Introduction

Palaeosoils record the climatic and environmental conditions at the time of the soil’s formation and are therefore considered proxy providers for palaeoclimatological and palaeogeographical reconstructions (Atalay 1996; Jeong et al. 2011; Meier et al. 2014; Sheldon and Tabor 2009). Studies related to the palaeosoils in Türkiye and their palaeoclimatological significance have been popular in recent years (Türkmenoğlu et al. 2010; Eren 2011; Küçükuysal 2011; Kaplan et al. 2013; Alçiçek and Alçiçek 2014; Kadir et al. 2014; Küçükuysal and Kapur 2014; Toker et al. 2015; Erginal et al. 2017; Gürel 2017; Tagliasacchi and Kayseri-Özer 2020; Kandemir et al. 2021; Bayer-Altın et al. 2021). In addition to the literature related to Anatolia, there is an important research related to the loess-palaeosoil successions on the northern shores of the Black Sea (Tsatskin et al. 2001; Dodonov et al. 2006; Gendler et al. 2006; Timar et al. 2010; Cordova et al. 2011; Tecsa et al. 2020; Hlavatskyi and Bakhmutov 2021; Hlavatskyi et al. 2021; Lanczont et al. 2022). Among these, the studies on the southern coast of the Black Sea are quite limited (Türkmenoğlu et al. 2010; Göktürk et al. 2011; Kandemir et al. 2021). Existing literature contributes to the understanding of sea-level changes and their related terrace deposits in the Black Sea in the Quaternary (Erinç 1954; İnandık 1956; Yalçınlar 1958, 1959; Öner 1990; Uzun 1995; Berndt et al. 2018) and the geomorphology, climate and natural vegetation of the Kürtün Creek Basin and its immediate vicinity (Çoban 1996; Başar 2003; Şahin 2009; Uzun 2000). This study presents multi-proxy data through a palaeosoil-sediment section of the lower Kürtün Valley on the Late Quaternary palaeoclimate and palaeovegetation, and it is expected to be the basis for future studies in the region.

Major aims of this study are (i) to determine the age of the palaeosoil-sediment section of the lower Kürtün Valley by Optically Stimulated Luminescence (OSL) and radiocarbon (14C) dating, (ii) to shed light on the changes of the palaeoclimatological and palaeovegetational conditions by combined mineralogical, geochemical and palynological analyses, and (iii) to place the results of the study in the wider context of regional and global climate change.

The Kürtün valley

Geographical setting and geomorphology

The area of interest is located on the southern coast of the Black Sea, within the borders of Türkiye’s Samsun province and the lower Kürtün Valley (41017’18” N − 36,012’07” E) (Fig. 1). This area and its surroundings are drained by the Kürtün Creek which originates from the north-facing slopes of the Black Sea Coastal Mountains and flows into the Black Sea from the western part of Samsun city. The Kürtün Creek has a water catchment area of approximately 332.0 km2, and there are occasionally floods during heavy rains.

Fig. 1
figure 1

Maps for the location of the study area together with local topographic map of the lower Kürtün Valley and the vicinity

Full size image

The Black Sea coastal mountains emerged during the Early Oligocene and have since then risen on the one hand and been eroded by erosional processes on the other (Saner 1980). Particularly during the Neogene, global climate changes and fluctuations in tectonic uplift rates led to the formation of extensive erosion surfaces in the coastal mountains (Uzun 1995). Tectonic uplift in the region persisted into the Quaternary with the average uplift rate over the past 545,000 years was calculated as 0.28 ± 0.07 m/thousand years (Berndt et al. 2018). Furthermore, global climate changes and associated sea-level fluctuations in the Quaternary led to the formation of river terraces within the valleys. The palaeosoil-sediment sequence that is the subject of this research is also located on such a fluvial terrace (Fig. 1).

Current temperature and precipitation values of the area were obtained from Samsun Meteorology Station data (1929–2021). The annual average temperature in Samsun is 14.6 oC, and the annual average precipitation is 706 mm at the present time (Table 1). Samsun experiences rainfall throughout the year, although the summer months (June, July and August) are comparatively drier due to increased evaporation. The highest precipitation occurs in November (84 mm), while the lowest precipitation is recorded in July (35 mm), (Table 1).

Table 1 Annual average temperature and precipitation values of Samsun (Türkiye) (General Directorate of Meteorology 2022)
Full size table

The study area is located in the Euxine district within the European-Siberian flora region. In this area, broad-leaved trees such as Quercus sp., Tilia sp., Acer sp., Alnus sp., Fraxinus sp., Acacia sp., Carpinus betulus and Platanus orientalis, shrubs such as Coryllus avellana, Erica arborea, Paliurus spina christi, Cornus mas, Cornus sanguinea, Phillyrea latifolia, Crataegus monogyna and Mespilus germanica, and also grass species are common. Furthermore, the area is inhabited by a range of wild animals including wild pigs, wolves, foxes, jackals, turtles, snakes, field mice and hedgehogs.

Brown forest soils are common in the lower course of the Kürtün Stream and are also present in the upper part of the palaeosoil sample. The upper layer (15 cm) exhibits a dark brown (10 YR 3/2) colour and the lower layer (30 cm) appears dark yellowish brown (10 YR 3/4). Additionally, colluvial soils are common on the foothills and alluvial soils predominate on the valley floor.

Geological setting

Türkiye comprises three main tectonic units: the Anatolide-Tauride block, Central Anatolian Crystalline Complex, and the Sakarya zone, each with distinct stratigraphic and structural features (Okay and Tüysüz 1999). The Central and Eastern Pontides are located in the northern margin of Türkiye and lie along one of the main tectonic units of the İzmir–Ankara–Erzincan suture zone (Şengör and Yılmaz 1981). The studied palaeosoil sequence is located in the Central Pontides. This sequence, located in Samsun, is positioned 10 km inside the mouth of the Kürtün Creek and on the left bank of the stream and rises 10 m above the valley floor (Fig. 1). The oldest unit in the study area is the Late Cretaceous–Late Palaeocene Akveren Formation, composed of sandstones, sandy limestones, limestones and marls (Gedik et al. 1984; Aydın et al. 1995; Görür and Tüysüz 1997; Temizel et al. 2016). This formation, which crops out in the southwest of the study area, is unconformably overlain by the Early-Middle Eocene Kusuri Formation which includes intercalation of limestone, sandstone and marl (Gedik et al. 1984; Aydın et al. 1995; Görür and Tüysüz 1997; Keskin 2011; Temizel et al. 2016). The Kusuri Formation is conformably covered by the Middle Eocene Tekkeköy Formation, which comprises basaltic-andesitic pyroclastics intercalated with mudstone, siltstone, limestone and marls (Gedik et al. 1984; Temizel et al. 2016). This formation is unconformably covered by the Late Miocene-Early Pliocene Samsun Formation (Doyuran et al. 1985). The lower part of the formation, named as the İlyas Member consists of marine marls intercalated with gypsum levels and siltstones (Doyuran et al. 1985). Based on the presence of Globigerine praebulloides, Giobigerinoides ruber, Globigerinoides sp., Globigenina sp., Globigerinita sp., Pulleniatina sp., Amphistegina sp., Spiroloculina sp., Pyrgo sp., Lenticulina sp. and Nodosaria sp., this member is dated to the Late Miocene-Early Pliocene. The Karasamsun Member in the upper part of the Samsun Formation contains sandstones, siltstones, claystones, marls and terrestrial conglomerates (Doyuran et al. 1985). This member transitions gradually to the upper part of the İlyas Member. All these formations are unconformably overlain by the Quaternary deposits and alluvium fed by the Kürtün River and coastal plains of Atakum (Temizel et al. 2016) (Fig. 2). The Karasamsun Member is exposed in thick successions along the western and eastern coastlines of the Kürtün Creek. In addition, unconsolidated alluvium (10–50 m thickness) is widespread along the stream bed and at the mouth of the Kürtün Stream (Şahin 2009).

Fig. 2
figure 2

A) Tectonic map of the northeastern Mediterranean region showing the major sutures and continental blocks (Okay 2008); B) Simplified geological map of the northwest Samsun (Keskin 2011)

Full size image

Materials and methods

Within the scope of this project, two measured stratigraphic sections (7.2 m and 5 m each) were extracted along the palaeosoil-sediment sequence. Samples exhibiting distinct colours and lithologies from each level were collected, with an approximate weight of 1.5 kg, and numbered from bottom to top as Ö-1 to Ö-15. Furthermore, three sediment samples were dated using OSL, while six calcrete-sediment samples were aged using 14C dating.

Dating analysis

Two different dating methods were used in the study area based on the lithological properties and organic matter contents of the samples. One of these methods, 14C dating, was performed using the AMS method at the Beta Analytical Laboratory (Canada), and age data were obtained from six samples (C1-C6). All results (excluding some inappropriate material types) falling within the range of available calibration data were calibrated to calendar years (cal BC/AD) and calibrated radiocarbon years (cal BP) (SD. 1). Calibration was performed using one of the databases associated with the 2013 INTCAL program (Bronk Ramsey 2009; Reimer et al. 2013). Three sediment samples (OSL1, OSL2 and OSL3) from the bottom part of the section were subjected to OSL dating at the Turkish Atomic Energy Agency Sarayköy Nuclear Research and Training Center (TAEK). For OSL dating, the Single Tablet Replication (SAR) method was applied to the quartz grains. As OSL dating determines the last time of sunlight exposure (Duller 2004; Bradley 2015), samples were immediately placed in black-coloured plastic bags after sampling. Steel pipes with a diameter of 10 cm and a length of 50 cm were used to extract the samples. OSL analysis was successful for OSL1 and OSL 2 samples; however, quartz from OSL3 sample was isolated three times, but luminescence efficiency could not be obtained.

Mineralogical and geochemical analysis

Samples were mineralogically analyzed using a Panalytical Expert Pro diffractometer equipped with a Cu tube at 40 kV voltage and 30 mA current, with a scanning rate of 2°/min in the Mineralogy-Petrography Laboratory at the General Directorate of Mineral Research and Exploration. Combined procedures of Thorez (1976), Chen (1977), Jackson (1979), Brindley (1980) and Moore and Reynolds (1989) were followed during the preparation of the samples for X-ray diffraction (XRD) analysis. The fine fraction (< 2 μm) was mounted as oriented aggregates on glass slides (Moore and Reynolds 1997). X-ray patterns were recorded as air-dried (AD), ethylene glycol solvated for 24 h (EG), and heated at 350 °C and 550 °C for 2 h. The diffractograms were plotted between 4°-70° 2ϴ for the whole rock samples and between 4°-30° 2ϴ for the fine fraction (< 2 μm). Semi-quantitative abundances of minerals in bulk compositions were determined using the method described by Brindley (1980). Mineral names are abbreviated according to Whitney and Evans (2010).

The organic matter content of the samples was analyzed using the modified Walkley Black Method, while the amount of calcium carbonate (CaCO3) was determined using the Scheibler Calcimeter (Çağlar 1949; Bouyocous 1951; Jackson 1958).

Geochemical analyses were performed using Inductively Coupled Plasma- Mass Spectrometry (ICP-MS) for whole-rock major elements. Samples were pulverized to 85% passing 200 mesh for whole-rock characterization and loss on ignition in ACME Analytical Labs. Weathering indices were determined using the Chemical Index of Alteration (CIA) equation proposed by Nesbitt and Young (1982) [CIA = 100×(Al2O3/Al2O3 + CaO* + Na2O + K2O)] and that of Maynard (1992) [CIA-K = 100×(Al2O3/Al2O3 + CaO* + Na2O)]. CaO* incorporating the silicate fraction was quantified according to the correction method of McLennan (1993). The mean annual precipitation (MAP) values in mm yr − 1 were calculated using the equation of MAP = 14.265 (CIA-K)-37.632 (Sheldon et al. 2002). Molecular weathering ratios of salinization [(K2O + Na2O)/Al2O3], calcification [(CaO + MgO)/Al2O3], clayeyness (Al2O3/SiO2), and base loss (Base/Ti) were calculated according to the given equations and compared with the reference values from Retallack (2001), Sheldon (2006), Sheldon and Tabor (2009).

Stable isotope analysis

δ18O and δ13C values of the samples were measured using an automated carbonate preparation device (KIEL-III) coupled to a gas-ratio mass spectrometer (Finnigan MAT 252) at the Environmental Isotope Laboratory at the University of Arizona. Powdered samples were reacted with dehydrated phosphoric acid under vacuum conditions at 70 °C. δ13C and δ18O values are expressed in parts per thousand (‰) relative to the VPDB standard. The isotope ratio measurement is calibrated based on repeated measurements of NBS-19 and NBS-18 with a precision of ± 0.1% for δ18O and ± 0.08% for δ13C (1sigma). In this study, stable isotope data from the measured stratigraphic section in the lower Kürtün Valley were obtained for 13 samples.

Palynological analysis

The palynological analysis was conducted on 15 palaeosoil samples at the Palynological Laboratory of the Direnç Mühendislik (Ankara-Türkiye). Nine of them yielded sufficient spores and pollen content for analysis. For the palynological investigation, all samples underwent treatments with HCl, HF and acetolysis following the standard procedures as described by Kaiser and Ashraf (1974). The resulting residue was sieved with a 10 μm mesh size. All palaeosoil samples, residues, and slides are stored at Dokuz Eylül University, İzmir, Türkiye. On each slide, a minimum of 250 pollen grains were counted under a transmitted light microscope. The pollen counts are expressed in percentage graphs using TILIA software (Grimm 2005).

Results

Geochronological data

For the geochronological study, three sediment samples from the lower part of the section and six calcrete-sediment samples along the palaeosoil-sediment section were collected. All age data are presented in Table 2. An age model for the palaeosoil-sediment section is provided in Fig. 3. The geochronological framework for this study is based on two major assumptions: (1) OSL data obtained from the quartz samples taken from the lower part of the section matched the expectations; (2) radiocarbon age data obtained from calcrete-sediment samples (C1-6) may be younger ages than the sediments in the section due to exposure to multiple carbon inputs in the terrestrial environment.

Table 2 The radiocarbon and OSL data of the samples from the palaeosoil-sediment section in the lower Kürtün Valley (Samsun)
Full size table
Fig. 3
figure 3

Age-depth model for the palaeosoil-sediment sequence in the study area with two insets of field photographs showing sampling points: A) from OSL1 (750 cm) to Ö11 (360 cm); B) from Ö12 (350 cm) to Ö13/C6 (150 cm) (age of Last Glacial Maximum from Mix et al. 2001; Younger Dryas from Broecker et al. 2010)

Full size image

The bottommost part of the sequence consists of the fluvial deposits with OSL1 sample from this unit dating back to 108,000 ± 9 ka (MIS5c). The coarse clastic deposits rise approximately 10 m from the Kürtün valley floor. No age data proving MIS3 could be obtained from the samples taken. This situation may be related to the occasional erosion of the colluvial material from the upper part due to excessive rainfall. However, a radicarbon age of 27,400 ± 120 yr BP was obtained from the C2 sample, indicating MIS2. In addition, similar age data were obtained from samples C3, C4 and C5, also attributed to MIS2 (Table 2). The radiocarbon age of the uppermost sample (C6) is 7570 ± 30 yr BP and is dated to MIS1 (Table 2; Fig. 3).

Mineralogical determinations

XRD analyses reveal the presence of quartz, feldspars, calcite, dolomite and phyllosilicates in the bulk composition of the samples (SF.1a). Quartz was determined by the intense and sharp reflections at 3.34 Å, 4.26 Å, 2.45 Å, 2.28 and 2.12 Å. K-feldspar with its prominent peak at 3.24 Å (27.5˚ 2Ɵ) and plagioclase with its major doublets at 3.18 Å and 3.15 Å are also observed on the diffractograms. Calcite was identified by its main reflection at 3.04 Å (29.4˚ 2Ɵ); while dolomite appears at 2.89 Å. In the bulk composition, the reflection at 4.44–4.48 Å refers to the presence of phyllosilicates. XRD study on the clay fraction of the samples indicates the presence of different clay minerals. The broad peak at 14 Å on the diffractogram of an air-dried oriented sample refers to two clay minerals; one shifts to 17 Å up on ethylene glycol solvation, while the other remains at 14 Å at the same angle. 17 Å peak collapses to 10 Å by heating to 350˚ and 550 ˚C, while 14 Å remains at 350 ˚C but becomes more intense at 550 ˚C (SF. 1b). The swelling peak points to the presence of smectite, while the latter displays chlorite. 7 Å reflection disappears upon heating to 550˚ C, indicating the presence of kaolinite in the clay fraction.

The results of semi-quantitative analysis on the bulk composition of the palaeosoil-sediment and calcrete samples are given in Table 3. Quartz and feldspar contents in the palaeosoil-sediment samples show very similar ranges, varying between 6 and 38% and 5–24%, respectively. Calcite shows a moderate range of abundance between 1 and 37%, while, dolomite is very low (1%). Total phyllosilicate abundance has a very broad range between 22 and 82%. In the calcrete samples, the most abundant mineral is determined as calcite with 70–90%, followed by quartz (3–26%). Feldspar and dolomite constitute less than 5% of the composition. Smectite and illite exhibit wide ranges of abundance in the clay fraction, between 10 and 87% and 6–74%, respectively. On the other hand, the relative abundance of kaolinite and chlorite falls between 6 and 22%.

Table 3 Results of semi-quantitative mineralogical analysis of the samples of the palaeosoil-sediment section in the lower Kürtün Valley (Samsun)
Full size table

Comparing the relative abundances of minerals along the palaeosoil-sediment section including those in the calcretes sheds light on the climatic interpretations in addition to the other proxies in this study. Quartz abundance between Ö2-Ö9 shows an increasing pattern and between Ö10-Ö12 a decreasing pattern which shows a very parallel trend to the feldspar abundance throughout the section (Fig. 4). Calcite with a general decreasing trend up to the top of section follows an opposite pattern to the quartz and feldspar abundances. Dolomite, on the other hand, shows enrichment between Ö9-Ö12, coinciding with high carbonate abundances (Fig. 4). Total phyllosilicates demonstrate an average increasing trend from bottom to top of the section. The relative abundance of clay minerals in the fine fraction exhibits variations along the section. Smectite as the most abundant mineral in the fine fraction plots an identical pattern to the total phyllosilicates. Illite, as a detrital phase, shows an enrichment trend between Ö2-Ö9, where smectite slightly decreases. The total amount of kaolinite and chlorite does not exhibit a regular pattern but decreases between Ö10-Ö12, where carbonate accumulation is very high. Mineral abundances in the calcrete compositions reveal very high calcite content (up to 90%) between Ö2-Ö12, with low amounts of quartz and feldspar. Dolomite appears between Ö10-Ö12 with an enrichment (Fig. 4).

Fig. 4
figure 4

Mineral abundances versus depth diagrams for the samples along the studied section in the lower Kürtün Valley

Full size image

According to the mineralogical data, MIS4 is characterized by higher calcite and illite, along with lower phyllosilicate and smectite (Fig. 4). Although there are no samples available to understand MIS3 along the section, MIS2 is represented by more samples showing lower calcite but higher quartz, feldspar and phyllosilicate (high smectite, low illite). Calcretes formed during MIS2 are highly enriched in calcite. The top of the section corresponds to MIS1 during which low calcite but high phyllosilicate (especially high smectite) typify this period.

Molecular weathering ratios and stable isotope values

Whole-rock major element compositions of the samples are provided in the Supplementary Table 1 (ST.1). Molecular weathering ratios, which are very important climofunctions used in many palaeoclimatic reconstruction studies, reveal the molecular behaviours under climate-driven conditions. All molecular weathering ratios and chemical index of alteration values for all samples are listed in Table 4. Relative changes in selected molecular weathering ratios, CIA and CIA-K with respect to depth are given in Fig. 5. The lowermost part of the section, represented by Ö1/C1samples, corresponds to MIS4 characterized by relatively higher salinization (> 1) and calcification (> 2) but lower hydration (< 1), infiltration, CIA-K (14.1%) and CIA (13.8%) (Table 4; Fig. 5). During MIS2, average increasing pattern for hydration, infiltration, CIA-K (19-40.2%) and CIA (18.5–38.8%) are all parallel to each other between Ö2-Ö12. Conversely, salinization (< 1) and calcification (1–3) exhibit a decreasing trend along MIS2 up to the top of section (Fig. 5). The sample Ö13/C6 dates back to MIS1, characterized by infiltration, CIA-K (52.7%) and CIA (50.5%) values slightly higher than the previous periods; similarly salinity (< 0.2) and calcification (< 1) values are much lower (Table 4; Fig. 5). The calculated hydrolysis and clayeyness values show the same tendency to increase and decrease at the same levels; but on average, they have an increasing value up to the top of section (Table 4).

Table 4 The stable isotope values, molecular weathering ratios and chemical index of alteration values of the samples of palaeosoil-sediment section in the lower Kürtün Valley (Samsun)
Full size table
Fig. 5
figure 5

Relative changes between stable isotopes values, molecular weathering ratios, and chemical index of alteration values with respect to depth

Full size image

The isotope composition of palaeosoil-sediment samples in the lower Kürtün Valley exhibits a slightly wider range in δ13C composition, ranging from − 3.29‰ to − 10.13‰ and a relatively narrower range in δ18O composition, ranging from − 5.91‰ to − 8.56‰ (Table 4). Changes in stable isotope values with respect to depth are plotted in Fig. 5. From the bottom to top of the section, the patterns followed by δ13C and δ18O are parallel to each other. Depletions and enrichments of both stable isotopes show a good covariance (Fig. 5). From MIS4 (Ö1) to the beginning of MIS2 (Ö2), the δ18O and δ13C values decrease, while within MIS2, the stable isotope values show a zigzag pattern. From Ö2 to Ö7, δ18O values increase from − 7.90‰ to -5.91‰, and similarly, δ13C values reach up to -3.29‰. This pattern is followed by a sudden drop in δ18O (-8.56‰) and δ13C (-9.90‰) values. Towards the end of MIS2, both stable isotopes are enriched to -6.30‰ (for δ18O) and − 5.36‰ (for δ13C). The average trend from MIS2 to MIS1 represents a decreasing pattern for both stable isotopes.

Palynoflora

A total of 15 sediment samples were collected from the measured stratigraphic section in the lower Kürtün Valey for palynological analysis. While several of them do not have enough palynomorphs (Ö6, Ö8, Ö10, Ö11 and Ö12), some samples include abundant pollen and spores. The described palynomorphs were grouped as spores, angiosperms, gymnosperm, arboreal pollen (AP), non-arboreal pollen (NAP) and non-pollen palynomorphs. A total of 19 sporomophs taxa (1 taxa of spores and 18 taxa of pollen) are identified as shown in the detailed palynological diagram (Fig. 6). Additionally, Pseudoschizaceae, Glomus, zooclasts, cuticul and fungal spores are recorded in the samples of the lower Kürtün Valey.

Fig. 6
figure 6

Detailed pollen diagram of the palaeosoil samples in the stratigraphic section of Kürtün Valley. Note that the distance between samples is not at scale (for detailed location of samples see Fig. 3)

Full size image

The samples were grouped into three different clusters according to their palynomorph content. The first subcluster (Ö1) of Cluster C in the stratigraphic section is only represented by abundance of Pinaceae-Pinus haploxylon type (Fig. 6). The other subcluster of Ö2, Ö3, Ö4 and Ö5 samples in Cluster C is characterized by moderate values (3–10%) of this gymnosperm species and abundance of Quercus evergreen type (5–20%). Additionally, Myrica of the riparian forest elements and Carpinus of the mixed mesophytic forest are less abundant in the palynoflora of Cluster C. The herbaceous species (NAP) represented by Asteraceae-Asteroideae-Tubuliflorea type and Artemisia have a low percentage (1–2%). Palynomorphs consist of zooclasts (1–2%) and cuticul (10–90%) are abundantly recorded (Fig. 6). In addition, abundant phytoclasts and a low percentage of amorphous organic matter were observed in sample Ö1 taken from the bottom of the section, and similar contents were also recorded in samples Ö2, Ö3, Ö4 and Ö5. According to the organic matter content, it could be said that the samples were deposited close to the stream source (Tyson 1995). The Cluster B, defined from the Ö7 and Ö9 samples, is characterized by an increase in the gymnosperm values (Pinaceae-Pinus haploxylon type) (from 5 to 40%). While a decreasing percentage of Quercus evergreen type is observed, values of Myrica and Fagaceae-Fagus increased in Cluster B. Furthermore, the grassland species is only characterized by Poaceae (1% in the Ö7 sample). Palynomorphs of the Ö7 consist of cuticul and fungal spores; however, these forms of the Ö9 are represented by Glomus, indicating reworking of Pseudoschizaceae (freshwater algae). Cluster A is significantly different from the other clusters. Relatively, although the AP ratio decreased, the NAP ratio and diversity increased. The species of AP are represented by Pinaceae-Pinus haploxylon type (5–15%) and Cupressaceae (1%) of the gymnosperm, Salix (1%) and Quercus evergreen type (1–2%). The NAP is characterized by Nymphaeaceae, Dipsacaceae, Geraniaceae, Brassicaceae, Asteraceae spp., Asteraceae-Asteroideae-Tubuliflorea and Cichorioideae-Ligulifloreae types, Centaurea, and Amaranthaceae. Besides, in this cluster, all palynomorphs are abundantly recorded (Fig. 6).

Interpretation of proxy data

The Black Sea coastal mountains emerged as land at the beginning of the Oligocene and have since then been uplifted and eroded by erosional processes. Changes in global climate and tectonic activity led to the formation of extensive erosion surfaces during the Neogene and the development of river terraces during the Quaternary. The palaeosoil sequence studied here is situated on one such fluvial terrace (Fig. 1). After the Kürtün Stream deepened its bed, colluvial materials covered this terrace forming a palaeosoil sequence.

In this area, the rate of sedimentary deposition was quite slow until the MIS2 period, as indicated by the age assessments obtained in the sequence containing palaeosoil levels in Kürtün Valley. No age evidence for MIS3 was obtained. The sequence during MIS5c is represented by a fluvial succession consisting of blocky conglomerate and coarse sands. A sequence containing palaeosoil levels deposited between the MIS4 and MIS2, yellowish brown horizontal bedding and calcrete levels were deposited towards the northeast on the fluvial succession (Fig. 7). Towards the southwest, a succession of sandstone and claystone containing grain-supported sandstone and dark grey coloured palaeosoil levels rich in organic matter deposited in MIS1, was observed on the fluvial sediments at the bottom of the sequence. Contemporary sediments were deposited over this entire sequence by eroding the old sequence in some places. Palaeoclimatic and palaeoenvironmental data obtained from this sequence, containing sedimentary deposits from different conditions in a narrow area play a very important role in understanding the events during the deposition. For this reason, each sequence was evaluated individually (Fig. 7).

Fig. 7
figure 7

Field view of the sedimentary sequence from which the stratigraphic section was measured

Full size image

The abundances of the minerals in the palaeosoils and calcretes show different trends through the section. Calcretes, with high amounts of calcite but low quartz and feldspar, reveal the dry conditions. In contrast, palaeosoils are enriched in quartz, feldspar and phyllosilicates but depleted in carbonate minerals, indicating wet climatic conditions. The reverse covariation between total clay (phyllosilicates) and calcite serves as a mineralogical proxy revealing humid climatic conditions intercalated with dry periods along the palaeosoil-sediment section (Singer 1980; Küçükuysal et al. 2013) (Fig. 5). An increase in total clay content implies intense hydrolysis and a higher degree of chemical weathering, favouring precipitation greater than evaporation (Ülgen et al. 2012). On the other hand, an increase in the amount of calcite refers to much drier conditions, with a degree of evaporation greater than precipitation (Singer 1984). Although few data could be obtained from the fluvial succession deposited during the beginning of MIS5c-MIS4, comparing sample Ö1 with the sample compiled after (Ö2) suggests an increasingly wet period at the base of the succession. In addition to the decrease in calcite amount, the increase in total clay, especially smectite, supports dominance of wet periods. Based on the XRD analysis, between Ö2-Ö12, an average increasing pattern for total clay is observed, with some small and short drops in the smectite amounts coinciding with the enrichments in calcite, especially for the calcrete samples (Fig. 4). Such periods in the MIS2 refer to drier conditions. In the same intervals, quartz, feldspar and illite contents also follow a decreasing trend in an average pattern. Dolomite, a carbonate mineral favouring drier conditions, shows an enrichment toward the end of MIS2, suggesting relatively much more drier conditions for calcrete formation (Küçükuysal and Kapur 2014) (Fig. 4). The MIS2 period with higher total clay, higher smectite but lower calcite and illite contents matches with higher CIA-K, CIA, and hydrolysis but lower salinization and calcification. Unfortunately, not represented with high-resolution samples, the transition from MIS2 to MIS1 suggests an increase in smectite; CIA-K, CIA and hydrolysis, with the decrease in δ18O, calcite, illite, kaolinite + chlorite, salinization, and calcification values, indicates the continuation of humid climatic conditions (Fig. 4). All these proxy fluctuations refer to the periodic wetting and drying conditions prevailing during the formation of soil and calcrete during MIS2 (Figs. 4 and 5).

Carpinus, Quercus evergreen type, Asteraceae-Asteroideae type and Artemisia, which prefer to live in cool, arid climatic conditions, were obtained from samples Ö2 to Ö5, according to palynological records obtained from especially clayey levels throughout the section. In addition, the decrease of smectite values and the increase of illite values support the cooling and drought in the climate, as determined by the pollen records at the same levels. During the deposition of the sequence from which samples Ö5 and Ö9 were collected, the pollen records show that humid conditions developed. The presence of Myrica indicates that riparian conditions developed in the depositional area due to increased precipitation. This interpretation coincides with the increase in smectite and total clay, but the decrease in the oxygen isotope values to reach the lowest values explains the increased palaeotemperature and precipitation.

In the palynological record, the abundance of Nymphaeaceae, Dipsacaceae, Brassicaceae, Geraniaceae, Salix, as well as Pseudoschizaceae and fungal spores, could be the result of increased moist conditions in MIS1. However, the presence of herbaceous plants (e.g. Amaranthaceae, Centaurea, Asteraceae-Asteroideae and Cichorioideae types) observed in samples Ö12-Ö15 indicates dry climatic conditions (Fig. 6). The increasing oxygen value from Ö10 to Ö12 samples and the presence of dolomite must be related to these dry periods (Fig. 4; Table 4). Glomus, recorded with a high percentage towards the upper levels of the succession, indicates erosion. Additionally, δ13C values of the calcrete samples are compared with the photosynthetic pathways (Cerling and Quade 1993). δ13C values of C1, C3, C5 and C6 samples ranges between − 24.2 and − 25.5‰, indicating higher input of 13C from soil respiration and correlate with C3 plants (Table 2). On the other hand, sample C2 with − 7.8‰ and sample C4 with − 9.9‰ refer to vegetation dominated by C4 plants. Comparing the other proxy data of C2 and C4 samples, relatively more arid conditions are evidenced by an increase in salinity and calcification but a decrease in CIA and CIA-K in association with carbon isotope values.

The stable isotope data of the samples collected from Samsun Kürtün Valley throughout the measured section were compared with the data obtained from the simultaneously deposited speleothem of Sofular Cave in the Western Black Sea, where important climate data were obtained. The O/C values of the samples from which age data were obtained were used for comparison (Fig. 8). The lack of age data for MIS 3 and MIS 5b indicates that localized erosion occurred during these periods. Oxidation levels at the base of the measured section support erosional conditions (Fig. 7). Throughout the section, the δ18O data of the Ö1 sample from the MIS 2 period were found to be more positive than the data from the Sofular Cave, indicating more arid conditions during the deposition of the sediments in the Kürtün Valley (Fig. 8). During this period, when the sea level in the Black Sea decreased, the drought conditions in the Kürtün Valley were consistent with global climate changes. Additionally, the data obtained from MIS 2 in the Kürtün Valley show that a drier climate persisted compared to the Sofular Area. However, a humid phase was observed within the arid period in the samples from the transition from MIS 2 to MIS 3 in the Kürtün Valley. The climate during deposition from the level where sample Ö13 was collected is compatible with the data from Sofular Cave which indicate that the general climatic conditions changed to arid from MIS2 to MIS1. As a result, during the deposition of the section deposited from the MIS 4 to MIS 2 to MIS 1 periods in the Kürtün Valley, more arid conditions existed compared to those during the precipitation of the speleothems in the Sofular Cave, where data from the Western Black Sea were obtained. Especially in the MIS 2 period, when the sea level of the Black Sea decreased and arid conditions were observed globally, it can be suggested that Samsun Kürtün Valley experienced a more severe drought.

Fig. 8
figure 8

Comparison of Kürtün Valley stable isotope records with sedimentary records from stacked Sofular δ18O data, the Black Sea and global sea level (Panin and Popescu 2007; Badertscher et al. 2011; Yanina 2014), a) stable isotope data of Samsun Kürtün Valey, b) stable isotope curve of the study area, c) Sofular δ18O records, with the colours and colour-coded dots with error bars denoting different stalagmites and 230Th-ages respectively (Badertscher et al. 2011). Solid horizontal line (black) indicates the threshold value of − 8.5 ± 1 h VPDB, d) Global sea-level curve (the error is ± 6 m) (Siddall et al. 2003; Rohling et al. 2010). Blue shaded area marks periods when global sea level was above the Bosphorus sill depth of ∼−35 mbsl, and e) LR04 stacked isotope record. Numbers denote marine isotope stages (MIS) (Lisiecki and Raymo 2005)

Full size image

Conclusion

Multi-proxy data obtained from a palaeosoil-sediment section located in the lower part of the Kürtün Valley (Samsun region, Northern Türkiye) provide a crucial contribution to the palaeoclimate and paleoenvironment of the Samsun coastal belt of the Late Pleistocene-Holocene period (from MIS5c to MIS1). The palaeosoil-sediment section dates back to the last interglacial period (108 ± 9 ka) and corresponds to MIS5c. At the end of this warm period, the climate cooled, sea level fell and river valleys deepened. In a similar way, the Kürtün Stream deepened its valley, leaving old river deposits 10 m above the stream bed. Meanwhile, colluvial material carried down the slopes covered these river deposits, forming a palaeosoil unit. An age date of 68 ± 9 ka (MIS4) was obtained from the sample taken from the lower part of this palaeosoil-sediment section. Multi-proxy data show that cold and dry conditions prevailed during this period. On the other hand, sand layers in this part of the section indicate occasional flood events. Age data between 28,959 and 25,768 cal BP (MIS2) were obtained from samples taken from the middle part of this palaeosoil-sediment section. Multi-proxy data indicate humid conditions with intermittent dry pulses, characteristic of MIS2 climates in the region. The palaeosoil-sediment section ends at 8381 cal BP, corresponding to the MIS1 period of increasing warming and humidity. The calcrete levels in the palaeosoil sequence indicate relatively dry conditions during a longer phase of hot-humid climate.

As a result, the climate findings for the MIS1 and MIS2 periods obtained from the palaeosoil-sediment sequence are comparable with nearby and distant speleothem records as well as sediment records, indicating their alignment with regional and global climate records. According to our findings, the palaeoclimatic conditions during the deposition of the clastic succession deposited in Samsun Kürtün Valley (Central Black Sea) from MIS 4 to MIS 1 were more arid than in the Western Black Sea (Sofular Cave). Consequently, this calcrete-palaeosol sequence has provided important data for the reconstruction of the Late Quaternary palaeoclimates and palaeoenvironments of the region. However, high-resolution studies must continue to understand better the connection between local climates and global climate phenomena. Additionally, the geosite represented by this palaeosoil-sediment section is suitable for further scientific research, education and geotourism purposes. It is recommended for preservation and conservation for future generations.