Geoarchaeological Investigations in the Apollonia-Region (Albania)

Michael E. Timpson

Geoarchaeological fieldwork in 1999 consisted of examination of the survey area surrounding the Classical to Roman city of Apollonia in central Albania, eight kilometers west of the industrial city of Fier. An additional visit was made to the archaeological site of Cakran and the surrounding region. The goals of this year’s investigations were to:

Travel to the project area required most of 4 days due to the political situation created by the hostilities in Kosovo. Although this reduced the time available for actual field work, considerable progress was achieved in 1999.


I began the process of familiarizing myself with the study region by walking the areas surveyed in 1998 in the company of the 1998 survey team leaders. This allowed me to examine the soils, geology, and landscapes while becoming familiar with the location and character of significant archaeological finds. The second step was to obtain copies of the Albanian Geological Service maps covering the entire study area. We obtained copies of these maps at a scale of 1:25, 000. These are detailed enough to provide good base maps for soil and geomorphological investigations and are the same scale as the topographic maps presently being used in the project GIS, thus they can easily be converted to a mapping layer in the GIS. Based on this field and map review, and the solid framework supplied by Dr. Eberhard Zangger in 1998 I will discuss the results and preliminary conclusions of the 1999 season.

General Setting and Physiography

The study area comprises the region stretching from the coast of the Adriatic Sea about thirty kilometers inland to the modern city of Ballsh and from there about twenty kilometers northward to the city of Fier, an area of about six hundred square kilometers. This region contains the western parts of the Mallakaster ridge bordered by the Myzeqe coastal plain in the north and west and the Vjose River flood plain in the south. The Mallakaster ridge rises from 104 meters above sea level at Apollonia in the west to 612 meters above sea level near Ballsh in the east. The valley of the Gjanices River parallels the eastern side of the ridge, running northwest from Ballsh to Fier (Zangger, 1998).

The entire area surveyed during archaeological fieldwork in 1998 and 1999 is underlain by Pliocene-aged nearshore sediments of the Rrogozhina Suite (composed of conglomerates, sandstones, and mudrocks), and undifferentiated Quaternary deposits. The Quaternary materials consist of shallow, valley-fill deposits, composed of local alluvium and colluvium, and fluvial deposits associated with the small streams that drain the local valleys. The Quaternary sediments have not been separated into Pleistocene and Holocene components. Most of the alluvium associated with local drainages is probably Holocene in age (i.e., less than 10, 000 years old). Much of the Myzeqe coastal plain also dates to the Holocene. In many locations, the Quaternary formations consist of relict paleosols (red beds) formed into the underlying Pliocene bedrock, and thus are not actual deposits, but rather Quaternary-aged surfaces. These surfaces may reflect as much as 100, 000 years of weathering and landscape genesis.

Soils and Soil Erosion in the Apollonia Region

The Kryegjata and Shtyllas valleys and associated ridges are covered by soils representing 4 soil orders, Alfisols, Entisols, Inceptisols, and Mollisols (USDA-NRCS, 1997 []). The Alfisols consist of brown- and red-colored soils (red beds) that contain pedogenic accumulations of clay in the subsoil. These soils occupy the most stable portions of the landscape in this area. The Inceptisols consist of brown and reddish-brown soils that exhibit minimal subsoil development (i.e., weak B-horizon development) They typically occupy less stable landscape positions (e.g., moderately steep ridge slopes and valley footslopes) where natural processes of erosion and sedimentation regularly provide fresh parent material for soil development. The Entisols consist of very young (recent) soils that exhibit only minimal soil development (typically A and C horizons) They are located in areas associated with major and minor flood plains where frequent additions of new flood deposits continually "reset" the soil development "clock." Mollisols in the study area occur primarily in the Myzeqe coastal plain. These soils exhibit thick, dark surface horizons characteristic of high amounts of organic matter. The formerly poorly-drained conditions in the coastal plain combined with the vegetation common prior to agricultural drainage (grasses, reeds, and sedges) have produced these soils.

Natural and anthropogenically enhanced soil erosion in the Apollonia region is minimal. The soils appear to be relatively intact, thus artifact distributions and the archaeological record should be well preserved. This is not to say that some movement of artifacts hasn’t occurred. Some of the Paleolithic materials at Kryegjata A and Kryegjata C have obviously been displaced, but compared with the marly regions in Greek Messenia, their context has not been greatly altered. One exception to this generalization is the southern side of the Shtyllas-Jaroi valley. Here, the abandoned terrace slopes are beginning to exhibit the affects of erosional processes. It should be interesting to contrast the artifact distributions on opposite sides of this valley and to compare the southern slope of this valley with the southern slopes of adjacent valleys.

Site Contexts of 1998 and 1999 Survey Areas

During the process of reviewing the general nature of each area surveyed in 1998 I paid extra attention to a few of the designated sites identified last year. Particular emphasis was placed on those sites that contained Paleolithic components.

Site 003 (Kryegjata B) contained the greatest abundance of middle and late Paleolithic artifacts identified in 1998. The site is situated on a low hill in an area underlain by Quaternary sediments and Pliocene Rrogozhina Suite deposits. This low hill is a stable upland position and aside from disturbances resulting from road construction and general agriculture, the site has suffered little damage. The soil underlying the site is a thick, well-developed alfisol consisting of approximately 2 meters of red, sandy loam and sandy clay loam overlying calcareous conglomerate and fine sand deposits. Based on the degree of soil development, I estimate the age of this surface at between 50, 000 and 80, 000 years. This age range agrees well with the Paleolithic materials identified from the site. My initial impression is that artifacts here may extend from the modern surface to 0.5 m or slightly more. Additional description and sampling will need to be undertaken at this site during the 2000 field season to verify the preliminary geologic and pedologic conclusions. This will be done in conjunction with the test excavations planned for this site. Site 017 (Kryegjata A) occupies a landscape position similar to Kryegjata B. Disturbance from road construction at Kryegjata A is more extensive than at site 003. Because of time constraints in 1999 I didn’t spend sufficient time at Site 017 to fully assess the context of this site. It appears to be underlain by a similar red alfisol that is also associated with a calcareous conglomerate of Pliocene age. Additional examination of this site will be undertaken in 2000.

During our review of 1998 survey areas an additional Paleolithic locality was identified (now designated site 024, Kryegjata C). Site 024 is located along a small field road that runs northeast into the village of Radostina. It lies at a slightly higher elevation than nearby site 003 (Kryegjata B), but appears to be underlain by similar soils and geologic deposits. Additional soil coring and sampling will need to be conducted here next year to test this hypothesis. Further down slope in this same area (at an abandoned military installation near Radostina) there was a quarry exposure of the red soil and geologic deposits that underlie sites 003 and 024. Provided this exposure is not destroyed by additional quarry activity, I plan to describe and sample this profile in 2000 for comparison with auger core and excavation samples from the adjacent Paleolithic sites.

The area surveyed by Team A in 1998 contained little Paleolithic material and no Paleolithic sites were designated in this area. The area surrounding Site 010 (a largely Hellenistic site) contains an intact wetland plant community. This vegetation is consistent with the verbal reports that a spring used to issue here. Shallow excavations, designed to provide waterholes for grazing stock, and the presence of extant hydrophytic vegetation provide evidence of the continued presence of shallow groundwater in this location. Soils associated with wetlands exhibit unique morphological properties related to their formation under conditions of constant saturation. These distinct properties are identifiable even after the soils dry out as a result of natural or anthropogenic causes. As a result, it should be possible, through the use of soil coring transects, to estimate the historic extent of this wetland, and therefore the historic usefulness of this area as a reliable water source. Mapping the extent of this spring area is planned for the 2000 field season.

During the survey activities carried out by Team C in 1999, a number of Middle Paleolithic artifacts were recovered from Tract C-011. This tract is located at mid-slope on the western end of Mt. Shtyllas. The area is underlain by 0.6 to 1 meters of brown-colored colluvial slope wash or debris flow deposits into which the current surface soil has formed. Below this is an older red-colored soil similar to those identified at Sites 003, 017, and 024. Soil development in the upper material suggests 10,000 to 20,000 years of relative slope stability. The lower, red-colored soil suggests 50,000 to 80,000 years of soil development. Paleolithic artifacts in this tract were found near the interface of these two sediment units and within the gully formed from erosion of an abandoned field road that bisects the tract. Additional artifacts were recovered from the modern surface in an area in which the modern brown soil has been eroded slightly due to traffic from livestock and carts. More thorough examination of this locality is planned for the 2000 field season to determine the exact nature of the contact between these two deposits, obtain more detailed soil descriptions and samples, and to verify the source and location of artifacts that may still be contained in the deposits.

In addition to the work conducted in the immediate vicinity of Apollonia, I also visited the archaeological site of Cakran and the surrounding area. This trip allowed me to examine the geology and soils between Apollonia and Ballsh and encompassed portions of the Gjanice and Vjose River valleys. As indicated by Zangger (1998), the Mallakaster ridge west of Cakran consists of Pliocene-aged, shallow marine sediments including marl, silt- and sandstone, and conglomerates. Between Cakran and Ballsh, the ridge is dominated by coarse-grained Oligocene deposits and fine-grained, marly or clayey Lower Miocene flysch deposits. All of these bedrock units are extremely soft and prone to erosion. As a result, the soils in this area are more susceptible to erosion and 20 to 30% of the slopes in this area may be moderately to severely affected by erosional disturbance. Conversely, the footslopes and valleys will be impacted by sedimentation. Therefore, archaeological survey in some of these areas will need to be interpreted in light of these disturbances.

The valley in which Cakran is located is underlain by Pleistocene or older valley-fill deposits from the surrounding ridges and local alluvium (probably Holocene-aged) derived from the small stream that flows through the site area. The alluvium consists of about 2 meters of fine-grained, overbank sediments that overly gravelly channel lag deposits. The stream has incised into the channel lag deposits by up to 0.4 meters. This incision appears to be a recent (6 100 years) phenomenon, probably associated with increased runoff from the surrounding uplands. Dr. Korkuti informed me that the upper 20 to 40 cm of the site were disturbed by modern agricultural practices. In spite of this, Mike Galaty and I observed a wide assortment of ceramic and lithic artifacts scattered across the site area.

The Gjanice River occupies a relatively narrow valley along its course from Fier to Ballsh. The river meanders between the Mallakaster ridge on the western side of the valley and a lower ridge on the eastern side. The modern channel appears to consist of gravel and cobble deposits, however, the flood plain and terrace system were being utilized for agricultural purposes, suggesting they consist of finer-grained overbank deposits. Most of the surrounding uplands did not appear to be too seriously impacted by erosion, however, there are some locations within the segment of the valley I examined that have been severely affected by erosion and sedimentation. The well-preserved system of terraces and narrow valley of the Gjanice River should provide an excellent opportunity for reconstructing the alluvial chronology and settlement history of the portion of this river valley that lies within the project area.

The Vjose River is larger than the Gjanice and its valley is characterized by a much wider flood plain. The present active flood plain consists of a bed of gravel- to cobble-sized material that is likely a relict of an older, higher energy flood regime than the river currently supports. Infrequent, high-energy floods probably still occur, but recent deposits range from medium and fine sands to silts. The flood plain and riverbed undoubtedly served as a source of cobbles for the manufacture of stone tools. Mike Galaty and I recovered a number of flakable lithic materials from the flood plain during our reconnaissance visit. The quality of the stones we recovered was good, but the cobbles were generally too small to be easily worked. It should be noted however, that we did not make an exhaustive search for these materials and only examined a small (ca. 1 hectare) area. The valley is too large to be included in detailed geomorphological and soils studies during the course of the project as presently proposed.

Proposed Geoscience Research for the Future

Year 2000

Beyond 2000



Zangger, E. 1998. Geoarchaeological Reconnaissance of the Apollonia-Region (Albania) 24.-31. May 1998. Report to The Mallakastra Regional Archaeological Project. Manuscript on file, Dept. of Classics, University of Cincinnati.