Navigation – Plan du site

AMS radiocarbon dating of carbonate cements in late Pleistocene alluvial conglomerates, Verdouble River. Palaeoenvironmental implications concerning the Palaeolithic site of Tautavel (Pyrénées-Orientales)

Datations AMS du radiocarbone des ciments carbonatés des conglomérats alluviaux fini-pléistocènes de la rivière Verdouble. Implications paléoenvironnementales relatives au site paléolithique de Tautavel (Pyrénées-Orientales)
Pierre Giresse et Michel Martzluff
p. 115-130

Résumés

Les conglomérats alluviaux pléistocènes de la vallée du Verdouble, proches du célèbre site paléolithique de Tautavel, ont fait l’objet de déterminations lithologiques des galets afin de déterminer leur provenance et d’analyses microstructurales des ciments calcitiques préliminaires aux datations AMS du 14C. Pour la première fois, ces ciments monophasés ou rarement polyphasés ont pu être datés du stade isotopique 3 ou peut-être 4 (sous-étages III et II du Würm) et non du stade 7 (interglaciaire Mindel-Riss) comme envisagé antérieurement par quelques auteurs. Les 15 âges obtenus se situent entre 10130 ans BP et plus de 48 000 ans. Ils témoignent d’une incision karstique relativement tardive des Gorges des Gouleyrous et, en corollaire, de l’ouverture récente de la communication entre les parties amont et aval du bassin du Verdouble, communication qui n’était pas encore achevée à l’époque de l’Homme de Tautavel, il y a 450 000 ans. Les érosions latérales des crues du Verdouble ont, dès lors, détruit les témoins de possibles sols d’habitat paléolithiques à l’aval des Gouleyrous : le seul modeste témoin conservé se situe précisément à l’amont des Gorges, au confluent du Ruisseau de Vingrau avec le Verdouble.

Haut de page

Notes de la rédaction

Article soumis le 16 janvier 2014, recu sous sa forme révisée le 24 mai 2014, accepté le 20 août 2014

Texte intégral

The research work reported upon this paper was partly funded by the French Minister of High Teaching in the framework of the Pluri-Formation Project (PPF) “The Tautavel Man on his territories: since the very former ages up to the present time developments”. Gary Fowler is acknowledged for checking the English and Thibaud Saos for technical help. Thoughtful editings by D.R. Bridgland and two anonymous reviewers have improved the manuscript and are gratefully appreciated.

1. Introduction

1In 1971 searches revealed the forehead of the skull of a Homo erectus tautavelensis or more exactly Homo heidelbergensis, since then more collectively called Tautavel Man, who lived in the vicinity of the cave of Caune Arago in the Verdouble valley (Lumley de and Lumley de, 1971). This skull was dated around 450,000 years ago according to 230Th/230U disequilibrium method (Lumley de et al., 1984; Falguère et al., 2004). Perched halfway up of a steep Urgonian limestone slope, this cave constituted a ideal observationary point from which to watch the movements of the herds of wild animals. Facing southward and shielded from prevailing wind, this site is also situated near the water source of the Verdouble, a necessity for both man and animal.

2This study documents the nature and history of Pleistocene coarse-grained alluvial deposits along the Verdouble River, i.e. upstream and downstream of the Gouleyrous Gorge, which crosses the Devèze upland. The deposits of the plain of Paziols and those of the Vingrau-Tautavel Depression will be described in succession (fig. 1, fig. 2A and fig. 2B).

Fig. 1General map of the course of the Verdouble River, tributary of the Agly River, from Padern until the Alzine loss.
Fig. 1Carte générale du cours du Verdouble, affluent de l’Agly depuis Padern jusqu‘à la perte de l’Alzine.

Fig. 1 – General map of the course of the Verdouble River, tributary of the Agly River, from Padern until the Alzine loss.Fig. 1 – Carte générale du cours du Verdouble, affluent de l’Agly depuis Padern jusqu‘à la perte de l’Alzine.

Numbers 2a and 2b refer to following maps .
Les numéros 2a et 2b correspondent aux cartes de la figure suivante.

Fig. 2 – Geologic sketch of the Verdouble catchment.
Fig. 2 – Carte géologique simplifiée du bassin du Verdouble.

Fig. 2 – Geologic sketch of the Verdouble catchment. Fig. 2 – Carte géologique simplifiée du bassin du Verdouble.

A: In the upstream of Gouleyrous Gorge (Paziols plain) according to the geologic map of Rivesalte (Berger et al., 1997). L: Lias, J: Jurassic, C: Cretaceous, Ci: Lower Cretaceous; g2-3: Oligocene conglomerates. Only the Pleistocene fluviatile deposits are exactly reproduced; the pebble accumulation (inner delta fan) in the upstream of Gouleyrous Gorge was added by the authors. River losses are indicated by arrows. B: In the downstream of Gouleyrous Gorge (depression of Vingrau-Tautavel) according to the geologic map of Tuchan (Berger et al., 1993). J: Jurassic, Ci: Lower Cretaceous; g: Oligocene conglomerates. Only the Pleiostocene fluviatile deposits are exactly reproduced.
A : À l’amont des Gorges des Gouleyrous (plaine de Paziols) d’après la carte géologique de Tuchan (Berger et al., 1997). L : Lias, J. Jurassic, C : Cretaceous, Ci : Lower Cretaceous, g2-3 : conglomérats oligocènes. Seuls les affleurements des dépôts fluviatiles pléistocènes sont précisément reportés, l’accumulation caillouteuse à l’amont des Gorges a été ajoutée par les auteurs. Les pertes de la rivière sont indiquées par les flèches. B : À l’aval des Gorges des Gouleyrous (dépression de Vingrau-Tautavel) d’après la carte géologique de Rivesaltes (Berger et al., 1993) ; J : Jurassique, Ci : Crétacé inférieur, g : conglomérats oligocènes.

3Pleistocene pebbly alluvium is locally hardened with calcite-rich cements. The location and nature of clastic sediment and the dating of its cements may provide constraints on the timing of calcrete formation and of aggradation and incision of the alluvial deposits. These accumulations and their cements involve a combination of a rather warm and wet environment. The cements are thought to have developed during very early diagenesis allowing the consolidation and the preservation of the deposit. Fifteen new AMS dates from these cements will lead us to propose a Late Pleistocene geomorphic and processes history for the Verdouble Valley, and for the Tautavel region, that is slightly different from that of a widespread assumption that Verdouble alluvium are mid-Pleistocene or older ages (Berger et al., 1997).

4The precipitation of calcium carbonate cement from soil water and shallow ground water provides a potential record of environmental conditions. For absolute age dating of inorganic carbon, thin sections of the crystal fabrics and textures preserved provided the preliminary documentation. Therefore, various examples of the 14C and 230Th dating of carbonate cements were provided through conglomerates of the proglacial valley fills of Austria (Ostermann et al., 2006), the alluvium terraces in New York State (Seleck and Baran, 2003), pedogenic layers (Rubin and Sieh, 1997; Vincent et al., 2007), calcretes in the soils (Geyh and Eitel, 1998; Srivastava, 2001), tufaceous carbonates (Haynes et al., 1989), pedogenic cutans (Courty et al., 1994; Pustovoytov, 1998), calcareous nodules (Taupin, 2003) or calcite from fine matrix material and coatings of glacial landforms (Yi et al., 2004). Various calcareous cement dates have also been undertaken for marine sediment, especially Holocene beach-rocks, and reconstituted sea level changes (Ramsay and Cooper, 2002; Desruelles et al., 2009). Although the cement can be contaminated by carbonate particles other than the calcite cement, radiocarbon 14C dating is still the most appropriate method for dating various Mediterranean beachrocks.

2. Study area

2.1. Geological and hydrogeological setting

5The Verdouble River is the main tributary of the Agly River, which it joins on his left bank downstream to Estagel (fig. 1). It flows through a subalpine valley located in the Corbières hilly area. The bedrock is Silurian and Devonian in the north of the catchment, Jurassic for uplands bounding the catchment in the south and finally Jurassic and Cretaceous for the downstream course of the Verdouble and the Vingrau-Tautavel plain. Due to the complex tectonics of the Corbières, strongly influenced by the Pyrenean orogeny, the Jurassic-Cretaceous limestones do not comprise a continuous aquifer. The depth of karstification, several hundred metres below sea level, especially in the Opoul area and in Lower Agly, is mainly linked to: (i) the Messinian salinity crisis, when the base level fell of 1500 m, and (ii) to the infiltration of meteoric waters with a strong dissolution power controlled by weather conditions (Dörfliger et al., 2008a; Dörfliger et al., 2008b). This system is partly cut off from the Mediterranean Sea by the Plio-Quaternary sedimentary cover (Ladouche and Dörfliger, 2004). Then during the various Quaternary glacial phases where the marine level is every time lowered towards – 120 m, the deep networks inherited from the crisis were reactivated at the same time as new dissolutions were developed.

6The Verdouble catchment extends over 305 km2 with an average slope between 10 and 40‰ in the upstream part and less than 10‰ in the downstream part (Gaume, 2002). In the study area, from Paziols to Tautavel, the Verdouble River decreases in altitude from 104 to 92 m but the surrounding hillslopes extend to summits that attain more than 600 m.

7The Verdouble River crosses two calcareous canyons of karstic origin, the first (Gouleyrous Gorge) is situated between the Paziols and Tautavel Basins in the side of which is Caune Arago (fig. 1 and fig. 3). The second, downstream, leads up to the main Agly River: the Verdouble disappears from the surface (Alzine loss) and then undergoes successive other losses in the crossing of the Jurassic massifs, in particular in the meanders of the Cimetière des Maures.

Fig. 3Location of the analysed conglomerates.
Fig. 3Localisation des conglomérats analysés.

Fig. 3 – Location of the analysed conglomerates. Fig. 3 – Localisation des conglomérats analysés.

The origin of the pebbles of Fya (mostly calcareous) is the western part of the Verdouble catchment (main course of the Verdouble), whereas the pebbles of Fyb (mostly siliceous) may come from the eastern part of the basin (Donneuve and Tarassac tributaries).
L'origine des galets de Fya (surtout carbonatés) est issue des zones occidentales du bassin (cours principal du Verdouble), celle des galets de Fyb (surtout siliceux) proviendrait de la partie orientale du bassin (affluents Donneuve et Tarassac).

8Surface waters penetrating into limestones of the Upper Jurassic and the Lower Cretaceous of the Verdouble catchment feed an important karstic network upstream and downstream from Gouleyrous Gorge. Near Paziols, a cave-in allows observation of the subterranean flow of the Verdouble. This karst is an allogenic system, recharged by both precipitation and river water infiltration through swallow holes.

9Lower Cretaceous formations making up the Devèze upland are especially constituted by jointed limestones. Swallow holes and caves are developed at various levels of the calcareous hill, the most important being the 35 m-long Caune of Arago. These limestones are the sites of temporary emergences, but long-lasting springs remain rare. To the South of Paziols, a hot and sulphated spring indicates an inflow of deep water. One temporary resurgence is observed near the confluence of Verdouble and Tarassac, on the western side of the Devèze, and indicates the existence of a drowned Lower Cretaceous karst (Berger et al., 1997).

10To the Mediterranean, the karst is in contact with the brackish Salses-Leucate lagoon. The two main springs of this karst system emerge a few kilometers from each other. The less transmissives Pliocene-Quaternary sedimentary formations play the role of a porous cork and induce a slowdown in the flow ; the two sources thus appear as overflow sources (Fleury et al., 2009). Moreover, both springs have a temperature anomaly of + 2°C to + 2.5°C which indicates an inflow of deep warm water (Dörfliger et al., 2008b ; Ladouche and Dörfliger, 2004).

11Peak streamflows of the Verdouble generally result from autumn storms. Three historical floods during the last century are noteworthy and both involve the Tautavel-Vingrau Depression site. The last big flood occurred in November1999, when the peak discharge was estimated at a little more than 900 m3 and reworked largely previous alluvial deposits (Gaume, 2002; Calvet and Lemartinel, 2002). In period of low water level, the bed of the Verdouble River is generally dry at the confluence with the Agly River.

2.2. Alluvial Pleistocene deposits

12According to the 1/50,000 geological map of Tuchan (Berger et al., 1997) only three levels of Pleistocene alluvial terraces (Fu, Fya, Fyb) have been reported (fig. 2A). On the map of Rivesaltes (Berger et al., 1993), the same Fya and Fyb were reported, but the older terraces called Fx are mainly out of the Verdouble catchment (Berger et al., 1993; fig. 2B). This is thus a more restricted succession than those of the nearby valleys of the Agly or the Têt, which are believed to date to the oldest glaciations of the Pleistocene (Lumley de, 1971; Collina-Girard, 1976; Lumley de et al., 1976; Berger et al., 1993; Calvet, 1996; Giret, 1997). It is necessary to specify that the notion of terraces is used here only in reference of the geological maps of both Tuchan and Rivesaltes. It does not involve the existence of a topographic shoulder on which deposits are overlaid.

13Fu, the highest and the oldest level (10-30 m above the Verdouble streambed) is ascribed to the Lower to Middle Pleistocene sensu lato. This “High terrace” amounts to some small residual coarse deposits that we observe only north and east of Paziols where they overlie residual Tertiary gravel. Despite the proximity of Jurassic calcareous hills, the pebble lithology is dominated (60 to 90%) by siliceous elements (schists, sandstones, lydian stones) derived from the Palaeozoic upland. Pebbles are cemented by several generations of sparites, sometimes partly dissolved and rubefied which testify to a long pedogenetic history. These outcrops disappear downstream of Paziols and this level is unknown in the Tautavel-Vingrau Depression where some small residual hills (Fx) are mapped.

14Fya represents a "Middle terrace", formed of rather numerous widespread conglomerates or coarse sandstones that are scattered only a few metres above the present bank of the Verdouble River, both upstream and downstream of the Gouleyrous Gorge. The maximum exposed thickness is 2 m. The pebbles are very heterometric and testify to the high flood discharges. Calcite cementation has allowed the fossilization of some deposits, but accumulations of largely scattered pebbles are even more frequent. A little upstream of Tautavel, alluvial loam interstratified within the conglomerate contains a temperate to fresh climate fauna of gastropods the 14C age of which is over 35,000 years BP (Berger et al., 1993). An early Würm age is suggested for this "Middle terrace".

15Fyb corresponds to the deposits of the "Lower terrace", which are relatively well preserved only 2 to 3 m above the present streambed, at least where cemented by calcite. Near the Mesozoic limestone cliffs, they are sometimes buried under accumulations of cryoclastic fragments and loessic silts with steppe malacofauna that allow their correlation with the recent Würm (Berger et al., 1997).

2.3. Archaeological setting of the Tautavel-Vingrau depression

16For half a century, the Tautavel-Vingrau Depression has benefited from intensive archaeological prospecting as well as excavations of open-air sites or caves (Martzluff et al., 2013). The occurrences of open-air Palaeolithic sites are particularly numerous (Abélanet, 1990). Very curiously, of 36 inventoried deposits, only one is attributable to the Lower Palaeolithic (fig. 4). This one is situated to the southeast of the Caune Arago, 300 m upstream of the confluence between the Vingrau tributary and the Verdouble (alt. 130 m NGF), namely out of reach of Verdouble floods. It overlies a break preceding an abrupt slope of 25 m deeply cut by the Vingrau tributary. 

Fig. 4Archaeological sites in Vingrau-Tautavel Depression.
Fig. 4 – Sites archéologiques de la dépression Vingrau-Tautavel.

Fig. 4 – Archaeological sites in Vingrau-Tautavel Depression. Fig. 4 – Sites archéologiques de la dépression Vingrau-Tautavel.

With dark grey colour, calcareous uplands and cliffs above the Depression. Broken line bounds the Upper Palaeolithic artefacts area.
En gris foncé, reliefs calcaires et falaises encadrant la dépression. 1 : Paléolithique inférieur, astérisque : outils moustériens erratiques ; carré : Paléolithique supérieur ancien (Gravettien, Solutréen) ; rond : Paléolithique supérieur récent (Magdalénien et Épipaléolithique). 2 : Site de la Teulera. La ligne en tiretés souligne les limites des sites du Paléolithique supèrieur.

17Flint and quartz tools, partially burned bones and teeths (goat, Mosbach horse and probable Merck’s rhinoceros), as well as a human canine tooth, were deposited within the hillside debris in the confines of the rocky cliff. The location and the good preservation of flint tools rule out an origin from a rock shelter deposit dismantled by erosion. All the finds belong to the Mediterranean Acheulean of the Middle Pleistocene, in its " Tautavelian " facies (Martzluff, 2006). Rare evidence concerns the end of the Old Palaeolithic and the Middle Palaeolithic periods, it has been well attested in the filling of the Caune de l'Arago by some Levalloisian artefacts. These artefacts are composed of quartz or brown jaspers and, more rarely, in the case of two examples, they are significantly rolled as a result of transport. These extremely rare records are always associated with other more recent archaeological remains.

18Thus, it is only from the Upper Palaeolithic that the open-air camps are well represented in the Tautavel-Vingrau depression during the coldest phases of the Würm (Martzluff and Abélanet, 1990; Baills, 2003, 2008; Martzluff et al., 2013) and also in cave deposits. However, the particular location of these deposits is remarkable. Up to the first Würm glacial maximum (MIS 4), the former phase of the Upper Palaeolithic is represented by Gravettian and Solutrean artefacts in marginal sectors, invariably far above the level of the Verdouble River (bonded by a broken line on figure 4). After the Last Glacial Maximum (LGM; MIS 2), the final Magdalenian is well represented on the left bank of the Verdouble, in particular at its confluence with the Vingrau (Teulera emplacement; Sacchi, 1986) although there are also Magdalenian sites in the Vingrau basin (Rec del Penjat; Martzluff and Abelanet, 1990). Nearby, two other concentrations of artefacts can relate to the end of the Palaeolithic, but also to the Epipalaeolithic. On the other hand, downstream of this confluence, we found only a few scattered and much eroded artefacts of possible Palaeolithic affinity in the fluvial deposit, while all the catalogued archeological sites belong to the Neolithic or to Antiquity. According to the current state of the knowledge, human occupation of a large sector of the valley situated downstream of the Gouleirous Gorge is restricted to the Holocene.

3. Methods

19The rough altitudinal datum of the fluvial gravel sites was taken from the 1/25,000 contour map and Google Earth, values are thus approximate except for the two profiles at the Tautavel Museum and the Alzine sites, where they were obtained from theodolite measurements (approximate resolution 10 cm). In each case, the altitude values correspond to the sampling point

20The lithologic composition of the cobbles was determined in the field by analogy with the outcrops mapped in the nearby upland. Several samples were the object of more precise determination in the laboratory (microscope examination, determination of carbonate content and X ray diffraction). In every station, between 50 and 100 lithoclasts were determined (tab. 1).

Tab. 1Lithologic composition of lithoclasts of the alluvial deposits of the Verdouble catchment.
Tab. 1Composition lithologique des galets des dépôts alluviaux du bassin du Verdouble.

Map

Locality

Altitude (m)

Pebbles number

Calcareous pebbles (%)

Siliceous pebbles (%)

Quartz (%)

Coated grains (%)

Fya

Moulin (1)

138

62

73

27

60

Fya

ND Faste (2)

140

50

86

16

4

28

Fya

Ford (3)

132

82

2.5

87.5

10

2.5

Fyb

Above ford (4)

135

83

22.5

62

9

2

Recent

Pond (5)

106

90

26

72

2

5.5

Fya

Seille (6)

117

67

11

87.5

1.5

22

Fya

Seille (7)

118

71

13

84

3

45

Fya

Museum (8)

111

55

69

29

2

89

Fya

Museum (8)

111

65

71

26

3

77

Fya

Alzine (9)

108

77

75

24

1

69

Fya

Alzine (9)

108

71

77

21

2

71

The parts related to the upstream and the downstream of the Gouleyrous Gorges are separated by an empty line.
Les parties à l’amont et à l’aval des Gorges des Gouleyrous sont séparées par une ligne vide.

21Thin sections of calcareous cement were examined under a polarizing microscope and cleaned and polished rock surfaces were observed under both a binocular and a scanning microscope. These observations permitted the characterization of the nature of the constituents of the cements and allowed us to select samples on which dating could be carried out. It also allowed us to choose the optimal method for the extraction of the cement. Fringes of cement were sampled with a microdrill under the microscope or simply selected by hand-picking lightly crushed materials for clean, unaltered calcite.

22Considering the good preservation of the sedimentary structures of the conglomerate or sometimes the associated sandstones (horizontal or oblique beds, some normally graded bedding without bioturbation), the basic premise of the study will be to assume that the carbonate cement was deposited at the same time as the gravel, or shortly afterwards, otherwise pebbles and gravels of layers frequently less than one metre thick would have been quickly and widely transported. Only some deposits protected by cementation can be dated, the others now being more or less scattered on the surface.

23The geochronology was determined from AMS radiocarbon dating of inorganic carbon from separated carbonate fractions. This method allowed the analysis of only one hundred milligrams of carefully selected calcite spars. Measurements were carried out at Poznan Radiocarbon Laboratory. For comparison with previous studies, and because the time interval prior to 25,000 BP is complex in terms of 14C atmospheric concentrations (Bard et al., 2004), all ages are reported in the text in conventional radiocarbon age rather than calibrated age. However, Table 2 presents the various conversions according to CalPal-2007 (online version) as described in Weninger and Jöris (2008). According to the same authors a nearly 2 kyr-reservoir correction was adopted.

Tab. 2Dating of inorganic carbon from carbonate cements in the Verdouble Valley.
Tab. 2Datations du carbone inorganique des ciments carbonatés de la vallée du Verdouble.

Field No

Longitude N

Latitude E

Height (m)

Laboratory No

14C yr BP

Calendric age BP

1 Mill

42°51’42.37

2°43’19.49

137

Poz-36407

>45,000

2 Ford

42°52’28.64’’

2°43’51.91’’

140

Poz-36408

31,400±600

35,527±685

3 Ford

42°52’28.64’’

2°43’51.91’’

140

Poz-36409

40,000±2000

44059±1552

4 upstream G.

42°52’11’’

2°43’53.13’’

134

Poz-36410

31,400±600

35527±685

5 upstream G.

42°51’32.86’’

2°43’51.70’’

137

Poz-36411

37,600±1200

42323±906

6 reach

42°51’32.86’’

2°43’51.70’’

137

Poz-36412

10,130±60

11738±202

7 Seille

42°50’12.23’’

2°45’11.40

117

Poz-36413

38,500±1300

42948±974

7 Seille

42°50’08.18’’

2°45’09.33’’

118

Poz-36493

34,600±900

39621±1130

8 Tautavel Museum

42°48’55.57’’

2°44’57.76’’

110

Poz-42580

20,220±130

24164±300

8 Tautavel Museum

42°48’55.57’’

2°44’57.76’’

110

Poz-42583

25,750±230

30811±388

8 Tautavel Museum

42°48’55.57’’

2°44’57.76’’

110

Poz-42582

14,210±80

17433±254

8 Tautavel Museum

42°48’55.57’’

2°44’57.76’’

110

Poz-42579

13,110±70

16023±391

9 Alzine

42°48’23.47’’

2°44’15.46’’

105

Poz-42585

26,400±250

31187±380

9 Alzine

42°48’23.47’’

2°44’15.46’’

105

Poz-42584

>48,000

upstream Tautavel

*

Gif.5051

>35,000

*: Dating from snail shells (Berger et al., 1997).
* : Datation de coquilles de gastéropode (Berger et al., 1997).

4. Observations and comments

4.1. Pleistocene alluvial deposits of the Paziols Basin (upstream of the Gouleyrous Gorge)

24Remnants of former flood plains less 1 metre thick were found along the Verdouble River, particularly those designated as map units Fyb and Fya. They comprise deposits formed from poorly-sorted clasts of variable size and variable petrographic composition, weakly stratified, generally sub-rounded to angular cobble gravel with a consequent sand matrix. The gravels are weakly imbricated and stratification dips about 1° towards the modern Verdouble channel. Some mainly sandy accumulations show well-preserved oblique stratifications. The cement fringes consist of scalenohedral calcite spars and generally, a only one-step cementation is observed. Most cemented beds are not laterally continuous for more than a few tens of metres. The exposures are currently dry and inactive.

25Near Paziols, the strongly cemented conglomerates of Fyb and Fya appear widely in the locality of Moulin (station 1). The conglomerate mapped Fyb forms a very small hill above the streambed (station 1, fig. 2A and fig. 4). The lithologic composition of the pebbles indicates a clear dominance of Cenozoic calcareous pebbles (red, beige, grey or dark grey) over siliceous pebbles from the Palaeozoic (dark grey or black) or from the Trias (reddish sandstones; tab. 1). More or less rubefied rind cement forms a continuous, isopachous and single-layer coating on lithoclasts (fig. 5A and fig. 5D) and consists of prismatic, elongated sparites 2-5 mm (fig. 5B). Its 14C age is earlier than 45,000 BP.

Fig. 5Examples of one-step cements.
Fig. 5Exemples des ciments déposés en une seule phase.

Fig. 5 – Examples of one-step cements. Fig. 5 – Exemples des ciments déposés en une seule phase.

A: Moulin, the skeleton of the conglomerate results from a single sparitic border (sp) linking the gravels of sandstones (s) and schists (sc), the void volumes (v) remain very important. B: Moulin, the first micritic cement (mi) developed preferentially on the surface of a calcareous clast (at the top) rather than on that of a siliceous clast (at the bottom), sparites (sp) attain a size from 1 to 2 mm. C: Upward Ford station, one single phase deposition of sparitic cement (sp) linking various gravels, frequent voids (v). D: Elongated cavity (c) between two calcareous lithoclasts (cl), the cavity is fringed by a single sparitic rim (sp) (A, D: surface sections; B, C: thin sections, cross polar).
A : Moulin, la charpente du conglomérat résulte d’un seul ourlet sparitique (sp) liant les graviers de grès (s) et de schistes (sc), les vides (v) sont demeurés très importants. B : Moulin, le premier ciment microsparitique (mi) s’est développé préférentiellement à la surface d’un claste calcaire (en haut) plutôt qu’à celle d’un claste silicaté (en bas), les sparites (sp) atteignent ici une taille de 1 à 2 mm. C : Amont du gué, une seule phase de cimentation sparitique (sp) ourle les vides (v) et lie différents graviers. D : La Seille, cavité étroite (c) développée entre deux lithoclastes calcaires (cl) et bordée par une seule frange de sparites (sp) (A, D : sections de surface ; B, C : lames minces, polarisation croisée).

26To the northeast of the Moulin locality, still 6-8 m above the streambed level, several other small mounds of presumed Fya crop out, but are not indicated on the map; the conglomerate beds are interstratified with sandstone benches with oblique stratification (Station 2). Near the confluence of the Donneuvé and Tarrassac tributaries (Station 3), the lithology of the pebbles, despite a separation of by about 1.5 km, is still similar to that of Moulin: Mesozoic calcareous pebbles are dominant over Palaeozoic siliceous pebbles and Trias sandstones (tab. 1). Homogeneous sparitic cements of the conglomerate (fig. 5C) and those of the sandstone were dated and gave ages of 31,400 and 40,000 BP, respectively.

27Lower down, the valley side, the same locality shows a mapped Fyb conglomerate only 3 m above the Verdouble streambed (Station 4). The lithologic assemblage is very different from that of nearby Fya, because it is dominated by silicified or siliceous elements of which nearly a third is shale-like and presumed to be Ordovician. Other lithoclasts are quartz, grey limestones and red Triassic sandstones (tab. 1). The homogeneous sparite cement here is dated to 31,400 BP. Lower still, the pebbles of the present streambed have a lithologic composition close to that of Fyb, with a very strong dominance of dark or grey-brown, more or less schistose siliceous rocks, some Triassic sandstones, quartz (with black siliceous veins) and limestones (tab. 1). The surfaces of several pebbles have traces of cement or coating that provide evidence of Fyb reworking.

28Moving downstream, near the entrance of the Gouleyrous Gorge, the Verdouble River flows at a height of 127 m. At approximately 10 m above the streambed, we observed a conglomerate not indicated on the map. Its homogeneous sparite cement supplied an age of 37,600 BP, whic is thus of an altitude and an age compatibles with Fya. The surface of the conglomerate is covered locally by a veneer of travertine, dated to 10,130 BP (Station 5, 6). A little before the entrance of Gouleyrous, the banks of the Verdouble and the nearby surfaces are covered with abundant non-cemented pebbles. This accumulation testifies to a sort of inner delta near the foot of the Devèze cliff.

4.2. Pleistocene alluvial deposits of the Tautavel-Vingrau depression (downstream of Gouleyrous Gorge)

29Gouleyrous Gorge is 700 m long and 150 m deep. It is a deeply incised gorge with nearly vertical Urgonian calcareous cliff. At the exit of this gorge, large erratic calcareous boulders of several meters diameter are observed near the foot of the cliff or at a few hundred metre distance farther indicating important collapses as a consequence of the nearby karstic erosion and probably of the recurrent earthquakes of the area (Gilli et al., 1999).

30In the reach immediately downstream from the exit of the Gorge, the unconsolidated gravel deposit of the lower alluvial accumulation reaches up to 3 m above the Verdouble streambed (Station 7, fig. 3). It contains bluish and light grey calcareous pebbles of Urgonian age, with the rest of the assemblage consisting of siliceous pebbles (ochre or brown Albian siltstones, beige Triassic sandstones, Ordovician schists) that are very widely scattered in the nearby area (tab. 1). This lower alluvial accumulation continues in a fragmentary state for 500 m along the right bank, breaks off, then reappears 500 m below at the approach to Tautavel while on the left bank it is reduced to some isolated pebbles in vineyard soils. A seismic study allowed us to observe the accumulation of one metre of gravel brought by the great flood of 1999 and covering the lower accumulation (Giresse et al., 2012). About 2 m in thickness this is in continuity indicating that we cannot characterize this alluvium between too narrow altitude limits. The lithologic composition of the pebbles of this recent flood is that one of the lower accumulation.

31Further above, between 115 and 120 m, and between 9 and 14 m above the Verdouble channel, the right bank is sprinkled with more or less scattered pebbles with some superficial cement remains. This deposit is typically 5-25 m wide, but locally may be as wide as 70 m. Some residual microcliffs of conglomerate testify to a cemented deposit (Fxb on the map, fig. 2B) that is less well preserved than in the Paziols plain. At the Seille locality (fig. 3; Stations 8, 9), this conglomerate still shows the dominance of siliceous pebbles over calcareous pebbles (tab. 1). More commonly, the dated isopachous fringes of calcite cement directly overlie lithoclasts. The remnant pore space above the cements is open. Millimetre-length sparite crystals compose a homogeneous and slightly rubefied cement which has been dated twice yielding two quite comparable ages of 34,600 and 38,500 BP, which classifies them within MIS 3. These ages thus confirm the estimation of > 35,000 BP obtained from the tests of gastropods from loam, slightly upstream of Tautavel (Berger et al., 1997).

32Further downstream another veneer of conglomerate, unknown on the map, was observed under the Museum of Tautavel (fig. 3, station 10); its altitude of 110-111 m places it 13-14 m above the Verdouble flowing here at 97 m. Calcareous pebbles are dominant over siliceous pebbles (especially sandstones and dark schists), which is a composition similar to that of the Fyb conglomerate, in the neighbourhood of Paziols (tab. 1). Generally, an isopachous single-layer phreatic cement was observed, but in contrast to the conglomerate further upstream, we note locally some multi-layer phreatic cement. In this last case, two distinct stages were noted: first a quite thick pink micritic cement coats the surface of clasts, in particular those of calcareous composition; the second cement, which is generally sparitic, is developed in certain cavities (fig. 6A and fig. 6C). The first generation micrites (fast crystallization) are dated to 20,220 BP and to 25,750 BP, the second-generation sparites (slow crystallization) are younger (14,120 and 13,110 BP). This accumulation is also contemporary with MIS 3, but that later cementation occurred, probably with the rewarming at the MIS 2 to 1 transition. A dating of the undifferentiated addition of these two cements would thus have supplied only an apparent age. 

Fig. 6Examples of two-steps cements. Tautavel Museum.
Fig. 6Exemples de ciments déposés en deux phases. Musée de Tautavel.

Fig. 6 – Examples of two-steps cements. Tautavel Museum. Fig. 6 – Exemples de ciments déposés en deux phases. Musée de Tautavel.

A: The first cementation, pigmented by ferric iron, is micritic (mi) and develops especially on the surface of calcareous pebbles, the second cementation is sparitic (sp) and respects some voids (v) edged by big palissadic crystals. B: On a white calcareous lithoclast (cl), passage (red arrow) from a first pink micritic cement (mi) to a white sparitic cement (sp). C: Higher magnification view of a sector of (A) showing the succession of a calcareous clast with a dark cover of oxides of manganese (m), the first red cementation compounds of micrites and microsparites, the second white cementation consisting of sparites ( sp ), the last 0.5 mm crystals constitute palissadic structure in border of geodes. D: Alzine: first step of cementation of the conglomerate: a 5 mm-long micritic bridge (mi) between two calcareous clasts (cl); the second step is a sparitic (sp) filling respecting some voids (v) (A, C, D: thin sections, plane light view; B: thin sections, cross polar view).
A : La première cimentation pigmentée par le fer ferrique est micritique (mi) et se développe plutôt à la surface des galets calcaires, la deuxième cimentation est sparitique (sp) et respecte quelques vides (v) ourlés de grands cristaux palissadiques. B : Sur un bioclaste blanc, passage (flèche rouge) d’un premier ciment micritique rose (mi) à un ciment sparitique blanc (sp). C : Fort grossissement d’un secteur de (A) montrant la succession d’un claste calcaire avec un revêtement sombre d’oxydes de manganèse (m). La première cimentation rouge est composée de micrites et de microsparites, la deuxième cimentation blanche est composée de sparites (sp), les derniers cristaux de 0,5 mm sont en structure palissadique en bordure de géodes. D : Alzine : première étape de cimentation constituée d’un pont micritique (mi) de 5 mm entre deux clastes calcaires (cl) ; deuxième étape constituée d’un remplissage sparitique (sp) respectant quelques vides (v), (A, C, D : lumière naturelle, B : lumière polarisée).

33Approximately 1 km further downstream, at the locality of Mas Alzine, we observe over Albian schists, the last downstream outcrop of conglomerates (Fxb on the map, fig. 2A) at approximately 15 m above the Verdouble (fig. 3, station 11). The slope of the hillside (120/1000) is comparable with those described in the Paziols Basin, but it is regular and like them, never presents any levelling of terrace (fig. 7). This 2 m-thick deposit is almost continuous for 50 m, but is cut locally by a reddish periglacial solifluction that packs angular fragments of Urgonian limestone. These rock flows are credibly contemporaries of the last glacial maximum (MIS 2), so postdate the alluvium conglomerate. The badly-sorted pebbles consist largely of limestone, as at the Museum station whereas silicated or siliceous elements are scarce (tab. 1). The cement is mainly sparitic: the first deposit presents a palissadic structure with microsparites pigmented by ferric iron or by manganese oxide coatings. The accumulation was interrupted locally by a thin fringe of micrite before the settlement of the second generation of larger size palissadic crystals intervenes. The first cement is estimated at over 48,000 BP; the second, which is the main cement, is dated to 26,400 BP, which is an age comparable to that of the Museum conglomerate. Locally, the first step of cementation is a micritic bridge between two calcareous clasts and the second step is a sparitic filling (fig. 6D). Neither of these cements showed indications of a vadose-like process as microstalactitic cement, silt vadose or late pustules on the underside of phreatic rind cement.

Fig. 7Topographic profile of the right-bank valleyside of the Verdouble at Mas Alzine.
Fig. 7Profil topographique du versant rive droite du Verdouble à hauteur du Mas d’Alzine.

Fig. 7 – Topographic profile of the right-bank valleyside of the Verdouble at Mas Alzine. Fig. 7 – Profil topographique du versant rive droite du Verdouble à hauteur du Mas d’Alzine.

Bed-rock is Albian schists and sandstones, which give rise to some minor irregularities of the relief. On the other hand, the pebbly deposit at + 15 m give rise to no corresponding terrace pattern.
Le soubassement est constitué par les schistes et les grès de l’Albien qui contrôlent quelques irrégularités mineures du relief. Par contre, le dépôt caillouteux à + 15 m ne correspond à aucun aplanissement du type terrasse alluviale.

5. Interpretation and discussion

5.1. The meaning of conglomerate cements and their dating

34Let us remember that the main part of the cementation, especially sparitic, of the conglomerates and sandstones was realized in a single stage even if it was incomplete (one-step cementation). In two cases only (the Museum and Alzine), the second generation of cement was able to settle locally in the voids (two-steps cementation).

35Cemented conglomerate exposures are spatially discontinuous and do not correspond to the present level of ground water emergence. This suggests that the geochemical conditions necessary for calcrete formation were not uniform in the catchment. Cementation may have been favoured by the exposure of large bank areas, leading to meteoric ground-waters with a relatively high content of dissolved calcium carbonate. The frequent lack of an impermeable layer beneath the calcareous conglomerate suggests that the carbonate did not form by the advection of water through the unit. Rind cements, especially in the Paziols area, are interpreted as having formed under water-saturated conditions – a phreatic environment (James and Choquette, 1987) – based on the continuous isopachous character and the lack of geopetal features. The distribution of cemented beds suggests that while the cements are phreatic in origin, the present water table is well below the exposures of cemented cobbles (Wright, 1989). This suggests that either the water table was higher in the past, during phreatic cement precipitation, or that local “perched” phreatic lenses developed during periods of high recharge. 

36The complete absence of organic matter indicates downward-fluxing meteoric waters without interaction with acidic soils: no dissolution feature was observed. We suggest that calcite cementation occurred when modest evaporation or degassing of CO2 occurred, promoting carbonate precipitation. Thus, it seems likely that some of the carbonate cement originated by deposition from downward-percolating calcium bicarbonate-laden waters after most of the overlying alluvium was deposited. If so, the conglomerate is unlike most soil caliche in origin (Rubin and Sieh, 1997).

37Another concern may be linked to the incorporation of “dead” calcareous carbon as recorded in stalagmite (Genty et al., 2001). Mesozoic calcareous lithoclasts are sporadically included in the alluvial fan. A resultant hard water effect would make the apparent age too old, but here, on the contrary, the measured ages turned out much younger than those generally suggested previously (Berger et al., 1997). The various MIS 3 and MIS 1 ages for the in situ cements are unsurprising and these ages suggest that the calcite does not contain a significant proportion of “dead” carbon from the dissolution of calcareous lithoclasts and that the hard water effect was negligible. Indirect proof is provided by ages obtained from conglomerates with dominant siliceous lithoclasts which are coherent with those of dominant calcareous lithoclasts.

38In summary, our 14C ages suggest that the cementation of both conglomerates of this upstream valley is of similar range (tab. 2): the identity and the continuity of cements do not seem to evoke a significantly earlier deposition, restricting chemical exchange with later atmospheric carbon dioxide. Based on similar deposition conditions, we can envisage that the estimates of > 45,000 yr BP at Moulin of Paziols and > 48,000 yr BP at Mas Alzine would be only marginally older than 45,000 BP. Although we did not observe secondary crystallization in these two cementations, these ages would indicate a main cementation at around 40,000 BP. On the other hand, we can consider that, in both the Museum and Alzine sites, the two distinct generations of cement were able to supply a closely connected and coherent chronology on the basis of a set of four dates. The age of the outer layer is younger than that of the inner layer, demonstrating that the environment has been closed and suggesting that 14C dating was useful.

5.2. Palaeoenvironmental conditions of conglomerate cementation

39Under slightly drier conditions, heavier meteoric waters due to higher temperatures during rainfall and the enhanced evaporation of soil water would be expected to favour the formation of the calcite cement.

40The cementation of the pebble deposits developed over a prolonged period between 40,000 and 25,000 yr BP. All cement ages appear to relate to the relatively warmer intervals of the Würm period in the ranges of 32,000-24,000 yr BP and 60,000-45,000 yr BP; these data are compatible with the snail shells data presented previously (Berger et al., 1997). The two stages of cementation in some conglomerates in the Tautavel-Vingrau Basin (Museum and Mas Alzine) can be understood in terms of intermittent interstitial water saturation in the downstream part of the valley. Thus, no cementation was recorded during the cold and dry paroxysm of the MIS 2. Lastly, somewhere around the MIS 2-1 transition, new carbonate precipitation was expressed by the localized deposition of travertine. These climatic intervals favourable to carbonated cementation were strengthened by ages recently obtained from the cements of alluvial conglomerates of another calcareous hillside, that of Roboul, a small tributary of Agly River, near Rivesaltes (study in progress by the authors).

41Arising from a total of 15 dates, it thus yields a coherent picture of nearly continuous carbonate cementation through the early temperate-climate of the Würm. This picture presents rather variable heights of the conglomerates, between 3 and 15 m above the current streamline (fig. 8). The previous distinction between “lower terrace” and “middle terrace” of the geological maps is no longer precise enough to be considered because it is about a continuous process of accumulation displayed enough in the time (20-25,000 years long). At most, we can notice the rarity or absence of cements in the lowest deposits (and presumably youngest) deposits, namely by going from MIS 3 and into MIS 2. The oldest Pleistocene conglomerates (Fu of the Tuchan map) show a cement which was altered and recrystallized at the same time, and is thus not datable; they are scarce in the Paziols Basin but totally absent in the Tautavel-Vingrau depression.

Fig. 8Altimetric positions of the dated alluvial deposits of the Verdouble catchment (stations 1-11) according to the upstream-downstream profile which leads from the Paziols Basin to that of Vingrau-Tautavel by crossing the Gouleyrous Gorge.
Fig. 8Positions altimétriques des dépôts alluviaux datés du bassin du Verdouble (stations 1 à 11) selon le profil amont-aval qui conduit du bassin de Paziols à celui de Vingrau-Tautavel en traversant les Gorges des Gouleyrous.

Fig. 8 – Altimetric positions of the dated alluvial deposits of the Verdouble catchment (stations 1-11) according to the upstream-downstream profile which leads from the Paziols Basin to that of Vingrau-Tautavel by crossing the Gouleyrous Gorge. Fig. 8 – Positions altimétriques des dépôts alluviaux datés du bassin du Verdouble (stations 1 à 11) selon le profil amont-aval qui conduit du bassin de Paziols à celui de Vingrau-Tautavel en traversant les Gorges des Gouleyrous.

In each station, the altitudes of the levels are evaluated with regard to the altitude of the nearby streambed of the Verdouble. 14C ages of the main phase of the conglomerate cementation are represented in red, those of late cementation and travertine are represented in green. Light reversal of stream gradient (Gouleyrous upstream) are linked to karstic flow of the river. On a profile of only approximately 10 km, the slope induced by the incision of the Verdouble is of the order of 35‰ and culminates at 150‰ at the crossing of the threshold of the Gouleyrous Gorge.
Dans chaque station, les altitudes des dépôts sont relatives à l’altitude moyenne du lit du Verdouble voisin. Les âges 14C de la phase principale de la cimentation carbonatée des poudingues sont figurés en rouge, ceux des cimentations tardives ou des travertins sont figurés en vert. Dans la partie à l’amont des Gouleyrous, de légères inversions de la pente sont liées à la circulation karstique dans le sous-sol calcaire crétacé ou jurassique. Sur un profil limité à environ 10 km, la pente conséquente de l’incision du Verdouble est de l’ordre de 35 et culmine à 150 ‰ dans la traversée du seuil des Gorges des Gouleyrous.

42Due to the absence of alluvial deposits dating to the Lower and Mid-Pleistocene, it is envisaged that the Verdouble flow through the Gouleyrous Gorge would have be totally effective during only the relatively recent Pleistocene (fig. 9). We could then envisage during the Middle Pleistocene, at the time of Tautavel Man, a still hospitable landscape with various springs and water supply points, albeit different from the current one.

Fig. 9Schematic proposal of Pleistocene evolution of Devèze upland, Gouleyrous Gorge incision and Verdouble River flow.
Fig. 9Proposition schématique d’évolution pléistocène du relief de la Devèze, de l’incision des Gorges des Gouleyrous et de l’écoulement du Verdouble.

Fig. 9 – Schematic proposal of Pleistocene evolution of Devèze upland, Gouleyrous Gorge incision and Verdouble River flow. Fig. 9 – Proposition schématique d’évolution pléistocène du relief de la Devèze, de l’incision des Gorges des Gouleyrous et de l’écoulement du Verdouble.

5.3. Trend of the Pleistocene alluvial dynamics of the Verdouble River

43More generally, the marked clastic accumulation of the Plain of Paziols downstream can be understood by a load loss connected to the morphological and hydraulic restriction represented by the calcareous upland of Devèze before crossing the barrier of the Gorge (fig. 9).

44On the scale of these alluvium deposits upstream of the Gorge, we notice a clear opposition between the mainly calcareous lithology of the pebbles of Fya and the especially siliceous lithology of the pebbles of Fyb (fig. 3). According to the dominance of the calcareous outcrops in the west of the basin and the siliceous outcrops of the Palaeozoic hills in the east, we can envisage dominant contributions of the main Verdouble during the early Würm and of the Ségure tributary during the late Würm. However, this first comment should perhaps be qualified to accord with the extensive occurrences of the conglomerates of the Tertiary Formation of Paziols where accumulations with siliceous pebbles are covered in the north by accumulations with calcareous pebbles (Calvet et al., 1991). It is possible that detrital reworking influenced the definition of the successive alluvium deposits, the dominant calcareous material thus representing distal origins (west) on a catchment scale, the siliceous dominance corresponding to more proximal sources (east).

45After the karst-induced excavation of Gouleyrous Gorge (fig. 9), the incision rapidly resulted in a new lower base level at the confluence with the Vingrau stream, which induced renewed regressive erosion of the upstream thalweg of this stream. Adjacent Pleistocene alluvial deposits, possibly including prehistoric remains, were probably eroded readily once the flow of the downstream Verdouble became more and more effective. The gradient of the Verdouble tends to become weaker as it reaches the Vingrau-Tautavel depression. Alluvial floods able to transport pebbles through the Gouleyrous Gorge seem to have been active only during the Early Würm, perhaps during MIS 4, certainly during MIS 3. The alluvial deposits represent roughly synchronous channel and flood-plain aggradation, followed by stream incision. In simple terms, streams aggraded because more sediment was supplied than could be carried downstream. However, the fans stopped forming because their streams incised when the Verdouble River base level was lowered according to the karstic dissolution. Under these conditions, only the rapid cementation allowed the preservation of the deposits. In any case, the absence of palaeosols in the fan gravels indicates that no significant pauses in deposition occurred between two alluvial depositions. On no account could any real fluviatile terrace morphology be formed, as the repetition of the morphodynamic processes was insufficient; so the accumulation came simply to overlap the hillsides with rather regular slopes.

5.4. Consequences for the palaeoenvironment of Tautavel Man

46Consequently to the Early Würm alluviation, in situ Lower Palaeolithic sites are preserved only in the upstream banks of the depression of Vingrau-Tautavel (Abélanet and Martzluff, 1990; Martzluff et al., 2013). Other extremely rare Old and Middle Palaeolithic artefacts, altered and rolled, are in marginal position and very scattered on the sides of the valley or in fluvial deposits downstream of the Gouleyrous Gorge. Considering the permanent settlement of hominids in the Caune Arago from 700 to 50 Ka and the abundant concentration of Lower and Middle Palaeolithic industries on the various old alluvial terraces of the plain of Roussillon, this almost total absence of open-air camps in the Tautavel-Vingrau Depression is unexpected. The most plausible explanation lies in a poor preservation of sites, these being totally destroyed by violent erosive floods. The same applies to the early phase of the Upper Palaeolithic, before the MIS 2: Gravettian and Solutrean sites are invariably in depression very distant from the torrential course of the Verdouble. It was during the MIS 2 that the relief seems to have stabilized in a form similar to that of the present day, so that camps of the Magdalenian peoples were able to be preserved on an alluvial deposit of the Verdouble, near its confluence with the Vingrau, while downstream, the alluvium contain only Holocene archaeological material.

47The preservation of Lower Pleistocene alluvial deposits of the Verdouble River is not confirmed. In their absence, we could envisage a crossing of the Gorge of Gouleyrous by the Verdouble River which would have been totally completed during the relatively late Pleistocene (fig. 9). Thus at the time of Homo erectus people, the river did not outflow as it does today. This landscape which was rather different from the present did not at all counter the likely presence of very attractive water sources close to the site that expressed the karstic resurgences varied by the underground flow of an important Palaeo-Verdouble.

48On the other hand, a periodic flow of the Vingrau stream might more plausibly represent the alluvial environment previously evoked; the alluvial deposits at + 25-30 m, situated just upstream of the confluence with the Verdouble, are associated with the only open-air site representing the lower Palaeolithic occupation of the valley and would come to the support of this hypothesis. Possible sites further downstream in the Tautavel depression, which could have represented other early Palaeolithic settlements, were probably destroyed during the MIS 3 alluviation and, possibly earlier, by the migration of the torrential Verdouble, which had just crossed the threshold of the karstic Gouleyrous Gorge.

6. Conclusions

49(1) Most of the Verdouble conglomerates correspond to one-step and, locally, to two-step calcite cementation resulting from downward-percolating calcium carbonate waters. Only, rapid cementation allowed the preservation of the deposits (very early diagenesis). AMS 14C ages of dominant calcareous lithoclasts are coherent with those of dominant siliceous lithoclasts suggesting a non-significant ''dead carbon'' or hard water effect.

50(2) The ages indicate recurrent interstitial water saturation in the downstream part of the valley between 60,000 and 45,000 yr BP and between 32,000 and 24,000 yr BP relating to relatively temperate intervals of MIS 3 and MIS 4, excluding the cold and dry MIS 2.

51(3) The previous distinction between “Lower terrace” and “Middle terrace” of the geological maps loses its significance because it is about a continuous process of alluvial accumulation without stepped terraces. Due to the absence of alluvial deposits dating back to the Lower and Mid-Pleistocene in the Vingrau-Tautavel depression, the connection between the upstream and downstream basin of the Verdouble River was initiated during the rather late (Early Würm) karst induced excavation and collapses of the calcareous Gouleyrous Gorge.

52(4) In situ lower Palaeolithic sites are preserved only in the upstream banks of the Vingrau-Tautavel depression attesting to erosion induced by the Würm alluviation. As a consequence, the downstream Verdouble landscape of Tautavel Man was attractive with various ponds and water sources linked to water resurgences. However, it was rather different from the present one without permanent drainage and without marked fluvial incision. Nevertheless, there is no reason to envisage a different landscape in the upstream valley of the Paziols area.

Haut de page

Bibliographie

Abélanet J. (1990) – Bilan de dix années de recherches sur les sites paléolithiques de plein air de la vallée de Tautavel-Vingrau. Travaux de Préhistoire catalane, Université de Perpignan, Centre d’Études Préhistoriques Catalanes, 6, 17-41.

Abélanet J., Martzluff M. (1990) – Note sur la découverte d’un campement de plein air de l’Homme de Tautavel, près de la Caune de l’Arago. Travaux de Préhistoire catalane, Université de Perpignan, Centre d’Études Préhistoriques Catalanes, 6, 37-39.

Baills H. (2003) – Les Conques. Des chasseurs et leur territoire, Héraul, 101, Université de Liège (Ed.), 222 p.

Baills H. (2008) – Un premier jalon gravettien dans les Pyrénées-Orientales : le Jas d’en Biel-1, L’Anthropologie, 112, 247-272.

Bard E., Rostek F., Ménot-Combes G. (2004) – A better radiocarbon clock. Science 303, 178-179.

Berger G.M., Fonteilles M., Leblanc D., Clauzon G., Marchal J.P., Vautrelle C. (1993) – Carte géologique de la France à 1/50 000. Feuille N° 1090 de Rivesaltes avec notice, BRGM, 119 p.

Berger G.M., Bessières G., Bilotte M., Viallard P. (1997) Carte géologique de la France à 1/50 000. Feuille 1078 de Tuchan avec notice, BRGM, 114 p.

Calvet M., Aguilar J.-P., Crochet J.-Y., Dubar M., Michaux J. (1991) Première découverte de mammifères oligocènes et burdigaliens dans les bassins de Paziols-Tautavel-Estagel (Aude et Pyrénées-Orientales). Implications géodynamiques. Géologie de la France, 1, 33-44.

Calvet M. (1996) Morphogenèse d'une montagne méditerranéenne, les Pyrénées orientales. Thèse doctorat d'Etat, Université Paris I, BRGM, doc. 255, 3 t., 1177 p.

Calvet M., Lemartinel B. (2002) – Précipitations exceptionnelles et crues éclair dans l’aire des Pyrénées méditerranéennes. Géomorphologie : relief, processus, environnement, 8-1, 35-49.

Collina-Girard J. (1976) Les industries archaïques sur galets des terrasses quaternaires de la Têt et du Tech (Catalogne française). Géologie Méditerranéenne 3, 183-190

Courty M.A., Marlin C., Dever L., Trembla Y., Vachier P. (1994) – The properties genesis and environmental significance of calcitic pendents from the High Arctic (Spitsbergen). Geoderma 61, 71–102.

Desruelles S., Fouache E., Ciner A., Dalongeville R., Pavlopoulos K., Kosun E., Coquinot Y., Potdevin J.-L. (2009) – Beachrocks and sea level changes since Middle Holocene: Comparison between the insular group of Mykonos-Delos-Rhenia (Cyclades, Greece) and the southern coast of Turkey. Global Planetary Change 66, 1-2, 19-33.

Dörfliger N., Fleury P., Le Strat P. (2008a) Caractérisation géologique et hydrogéologique des aquifères carbonatés karstiques sous couverture. Méthodologie, synthèse bibliographique et synthèse géologique régionale. Rapport intermédiaire BRGM-RP-56375-FR, 49 p.

Dörfliger N., Fleury P., Ladouche B. (2008b) Inverse modeling approach to allogenic karst system characterization. Ground Water 47, 3, 414-426.

Falguères C., Yokoyama Y., Shen G., Bischoff J.L., Ku T.-L., de Lumley H. (2004) – New U-series dates at the Caune de l’Arago, France. Journal of Archaeological Science 31, 941-952.

Fleury P., Ladouche B., Dewandel B., Dörfliger N.P., Le Strat P., Grunenwald G., Izac J.-L., Cubizolles J. (2009) Evaluation des ressources en eau souterraine des systèmes aquifères karstiques des Corbières. Phase III – Démonstration de la ressource. Rapport « final». BRGM/RP-57612-FR, 148 p.

Gaume E. (2002) – Éléments d’analyse sur les crues éclair. Mém. Doct. École Nationale Génie Rural Eaux et Forêt et Institut National Recherche Scientifique, Université du Québec, Canada, 359 p.

Genty D., Baker A., Massault M., Proctor C., Gilmour M., Pons-Branchu E., Hamelin B. (2001) Dead Carbon in stalagmites : Carbonate bedrock paleodissolution vs. ageing of soil organic matter. Implications for 13C variations in speleothems. Geochimica et Cosmochimica Acta 65, 20, 3443-3457.

Geyh M.A., Eitel B. (1998) – Radiometric dating of young and old calcrete. Radiocarbon 40, 2, 795–802.

Gilli I., Levret A., Sollogoub R., Delange P. (1999) Research on the february 18, 1996 earthquake in the caves of St-Paul-de-Fenouillet area (Pyrénéees Orientales). Geodinamica Acta 12, 3-4, 143-158.

Giret A. (1997) L'Agly et ses terrasses dans la traversée du Fenouillèdes. Quaternaire, 8, 4, 409-418.

Giresse P., Benkhelil J., Courp T., Pauc H., Saos T., Bassetti M.-A., Berné S., Gaullier V., Loncke L., Robin N. (2012) L’environnement géologique de la Caune de l’Arago, site préhistorique de l’homme de Tautavel. In Martzluff M., Catafau A., Galinier M. (Eds) : Tautavel, des hommes dans leur vallée. Presses Universitaires de Perpignan, 19-44.

Haynes Jr C.V., Eyles C.H., Pavlish L.A., Ritchie J.C., Rybak M. (1989) Holocene palaeoecology of the eastern Sahara; Selima oasis. Quaternary Science Review 8, 109-136.

James N.P., Choquette P.W. (1983) – Diagenesis - Limestones. Geoscience Canada 10, 159-161.

Ladouche B., Dörfliger N.P. (2004) – Synthèse de la caractérisation des systèmes karstiques des Corbières Orientales – Rapport final de la phase 1 du Projet « Evaluation des ressources en eau des Corbières », Vol 2-Caractérisation géologique et hydrogéologique du système karstique du synclinal du Bas-Agly. BRGM Report RP-529119-Fr. Orléans, France, www.brgm.fré.

Lumley de H., (1971) – Le Paléolithique inférieur et moyen du Midi méditerranéen dans son cadre géologique. T..II: Bas-Languedoc, Roussillon, Catalogne. Paris, CNRS (Ed), 433 p.

Lumley de H., Lumley de M.A. (1971) – Découverte de restes humains anténéanderthaliens datés du Riss à la Caune de l’Arago (Tautavel, Pyrénées-Orientales). Comptes Rendus Académie Sciences 272, série D, 1739-1742.

Lumley de H., Collina-Girard J., Abelanet J., Bazile F., Meignen L. (1976) – Les premières industries humaine en Languedoc méditerranéen et en Roussillon. In de Lumley H. (Ed.) : La Préhistoire française, Paris, CNRS, 1, 2, 777-794.

Lumley de H., Fournier A., Park Y.C., Yokoyama Y, Demouy A. (1984) Stratigraphie du remplissage pléistocène moyen de la caune de l’Arago à Tautavel. Etude de huit carottages effectués de 1981 à 1983. L’Anthropologie 88, 1, 5-18.

Martzluff M., Abélanet J. (1990) – Le campement magdalénien du Rec del Penjat (commune de Vingrau, P.-O). Travaux de Préhistoire Catalane 1989-90, C.E.P.C., Université de Perpignan, 6, 43-62.

Martzluff M. (2006) Pebble Culture, bifaces et érosion: le « Tautavélien » des terrasses quaternaires en Roussillon, Archéo 66, Bulletin de l’A.A.P.O., Perpignan, 21, 89-112.

Martzluff M., Grégoire S., Giresse P. (2013) Le Solutréen des Espassoles (Vingrau, Pyrénénées-Orientales). In Martzluff M., Catafau A., Galinier M. (Ed.) : Tautavel, des hommes dans leur vallée. Presses Universitaires de Perpignan, 59-196.

Ostermann M., Sanders D., Kramers J. (2006) – 230Th/234U ages of calcite cements of the proglacial valley fills of Gamperdona and Bürs (Riss ice age, Vorarlberg, Austria): geological implications. Austrian Journal Earth Science 99, 31-41.

Pustovoytov K. (1998) Pedogenic carbonate cutans as a record of the Holocene history of relic tundra-steppes of the upper kolyma valley (north-eastern Asia). Catena 34,185–195.

Ramsay P.J., Cooper J.A.G. (2002) Late Quaternary sea-level change in South Africa. Quaternary Research 57, 82-90.

Rubin C.M., Sieh K., (1997) Long dormancy, low slip rate, and similar slip-per-cent- event for the Emerson fault, eastern California shear zone. Journal of Geophysical Research 102, B7, 15319-15330.

Sacchi D. (1986) Le Paléolithique Supérieur du Languedoc et du Roussillon. Gallia Préhistoire, 21ème supplément, 284 p.

Selleck B., Baran J. (2003) Petrology and stable isotope geochemistry of Pleistocene and Holocene calcite cement in Kame Terrace Gravel, Central New York State, Northeastern. Geology and Environmental Sciences 25, 3, 186-196.

Srivastava P. (2001) – Paleoclimatic implications of pedogenic carbonates in Holocene soils of the Gangetic Plains, India. Palaeogeography, Palaeoclimatology, Palaeoecology 172, 207-222.

Taupin J.-D. (2003) Palaeoclimatic and paleohydrological interpretation on the origin of carconeos nodules in the Niger River terraces close to Niamey (Niger) using isotopic and geochemical tools. Hydrological Sciences Journal 48, 3, 473-487.

Vincent K.R., Church S.E., von Gerard P., Finger S.E. (2007) Geomorphology of Cement Creekand its Relation to Ferricrete Deposits. In Church S.E., von Guerard P., Finger S.E. (Eds): Integrated Investigations of Environmental Effects of Historical Mining in the Animas River Watershed, San Juan County, Colorado U.S. Department of the Interior, U.S. Geological Survey, Professional Paper 1651, 2, E16, 747-79.

Weninger B., Jöris O. (2008) Towards an Absolute Chronology at the Middle to Upper Palaeolithic Transition in Western Eurasia: A New Greenland Hulu Time-scale Based on U/Th Ages. Journal of Human Evolution 55, 5, 772-782.

Wright V. (1986) The polyphase karstification of the Carboniferous Limestone in South Wales. In Paterson K. and Sweeting M.M. (Eds): New Direction in Karst. Proceedings of the Anglo-French Karst Symposium, Geoabstract, Norwich, 569-580.

Wright V. (1989) A micromorphological classification of fossil and Recent calcic and petrocalcic microstructures. Proceedings International Wkg. Mig. Sol Micromorphology, San Antonio, Elsevier, Amsterdam.

Yi Ch, Liu K., Cui Z., Jiao K., Yao T., He Y. (2004) – AMS radiocarbon dating of late Quaternary glacial landsforms, source of the Urumqi River, Tien Shan - a pilot study of 14C on inorganic carbon. Quaternary International 121, 1, 99-107.

Haut de page

Annexe

Version française abrégée

Le Verdouble est une petite rivière affluente de l’Agly qu’elle rejoint après avoir incisé perpendiculairement des couches calcaires mésozoïques. La plus grande partie de ces calcaires a été l'objet d'importantes dissolutions karstiques dont témoignent le creusement des Gorges des Gouleyrous et la perte de l'Alzine aboutissant aux résurgences de l'étang de Salses-Leucate (Dörflinger et al., 2008b). Ainsi l'écoulement du Verdouble se fait à l’air libre dans son cours supérieur et devient souterrain dans son cours inférieur (fig. 1).

L’abri sous-roche préhistorique de Tautavel se situe à mi-hauteur de la colline de la Devèze et près du débouché des Gorges des Gouleyrous où le Verdouble est profondément encaissé dans les calcaires urgoniens. Ce site est devenu célèbre grâce aux découvertes successives des restes de crâne de l’Homo erectus tautavelensis (l’Homme de Tautavel) datés vers 450000 ans (de Lumley et al., 1984 ; Falguère et al., 2004). L’attractivité de cet abri fût liée à son exposition favorable et à la proximité des eaux du Verdouble appréciée tant par les animaux que par les hommes. Toutefois cette vision repose surtout sur les caractères actuels du paysage, caractères qui pouvaient être substantiellement différents, il y a 450 000 ans.

L’étude concerne les alluvions conglomératiques du Verdouble tant à l’amont (bassin de Paziols) qu'à l’aval (dépression de Vingrau-Tautavel) des Gorges (fig. 3). Elle aborde la pétrographie des lithoclastes afin de reconnaître leur origine et surtout l’étude microscopique de la texture et de la structure des ciments calcitiques. Ces ciments sont le plus souvent définis par une seule phase de cristallisation, rarement deux phases. Ils ont fait l’objet de microprélèvements à l’aide d’une fraise de dentiste, les quantités sélectionnées de l’ordre de la centaine de milligrammes ont été datées par AMS au 14C.

Les cartes géologiques au 1/50 000 de Tuchan (Berger et al., 1997) et de Rivesaltes (Berger et al., 1993) ne proposent que trois niveaux de dépôts alluviaux pléistocènes. Le plus ancien et le plus haut (Fu), assimilé au Pléistocène inférieur et moyen sensu largo, est observé dans la vallée amont du Verdouble, mais il disparaît totalement à l'aval des Gouleyrous. Par contre, les niveaux plus jeunes et plus bas (Fya, Würm ancien, Fyb, Würm récent) sont reconnus à la fois en amont et en aval des Gouleyrous (fig. 2A et fig. 2B).

Des inventaires archéologiques exhaustifs de la dépression de Vingrau-Tautavel, en contrebas de l'abri de l'Homme de Tautavel, n'ont mis en évidence qu'un seul site de campement du Paléolithique inférieur situé juste à l'aval de l'arrivée du Verdouble, exactement à 300 m à l'amont du confluent du ruisseau de Vingrau avec le Verdouble (Abélanet, 1990 ; Abélanet et Martzluff, 1990). Par contre, les restes d’industries du Paléolithique supérieur et du Néolithique s'avèrent abondants dans toute la dépression, mais cependant à distance du cours du Verdouble (fig. 4).

Dans le bassin de Paziols, les conglomérats de graviers et de galets forment des bancs de moins d'un mètre d'épaisseur qui affleurent à une altitude de 6 à 8 m au dessus du lit du Verdouble (Fya). Ils sont principalement composés de lithoclastes calcaires et ont conservé leur stratification oblique. Les ciments monophasés de grandes sparites calcitiques (fig. 5) ont été datés entre 31400 BP et 40000 BP, certains âges étant supérieurs à 45000 BP. Plus en aval, à 10 m au dessus du lit (Fya), les ciments de conglomérats à galets ici surtout siliceux sont datés à 37600 BP, d'autres à 3 m au-dessus de ce même lit (Fyb) sont datés à 31400 BP (tab. 1 et tab. 2). Il est à noter que d'abondants galets non cimentés sont trouvés épars sur les rives et laissent présumer de pertes de charge répétées à l'approche du relief de la Devèze.

Dans la dépression Vingrau -Tautavel, à quelques dizaines de mètres du pied de la falaise, de grands blocs calcaires plurimétriques témoignent d'effondrements anciens des voutes ou des parois des Gorges. Plus bas, des galets non cimentés sont accumulés sur près de deux mètres d'épaisseur sur la rive droite du Verdouble (Fyb). Les conglomérats affleurent sur la rive droite entre 9 et 14 m au-dessus du fleuve où ils sont moins bien préservés que dans le bassin en amont (Fya). Les lithoclastes à dominante siliceuse sont cimentés par une frange isopaque de sparites datée à 34600 et à 38500 BP (tab. 1 et tab. 2). En aval, à hauteur du Musée de Préhistoire, un conglomérat est observé à 13-14 m au dessus du Verdouble (Fya). Ses galets à dominante calcaire sont souvent cimentés par une seule enveloppe sparitique, mais localement, on peut parfois observer un premier remplissage micritique (première cristallisation rapide) daté à 20220 et à 25750 BP et des remplissages sparitiques plus tardifs (cristallisation lente) qui sont datés à 14120 et à 13110 BP (fig. 6). Enfin, juste à l'amont de la perte de l'Alzine, un conglomérat de 2 m d'épaisseur est observé vers 15 m au-dessus du fleuve (Fya); son ciment micritique est estimé au delà de 48000 BP alors qu'une seconde génération de ciment sparitique est datée à 26400 BP, âge compatible avec ceux obtenus pour les conglomérats voisins du Musée (fig. 6, tab. 1 et tab. 2). Un profil topographique de ce versant de l'Alzine ne présente aucun gradin qui puisse être assimilé à un nivellement de terrasse fluviatile (fig. 7).

Les analyses lithologiques des galets démontrent leur origine plutôt occidentale (calcaires mésozoïques) pendant le Würm ancien et plutôt orientale (grès paléozoïques) pendant le Würm plus récent.

Les cimentations sparitiques, généralement isopaques, traduisent une sursaturation des eaux interstitielles en conditions phréatiques (James et Choquette, 1983). Il s'agissait de nappes moins profondes qu'aujourd'hui ou encore de nappes perchées. La cohérence des âges, notamment entre ciments avec galets calcaires et avec galets siliceux, n'indique pas d'intervention d'un carbone « mort » qui, par ailleurs, aurait « vieilli » les âges. En particulier, cette cohérence est aussi respectée dans les cas particuliers où deux cimentations se sont succédées dans les mêmes pores. Seule une cimentation intervenant juste après le dépôt (diagenèse très précoce) a permis la conservation de celui-ci alors qu'aux environs de nombreux exemples de galets épars témoignent de son démantèlement.

Ce sont des interstades tempérés et relativement humides du MIS 3 et/ou du MIS 4 qui ont été les plus favorables à ces cimentations, à l'exclusion du MIS 2 plus sec et plus froid. Ces interstades, ici entre 60000 et 45000 BP et entre 32000 et 24000 BP, viennent aussi d'être mis en évidence dans les ciments des conglomérats observés sur les versants du Roboul, autre affluent de l'Agly (travaux en cours des auteurs). Ces datations rendent obsolètes les étagements proposés de « basse terrasse »' et de « moyenne terrasse » : pendant le Würm, les crues ravageuses du Verdouble ont accumulé des nappes d'alluvions caillouteuses, mais elles les ont aussi remaniées et seuls quelques témoins cimentés ont survécu à distance du cours principal. Ces crues n'ont pu se propager qu'à partir du moment où le couloir des Gorges s'est trouvé dégagé. Leur irruption a pu s'initier et se développer en fonction des effondrements des voûtes des cavités karstiques (fig. 8 et fig. 9). Le creusement karstique des Gorges et, par corollaire, l'abaissement du niveau de base du Verdouble, ont été favorables à l'activité alluviale qui a probablement détruit les vestiges des campements du Paléolithique inférieur à l'aval de la dépression de Vingrau-Tautavel. Les environnements anciens, et notamment celui de l'Homme de Tautavel, offraient des conditions attractives avec de multiples sources et résurgences qui devaient alimenter des marécages au pied des falaises, mais ce paysage était sensiblement différent de celui que nous propose aujourd’hui la vallée du Verdouble.

Haut de page

Table des illustrations

Titre Fig. 1 – General map of the course of the Verdouble River, tributary of the Agly River, from Padern until the Alzine loss.Fig. 1Carte générale du cours du Verdouble, affluent de l’Agly depuis Padern jusqu‘à la perte de l’Alzine.
Légende Numbers 2a and 2b refer to following maps .Les numéros 2a et 2b correspondent aux cartes de la figure suivante.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-1.jpg
Fichier image/jpeg, 3,6M
Titre Fig. 2 – Geologic sketch of the Verdouble catchment. Fig. 2 – Carte géologique simplifiée du bassin du Verdouble.
Légende A: In the upstream of Gouleyrous Gorge (Paziols plain) according to the geologic map of Rivesalte (Berger et al., 1997). L: Lias, J: Jurassic, C: Cretaceous, Ci: Lower Cretaceous; g2-3: Oligocene conglomerates. Only the Pleistocene fluviatile deposits are exactly reproduced; the pebble accumulation (inner delta fan) in the upstream of Gouleyrous Gorge was added by the authors. River losses are indicated by arrows. B: In the downstream of Gouleyrous Gorge (depression of Vingrau-Tautavel) according to the geologic map of Tuchan (Berger et al., 1993). J: Jurassic, Ci: Lower Cretaceous; g: Oligocene conglomerates. Only the Pleiostocene fluviatile deposits are exactly reproduced.A : À l’amont des Gorges des Gouleyrous (plaine de Paziols) d’après la carte géologique de Tuchan (Berger et al., 1997). L : Lias, J. Jurassic, C : Cretaceous, Ci : Lower Cretaceous, g2-3 : conglomérats oligocènes. Seuls les affleurements des dépôts fluviatiles pléistocènes sont précisément reportés, l’accumulation caillouteuse à l’amont des Gorges a été ajoutée par les auteurs. Les pertes de la rivière sont indiquées par les flèches. B : À l’aval des Gorges des Gouleyrous (dépression de Vingrau-Tautavel) d’après la carte géologique de Rivesaltes (Berger et al., 1993) ; J : Jurassique, Ci : Crétacé inférieur, g : conglomérats oligocènes.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-2.jpg
Fichier image/jpeg, 2,8M
Titre Fig. 3 – Location of the analysed conglomerates. Fig. 3Localisation des conglomérats analysés.
Légende The origin of the pebbles of Fya (mostly calcareous) is the western part of the Verdouble catchment (main course of the Verdouble), whereas the pebbles of Fyb (mostly siliceous) may come from the eastern part of the basin (Donneuve and Tarassac tributaries).L'origine des galets de Fya (surtout carbonatés) est issue des zones occidentales du bassin (cours principal du Verdouble), celle des galets de Fyb (surtout siliceux) proviendrait de la partie orientale du bassin (affluents Donneuve et Tarassac).
URL http://geomorphologie.revues.org/docannexe/image/10956/img-3.jpg
Fichier image/jpeg, 1,5M
Titre Fig. 4 – Archaeological sites in Vingrau-Tautavel Depression. Fig. 4 – Sites archéologiques de la dépression Vingrau-Tautavel.
Légende With dark grey colour, calcareous uplands and cliffs above the Depression. Broken line bounds the Upper Palaeolithic artefacts area.En gris foncé, reliefs calcaires et falaises encadrant la dépression. 1 : Paléolithique inférieur, astérisque : outils moustériens erratiques ; carré : Paléolithique supérieur ancien (Gravettien, Solutréen) ; rond : Paléolithique supérieur récent (Magdalénien et Épipaléolithique). 2 : Site de la Teulera. La ligne en tiretés souligne les limites des sites du Paléolithique supèrieur.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-4.jpg
Fichier image/jpeg, 4,0M
Titre Fig. 5 – Examples of one-step cements. Fig. 5Exemples des ciments déposés en une seule phase.
Légende A: Moulin, the skeleton of the conglomerate results from a single sparitic border (sp) linking the gravels of sandstones (s) and schists (sc), the void volumes (v) remain very important. B: Moulin, the first micritic cement (mi) developed preferentially on the surface of a calcareous clast (at the top) rather than on that of a siliceous clast (at the bottom), sparites (sp) attain a size from 1 to 2 mm. C: Upward Ford station, one single phase deposition of sparitic cement (sp) linking various gravels, frequent voids (v). D: Elongated cavity (c) between two calcareous lithoclasts (cl), the cavity is fringed by a single sparitic rim (sp) (A, D: surface sections; B, C: thin sections, cross polar). A : Moulin, la charpente du conglomérat résulte d’un seul ourlet sparitique (sp) liant les graviers de grès (s) et de schistes (sc), les vides (v) sont demeurés très importants. B : Moulin, le premier ciment microsparitique (mi) s’est développé préférentiellement à la surface d’un claste calcaire (en haut) plutôt qu’à celle d’un claste silicaté (en bas), les sparites (sp) atteignent ici une taille de 1 à 2 mm. C : Amont du gué, une seule phase de cimentation sparitique (sp) ourle les vides (v) et lie différents graviers. D : La Seille, cavité étroite (c) développée entre deux lithoclastes calcaires (cl) et bordée par une seule frange de sparites (sp) (A, D : sections de surface ; B, C : lames minces, polarisation croisée).
URL http://geomorphologie.revues.org/docannexe/image/10956/img-5.jpg
Fichier image/jpeg, 6,7M
Titre Fig. 6 – Examples of two-steps cements. Tautavel Museum. Fig. 6Exemples de ciments déposés en deux phases. Musée de Tautavel.
Légende A: The first cementation, pigmented by ferric iron, is micritic (mi) and develops especially on the surface of calcareous pebbles, the second cementation is sparitic (sp) and respects some voids (v) edged by big palissadic crystals. B: On a white calcareous lithoclast (cl), passage (red arrow) from a first pink micritic cement (mi) to a white sparitic cement (sp). C: Higher magnification view of a sector of (A) showing the succession of a calcareous clast with a dark cover of oxides of manganese (m), the first red cementation compounds of micrites and microsparites, the second white cementation consisting of sparites ( sp ), the last 0.5 mm crystals constitute palissadic structure in border of geodes. D: Alzine: first step of cementation of the conglomerate: a 5 mm-long micritic bridge (mi) between two calcareous clasts (cl); the second step is a sparitic (sp) filling respecting some voids (v) (A, C, D: thin sections, plane light view; B: thin sections, cross polar view). A : La première cimentation pigmentée par le fer ferrique est micritique (mi) et se développe plutôt à la surface des galets calcaires, la deuxième cimentation est sparitique (sp) et respecte quelques vides (v) ourlés de grands cristaux palissadiques. B : Sur un bioclaste blanc, passage (flèche rouge) d’un premier ciment micritique rose (mi) à un ciment sparitique blanc (sp). C : Fort grossissement d’un secteur de (A) montrant la succession d’un claste calcaire avec un revêtement sombre d’oxydes de manganèse (m). La première cimentation rouge est composée de micrites et de microsparites, la deuxième cimentation blanche est composée de sparites (sp), les derniers cristaux de 0,5 mm sont en structure palissadique en bordure de géodes. D : Alzine : première étape de cimentation constituée d’un pont micritique (mi) de 5 mm entre deux clastes calcaires (cl) ; deuxième étape constituée d’un remplissage sparitique (sp) respectant quelques vides (v), (A, C, D : lumière naturelle, B : lumière polarisée).
URL http://geomorphologie.revues.org/docannexe/image/10956/img-6.jpg
Fichier image/jpeg, 6,7M
Titre Fig. 7 – Topographic profile of the right-bank valleyside of the Verdouble at Mas Alzine. Fig. 7Profil topographique du versant rive droite du Verdouble à hauteur du Mas d’Alzine.
Légende Bed-rock is Albian schists and sandstones, which give rise to some minor irregularities of the relief. On the other hand, the pebbly deposit at + 15 m give rise to no corresponding terrace pattern. Le soubassement est constitué par les schistes et les grès de l’Albien qui contrôlent quelques irrégularités mineures du relief. Par contre, le dépôt caillouteux à + 15 m ne correspond à aucun aplanissement du type terrasse alluviale.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-7.jpg
Fichier image/jpeg, 1,4M
Titre Fig. 8 – Altimetric positions of the dated alluvial deposits of the Verdouble catchment (stations 1-11) according to the upstream-downstream profile which leads from the Paziols Basin to that of Vingrau-Tautavel by crossing the Gouleyrous Gorge. Fig. 8Positions altimétriques des dépôts alluviaux datés du bassin du Verdouble (stations 1 à 11) selon le profil amont-aval qui conduit du bassin de Paziols à celui de Vingrau-Tautavel en traversant les Gorges des Gouleyrous.
Légende In each station, the altitudes of the levels are evaluated with regard to the altitude of the nearby streambed of the Verdouble. 14C ages of the main phase of the conglomerate cementation are represented in red, those of late cementation and travertine are represented in green. Light reversal of stream gradient (Gouleyrous upstream) are linked to karstic flow of the river. On a profile of only approximately 10 km, the slope induced by the incision of the Verdouble is of the order of 35‰ and culminates at 150‰ at the crossing of the threshold of the Gouleyrous Gorge. Dans chaque station, les altitudes des dépôts sont relatives à l’altitude moyenne du lit du Verdouble voisin. Les âges 14C de la phase principale de la cimentation carbonatée des poudingues sont figurés en rouge, ceux des cimentations tardives ou des travertins sont figurés en vert. Dans la partie à l’amont des Gouleyrous, de légères inversions de la pente sont liées à la circulation karstique dans le sous-sol calcaire crétacé ou jurassique. Sur un profil limité à environ 10 km, la pente conséquente de l’incision du Verdouble est de l’ordre de 35 et culmine à 150 ‰ dans la traversée du seuil des Gorges des Gouleyrous.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-8.jpg
Fichier image/jpeg, 2,4M
Titre Fig. 9 – Schematic proposal of Pleistocene evolution of Devèze upland, Gouleyrous Gorge incision and Verdouble River flow. Fig. 9Proposition schématique d’évolution pléistocène du relief de la Devèze, de l’incision des Gorges des Gouleyrous et de l’écoulement du Verdouble.
URL http://geomorphologie.revues.org/docannexe/image/10956/img-9.jpg
Fichier image/jpeg, 2,8M
Haut de page

Pour citer cet article

Référence papier

Pierre Giresse et Michel Martzluff, « AMS radiocarbon dating of carbonate cements in late Pleistocene alluvial conglomerates, Verdouble River. Palaeoenvironmental implications concerning the Palaeolithic site of Tautavel (Pyrénées-Orientales) », Géomorphologie : relief, processus, environnement, vol. 21 – n° 2 | 2015, 115-130.

Référence électronique

Pierre Giresse et Michel Martzluff, « AMS radiocarbon dating of carbonate cements in late Pleistocene alluvial conglomerates, Verdouble River. Palaeoenvironmental implications concerning the Palaeolithic site of Tautavel (Pyrénées-Orientales) », Géomorphologie : relief, processus, environnement [En ligne], vol. 21 – n° 2 | 2015, mis en ligne le 01 janvier 2016, consulté le 21 octobre 2017. URL : http://geomorphologie.revues.org/10956 ; DOI : 10.4000/geomorphologie.10956

Haut de page

Auteurs

Pierre Giresse

Centre de Formation et de Recherches sur les Environnements Méditerranéens – UMR 5110 – Université de PerpignanVia Domitia, 52, Avenue Paul Alduy66860, PerpignanFrance (giresse@univ-perp.fr). Tél.:  33 (0)4 68 66 20 56 ; Fax : 33 (0)4 68 66 20 96.

Michel Martzluff

Mediterra – Université de PerpignanVia Domitia, 52 avenue Paul Alduy – 66860, Perpignan – France (michel.martzluff@orange.fr).

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