of Ambudifutatra Harbour (Serapis Project 2008). In May 2004, a team of archaeologists and historians led by Michael Tuttle, Prof. Jean Aime Rakotoarisoa and Dr. Chantal Radimilahy from the University in Madagascar, recommenced research on the site. An area of copper lining from the hull and an associated ballast pile was located, within which numerous artefacts were observed. These features were mapped and recorded, with a limited number being lifted for diagnostic purposes and further analysis. These have been deposited at the University of Antananarivo Museum of Art and Civilization for conservation (Society for Historical Archaeology 2004).
Other Wreck Sites
Numerous twentieth century wrecks, including some sunk during both World Wars are known in the southern Red Sea and WIO. Many of these are visited on a regular basis by sports divers. In previous decades, sports divers commonly removed ‘souvenirs’ from such sites. Although such practices continue, self-policing by sports divers and the dive companies that operate in the region has encouraged a significant reduction in this kind of threat to the integrity of the underwater heritage. Far greater threats are posed by commercial treasure hunters, who often style themselves as ‘maritime archaeologists’ while exploiting loopholes in the existing international and national legislation governing the protection of the underwater cultural heritage, salvage and the sea (see Lane 2007 for a synopsis).
List of known Portuguese Indiamen shipwrecks in the Western Indian Ocean
Location and date wrecked
Coast of Malindi, 1501
Parcel de Sofala, 1502
Mozambique or Kilwa, 1505
Leaving the Kilwa barra, April 6, 1506 & two others
N Sra da Luz
Santa Maria das Virtudes
Between Sofala and Mozambique, 1507
Between Sofala and Mozambique, 1507
Between Sofala and Mozambique, 1507
Between Sofala and Mozambique, 1507
Indian Ocean? 1508–1509?
Malindi/Kilwa? or Mafia? 1519–1520?
Between Mozambque and India, 1524
Low waters off Malindi, 1524
Mozambique, August 1524?
Close to Mozambique, 1524
S Asanto Antonio?
Parcel de Sofala, 1528
S. Dinis or Ajuda?
Socotra or Mozambique, 1530?
Comoro Islands, 1547
Socotra, Sept 1547
Close to Mozambique, 1549
Mombasa, Aug 20, 1561
Flor de la Mar
Mozambique, May 16, 1566
Entrance of Mozambique barra, 1576
Parcel de Sofala, 1583
Rio de Quelimane/Parcel de Sofala, 1582
Between Mozambique and India, 1589
N Sra da Nazaré
Mozambique, July 15, 1593
Between Mozambique and Goa, Aug 17, 1594
N Sra da Victória
N Sra do Rosário
Mozambique, April 8, 1597
Socotra, end 1601
Mozambique, Sept 1, 1607
N Sra da Consolação
Mozambique, July 25, 1608
N Sra da Salvação
Mombasa, April 17, 1609
Mozambique, Aug 17, 1608
Maldive Islands, March 22, 1615
Comoro Islands, Aug 18, 1616
São João Evangelista
Close to Mombasa, 1620
Santa Teresa deJesus
Entrance of Mozambique barra, July 25, 1622
Mozambique, July 25, 1622
Mozambique, Jan 28, 1624
Mozambique, Jan 28, 1624
Mozambique, Jan 28, 1624
N Sra da Guia
Between Mozambique and Muscat, 1623/24?
Mozambique, Dec 27, 1642
Close to Mozambique, 1647
N Sra do Bom Sucesso
N Sra da Salvação
N Sra dos Remédios
Mombasa or Mozambique, 1670?
Confirmed pre-twentieth century wrecks around Madagascar
Location and material remains
Southern Madagascar; numerous cannon
Sainte-Marie Island, NE Madagascar (vessel used by the pirate William Kidd; others reported from vicinity are Flying Dragon (1721), Rouparelle (1698), Mocha (1699) & New Soldado (1699)).
Nossa Senhora do Monte do Carmo
Salara reef, west coast; cannon and anchors, artefact scatter
South Madagascar; iron cannon, ship’s bell
? Star Bank, southern Madagascar; large anchors and cannon
Sainte-Marie Island, NE Madagascar—copper box, copper hull lining, ballast, artefact scatter
Point St Felix (Ambatomifoka); anchors, cannon, ballast and artefact scatter
Near Salara, SW Madagascar; anchors, hull fragments, artefacts
Nosi Mahampana (Barracouta Islands) NE Madagascar; steel hull
Known Dutch East India Company (VOC) wrecks
Approximate location and reason for loss
Date of wreck
Ran aground, Cabacella, Mozambique
26th May 1607
Near Maldive Islands
Lost near Socotra
Lost near Cape Mauritius
Lost near Mauritius
South-east coast of Madagascar
29th Aug. 1639
2nd Dec. 1656
‘Cape Hangklip’, East African coast
20th Feb. 1673
Capsized near Maldive Islands
Lost off Mozambique coast
9th June 1711
8th May 1726
Foundered on reef off Mozambique coast during hurricane
6th June 1753
Foundered on a reef near Mauritius during hurricane
6th March 1615
Mauritius, ran shore during hurricane
6th March 1615
Near Mauritius during hurricane
6th March 1615
Near Madagascar, lost during a storm
18th March 1625
Brandao, east of Madagascar
12th Feb. 1662
Raadhuis van Middelburg
Huis te Forest
Delagoa Bay, deliberately set on fire as no longer seaworthy
1st Jan. 1757
Between the Cape and Mauritius
Almost 400 years after Sao Rafael foundered, 150 Arab and Indian dhows and many European ships were sunk in and around Zanzibar Harbour following a major hurricane in 1872 (Sheriff 1987:234). Since many of the dhows were laden with cargo, one might expect that the remains of several of these would have survived on the seabed. In fact, recent marine geophysical surveys by the Centre for Maritime Archaeology, University of Ulster, has established that conditions for wreck preservation in this area seem to have been extremely poor, and only minimal traces survive (or, perhaps are not detectable with current technology) (pers. comm. Colin Breen 2/10/07). This underlines the fact that while maritime trade is known to have been extensive for many centuries, and ships are known to have sunk for various reasons, this does not guarantee that all such wrecks will have left archaeological traces.
Most of the published information concerning European wrecks in the WIO and southern Red Sea has been compiled by enthusiasts and amateur historians. Details are mostly published on the Internet, although occasionally also in book form. Good examples of the latter are the books by Patience (2006) which covers East Africa and van den Boogaerde’s (2009) book on shipwrecks off Madagascar. The latter lists a total of 10 Portuguese East India Company vessels, eight VOC ships, four British East India Company ships, 21 French East India Company vessels, five Royal Navy ships, five French navy vessels, 17 pirate ships and 88 other vessels (including a German U-boat from WW2 and two Japanese midget submarines) known to have been wrecked, scuppered or otherwise sunk in Madagascan waters since ca. AD 1500. The remains of only a handful of these have been relocated by divers, however (Table 2).
The following brief summary of some of the better known wrecks is intended to illustrate the nature of this resource, the potential for further research and the ongoing threats from treasure hunters. It is not intended to be comprehensive. It also needs to be reiterated that while the archaeology of the ‘deep sea’ is restricted to wrecks (including those of aircraft) and materials jettisoned from vessels, the underwater archaeology of the continental shelf also comprises the remains of former terrestrial sites that were subsequently inundated following changes in sea-level at different times in the past (Åse 1987; Ramsay and Cooper 2002; Camoin et al. 2004). No such sites have yet been located, but as demonstrated by recent work off the Cape of Good Hope, South Africa (Werz and Flemming 2001), and on-going research in the Red Sea (see below), it is probably only a matter of time before similar remains are located off the East African coast.
Eighteenth-Nineteenth Century European Wrecks
St. Geran was commissioned by the French Compagnies des Indes Orientales and launched at Orient, France in 1736. She was wrecked on her maiden voyage during a violent storm off the north coast of Mauritius in August 1736. Nine members of the crew were rescued (out of a total of 149), and the heroic efforts of the islanders later inspired the novel ‘Paul et Virginie’ by the French author Bernadin de Saint-Pierre. At the time St. Geran was wrecked, she was carrying a cargo of textiles and metal-ware to be exchanged for spices on reaching her destination, and a batter of 18 cannons (Blot 1979). Pieces have been salvaged from the wreck over the years, and while most artefacts are now in private collections some are on display in Mahebourg Museum on Mauritius (Tirvengadum 1983).
Known English East India Company wrecks
Island of João da Nova, Primeira Isles, off Mozambique
60 miles from Mozambique
Wrecked off Madagascar
Lost with all hands off Mauritius
Near town of Mozambique
50 leagues east of Reunion
Parted company on 14th March 1809 with the main convoy of East Indiamen off Mauritius in a gale and not heard of again
Jane Duchess of Gordon
Wrecked on a coral reef in the Mozambique Channel on 20th June 1817
Cargados Carajos (St Brandon shoals) off Mauritius
Nossa Senhora do Monte do Carmo
This Portuguese ship sank off the west coast of Madagascar on 18th August 1774. It was re-located and ‘salvaged’ by a Belgian self-styled ‘marine archaeologist’ Robert Stenuit in the late 1980s and early 1990s (Anon 1992). Little has been reported regarding what was found or recovered, however.
World War I Wrecks
Naval engagements between British and German forces during the course of WWI resulted in vessels on both sides being sunk (Charlewood 1960a, b; Hatchell 1954). These include the German battleship S.M.S. Königsberg, which was eventually sunk after extended naval action in the Rufiji River Delta in 1915. Over time, the hulk sank gradually into the soft mud and parts of the superstructure were salvaged in 1962, although up to 1965 the freeboard remained visible above water. It is now completely buried, although potentially still quite well preserved. Various items have been salvaged from the ship over the years. These include her main guns which were recovered by the Germans and modified for use on land. After the defeat of German troops, captured guns were taken as trophies to Pretoria, Leopoldville, Stanleyville, Kampala, Dar es Salaam and Mombasa. The latter example stands on display outside Fort Jesus Museum, alongside one of the guns salvaged from the British cruiser H.M.S. Pegasus, which was sunk by S.M.S.Königsberg off Zanzibar town during a surprise attack on 20th September 1914 (Hatchell 1954). Other items in public collections include a wardroom dining set (National Railway Museum, Nairobi, Kenya), zinc cordite cases (Zanzibar Museum), a porthole (Dar es Salaam Museum, Tanzania) and a torpedo director and shell casings (Imperial War Museum, London). The wreck of Pegasus was sold in 1955 and broken up for scrap, however parts remain on the seabed and it is a popular dive site. There is one other vessel associated with Königsberg that can still be seen. This is S.S. Somali, which served as the supply ship for Königsberg, which was sunk during the same naval engagement slightly further north in the Rufiji Delta near Salale. For years it was obscured by vegetation, but some of this is reported to have been cleared sometime between 1995 and 2000, to expose the greater part of the hull and surviving superstructure.
Seascapes and Landscapes: Recent Maritime Archaeology Projects
On mainland Eastern Africa from Somalia to Mozambique, the offshore islands of Pemba, Zanzibar and the Mafia archipelago, and the Comoro Islands, there is a long and distinguished tradition of ‘coastal’ or ‘Swahili’ archaeology (for overviews see Kusimba 1999; Horton and Middleton 2000). Similar research is also well established on Madagascar (Wright 1993). In all these areas, scholars have long recognised that access to the sea and the utilisation of maritime resources, as well as trans-oceanic trade were important contributory factors to the rise of Swahili towns and the evolution of a distinctive Swahili culture (e.g., Prins 1965; Allen 1993; Pearson 1998; Horton and Middleton 2000), and quite possibly pre-Swahili communities as well (e.g., Chittick 1979; Smith and Wright 1988; Chami 1999a; Blench 1996; Sinclair 2007; Boivin and Fuller 2009; Boivin et al. 2009). However, despite the importance of trans-oceanic trade to these communities, there has been no systematic attempt to establish whether any shipwrecks (or any other underwater remains) associated with the Indian Ocean trade during this era survive, despite occasional reports of finds of Chinese porcelain and similar materials being dredged up by fisherman at various points off the East African coast (see above). Also, until the last decade, research on specifically maritime themes was limited to preliminary studies of the information available from ethnography, historical sources and ‘grafitti’ as to the types of boats used (Garlake and Garlake 1964; Chittick 1980; Prins 1982; Gilbert 1998; Poumailloux 1999), and a few studies of fishing strategies and technologies (Horton and Mudida 1993; Kleppe 1996, 2001; Van Neer 2001), and the exploitation of maritime resources such as mangroves (Whitehouse 2001; Radimilahy 2001) and salt (Chittick 1975).
Over the last decade, this situation has begun to change and the value of adopting a ‘cultural landscape’ approach to the investigation of the maritime archaeology of the region, both for the purposes of compiling a more comprehensive inventory of the range of archaeological remains that survive on the coastal foreshore, inter-tidal zone and underwater along the East African coast and around the offshore islands, and for changing research paradigms, is increasingly recognised (Breen and Lane 2003; Lane 2005). Particularly influential here has been the work of staff and research students at the CMA at the University of Ulster in developing integrated approaches, often in collaboration with local research bodies including the Coastal Archaeology Unit of the NMK, the Tanzania Department of Antiquities, the Department of Archaeology, University of Khartoum, the National Corporation of Antiquities and Museums of Sudan (NCAM), and the BIEA.
The CMA’s first project was on Mombasa Island on the southern Kenyan coast, which has been witness to historical settlement and port activity for nearly two thousand years (Berg 1968; Sassoon 1980, 1982). As outlined above, the primary aim was to study the port-town and island of Mombasa using contemporary integrated landscape approaches. It was interdisciplinary in scope and involved archaeologists, geologists, geophysicists and historians, involving systematic foreshore and inter-tidal zone surveys, marine geophysical surveys supplemented by targeted inspection of sub-surface anomalies by divers, and some small scale excavation on land and use of various archival sources (Breen et al. 2001; Breen and Lane 2003; Forsythe et al. 2003; McConkey and McErlean 2007; Quinn et al. 2007). Subsequent work by staff and students of CMA have included a marine geophysical survey of the Zanzibar channel between Zanzibar Stone Town and Bagamoyo on the mainland (Breen in prep.); detailed investigation of the maritime landscapes of the medieval town and port of Suakin on the Red Sea coast of Sudan; and two maritime archaeological landscape projects carried out by CMA students as part of their PhD research (Pollard 2007; Rhodes 2008).
These recent excavations also demonstrate that prior to the establishment of Ottoman control, off-loading from ships was typically by landing bow forward to the shore, giving easy access to the nearby merchants’ houses (Mallinson 2012:166). Subsequently, however, a quay and accompanying custom’s office were built, which was later to form a larger quayside. Trade at the port also expanded rapidly, prompting the Portuguese explorer João de Castro, who visited in 1541, to liken the scale of activity to that seen in Lisbon (Smith et al. 2012:179), which is also attested by the diverse range of ceramics from China, India, the Persian Gulf, Egypt and other parts of the Red sea zone that have been recovered (Mallinson et al. 2009; Smith et al. 2012). As well as this more general work, detailed investigations of the developments under British colonial rule in the late nineteenth century have been carried out by Rhodes (2011). This aspect of the project focused on the creation and modification of a customs complex, the development of the waterfront, and the colonial administrative buildings on the main island, and related changes on the smaller Condenser/Quarantine Island, with its piers and associated railway.
Adulis, on the Eritrean coast ca. 50 km south of Massawa (see Fig. 4), was another important Red Sea port and has also been the focus of recent studies from a maritime archaeology perspective. Adulis is mentioned in the Periplus as a source of ivory, rhinoceros horn and tortoise-shell exports and as offering a safe harbour for mariners en route to India, while the sixth century ‘Christian Topographies’ by Cosmas Indicopleustes refers to two other places, Gabaza and Samidi (Peacock and Blue 2007). Although the location of Aksumite Adulis has been known since Henry Salt’s research in 1810, until the recent studies by a joint Eritro-British team coordinated by Southampton University and the University of Asmara, there were several uncertainties about whether this was also the Roman port. Specifically, the Periplus describes Adulis as ‘a fair sized village’ 20 ‘stades’ (3.3 km) from the sea, and that previously ships had moored off Diodorus Island which was approached from the mainland by a causeway, but following attacks by local barbaroi an alternative offshore anchorage at Oriene was used (Blue et al. 2008:301–302). Today, Adulis is situated 7 km from the sea and most of the surface remains at the site date to the period of Adulis’s prominence as the main port for the Aksumite kingdom, thus raising doubts that the ruins also mark the location of the town during the Classical era. However, by using a combination of sediment coring, remote sensing, and survey the recent project managed to establish that Roman Adulis underlies the main Aksumite town, and, that sedimentation and other coastal geomorphological processes had caused a seaward extension of the shoreline. Identification of the likely location of the shoreline in the early first millennium AD from satellite imagery coupled with detailed sedimentological analyses, also resulted in the identification of the former Diodorus Island in the Galla Hills to the south. The later Aksumite places of Gabaza, also in the Galla hills, and Samidi some 7 km to the north, were also relocated (Peacock and Blue 2007).
Further south, recent maritime cultural landscape studies have focused in particular on the development of waterfront and harbour facilities under different economic, socio-cultural and political conditions, and on the exploitation of marine resources. Although closely modelled on the Mombasa Harbour study, only one of these has involved the integrated use of marine geophysics and diver surveys (Breen in prep.). This was conducted in 2005 in the area of sheltered water that lies between Zanzibar Stone Town and a series of small islands a few kilometres offshore that protect the harbour. Zanzibar Island has been a centre of Indian Ocean commerce and cultural developments for centuries, and previous archaeological research along this stretch of coastline has suggested the presence of a number of small fishing villages since at least the early centuries AD. Those in the immediate area of what became Stone Town, however, only grew in size after the arrival of the Portuguese in the late fifteenth century.
For his doctoral research, Edward Pollard (2007) investigated the maritime cultural landscapes around Bagamoyo and Kaole in central Tanzania, and Kilwa in the southern part of the country (see also references in Pollard et al., this volume). The aim was to build on the well established body of work on the archaeology of Swahili and earlier ‘Early Farming and Metal Working’ (or ‘Iron Age’) sites along the East African coast, by adopting a much more explicit focus on identifying the ‘maritime’ elements of these. In particular, this entailed documentation of any evidence for a nautical infrastructure associated with these sites (e.g., landing places, wharves, jetties, lighthouses, navigational guides), evidence relating to the marine economy (e.g., fish-traps, fishing equipment) and exploitation of marine resources (e.g., fish, mangroves, shell-fish, salt, coral), transportation facilities (shipwrecks, ballast, coastal access routes), and social and religious manifestations of a ‘maritime culture’ (such as manifested by coastal shrines, mosques, settlement layout) (Pollard 2008a:36–42). A combination of inter-tidal and coastal surveys supplemented by selected test-excavations at several sites and ethnoarchaeological studies of the contemporary maritime landscape generated a wealth of new insights (see Pollard 2008a), especially with regard to the maritime landscape around Kilwa Kiswani, southern Tanzania, during first half of the second millennium AD. Among other features, the inter-tidal surveys identified a sequence of previously unrecognised, yet substantial, coral causeways and platforms of likely thirteenth-sixteenth century date. Some of these were used to access low water perhaps for unloading ships but also perhaps to facilitate exploitation of marine resources including shell-fish beds (Pollard 2008b); others may have been used as navigation markers (Pollard 2011) or breakwaters.
Pollard’s surveys also identified the social and religious significance of the sea to the Kilwa community, as evidence by, for instance the location of particular mosques (Pollard 2008b, c), and also the diversity of economic activities, from salt extraction and lime burning to fishing that took place along the foreshore and in the intertidal zone (see also Lane 2005). Although smaller in scale, comparable studies have now been conducted by Annalisa Christie around Malindi (Christie 2007) and in the Mafia archipelago, especially around the ca. twelfth to seventeenth century AD site of Kua on Juani Island (Christie 2011). A particular feature of this study was that it sought to explore what archaeologists mean by referring to sites, landscapes or cultures as ‘maritime’, concluding that in many cases the use of the term simply implies a use of marine resources or settlement beside the sea, neither of which are especially insightful observations. Instead, Christie urges the development of a more socially informed perspective that recognises that not all societies that live beside the sea have a ‘maritime culture’ as well (Christie 2011, Christie, in press a; see also Westerdahl 1992). In the Kua case, a combination of archaeological survey and excavation supplemented by ethnoarchaeological studies of fishing practices (the first to be conducted in the region) and oral histories, Christie was also able to demonstrate evidence for differential access to maritime resources between members of the Swahili elite and ‘commoners’ living at the site (Christie 2011, in press b).
Mafia was also a focus of Daniel Rhodes’ recent maritime landscape study, along with a number of coastal towns (Mombasa, Tanga, Pangani, Bagamoyo, Unguja, Dar es Salaam, Chole and Kilwa Kivinje) along the Tanzania and Kenya coasts. In contrast to the work by Pollard and Christie, Rhodes’s focus was on the nineteenth century archaeology and architectural history of these towns and especially following the establishment of European colonial rule (British and German). This remains a neglected area of study in the region, despite the obvious potential given the wealth of material evidence, documentary sources and scope for the collection of oral histories. A specific focus of this comparative study was on how manipulation and transformation of the waterfronts of these pre-European settlements, such as through the addition of quays, seawalls, slipways, new harbours and customs houses, served to reinforce the new social distinctions and power relations that emerged within the context of European colonial rule (Rhodes 2010). Another noticeable feature was the increased militarisation of some of these settlements as the European powers themselves struggled for control over different territories and seaways, as is well documented around Mombasa Harbour—although this process began soon after the arrival of the Portuguese (McConkey and McErlean 2007) and also included Omani activity as well (see Lane 2005).
Conclusion: Future Prospects and Trends
The excavation of the Mombasa wreck in the 1970s was a promising start to underwater archaeology in the western Indian Ocean and southern RSZ. However, as this review has highlighted, despite the obvious potential for sustained archaeological surveys and even excavations, underwater and maritime archaeology in region was virtually suspended for the next two decades. There seem to be several reasons for this, including the widespread perception among archaeologists in the region that this field of archaeology has little to contribute to understanding the African past, coupled also by the obvious expense of underwater excavations and subsequent artefact conservation projects, and the lack of investment in local capacity building. To its credit, and despite other demands on a limited budget, the NMK has sustained the excellent marine conservation facilities that were set up at Fort Jesus Museum to deal with the finds from the Mombasa wreck. These represent a very real resource but are presently under utilised and without additional support maintaining the laboratory may eventually come to be regarded as a financial burden that can no longer be supported.
It is only in the last decade that maritime archaeology has been revived in the region. As reviewed above, this has been stimulated by projects that have taken a more holistic approach that extends well beyond an older, narrower focus on shipwrecks. Two, equally important, additional objectives of these recent projects have been to provide training and capacity building for local archaeologists and cultural heritage managers, and to raise the profile of maritime and underwater archaeology in the region. The success of this strategy can be gauged by the range of subsequent activity, the increased level of interest among archaeologists based in the region, and the beginning of integration of maritime and underwater archaeology into national archaeological services. This is important given the very long history of engagement with the sea by the diverse populations that have occupied the coastal littorals of the southern Red Sea and western Indian Ocean. This activity has prompted a change in perspective, also, on the part of local archaeologists, who have begun to recognise wrecks and other underwater remains (where they occur in their territorial waters) as being a part of their heritage assets. The launch of the UNESCO CPUCH in 2001 has also helped, although it is a matter of regret that at the time of writing none of the coastal states bordering either the southern Red Sea or the western Indian Ocean have ratified this convention.
Nonetheless, other international partners are becoming more actively involved in supporting the development of underwater archaeology in the region. For instance, since mid-2008 the Centre for International Heritage Activities in the Netherlands has been assisting the Antiquities departments in Tanzania, South Africa, and Mozambique to develop a programme in Maritime and Underwater Cultural Heritage (MUCH) (Centre for International Heritage Activities n.d.; see Sharfman et al., this volume). This has included providing practical training for selected personnel as divers, training in maritime and underwater archaeology and legislation, and conducting shoreline and underwater surveys around Kilwa Kisiwani, Songo Mnara, Kismania Mafia and Ilha de Moçambique. The possibility of developing some sites, including underwater ones, for heritage tourism is also being investigated. NMK have gone further, with the creation of a new position for a ‘marine archaeologist’ at Fort Jesus, and also now maintain a historical shipwreck inventory. As reported above, the governments of China and Kenya have also signed an agreement to jointly explore the Kenyan coast for wrecks and terrestrial remains associated with the voyages to East Africa led by Admiral Zheng He during the fifteenth century. The project involves collaboration between the NMK Coastal Archaeology Unit and the School of Archaeology and Museology at Peking University, Beijing. This research, with planned terrestrial investigations and underwater surveys at Malindi as well as further work around Lamu was announced via the international media in mid-2010 (Murphy 2010). New underwater surveys have also been conducted recently around the island of Reunion as part of a wider archaeological resources assessment project coordinated by Édouard Jacquot for the Institut national de recherches archéologiques préventives (INRAP), the French national organisation for preventive archaeology (INRAP 2012). The CMA also continues to act as an official advisor to the Tanzania Antiquities Department and NMK’s Coastal Archaeology Unit on the management of submerged cultural resources and also sustains its partnership with NCAM in Khartoum, Sudan (Colin Breen, pers. comm. 8/2010).
Alongside these explicitly maritime archaeology initiatives, there is growing interest among terrestrial archaeologists in the exploitation of marine resources during different periods, as well as in the possible impacts of sea-level change on coast-dwelling societies. For example, new research suggests that coastal routes were critical to the subsequent dispersal of early Anatomically Modern Humans (AMH) to other parts of the world, and their abilities to make deliberate open water crossings and exploit marine resources are now being assessed (e.g., Oppenheimer 2009). Thus far, most research has concentrated in the RSZ (e.g., Walter et al. 2000; Bailey et al. 2007a, b). This is perhaps because the sequence of marine terraces of Pleistocene origin that border the modern coastline are relatively well dated (Lambeck et al. 2011) in contrast to those found along the East African littoral. In the RSZ one of the most important is a former reef terrace lying between 6 and 10 m above current ASL, on which archaeological surveys have located frequent in situ Middle Stone Age (MSA) artefacts dating to around 125 kya (Walter et al. 2000; Buffler et al. 2010). While raised terraces with MSA artefacts are also known along the East African littoral (e.g., at Kilwa, see Chittick 1974, and around Mombasa, Oosterom 1988) the dating of these terraces is much less secure. Accordingly, one priority for future research would be to develop a much better dated sea-level curve for the region, which takes into account the complicating factor of local tectonic activity. More systematic and coordinated survey of the different marine terraces would also be beneficial, including those now submerged.
With reference to the early Holocene, most of the recent research has been conducted in the RSZ, especially Eritrea where surveys and excavations have revealed several early-middle Holocene human occupations, notably at the sites of Gelalo NW, Misse East (both occupied during the eighth millennium BC), and Asfet (occupied in the sixth millennium BC) (Beyin 2010, 2011; Bar-Yosef Mayer and Beyin 2009). It is suggested that occupation of coastal zones during the early Holocene, and the practice of gathering shell-fish that emerged as a consequence, may represent an adaptive response to increasing aridity in the interior as a result of regional and global climatic changes (Bar-Yosef Mayer and Beyin 2009). Sites of an equivalent date and nature further south have yet to be investigated in any detail, although shell-fish gathering was clearly an important subsistence practice for some pre-farming communities—such as those at Chibuene (Sinclair 1982) and around Sofala (Liesegang 1972) and elsewhere in southern Mozambique (Derricourt 1975). As well as being well-documented on the South African coast (Steele and Klein 2008; Jerardino 2010), Neville Chittick recovered an LSA long-blade industry associated with shell-fish, some of which showed clear signs of human modification, at Ras Hafun, Somaliland (Chittick 1976).
For later periods, more work has been conducted along the East African coast and nearby offshore islands (e.g., Chami 1999b; Helm 2000). Of particular current interest is the dating of the initial introduction of SE Asian domesticates, such as banana, taro and chicken. While this is not a new topic, it has been given particular emphasis by the Sealinks Project, led by Nicole Boivin and funded by the European Research Council. This also aims to resolve the many uncertainties concerning the timing of long-distance, trans-oceanic connections between different communities occupying the Indian Ocean rim in the last few millennia BC and the emergence of early seafaring (Sealinks 2008). An important aspect of this research has been the systematic sampling for palaeobotanical remains at coastal sites spanning the transition from hunting and gathering to food production in both Kenya (Helm et al. 2012) and Tanzania (Sealinks 2008, Project news). A related issue is when the close (Pemba, Unguja, Mafia) and more distant offshore islands (Madagascar, Comoros Islands, Seychelles) were first occupied and by whom (Beaujard 2005, 2011; Chami 2009). Here too, providing well dated sea-level curves for the East African coast and the offshore islands, coupled with systematic survey of the inter-tidal zones and appropriate sections of the seabed should be regarded as a priority for future work (see also Pollard 2009).
In summary, the past decade has helped lay the basis—in terms of awareness, capacity and intellectual frameworks—for more sustained and sustainable programmes of maritime and underwater archaeology in both regions. This heightened sensitivity has also encouraged greater awareness among national archaeological agencies of the need to extend impact assessments and mitigation work to include underwater contexts as well as terrestrial ones, as for example illustrated by the surveys in Tudor Creek around Fort Jesus in 2010 in advance of the laying of a fibre optic cable (Wanyama 2010). Previous, shipwreck-oriented research has also highlighted the likely potential of the underwater archaeology, although thus far focus has been exclusively on ships of European origin. While there is certainly scope for further investigation of these, and a pressing need to accord them better protection from commercial salvage and treasure hunters, a more integrated approach to the entire range of underwater and maritime archaeological resources is now needed (Breen and Lane 2003). This must include the compilation of national inventories of all known underwater archaeological sites; the introduction of a regular system for monitoring the more significant of these; the introduction of systems that facilitate the reporting of new finds made by sports divers, fishermen and other members of the public; public awareness campaigns targeted at different audiences ranging from local residents to tourists and government officials to sensitise them as to the historical and cultural value of such remains; a moratorium on all commercially oriented surveys for historic shipwrecks, irrespective of their origins; the adoption of legislation that enforces requirements for Archaeological Impact Assessments and if necessary mitigation work based on the ‘polluter pays’ principle; and the development of national and regional strategies for maritime and underwater archaeology which identify research priorities and identify the themes and geographical locations that have greatest potential. In all this, international partners will probably be needed to provide training, capacity building and even financial support, but it must be the locally based archaeologists who set the agendas.
I would like to thank the editor Athena Trakadas for her helpful comments and suggestions on an earlier draft of this paper, and Colin Breen for preparing the illustrations and his long-standing commitment to the promotion of maritime archaeology in eastern Africa and comments on earlier versions of this paper.