The Messinian salinity legacy: 50 years later
- 228 Downloads
The so-called Messinian salinity crisis, which dramatically affected the Mediterranean Sea at the end of Miocene, and resulted in the accumulation of important quantities of both layered and massive salt deposits within the basin, also generated a thick pile of research papers, as well as endless debates within the scientific community. We believe that the body of evidence available from geological, hydrological, and geochemical data collected during the last 50 years provides a basis to identify a relatively coherent scenario to explain this extraordinary event, which led to the Mediterranean Sea being partly desiccated and which has strongly impacted its post Miocene evolution. We also stress how the late, or the limited understanding of several fundamental parameters, such as evidence of deep outflows of salty waters, or of realistic paleogeographic reconstructions of the Messinian basins or of recognition of their frequent syn-tectonic origin, has contributed to delay a general agreement by the scientific community of a coherent global model.
KeywordsMediterranean Sea Messinian crisis Evaporites
Near the end of the 1960s, the world of academic geosciences became aware of numerous salt structures (already known by oil industry companies) detected in the deep Mediterranean basins thanks to seismic refection profiling. Initially attributed to Triassic formations (Glangeaud et al. 1966), a Late Miocene age (Messinian, Pontian) was soon proposed (Cornet 1968; Montadert et al. 1970; Auzende et al. 1971; Ryan et al. 1973) and confirmed during the first drilling operations within the Mediterranean Sea (DSDP Leg 13 in 1971).
Recognizing the Messinian record
The Messinian geological stage was definitively validated; its base and its summit (GSSP) were, respectively, characterized in the sections of the Oued Akrech in Morocco (Hilgen et al. 2000a, b) and of Eraclea Minoa in Sicily (Van Couvering et al. 2000).
- (2)It has been possible to establish a quite precise chronology of the events (Fig. 2), thanks to robust coupling between astrochronology, paleomagnetism, micro and nano-paleontology, and some radiochronological dating (Hilgen et al. 2000a, b, 2007; Krijgsman et al. 1999, 2004; Krijgsman 2001; Hilgen et al. 2007; Manzi et al. 2013). The Messinian stage initiated at 7.246 Ma and ended at 5.33 Ma; the salinity crisis started itself around 5.96–5.97 Ma; a clear discontinuity, occurring around 5.6 Ma, separates two different Messinian sections (Decima and Wezel 1971, 1973; Roveri et al. 2008): the Lower and Upper evaporites.
- (3)A large set of seismic profiles has made possible the imaging of various Messinian sequences (Fig. 3) deposited within the different Mediterranean basins; these data facilitated an evaluation of the volumes of Messinian formations, demonstrated the existence of wide areas of intra-Messinian erosion, and showed that a strong unconformity is observed between an Upper and a Lower units almost everywhere within the deep Mediterranean basins as illustrated by Lofi et al. (2005, 2011a, b, 2019).
Competing models for the Messinian crisis
The largest part of the literature was attempting to explain the genesis and the evolution of the Messinian events. It has been thus well established that:
The water balance of the Mediterranean (evaporation vs. inflows by rainfall and Mediterranean rivers) was (and is still) quite deficient; this disequilibrium is today compensated by a strong inflow (1.5 m/s) of less salty Atlantic waters through the Strait of Gibraltar, a phenomenon known since ancient times.
The total evaporation of the actual Mediterranean would only result in an evaporite layer about 30 m thick, far less than the observed masses of salt, which may locally exceed 1 km in thickness.
A first model assumed that the evaporitic sections were deposited in shallow basins that subsequently collapsed (Nesteroff 1973). This hypothesis was soon rejected as being incompatible with both the subsidence history of the Algeria–Provence Basin on one hand, and on the other hand with the history of the Eastern Mediterranean, which was known to have been a deep basin a long time before the Messinian stage.
- Subsequent models can be grouped in two main types (Fig. 4):
Shallow evaporite basins deposited within deep depressions. This “Sebkha”-type model, involving shallow water salty sedimentation on the floor of deep depressions, implies either a recurrent feeding mechanism of these depressions, including alternation of dewatering and filling events, or, a system of basins located at different altitudes but connected by waterfalls. One may effectively observe, within the Upper evaporites of Sicily, repetitive cycles of marls and gypsum (already reported by Ogniben 1954 and 1957) and interpretable in terms of alternating flooding and dewatering. However, the Lower unit does not generally display such characteristics except when syn-sedimentary tectonism occurred as in the Apennines (Mostardini et al. 1966; Ricci Lucchi 1973) or in Sicily (Ogniben 1954, 1957). Moreover, chronological correlations, established at the scale of the entire Mediterranean domain, do not support such a complex system of connected basins (Lofi et al. 2011a, b, 2019).
Already deep and hypersaline (brine-type), basins.
Such a deep, hypersaline, basin model, was inspired by a concept already proposed by Schmalz (1969), but for the Devonian evaporites of Alberta in Canada. Using the Schmalz’s hypothesis of a deep hypersaline basin was applied to the Mediterranean Messinian event by Cita (1973) and Hsü et al. (1973a, b, b), who faced a strong skepticism, and even categorical rejection, from many scientists, chiefly based on the absence of any modern analogue (uniformitarianism) as emphasized by Selli (1973).
Recent modeling and detailed analysis (Ryan 2008; Christeleit et al. 2015; Simon and Meijer 2017; Meilijson et al. 2017) make it seems quite possible that a large volume of over-salty water existed within one or several deep basins of the Mediterranean Sea and that precipitation of salts could have occurred, supporting thus the Schmalz’s model also considered by Roveri et al. (2014a).
Significance of the Gibraltar strait
In fact, the problem of water exchanges between the two oceanic spaces (Mediterranean Sea and World ocean) was not correctly addressed, since it was still not yet clearly understood in the 1990s that the present water balance within the Mediterranean is strongly connected to the existence of the Gibraltar undercurrent.
It is now well recognized that the interruption of this deep current, prohibiting the export of Mediterranean waters (and of its salt content), has led during the Messinian stage to over-salinity in the Mediterranean basins. Denser salty water accumulates at the bottom of the basins as brines until salt precipitation are reached; such a phenomenon is likely to have played a role in the initiation of the Messinian crisis: interruption of the deep salty Mediterranean waters outflow current, not necessarily involving interruption of Atlantic waters surface current. Despite having been suspected by Selli (1973), the fundamental importance of such a mechanism has only relatively recently been perceived (Krijgsman and Meijer 2008; Meijer 2012; Meijer et al. 2004; Hilgen et al. 2007; Ryan 2008; Roveri et al. 2014a).
An immediate corollary is that the level of the Mediterranean Sea remained practically identical to that of the Atlantic; in others words at this stage, the Mediterranean was not a series of deep and desert-like depressions! The salinity of the waters gradually increased, water stratification initiated, and finally, salt precipitation started within the deepest basins. This geochemical event occurred between 5.96 and 5.5 Ma. Meijer et al. (2004) and Hilgen et al. (2007) place the onset of salt deposition at 5.6 Ma.
This was followed by a real interruption of the Atlantic waters input current. This event occurred only around 5.55 Ma. A direct consequence was a very sudden and considerable sea-level drop, on the order of 1500 m; at that time were probably formed deep desiccated depressions.
Finally, one may consider that both types of models are relevant: between 5.96 and 5.55 Ma, the Lower Unit and the massive salt are sedimented in deep basins (Schmalz’s type). Between 5.55 and 5.33 Ma, the Upper Unit is deposited in a sebkha-type environment on the bottom of some basins in Sicily, Crete, and Cyprus now at shallow levels. Note that this sequence of events is somewhat similar to that proposed by Clauzon et al. (1996), although it differs in several respects, for example, the upholding of the Atlantic sea level within the Mediterranean realm during the first stage. Such chronological evolution leads, however, to various consequences. First, the very short duration (about 200 ka) of the drying event precludes stable isostatic equilibrium being reached. As a consequence of water weight disappearing, or at least its important reduction, the Central Mediterranean uplifted, and consequently, the surrounding lands started subsiding, but without reaching a stable equilibrium. In this perspective, all previous attempts to evaluate the real depth of basins and the subsidence of their margins require revision. The elevation reached during the Early Pliocene transgression in stable areas, for example, the Languedoc in southern France (160 m), may reflect the amount of the isostatic variations on land.
A second consequence regards the climatic evolution during Messinian times. There is no real justification for a durable climatic crisis during the whole Messinian. If a strong climatic change occurred in the Mediterranean area, it was a very short one and only during about 200 ky before the Pliocene re-inundation.
Paleogeographic reconstructions and tectonic movements
It is clear that during Messinian times, the Eastern Mediterranean Sea was much wider than today and in open connections westward with the Ionian and Adriatic basins; the Mediterranean ridge, running in its middle, did not interrupt deep water circulation between the Levantine Basin and the Ionian Sea, while a seawater connection was open with the Adana Basin in Turkey (Cipollari et al. 2013; Radeff et al. 2016). In fact, by Latest Miocene, the Eastern Mediterranean Sea was about twice as wide than today (Fig. 10) (Mascle 2008). This, relatively wide, oceanic space was bounded by two active subduction zones: a northern one, linking the Adana gulf to the Adriatic Sea, and a western one extending between Tunisia and Tuscany; both subduction zones were bordered by accretionary wedges. In the present-day Caltanissetta basin, in Sicily, numerous tectonic intercalations made of chaotic allochthonous materials are observed within the Messinian sequences (Ogniben 1954) and their occurrence indicates that the area was a portion of an accretion complex that connected northward with the Bradanic Trench in Southern and Central Apennines. Therefore, the salt sequences from the Caltanissetta basin, and similarly from Calabria, may indicate a continuity of the deep Ionian Basin rather than a tectonic setting related to potential perched basins. The Messinian sequences were simply progressively incorporated within the accretionary wedge, either during Messinian times or even later on during the Pliocene. Such mechanisms are clearly occurring today along the southern borders of both the Calabria external arc and the Mediterranean Ridge, respectively, in the Central and Eastern Mediterranean Sea. Evidence for this can easily be observed on the “Geological and Morpho-Tectonic Map of the Mediterranean Domain” published by Mascle and Mascle (2012). Puzzling that such a fundamental paleogeographic constraint (Fig. 10) has remained more or less unaddressed, since it first assessment more than 10 years ago (Mascle 2008). Recently, however, it has finally been taken into account (see Krijgsman et al. 2018).
- Differing tectonic configurations in the past will clearly have impacts on mass balance estimates. It is generally accepted (Warren 2010; Ryan in Lofi et al. 2011a, b) that the Messinian event resulted in the sequestration, within the Mediterranean Sea, of almost 5% of the salt of the world ocean. This corresponds to an estimated decrease of the average salinity of the world’s ocean seawater on the order of 2 g/l. Considering a saltwater-phase diagram (Fig. 11), such a decrease in salinity implies an increase in the freezing temperature of the salt water on the order of 1 °C. This in turn implies an easier sea ice creation at high latitudes, and so major developments of ice pack in the polar areas, with drastic climatic and eustatic consequences. Accordingly, one may question the relations between the salinity crisis and climatic changes during Late Neogene and Quaternary times. Did the salinity crisis have induced global climatic alterations? Or did climatic alterations (initiating in high latitude glaciations) have caused the salinity crisis? Or finally was there a complex interaction between climatic alterations and salinity crisis? More data are needed to definitely answer these questions.
Concerning the water connection between the Mediterranean Sea and the Atlantic, the model, proposed by Benson et al. (1991), based on a South-Rif entry corridor and a Betic exit corridor, remains valid, provided that it is in a restored Messinian geometry. On the path of the Betic corridor, the Upper Miocene series exhibits contourite features implying a westward current of up to 1.5 m/s (Hernandez-Molina et al. 2014). This corridor must have become ineffective, closed or not deep enough, around 5.96 Ma. It seems difficult to link the cause of this interruption to a simple ice-driven eustatic mechanism. Indeed, during the most recent glaciations, the about 120 m sea-level lowering did not interrupted the Gibraltar undercurrent. Therefore, a tectonic origin of the closure seems more likely. The South-Rif corridor, meanwhile, ceased being active around 5.5 Ma. Its closure may have been driven by differential vertical movements of the crust, which can themselves have resulted from a regional asthenospheric uplift at the origin of a magmatic axis known as “Morocco Hot Line” as proposed by Frizon de Lamotte et al. (2008).
As for the re-flooding of the Mediterranean, occurring at 5.33 Ma, a model proposed by Blanc (2002, 2006) of a regressive erosion able to break the structural barrier at the level of the actual Strait of Gibraltar seems quite coherent, considering evidence that the region at that time was characterized by steep relief and strong erosive capacity. The Nile and Rhône canyons are good examples of very deeply incised valleys that imply extremely important regressive erosion rates occurring at that time. Similarly, a strong erosive capacity of rivers at the level of the Gibraltar strait has been supported by numerical modeling (Loget et al. 2005), as well as by seismic reflection data recorded in the western Alboran Sea (Garcia-Castellanos et al. 2009; Campillo et al. 1992; Martinez del Olmo and Comas 2008; Estrada et al. 2011).
This paper has benefited from the critical reviews and improvements from Dr. D. Cosentino and an anonymous reviewer. We thank Dr. D. Praeg for both scientific and English wording improvements.
- Arambourg C (1925) Révision des poissons fossiles de Licata (Sicile). Ann Paléontol 14:37–96Google Scholar
- Arambourg C (1927) Les poissons fossiles d’Oran. Mat Carte géol Algérie, 1ère série, Paléontologie, 6:1–298, atlasGoogle Scholar
- Auzende JM, Bonnin J, Olivet JL, Pautot G (1971) Upper Miocene salt layer in the western Mediterranean. Nature 230:82–84Google Scholar
- Campillo A, Maldonado A, Mauffret A (1992) Stratigraphic and tectonic evolution of the western Alboran sea: late Miocene to recent. Geo–Mar Lett 12:165–172Google Scholar
- Chumakov IS (1973) Pliocene and pleistocene deposits of the Nile valley in Nubia and upper Egypt. In: Ryan WBF, Hsu KJ et al (eds) Initial reports of the deep sea drilling project, 13. US Gov. Printing Office, WashingtonGoogle Scholar
- CIESM (2008) The Messinian Salinity Crisis from me–ge–deposits to microbiology: a consensus report. In: Briand F (eds) CIESM workshop monographs, vol 33, Monaco, p 168Google Scholar
- Cita MB (1973) Mediterranean evaporite: paleotological arguments for a deep–basin desiccation model. In: Drooger CW (ed) Messinian events in the Mediterranean. North Holland Pub., Amsterdam, pp 226–228Google Scholar
- Cornet C (1968) Le graben médian (zone A) de la Méditerranée occidentale pourrait être pontien. C R Somm Soc Géol France 5:149Google Scholar
- Decima A, Wezel FC (1971) Osservazioni sulle evaporiti messiniane della Sicilia centro– meridionale. Riv Min Sicil 130–132:172–187Google Scholar
- Decima D, Wezel FC (1973) Late Miocene evaporites of the central Sicilian Basin, Italy. Initial Rep Deep Sea Drill Proj 13:1234–1241Google Scholar
- Estrada F, Ercilla G, Gorini C, Alonso B, Vasquez JT, Garcia-Castellanos D, Juan C, Maldonado A, Ammar A, Elabbassi M (2011) Impact f pulsed Atlantic water inflow into the Alboran basin at the time of Zanclean flooding. GeoMarine Lett 31:361–376Google Scholar
- Frizon de Lamotte D, Zizi M, Missenard Y, Hafid M, El-Azzouzi M, Maury RC, Charrière A, Taki Z, Benammi M, Michard A (2008) The Atlas system. In: Michard A, Saddiqui M, Chalouan A, Frizon de Lamotte D (eds) Continental evolution: the geology of Morrocco. Springer, Berlin, pp 133–202CrossRefGoogle Scholar
- Glangeaud L, Alinat J, Polvèche J, Guillaume A, Leenhardt O (1966) Grandes structures de la mer Ligure et leurs relations avec les chaînes continentales. Bull Soc Géol France 7:921–937Google Scholar
- Hernandez-Molina FJ, Stow DAK, Alvarez-Zarikian CA, Acton G, Bahr A, Balestra B, Ducassou E, Flood R, Flores JA, Furota S, Grunert P, Hodell D, Jimenez-Espero F, Kim JK, Krissek L, Kuroda J, Li BH, Llave E, Lofi J, Lourens L, Miller M, Nanayama F, Nishida N, Richter C, Roque C, Pereira H, Sanchez-Goni MF, Sierro FJ, Singh AD, Sloss C, Takashimizu Y, Tzanova A, Voelker A, Williams T, Xuan C (2014) Onset of Mediterranean outflow into the North Atlantic. Science 344:1244–1250CrossRefGoogle Scholar
- Hilgen FJ, Iaccarino S, Krijgsman W, Villa G, Langereis CG, Zachariasse WJ (2000b) The global boundary stratotype section and point (GSSP) of the Messinian stage (uppermost Miocene). Episodes 23:172–177Google Scholar
- Hilgen FJ, Kuiper KF, Krijgsman W, Snel E, van der Laan E (2007) High-resolution integrated stratigraphy and astronomical tuning as prerequisites for deciphering the intricate history of the Messinian Salinity Crisis. Stratigraphy 4:231–238Google Scholar
- Hsü KJ, Cita MB, Ryan WBF (1973a) The origin of the Mediterranean evaporites. In: Ryan WBF, Hsü KJ, Cita MB (eds) Initial reports of the deep sea drilling project 13, Part 2. U.S. Gov. Print. Office, Washington, pp 1203–1231Google Scholar
- Kastens K, Mascle J, Auroux C, Bonnati E, Broglia C, Channell J, Curzi P, Emeis KC, Glacon G, Hasegawa S, Hieke W, Mascle G, Mc Coy F, Mc Kenzie J, Mmendelson J, Muller C, Rehault JP, Robertson A, Sartori R, Sprovieri R, Torii M (1988) ODP leg 107 in the tyrrhenian sea: insights into passive margin and back-arc basin evolution. Geol Soc America Bull 100:1140–1156CrossRefGoogle Scholar
- Knauss JA (1978) Introduction to physical oceanography. Prentice-Hall, NJ, p 338Google Scholar
- Krijgsman W, Gaboardi S, Hilgen FJ, Iaccarino S, de Kaenel E, van der Laan E (2004) Revised astrochronology for the Ain el Beida section (Atlantic Morocco): no glacio–eustatic control for the onset of the Messinian Salinity Crisis. Stratigraphy 1:87–101Google Scholar
- Levitus S, Burgett R, Boyer TP (1994) World Ocean Atlas 1994, Vol 3: salinity. In: NOAA Atlas NESDIS 3. U.S. Department of Commerce, Washington, DCGoogle Scholar
- Lofi J, Deverchère J, Gaullier V, Gillet H, Gorini C, Guennoc P, Loncke L, Maillard A, Sage F, Thinon I (2011) Seismic atlas of the “Messinian Salinity Crisis” markers in the mediterranean and black seas, vol I. Commission for the Geological Map of the World, Paris, France. Mém. Soc. Géol. France, n.s., 179, 1 CDGoogle Scholar
- Lofi J, Camerlenghi A, Del Ben A, Do Couto D, Estrada F, Gallais F, Garcia M, Gaullier V, Maillard A, Micallef A, Rossi M, Sage F, Schattner U, Tass A, Urgeles R (2019) Seismic atlas of the “Messinian Salinity Crisis” markers in the mediterranean and black seas, vol II. Commission for the Geological Map of the World, Paris, France. Mém. Soc. Géol. France, n.s., 179, 1 CDGoogle Scholar
- Lugli S, Manzi V, Roveri M, Schreiber CB (2015) The deep record of the Messinian salinity crisis: evidence of a non dessicated Mediterraneab Sea. Pal Pal Pal 433:201–218Google Scholar
- Madelain F (1970) Influence de la topographie du fond sur l’écoulement méditerranéen entre le Détroit de Gibraltar et le Cap Saint-Vincent. Cah Oceanogr 22:43–61Google Scholar
- Martinez del Olmo W, Comas MC (2008) Arquitectura sismica, olistostromas y fallas extensionales en el Norte de la cuenca oeste del Mar de Alboran. Rev Soc Geol Espana 21:151–167Google Scholar
- Mascle G (2008) Les roches: mémoire du temps. Grenoble Sciences Ed, p 288Google Scholar
- Mascle J, Mascle G (2012) Geological and morpho-tectonic map of the mediterranean domain. Publication CCGM/CGMW, ParisGoogle Scholar
- Montadert L, Sancho J, Fial JP, Debyser J (1970) De l’âge tertiaire de la série salifère responsable des structures diapiriques en Méditerranée Occidentale (Nord–est des Baléares). CRAS 271:812–815Google Scholar
- Mostardini F, Pieri M, Pirini C (1966) Stratigrafia del foglio 212, Montalbano Jonico. Boll Serv Geol Italia 87:541–547Google Scholar
- Nesteroff W D (1973) The sedimentary history of the Mediterranean area during the Neogene. In: Ryan WBF, Hsu KJ et al (eds) Initial report of DSDP, vol 13, pp 1257–1261Google Scholar
- Ogniben L (1954) Le « Argille brecciate » siciliane. Con i rilievi di dettaglio di Grottacalda (Valguarnera, Enna), Pasarello (Licata, Agrigento), Zubbi (San Cataldo, Caltanissetta). Mem Ist Geol Min Univ Padova 18:1–92Google Scholar
- Ogniben L (1957) Petrografia della serie solfifera siciliana e considerazioni geologiche relative. Mem Descr Carta Geol Italia 33:1–275Google Scholar
- Radeff G, Cosentino D, Cipollari P, Schildgen TF, Iadanza A, Strecker M (2016) Stratigraphic architecture of the upper Messinian deposits of the Adana Basin (southern Turkey): implications for the Messinian salinity crisis and the Taurus petroleum system. Italian J Geosci 135(3):408–424CrossRefGoogle Scholar
- Regione Siciliana (1962–1964) Studi ed indagini per ricerche solfifere. Riv Min Siciliana, 76–78, 79–81, 85–87Google Scholar
- Ricci Lucchi F (1973) Resedimented evaporites: indicators of slope instability and deep–basin conditions in Periadriatic Messinian (Apennines Foredeep, Italy). In: Drooger CW (ed) Messinian events in the Mediterranean. North–Holland Pub. Cie, Amsterdam, pp 142–149Google Scholar
- Roveri M, Manzi V, Gennari R, Iaccarino SM, Lugli S (2008) Recent advancements in the Messinian stratigraphy of Italy and their Mediterranean-scale implications. Boll Soc Pal Ital 47:71–85Google Scholar
- Roveri M, Gennari R, Lugli S, Manzi V, Reghizi M (2019) The synthetic seismic expression of the Messinian Salinity Crisis: implications for shallow—to deep-water correlations. Basin Res 32:1–32Google Scholar
- Ryan WBF (2008) Modeling the magnitude and timing of evaporative drawdown during the Messinian salinity crisis. Stratigraphy 5:227–243Google Scholar
- Ryan WBF, Hsu KJ, Cita MB, Dumitrica P, Lort JM, Maync W, Nesteroff WD, Pautot G, Stradner H, Wezel FC (1973) Initial reports of the deep sea drilling project, vol 13. US Government Printing Office, WashingtonGoogle Scholar
- Schmalz RF (1969) Deep–water evaporite deposition: a genetic model. Bull Am Assoc Pet Geol 53:798–823Google Scholar
- Selli R (1960) Il Messiniano Mayer Eymar 1867, proposta di un neostratotipo. Giorn Geol 2(28):1–33Google Scholar
- Selli R (1973) An outline of the Italian Messinian. In: Drogger CW (ed) Messinian events in the Mediterranean. North–Holland Pub. Cie, Amsterdam, pp 124–141Google Scholar
- Sverdrup HU, Johnson MW, Fleming RH (1942) The oceans, their physics, chemistry, and general biology. Prentice-Hall, New York, p 1087Google Scholar
- Van Couvering JA, Castradori D, Cita MB, Hilgen FJ, Rio D (2000) The base of the Zanclean stage and of the Pliocene series. Episodes 23:179–186Google Scholar