Content uploaded by Eric Gilli
Author content
All content in this area was uploaded by Eric Gilli
Content may be subject to copyright.
1
ICS - ISLAND SYMPOSIUM - July 2009
KERVILLE - TEXAS
accepté le 24 février 2009
Mechanism of salt contamination of karstic springs related to the Messinian deep stage.
The speleological model of Port Miou (France).
by Eric Gilli1 & Thomas Cavalera2
1University Paris 8 and UMR Espace 6012 Nice.
2University of Provence
Abstract
Submarine karst springs are common on the Mediterranean shore but most of them are brackish, which limits
their usefulness. Various experiments to limit freshwater/seawater mixing were attempted in the past but have failed.
This study examined Port Miou (Cassis, France) showing that salinity is present at 2300 m from the entrance and at a
depth of 179 m bsl. The geometry of this setting is inherited from a complex palaeogeography. The lowering of the
Mediterranean during the Messinian salinity crisis has made possible the existence of caves several hundred meters
below present sea level. The presence of titanium in the sediment of the cave seems to be a residual product of an
alumina factory that is discharged at a depth of 300 m, 6 km south of the cave. This supports the hypothesis of a deep
aspiration of seawater by a Messinian gallery. A similar example exists in Kefalonia Island, where a marine intrusion is
observed in coastal sinkholes.
1. Introduction
Submarine karst springs are common on the Mediterranean shore but most of them are brackish, which limits
their usefulness. For the southeast coastline of France, these aquifers are estimated at 10 m3/s. For the entire
Mediterranean basin, the potential is approximately 1000 m3/s which theoretically supports the domestic needs of 350
million persons. Numerous attempts to catch these springs (dams, flexible or rigid artificial insulations, pumping, etc.)
ended in failure (Gilli, 2003). The principle was generally to artificially augment the hydraulic gradient in order to
lower the contact between fresh water and sea water. Recent explorations of deep submarine caves explain why these
attempts failed.
Cave divers or ROV have reached important depths in several places : 308 m (224 m b.s.l.) in the Fontaine de
Vaucluse (France), 179 m (179 m b.s.l.) in the Port Miou spring (France). Similar examples exist also in Greece, Italy,
Spain and Turkey. These depths are more important than the lowering of the water table related to the Quaternary
glacio-eustatism (120 m b.s.l.) and several authors imagine a possible settlement of the karstic systems during the
Messinian stage [-5.95 to –5.35 Ma] when a large drop in sea-level, down to 1500 m, occurred in the Mediterranean
Sea (Ryan et al, 1973 ; Gautier et al, 1994 ; Krijgsman et al, 1999). It caused the sedimentation of important evaporitic
levels on the bottom of the sea, and the presence of deep submarine canyons on the margins of the basin. Deep canyons
were also formed inland that are now filled with alluviums. This drastic drop in base level affected the local karst
systems and caused the deepening of the water circulations and the possible reopening of palaeokarsts. Several authors
pointed out the influence of the Messinian event on the French karst (Julian and Nicod, 1984 ; Audra et al, 2004) and it
has been widely recognized in Italy (Bini, 1994), in Greece (Arfib, 2001) and in Spain (Fleury, 2005). Recent works on
karstic springs suggest the possibility of a deep salt contamination of the main Mediterranean brackish springs, due to
the presence of deep karstic galleries related to that Messinian model (Gilli, 2001). The current study of the shoreline
aquifers of southeastern France has enabled us to propose an operating model.
2 . The Port Miou system
In Cassis ( South-eastern France) the springs of Port-Miou and Bestouan are the outlets of an important system
of submarine karstic galleries explored since the 50's (table 1). The average discharge is between 2 to 5 m3s-1 but the
water is brackish and cannot be used for water supply. The Port Miou cave is a 2.4 kilometers long gallery that extends
in the limestone series of Calanques. The total length of Bestouan is 3.7 kilometers. Dye tests and monitoring have
proved the relation between the two caves but no connecting passage has been discovered. At Port Miou a dam was
built in the 70's inside the cave, to prevent the marine intrusion, but in spite of a noticeable decrease of the salinity
down to 3 gL-1, it had not been possible to obtain drinkable fresh water upstream of the dam (Potié, 1974) (fig. 1). The
use of helium and later rebreathers by cave divers made possible the exploration of a vertical pit, down to 179 m below
the sea level, at the end of the cave. At that depth, the water is still brackish.
2
Table 1 : History of explorations and studies in Port Miou and Bestouan (Cassis, France)
Figure 1 : Cross section of the Port Miou gallery.
Discharge, temperature and salinity of the spring have been collected for several months in the 70's, the
discharge varied from 2 m3s-1 to 100 m3s-1 and the salinity was from 20 gL-1 during low water to 0 gL-1during the
floods. New data is recorded since December 2004 with an average discharge of 3 m3s-1 and salinity between 14 and
3.4 gL-1. The average values indicates a medium flow of seawater close to 1 m3s-1 but the mechanism of the
contamination is still unknown. Two hypotheses are possible that offer two different possibilities to reduce the salinity
:
- a classical hypothesis where the fresh water circulates in a deep karstic gallery connected to a matricial
fissured zone polluted by sea water. The diminution of the salinity is possible by augmenting the pressure of fresh
water in the conduit (Arfib, 2001). In Port Miou the study of the discharge/salinity graphs did not make it possible to
see such a mechanism (Arfib et al, 2006 ; Cavalera, 2007).
- a speleological hypothesis where the fresh water circulates in a deep karstic gallery connected to the sea by
another gallery (Gilli et al, 2004). This suggests that the permeability of the limestone is low.
In the area of Port Miou, this second hypothesis is supported by the existence of karst features below current sea
level. The bathymetric map of Lion gulf (Berne et al, 2002) and a morpho-bathymetric study (Collina-Girard, 1996)
reveal the existence of a limestone plateau that extends a few kilometers south to the Calanques, with dolines at a depth
1953 1s
t
true exploration in the Port Miou cave by the EOLE team (J. Blanc, J. Picard, et M. Galerne)
1955-1956
Exploration and study of submarine galleries by O.F.R.S. (Cdt Cousteau) : the team reach 280 m in
Port Miou and 40 m in Bestouan. Topographic, thermographic and faunistic data is collected
(CORROY et al, 1958)
1960 A US scientific diver (C. Limbaugh) get lost and die during a photography session in Port Miou
1964-1966 Divers of G.E.P.S. (Claude Touloumdjian) reach 400 m in Port Miou et 450 in the Bestouan.
1968 Discovery in Port Miou of an aerial zone at 530 m from the entrance et exploration of the gallery up to
870 m.
1968-1973 Study and construction of an underground dam by the SRPM. (POTIÉ,1974)
1978 C. Touloumdjian reaches 1365 m in Port Miou and Francis Leguen 1400 m in Bestouan
1981 Discovery in Port Miou of a vertical shaft at 2230 m from the entrance by B.Léger. The terminal depth
is -82 m bsl
1982 Two Italian and Swiss divers get lost and die in Port Miou.
1980-1983 The divers C. Touloumdjian and F. Leguen reach 2290 m in Bestouan
1990-1993 Series of explorations by C.R.P.S. (Comité Régional de Plongée Souterraine) during which the divers
reach 3000 m in Bestouan. In Port Miou M.Douchet goes down to -147 m bsl (DOUCHET,1993).
2001 Starting of new geological studies (GILLI, 2001)
2005 J Meynié, using rebreather, reaches -172 m bsl in Port Miou.
2008 X. Meniscu, using rebreather, reaches -179 bsl in Port Miou
3
of 150 m below present sea level. A deep submarine valley, the canyon of Cassidaigne, whose bottom is at a depth of
1000 m, cuts this plateau. This canyon looks like a karst pocket valley and is not connected to a continental valley.
Several dives with small submarines have revealed on the walls the existence of caves with speleothems (oral
communication from COMEX). We assume that, during the Messinian deep stage, the underground river of Port-Miou
was flowing 200 or 300 hundred meters below its current position and has excavated the canyon of Cassidaigne. At the
end of the Messinian deep stage, the system was breached by the sea, causing the fresh water to flow through an upper
gallery. Now the presence of a deep paleo-drain, filled by seawater, could provoke a saltwater intrusion into the karst
system.
3. Three pipes model
A three pipes model can be used to understand the mechanism in static conditions (Drogue,1993) (fig.2, left
panel). At equilibrium :
(H+∆H1)ρ1 = (H+∆H2)ρ2 = (H+∆H3)ρ3
With H depth of mixing zone, ∆H1 karst hydraulic gradient, ρ1 density of fresh water,
∆H2 altitude of brackish spring, ρ2 density of brackish water,
∆H3 variation of sea level, ρ3 density of sea water.
We have developed a laboratory model of the system, with 3 pipes respectively filled with colored sea water,
fresh water and brackish water (fig. 2-right). It perfectly simulates the mechanism. When fresh water is injected in the
fresh water pipe (∆H1 augments), a current moves towards the brackish outlet where the density is lower than in the sea
pipe, ρ2 decreases by dilution and this creates a dilution and an aspiration (negative ∆H3) in the seawater pipe.
Figure 2 : Three pipes model.
In Port Miou, the presence of an anthropic tracer supports this affirmation. In the area of Gardanne a factory of
alumina generates important quantities of residual products locally called "red mud". The red mud contains an
important quantity of heavy metals notably titanium and chromium. This mud is mixed with water and is transported
since the 70's by a submarine pipe towards the canyon of Cassidaigne where it is discharged into the sea, at a depth of
300 m (fig. 3).
4
Figure 3 : The speleological model of Port Miou (Cassis, France).
Several samples of sediments (surface and core samples) were collected in the Port Miou gallery upstream and
downstream of the dam. Most of samples upstream the dam present a thin level of red mud laying on a thick level of
gray to brown marine silt. The upper layer contains a much more important quantity of titanium than samples collected
in this area (table.2). When analyzing the cave sediment a few centimeters below its surface the concentration of
titanium is much lower. Downstream of the dam the concentration is also lower. This shows that red particles, rich in
titanium, have been recently transported by a current of brackish water, from the bottom of the cave to the entrance.
This supports an aspiration of sea water with particles of red mud by a deep gallery connected to the canyon of
Cassidaigne. Other possible explanations have been unsuccessfully studied. The factory, where the alumina is
processed, is on an impervious substratum. The inland wastes have low concentration of titanium. There are no leaks
on the pipeline. Environmental studies (Arnoux and Stora, 2003) show that the red mud, discharged in the canyon,
remains at an important depth and is never present close to the spring in the continental platform. One explanation
could be an underground sea current in a karstic conduit that transports the particles of red mud towards the north. New
analysis is underway to confirm it.
Sample Cr Cu Ti V Fe Mn Ni Pb Zn
"Red muds" (in factory) 1815 23,4 32 415 717 184 500 351 n.d. 94,5 38,3
Marine sediment with red muds in
Cassidaigne canyon 64 31.5 168 37,4 26 003 1113 n.d. 25,4 81
PM100705 surface sediment of Port Miou gallery 36.6 12,9 1992 71,3 21 101 888,65 36,3 22,6 197,2
PM100705_1 surface sediment of Port Miou gallery 1600
PM 191105_1 surface sediment of Port Miou gallery 400
PM 191105_2 surface sediment of Port Miou gallery 640
PM160405 sediment of Port Miou : 10 cm depth 0.48
PM160405 sediment of Port Miou : 40 cm depth 0.26
PM160405 sediment of Port Miou : 80 cm depth 0.22
BE140606 sediment of Bestouan gallery : surface 40 n.d. 680 0,07 27 1500 40 20 160
BE140606 sediment of Bestouan : 5 cm depth 360
BE140606 sediment of Bestouan : 15 cm depth 260
BE140606 sediment of Bestouan : 30 cm depth 220
5
BE140606 sediment of Bestouan : 38 cm depth 430
Continental karstic sediment (terra rosa) 610
Roucas Blanc (sediment of a brackish karstic spring) <10 <10 20 10 1100 -. -. <10 -.
Continental waste of La Barasse (ground) 490 20 17 0,6 275 000 -. - 370 -
Continental waste of La Barasse (leachate) 60 30 1,1 0,13 38 000 - - 60 -
Table 2 : Concentration of heavy metals in Port Miou and the area of Cassis (France).
4. The Argostoli system
Examples of marine aspiration are already known in Mediterranean sea at Argostoli (Drogue, 1989, 1993 ;
Maurin and Zötl, 1965) and Bali (Greece) (Arfib, 2001) or Moraig-Toix (Spain) (Cortes et al, 2000). Near the small
city of Argostoli, on the western coast of Kefalonia island, a sea intrusion is observed in coastal sinkholes. The flow
rate of the seawater is sufficient to power mills (fig.4). A dye test conducted in 1963 showed a communication, after 16
days, with the brackish springs of Sami on the eastern coast of the island (Maurin and Zötl, 1965). In the sinkholes, the
level of the seawater sumps is 1 m bsl. In the Sami springs it is 1 m asl.
Figure 4 : Argostoli mill in Kefalonia (Greece).
5. Conclusion
This speleological model can probably be extended to other karstic springs in the Mediterranean sea. It opens
two main directions for the catchment of water : drilling inland, far enough from the coast, to reach the aquifer
upstream of the mixing with sea water, or obstructing the deep gallery to prevent the seawater intrusion. A comparison
with the Floridian karst, where geothermal conditions are evoked (Kohout et al, 1977) to explain the presence of deep
karstic galleries should be very interesting to do.
Acknowledgement
Acknowledgement : This study was realized with the help of the Conservatoire du littoral (Mr Estève), the ONF (Mr
Vincent), the Société des Eaux de Marseille( MM. D'Aspe, Onatsky and Lieutaud ), the Alcan-Rio Tinto society (Mrs
Raignault), the Camargo fundation (Mr Dautricourt). Monitoring and sampling was done by the divers : C.
Touloumdjian, M. Douchet, J. Meynié, F. Tessier and by our colleagues B. Arfib, D. Chevaldonné.
References
Arfib, B. (2001) Etude des circulation d’eaux souterraines en aquifère karstique côtier: observations et modélisation de
la source saumâtre Almyros d’Héraklion, Crète (Grèce). Thèse de troisième cycle, Université Paris VI, p. 337.
Arfib, B., Cavalera, T. and Gilli, E., (2006) Influence de l'hydrodynamique sur l'intrusion saline en aquifère karstique
côtier. Comptes Rendus Geoscience, 338,757-767.
6
Arnoux, A. & Stora, G. (2003) Analyses granulométriques et chimiques des sédiments prélevés dans la zone de rejet
des boues résiduaires de l'industrie de l'aluminium (Campagne ALPECAST 2 sept-oct 2002), Comité de suivi
scientifique résidus de Bauxite Aluminium Péchiney.
Audra P., Mocochain L., Camus H., Gilli, E. and Clauzon, G. (2004) The effect of the Messinian Deep Stage on karst
development around the Mediterranean Sea. Examples from Southern France. Geodinamica Acta, 17(6), 27-38.
Berné, S., Et Al., (2002) Cartes morpho-bathymétriques du Golfe du Lion au 1/100000, IFREMER, Région
Languedoc-Roussillon.
Bini, A., (1994) Rapports entre la karstification périméditerraneenne et la crise de salinité messinienne, l’exemple du
karst lombard (Italie). Karstologia 23, 33–53.
Breznik, M., (1998) Storage Reservoirs and Deep Wells in Karst Regions. Balkema, Rotterdam, Brookfield, p. 251.
Calvino, F. & Stefanon, A., (1969) The submarine springs of fresh water and the problem of their capture, Rapp.
Comm. Int. Mer Médit.
Cavalera, T., (2007) Etude du fonctionnement et du bassin d'alimentation de la source sous-marine de Port Miou
(Cassis, Bouches-du-Rhône). Approche multicritère. Mémoire de thèse, Marseille, 397 pp.
Collina-Girard, J. (1996) Préhistoire et karst littoral : la grotte Cosquer et les calanques marseillaises, Karstologia ,
27,27-40.
Corroy, G., Gouvernet, C., Chouteau, J., Sivirine, A., Gilet, R. and Picard, J. (1958) Les résurgences sous-marines de la
région de Cassis - La fontaine de Vaucluse - Résultats scientifiques des explorations de 1955 et 1956, Bull. Inst. Océan.
1131 1-35
Cortes, J.M., Antoranz, A., Menvielle, S., Ratsimandresy, A., Cisneros, J., Ramos, S., Cabrera, M.C., Serrano, V. and
Mateu J. (2000) La intrusion marina en el sistema de cuevas Moraig-Toix. Un estudio para la recuperacion del acuifero
de la depresion de Benissa (Marina Alta – Alicante). Ve Congresso Geologico de Espana, Alicante, p. 26.
Douchet, M. (1993) Port Miou : Plongée dans l'eau delà, Subaqua, 127, 1-7.
Drogue, C., (1989) Continuous inflow of seawater and outflow of brackish water in the substratum of the karstic island
of Cephalonia, Greece. Journal of Hydrology 106, 147–153.
Drogue, C., (1993) Absorption massive d'eau de mer par des aquifères karstiques cotiers. Hydrogeological Processes in
Karst Terranes (Proceedings of the Antalya Symposium and Field Seminar, October 1990) IAHS 207
Fleury, P. (2005) Sources sous-marines et aquifères karstiques côtiers méditerranéens. Fonctionnement et
caractérisation. Thèse de troisième cycle, Université Paris VI, 286p.
Gautier, F., Clauzon, G., Suc, J.-P., Cravatte, J. and Violanti, O. (1994) Âge et durée de la crise de salinité messinienne
(Age and duration of the Messinian salinity crisis), Compte-rendus à l’Académie des Sciences II, 318 1103-1109
Gilli, E. (2001) Compilation d'anciennes mesures de débit à Port Miou. Apport à l'hydrogéologie de la Basse Provence,
7ème colloque d'hydrologie en pays calcaire et en milieu fissuré. Sciences et techniques de l'environnement, Besançon,
pp. 157-160.
Gilli, E. (2003) Les karsts littoraux des Alpes Maritimes : inventaire des émergences sous-marines et captage
expérimental de Cabbé. Karstologia, 40, 1-12
Gilli, E., Mangan, C. And Mudry, J. (2004) Hydrogéologie. Objets, méthodes, applications. Dunod edit., Paris, 303 p.
Julian, M. and Nicod, J. (1984) Paléokarsts et paléo-géomorphologies néogènes des Alpes occidentales et régions
adjacente (Neogene paleokarsts and paleogeomorphology in the Western Alps, Jura and Provence), Karstologia, 4, 11-
18.
Kohout, F.A., Henry, H.R. and Banks, J.E. (1977) Hydrogeology related to geothermal conditions of the Floridian
Plateau. in The Geothermal Nature of the Floridan Plateau, Smith D.L., Griffin G.M. (Eds.), vol. 21. Florida Bureau
of Geology Special Publication, 1–41.
Krijgsman, W., Hilgen, F. J., Raffi, I., Sierro, F. J., Wilson, D. S. (1999) Chronology, causes and progression of the
Messinian salinity crisis, Nature, 400, 652-655.
Maurin, V. And Zötl, J. (1965) Salt Water Encroachment in the Low Altitude Karst Water Horizons of the Island of
Kephallinia (Ionian Islands), Hydrologie des roches fissure´es, vol. 2. AIHS, Dubrovnik, pp. 423–438.
Potié, L., (1974) Captage des résurgences sous-marines d’eau douce. Observations sur les effets du barrage
expérimental de Port-Miou. Internal report 74 RT 1001 PM, Syndicat de recherche de Port-Miou, Société des eaux de
Marseille.
Ryan, W. B. F., Hsu K.J. et al. (1973) Initial Reports of the Deep Sea Drilling Project, 13, 1-2, U.S. Government
Printing Office, Washington D. C. (1973)1203-1231