Albania Survey Project: 2007 Field Season
In July of 2007 the first large-scale survey of the Albanian coast commenced, a project undertaken by RPM Nautical Foundation (RPMNF) in conjunction with Drs. Adrian Anastasi and Neritan Ceka of the Albanian Institute of Archaeology (AIA) and Auron Tare, Director of the Albanian National Trust. This project was also made possible by the cooperation of the Albanian Ministry of Defense and Ministry of Culture and Tourism. The survey began at Albania’s southern border with Greece with a goal to complete a multibeam survey of the seafloor between the 25- and 100-m contour; this initial stage reached the area around Sarande.
In addition to remote sensing dive survey along the shoreline was also conducted; these relegated to the 5-30-m contours. A total of 15 wrecksites and other finds were discovered during operations; which are discussed here. Multibeam survey was conducted by a RPMNF’s survey vessel, the 33-m R/V Hercules. Among the remote sensing equipment on this vessel there is a hull-mounted multibeam echosounder, model type EM3002D, from Kongsberg Maritime division. After multibeam data was collected, processed, and cleaned of aberrant readings, it was reviewed as a three-dimensional model of the seafloor. Software applications allowed the visualization and manipulation of each individual beam reading in order to attain the best possible seafloor image and the identification of anomalies. Anomalies were assessed for their association with surrounding geologic formations or products of recent human disturbance. Those meeting this later criterion were marked on a planning map and plotted with navigation and spatial software. Once plotted it was possible to navigate on an electronic chart both the vessel and ROV (equipped with a transponder) to each of the anomaly locations.
Once the research vessel was positioned over an anomaly location, the ROV was deployed. Anomalies were verified primarily with the ROV except for a few in shallow areas, c. 25 m of depth, where divers were utilized. A forward-scanning sonar affixed to the ROV facilitated the location of each anomaly and aided in the examination of the area surrounding each anomaly for scattered objects. For example, the power and frequency of the sonar allows the detection of a single amphora at a distance of 40 m. After it was determined that an anomaly was located, and a site was present, recording ensued with the use of a video camera. When sites were located, objects were raised when it was deemed necessary to assist in determining date and nature of the site. When geologic formations were encountered, they were scanned for cultural material that often is trapped when drag nets or currents push them onto rocks. Although such displaced material does not constitute a site in itself, a distribution map of these finds can assist in locating sites.
2007 Field Season Results
During multibeam survey 127 anomalies were identified in the multibeam data of which 67 were verified; time constraints prevented the completion of verification this season. These anomalies were distributed throughout the survey area, roughly grouped in the large bay S of Sarande, off a point at the south end of the aforementioned bay in front of the small islands, and at the mouth of the Butrint River. During the verification of anomalies in the bay S of Sarande 12 modern wrecksites were located. These wrecks were all exposed to some degree, extending out of the mud, but were obviously undergoing rapid burial. Based on the number of modern wrecks in this bay, wrecks from the medieval and ancient eras undoubtedly lie beneath this sediment deposit. Also in this area, there was a late medieval vase lying loose on the seafloor and numerous Roman-era amphora fragments near the coast. It is clear that the sedimentation rate that results from the river’s outflow has buried any wrecksite from the ancient to medieval periods. Further investigation of this area, as well as the area in the bay south of Sarande, will require the use of magnetometers and/or sub-bottom profilers to detect buried wrecksites.
One of the modern wrecks discovered during the 2007 survey season. Based on the number of modern wrecks in this bay, wrecks from the medieval and ancient eras undoubtedly also lie beneath this sediment deposit.
Ottoman Pottery Deposit
On the suggestion of Auron Tare who had dived in the area of the three islands previously, a dive investigation around the islands was conducted. A brief dive survey on the southernmost side of the outermost island noted a large pottery deposit, primarily of Ottoman-period ceramics. Dr. Anastasi collected samples for analysis which were taken to the AIA office for conservation. Prior to conservation, the sherds will be analyzed by Katie Johnson, a graduate student at the Univ. of Chicago who is documenting and studying Ottoman material culture.
Finds at Gjri Para Kakomese Bay
A small bay is situated at the northern portion of the survey area, just N of Sarande, which was investigated by divers over a three day period. The bay is tucked behind a promontory that hides its presence from a vantage point at sea and offers exceptional protection from weather for ships moored here. The seafloor within the bay is covered with a layer of sand and has large patches of sea grass, apparently posidonia. There are also areas of rock outcrops and loose stone much like the rugged hillside surrounding the bay.
During searches along both the northern and southern sides of the bay, divers encountered a wide scatter of pottery. These pottery finds included both amphoras and tableware, many intact, that are located in the shallow portions of the bay (10-20 m of depth). Based upon this initial investigation, the artifact deposit does not appear to form a shipwreck site as there is no discernable artifact concentration. Furthermore, the dispersion of the pottery is along both sides of the bay; further investigation of the deeper portions of the bay with the ROV will be undertaken in 2008. Two types of amphoras were present among the pottery. One distinct type is a pyriform-shaped amphora with a moderately constricting neck. Its arched handles rise from a rounded shoulder high above the mouth and fall to join just below the rim.
At the junction with the neck, the thick handles merge with neck and displace the rim on each side. The handles vary from round to ovoid in cross-section over their length. The body constricts near the rounded bottom and has tightly-spaced ridges along the body. The second type of amphora features small arched handles that rise from the shoulder to above the mouth and fall to join the short neck below the rim. Its handles are ovoid in cross-section with a slight ridge running along their outer side. The rime appears to be a slight beaded one, but only slightly pronounced. The body gradually tapers towards a flat base. Good parallels for both types of amphoras were found on the Çamalti Burnu wrecksite from the NW coast of Marmara Island, Turkey, that is dated to the 13th century CE. The first type of amphoras are clearly a Gunsinen type 3 amphora, while the flat based amphoras are small varieties of the Gunsinen type 4 amphoras. A small, intact table pitcher was located among these amphora finds; it also has parallels with those found on the Çamalti Burnu site. Hence, he pitcher is likely associated with the amphoras, with the overall ceramic deposit having a tight chronological period probably in the early 13th century CE.
Another set of artifacts located among the ceramic finds were stone balls; these were present on both the southern and northern shores of the bay. Many of stone balls were loose in the sand, while others were adhered to the sandstone rock that forms outcrops in the bay. These stone balls were regular in shape but not homogeneous in their dimensions; however, the majority are between 12-18 cm in diameter. A possible identification of the stone balls is their being shot for a type of mangonel such as the Turkish manjaniq or Byzantine petrabolos. These artillery pieces were human-torsion powered devices that made their entrance into the Byzantine world from further east in the 12th/13th century CE. A pre-cursor to counter-weight artially, these machines were effective devices for a siege or defense of fortifications, and were also incorporated on board ships. The defense against ships, fired wither from fortifications or other vessels, such mangonels were recorded as effective in damaging wooden ships. Hence, the presence of stone mangonel shot is consistent with the date and context of the pottery. Further analysis of the stone will be necessary to determine its origin, and address their presence in the bay.
Good parallels for both types of amphoras were found on the Çamalti Burnu wrecksite from the NW coast of Marmara Island, Turkey, that is dated to the 13th century CE. The first type of amphoras are clearly a Gunsinen type 3 amphora, while the flat based amphoras are small varieties of the Gunsinen type 4 amphoras. A small, intact table pitcher was located among these amphora finds; it also has parallels with those found on the Çamalti Burnu site. Hence, he pitcher is likely associated with the amphoras, with the overall ceramic deposit having a tight chronological period probably in the early 13th century CE.
Site AB-7-AO The South Butrint I Wrecksite
Dive survey along the southern portion of the survey area resulted in the discovery of a wrecksite on a steep slope. The seafloor on this slope is features many rock outcrops with pockets of sands inter-dispersed among them. Many of these pockets are several meters in width; it is within these pockets that amphora fragments and intact vessels are found. The site was scattered amongst the sandy deposits and many rock outcrops, and spread out over an area of approximately 50 m at its deepest section. The site narrowed along its midline as it progressed up the slope. These deposits were amphoras, many whole or in large fragments, grouped on the surface and/or buried in the sand and hand-fanning revealed groups of intact amphoras below the surface. One of these intact examples was raised for analysis.
Material Deposits at the S. Butrint I Wrecksite
There are also smaller Corinthian variants, between 35-45 cm in height, that have a very similar features to the larger Type B forms except that there bodies are much shorter. A lesser number of Corinthian Type A amphoras are also present on the site, possessing a bulbous body, curved handles, and slightly everted rim. These Corinthian types comprise the majority of the examples found on the site and are undoubtedly the cargo of the vessel. Considering the nature of this site, the extensive individual finds, and the state of deeper material, the submerged cultural material along the Albanian coast is relatively untouched compared to the majority of Mediterranean countries.
Corinthian B Amphora Raised
The South Butrint wrecksite is, therefore, not likely to be a unique find during the program of survey. From an academic interest, such sites offer a wealth of information to the questions of overseas trade involving such ancient cities as Butrint, Apollonia, and Dyrrachium, as well as the Illyrian and Epiriote states.