Navigation – Plan du site

Geoarchaeological study of karst depressions integrating geophysical and sedimentological methods: case studies from Zominthos and Latô (Central and East Crete, Greece)

Étude géoarchéologique des dépressions karstiques intégrant la géophysique et la sédimentologie : exemples des sites de Zominthos et de Latô (Crète centrale et orientale, Grèce)
Christoph Siart, Matthieu Ghilardi et Ingmar Holzhauer
p. 241-256

Résumés

Cet article a pour but de présenter les dynamiques morphologiques holocènes de dépressions karstiques du centre et de l’est de la Crète (Grèce), en liaison avec l’occupation humaine. Deux sites différents, Zominthos et Latô, ont été privilégiés et étudiés en raison de la présence de structures archéologiques datées des époques minoenne à hellénistique. Pour la première fois en Crète, l’approche géoarchéologique, associant une étude géophysique et des analyses sédimentologiques, a été privilégiée dans des contextes géologiques et géomorphologiques similaires. Des observations de terrain, associées à des analyses de résistivité-tomographie ont permis de révéler l’épaisseur des dépôts hétérogènes (20 m) et de mettre en évidence l’existence de structures archéologiques. D’après les résultats, il est évident que les dépressions karstiques concernées furent occupées et exploitées pour des besoins agricoles. Un Système d’Information Géographique a été mis en place pour créer une base de données commune et pour produire des cartes tridimensionnelles. En complément, trois carottages (Z1, Zominthos ; D4A et D4B, Latô) ont été réalisés, d’une profondeur maximale de 9 m. Les sédiments prélevés ont ensuite fait l’objet d’analyses de granulométrie, de susceptibilité magnétique et de diffraction des rayons X. Les résultats mettent en évidence, en plus de l’activité colluviale observée dans la doline (Latô) et l’ouvala (Zominthos), des dynamiques liées à des écoulements intermittents et à des processus de décantation d’argiles provenant de terra rossa. L’alternance de phases d’activité et de stabilité morphologique au cours de l’Holocène récent a contribué à modeler les paysages actuels du centre et de l’est de la Crète, où les Hommes ont su s’adapter au relief escarpé en s’installant à l’intérieur et/ou à proximité des dépressions karstiques.

Haut de page

Notes de la rédaction

Article soumis le 8 juin 2009, accepté le 7 octobre 2009

Texte intégral

The authors would like to thank Stefan Hecht, Bernhard Eitel, Gerd Schukraft and Olaf Bubenzer (Geographical Institute, University of Heidelberg) for their support concerning geophysical prospection, field surveys and laboratory analyses. Moreover, thanks are due to Barbara Brilmayer-Bakti who was in charge of 3D visualistions and IT-based data processing. For helpful information on local archaeology and geology as well as permanent support during the last few years, a special thank you is dedicated to Diamantis Panagiotopoulos (Institute of Cassical Archaeology, University of Heidelberg) and Charalampos Fassoulas (Natural History Museum of Crete, Heraklion). The authors kindly thank Irene Zananiri (Institute of Geology and Mineral Exploration, Athens) for the authorisations for the field campaigns. Thanks to Eric Fouache for providing his coring equipment, as well as to Ms. Apostolakou (Prehistorical and Classical Ephoria of Aghios Nikolaos) for her great interest in the palaeoenvironmental researches undertaken in the area of ancient Latô. Stéphane Kunesch is acknowledged for his help during the field campaigns, his kindness and constant support. The French school of Archaeology in Athens and its Directors (Dominique Mulliez and Veronique Chankowski) financially supported the fieldwork in the Latô/Dreros area. Sincere thanks to the Editor-in-Chief, Gilles Arnaud-Fassetta, as well as to Andy Bicket, George A. Brook, Andrej Mihevc, and two anonymous referees for remarks and helpful comments on the paper. For corrections on the manuscript, a special thank you is dedicated to Kerstin Mewes. This paper is a contribution to the interdisciplinary Zominthos project (University of Heidelberg) and to the research program led by Alexandre Farnoux (university of Paris-Sorbonne Paris 4) to better understand the relationships between the human occupation and the geomorphologic changes occurred during the Holocene in the Mirambello area.

Introduction

1Sediment-filled karst landforms like dolines, uvalas and poljes are of great interest for interdisciplinary research at the human-environmental interface, as they have been of essential economic significance for thousands of years. They serve as favourable agricultural areas, as well as predominant locations for settlement activities and furthermore represent traps for eroded sediments which are accumulated into the dolines and preserved from transport from other regions (Siart et al., 2008 a and b). Important research has been undertaken in the southern France to reveal the links between anthropogenic occupation and land use. They demonstrate that karst depressions were important settlement locations and therefore of high palaeoenvironmental interest (Bruxelles, 2001; Bruxelles et al., 2006 and 2007). However, karst depressions in Greece have barely been used for palaeoenvironmental studies so far due to the lack of information about the thickness of infills along with the assumption of open systems characterised by dynamic sediment flux and storage. This fact also applies to the Cretan Mountains, where filled dolines must be regarded as the only sedimentary records because of area-wide karstification and massive soil erosion (Siart et al., 2009a). Besides some general geomorphological studies (e.g., Fabre and Maire, 1983; Bartels, 1991; Egli, 1993), they have not been investigated in terms of subsurface geometry and sediment fill. Furthermore, no detailed prediction was possible concerning their actual depth structure. As demonstrated by L. Hempel (1991) and H. Poser (1957 and 1976), larger depressions may sometimes be filled with loose sediments to a significant degree, but this finding has not been investigated further. These circumstances raise the question of how much material is deposited within Cretan karst depressions and if the associated sediments generally qualify as geoarchives. The results certainly need to be examined in respect of human-environmental interactions and geoarchaeological implications. Hence, our objective is to apply a multi-method approach in order to scrutinize the geoecological setting of distinct dolines in different regions in Crete, to define their main morpho-structural features and to analyse the geometrical (thickness, subsoil topography), geophysical (electrical resistivity) and geoarchaeological characteristics (artefacts) of their infills. Based on the results, a 3D model of the subsurface topography is to be developed, since no concept of the local buried karst relief exists so far.

Study sites

2In order to obtain supra-regional results, our study focuses on two distinct areas of investigation, both situated in coast-distal regions within the Cretan mountains. The site of Zominthos (fig. 1) is located in the Ida-Oros (Mount Ida), a massif predominantly consisting of Palaeozoic to Mesozoic limestones and dolomites (Fassoulas, 2000). All altitudinal levels from the base up to the highest peak at 2456 m a.s.l. are affected by intense karstification including almost every type of corresponding landforms (e.g., sinkholes, dolines, uvalas and poljes; Siart et al., 2009a). The investigated site itself can be defined as an enclosed depression at an altitude of 1200 m a.s.l. filled with loose sediments and overgrown with an almost abundant vegetation cover. Due to its large spatial extension it tends to form an uvala-type cluster of adjoining hollows. Considering the local topography, the catchment area of Zominthos (approx. 75 ha) is characterised by an alternation of small ridges and several isolated limestone hills flanked by dry valleys with a dominant trend parallel to the orientation of Crete (east-west). They have been formed at low altitudes above sea level prior to the tectonic uplift of the island during the Pliocene and represent the most prevalent locations for the formation of karst depressions (Creutzburg, 1958). Since the doline of Zominthos is in close proximity to a major normal detachment fault (thrust) that separates two different petrographical units (Platy limestone, Tripolitza limestone) forming an acquiclude, it is episodically influenced by surface drainage of several strata bound springs. As for the local settlement history, some of the most outstanding features are multiple remains of the Minoan period, such as the villa of Zominthos (Neopalatial period, ~1650 BC). Not just the mere existence of this huge settlement complex constructed by a culture that colonized and cultivated the area during the Bronze Age, but its location in the immediate vicinity of one of the largest and most impressive karst depressions in the region attracts special geoarchaeological interest. Thus, Mount Ida ideally qualifies for investigating the human-environmental interactions of the past. In contrast to the second millennium BC, the area was abandoned and fallow from the beginning of the Dark Ages (ca. 1050 BC) until recent times. The current climatic and geoecological conditions are very unfavourable for human land use and accordingly the upper limit of modern settlement is located only at about 700 m a.s.l. (Rackham and Moody, 1996; Sakellarakis and Panagiotopoulos, 2006).

Fig. 1 – Geological maps of the Cretan Ida Mountains (A) and Dikti Mountains (B). Major geological units.
Fig. 1 – Cartes géologiques des montagnes crétoises de l'Ida (A) et du Dikté (B). Unités géologiques principales.

Fig. 1 – Geological maps of the Cretan Ida Mountains (A) and Dikti Mountains (B). Major geological units.Fig. 1 – Cartes géologiques des montagnes crétoises de l'Ida (A) et du Dikté (B). Unités géologiques principales.

1: Neogene/Quaternary; 2: ophiolites; 3: Tripolitza unit; 4: phyllite-quartzites; 5: Kalavros schists; 6: Plattenkalk unit; 7: main detachment fault; 8: road system (2 and 3 belong to the upper nappes; 4, 5 and 6 belong to the lower nappes. Subset A: Ida Mountains. Subset B: Dikti Mountains (data source: IGME geological map of Greece, 1983).
1 : Néogène/Quaternaire ; 2 : ophiolites ; 3 : unité de Tripolitza ; 4 : Phyllithe-quartzites ; 5 : schistes de l'unité de Kalavros ; 6 : unité de Plattenkalk ; 7 : décrochement de faille principal ; 8 : système routier (2 et 3 appartiennent aux nappes supérieures ; 4, 5 et 6 appartiennent aux nappes inférieures. A : Montagnes de l'Ida. B : Montagnes du Dikté (données : cartes géologiques IGME, 1983).

3Latô, the second region of investigation, is located in the Dikti-Oros in East Crete (fig. 1) at a distance of about 75 km from Zominthos. Analogous to Mount Ida, the geological units largely consist of limestones, phyllite-quarzites and Neogene deposits (Institute of Geology and Mineral Exploration, 1983). The main study site is well known from 19th century travellers who mentioned the name of Goulas (or Goolas) to identify the settlement of Latô (Spratt, 1865), occupied from Mycenian to Hellenistic times. However, there is no evidence of a subsequent occupation during medieval times. Similar to Zominthos, there is both a close and a spatial relationship between a karstic geoarchive and past human occupation. The study area is situated on the margins of a deeply incised solution doline (approx. 70 m deep) at 278 m a.s.l. While first investigations were conducted by English and French archaeologists in the end of the 19th century (Evans, 1894 and 1895-1896; Demargne, 1901) and the French school of archaeology (Athens, Greece) in the 1960s, recent geomorphological studies focused on several dolines in the south east of ancient Latô (Ghilardi, 2006). Particularly, pottery shards from Minoan to modern times attracted geoarchaeological interest. They were discovered in one of the depressions (D4; fig. 2) along with some anthropogenic infrastructure remains buried under a thick sedimentary overburden. As revealed by the discovery of a ceramic workshop and several ovens, potteries were built in Zominthos as well as Latô (Ducrey and Picard, 1996; Panagiotopoulos, 2007; Wurmser, 2008). Evidently, inhabitants used local material for human purposes, most likely clays extracted from the surrounding depressions. Regarding the recent climatic conditions, both study areas are characterised by the typical Mediterranean seasonality in terms of winter rain and summer aridity. However, due to their elevated locations, a regional climatic modification can be observed, showing increasing amounts of precipitation (approx. 800-1200 mm/a) and lower mean annual temperatures than in the lowlands (Rackham and Moody, 1996). Even though Zominthos and Latô exhibit rather periodical geomorphodynamic activity, corresponding processes may be less accentuated than in coastal areas and subtropical weathering conditions are fairly moderate.

Fig. 2 – Location of the depressions Z (Zominthos) and D4 (Latô) with the coring sites.
Fig. 2 – Localisation des dépressions karstiques Z (Zominthos) et D4 (Latô) et emplacement des carottages.

Fig. 2 – Location of the depressions Z (Zominthos) and D4 (Latô) with the coring sites.Fig. 2 – Localisation des dépressions karstiques Z (Zominthos) et D4 (Latô) et emplacement des carottages.

The depression of Zominthos (image 1, upper left) lies in the north of the Minoan villa, while doline D4 is located 1.4 km SE of the ancient settlement of Latô (image 2, upper right). Detail a (lower left) displays the bottom of the Zominthos uvala including the ERT array E3 (see also fig. 3). Image b (center) indicates the presence of a wall buried by thick sediment accumulations in D4. In the subsurface, two sections can be distinguished, i.e. an upper horizon with badly carved and poorly arranged stones and second layer between 1 and 2 m b.s., in which limestone boulders are well arranged. Detail c (lower right) shows an irrigation system branching to a probable dry spring in D4 (35°10’20.9’’N/25°40’14.9’’E).
La dépression de Zominthos (image 1, en haut à gauche) se situe au nord de la villa minoenne, tandis que la doline D4 se trouve à 1,4 km au sud-est du site archéologique de Latô (image 2, en haut à droite). Le détail marqué « a » révèle la partie basse de l’ouvala de Zominthos (en bas, à gauche), comme le fait le relevé E3 ERT (voir aussi fig. 3). L’image b (au centre) indique la présence d’un mur recouvert par une épaisse couche de sédiments dans la doline D4. Sous la surface, deux sections peuvent être identifiées : dans la partie supérieure, des pierres taillées de manière grossière sont mal agencées, et dans la partie inférieure, entre 1 et 2 m sous la surface, des blocs sont ordonnés de manière géométrique. Le détail c (en bas à droite) indique la présence d’un système d’irrigation relié à une probable source tarie dans la doline 4 (35°10’20,9’’N/ 25°40’14,9’’E).

Materials and methods

4A multi-method approach was undertaken by focusing on initial remote sensing and GIS-applications. Land surface structure, geological setting and petrographic variables were mapped and integrated into a digital database for further analysis. Fieldwork consisted of topographical mapping, geophysical prospection and percussion drilling of sediment cores. Sedimentological and mineralogical analyses were conducted at the Laboratory of Geomorphology and Geoecology (University of Heidelberg, Germany), the Institute of Mineralogy (University of Heidelberg, Germany) and the Laboratoire de Géographie Physique de Meudon (CNRS-UMR 8591, France).

5Geophysical prospection and topographic mapping were associated to GIS-based studies. As demonstrated by S. Hecht (2009), geophysical prospection methods can serve as valuable tools for carrying out geoarchaeological research and investigating the environmental history. Thus, 2D-Earth resistivity tomographies (ERT) based on a 100 electrode system (Geotom) were measured in the depression of Zominthos to evaluate the thickness of the doline fills, their electrical properties and the corresponding stratification by means of Schlumberger configurations due to their high lateral precision (Siart et al., 2010). Processing of resultant data was carried out with RES2DINV software. In order to capture the current topography and to provide a basis for subsequent visualisation, relief data were collected by total station mapping (Leica TPS 700) and GPS measurements during several field campaigns and converted into a high resolution digital elevation model (DEM-spacing: 1.5 m). ArcMap was used for spatially linking the tomographical results to each other and delineating different vertical levels of loose sediment accumulations by vectorisation, before transforming them into the 3rd dimension. By processing the DEM on the basis of ERT-transects, different depth levels were digitally isolated to expose the approximate course of the bedrock and a probable shatter zone of coarse detritus without overlying loose substrate, respectively. In case of areas without geophysical data, resistivities measured in the surroundings were interpolated, thereby creating the subsurface topography of the Zominthos depression with all its features (sinkholes, pinnacles, etc.). In this context, resistivity values of R>1200 Ωm served as crucial thresholds to identify the boundary between bedrock and overburden (Hecht, 2007). The modified DEM was converted into *.tif-format. Boundary walls extending to depth were attached and then textured with 2D profiles in Visual nature studio and 3D studio max in order to generate perspective block diagrams. For subsequent visualisation purposes, additional environmental variables of Mount Ida (e.g., petrography, geomorphology, hydrology and vegetation) were both digitized from maps or satellite images and ground-truthed during fieldwork (Siart and Eitel, 2008). High resolution Quickbird images (pixel size: 0.61 m) were processed with Erdas Imagine 9.1 and used as surface drapes (Siart et al., 2008a and 2009b).

6Vibracoring and sedimentological analysis were based on one 50-mm diametre vibracore (Z1; recovery depth: 9 m b.s.) and two 40-mm diametre cores (D4A and D4B) up to a maximum depth of 5 m b.s. drilled in May 2007 and September 2008. The coring sites were located within two karst depressions (tab. 1) adjacent to the ancient settlements of Zominthos (northward) and Latô (southeastward). The boreholes were documented accurately with GPS. Measurements and the absolute elevation were calculated with Hellenic Military Geographic Service (HMGS) data (accuracy: ±0.70 m). Sediment samples (grain-size fraction <2mm) were collected at 10 cm (80 samples; cores D4A and D4B) and 20 cm intervals (40 samples; core Z1) through the cores and packed into 10 cm3 plastic boxes. Their magnetic susceptibility was measured with a Bartington MS2B dual frequency sensor, at low and high frequencies (χlf, χhf). All measurements were made using the measurement setting of 0.1, with a resolution of 10-5 SI units. Subsequently, the results were related on specific mass of the samples and processed with the Origin and Slater software packages. Grain-size analyses were conducted at the above mentioned intervals on the samples studied for magnetic susceptibility measurements. Size fractions <2 mm and >2 mm were separated by sieving and each was weighed. Particles <2 mm were then dispersed using 0.5% of sodium hexametaphosphate and left in deionised water for two hours (dispersal of clay particles). An exposure to ultrasound was conducted. The grain-size distribution (116 fractions) was measured using a Coulter LS 230 laser granulometre with a range of 0.04 to 2000 μm. The calculation model (software version 2.05) uses Fraunhofer and Mie theory, which is applicable down to a grain-size of about 0.04 µm. For the calculation model, water was used as the medium (RI=1.33 at 20°C). A refractive index in the range of kaolinite for the solid phase (RI=1.56) and absorption coefficients of 0.15 for the 750-nm laserwave length and 0.2 for the polarized wavelengths were applied (Buurman et al., 1996). Samples containing fine particles were diluted with deionised water, while measurements were conducted between 6 and 10% of obscuration and between 50 and 57% PIDS (Polarization Intensity Differential Scattering) obscuration. Clay mineral spectra of selected horizons were analysed by X-ray diffraction (XRD). The mineralogical composition of samples (grain size <63 μm; including silt to avoid bulging at the margins of texture samples with preferred orientation) was determined using a Philips PW 3710 diffractometer (Cu Kα-radiation, 40 kV/30 mA) with scattering angles from 3°2θ to 21.9°2θ. Air-dried specimens should scanning occur after sample preparation, exposed to ethylene glycol vapours for 24 h and heated at 550°C for 16 h. Subsequent analysis and interpretation were based on the EVA 5.0/rev 1 Diffrac Plus software package (Bruker Analytical X-ray systems) and semi-quantitative estimation of components by peak intensity values (e.g., Heim, 1990; Pye, 1992; Nihlén and Olsson, 1995). To help determine the provenance of the clays, insoluble residues from carbonate rocks were investigated by methods described by R. Perrin (1964) and the resulting fraction (<63 μm) was analysed by XRD as described above. Samples of different petrographical units around Zominthos, which also occur in the vicinity of Latô, were used as reference parent materials (Platy- and Tripolitza-limestones, Kalavros schists; see fig. 1 for location of outcrops).

Tab. 1 – Location of sediment cores.
Tab. 1 – Localisation des carottages.

Core id

Absolute elevation (m a.s.l.)

Latitude  (D°M'S''/WGS84 - NUTM34)

Longitude (D°M'S''/WGS84 - NUTM34)

Recovery depth

 (m b.s.)

Z1 (Zominthos)

1176

35°14’58.1’’N

24°53’11.1’’E

9

D4A (Lato)

278

35°10’21.0’’N

25°40’18.0’’E

5

D4B (Lato)

278

35°10’20.2’’N

25°40’17.7’’E

5

Results

7The presentation of results pursues our top-down approach by preliminarily focusing on the mesoscale properties of karst depressions (buried relief and geometry), followed by microscale investigations (local sediment archives, stratification, characteristics of infills).

Geophysical prospection, GIS-based analysis and 3D visualisation

8In order to explore the maximum thickness of the sedimentary overburden, ERT transect E3 was measured on a total length of 100 m within the karst depression of Zominthos (unit electrode spacing: 1 m; Schlumberger array). The resulting data show marked variations in resistivity values in the subsurface and provide information about the composition of sediments, e.g., by designating the transition between two different subsoil media and different substrate characteristics (different colour shades infig. 3). Several zones of distinct resistivities can be detected, such as low values in the shallow subsurface (R<150 Ωm), which generally represent fine-grained sediments, underlain by a higher resistivity zone (R>400 Ωm) in a horizontal alignment (potential anthropogenic wall remains or single limestone slabs and boulders). In fact, the central part of E3 is dominated by extremely low resistivities (R<100 Ωm) that extend almost from the top to the bottom of the cross section to a depth of at least 18 m below surface. As for the subsurface morphology, one can infer a large buried ponor (katavothra in Greek) in the central part having a good underground drainage while being bordered by a bedrock pinnacle to the north. As indicated by the low resistivity values, the subterranean katavothra is almost completely filled by mainly fine-grained loose substrate. The basal sub-layer has higher resistivities (R>1200 Ωm; actual limestone bedrock). This applies to the southern sector of the tomography, which is featured by a subsurface declivity starting from a tributary side valley and dipping straight into the filled ponor. According to the geophysical data from Zominthos (total of 16 cross sections), an approximate boundary zone between loose sediments and fractured underlying bedrock detritus can be identified. Based on this interface, a virtual model displaying the bottom level of the karst depression can be generated. As demonstrated by the resulting 3D model, several buried channels and underground trenches flank the settlement hill with the Minoan villa on top (fig. 4). The general subsurface structure is characterised by numerous solution cavities, sinkholes, bedrock pinnacles and chimneys, accompanied by undulating bedrock ridges typical of karst terrains (Gautam et al., 2000; Roth et al., 2002). According to this ERT-based model and the depth of detected sediment fills, which increases towards the central part of the doline, a hypothetical subsurface runoff pattern can be presumed extending from the south-eastern part of the catchment area to several buried katavothres (fig. 4). It illustrates the preferential underground water flow in the doline fill and highlights areas of high susceptibility to drainage. However, its impact on the formation of the epikarst morphology must not be overrated, since irregularities of the bedrock rather result from cryptokarst processes.

Fig. 3 – Block diagram showing ERT profile E3 as a vertical texture on the high resolution DEM in twofold vertical exaggeration (profile length: 100 m).
Fig. 3 – Bloc diagramme superposant le profil ERT E3 sur le modèle numérique de terrain de haute résolution avec un facteur d’exagération double (longueur du profil : 100 m).

Fig. 3 – Block diagram showing ERT profile E3 as a vertical texture on the high resolution DEM in twofold vertical exaggeration (profile length: 100 m).Fig. 3 – Bloc diagramme superposant le profil ERT E3 sur le modèle numérique de terrain de haute résolution avec un facteur d’exagération double (longueur du profil : 100 m).

The cross section records the deepest part of the karst depression, which consists of a buried sinkhole indicated by the blue colours in the middle of the array. It clearly demonstrates the profoundly heterogeneous and unpredictable subsurface karst morphology, which has also been proved by further ER-tomographies from Zominthos (surface texture: Quickbird MS data; 3D model of Minoan villa).
Le profil E3 révèle la partie la plus profonde de la dépression. Indiquée par la couleur au centre du profil, la dépression est en fait un « entonnoir » recouvert de sédiments. Cela démontre clairement le caractère hétérogène et imprévisible de la surface souterraine, ce qui a été confirmé par d’autres ER-tomographies réalisées à Zominthos (logiciel utilisé : Quickbird MS data ; modèle 3D de la villa minoenne).

Fig. 4 – 3D model of the subsurface karst topography in Zominthos as inferred and interpolated from geophysical tomographies.
Fig. 4 – Modèle 3D de la topographie souterraine de Zominthos, interprétée à partir des résultats géophysiques.

Fig. 4 – 3D model of the subsurface karst topography in Zominthos as inferred and interpolated from geophysical tomographies.Fig. 4 – Modèle 3D de la topographie souterraine de Zominthos, interprétée à partir des résultats géophysiques.

A. Digitally excavated DEM of the depression in a perspective view looking to the NW. B. View to the SE (dashed white lines: potential underground drainage pattern; striped polygon below Minoan villa: backfilling of the settlement hill). The indicated surface (greyish shaded texture) represents the approximate boundary zone between the sedimentary overburden and the solid limestone bedrock. Quickbird MS satellite data were draped on the DEM in the exterior of the depression, while a 3D model of the Minoan villa was inserted on top of the settlement hill using ArcGIS and Visual nature studio software.
A. Modèle numérique de terrain dépourvu de la couverture sédimentaire, bloc orienté au nord-ouest. B. Réseau hydrographique souterrain supposé (le polygone rayé correspond au remplissage sédimentaire localisé derrière le site de Zominthos ; figure orientée au sud-est, lignes avec tirets blancs). La surface surlignée par la couleur grise (ombrée) représente le contact approximatif entre la couverture sédimentaire et le substrat calcaire. Les données satellites traitées grâce au logiciel Quickbird MS et un modèle 3D de la villa minoenne ont été utilisés comme texture de surface autour de la dépression, insertion des données dans ArcGIS et Visual nature studio software.

Sedimentological properties

9Grain-size investigations indicate a homogeneous and dominantly fine-grained substratum of a brownish colour (10 YR 4/3; Munsell soil color in wet condition) and the same granulometrical composition with only minor variations in the uppermost parts of all cores. Only at about 1.5-2 m depth does the sediment become a reddish color (7.5 YR 4/6). However, Z1 is composed of remarkable amounts of gravel-rich sediments, which are totally absent in the sediments from Latô. Besides the uppermost soil horizons that mainly consist of fine silts (fig. 5A; mean size of 30 µm with a leptokurptic curve on the left indicating the presence of coarser particles, i.e., medium silts to medium sands ranging from 40-200 µm), this applies to the complete sediment column of Z1. These poorly sorted strata between 2 and 7.7 m b.s. contain infrequent fine-grained and laminated layers, which are characterised by higher amounts of sand while gravel is almost absent (e.g., Z1-2.1 or 2.9). Both sediment varieties constitute the major part of the sediment core and must be considered representative for the deposits at Zominthos. At the base of the section, massive limestone detritus (bench of lithoclasts) was detected. The near surface horizons of D4A and D4B are quite similar to Z1, showing brownish fine-grained properties. However, they include a high concentration of rounded nodules, a fact which can be well observed by the grain-size distribution showing small peaks in the coarser fractions (50-200 µm; fig. 5B). The black nodules originate from local petrographical units, particularly from a limestone breccia with carbonate cement rich in heterogeneous pebbles and gravels. The grain-size distribution generally shows poorly-sorted clay material throughout the entire sediment columns of the Latô cores with only slight variation. The modal value does not exceed 6 µm and the mean size is commonly below 4 µm in diametre.

Fig. 5 – Grain-size distribution of selected samples analysed by laser granulometre.
Fig. 5 – Distribution granulométrique des échantillons analysés par la méthode laser.

Fig. 5 – Grain-size distribution of selected samples analysed by laser granulometre.Fig. 5 – Distribution granulométrique des échantillons analysés par la méthode laser.

A. Grain-size distribution for core Z1. Within the first two metres (from Z1-010 to Z1-190), the samples close to the surface indicate a modal index comprised between 10 and 40 µm. The peaks observed in the coarsest fractions are sandy particles (poorly rounded). B. Cores D4A and D4B and displaying a modal index between 1 and 6 µm on average. The peaks observed in the coarsest fractions are induced by small black rounded nodules.
A. Distribution granulométrique pour le sondage Z1. Dans les deux premiers mètres (de Z1-010 à Z1-190), le mode granulométrique indique des valeurs comprises entre 10 et 40 µm. Les pics observés dans les fractions les plus grossières correspondent à des sables (émoussés). B. Carottages D4A et D4B dans lesquels le mode granulométrique oscille entre 1 et 6 µm. Les pics dans les fractions les plus grossières sont liés à la présence de nodules très émoussés de couleur noire.

10The magnetic susceptibility measurements confirm similarities between all cores by showing high signals in the brownish near surface horizons (fig. 6 and fig. 7; Z1: 35-130x10-8 m3/kg; D4A and D4B: 25-85x10-8 m3/kg). Underneath, susceptibility levels have medium values of about 40x10-8 m3/kg on average. Particularly Z1 exhibits this two parts structure, having a remarkable change of values at 2 m b.s. with oscillating signals between 40 and 65x10-8 m3/kg, which occur in the reddish strata down to a depth of 7.7 m (fig. 6). However, as several different peaks are apparent, e.g., at 3.1 m b.s. (χlf: 65x10-8 m3/kg) and 7.7 m b.s. (χlf: 50x10-8 m3/kg), the presence of fine ‘viscous’ ferrimagnetic grains (~0.02 μm in diametre; Oldfield, 1991; Maher, 1988) is suggested. Within the basal part of the core (7.7-9 m b.s.), a bench of lithoclasts produces the lowest signals with values below 40x10-8 m3/kg. The χfdlfhf) for core Z1 indicates a high susceptibility ranging from 3 to 12. As for depth variations, the magnetic parameters measured in cores D4A and D4B (fig. 7) resemble those of Z1, attaining 25-85x10-8 m3/kg. Between 2 and 5 m b.s., there is a lower signal by tendency (D4A: 25-65x10-8 m3/kg; D4B: 20-70x10-8 m3/kg), which slightly re-increases depthwise. In both cases, a maximum of magnetic susceptibility values is displayed at 3.75 m b.s. (χlf: 55x10-8 m3/kg) and 4.75m b.s. (χlf: 55 x10-8 m3/kg for core D4A; χlf: 75 x10-8 m3/kg for core D4B). In this context, the oscillating signal is not related to the grain size distribution, since clay-rich sediments were detected throughout the profile almost exclusively. It is also disconnected from the geological background. The χfdlfhf) of D4A and D4B indicates high values between 4 and 6 in depths of 3.75 and 4.75 m b.s., which point to fine ‘viscous’ ferrimagnetic grains analogous to Z1. The comparison of values at low frequency clearly shows similar features in all three cores. The uppermost horizons (0 to 2 m b.s.) are characterised by an enhanced signal (fig. 8), oscillating from 20-135x10-8 m3/kg. In contrast, the underlying strata show lower values, with several infrequent susceptibility maxima. Most notably, these χlf peaks occur in all cores (~3.5 and 4.5 m b.s.). As for their origin, allochthonous contribution of sediment with higher contents of magnetic minerals (aeolian input of volcanogenic particles) or development of palaeosoils (land use at different periods) must be considered.

Fig. 6 – Core profile and magnetic parameters of core Z1 (Zominthos).
Fig. 6 – Profil stratigraphique et paramètres magnétiques pour le sondage Z1 (Zominthos).

Fig. 6 – Core profile and magnetic parameters of core Z1 (Zominthos).Fig. 6 – Profil stratigraphique et paramètres magnétiques pour le sondage Z1 (Zominthos).

From left to right: stratigraphic information based on colour as well as magnetic susceptibility at low (χlf) and high (χhf) frequencies are plotted with regard to depth. χ is a proxy for the magnetic mineral concentration, which is often approximately equated to the concentration of magnetite. Moreover, it can be considered as grain-size dependent (Dekkers, 1997; Dearing, 1999). χlfhffd) measurements are sensitive to ultrafine magnetic grains and high values designate the presence of fine ‘viscous’ ferrimagnetic grains (~0.02 μm in diametre). Fd (in%) is the relative contribution of fine viscous particles to the total ferromagnetic assemblage, high corresponding percentage values point to wide distributions of ultrafine grains. The absolute elevation in m a.s.l. is indicated in brackets.
De la gauche vers la droite : la stratigraphie générale est fondée sur des différences de couleurs ; les valeurs de susceptibilité magnétique à basse (χlf) et haute fréquence (χhf) sont exprimées en fonction de la profondeur. χ est un paramètre pour étudier la concentration des minéraux magnétiques, qui est généralement égale à la concentration de magnétite, il existe un lien avec la granulométrie (Dekkers, 1997 ; Dearing, 1999). Les mesures pour χlfhffd) sont sensibles à la présence de grains magnétiques ultrafins et les valeurs élevées indiquent l’existence de grains ferrimagnétiques de très petite taille (~0,02 μm de diamètre). Fd (en %) est la contribution relative de fines particules dans l’assemblage total des minéraux ferromagnétiques : de forts pourcentages indiquent la présence de grains très fins. L’altitude absolue en mètres par rapport au niveau de la mer est indiquée entre parenthèses.

Fig. 7 – Core profiles and magnetic parameters of cores D4A and D4B (Latô).
Fig. 7 – Profils stratigraphiques et paramètres magnétiques pour les carottages D4A et D4B (Latô).

Fig. 7 – Core profiles and magnetic parameters of cores D4A and D4B (Latô).Fig. 7 – Profils stratigraphiques et paramètres magnétiques pour les carottages D4A et D4B (Latô).

From left to right: general stratigraphy based on colour as well as magnetic susceptibility at low (χlf) and high (χhf) frequencies are plotted with regard to depth (for details on χ, χfd and Fd (in%) see caption of fig. 6); absolute elevation in m a.s.l in brackets).
De la gauche vers la droite : la stratigraphie générale est fondée sur les différences de couleurs ; les valeurs de susceptibilité magnétique à basse (χlf) et haute fréquence (χhf) sont exprimées en fonction de la profondeur (pour les détails concernant χ, χfd, and Fd (en %), se référer à la légende de la fig. 6) ; l’altitude absolue en mètres par rapport au niveau de la mer est indiquée entre parenthèses.

11As shown by the XRD-analyses of clay mineral spectra, the sediment samples from Z1, D4A and D4B contain quartz, micaceous clay minerals (illite, muscovite), kaolinite, chlorite, mixed-layer minerals (e.g., corrensite) and vermiculite. There is little change with depth with only minor variations of trace components (tab. 2). All analysed specimens are characterised by phyllosilicates like muscovite or illite, which originate from parent limestones or mica weathering (Moore and Reynolds, 1997). Constituting a major component in Z1 and a minor component in D4A and D4B, they were identified in the compounds due to typical 5 Å and 10 Å peaks. The significant 7 Å peak in most samples collapsed due to the heat treatment at 550°C and therefore indicates the presence of kaolinite. Contrary to the sediment samples, all limestone residuals lack kaolinite, while micas dominate. Consequently, kaolinite minerals are either neoformed by crystallisation from solutions, transformed from existing minerals or wind-blown from other regions (Martín-García et al., 1998). Chlorite was only detected in Z1, indicating inheritance from parent rocks (primary chlorites), in-situ formation (weathering of illites and vermiculites) or autochthonous origin (external input of primary and/or secondary chlorites). It was identified in the XRD plots subsequent to heat treatment, which, in contrast to the behaviour of pedogenic chlorites, amplifies the 14.2 Å peak of primary chlorites while extinguishing or weakening the reflexes of higher order (Heim, 1990). Further common clay minerals must be considered as poorly defined mixed-layer varieties that produce indistinct peaks and a raised background in the 10-14 Å area (e.g., corrensite). Even though quartz is ubiquitous in all samples, its total amount is difficult to assess as only the corresponding 4.26 Å peak is visible in the diffractograms due to maximum scattering angles of 21.9°2θ. Vermiculite appears in traces or as a minor constituent in the samples of Z1, as highlighted by the occasional 14-15 Å reflexes and their collapse/shift to about 9.5 Å after heat treatment. Being a characteristic mineral in Mediterranean soils (Moore and Reynolds, 1997), it mostly appears in well drained locations where the parent material contains micas, illites or micaceous schists. Vermiculite also originates from destruction of primary chlorites under subtropical weathering conditions and moderate acidification (Macleod, 1980; Scheffer and Schachtschabel, 2002). The insoluble residuals of parent rocks are predominantly composed of mica and chlorite with traces of quartz whereas kaolinite was absent.

Tab. 2 – Semi-quantitative estimation of the clay-mineralogical composition
Tab. 2 – Estimation semi-quantitative de la composition minéralogique des argiles

Sample no. &

 depth (m)

Illite - Mica

Kaolinite

Chlorite

Mixed-layer

Quartz

Vermiculite

Z 1 – 0.15

x

xx

x

  

x

  

Z 1 – 0.80

xx

xx

x

  

x

  

Z 1 – 1.70

x

xx

  

x

x

  

Z 1 – 2.15

x

xx

  

o

x

  

Z 1 – 2.65

xx

xx

  

o

o

  

Z 1 – 4.20

xx

x

  

x

x

Z 1 – 5.65

xx

x

  

x

x

Z 1 – 7.40

xx

xx

  

x

x

  

Z 1 – 8.85

xx

  

  

  

xx

  

Z 1 – 8.95

xx

  

xx

  

xx

  

D4A – 1.90

x

xx

  

o

xx

  

D4A – 3.00

x

xx

  

o

xx

  

D4A – 3.50

x

xx

  

o

xx

  

D4A – 4.12

x

x

  

o

xx

  

D4B – 1.35

x

x

  

  

xx

  

D4B – 2.45

x

xx

  

  

xx

  

D4B – 2.60

x

xx

  

o

xx

  

D4B – 3.00

x

xx

  

o

xx

  

D4B – 3.60

x

xx

  

  

xx

  

D4B – 4.12

x

xx

  

o

xx

  

KAL

xx

  

xx

  

x

  

PL

xx

  

xx

  

xx

  

TL

xx

  

  

  

x

  


(fractions argileuses et limoneuses des horizons sélectionnés ; xx : composant majeur ; x : composant mineur ; o : traces ; abréviations : KAL =schistes de la nappe de Kalavros, PL =calcaires de la nappe de Plattenkalk, TL =calcaires de la nappe de Tripolitza).

Fig. 8 – Magnetic susceptibility signal at low frequency (χlf) for cores Z1, D4B and D4A.
Fig. 8 – Valeurs de susceptibilité magnétique à basse fréquence (χlf) pour les carottages Z1, D4B et D4A.

Fig. 8 – Magnetic susceptibility signal at low frequency (χlf) for cores Z1, D4B and D4A.Fig. 8 – Valeurs de susceptibilité magnétique à basse fréquence (χlf) pour les carottages Z1, D4B et D4A.

All values are expressed according to depth (ordinate). The high values in the upper horizons are in significant contrast to the lower susceptibilities in deeper profile parts. This general trend can be observed in both the cores from Zominthos and Latô.
Toutes les valeurs sont exprimées par rapport à la profondeur. Les valeurs élevées observées dans les horizons supérieurs contrastent avec les faibles valeurs obtenues dans la partie inférieure. Cette tendance générale peut être observée dans les carottages de Zominthos et de Latô.

Discussion

Subsurface morphology of karst depressions

12Sediment fills in Cretan karst depressions have not previously been studied in detail of little information on their thickness and value as geoarchives. Rather, extensive areas of heavily fissured limestones suggested that dolines and poljes were open systems with a considerable loss of sediment. On the contrary, our results from ERT and vibra-coring reveal a very different situation, since in Mount Ida overburdens of up to 20 m of loose materials can be expected. According to these findings, the subterraneous karst relief is buried under a thick sediment cover. This fact can be confirmed by a natural sediment outcrop within the katavothra of the Latô depression (fig. 2), and therefore proves the important reservoir function of dolines both in the central and the east Cretan mountain ranges. Regarding the methodological interface between geophysical and geomorphological data, their GIS-based processing and their digital visualisation, some hypotheses need to be considered because virtual 3D models are not able to display the actual buried karst relief in detail. Furthermore, such digital images cannot (and should not) represent a stage of landscape development at a certain point in time. Instead, they provide an impression of the heterogeneous shape of the subrelief, allow for a synopsis of results obtained by fieldwork and represent a valuable further step of interpretation, which acknowledges the third dimension. Most notably, these models help to assess the spatial distribution of karst features as well as their sedimentary infills. As for the enclosed depression of Zominthos, the subsurface relief appears totally different from the local surface topography with its rather smooth and slightly inclined slopes. Our results reveal an extremely heterogeneous subsurface morphology by highlighting strong variations in resistivity along with remarkable differences in the depth of the underlying bedrock, which must be considered as characteristic for karst terrains (Deceuster et al., 2006). The detected subterranean declivities of the epikarst zone are occasionally very steep and form scarp-like slopes (fig. 3). Regarding the genesis of the buried doline, a significant impact of cryptokarst processes must therefore be taken into account. Filled depressions may consequently exhibit completely distinct morphologic properties at even the smallest scales. As shown by adjacent areas where soils are eroded and washed out whereas cavities are increasingly filled with loose materials, sediment dynamics are frequently very intense. Hence, the investigated dolines must be regarded as unpredictable and highly active geomorphodynamic systems, while attention has to be paid to the fact that the sedimentary fills may be discontinuous and interrupted by hiatuses. Seen from both a geomorphological and a mesoscale viewpoint, the karst depression of Zominthos is one of the major sediment reservoirs in the northern Ida Mountains, a fact which is documented by the large amount and the impressive depth of loose infills. Due to its specific location bordered and encompassed by the major detachment fault to the south and the east (fig. 1), it represents the main sediment trap within the catchment area and, thus, prevents colluvials and soils from further transport into lower parts of the island. The same applies to doline D4, which serves as the principal sediment reservoir in the Latô area. As a consequence, both sites qualify for investigating the environmental history and offer promising prospects for further studies.

Sedimentological characteristics and provenance of doline fills

13The existence of two different sediment types with distinct granulometrical properties must be regarded as the specific feature of Z1, as both have a separate origin. Grain-size analyses provide evidence for the dominating heterogeneous strata with poorly sorted material from top to bottom. Therefore, the term colluvial-type fills applies to the karstic depressions around Zominthos because they share characteristic attributes such as loose, barely or non-stratified, heterogeneously sized and generally unsorted materials that suggest former pedogenetic processes (Leopold and Völkel, 2007). This fact is supported by their heterogeneous components of autochthonous and allochthonous provenance (e.g., volcanogenic heavy minerals and glass shards; Siart et al., 2008b). Corresponding strata predominate in the basal part of Z1, particularly in depths below 4.3 m below surface, and record geomorphologic activity under dry conditions with increased periodicity of annual precipitation and a sparse vegetation cover (Günster and Skowronek, 2001). In contrast, occasional interbedded fine-grained materials clearly indicate semi-fluvial dynamics, which are shown by lamination textures resulting from fining-upward sequences in consequence of sedimentation in a stagnant water body. They mainly occur in the central part of Z1, where they alternate with coarse horizons while being less frequent in greater depths. Indeed, the term ‘fluvial’ applied in this study relates to low-energy and water-induced short-distance transport in combination with filled shafts, e.g., superficial discharge caused by insoluble residuals and/or clayey material. One of its origins can be traced back to strata-bound springs near the uvala of Zominthos (Agia Marina). The graded layers must have been accumulated during periods of geomorphologic stability with reduced erosion intensity and slow surface discharge velocities (wetter conditions, lower periodicity of annual precipitation, closed vegetation cover). Such opposed geomorphodynamic phases, which show alternating regimes rather than just a singular change during the mid- and late-Holocene, occurred episodically in the Ida Mountains (Siart et al., 2009b). However, ponding of water within karst depressions is also a recent phenomenon that commonly occurs during wintertime due to local precipitation maxima. Considering Zominthos, the existence of an autochthonous and clayey loam derived from in-situ weathering of carbonates cannot be confirmed, as with the doline of Latô. The results from D4A and D4B highlight the predominance of extremely fine-grained materials that are only occasionally mixed with coarser detritus (black nodules) from local breccia outcrops. After chemical dissolution of the carbonated breccia cement, the residual nodules are transported into the lower part of the doline subsequent to stronger precipitation or torrential events and finally buried by the red clay sediments. Even though such sand-sized components indicate colluvial transport to some extent, the actual process of colluviation is generally less pronounced in Latô. These distinctions between the two study sites are related to different geomorphodynamic processes and petrographic factors, proving that Zominthos experienced rather stronger colluvial events or phases, a fact that might be supported by the existence of its larger catchment area (approx. 75 ha) and large amounts of relatively coarse-grained detritus in the sediment column derived from weathering of chert-bearing limestone outcrops nearby. Instead, Latô is characterised by a fairly isolated location that has been protected from massive erosion as well as transport and accumulation of larger quantities of sediments (catchment area about 10 ha). Residual clays, which represent typical remains of non-carbonate impurities from limestones, and associated terra rossa soils constitute the major proportion of the corresponding sediment fills. Even though both investigated sites have their origin in limestone dissolution and massive karstification, they must be considered as different types of landforms (e.g., in terms of size, catchment area, geomorphodynamic processes), which have a major impact and a distinguishing influence on the accumulation as well as nature of their sedimentary infills.

14Our magnetic susceptibility measurements strongly suggest that the high signals are caused by the presence of fine ‘viscous’ ferrimagnetic grains (~0.02 μm in diametre). Furthermore, the Fd (in%) shows large percentages, which vary throughout the entire profile of each core. This finding indicates a widespread distribution along with a massive contribution of ultrafine grains to the total ferromagnetic assemblage (Oldfield, 1991; Dearing et al., 1996). Increased values in the near surface must be considered as a result of recent land use and current soil formation, which generally manifests itself in pronounced susceptibilities due to the genesis of ferromagnetic minerals (Hanesch and Scholger, 2005). Further maxima (Z1-3.1 and 7.7 m b.s.; D4A/D4B-3.75 and 4.75 m b.s.) can, on the one hand, be explained by the development of palaeosols within the sediment column related to former land use practices within the depressions (e.g., horticulture, irrigation and animal husbandry; Sakellarakis and Panagiotopoulos, 2006; Siart et al., 2008b). Both the geophysical prospection of the subsurface in Zominthos (fig. 4) and several archaeological field observations from doline D4 in Latô provide evidence for the capture of water from springs and/or slopes (fig. 2c), thus proving the significance of karst depressions for human purposes like agriculture in ancient times. On the other hand, the presence of volcanogenic sediments with higher contents of magnetic minerals might produce peaks in susceptibility, since the method is very sensitive to the Fe-rich properties of tephra. As stated by C. Siart et al. (2008b), volcanogenic ash from the 3.6 ka Minoan eruption of Santorini was, contrary to the hitherto established opinion, also transported into the Cretan Mountain ranges and contributed to the composition of local sediments around Zominthos. Corresponding heavy minerals and glass shards were detected in sediment cores adjacent to Z1 in a depth of up to 10 m b.s. and therefore prove the relatively young age of doline infills. Despite the scarcity of datable material in the cores Z1, D4A, and D4B - a fact that may complicate the reconstruction of the environmental history to a certain degree -, such mineralogical results serve as alternative time markers. They qualify as essential geochronological records and suggest an estimated sedimentation rate of ~2.7 m/ka (Siart et al., 2009b). As ERT-results from Zominthos prove a very heterogeneous morphology of the bedrock, which has to be seen in conjunction with a potential discontinuity of sediment archives, geochronological methods are indispensable for palaeoenvironmental investigations. Since the fallout of tephra from Thera predominantly occurred in eastern Crete as well as in coastal-proximate regions at lower altitudes (Boekschoten, 1971; Hempel, 1994; Bruins et al., 2008), its appearance in the doline-fills of Latô (red clays) seems possible as well. However, the igneous minerals found in the loose sediments of Zominthos are dispersed through the entire sediment column of the depression and do not occur in one single ash layer (Siart et al., 2009b). This fact must be considered as a further proof for the massive colluvial activity in Mount Ida subsequent to the Minoan eruption during the late Bronz Age. In this context, the peaks in the magnetic susceptibility measurements of several strata from Z1, D4A and D4B are most likely the result of the colluvial enrichment of volcanogenic minerals in comparison to over- and underlying horizons.

15As demonstrated by XRD analyses of clay minerals, the polygenetic character of the pedo-sedimentary complexes can be confirmed. Mica, chlorite and quartz occurring in both the parent material and the loose sediments suggest an authigenic provenance. Kaolinite minerals that occur in the colluvial fills but are absent in limestone residuals must either be derived from neoformation or deposited by wind as they are absent in parent limestones (Durn et al., 1999; Martín-García et al., 1998). According to previous investigations on sediments from Mount Ida, both the moderate pH values in local sediment cores and the currently temperate climatic conditions in the study area due to its high elevation at 1200 m a.s.l. point to little or no kaolinitization (Nihlén and Olsson, 1995; Siart et al., 2008b). Since several studies verified the polygenetic attributes of terra rossa type sediments, the formation of kaolinite could also be of Neogene age (Bronger and Bruhn-Lobin, 1997; Jahn, 1997). In particular, this fact has to be considered for Latô due to the quantitative prevalence of the 1:1 layer phyllocsilicate in D4A and D4B. However, given the fact that it occurs in remarkable amounts while it is dispersed all over the sediment column, an exclusive in-situ genesis is not plausible. Considerable quantities must derive from windblown dust (comp. Macleod, 1980; Rapp and Nihlén, 1986; Pye, 1992; Mattson and Nihlén, 1994), a phenomenon which can be observed annually in the kokkinovrokhí (red sediment rain) on snow patches in Crete (Rackham and Moody, 1996). Vermiculite and interstratified minerals were not found in D4A, D4B or local rock samples. Consequently, they must have been neoformed and/or transformed during mica weathering under the mountainous climatic conditions in Mount Ida. The distinctive peak for chlorite indicates the presence of mainly primary types. They are related to the Kalavros schists, which crop out around Zominthos but are not present at Latô. All in all, the clay mineral spectra only show slight variations. In the case of in-situ soil formation, a stronger diversification between horizons should be observable (Delgado et al., 2003; Priori et al., 2008), a fact that points significantly to the colluvial properties of the fills.

Conclusion

16Regarding the investigation of karst depressions in Crete and their subsurface structure in particular, the application of geophysical techniques performs excellently and represents a valuable methodical approach. ERT and SRT profiles from Zominthos prove that the dolines and uvalas in Mount Ida are filled with remarkable amounts of loose materials several decametres thick. Contrary to the hitherto established opinion, these hollows are not just open systems but also serve as storage basins for colluvial sediments and preserve them from further transport. Hence, they are important palaeoenvironmental archives. The measured tomographies also reveal an unpredictable buried karst morphology, which requires certain precautions and a versatile methodologic strategy in order to investigate. Moreover, the combined mineralogical and sedimentological investigations underline the heterogeneous characteristics of sediment fills in Cretan karst depressions. Our findings demonstrate that petrographical conditions and the local relief determine the genesis and also the type of resulting landform in terms of size, catchment area, type of sedimentary infill and local geomorphodynamic processes. The sediments in micro-scale dolines and meso-scale uvalas may consequently differ with regard to grain-size, magnetic susceptibility and mineralogical composition. Besides prevalent residual clays (Latô), colluvial and polygenetic attributes can be established for the loose materials, particularly in Zominthos. The reason for the multi-provenance character of these colluvials lies in climatic changes, tectonic activity, heterogenic petrographical units, anthropogenic land use and deforestation (Durn et al., 1999; Siart et al., 2008b). In addition, investigating the magnetic parameters of the sediments produces information on the contribution of very fine grain-sizes to the total of the ferromagnetic assemblage. As for their origin in Latô, the aeolian input of volcanic tephra must be taken into consideration, since the ash fallout of the Minoan 3.6 ka event has already been verified for Zominthos.

17With regard to the geoarcheological significance of the sites, our investigations reveal clear evidence for anthropogenic land use within both depressions. The results provide an important step towards reconstructing the human-environmental interactions in Central and East Crete, since the mineralogical records as well as the subsurface archaeological remains relate to the Minoan and the Hellenistic period and therefore absolutely coincide with the climax of human occupation in Zominthos and Latô. The corresponding dolines and uvalas obviously represent the most favourable locations fur human purposes because they are uniquely qualified regarding hydrology (water capture of springs for drainage and irrigation), geomorphology (plane bottom suitable for agriculture, ubiquitous clay deposits for pottery manufacturing), ecology (sufficient amounts of soil for arable farming) and topography (protective location). This fact becomes even more important, as large parts of Mediterranean karst regions are generally unsuitable for land use due to massive lack of soils, extensive bedrock outcrops, severe land degradation and scarcity of surface waters (Siart et al., 2008a). Our interdisciplinary approach based on geophysical investigations, GIS studies, field work and sedimentological analyses contributes to a better understanding of the function of karst depressions as geoarchives and provides new insight into the adaptation of ancient human societies to the rough mountain conditions in the Ida and Mirambello/Dikti area. In order to validate and confirm our findings, to gain more precise information about the environmental history and to develop a chronostratigraphical record, datings are required in addition to further mineralogical, geomorphological and geochemical analyses. However, the geophysical and sedimentological results obtained (thus far) offer promising prospects for future geoarchaeological research in karstic regions by proving the suitability of dolines and uvalas as important sediment reservoirs.

Haut de page

Bibliographie

Bartels G. (1991) – Karstmorphologische Untersuchungen auf Kreta. Erdkunde 45, 27-37.

Boekschoten G. (1971) – Quaternary tephra on Crete and the eruptions of the Santorin Volcano. In Strid A. (Ed.): Evolution in the Aegean. Opera Botanica 30, 40-48.

Bronger A., Bruhn-Lobin N. (1997) –Palaeopedology of Terrae rossae-Rhodoxeralfs from Quaternary calcarenites in NW Morocco. Catena 28, 279-295.

Bruins H., MacGilivray J., Synolakis C., Benjamini C., Keller J., Kisch H., Van der Klügel A., Pflicht J. (2008) –Geoarchaeological tsunami deposits at Palaikastro (Crete) and the Late Minoan IA eruption of Santorini. Journal of Archeaological Science 35, 191-212.

Bruxelles L. (2001) – Reconstitution morphologique du Causse du Larzac (Larzac central, Aveyron, France) - Rôle des formations superficielles dans la morphogenèse karstique. Karstologia, 38-2, 25-40.

Bruxelles L., Colonge D., Salgues T. (2006) – Morphologie et remplissage de dolines du Causse Martel. Karstologia, 47, 21-32.

Bruxelles L., Simon-Coiçon R., Guendon J.-L., Ambert P. (2007) – Formes et formations superficielles de la partie ouest du Causse de Sauveterre. Karstologia, 49, 1-14.

Buurman P., Pape T., Muggler R.C.C. (1996) – Laser grain-size determination in soil genetic studies: Practical problems. Soil Science 162, 211-218.

Creutzburg N. (1958) – Probleme des Gebirgsbaues und der Morphogenese auf der Insel Kreta. Freiburger Universitätsreden N.F. 26, Freiburg, 47 p.

Deceuster J., Delgranche J., Kaufmann O. (2006) – 2D cross-borehole resistivity tomographies below foundations as a tool to design proper remedial actions in covered karst. Journal of Applied Geophysics 60, 1, 68-86.

Dearing J.A. (1999) – Magnetic susceptibility. In Walden J., Oldfield F., Smith J. (Ed.), Environmental Magnetism, a practical guide. Quaternary Research Association Technical Guide n° 6, London, 35-62.

Dearing J.A., Dann R.J.L., Hay K., Lees J.A., Loveland P.J., Maher B.A., O'Grady K. (1996) – Frequency-dependent susceptibility measurements of environmental materials. Geophysical Journal International 124-1, 228-240.

Dekkers M.J. (1997) – Environmental magnetism: an introduction. Geologie en Mijnbouw 76, 163-182.

Delgado R., Martín-García J., Oyonarte C., Delgado G. (2003) – Genesis of the terrae rossae of the Sierra Gádor (Andalusia, Spain). European Journal of Soil Science 54, 1-16.

Demargne P. (1901) – Les ruines de Goulas ou l’ancienne ville de Latô en Crète. Bulletin de Correspondance Hellénique, 25, 282-307.

Ducrey P., Picard O. (1996) Recherches à Latô VII. La rue Ouest. Habitations et défense. Bulletin de Correspondance Hellénique, 120, 721 p.

Durn G., Ottner F., Slovenec D. (1999) – Mineralogical and geochemical indicators of the polygenetic nature of terra rossa in Istria, Croatia, Geoderma 91, 125-150.

Egli B.R. (1993)Ökologie der Dolinen im Gebirge Kretas (Griechenland). Ph.-D thesis, University of Zürich, 276 p.

Evans A. (1894) – From Crete to Peloponnese. Journal of Hellenic Studies 14, 277-280.

Evans A. (1895-1896) – Goulas: the city of Zeus. Annual of the British school at Athens 2, 169-194.

Fabre G., Maire R. (1983) – Néotectonique et morphogenèse insulaire en Grèce : Massif du Mont Ida (Crète). Méditerranée, 2, troisième série, tome 48, 39-49.

Fassoulas C. (2000)Field guide to the geology of Crete. Natural History Museum of Crete, Heraklion, 103 p.

Gautam P., Surendra R., Hisao A. (2000) – Mapping of subsurface karst structure with gamma ray and electrical resistivity profiles: a case study from Pokhara valley central Nepal. Journal of Applied Geophysics 45, 97-110.

Ghilardi M. (2006)Contribution à l’étude géomorphologique du site de Latô (Crète – Grèce), dates : 31 juillet / 11 août 2006. Report for the French School of Archaeology in Athens (Greece), 22 p.

Günster N., Skowronek A. (2001) – Sediment-soil sequences in the Granada Basin as evidence for long- and short-term climatic changes during the Pliocene and Quaternary in the Western Mediterranean. Quaternary International 78, 17-32.

Hanesch M., Scholger R. (2005) –The influence of soil type on the magnetic susceptibility measured throughout soil profiles. Geophysical Journal International 161, 50-56.

Hecht S. (2007) – Sedimenttomographie für die Archäologie, Geoelektrische und refraktionsseismische Erkundungen für on-site und off-site studies. In Wagner G.A. (Ed.) : Einführung in die Archäometrie. Berlin, Heidelberg, 95-112.

Hecht S. (2009) – Viewing the subsurface in 3D: Sediment tomography for (Geo-)Archaeological prospection in Palpa, Southern Peru. In Reindel M., Wagner G.A. (Ed.) : New Technologies for Archaeology, Heidelberg, 87-102.

Heim D. (1990)Tone und Tonminerale. Enke, Spektrum, Berlin, Heidelberg, 157 p.

Hempel L. (1991) – Forschungen zur physischen Geographie der Insel Kreta im Quartär. Ein Beitrag zur Geoökologie des Mittelmeerraumes. Abhandlungen der Akademie der Wissenschaften in Göttingen 42, Göttingen, 171 p.

Hempel L. (1994) – Der Vulkanismus von Santorin und die minoische Kultur Kretas : Eine geoökodynamische Analyse für das 17. Jahrhundert vor Christus. Petermanns Geographische Mitteilungen 138, 131-146.

Institute of Geology and Mineral Exploration (1983) – Geological Map of Greece 1 :500.000, Athens (Greece).

Jahn R. (1997) – Bodenlandschaften subtropischer mediterraner Zonen. In Blume H.P., Felix-Henningsen W., Fischer H.G., Frede R., Horn R., Stahr K. (Ed.) : Handbuch der Bodenkunde (3.4.5.4), 1-27.

Leopold M., Völkel J. (2007) – Colluvium: definition, differentiation and possible suitability for reconstructing Holocene climate data. Quaternary International 162-163, 133-140.

Macleod D. (1980) – The origin of the red Mediterranean soils in Epirus, Greece, Journal of Soil Science 31, 125-136.

Maher B.A. (1988) – Magnetic properties of some synthetic sub-micron magnetites. Geophysical Journal-Royal Astronomical Society 94-1, 83-96.

Martín-García J., Delgado G., Párraga J., Bech J., Delgado R. (1998) – Mineral formation in micaceous Mediterranean Red soils of Sierra Nevada, Granada, Spain. European Journal of Soil Science 49, 253-268.

Mattson J., Nihlén T. (1994) – The transport of Saharan dust to southern Europe: a scenario. Journal of Arid Environments 32, 111-119.

Moore D., Reynolds R. (1997)X-ray diffraction and the identification and analysis of clay minerals. Oxford University Press, Oxford, 400 p.

Nihlén T., Olsson S. (1995) – Influence of aeolian dust on soil formation in the Aegean area. Zeitschrift für Geomorphologie N.F. 39, 341-361.

Oldfield F. (1991) – Environmental magnetism: a personal perspective. Quaternary Science Reviews 10, 73-85.

Panagiotopoulos, D. (2007) – Minoische Villen in den Wolken Kretas. Antike Welt 38, 17-24

Perrin R. (1964) – The analysis of chalk and other limestones for geochemical studies. In Society of the chemical industry (Ed.): Analysis of calcareous materials. Journal of the Society of Chemical Industry, 18, London, 207-221.

Poser H. (1957) – Klimamorphologische Probleme auf Kreta. Zeitschrift für Geomorphologie N.F. 1, 113-142.

Poser H. (1976) – Beobachtungen über Schichtflächenkarst am Psiloriti (Kreta). Zeitschrift für Geomorphologie N. F. Supplement Band 26, 58-64.

Priori S., Constatini E., Capezzuoli E., Protano G., Hilgers A., Sauer D., Sandrelli F. (2008) – Pedostratigraphy of Terra Rossa and Quaternary geological evolution of a lacustrine limestone plateau in central Italy. Journal of Plant Nutrition and Soil Science 171, 509-523.

Pye K. (1992) – Aeolian dust transport and deposition over Crete and adjacent parts of the Mediterranean Sea. Earth Surface Processes and Landforms 7, 271-288.

Rackham O., Moody J. (1996)The making of the Cretan landscape. Manchester University Press, Manchester-New York, 237 p.

Rapp, A., Nihlen, T. (1986) – Dust storms and eolian deposits in North Africa and the Mediterranean. Geoökodynamik 7, 41-62.

Roth M., Mackey J., Mackey C., Nyquist J. (2002) – A case study of the reliability of multielectrode earth resistivity testing for geotechnical investigations in karst terrains. Engineering Geology 65, 225-232.

Sakellarakis Y., Panagiotopoulos D. (2006) – Minoan Zominthos. In Gavrilaki I., Tzifopoulos Y. (Ed.): Mylopotamos from Antiquity to the Present. Environment, Archaeology, History, Folklore, Sociology. Historical and Folklore Society of Rethymnon, Rethymnon, 47-75.

Scheffer F., Schachtschabel P. (2002)Lehrbuch der Bodenkunde. Spektrum, Heidelberg, 593 p.

Siart C., Eitel B. (2008) – Geoarchaeological studies in Central Crete based on Remote Sensing and GIS. In Posluschny A., Lambers K., Herzog I. (Ed.): Layers of Perception. Proceedings of the 35th International Conference on Computer Applications and Quantitative Methods in Archaeology (CAA), Berlin, Germany, April 2-6, 2007, Kolloquien zur Vor- und Frühgeschichte vol. 10, 299-305.

Siart C., Eitel B., Panagiotopoulos D. (2008a) – Investigation of past archaeological landscapes using remote sensing and GIS: a multi- method case study from Mount Ida, Crete. Journal of Archaeological Science 35, 2918-2926.

Siart C., Holzhauer I., Hecht S., Eitel B., Schukraft G., Bubenzer O. (2008b) – Karstdepressionen als Archive der Landschaftsgeschichte. Geomorphologische, geophysikalische und sedimentologische Untersuchungen auf dem Plateau von Zominthos, Zentralkreta. In Mächtle B., Nüsser M., Schmid H., Siegmund A. (Ed.) : Inszenierte Landschaften und Städte. Heidelberger Geographische Gesellschaft, Journal, 22, Heidelberg, 205-219.

Siart C., Bubenzer O., Eitel B. (2009a) – Combining digital elevation data (SRTM/ASTER), high resolution satellite imagery (Quickbird) and GIS for geomorphological mapping : a multi-component case study on Mediterranean karst in Central Crete. Geomorphology 112, 106-121.

Siart C., Hecht S., Holzhauer I., Altherr R., Meyer H.P., Schukraft G., Eitel B., Bubenzer O., Panagiotopoulos D. (2009b) –Karst depressions as geoarchaeological archives : the palaeoenvironmental reconstruction of Zominthos (Central Crete) based on geophysical prospection, sedimentological investigations and GIS. Quaternary International, in press.

Siart C., Hecht S., Brilmayer Bakti B., Holzhauer I. (2010) – 3D visualisation of Mediterranean subsurface karst features based on tomographic mapping (Zominthos, Central Crete). Zeitschrift für Geomorphologie, in press.

Spratt T. (1865)Travels and researches in Crete. vol. 1., Editor J. Van Voorst, London, 387 p.

Wurmser H. (2008)Etude de morphologie urbaine : le cas de Latô (massif du Mirambello, Crète). Mémoire de troisième année de l’Ecole française d’Athènes, 121 p.

Haut de page

Annexe

  

Version française abrégée

Les dépressions karstiques se révèlent être d’un grand intérêt pour les études géoarchéologiques en Grèce, ces formes du relief ayant probablement joué un rôle important dans l’économie des sociétés crétoises anciennes. Si, dans le sud de la France, des études complètes ont été menées à partir de carottages et de tranchées géologiques et ont permis de reconstituer l’évolution paléoenvironnementale de certaines dolines, elles ont aussi mis au jour l’aménagement de ces dernières à des fins agricoles par les populations locales (Bruxelles, 2001 ; Bruxelles et al., 2006 et 2007). En Grèce, et en Crète plus particulièrement, peu d’études géoarchéologiques ont été réalisées ; pour l’essentiel seule la morphologie karstique du milieu a été décrite (Fabre et Maire, 1983 ; Bartels, 1991 ; Egli, 1993). Aujourd’hui encore, les conditions aynat prévalues au moment des remplissages sédimentaires de ces dépressions restent méconnues dans cette partie de la Méditerranée. Le présent article étudie simultanément le cadre morphologique et le contexte paléoenvironnemental de deux sites archéologiques majeurs de Crète centrale et orientale, Zominthos et Latô (fig. 1). Le site de Zominthos est situé en Crète centrale, dans la chaîne de montagnes de l’Ida, à environ 1200 m d’altitude. La géologie du secteur est caractérisée par l’omniprésence de roches carbonatées présentant les traces d’une forte dissolution karstique. Le site a été occupé à l’époque minoenne (~1650 BC) et abandonné vers 1050 BC. Le site de Latô, localisé à environ 300 m d’altitude, se situe dans la région du Mirambello, à l’est de la chaîne montagneuse du Dikté. Le site principal a été bâti sur le pourtour d’une doline et a été intensément occupé à l’époque hellénistique. Dans le secteur de Latô, vers le sud-est, on retrouve de nombreuses dolines présentant le même profil topographique et développées dans des roches carbonatées. Dans l’une d’entre elles, la doline D4 (Ghilardi, 2006), des structures archéologiques recouvertes d’argiles de décalcification ont été identifiées et des tessons de céramique de différentes périodes ont été également repérés.

Associée à la réalisation d’un sondage d’une longueur de 10 m, une étude géophysique (méthode d’Electricité Résistivité Tomographie) a été menée dans l’ouvala de l’Ida. C’est sur les marges de cette dernière qu’a été bâtie la villa minoenne de Zominthos. Les résultats ont notamment permis de révéler la stratigraphie des formations superficielles et ont mis en évidence la présence de structures archéologiques recouvertes par d’importantes quantités de sédiments : un mur est probablement identifié au pied du site archéologique, dans l’ouvala. En complément, les résultats de la prospection géoélectrique ont révélé la morphologie du fond de l’ouvala où un point d’arrivée des eaux semble avoir été identifié. En parallèle, des observations de terrain et deux carottages d’une longueur identique de 5 m, ont été réalisés en septembre 2008 dans la doline D4 du secteur de Latô, à environ 1,4 km au sud-est du site archéologique, Dans le fond de cette doline, des structures archéologiques sont recouvertes par d’épaisses couches d’argiles rubéfiées. Les analyses sédimentologiques réalisées comprennent des mesures de susceptibilité magnétique (à haute et basse fréquences), de granulométrie laser (pour la fraction inférieure à 2 mm), et de diffraction des rayons X. L’ensemble des résultats a facilité la caractérisation de la texture, de la teneur en minéraux ferromagnétiques ainsi que de la composition minéralogique des sédiments. Les observations de terrain réalisées à Latô ont également permis de découvrir une probable ancienne source, tarie de nos jours, ainsi qu’un réseau d’irrigation indiquant une occupation humaine et une utilisation des eaux de ruissellement au cours de la période historique.

Les analyses granulométriques effectuées à partir des carottages Z1 (Zominthos), D4A et D4B (Latô) indiquent la présence de sédiments généralement homogènes et de couleur relativement similaire, entre la surface et environ -2 m (code Munsell 10 YR 4/3). Entre 2 m de profondeur et la partie inférieure des carottages Z1 et D4A/D4B, on relève la présence de dépôts rubéfiés (code Munsell 7.5 YR 4/6) présentant une granulométrie variée. Un tamisage à 2 mm a permis d’étudier la fraction la plus fine (<2 mm) par granulométrie laser. Les paramètres granulométriques varient de manière significative entre les deux secteurs étudiés en raison de la morphologie et des dimensions des dépressions karstiques : dans Z1 (ouvala), il semble que des apports fluvio-torrentiels (graviers/galets à matrice limono-sableuse) sédimentent périodiquement tandis que dans D4A et D4B, il s’agit essentiellement d’argiles issues de la décantation des eaux de ruissellement de versant (présence continue d’argiles homogènes entre la surface et 5 m de profondeur). Les analyses de susceptibilité magnétique réalisées sur l’ensemble des échantillons présentent de fortes similitudes entre les deux secteurs concernés : entre -2 m et la surface, la susceptibilité magnétique augmente de façon continue alors qu’entre 2 m de profondeur et la partie inférieure des carottages, les valeurs sont plus faibles et oscillent (pics de plus forte valeur) de manière identique dans les trois carottages. L’hypothèse d’apports éoliens d’origine volcanique (cendres d’éruptions volcaniques) est avancée pour expliquer ces variations. La diffraction des rayons X réalisée sur les sédiments limono-argileux des trois carottages montre la présence de quartz, d’illite, de muscovite, de kaolinite (forte représentation pour D4A et D4B), de chlorite, de vermiculite et d’un mélange non identifié (corrensite ?). Les proportions sont toujours très homogènes. L’origine de ces minéraux est essentiellement locale et les phyllosilicates retrouvés (muscovite, illite) proviennent de la dissolution des calcaires locaux.

En conclusion, l’approche développée, à la fois pluridisciplinaire et multisites, a permis d’analyser les dynamiques de remplissage sédimentaire des dépressions karstiques de Crète centrale et orientale, dans les régions de l’Ida et du Mirambello/Dikté. Le relief accidenté et en apparence inhospitalier a tout de même permis l’installation des sociétés sur les marges des dépressions karstiques quelle que soit leur taille. Une position à l’abri des envahisseurs potentiels a certainement profité au développement important de ces sites. L’exploitation des dolines et des ouvalas de Crète centrale et orientale s’est amorcée dès l’époque minoenne et s’est poursuivie jusqu’à l’époque hellénistique, voire au début du XXe siècle. L’utilisation des argiles (forte teneur en kaolinite) pour la confection de céramiques est attestée par la présence de fours sur le site de Latô et il ne fait aucun doute que le matériel avait une origine locale, certainement issu de la doline étudiée dans cet article.

Haut de page

Table des illustrations

Titre Fig. 1 – Geological maps of the Cretan Ida Mountains (A) and Dikti Mountains (B). Major geological units.Fig. 1 – Cartes géologiques des montagnes crétoises de l'Ida (A) et du Dikté (B). Unités géologiques principales.
Légende 1: Neogene/Quaternary; 2: ophiolites; 3: Tripolitza unit; 4: phyllite-quartzites; 5: Kalavros schists; 6: Plattenkalk unit; 7: main detachment fault; 8: road system (2 and 3 belong to the upper nappes; 4, 5 and 6 belong to the lower nappes. Subset A: Ida Mountains. Subset B: Dikti Mountains (data source: IGME geological map of Greece, 1983).1 : Néogène/Quaternaire ; 2 : ophiolites ; 3 : unité de Tripolitza ; 4 : Phyllithe-quartzites ; 5 : schistes de l'unité de Kalavros ; 6 : unité de Plattenkalk ; 7 : décrochement de faille principal ; 8 : système routier (2 et 3 appartiennent aux nappes supérieures ; 4, 5 et 6 appartiennent aux nappes inférieures. A : Montagnes de l'Ida. B : Montagnes du Dikté (données : cartes géologiques IGME, 1983).
URL http://geomorphologie.revues.org/docannexe/image/7709/img-1.jpg
Fichier image/jpeg, 1004k
Titre Fig. 2 – Location of the depressions Z (Zominthos) and D4 (Latô) with the coring sites.Fig. 2 – Localisation des dépressions karstiques Z (Zominthos) et D4 (Latô) et emplacement des carottages.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-2.jpg
Fichier image/jpeg, 1,9M
Titre Fig. 3 – Block diagram showing ERT profile E3 as a vertical texture on the high resolution DEM in twofold vertical exaggeration (profile length: 100 m).Fig. 3 – Bloc diagramme superposant le profil ERT E3 sur le modèle numérique de terrain de haute résolution avec un facteur d’exagération double (longueur du profil : 100 m).
Légende The cross section records the deepest part of the karst depression, which consists of a buried sinkhole indicated by the blue colours in the middle of the array. It clearly demonstrates the profoundly heterogeneous and unpredictable subsurface karst morphology, which has also been proved by further ER-tomographies from Zominthos (surface texture: Quickbird MS data; 3D model of Minoan villa).Le profil E3 révèle la partie la plus profonde de la dépression. Indiquée par la couleur au centre du profil, la dépression est en fait un « entonnoir » recouvert de sédiments. Cela démontre clairement le caractère hétérogène et imprévisible de la surface souterraine, ce qui a été confirmé par d’autres ER-tomographies réalisées à Zominthos (logiciel utilisé : Quickbird MS data ; modèle 3D de la villa minoenne).
URL http://geomorphologie.revues.org/docannexe/image/7709/img-3.jpg
Fichier image/jpeg, 920k
Titre Fig. 4 – 3D model of the subsurface karst topography in Zominthos as inferred and interpolated from geophysical tomographies.Fig. 4 – Modèle 3D de la topographie souterraine de Zominthos, interprétée à partir des résultats géophysiques.
Légende A. Digitally excavated DEM of the depression in a perspective view looking to the NW. B. View to the SE (dashed white lines: potential underground drainage pattern; striped polygon below Minoan villa: backfilling of the settlement hill). The indicated surface (greyish shaded texture) represents the approximate boundary zone between the sedimentary overburden and the solid limestone bedrock. Quickbird MS satellite data were draped on the DEM in the exterior of the depression, while a 3D model of the Minoan villa was inserted on top of the settlement hill using ArcGIS and Visual nature studio software.A. Modèle numérique de terrain dépourvu de la couverture sédimentaire, bloc orienté au nord-ouest. B. Réseau hydrographique souterrain supposé (le polygone rayé correspond au remplissage sédimentaire localisé derrière le site de Zominthos ; figure orientée au sud-est, lignes avec tirets blancs). La surface surlignée par la couleur grise (ombrée) représente le contact approximatif entre la couverture sédimentaire et le substrat calcaire. Les données satellites traitées grâce au logiciel Quickbird MS et un modèle 3D de la villa minoenne ont été utilisés comme texture de surface autour de la dépression, insertion des données dans ArcGIS et Visual nature studio software.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-4.jpg
Fichier image/jpeg, 1,7M
Titre Fig. 5 – Grain-size distribution of selected samples analysed by laser granulometre.Fig. 5 – Distribution granulométrique des échantillons analysés par la méthode laser.
Légende A. Grain-size distribution for core Z1. Within the first two metres (from Z1-010 to Z1-190), the samples close to the surface indicate a modal index comprised between 10 and 40 µm. The peaks observed in the coarsest fractions are sandy particles (poorly rounded). B. Cores D4A and D4B and displaying a modal index between 1 and 6 µm on average. The peaks observed in the coarsest fractions are induced by small black rounded nodules. A. Distribution granulométrique pour le sondage Z1. Dans les deux premiers mètres (de Z1-010 à Z1-190), le mode granulométrique indique des valeurs comprises entre 10 et 40 µm. Les pics observés dans les fractions les plus grossières correspondent à des sables (émoussés). B. Carottages D4A et D4B dans lesquels le mode granulométrique oscille entre 1 et 6 µm. Les pics dans les fractions les plus grossières sont liés à la présence de nodules très émoussés de couleur noire.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-5.jpg
Fichier image/jpeg, 616k
Titre Fig. 6 – Core profile and magnetic parameters of core Z1 (Zominthos).Fig. 6 – Profil stratigraphique et paramètres magnétiques pour le sondage Z1 (Zominthos).
Légende From left to right: stratigraphic information based on colour as well as magnetic susceptibility at low (χlf) and high (χhf) frequencies are plotted with regard to depth. χ is a proxy for the magnetic mineral concentration, which is often approximately equated to the concentration of magnetite. Moreover, it can be considered as grain-size dependent (Dekkers, 1997; Dearing, 1999). χlfhffd) measurements are sensitive to ultrafine magnetic grains and high values designate the presence of fine ‘viscous’ ferrimagnetic grains (~0.02 μm in diametre). Fd (in%) is the relative contribution of fine viscous particles to the total ferromagnetic assemblage, high corresponding percentage values point to wide distributions of ultrafine grains. The absolute elevation in m a.s.l. is indicated in brackets.De la gauche vers la droite : la stratigraphie générale est fondée sur des différences de couleurs ; les valeurs de susceptibilité magnétique à basse (χlf) et haute fréquence (χhf) sont exprimées en fonction de la profondeur. χ est un paramètre pour étudier la concentration des minéraux magnétiques, qui est généralement égale à la concentration de magnétite, il existe un lien avec la granulométrie (Dekkers, 1997 ; Dearing, 1999). Les mesures pour χlfhffd) sont sensibles à la présence de grains magnétiques ultrafins et les valeurs élevées indiquent l’existence de grains ferrimagnétiques de très petite taille (~0,02 μm de diamètre). Fd (en %) est la contribution relative de fines particules dans l’assemblage total des minéraux ferromagnétiques : de forts pourcentages indiquent la présence de grains très fins. L’altitude absolue en mètres par rapport au niveau de la mer est indiquée entre parenthèses.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-6.jpg
Fichier image/jpeg, 668k
Titre Fig. 7 – Core profiles and magnetic parameters of cores D4A and D4B (Latô).Fig. 7 – Profils stratigraphiques et paramètres magnétiques pour les carottages D4A et D4B (Latô).
Légende From left to right: general stratigraphy based on colour as well as magnetic susceptibility at low (χlf) and high (χhf) frequencies are plotted with regard to depth (for details on χ, χfd and Fd (in%) see caption of fig. 6); absolute elevation in m a.s.l in brackets).De la gauche vers la droite : la stratigraphie générale est fondée sur les différences de couleurs ; les valeurs de susceptibilité magnétique à basse (χlf) et haute fréquence (χhf) sont exprimées en fonction de la profondeur (pour les détails concernant χ, χfd, and Fd (en %), se référer à la légende de la fig. 6) ; l’altitude absolue en mètres par rapport au niveau de la mer est indiquée entre parenthèses.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-7.jpg
Fichier image/jpeg, 744k
Titre Fig. 8 – Magnetic susceptibility signal at low frequency (χlf) for cores Z1, D4B and D4A.Fig. 8 – Valeurs de susceptibilité magnétique à basse fréquence (χlf) pour les carottages Z1, D4B et D4A.
Légende All values are expressed according to depth (ordinate). The high values in the upper horizons are in significant contrast to the lower susceptibilities in deeper profile parts. This general trend can be observed in both the cores from Zominthos and Latô.Toutes les valeurs sont exprimées par rapport à la profondeur. Les valeurs élevées observées dans les horizons supérieurs contrastent avec les faibles valeurs obtenues dans la partie inférieure. Cette tendance générale peut être observée dans les carottages de Zominthos et de Latô.
URL http://geomorphologie.revues.org/docannexe/image/7709/img-8.jpg
Fichier image/jpeg, 251k
Haut de page

Pour citer cet article

Référence papier

Christoph Siart, Matthieu Ghilardi et Ingmar Holzhauer, « Geoarchaeological study of karst depressions integrating geophysical and sedimentological methods: case studies from Zominthos and Latô (Central and East Crete, Greece) », Géomorphologie : relief, processus, environnement, vol. 15 - n° 4 | 2009, 241-256.

Référence électronique

Christoph Siart, Matthieu Ghilardi et Ingmar Holzhauer, « Geoarchaeological study of karst depressions integrating geophysical and sedimentological methods: case studies from Zominthos and Latô (Central and East Crete, Greece) », Géomorphologie : relief, processus, environnement [En ligne], vol. 15 - n° 4 | 2009, mis en ligne le 01 janvier 2012, consulté le 19 août 2017. URL : http://geomorphologie.revues.org/7709 ; DOI : 10.4000/geomorphologie.7709

Haut de page

Auteurs

Christoph Siart

Geographical Institute, Laboratory for Geomorphology and Geoecology, University of Heidelberg, Im Neuenheimer Feld 348, D-69120 Heidelberg, Germany (christoph.siart@geog.uni-heidelberg.de; ingmar.holzhauer@geog.uni-heidelberg.de)

Articles du même auteur

Matthieu Ghilardi

CEREGE UMR 6635, CNRS, Europôle Méditerranéen de l’Arbois BP 80, 13545 Aix-en-Provence Cedex 04, France (ghilardi@cerege.fr)

Articles du même auteur

Ingmar Holzhauer

Geographical Institute, Laboratory for Geomorphology and Geoecology, University of Heidelberg, Im Neuenheimer Feld 348, D-69120 Heidelberg, Germany (christoph.siart@geog.uni-heidelberg.de; ingmar.holzhauer@geog.uni-heidelberg.de)

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • Revues.org