Central Turkey represents the only orogenic plateau in the Mediterranean region. Also, the largest closed drainage basin and the largest intracontinental basin of Turkey, the Lake Tuz Basin, is located in this region. Results from a three-dimensional (3-D) computer modeling study of the Lake Tuz Basin indicate a southward-deepening freshwater lake basin with great depth in the Mio–Pliocene, which regressed toward the north during the Plio–Quaternary into the shallow saline lake basin it is today. The spatio-temporal variations of Neogene and Quaternary deposits reflect the main effects of internal forces (isostasy>volcanism>faulting) that were caused by lithospheric slab breakoff and subsequent asthenospheric upwelling under central Turkey. Climatic change played a relatively minor role during these periods and was closely associated with the results of internal forces.


Orogenic plateaus represent high-elevation regions with low-relief topography bounded by mountain ranges. The most prominent examples of such orogenic plateaus are the Tibetan and Puna-Altiplano Plateaus. Inner areas of plateaus are generally associated with closed (internally drained, endorheic, or terminal) basins. Although there is a consensus on the formation process of closed basins, several discussions have been carried out in the literature on the origin of basin closure processes in these areas. Whereas some of the works indicate a major role of external forces (such as climate and/or surface processes; e.g., Heller et al., 2011) in initiating basin closure, some others suggest that the role of internal forces (such as isostasy and/or tectonics; e.g., Isacks, 1988) is much stronger. In other settings, there is an apparent coupling between internal and external forces (e.g., Masek et al., 1994; Garcia-Castellanos, 2006; Strecker et al., 2009). In every case, closed basins form when the rates of sediment supply derived from the surrounding mountains is less than the rate of accommodation by tectonics in the basin (e.g., Carroll and Bohacs, 1999; Garcia-Castellanos, 2007; Heller et al., 2011).

Turkey includes the only orogenic plateau in the Mediterranean region, with the largest closed drainage and intracontinental basin within its central domain. The 15,000 km2 Lake Tuz Basin is an endorheic topographic depression situated at an altitude of ∼1000 m above sea level (asl) in the middle part of central Turkey (Fig. 1A). The basin’s closed drainage consists of a few perennial and many more ephemeral streams that flow into the largest salt lake of the Mediterranean region (Lake Tuz = ∼1500 km2; “tuz” means “salt” in Turkish) and other small playas, or that disappear beneath sheet deposits. These arid landforms are consistent with the modern climatic characteristics of the basin (Fig. 1B). The Lake Tuz Basin represents the most arid part of Turkey today, with a precipitation rate of 350 mm/yr (Fig. 1B). However, according to sedimentological, paleontological, and geochemical studies, the Central Anatolia region was a relatively humid region during the early Neogene, and it gradually turned into an arid region (e.g., Erol, 1969; Akgün et al., 2007; Akkiraz et al., 2011; Lüdecke et al., 2013). In this scene, some authors have suggested that the most important factor controlling this climatological evolution was the global climatic changes during the Neogene–Quaternary periods (e.g., Erol, 1969).

Despite its impressive features, the Neogene–Quaternary evolution of this area has not been well investigated within the scope of basin analysis except a few studies (i.e. Erol, 1969; Fernandez-Blanco et al., 2013), which must consider both the climatic and tectonic features of central Turkey. This study uses an integrated approach to understanding the processes behind formation of the closed basin system. Based on surface and subsurface data, we created a three-dimensional (3-D) model of the Lake Tuz Basin that helps us to reconstruct the basin geometry through time and space. The results indicate a southward-deepening freshwater lake basin with great depth in the early Neogene, regressing toward the north into the shallow saline lake basin that it is today. The spatio-temporal variations of Neogene and Quaternary deposits reflect the main effects of internal forces (isostasy>volcanism>faulting) on the basin closure process, with a relatively minor influence of external forces (climatic change).


Central Turkey is composed of accreted continental blocks and associated suture zones related to closing of the strands of the Neotethys Ocean during the Cretaceous–Eocene period (e.g., Şengör and Yılmaz, 1981; Görür et al., 1998; Gürer and Aldanmaz, 2002; Okay, 2008; Dilek and Sandvol, 2009). Convergence between the Laurasian- and Gondwanan-derived plates during this period resulted in the accretion of the Pontide and Tauride orogenic belts in northern and southern Turkey, respectively (e.g., Şengör and Yılmaz, 1981; Fig. 1A). The Lake Tuz Basin was originally a part of the forearc basins that developed during that time span (Görür et al., 1998). By the last stage of closure of the Neotethys Ocean between the Arabian and Eurasian plates in the east, regional surface uplift of Turkey has started (e.g., Şengör and Kidd, 1979). This last phase of deformation is represented by the broadly coeval onset of extension in central Turkey (e.g., Whitney and Dilek, 1997; Dilek et al., 1999; Cosentino et al., 2012; Schildgen et al., 2012, 2014).

Central Turkey is surrounded by the Taurides and Pontides, and it is made up of an ∼350-km-wide, 1–1.5-km-high region (Fig. 1) dominated by normal and strike-slip faulting, sedimentary basins, and localized volcanism with wide-ranging compositions (Cosentino et al., 2012, and references therein). The Lake Tuz Basin alone includes all these features.

The basin is characterized by NW-SE–trending structural features (Fig. 2A). NW-SE–trending Lake Tuz coincides with the lowest elevations in the basin (Fig. 1B). The most prominent of these structures based on morphology is the Tuzgölü fault. It bounds the basin to the east today with a 200 km length and works as a normal fault with a minor right-lateral component (e.g., Gürbüz, 2012; Fernandez-Blanco et al., 2013; Özsayın et al., 2013; Kürçer and Gökten, 2014; Yıldırım, 2014). Another NW-SE–directed structure, the Tersakan-Sultanhanı fault, is not well expressed morphologically, but it has a well-described structure within geophysical data (e.g., Aydemir and Ateş, 2006). While there is little tectonic activity in the area compared to the seismically highly active framework of Turkey, there is evidence of seismicity and young normal faulting along these structures (e.g., Özsayın and Dirik, 2011; Kürçer and Gökten, 2014). The other small playas in the basin that were formed by the retreat of Lake Tuz during the Quaternary are separated from the lake by some small sills that correspond to these faults (e.g., Gürbüz and Kazancı, 2014).

The Lake Tuz Basin consists of sedimentary units with an age range from the Late Cretaceous to Quaternary over a Paleozoic–Mesozoic crystalline basement. This sedimentary succession generally represents a deep- to shallow-marine environment that was interrupted by some unconformities due to collisional and postcollisional tectonics related to the aforementioned closing processes of the Neotethys Ocean up until the Oligocene, by which time the basin was characterized by a terrestrial environment. During the Oligo–Miocene, possible marine connections between the fluvio-lacustrine environment and Neotethys Ocean in the south have been reported by several authors (e.g., Erol, 1984; Huvaz, 2009). In this case, the Lake Tuz Basin was not completely endorheic until at least the late Miocene.

The Oligo–Miocene succession rests with an angular discordance on the highly deformed pre–Oligo–Miocene rocks. They are in turn unconformably overlain by the Mio–Pliocene fluvio-lacustrine deposits. The study area was covered by a widespread freshwater lacustrine environment during the Mio–Pliocene, represented mostly by carbonate rocks. The Plio–Quaternary sedimentary deposits consist of fluvio-lacustrine deposits including alluvial fan, fan-delta, freshwater lacustrine, and saline lacustrine sediments together with pyroclastic material above the Mio–Pliocene deposits (e.g., Gürbüz and Kazancı, 2014; Fig. 3).

The Cappadocian volcanic province is situated to the southeast of the Lake Tuz Basin, with an age range of middle Miocene–Quaternary (Pasquaré et al., 1988; Ercan et al., 1990; Fig. 2B). It contains stratovolcanoes (i.e., Mount Hasan, Mount Melendiz), cinder cones, calderas, extensive lava flows, and volcaniclastic units. Whereas the first products of volcanic activity indicate a calc-alkaline composition, the later products are characterized by bimodal and alkaline compositions (e.g., Ercan, 1986; Dilek et al. 1999).


In our modeling study, we used Rockworks (Rockware earth science and geographic information system [GIS] software: www.rockware.com) software to create a 3-D stratigraphic model of the Lake Tuz Basin. Data from over 250 boreholes were compiled from sources mainly including State Water Works of Turkey, Turkish Petroleum Corporation (e.g., Turgut, 1978), and wells drilled by independent researchers (e.g., Kashima, 2002; Gürbüz, 2012). After an elimination of some boreholes due to their proximity to one another, a digital database of lithologic information from 119 boreholes was compiled by manually entering data from the aforementioned sources (Fig. 2A). Borehole depths vary from 10 m to 4609 m. While 66 of them reached a depth of over 100 m, 17 boreholes are >1000 m in total depth. The deepest boreholes were drilled for oil and gas exploration, but we used only the first 800 m of these boreholes to obtain a conservative lower limit for our model. The basement surface is defined as the Oligo–Miocene and older rocks in the Lake Tuz Basin, because the closing process occurred after that period. In addition, six measured stratigraphic sections along the eastern flank of the basin were added to the database as surface data. Those data represent the Oligo–Miocene and older units that were exhumed by the tectonics of the Tuzgölü fault zone.


According to our modeling results, the Mio–Pliocene fluvio-lacustrine deposits represent a S-SE–thickening package (Fig. 4). The thickness of this unit (up to 500 m) represents high sediment deposition rates. This scene indicates the presence of a large depression to the south of the basin during the Mio–Pliocene period, in agreement with Toprak and Göncüoğlu’s (1993) suggestion. In the last stages of this environment, the basin filled up rapidly. The Plio–Quaternary fluvio-lacustrine units were deposited along a significant new depression and are characterized by relatively thin thicknesses (up to 190 m). This shift in the locus of deposition in the Plio–Quaternary implies that normal motion along the Tersakan-Sultanhanı fault and especially the Tuzgölü fault occurred in the Plio–Quaternary, in support of several previous interpretations (e.g., Koçyiğit, 2003; Dirik and Erol, 2003; Özsayın et al., 2013; Kürçer and Gökten, 2014; Yıldırım, 2014), and not in the Late Cretaceous or Miocene as claimed by others (e.g., Dellaloğlu and Aksu, 1984; Çemen et al., 1999). The NW-SE–trending depression is aligned diagonal to the overall trend of the Lake Tuz Basin with a gradient toward the southeast. The spatial distribution of sediment thickness is consistent with the published gravity data of the region (e.g., Aydemir and Ateş, 2005, 2006).


Impact of Internal Forces

Subdued relief, aridity, and closed drainage characterize the interiors of orogenic plateaus. Development of closed basins in these contractional provinces has been related to successive marginal uplifts due to reverse faulting (e.g., Sobel et al., 2003). Orographic barriers that develop due to these marginal uplifts are oriented orthogonal to moisture-bearing winds and cause regional aridity (e.g., Sobel et al., 2003). In contrast with this general view, the closure process of the Lake Tuz Basin developed within an extensional domain on an ancient forearc and foreland basin system bounded by orogenic belts that stored thick accumulations of sediment for over 50 m.y.

In central Turkey, while the Oligo–Miocene deposits represent folded and terminally reverse faulted units due to a postcollisional compressional regime, the Mio–Pliocene deposits are characterized by subhorizontal stratification. Interestingly, the depositional environments of the Miocene deposits are very different to the north and south of the Taurides. To the north, a terrestrial environment is represented by fluvio-lacustrine deposits now at an elevation of ∼1 km asl (e.g., Erol, 1969; Ulu et al., 1994; Jaffey and Robertson, 2005); however a marine environment is represented by early to late Miocene neritic limestones now at ∼2 km elevation to the south, on top of the central Taurides (e.g., Şafak et al., 2005; Cosentino et al., 2012; Fig. 2B). Although an extensional tectonic regime has controlled the middle part of central Turkey (e.g., Özsayın et al., 2013), the southern margin of the orogenic plateau represents a high amount of surface uplift since the latest Miocene (Dilek et al., 1999; Cosentino et al., 2012; Schildgen et al., 2012). This has been explained by upwelling asthenospheric mantle (e.g., Genç and Yürür, 2010) and isostatic uplift following the lithospheric slab breakoff under the Taurides (e.g., Schildgen et al., 2012; Cosentino et al., 2012). The extensional faulting in the plateau interior may have also been triggered by this geodynamic setting (Schildgen et al., 2014).

Under this condition, sensu lato, isolation of the inner part of Turkey from the marine environment to the south was likely caused due to latest Miocene surface uplift of the southern margin (step I in Fig. 5A). After this phase, especially during the Plio–Quaternary, widespread basaltic volcanism with large amounts of pyroclastics separated the basin into two parts, and the lacustrine environment continued its regression toward the north as a result of tilting processes (step II in Fig. 5B). As another factor, movement along the Tuzgölü and Tersakan-Sultanhanı faults in the region defined the boundaries of today’s Lake Tuz Basin (step III in Fig. 5B).

What about External Factors?

According to palynological and isotopic data, central Turkey experienced a humid subtropic climate during the Oligo–Miocene period (Akgün et al., 2007; Lüdecke et al., 2013). The mean annual paleoprecipitation values exceeded 1300–1200 mm/yr during the early Miocene and decreased to below 1000 mm/yr in the late Miocene, which represents the lowest values of the country today, with amounts of 300–600 mm/yr (Akkiraz et al., 2011; Fig. 1B). Lüdecke et al. (2013) attributed the earlier, wetter conditions to the absence of significant orographic barriers to precipitation along the northern and southern plateau margins during the early to middle Miocene. This interpretation is in concert with the aforementioned view of major uplift of the central Taurides in the south, which started in the latest Miocene. In the Plio–Pleistocene, the climate was characterized by a significant global cooling in many mountain ranges (Barry, 1992). For the uplifted southern margin of central Turkey, evidence of such cooling is represented by extensive glacial physiographic features (Dilek et al., 1999; Zreda et al., 2011). In the inner part of the plateau, the results of such climatic changes can be traced locally through lake-level fluctuations with relatively long wavelengths (e.g., Erol, 1969; Gürbüz and Kazancı, 2014).


Our results indicate that the development of the largest closed basin on the only orogenic plateau in the Mediterranean region is related primarily to internal forces. Our 3-D modeling of the basin geometry reveals the existence of a large topographic depression to the south of the Lake Tuz Basin, where mostly fluvio-lacustrine sediments were deposited in the Mio–Pliocene. This deposition was related to subsidence of the southern part of the basin in front of the uplifted Taurides during this period. Isopach maps clearly support the development of the basin-bounding Tuzgölü and Tersakan-Sultanhanı faults in the Plio–Quaternary, as demonstrated by recent field-based studies (e.g., Özsayın et al., 2013; Kürçer and Gökten, 2014), rather than previously suggested ages of Late Cretaceous or Miocene. Regarding the basin closure process, uplift of the southern margin of the Central Anatolian plateau in the latest Miocene was responsible for the first stage, potentially cutting off a south-directed drainage system. In the second stage, widespread Plio–Quaternary volcanism was liable for the division of the very large paleolake basin into two parts: the Konya Basin to the south and the modern Lake Tuz Basin to the north. Over the same period, volcanism forced the depocenter of the basin to migrate northward. During the third stage, the development of the basin-bounding faults of the Lake Tuz Basin defined today’s physiographic framework. All these processes may be related to the uplift of the region, especially in the southern part due to lithospheric slab breakoff and subsequent asthenospheric upwelling. As an external agent, climate played a relatively minor role during the basin closure process, and it is closely associated with the results of internal forces (i.e., the development of orographic barrier to rainfall). As stated by Sobel et al. (2003), the increased aridity aids basin closure by decreasing the drainage system’s ability to incise through growing topography at the basin outlet.

Erratum to this article.

Genetic framework of Neogene–Quaternary basin closure process in central Turkey

Alper Gürbüz and Nizamettin Kazancı

(this issue, v. 7, no. 4, p. 421–426; doi: 10.1130/L408.1)

“Mediterranean region” should read “eastern Mediterranean region” in the following location: Page 421, column 2, paragraph 1, line 6.

We are grateful to İ.Ö. Yılmaz, G. Seyitoğlu, Ö.F. Gürer, B. Varol, Y.K. Kadıoğlu, A. Ateş, F. Şaroğlu, and L. Royden for their contributions and fruitful discussions. The constructive reviews of E. Kirby, Y. Dilek, and T. Schildgen improved the manuscript significantly. This study was supported by the AÜBAP (Ankara University Scientific Research Projects, Grant no: 09B4343017).