Chronological evidence for a pre-Minoan age of pyroclastic deposits on Anafi Island, Cyclades, Greece
- 162 Downloads
Pyroclastic deposits on Anafi Island, Cyclades, Greece, were initially related to the Minoan eruption (1613 BC), but later geochemical studies favored an older chronology that points at the Lower Pumice eruption of Santorini (172 ka). Here we attempt through OSL dating to provide a minimum age by dating sand-sized quartz from colluvium overlying the deposits. The chronological results excluded the possibility of being Minoan.
KeywordsAegean Anafi Pumice Santorini eruption OSL dating Colluvium deposits
The Anafi Island, 30 km east of Thera, has been intensively investigated for the presence of pumice deposits related to the Minoan explosive activity of Thera’s volcano (Keller 1980, 1996; Francaviglia 1990; Marinos 1963). The ‘Minoan’ eruption, occurred in the seventeenth century BCE (1627–1600 BCE, Friedrich 2013) and had widespread impacts on the civilization of the Aegean and Eastern Mediterranean (Marinatos 1939; Nixon 1985; Doumas 1990).
At the beginning of the Late Bronze Age (LBA), major civilizations had flourished in the region: The Minoan culture in Crete and Dodecanese, the Hattian, Hittite, Assyrian and Luwian kingdoms in Asia Minor (Collins 2007) and Egypt were undergoing the transition from the Middle to the New Kingdom (Ryholt 1997). The Minoan eruption showered tephra over these major archeological sites of the Eastern Mediterranean and might have had intense impacts on them.
The tephra dispersion modeling (Bonadonna et al. 2005) or geostatistics (Athanassas et al. 2018) have been used to predict the thickness and spread of the Minoan tephra in adjacent areas. Some deterministic models (Sewell 2001) have shown that the dispersal of Minoan tephra had a W–E trend, while a geostatistical prediction map (Athanassas et al. 2018) suggests a NE–SW axis stretching from approximately the Bosporus to the eastern tip of Crete, and exhibits an eastward drift away from Santorini as far as Central-Western Anatolia. Recent studies in western Turkey certified the most eastward occurrence of Minoan tephra on land and the first in continental deposits except from lacustrine sediments (Sulpizio et al. 2013).
The Anafi Island could be a key site for the dispersal of Minoan tephra as it is the most nearby island of Santorini. Despite its proximity, only a few spots with pyroclastic deposits have been found on the island. Therefore, the same situation is expected to be observed in adjacent islands, such as Melos, Astypalaia, Ios, etc. The occurrence of tephra layers of the Minoan eruption on Rhodes, Kos and western Turkey (Doumas and Papazoglou 1980; Keller 1980; Eastwood et al. 1999), suggests that Anafi must have been covered by Minoan tephra, which was probably eroded and swept away afterwards (Keller et al. 2014).
In recent years, a debate has sparked regarding the relation of the Anafi deposits to the Minoan eruption. In early 1960s Marinos (1963) and later McCoy and Dunn (2002) associated the thick pumice deposits on Anafi with the Minoan eruption, suggesting “a far larger (Minoan) eruption than previously thought”. On the contrary, Francaviglia (1990) and Keller et al. (2014), after geochemical determination of glass composition by electron microprobe analysis on the samples from Anafi, pointed out a strong geochemical similarity with those analyzed from LP2 eruption of Thera at 172 ka, as has been determined by Druitt et al. (1999), excluding any correlation with the Minoan eruptive phase. The deposits of the Lower Pumice eruptions are noticeable in the caldera cliffs of central and southern Thera (Gertisser et al. 2009). A definite comparison of the composition of pumice glass with major eruptive events of Thera, gave a strong correlation of the pumice deposits in eastern and western Anafi, and their correlation with the proximal Plinian LP2 deposits in the caldera wall of Thera.
In this paper, we attempt to settle the debate of the tephra chronology on Anafi on the basis of numerical dating and specifically by applying the OSL dating method to colluvial deposits that bracket the pumice deposits, reported by McCoy and Dunn (2002) and Keller et al. (2014). As the underlying layers on which the pyroclastic deposits have been deposited are not exposed in the studied sections, OSL dating was limited to the alluvium layers which directly cap onto the pyroclastic deposits. It is noted that it is the first time that the pyroclastic deposits of Anafi will be chronologically constrained, providing indirectly the provenance (Minoan or of an earlier eruptive phase) of the pyroclastic deposits.
Anafi is a non-volcanic island, located ca. 30 km east of Santorini, with the geological structures differing from other islands of the Hellenic Arc in some essential aspects. Regardless of its position at the center of the volcanic arc, no important volcanic activity has so far been reported, except some recent field observations and petrological studies confirming the existence of Neogene or Pleistocene volcanism on the island (Leichmann and Hejl 2006) related to an extensional basin, which became filled with siliciclastic sediments and to a lesser extent by volcanogenic deposits.
Anafi is located at the southeastern edge of the Atticocycladic metamorphic core complex, north of the Cretan back-arc basin. It shows a remarkable lithological variety, consisting mainly of metamorphosed and non-metamorphosed Alpine units, as well as post-Alpine sediments (Soukis and Papanikolaou 2004).
Materials and methods
At Prassa, along a new road-cut, the pyroclastic deposit is exposed for about 10 m and has a thickness of 1–1.5 m. This deposit is stratified and subdivided into three units (Keller et al. 2014): (a) the lowest, a white layer of pumice, well sorted, with a thickness of 5 cm and maximum pumice diameter of 1 cm; (b) in the middle a 35-cm-thick, white, non-stratified layer and (c) the upper, consisting of a gray, poorly sorted layer of pumice ~ 40 cm thick with maximum pumice (MP) diameter of 5 cm and maximum lithic size (ML) of 4 cm.
Vounia is located in the southeastern part of the island and at about 250 m a.s.l. These deposits are subdivided into two main units: (a) the lower part consisting of well-sorted fine grain gray pumice and the upper one, consisting of white pumice poorly sorted. Colluvial deposits cover the pumice unit.
The eligibility of OSL dating on colluvial deposits is shown in numerous studies (Fuchs and Lang 2009; Kadereit et al. 2010; Fuchs et al. 2011) despite cases of partial bleaching. Partial bleaching is usually due to limited exposure of sediment grains to daylight because of rapid transport of slope deposits or low-luminescence sensitivity of quartz in many lithologically immature deposits. Τo avoid modern bleaching by bioturbation, soil material from the upper 50 cm of the profiles was not sampled for OSL dating (Lang and Hönscheidt 1999).
The sample preparation was carried out at the Department of Geology and Geoenvironment, University of Athens, and aimed at extracting quartz grains from the sampled colluvial.
Pure quartz grains were separated according to standard preparation procedures (e.g., Athanassas 2011) that involve treatment with (i) 10% hydrochloric acid to extract calcite minerals; (ii) 20% hydrogen peroxide (to remove any amount of organic matter); (iii) sieving to isolate grains of sizes ranging between 125 and 212 μm; (iv) density separation using sodium polytungstate (with densities of 2.769 and 2.62 g cm−3) to separate quartz grains from the feldspathic content) hydrofluoric acid (40%) etching to remove the alpha-particle influenced outer rim of the quartz grains and (vi) a final rinsing in 10% HCl to remove any residual soluble fluoride salts.
OSL measurements were carried out at the OSL dating laboratories at CEREGE, Aix-Provence, France, and the Archaeometry Center, University of Ioannina, Greece. Equivalent doses were estimated using the standard single aliquot re-generated (SAR) dose protocol by Murray and Wintle (2000), while for the estimation of the dose rate, measurements of U, Th and K were conducted by ICP-MS and then converted to dose rate units (Gy/ka) using conversion factors by Guerin et al. (2011).
In the case of significant skewness of the equivalent dose distribution, a minimum age model (Cunningham and Wallinga 2012) was used. Skewness can result from partial bleaching, e.g., by bioturbation. When incompletely reset grains get involved, the De specific to population representing the event can be estimated by the ‘minimum age model’ (Galbraith et al. 1999; Galbraith and Roberts 2012). This statistical basis was used for the samples measured here to calculate equivalent doses, as most representative of the deposition as possible.
Summary of the OSL dating results
Grain size (μm)
Dose rate (Gy/Ka)
MAM De (Gy)
OSL age (ka)
2.0 ± 0.1
42.0 ± 2.8
21.0 ka ± 1.4
1.8 ± 0.1
54.1 ± 2.9
30.1 ka ± 1.6
The results derived by the application of OSL dating, demonstrated that the colluvium above the pumice deposition at Prassa was dated about 21 ka and in Vounia about 30 ka.
It is clearly noticed that the pyroclastic deposits could not be of Minoan origin, as proposed by McCoy and Dunn (2002) and that the colluvium sediments, cover over eruptions preceding the Minoan. Since there is no facility to access to the base of pyroclastic deposits, these ages could be considered as minimum ages for the pyroclastic deposition.
The OSL dates of Prassa and Vounia both well predate the Minoan eruption. Despite the fact that the OSL age of Prassa seems to be concordant with the Cape Riva eruption (21.8 ± 0.4 ka; Druitt et al. 1999), we refrain from relating the deposits to the latter, as the underlying volcanic deposits could be indefinitely older, if a hiatus has occurred. Since sediment deposition is controlled by local factors (i.e., topographic gradient, clastic material supply, type of parent rock, precipitation rate, etc.), initiation of colluvium deposition could be non-simultaneous, accounting for the difference between the two ages.
Keller et al. (2014), based on their geochemical analysis on these deposits, excluded the possibility of being related to the Minoan or Cape Riva eruption. An element-by-element comparison made by them, showed an average composition of 71.55 and 72.17 wt% SiO2 for LP2 (on Thera and Anafi, respectively), while samples from the Minoan eruptive phase are usually characterized, by an average SiO2 content of 73.81 wt% (Keller et al. 2014). Additionally, measurements of FeO, which for Anafi’s samples showed an average composition of 2.56 and 2.79 wt%, excluded any possibility for Cape Riva eruptive phase (22 ka), where FeO rates fluctuate around ~ 3.5 wt% (Druitt et al. 1999). However, other chemical analyses of FeO from Cape Riva eruption on tephra layers from the Black sea fluctuate between 3 and 3.3 wt% (Wulf et al. 2002). These assertions identify both deposits with the Lp2 eruptive activity of Santorini and can be questioned, since they are based exclusively on small geochemical variations of SiO2 and FeO.
The Minoan provenance of tephra is also excluded by Keller et al. (2014) because of its high SiO2 content (~ 73.81 wt%) compared to Anafi’s deposits which range between 71.55 and 72.17 wt%. Despite the fact that these differences are small, and within the limit of the statistical error, other analyses on Minoan tephra deposits estimated SiO2 content lower than 73 wt%, between 68.91 and 69.49% (Warren and Puchelt 1990; Vitaliano et al. 1990; Francaviglia 1990). Cape Riva (CR) eruption unleashed several km3 or more of magma, generating four eruption units: a basal Plinian fall deposit (CR-A) and three pyroclastic flow deposits (CR-B to CR-D upwards). CR-B and CR-D are welded ignimbrites (Druitt 1985). It tapped a zoned magma chamber containing rhyodacite overlying andesite and produced a rhyodactic tephra (Wulf et al. 2002; Druitt 1985). It is already recorded in deep-sea sediments from the Black Sea (Kwiecien et al. 2008), the Marmara Sea (Wulf et al. 2002), the Aegean Sea, in eastern Mediterranean deep-sea sediments near western Cyprus (Wulf et al. 2002); in Tenaghi Philippon in Greece (Seymour et al. 2004) and in lake Iznik in Turkey (Roeser et al. 2012).
According to our new chronological framework from the two sites at Vounia and Prassa, we conclude that tephra deposits in Anafi could not be of Minoan origin as McCoy and Dunn (2002) have suggested, given that the age of the stratum above is about 21 ka. The absence of Minoan pyroclastic deposits has also been confirmed in neighboring islands of Santorini.
On the other hand, we could not unquestionably exclude its LP2 origin as suggested by Keller et al. (2014), as we did not date the base of pyroclastic deposits. If one attempted to refute McCoy and Dunn (2002) theory in favor of the LP2 origin in Anafi proposed by Keller et al. (2014), based exclusively on subtle deviations between the Minoan deposits in Santorini and the pyroclastics in Anafi, would be precarious, since the deviations in the same data for the same explosion (Minoan) from other analyses (e.g., Vitaliano et al. 1990) are quite different. On the other hand, the field characteristics of tephra deposits, such as the stratigraphy that matches the sub-units of LP2 identified in the proximal deposits in the Santorini cliffs, their color, thickness and the (rare) occurrence of mafic cauliform blebs, could be correlated with LP2 eruption.
We come here with an independent and pure geochronometric approach to definitively exclude their Minoan origin. This work is useful inasmuch as it reinforces a pre-Minoan age of these deposits on the basis of numerical chronology enhancing the qualitative character of former field observations (Keller et al. 2014). More filed observations, geochemical and chronological analyses on pyroclastic deposits are needed, for the more efficient determination of their provenance and for the better understanding of the effects of Minoan eruption in the Aegean and eastern Mediterranean.
This research is innovative, as it attempted the chronological determination of pyroclastic deposits in the Anafi Island and the clarification of the effects of Minoan eruption on the island. According to our results stemming from OSL dating on colluvial sediments covering pyroclastic deposits, we excluded any possibility of them being Minoan, since the age of the colluvium deposition is about 21–30 ka. Thus, it appears that the deposition of tephra is much older than 20 ka. The conclusion that we want to convey is that, based on OSL dating, these deposits ought to be pre-Minoan (be it LP2 or other). More geochemical and chronological analyses on tephra deposits are needed to clarify their provenance.
Katerina Theodorakopoulou was entitled by the State Scholarships Foundation (IKY) (7376) for a post-doctoral fellowship (MIS 5001552). Dating of the deposits was funded through ENVI-Med/Mistrals (Acronym: Med-HolVol), granted to CEREGE, France. The authors are grateful to the comments and suggestions of guest editor and two anonymous reviewers who have improved this contribution.
Compliance with ethical standards
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
- Bonadonna C, Phillips JC, Houghton BF (2005) Modeling Tephra Fall from a Ruapehu Weak Plume Eruption”. J Geophys Res 110:B8Google Scholar
- Boynton WV (1984) Cosmochemistry of the rare earth elements: meteorite studies. Developments in geochemistry, vol 2. Elsevier, Oxford, pp 63–114Google Scholar
- Collins BJ (2007) The Hittites and their world. The Society of Biblical Literature, New York, p 272Google Scholar
- Doumas C (1990) Archaeological observations at akrotiri relating to the volcanic destruction. In: Hardy DA, Renfrew AC (eds) thera and the aegean world III/3. The Thera Foundation, London, pp 48–50Google Scholar
- Francaviglia V (1990) Sea-Borne pumice deposits of archaeological interest in Aegean and Eastern Mediterranean Beaches. In: Hardy DA, Renfrew AC (eds) Thera and the Aegean World III, volume three, chronology. The Thera Foundation, London, pp 127–134Google Scholar
- Friedrich WL (2013) The Minoan Eruption of Santorini around 1613 BC and its consequences. Tagung en des Landesmuseums für Vorges chichte Halle 2013:9Google Scholar
- Guerin G, Mercier N, Adamiec G (2011) Dose-rate conversion factors: update. Ancient TL 29:5–8Google Scholar
- Keller J (1980) Prehistoric pumice tephra on Aegean islands. In: Doumas C (ed) Thera and the Aegean World II, vol 1. The Thera Foundation, London, pp 49–56Google Scholar
- Keller J (1996) Significance of recent explosive activity at Santorini (Minoan and Post-Minoan) to hazard assessment. In: Keller J, Druitt T, Pyle D, Sbrana A (eds) The European Laboratory Volcanoes: volcanology of santorini: magma evolution and physical volcanology of historic, prehistoric and quaternary eruptions at Santorini, final report, ESF Strasbourg, vol 1, pp 1–16Google Scholar
- Leichmann J, Hejl E (2006) Volcanism on Anafi island: short living, extensional, hydromagmatic volcanism in the central part of the South Aegean volcanic chain (Greece). Neues Jahrbuch für Mineralogie-Abhandlungen J Mineral Geochem 182(3):231–240Google Scholar
- Marinos GN (1963) The geology of Anaphi island (in Greek). IGME VIII:3Google Scholar
- McCoy FW, Dunn S (2002) Modelling the climatic effects of the LBA eruption of Thera: new calculations of tephra volumes may suggest a significantly larger eruption than previously reported (abstract). In: Proceedings of the Chapman Conference on Volcanism and the Earth’s Atmosphere, American Geophysical Union, Santorini, Greece, pp 21–22Google Scholar
- Melidonis N (1963) Die Geologie der insel Anafi, Athens. Geol Geophys Res 8(3):53–308Google Scholar
- Roeser PA, Franz SO, Litt T, Ulgen UB, Hilgers A, Wulf S, Wennrich V, Ön SA, Viehberg FA, Çağatay MN, Melles M (2012) Lithostratigraphic and geochronological framework for the paleoenvironmental reconstruction of the last 36 ka cal BP from a sediment record from Lake Iznik (NW Turkey). Quatern Int 274:73–87CrossRefGoogle Scholar
- Ryholt KSB (1997) The Political Situation in Egypt During the Second Intermediate Period c. 1800–1550 B.C. Carsten Niebuhr Institute Publications, Copenhagen, Museum Tusculanum Press, Copenhagen, p 463Google Scholar
- Sewell DA (2001) Earth, air, fire and water: an elemental analysis of the Minoan eruption of Santorini volcano in the Late Bronze Age. Unpublished PhD thesis, University of Reading. Accessed 29 Nov 2016Google Scholar
- Vitaliano CJ, Taylor SR, Norman MD, McColloch MT, Nicholls IA (1990) Ash layers of the Thera volcanic series: stratigraphy, petrology and geochemistry. In: Hardy DA, Keller J, Galanopoulos VP, Flemming NC, Druitt TH (eds) Thera and the Aegean World III: volume two—earth sciences, proceedings of the third international congress, Santorini, Greece. Thera Foundation, London, pp 53–78Google Scholar
- Vorwerk W (1979) Das Neogen im Nordwestteil der Insel Anafi/Kykladen/Griechenland. Diploma Thesis, Univ. Kiel, p 83Google Scholar
- Warren PM, Puchelt H (1990) Stratified pumice from Bronze Age Knossos. In: Hardy DA, Renfrew AC (eds) Thera and the Aegean World III: volume three—chronology. Proceedings of the third international congress, Santorini, Greece. Thera Foundation, London, pp 71–81Google Scholar