Fig 1 - uploaded by Jonathan Benjamin
Content may be subject to copyright.
A topographic map of the Mediterranean Sea region with bathymetric data derived from the European Marine Observation and Data Network (http://www.emodnet.eu/). Topographic data derived from Shuttle RADAR Topographic Mission (SRTM, srtm.csi.cgiar.org). Key sites as mentioned in text. (credit: A. Fontana).
Source publication
This article reviews and discusses key data, literature, debates and discussions focussed on relative sea-level change since the Last Interglacial (approximately last ~132,000 years) in the Mediterranean Basin. Special reference is given to the geomorphological (physical and environmental) and archaeological (human and cultural) aspects of central...
Contexts in source publication
Context 1
... studies have reconstructed past sea levels from rock- cut coastal fish tanks (Fig. 12) in Italy (Dreghorn, 1981;Auriemma and Solinas, 2009;Evelpidou et al., 2012b), Greece ( Kolaiti and Mourtzas, 2016), Israel ( Galili and Sharvit, 1998;Toker et al., 2012 with sluice gates (cataracta) that controlled the water exchange between the tanks and the open sea, preventing the fish from escaping. Water exchange took place through multiple channels, sometimes carved in bedrock. A breakwater is often built around the fish tank to protect the inner basin from sea waves. The latter are often delimited by foot walks (crepido), generally occurring at two or three levels that were not recognised or interpreted in earlier studies. These levels, together with the sluice gates, are key in interpretations of the position of former sea level in relation to the fish tanks. Lambeck et al. (2004b) demonstrated that the top of the sluice gate corresponds to the elevation of the lowest level foot-walk (crepido) (Fig. 12). According to the Latin treatise De Re Rustica XIII (Columella, early 1st century AD), the crepido should lie above the highest tidal level, as also reported in the description by Pliny the Elder (23e79 A.D., in Naturalis Historia) in a constructional part that looks at the water (marginum eam partem, quae aquas spectat). Using sites with complete preservation of channels, sluice gates and foot-walks, Lambeck et al. (2004b) estimated that the palaeo high tide in Roman time was about 0.2 m below the lowest crepido. Further, Lambeck et al. (2004b), and subsequently Auriemma and Solinas (2009) and Mourtzas (2012a, b), suggested that the flow of water inside the fish tanks was tidally controlled and that the palaeo mean sea level was placed at the middle of the sluice gate, while mean low tide was denoted by the channel thresholds, often corresponding to the base of the mobile cataracta. Past RSL was then constrained by these structural features of the fishtanks (Fig. 12). Such interpretation was also applied at a number of sites throughout the Mediterranean (e.g. Antonioli et al., 2007;Anzidei et al., 2011aAnzidei et al., , 2014a. Evelpidou et al. (2012b) also performed a detailed survey of some Tyrrhenian Sea fish tanks, previously observed by Schmiedt (1972) and Lambeck et al. (2004b) and proposed that RSL during the Roman period ranged between ~ À0.6 and ~À0.3 m. Evelpidou et al. (2012b) disagreed with the interpretation of an original supratidal position of the lowest crepido proposed by Lambeck et al. (2004b). Further, they stated that the height of the cataracta pro- posed by Lambeck et al. (2004b) would not be sufficient for the fish tanks to function properly. They suggested that the upper part of the cataracta corresponds to the upper crepido instead of the lower one. Evelpidou et al. (2012b) assumed that the top of the upper crepido was high enough to prevent the fish tank from flooding by sea surges, which could lead to a loss of fish. Pirazzoli and Evelpidou (2013) stated that the heights of the channel threshold and the base of the cataracta can vary and they argued, therefore, that this structural feature was too unreliable to precisely recon- struct the palaeo ...
Context 2
... studies have reconstructed past sea levels from rock- cut coastal fish tanks (Fig. 12) in Italy (Dreghorn, 1981;Auriemma and Solinas, 2009;Evelpidou et al., 2012b), Greece ( Kolaiti and Mourtzas, 2016), Israel ( Galili and Sharvit, 1998;Toker et al., 2012 with sluice gates (cataracta) that controlled the water exchange between the tanks and the open sea, preventing the fish from escaping. Water exchange took place through multiple channels, sometimes carved in bedrock. A breakwater is often built around the fish tank to protect the inner basin from sea waves. The latter are often delimited by foot walks (crepido), generally occurring at two or three levels that were not recognised or interpreted in earlier studies. These levels, together with the sluice gates, are key in interpretations of the position of former sea level in relation to the fish tanks. Lambeck et al. (2004b) demonstrated that the top of the sluice gate corresponds to the elevation of the lowest level foot-walk (crepido) (Fig. 12). According to the Latin treatise De Re Rustica XIII (Columella, early 1st century AD), the crepido should lie above the highest tidal level, as also reported in the description by Pliny the Elder (23e79 A.D., in Naturalis Historia) in a constructional part that looks at the water (marginum eam partem, quae aquas spectat). Using sites with complete preservation of channels, sluice gates and foot-walks, Lambeck et al. (2004b) estimated that the palaeo high tide in Roman time was about 0.2 m below the lowest crepido. Further, Lambeck et al. (2004b), and subsequently Auriemma and Solinas (2009) and Mourtzas (2012a, b), suggested that the flow of water inside the fish tanks was tidally controlled and that the palaeo mean sea level was placed at the middle of the sluice gate, while mean low tide was denoted by the channel thresholds, often corresponding to the base of the mobile cataracta. Past RSL was then constrained by these structural features of the fishtanks (Fig. 12). Such interpretation was also applied at a number of sites throughout the Mediterranean (e.g. Antonioli et al., 2007;Anzidei et al., 2011aAnzidei et al., , 2014a. Evelpidou et al. (2012b) also performed a detailed survey of some Tyrrhenian Sea fish tanks, previously observed by Schmiedt (1972) and Lambeck et al. (2004b) and proposed that RSL during the Roman period ranged between ~ À0.6 and ~À0.3 m. Evelpidou et al. (2012b) disagreed with the interpretation of an original supratidal position of the lowest crepido proposed by Lambeck et al. (2004b). Further, they stated that the height of the cataracta pro- posed by Lambeck et al. (2004b) would not be sufficient for the fish tanks to function properly. They suggested that the upper part of the cataracta corresponds to the upper crepido instead of the lower one. Evelpidou et al. (2012b) assumed that the top of the upper crepido was high enough to prevent the fish tank from flooding by sea surges, which could lead to a loss of fish. Pirazzoli and Evelpidou (2013) stated that the heights of the channel threshold and the base of the cataracta can vary and they argued, therefore, that this structural feature was too unreliable to precisely recon- struct the palaeo ...
Context 3
... studies have reconstructed past sea levels from rock- cut coastal fish tanks (Fig. 12) in Italy (Dreghorn, 1981;Auriemma and Solinas, 2009;Evelpidou et al., 2012b), Greece ( Kolaiti and Mourtzas, 2016), Israel ( Galili and Sharvit, 1998;Toker et al., 2012 with sluice gates (cataracta) that controlled the water exchange between the tanks and the open sea, preventing the fish from escaping. Water exchange took place through multiple channels, sometimes carved in bedrock. A breakwater is often built around the fish tank to protect the inner basin from sea waves. The latter are often delimited by foot walks (crepido), generally occurring at two or three levels that were not recognised or interpreted in earlier studies. These levels, together with the sluice gates, are key in interpretations of the position of former sea level in relation to the fish tanks. Lambeck et al. (2004b) demonstrated that the top of the sluice gate corresponds to the elevation of the lowest level foot-walk (crepido) (Fig. 12). According to the Latin treatise De Re Rustica XIII (Columella, early 1st century AD), the crepido should lie above the highest tidal level, as also reported in the description by Pliny the Elder (23e79 A.D., in Naturalis Historia) in a constructional part that looks at the water (marginum eam partem, quae aquas spectat). Using sites with complete preservation of channels, sluice gates and foot-walks, Lambeck et al. (2004b) estimated that the palaeo high tide in Roman time was about 0.2 m below the lowest crepido. Further, Lambeck et al. (2004b), and subsequently Auriemma and Solinas (2009) and Mourtzas (2012a, b), suggested that the flow of water inside the fish tanks was tidally controlled and that the palaeo mean sea level was placed at the middle of the sluice gate, while mean low tide was denoted by the channel thresholds, often corresponding to the base of the mobile cataracta. Past RSL was then constrained by these structural features of the fishtanks (Fig. 12). Such interpretation was also applied at a number of sites throughout the Mediterranean (e.g. Antonioli et al., 2007;Anzidei et al., 2011aAnzidei et al., , 2014a. Evelpidou et al. (2012b) also performed a detailed survey of some Tyrrhenian Sea fish tanks, previously observed by Schmiedt (1972) and Lambeck et al. (2004b) and proposed that RSL during the Roman period ranged between ~ À0.6 and ~À0.3 m. Evelpidou et al. (2012b) disagreed with the interpretation of an original supratidal position of the lowest crepido proposed by Lambeck et al. (2004b). Further, they stated that the height of the cataracta pro- posed by Lambeck et al. (2004b) would not be sufficient for the fish tanks to function properly. They suggested that the upper part of the cataracta corresponds to the upper crepido instead of the lower one. Evelpidou et al. (2012b) assumed that the top of the upper crepido was high enough to prevent the fish tank from flooding by sea surges, which could lead to a loss of fish. Pirazzoli and Evelpidou (2013) stated that the heights of the channel threshold and the base of the cataracta can vary and they argued, therefore, that this structural feature was too unreliable to precisely recon- struct the palaeo ...
Context 4
... sea-level curves have been constructed in Italy (e.g., Lambeck et al., 2004aLambeck et al., , 2011), Croatia and Slovenia (e.g., Antonioli et al., 2007;Faivre et al., 2013), southern France and Corsica (e.g., Laborel and Laborel-Deguen, 1994;Vacchi et al., 2016a); Turkey (Anzidei et al., 2011a), Greece (e.g., Pirazzoli, 2005;V€ ott, 2007;Pavlopoulos et al., 2011;Vacchi et al., 2014;Kolaiti and Mourtzas, 2016;, Tunisia and Libya (Anzidei et al., 2011b), the Aeolian Islands ( Anzidei et al., 2014aAnzidei et al., , 2016a, b), Israel ( Sivan et al., 2001Sivan et al., , 2004Toker et al., 2012;Galili et al., 1988Galili et al., , 2005) and Lebanon ( Morhange et al., 2006;Sivan et al., 2010). Data collected from tectonically stable regions, some characterised by negligible isostatic effects ( Sivan et al., 2001Sivan et al., , 2004Toker et al., 2012) for the last 4000 years, indicate that sea level was close to present levels by 4000e3600 BP ( Galili et al., 2005;Galili and Sharvit, 1998;Porat et al., 2008). Depending on the location in the Mediterra- nean, RSL fluctuated either below or slightly above the present since that time ( Sivan et al., 2004;Toker et al., 2012;Vacchi et al., 2016a, b). As an example, RSL along the coastlines of Israel rose from À7 m to the present level at a rate of 2.5e3.5 mm/yr between 6800 and 4000 BP. At the same location, RSL was approximately between À2.5 m and À5 m during the Chalcolithic period (6000e5700 BP). By the Middle Bronze Age (~4000 BP) the sea had reached its present level and the coastline reached its current form. Since then, RSL has been relatively stable with possible fluctuations of no more than 0.5 m vertically ( Galili et al., 2005;Sivan et al., 2001Sivan et al., , 2004Anzidei et al., 2011a). A recent notch study by Goodman-Tchernov and Katz (2015) does however question this stability and theorises that a more punctuated rise may have occurred during the ...
Context 5
... the Holocene sea-level rise had slowed down, various areas throughout the eastern Mediterranean continued to undergo significant regional and micro landscape changes owing to adjustment in land level caused by tectonic activity. The Bronze Age site at Pavlopetri (Greece, #27 in Fig. 1), was first investigated in the 1960s for its archaeological implications and for its contri- bution to sea-level studies in the Peloponnese (Flemming, 1978) and later revisited and systematically mapped by Henderson et al. (2011). The changes in landscape around similar coastlines would certainly not have gone unnoticed by local populations and oral traditions are likely to persist in the region's modern collective ...
Context 6
... is also evidence that interactions with coastlines during MIS 5 were not limited to consumption of resources. Shell tools were manufactured by Neanderthals during MIS 5.1 in Mediter- ranean cave sites such as Grotta del Cavallo (#8 in Fig. 1, Romagnoli et al., 2015). That technology persisted until ~50 ka (Douka and Spinapolice, 2012). Perforated shell beads from early MIS 5 cave contexts associated with H. sapiens dated to 100e135 ka (MIS 5.5e5.3) are known from Skhul (e.g. Vanhaeren et al., 2006) as well as Qafzeh where, dated to ~90e100 ka (MIS 5.3), they were stained with pigment ( Bar-Yosef Mayer et al., 2009). Both cave sites were ca. 35 km from the coastline, indicating specific transport of personal ornaments away from the coast ( Bar-Yosef Mayer et al., 2009). Similar beads have also been found in cave sites in the Maghreb around 82 ka, again linked to H. sapiens populations ( Bouzouggar et al., 2007;D'Errico et al., 2009). There are no known symbolic uses of marine shells by Neanderthals until MIS 4 (see below). In North Africa these or- naments seem to disappear from the archaeological record after 70 ka, remaining absent until around 50e40 ka, indicating that they may have been part of specific symbolic adaptations, the disappearance of which may have marked a cultural disconti- nuity (D'Errico et al., ...
Context 7
... Protohistoric coastal structures of the Mediterranean Basin and their archaeological signatures, have suffered from the devel- opment of later societies: many sites were destroyed, incorporated or generally transformed the pre-existing archaeological evidence. In particular, widespread diffusion and large stone construction (especially during the Classical periods) have resulted in the loss of the protohistoric features, through human reuse and recycling of materials. Thus, information on sea levels in the Bronze and Iron Age based on coastal settlements' remains is not as well docu- mented as in later periods and is often related to a few selected sites, as in the case of the harbour of Marseille ( Morhange et al., 1996Morhange et al., , 2001), the Northern Cyclades, the central and eastern Crete (Mourtzas and Kolaiti, 2016;, or along the coast of modern day Israel ( Sivan et al., 2001). The rare finding of a protohistoric vessel that would have entered a harbour might offer an extraordinary window on maritime life and seafarers ways (e.g. Uluburun in Turkey, #25 in Fig. 1; Bass et al., 1984), however shipwrecks of the open seas do not provide good indication of past sea levels. Conversely, relicts of beached or abandoned vessels (particularly vernacular vessels used for every day short-range activity), which can be confidently determined to have been left at or near sea level at the time, may contribute information related to sea level and ...
Context 8
... the importance and use of marine resources in open air sites remains difficult. In the Levant, activity in coastal environ- ments during early MIS 5 is shown in the form of MP stone in- dustries in several MIS 5.5 beach deposits on the Israeli coast ( Galili et al., 2007Galili et al., , 2017Ronen et al., 2008), where assemblages con- taining molluscs, animal bones and Middle Palaeolithic flint im- plements were recovered from such deposits on the Carmel coast at Nahal Bir Ibdawiya and Nahal Me'arot ( Fig. 6; #6 in Fig. 1) and on the Galilee coast. Whilst the molluscs in these beach deposits may be naturally occurring, the deposition of stone tools and animal bones within them highlights an occupation and use of the wider coastal landscape. It appears that whilst marine resources were exploited by both H. sapiens and Neanderthal populations, these economies likely used marine molluscs as part of a range of re- sources available during this period of relative environmental sta- bility and high sea levels around the ...
Context 9
... the eastern Mediterranean, the depositional sequences of deep waters are characterised by the quasi-cyclical occurrence of dark layers, rich in organic carbon, called 'sapropels' (Negri et al., 2012;Rohling et al., 2015;Grant et al., 2016 and reference therein). They correspond to hypoxic or anoxic episodes that are recorded east of the Sicily Strait ( Fig. 1) and during which oxygen starvation occurred in deep basins and caused the collapse of the deep ecosystems, but affected the entire water column (e.g. Cramp and O'Sullivan, 1999). The most recent sapropel has a Holocene age (Fig. 7), while another 4 sapropels are documented during the Late Pleistocene (Fig. 4). In different parts of the Mediterranean they can display some noticible differences in their age limits and duration (De Lang et al., 2008;Grant et al., 2016). The causes that led to the sapropel formations are still a question of debate, but their depo- sition was influenced by astronomical forces and generally corre- spond to periods of enhanced monsoon rainfall (e.g. RossignolStrick et al., 1982;Grant et al., 2016). In the late Quaternary, the sapropels generally caused notable sedimentation during periods of major fresh-water input, probably in connection with enhanced discharge of the Nile River linked with monsoon activity (e.g. Rohling et al., ...
Context 10
... the Middle and Late Bronze Ages (ca. 3600e3100 BP) of northwestern Adriatic, society flourished along the rims of the la- goons of Venice, Carole and Grado-Marano. Villages developed on slightly elevated fluvial ridges entering in the lagoon, and also on salt marshes. Several archaeological structures constrain sea level to À3.0 ± 0.6 m around 4000 BP and À2.0 ± 0.6 m at 3000 BP ( Fontana et al., 2017). This symbiotic relationship between lagoon and dwelling sites existed in the area also in the Iron Age, as clearly depicted in the 1st century BC by the geographer Strabo in his description of the cities of the Venetian people, that "stand in the midst of water like islands, others are only partially surrounded. Such as lie above the marshes in the interior are situated on rivers navigable for a surprising distance" (Strabo, Geografia, V, 1, 5). Thus, it seems that the settlement system of the Bronze Age represents the early evidence of an Adriatic culture strongly related to the brackish environments. This was later developed in the same area during the Iron Age and the Roman period with the harbour cities of Aquileia, Concordia Sagittaria, Altinum, Adria, Spina and Ravenna (#22 in Fig. 1), and later, during the early Middle Age, by Venice and its ...
Context 11
... Zambratija Bay site, in northern Croatia, (#17 in Fig. 1), which remains to be explored in detail, represents a submerged settlement in the northern Adriatic (Benjamin et al., 2011, Fig 16.4). It is still unclear as to why the site was abandoned, though early indications do not exclude sea-level rise as a direct cause. The site represents an important opportunity in this respect, and further, detailed study will be required to resolve the abandonment ques- tion ( Benjamin and Bonsall, 2009;Benjamin, 2010). The Grotta Verde in Sardinia (Italy, #24 in Fig. 1) has yielded submerged archaeological material in the form of cardial ceramics at À10 m depth and human remains at À8 m ( Antonioli et al., 1996) in what appears to be a submerged grave, dated to approximately 7300 BP (Fig. ...
Context 12
... Zambratija Bay site, in northern Croatia, (#17 in Fig. 1), which remains to be explored in detail, represents a submerged settlement in the northern Adriatic (Benjamin et al., 2011, Fig 16.4). It is still unclear as to why the site was abandoned, though early indications do not exclude sea-level rise as a direct cause. The site represents an important opportunity in this respect, and further, detailed study will be required to resolve the abandonment ques- tion ( Benjamin and Bonsall, 2009;Benjamin, 2010). The Grotta Verde in Sardinia (Italy, #24 in Fig. 1) has yielded submerged archaeological material in the form of cardial ceramics at À10 m depth and human remains at À8 m ( Antonioli et al., 1996) in what appears to be a submerged grave, dated to approximately 7300 BP (Fig. ...
Context 13
... Zambratija Bay site, in northern Croatia, (#17 in Fig. 1), which remains to be explored in detail, represents a submerged settlement in the northern Adriatic (Benjamin et al., 2011, Fig 16.4). It is still unclear as to why the site was abandoned, though early indications do not exclude sea-level rise as a direct cause. The site represents an important opportunity in this respect, and further, detailed study will be required to resolve the abandonment ques- tion ( Benjamin and Bonsall, 2009;Benjamin, 2010). The Grotta Verde in Sardinia (Italy, #24 in Fig. 1) has yielded submerged archaeological material in the form of cardial ceramics at À10 m depth and human remains at À8 m ( Antonioli et al., 1996) in what appears to be a submerged grave, dated to approximately 7300 BP (Fig. ...
Context 14
... Holocene archaeological features used as sea-level markers can be divided into two broad categories (Flemming, 1978;Blackman, 1973;Galili and Sharvit, 1998;Sivan et al., 2001;Lambeck et al., 2010;Morhange and Marriner, 2015): i) Features that need to be at or partially below sea level in order to function properly. These include pools that are fed by seawater driven by gravity, slipways, harbour installations, salt produc- tion installations, etc. These structures typically mark the up- permost or lowermost sea level at the time of construction (Fig. 11). ii) Features that are located normally only on dry land, including dwellings, quarries, roads, water-wells, freshwater pools, etc. These structures usually provide the uppermost sea level at the time of construction (Pirazzoli, 1976(Pirazzoli, , 1986Galili and Sharvit, ...
Context 15
... Caesarea, south of Haifa in Israel, there are ample indications for relative sea-level and tectonic stability in the past 2000 years ( Sivan et al., 2001Sivan et al., , 2004Goodman-Tchernov and Katz, 2016). One of the best indicators is the presence throughout the site within Roman and later coastal features of notches in both pools and harbours. However, while the portions of the large Roman harbour of Caesarea that were built directly on nearshore bedrock are stable, the offshore portions of it are today submerged between À1 and À5 m (Fig. 10). This is most likely due to either storm-wave pounding causing subsidence of the breakwaters in the sand (Galili and Sharvit 1998; Morhange et al. 2014) or the combined effects of tsunami damage and liquefaction (Reinhardt et al., 2006;Goodman-Tchernov et al., 2009a;Goodman-Tchernov, 2010, Goodman-Tchernov andAustin, 2015). Such variables, in an otherwise tectonically stable environment, serve as a reminder that local dynamic processes can influence the archaeological inter- pretation of key processes and ...
Context 16
... lowstand deposits produced by the Po River mainly consist of a sequence of prodeltaic deposits that prograded for 40 km in the foredeep basin and reached a maximum thickness of about 350 m in the Middle Adriatic Depression and of 70 m on the shelf ( Trincardi et al., 2004;Pellegrini et al., 2017). Topset beds of the LGM delta can be recognised through geophysical soundings from a depth of À100 m and below, and this elevation is a constraint for the LGM sea-level position ( Ridente et al., 2008;Trincardi et al., 2011aTrincardi et al., , 2011bAmorosi et al., 2014;Pellegrini et al., 2017). In the Tyrrhenian Sea, submerged depositional terraces are documented at variable depths between À50 m and À200 m. Between À90 m and À150 m they are interpreted as the evidence of the LGM shoreline ( Chiocci et al., 1997;Milli et al., 2016). Direct evidence of the LGM lowstand is known from offshore near Termini Imerese in northern Sicily, where a piston core collected a sample dated to 21.8 ka at a depth of À127 m (Caruso et al., 2011). Another location ...
Context 17
... analysis suggests that H. sapiens from across Europe exploited marine and freshwater resources between 40 and 24 ka, with freshwater fishing at Pes¸teraPes¸Pes¸tera cu Oase (#15 in Fig. 1)
Fig. 1), and the Gibraltar caves (Rodriguez Vidal et al. 2004) important coastal archeological records for this period were preserved above modern sea level due to their geographic setting. The now submerged landscapes and coastlines accessible during MIS 4 and the LGM are one of the main areas where future research must ...
Context 18
... analysis suggests that H. sapiens from across Europe exploited marine and freshwater resources between 40 and 24 ka, with freshwater fishing at Pes¸teraPes¸Pes¸tera cu Oase (#15 in Fig. 1)
Fig. 1), and the Gibraltar caves (Rodriguez Vidal et al. 2004) important coastal archeological records for this period were preserved above modern sea level due to their geographic setting. The now submerged landscapes and coastlines accessible during MIS 4 and the LGM are one of the main areas where future research must ...
Context 19
... beads were manufactured in large numbers in the Upper Palaeolithic layers at Üça gızlı Cave II in southern Turkey (#13 in Fig. 1) between 40 and 23 ka (Stiner et al., 2013) and became a feature of Upper Palaeolithic contexts across Europe ( Bar-Yosef, 2002). At Cueva de los Aviones and Cueva Ant on, Spain, perfo- rated shells and pigments have also been found dating from c.50 ka (Zilh~ ao et al., 2010) Fig. 1; Darlas, 2007;Douka and Spinapolice, ...
Context 20
... beads were manufactured in large numbers in the Upper Palaeolithic layers at Üça gızlı Cave II in southern Turkey (#13 in Fig. 1) between 40 and 23 ka (Stiner et al., 2013) and became a feature of Upper Palaeolithic contexts across Europe ( Bar-Yosef, 2002). At Cueva de los Aviones and Cueva Ant on, Spain, perfo- rated shells and pigments have also been found dating from c.50 ka (Zilh~ ao et al., 2010) Fig. 1; Darlas, 2007;Douka and Spinapolice, ...
Context 21
... MP/MSA archaeological record in the Levant contains the first evidence for H. sapiens dispersals into Eurasia during MIS 5.5e3, with their remains preserved in terrestrial contexts at Skhul and Qafzeh caves dated to 100e130 ka (#7 in Fig. 1, Grün and Stringer, 1991;Grün et al., 2005). Both Neanderthals and H. sapiens populations utilised MP/MSA technology in the Levant during this period (Shea, 2003). The extent to which these pop- ulations exploited the coastal regions and the resources they con- tained is becoming clearer with targeted research into the origins of marine exploitation. Outside of the Mediterranean, evidence for exploitation of marine resources (molluscs, mammals etc.) by H. sapiens is known from MIS 6 contexts in southern Africa, with populations there developing a full 'coastal adaptation' by ~110,000 ka (Marean, 2014). Possible occupation of coastal environments during MIS 5.5, presumably by H. sapiens, are also suggested from the Red Sea region ( Walter et al., 2000;Bailey et al., ...
Context 22
... overarching aim of this article is to define the state of the art of Mediterranean sea-level studies, a century after its inception, and to consider the impacts of past sea-level and coastal changes on human-environment interaction. We identify and highlight the major on-going discussions and gaps in knowledge which we expect to, at least partially, define the next decade of integrated sea-level research into past coastal environments and archaeology ( Fig. 1). In doing so, we aim to bring together the research of the geomorphological and archaeological communities and promote interdisciplinary work specifically related to sea-level ...
Similar publications
The fluvial systems of Saurashtra, Gujarat, India have archived signatures of sea-level stands for the Late Quaternary Period. Based on geomorphology, sedimentology and optical dating, the lower reaches of Noli river in southwestern Saurashtra reveal the presence of marine, fluvially reworked and aeolian forms of miliolite. Optical chronology permi...
An approach to the contribution of anthropogenic actions on the Araruama lagoonal system (SE Brazil) geomorphological evolution. Journal of Sedimentary Environments, 3 (2): 108-120.
Abstract: The first studies regarding the constitution of the Araruama lagoonal system, located in the Rio de Janeiro state (SE Brazil), were carried out by Alberto Ri...
Citations
... Many small-sized Mediterranean islets have 'disappeared' or 'appeared' even during the last millennia due to various geological processes and gradual sea-level changes (Benjamin et al., 2017;Kopaka & Kossyva, 1999;Vacchi et al., 2018). Our lack of knowledge probably leads us to underestimate the effects on these microinsular landscapes of coastal dynamics and of the presence of humans since their early colonizations. ...
We provide a synthesis of the history and the extent of human impact on the terrestrial ecosystems of the small islands and islets of the western Mediterranean. We focus on the exploitation of islands smaller than a few thousand hectares. The use of microinsular biotic resources (e.g., timber, soda ash, wild mammals) began as early as the Upper Palaeolithic, intensified during the Neolithic, and expanded during the Iron Age, peaking around the end of the nineteenth to the mid-twentieth century. As a consequence of long-lasting land exploitation, in many cases local terrestrial ecosystems were severely damaged and some of their living components disappeared. In other cases, even with uninterrupted human occupation over millennia, several insular resources have been sustainably managed. Although these small islands are currently under less pressure compared with neighbouring mainland coastlines, their protection or even the restoration is unquestionably necessary.
... These coastal zones are further experiencing dramatic beach retreat, coastal erosion, marine flooding and salinization of the water tables that is causing 85% of the estimated damage costs. Natural and anthropogenic land subsidence is increasing the risk of flooding along the Mediterranean coasts (Anzidei et al. 2021;Vecchio et al. 2023), causing the relatively slow submergence of ancient coastal settlements of the Roman or pre-Roman age (Benjamin et al. 2017 and references therein), historical cities like Venice (Vecchio et al. 2019), and small islands . ...
Sea level rise (SLR) is among the major climate change effects threating the coasts of the Mediterranean basin, which are increasingly exposed to coastal flooding, especially along the low lying coastal plains, river deltas, lagoons and reclamation areas. Coastal erosion, beach retreat and marine flooding are already causing unprecedented environmental and socio-economic impacts on coastal populations. According to the Intergovernmental Panel on Climate Change (IPCC) these effects are expected to worsen by 2100 and beyond with a projected global SLR up to about 1 m above the current level. This study provides an overview of the Mediterranean basin, focusing on the vulnerable city of Venice, which is particularly exposed to marine flooding due to SLR and land subsidence. We show the current and future sea level trend as well as a flooding scenarios in the absence of the Experimental Electromechanical Module (MoSE), which is protecting the city of Venice since 2020. To understand the awareness of citizens in Venice to address SLR, we have engaged a group of stakeholders through a structured participatory process to develop solution-oriented, case-specific and site-specific Policy Tools. Our results show that the Policy Tools contain relevant, effective and implementable actions stemming from stakeholder interaction and consensus building, identifying relevant issues that should be considered for SLR adaptation policies. A more extensive participation in public processes is required to materialize the Policy Tools into concrete actions to help vulnerable areas adapt to the expected SLR by the end of this century.
... To better understand how humans have responded to past rapid climate and sea-level change and inform contemporary responses, the intersect between human and social evolution and environmental change has become a focus of interdisciplinary research (Benjamin et al., 2017;Dong, 2018;Davis, 2020). The Balzi Rossi record is of particular significance for its potential to inform how Palaeolithic coastal inhabitants interacted and responded to coastal changespecifically how Middle and Upper Palaeolithic populations interacted with and responded to Mediterranean coastline shifts driven by global climatic change. ...
... It is characterised by tropical, warm water taxa including the well-recognized Thetystrombus latus [formerly Strombus bubonius]. Senegalese faunal assemblages are mainly attributed to the Last Interglacial (MIS 5e) highstand within the western Mediterranean; although, some deposits associated with the MIS 7 highstand have been identified in Spain (Zazo et al., 2013;Benjamin et al., 2017;Cerrone et al., 2021). Interestingly, one taxon typical of the Senegalese group, Patella ferruginea, has been identified within the M1 fossil assemblage from the Prince of Monaco's cave and could indicate a highstand of elevated water temperatures within the Mediterranean predating the Last Interglacial. ...
The Balzi Rossi archaeological complex (comprised of caves, rock shelters, and open-air sites) is a globally significant site for Palaeolithic culture and understanding the transition from Neanderthal to Anatomically Modern Human populations in Europe. It also retains some of the earliest evidence of human interactions with their coastal environment. Balzi Rossi has been subject to excavation for over 150 years – traditionally as individual site locations – with most deposits removed when the discipline of archaeology was nascent, and the science not yet developed. The consequence was the unfortunate loss of materials and critically important stratigraphic context. However, valuable information regarding the Palaeolithic population, their coastal environment, and earlier sea-level change, remains in the literature and in museum repositories. In this work we have compiled and reviewed the extensive resources, available largely in French and Italian, to provide a summary and catalogue for each individual site. These ‘Site Summaries’ are available as appendices to this review, which provides a comprehensive synopsis of the history of excavations conducted at Balzi Rossi, a reconstruction of stratigraphy where possible, the evidence of Palaeolithic occupations, the evidence of Pleistocene sea-level fluctuations, and an assessment of the chronological constraint available for both the Palaeolithic populations and sea level. Finally, this synopsis identifies gaps in knowledge and provides comments on pathways for future research, suggesting a consilient approach that can be applied in other archaeological contexts.
... cal. BC), regional estimations suggest that the RSL was c. − 15 m in the Mediterranean after a very sharp rise during the Early Holocene (Lambeck et al., 2014;Benjamin et al., 2017). During the 6th millennium cal. ...
This paper presents an integrated Geoarchaeological approach to Holocene landscape change and socioenvironmental interaction around ancient Abdera, a Greek colony in Aegean Thrace. A combination of remote sensing, geomorphological mapping, sedimentary coring, and radiocarbon dating was used to build the first detailed Holocene sea-level curve in the Northern Aegean, and to reconstruct the palaeogeographic evolution around Abdera from the Neolithic onwards. The discussion of these results, alongside the available archaeohistorical data, sheds new light on the role of historical and environmental factors in the rise and eventual
decline of Abdera, thereby challenging previous narratives. This study is the first to detect evidence of the Neolithic landscapes in this coastal area being submerged by marine transgression c. 5000 cal. BC. A lagoonal landscape developed from the Neolithic until Greek colonists settled in the area c. 654 BC. The results presented
here suggest that the relocation of the city in the 4th century BC was not caused by the silting up of the bay but was due to historical events. The coastal progradation only became significant after 300 cal. BC and accelerated in the following centuries. Based on the data acquired, it is hypothesized that this was due to anthropogenic
forcing of the sedimentary systems both at local and regional levels, caused by the development of the productive activities of the colony. The evidence presented here demonstrates that while Abdera retained good access to the sea, its decline in the 4th c. AD was certainly due to the gradual shift of the economic axis inland. By 400 cal. AD the coastline was very close to its current position. The results of this study also add to the catalogue of known tsunamis in the Northern Aegean, by providing new evidence of a tsunami on this section of the Thracian coast in 544 AD.
... Although the Panthera Cave was not far from the coast, the game brought to the site consisted of medium-sized ungulates, suggesting a clear orientation to terrestrial resource acquisition. Beyond the Ionian, in the Eastern Adriatic, karstic landscapes with a number of submerged caves are also found on the western coasts of Albania and further north on the Croatian and Slovenian coasts (Benjamin et al. 2011(Benjamin et al. , 2017Rossi et al. 2020). The present-day Dalmatian islands would have been part of the Adriatic mainland for most of the Pleistocene (Markovic-Marjanovic 1971; Shackleton et al. 1984;Surić et al. 2009). ...
The strategic geographical position of the Balkan Peninsula, at the crossroads between southwest Asia and central and western Europe, make of this territory a key area for understanding the different human migrations into Europe during the Pleistocene. This long-time neglected area for the Palaeolithic research, last years has experienced a blossoming in terms of research projects and key discoveries. Only in the past decade, sites from the Balkan Peninsula have yielded, for instance, the oldest anatomically modern human occupations in Europe, the first human remains of our species in the continent, the first confirmation of interbreeding between us and Neanderthals and evidence of Palaeolithic rock art, a phenomenon traditionally restricted to South-western Europe. This volume provides a comprehensive overview of the previous data and new discoveries, addressed by an international list of contributors among the most renowned scholars developing archaeological researches in this territory. It summarises the state of the art for the Early Prehistory Archaeology of one of the most important emerging territories for the discipline.
... However, in the north-western and southern parts of the region, as in Slovenia (Brodar & Osole 1979), Bosnia and Herzegovina (Montet-White et al. 1986;Montet-White 1996;Jovanović et al. 2014) and Greece (Tourloukis 2010), Middle Palaeolithic sites are not yet known before MIS 5. A similar situation is present in Croatia, where the lack of analysis of the assemblages from the Middle and upper Pleistocene deposits dated to MIS 6 and 5 in Vindija (Wild et al. 2001;Karavanić et al. 2021b) hinders the clarification of the earliest occurrence of the Middle Palaeolithic. This absence of the Middle Palaeolithic record in some parts of South-eastern europe prior to MIS 5 is explainable in terms of research history and geologic processes (e.g., Tourloukis 2010), but it is also conceivable that some mountainous regions were depopulated during the MIS 6 glacial expansion (Bavec & Verbič 2011), and that the transgression during MIS 5e destroyed sites from the preceding glacial in what is today the Adriatic (Benjamin et al. 2017). The two most widespread industrial variants during MIS 5 in South-eastern europe are the Typical Mousterian and the South-eastern/ Balkan Charentian (hereafter charentian) ( figure 4.3). ...
The strategic geographical position of the Balkan Peninsula, at the crossroads between southwest Asia and central and western Europe, make of this territory a key area for understanding the different human migrations into Europe during the Pleistocene. This long-time neglected area for the Palaeolithic research, last years has experienced a blossoming in terms of research projects and key discoveries. Only in the past decade, sites from the Balkan Peninsula have yielded, for instance, the oldest anatomically modern human occupations in Europe, the first human remains of our species in the continent, the first confirmation of interbreeding between us and Neanderthals and evidence of Palaeolithic rock art, a phenomenon traditionally restricted to South-western Europe. This volume provides a comprehensive overview of the previous data and new discoveries, addressed by an international list of contributors among the most renowned scholars developing archaeological researches in this territory. It summarises the state of the art for the Early Prehistory Archaeology of one of the most important emerging territories for the discipline.
... 87 Sr/ 86 Sr in carbonates has therefore been used to conduct stratigraphic correlations and marine carbonate dating (Elderfield, 1986). It has also been used as a tracer of rock source type/age (Bayon et al., 2021). ...
... Previous interglacial marine highstands stood within a few meters from the current highstand (Antonioli et al., 2004;Benjamin et al., 2017). Therefore, at places that undergo moderate amounts of coastal progradation or retrogradation highstand coastal islands reform at the same place from one interglacial to the next, until they become either permanently attached to the land, or eroded away. ...
... The offshore ridge north of Akko, emplaced seaward of the MIS 5.5 ridge, was therefore likely emplaced during interstadial 5.3 and 5.1. The ridge of Tyre, which rises 30 m above the seafloor therefore most likely formed during such interstadials, when sea level stood 20 to 30 m bsl (Benjamin et al., 2017). It is also consistent with the age of the flooded 99 ka calcarenite ridge further north, which supports the island and Phoenician city of Arwad (Dodonov et al., 2008). ...
The renowned Phoenician city of Tyre was settled on a small coastal island from which it resisted invasions and
sieges for centuries. In 332 BCE, the city was taken when Alexander-the-Great built a 750 m-long causeway to
seize the island. The causeway interrupted longshore sediment transport, forcing sand to accumulate along the
causeway, creating a isthmus that today still connects the island to the mainland. This isthmus is no less than
530 m wide and 13 m high. We studied the impact of isthmus growth on city development during Antiquity, by
combining archeological data, core stratigraphy, and multibeam bathymetry to track the paleogeographic evolution
of the land-facing coast of Tyre, from the Holocene marine transgression up to today. In Phoenician time
(900-300 BCE), sea-level markers indicate that relative sea level lied more than 2.5 m below current sea level,
defining a 1500 × 600 m rocky island, 450 m longer in the south than the modern rocky headland. Between the
island and the mainland, diffracted/refracted waves built a submerged sand bank that rests on eroded transgressive
lagoonal clays. Along the mainland-facing coast of Tyre Island, Phoenician walls were built on the
emerged end of the sand bank. Middle infrared spectroscopy (MIRS) shows that the sand of the bank and its
successor isthmus is made of carbonate clasts (90–100%). 36Sr/37Sr ratios further indicate that these are modern
bioclastic sands. Coeval, heterogeneous clay-rich sediments were deposited between the rocky island, and the
sand bank. We interpret these as harbor sediments, deposited behind buried and/or submerged Phoenician
breakwaters. The sand bank inflated dramatically after the building of Alexander’s causeway, rising 5 m during
the Hellenistic and Roman periods. It rose another 4 m–8 m in successive stages during Byzantine and Medieval
times, up to the 19th century. The rapid growth of the sandy isthmus after 332 BCE led to the demise and burial
of the harbors that had been developed along the mainland-facing coast. The Phoenician southern harbor of Tyre
appears to have been buried in Roman times during the repurposing of the SE corner of the island for monumental
baths. We interpret this repurposing as a response to the growth of the sandy isthmus, and suggest that
the northern Phoenician harbor was then relocated to its present position. Subsequent relative sea level rise led to
the erosion of the landfill overlying the southern harbor, exposing the formerly buried port on the seafloor.
... The most recent geo-archaeological studies related to the Mediterranean Sea show that the main reason for the submersion of underwater heritage is linked to a slow increase in the sea level, which has caused a slight but noticeable regression of the coastline over the last 2000 years [32][33][34], with some more important tectonic movements (mainly local) such as the case of Campi Flegrei, which have caused a more significant submersion of several meters below the current sea level [35]. ...
Underwater cultural heritage (UCH) is a diverse and valuable resource that includes shipwrecks, sunken cities, and other submerged archaeological sites. It is an important part of human history and culture and can significantly benefit society. However, various factors often neglect and threaten UCH, including climate change, pollution, and human activities. Several factors, including technological advances, the development of international law, and the growing awareness of the importance of cultural heritage, have influenced the evolution of the concept of UCH. In the early days of underwater archaeology, the focus was on recovering artifacts and treasures from shipwrecks. However, over time, there has been a shift towards a more holistic approach to the management of UCH, which emphasizes the importance of in situ preservation and the involvement of local communities. This review provides a chronological analysis of the evolution of the concept of UCH over the past 70 years and examines the main international conventions and charters developed to protect UCH. The review also discusses the relationship between UCH and marine protected areas (MPAs), the marine environment, and the coastal landscape.
... According to Desruelles et al. (2009), the mean sea level in Mykonos-Delos-Rhenia was approximately 2.50 (±0.50) meters below the present sea level (b.s.l.) around 400 BCE. This evolution over time is due to eustatism and glacio-hydro-isostasy, whose effects are enhanced by the subsidence trend (Lambeck and Purcell, 2005;Pavlopoulos, 2010;Pavlopoulos et al., 2011;Sakellariou and Galanidou, 2016;Benjamin et al., 2017;Roy and Peltier, 2018). ...
... The southern part of the bay was more accessible to boats due to the shape of the bedrock. During the last glacial maximum (17,000 to 21,000 years ago), when the sea level was about 120 m lower than today (Lambeck et al., 2014;Benjamin et al., 2017), the Inopos stream carved a riverbed in the extension of its emerged valley, heading south (Fig. 7a). The part of the strait, located west of the two Rematiaris islets, was relatively deep and more suitable for navigation. ...
... However, in the north-western and southern parts of the region, as in Slovenia (Brodar & Osole 1979), Bosnia and Herzegovina (Montet-White et al. 1986;Montet-White 1996;Jovanović et al. 2014) and Greece (Tourloukis 2010), Middle Palaeolithic sites are not yet known before MIS 5. A similar situation is present in Croatia, where the lack of analysis of the assemblages from the Middle and upper Pleistocene deposits dated to MIS 6 and 5 in Vindija (Wild et al. 2001;Karavanić et al. 2021b) hinders the clarification of the earliest occurrence of the Middle Palaeolithic. This absence of the Middle Palaeolithic record in some parts of South-eastern europe prior to MIS 5 is explainable in terms of research history and geologic processes (e.g., Tourloukis 2010), but it is also conceivable that some mountainous regions were depopulated during the MIS 6 glacial expansion (Bavec & Verbič 2011), and that the transgression during MIS 5e destroyed sites from the preceding glacial in what is today the Adriatic (Benjamin et al. 2017). The two most widespread industrial variants during MIS 5 in South-eastern europe are the Typical Mousterian and the South-eastern/ Balkan Charentian (hereafter charentian) ( figure 4.3). ...
This paper reviews the Middle Palaeolithic of South-eastern Europe in light of the results obtained by recent research. A large territory from Slovenia in the north to Greece in the south and from Croatia in the west to Bulgaria and Romania in the east is covered by the chapter. The chronology and characteristics of defined technocomplexes from MIS 7 till MIS 3 are presented and discussed in their functional and environmental context. Examples of these technocomplexes include the so-called South-eastern European (Balkan) Charentian, so-called Micromousterian as both Charentian and a kind of Denticulate Mousterian, Typical Mousterian with Levallois technology and often with bifacial points and other industrial variants that appeared in the region during this period. We primarily consider aspects of technology, typology and stone raw-material procurement in chronological and regional context but also included are aspects of palaeoecology and other patterns in our interpretations.
