Late Cenozoic fluvial development within the Sea of Azov and Black Sea coastal plains

Global and Planetary Change 68 (2009) 270–287

Contents lists available at ScienceDirect

Global and Planetary Change

j ourna l homepage: www.e lsev ie r.com/ locate /g lop lacha

Late Cenozoic fluvial development within the Sea of Azov and Black Sea coastal plains

A. Matoshko ⁎, P. Gozhik, V. SemenenkoInstitute of Geological Sciences, National Academy of Sciences of Ukraine, 55B Gonchara Street, 01054 Kiev, Ukraine

⁎ Corresponding author.E-mail address: [email protected] (A. Mato

0921-8181/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.gloplacha.2009.03.003

a b s t r a c t
a r t i c l e i n f o
Article history:Accepted 20 February 2009Available online 16 March 2009

Keywords:Late CenozoicParatethysBlack SeaSea of Azovfluvial developmentcoastal plains

Late Cenozoic terrestrial deposits are widespread across the northern coastal regions of the Black Sea and theSea of Azov and represent diverse fluvial, estuarine and deltaic environments. The dating and correlation ofthese deposits rely on stratigraphically-associated marine index beds, mammalian and molluscan faunas andmagnetostratigraphy. In detail the geometries of these sediment bodies are extremely complex, typicallyvarying between localities and representing many cycles of incision and aggradation. However, the overalldisposition of the sediments reflects the transition from the uplifting sediment source region to the northand the subsiding depocentre in the interior of the Black Sea to the south. Since the Middle Miocene the areaof the Paratethys/Black Sea depocentre has decreased significantly, but since the Middle Pliocene the hingezone between uplift and subsidence has been located close to the modern coastline. A combination ofregional and local differential crustal movements has given rise to the great variety of fluvial sedimentbodies, to the erosion–aggradation cycles, different phases and river activity and to the various fluviallandforms that have all been important in landscape development in this region during the past 12 Ma. Thefluvial erosion–accumulation cycles (during the upper Serravillian–Messinian, the Zanclean–late Gelasian,and the Pleistocene) and corresponding cycles of relief dissection and planation are reconstructed against abackground of local sea-level changes and climatic variations determined from palaeobotanical data. Themaximum fluvial incision occurred in the early Zanclean time with alluvial coastal plains, unique in this area,developing in the Gelasian. Increased climatic aridity during the Pleistocene caused a reduction of fluvialactivity in comparison with the Late Miocene and Pliocene. The sea-level oscillations and Pleistoceneglaciations affected fluvial processes in different ways. The most remarkable events were the substantialreduction of fluvial activity during the Messinian dessication in the Black Sea and drainage of the shelf, withintensive dissection, coeval with the Last Glaciation.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

The chronology and disposition of the Late Cenozoic fluvialdeposits of the main rivers of Eastern Europe, the Don, Dniester,Dnieper and Volga, have been reviewed by Matoshko et al. (2002,2004). The present study extends that review to the buried fluvialdeposits around the Black Sea and Sea of Azov, including areas beyondthe valleys of the above-mentioned rivers (Fig. 1). This report presentsa brief summary of the vast amount of material held by nationalgeological surveys, most of which is unpublished, along with datafrom published sources and results of the authors' own investigations.

The study region forms part of the Paratethys Basin and itsterrestrial surroundings. This basin was a large landlocked former seathat became isolated from the global marine environment during theMiddle Miocene and therefore developed unique endemic faunas.Palaeosalinity inferred from foraminifera, ostracods, and calcareousnannofossils indicates a variety of marine, brackish, and freshwaterenvironments within the basin (Jones and Simmons, 1997). In

shko).

ll rights reserved.

accordance with local custom, sediments deposited in the ParatethysSea and its modern remnants, the Black Sea and Sea of Azov, areclassified as marine, although during much of the Late Cenozoic thesewater bodies have more closely resembled lacustrine environments.The isolation of the Paratethys Basin means that, from the late MiddleMiocene onwards, stratigraphic correlation with global chronology isproblematic, because the marine taxa that define the globalchronology are typically absent. However, magnetostratigraphy andmammalian and molluscan biostratigraphy can be used for strati-graphic correlation, as the present study illustrates.

The local stratigraphic nomenclature in the study region hasreflected the conceptions of many different authors over more than acentury and is thus complex. This nomenclature is expressed in termsof ‘beds’, ‘suites’ and ‘series’ (NSCU, 1997). A suite is a mappablestratigraphic unit that is broadly equivalent to a formation; a series,which may comprise a number of designated suites, is broadlyequivalent to a group. In this paper preference is given to the presentregional Ukrainian stratigraphical scale (NSCU, 1997) for the Mioceneand Pliocene, which is correlated with subdivisions of the Gradsteinet al. (2004) global time scale. Some historic terminology is also used,and some new nomenclature is proposed. The resulting interpretation

mailto:[email protected]

http://dx.doi.org/10.1016/j.gloplacha.2009.03.003

http://www.sciencedirect.com/science/journal/09218181

Fig. 1.Map of the study region, showing the locations of sediments, landforms, and localities discussed in the text, based on data from Zamorii (1940), Moliavko (1960), Mulika and Vyrvyklenko (1966), Goretsky (1970), Chirka (1974), Matsuiet al. (1981), Bilinkis (1992), Lericolais et al. (1998), Rekovets (1994), Shnyukov et al. (1999) and new data. Liman and delta localities are: 1, Kakhul Liman; 2, Yalpukh Liman; 3, Katlabukh Liman; 4, Danube Delta; 5, Sasyk Liman; 6, DniesterLiman; and 7, Dnieper Liman. Localities with palaeontological and magnetostratigraphic data are: 1, Baimakliia; 2, Teteresht; 3, Yetuliia Nuoe; 7, Roksolany; and 13, Kairy. Localities with palaeontological data are: 4, Nagorne; 5, Ozerne;6, Suvorovo; 8, Kuialnyk; 9, Kryzhanivka; 14, Obitochnaya; 15, Urzuf or Kulikovskoe; 16, Shirokino; and 17, Khapry. Other localities are: 10, Cape Karabush; 11, Polovinka; 12, Parutino; and 18, Parkan. The coastline during the maximumtransgressions in the Early Kuialnyk stage appears to have typically been near the modern coastline (Semenenko, 1975) and so is not illustrated in this figure.

271A.M

atoshkoet

al./Global

andPlanetary

Change68

(2009)270

–287

Fig. 2. Correlation table for Miocene and Pliocene deposits in the study region. Magnetostratigraphy is after Mankinen and Dalrymple (1979) and Gradstein et al. (2004), with darkornament indicating normal magnetization and using notation from Cande and Kent (1995); M, Ga and Gi indicate the Matuyama, Gauss and Gilbert chrons; Ol, Ka, Ma, Co, Nu, Si, Thindicate the Olduvai, Kaena, Mammoth, Cochiti, Nunivak, Sidufjall, and Thvera subchrons. The beginning of the Pontian stage is at 7.5 Ma and the start of the preceding Meotian stage(end of the Sarmatian stage) is younger than 10.7 Ma (Pevzner et al., 2003). The Balta Series spans the Upper Sarmatian and Meotian, as well as the Pontian, with its base at 12.5–12.0 Ma (Zosimovich et al., 1975).

272 A. Matoshko et al. / Global and Planetary Change 68 (2009) 270–287

of the regional stratigraphy is summarized in Fig. 2 for the LateMiocene and Pliocene and in Fig. 3 for the Pleistocene.

The Paratethys Basin became isolated during the Sarmatian stageof the local stratigraphy (late Middle Miocene; Serravallian) (e.g.,Jones and Simmons, 1997). Much of the interior of the basinsubsequently uplifted above the level of the Paratethys ‘Sea’, suchthat marine deposits are overlain by fluvial deposits that may bedisposed as river terraces or as deltas or may infill erosional incisions.Some parts of the Paratethys Basin (e.g., the Pannonian Basin andcentral Black Sea) have been characterized by Late Cenozoicsubsidence, resulting in the accumulation of thick stacked sequences.Following the Sarmatian stage, the Meotian and Pontian stages ofParatethys are broadly equivalent to the Late Miocene (they span theUpper Serravillian, Tortonian, and Messinian); the Kimmerian stage isbroadly equivalent to the Early Pliocene (Zanclean and lowerPiacenzian) and the Akchagyl stage to the Late Pliocene (upperPiacenzian and Gelasian) (Fig. 2), being also an analogue of theAkchagyl stage in the Caspian Sea basin, another part of the ParatethysBasin. Local stage names also exist for subdivisions of the Pleistocene(Fig. 3).

The stratigraphical positions of the fluvial deposits are establishedwith reference to palaeontologically studied Late Miocene–Pliocene

marine index beds supported by magnetostratigraphic evidence (e.g.,Didkovskii and Nosovskii 1975, Karmishina 1975; Semenenko, 1975,1989; Paramonova et al., 1979; Ananova et al., 1985; Gozhik, 1992;Bilinkis, 1992; Krakhmalnaya et al., 1993, Gozhik, 2002; Gozhik et al.,2006) as well as Pleistocene index beds of the Sea of Azov and BlackSea (e.g., Nevesskaya, 1965; Fedorov, 1978; Gozhik and Novoselsky,1989; Mihailescu and Markova, 1992). Terrestrial sections with richmammalian and molluscan faunas also provide age control.

The chronology of the marine index beds of the Paratethys Basinhas been debated for decades, given the available magnetostrati-graphic, biostratigraphic and nanoplankton evidence (e.g., Jones andSimmons,1997). In particular, this debate has concerned the age of thePontian stage. For instance, Steininger (1999), Chumakov (2000) andmany others have correlated the Pontian with the Messinian (or,indeed, have defined the Pontian and Messinian as equivalent). Incontrast, Zubakov (2000) correlated the Pontian with the UpperTortonian and Lower Messinian and Popov et al. (2006) correlated itwith the Upper Messinian.

New data and new approaches enable revision and more precisedefinition of the previous incompatible schemes. Within the CentralParatethys (Western Hungary) after the introduction of the Transda-nubian stage (9.0–7.4 Ma; defined in terms of biostratigraphy,

Fig. 3. Correlation table for Pleistocene and Holocene deposits. Magnetostratigraphic data are from the same sources as for Fig. 2. The same notation is also used, with B, and Mdenoting the Brunhes and Matuyama chrons and Ja and Old the Jaramillo and Olduvai subchrons. Note that the ‘Upper Palaeolithic’ faunal zone (abbreviated to Up. Pal.) is unrelatedto the archaeological definition of the Upper Palaeolithic as the span of time during which anatomically modern humans have lived in the region. The allocation of fluvial deposits totimescale is based on the data of Goretsky (1970), Fedorov (1978), Gozhik and Novoselsky (1989), Mihailescu andMarkova (1992), Rekovets (1994), Markova (1998), Matoshko et al.(2002) and Dodonov et al. (2005). Details see in the text. The stages of the Black Sea are cited according to Fedorov (1978). The oxygen isotope record is after Shackleton et al. (1990).

273A. Matoshko et al. / Global and Planetary Change 68 (2009) 270–287

lithostratigraphy and magnetostratigraphy) the Pontian Stage becameequivalent to the uppermost Tortonian and the entire Messinian(Sacchi and Horvath, 2002). Data from the Black Sea and Caspian Searegions have led to the development of an alternative chronologywiththe base Meotian at 9.6 Ma, the Meotian–Pontian boundary at 7.5 Ma,and the top Pontian at 6.7–6.6 Ma (Pevzner et al., 2003). In contrast,magnetostratigraphic and other data from the Mio-Pliocene Para-tethyan sedimentary sequences of the Focsani basin in the CarpathianForedeep, Romania (Vasiliev et al., 2004), place the Meotian–Pontian

boundary at ~5.8 Ma, the Pontian–Dacian boundary at 4.8 Ma, and theDacian–Romanian boundary at 4.1 Ma. Seismic data from the BLaSONsurveys in the Black Sea (Gillet et al., 2007) and their correlation withDSDP borehole data also indicate that the Pontian stage overlaps theMiocene–Pliocene boundary. These differences (resolution of which isbeyond the scope of this study) should be borne in mind throughoutthe present review.

The text begins with a chronological account of the development offluvial systems, of associated environmental conditions and of

Fig. 4. (a) Cross-section through the Balta Series and Upper Kuialnyk Suite within the South Bukh–Tiligul interfluve in SW Ukraine (see Fig. 1, profile 1, for location). Ages of rocks areabbreviated thus: Ar-PR, Archaean–Proterozoic; K2, Upper Cretaceous; E, Eocene; N1s Sarmatian; N1m, Meotian; N1p, Pontian, N1b, Balta Series; N2ku, Kuialnyk Series; Pl1, LowerPleistocene; Pl1–3, Lower Pleistocene–Upper Pleistocene; Pl3-H, Upper Pleistocene–Holocene. Sediments assigned to the Upper Kuialnyk Suite have been shaded for emphasis.(b) Locations of boreholes used to compile part (a).


landscape evolution, starting in the Late Miocene. It then discusses theavailable age control evidence and the stratigraphical relationsbetween sediments of different ages that demonstrate vertical crustalmotions in the study region. An important data source, madeaccessible in the international literature for the first time, is providedby five cross-sections at key localities (profiles 1 to 5, at localitiesindicated in Fig. 1), which have been prepared using boreholeevidence (Figs. 4–8). The changing environmental conditions andtheir relation to the development of the Black Sea basin are thenconsidered.

2. The Late Miocene

The Upper Miocene Balta Series occurs mainly within the basin ofthe River Southern Bukh (or Bug), occupying an area of ~40,000 km2,part of which is shown in Fig. 1. Its typical andmaximum thickness are40–50 m and 135m; its top is 100–140 m a.s.l. (above sea level) in thesouth, rising to 200–250 m in the north, northeast and northwest.

In some localities the Balta Series unconformably overlies the erodedtop of marine deposits of the Novopetrivtsi stage (Lower and MiddleMiocene). Elsewhere it overlies Precambrian basement or residualdeposits (‘wastemantle’ derived from theweathering of the basement).The uneven base of the Balta Series descends to 60–50 m a.s.l. in thesouth, where these deposits infill incisions and their middle and upperparts are replaced by Middle–Upper Sarmatian, Meotian and Pontianmarine deposits (Melnik,1970; Zosimovich et al., 1975). The Balta Seriesis overlain by red-brown and brown clays (presumed to be products ofPliocene and Early Pleistocene weathering) as well as Pleistocenesubaerial silts and loesses with palaeosols, which are widespread in theAzov–Black Sea coastal area. The fluvial Stolnichen Series (LowerKimmerian) rests upon the Balta Suite on the right bank of the MiddleDniester, with four Pliocene suites (the Kuchurgan, Runkashiv, Vadul-

lui-Vode, and Rashkiv suites) inset into the Balta Series (Matoshko et al.,2004); collectively these deposits form the highest terraces of the RiverDniester.

Although themselves largely unfossiliferous, the Balta Seriesdeposits can be assigned to the Upper Sarmatian, Meotian andPontian stages on the basis of fossils from the interbedded marinedeposits. Reported taxa include Mactridae (bivalve molluscs endemicto the Paratethys Basin) indicative of the Middle (Upper, in theopinion of the present authors) Sarmatian (Monodacna caspia andMactra bulgarica) in such marine interbeds in the south (Zosimovichet al., 1975). Also reported areMeotian freshwater molluscs (includingPsilunio radiatodentatus, P. novorossicus, and Plicatibaphia flabellatusrossicus) (Zosimovich et al., 1975; Gozhik, 2002) andmammals of LateSarmatian or Meotian age, including rare remains of Hipparion sp.,Mastodon. borsoni Hays, Deinotherium giganteum Kaup, Aceratheriumsp., and Macharoides sp. (Moliavko, 1960).

The Balta deposits consist mainly of alternations of sand, clay andsilt, with gravel, sandstone and conglomerate. Althoughmainly fluvial,deltaic and nearshore marine deposits also occur, and becomeincreasingly important southward, where carbonate interbeds arealso observed. Aeolian and slope-wash deposits are less significant.According to Melnik (1970), channel and overbank as well asabandoned channel facies can be identified within the fluvial andnearshore deposits, which include beach, lagoonal and bar facies. Insome sections up to eleven stacked fluvial suites have been resolved(Hubka, 1969); hence the designation of the Balta deposits as a series.Sections in the Balta Series typically demonstrate complex internalstructure, characterized by incision and superimposition.

In the west, in the Dniester–South Bukh interfluve, the Baltadeposits mainly consist of sand, with gravel also frequently observed;east of the South Bukh the deposits are typically finer. The deposits inthewest are characterized by jasper-like rocks or so-called ‘Carpathian

Fig. 5. (a) Cross-section through the Near-Danube Series and Upper Porat Series betweenthe towns of Reni and Izmail in the Ukraine–Moldova–Romania border region (see Fig. 1,profile 2, for location). Ages of rocks are abbreviated thus: PR3, Upper Proterozoic; Mz-Cz,Mezozoic–Cenozoic; N1, Miocene; N1s, Sarmatian; N1m, Meotian; N1p, Pontian; N2,Pliocene; N2k, Kimmerian; N2nd, Near-Danube Suite; N2pr1, Lower Porat Suite; N2pr2,Upper Porat Suite; N2-Pl1, Pliocene–Lower Pleistocene; Pl2, Middle Pleistocene; Pl2oz,Ozerne Suite; Pl2eu, Palaeo-Euxinian Horizon; Pl2–3, Middle Pleistocene–Upper Pleisto-cene; Pl3-H, Upper Pleistocene–Holocene. Note that the southward decrease in altitude ofthedeposits is interpreted as inpart due to southward tilting and inpart due to downthrowacross active faults. (b) Locations of boreholes used to compile part (a).


pebbles’, although the true source area is unknown (i.e., they are notnecessarily derived from erosion of the Carpathian Mountains). Flintderived from local Upper Cretaceous chalk and clasts of Palaeogeneand Silurian rocks also occur. The sands consist mainly of quartz andfeldspar; although abrupt changes in composition are found, overall itis close to the composition of the underlying Sarmatian and olderMiocene (Paratethyan) marine rocks (Melnik, 1970).

Although widespread elsewhere, Miocene and Pliocene fluvialdeposits are absent in themodern Lower Don (Goretsky, 1982). At this

time the mouth of palaeo-Don was thus probably farther east,somewhere in the Caucasus foreland. South of the Sea of Azov, theLate Miocene succession includes the marine Upper Pontian lime-stones and clays of the Novorossiisk Beds. These include near-shoresands, which are not easily distinguished from their terrestrialanalogues.

3. The Early Pliocene

In the regional stratigraphy the Pliocene is subdivided into theKimmerian and Akchagyl stages, the boundary between them beingplaced at 3.0 Ma (NSCU, 1997; Fig. 2), somewhat earlier than in someprevious interpretations. Fragmentary fluvial deposits of Kimmerianage are widespread (Fig. 1), being characterized by polymictcomposition (Bilinkis, 1992). These deposits occur between theDniester and Prut (the Stolnichen Series), within the MiddlePleistocene delta of the Dnieper (the Tendra Suite), along the LowerPrut and Lower Danube (the Near-Danube Series, in previousdefinitions, Fig. 2: the lower part of Porat or Nizhneporat deposits),in the Prut–Yalpukh interfluve (the Karboliia Series), and between theDniester and South Bukh (the Kuchurgan and Ananiiv suites).

The Stolnichen Series is preserved solely in modern interfluves,overlying Middle–Upper Sarmatian, Meotian, Balta Series and Pontiandeposits, including the Upper Pontian Novorossiisk Beds. Its thicknessvaries from 6–7m to 16m, its altitude varying between 150 and 400ma.s.l. Deposits of channel facies consist of cross-bedded sand andgravel, with clasts of flint, quartz and jasper, redeposited clayey andcarbonate concretions, and debris of local rocks. Deposits of overbankfacies, consisting of greenish grey silts and clays, locally overlie thechannel deposits (Bukatchuk and Negadaev-Nikonov, 1963;Bukatchuk et al., 1983). Sediments of the Ananiev Suite are consideredequivalent to the Stolnichen Series due to the similar lithology and thepresence of shells of Plicatibaphia flabelatiformis (Bukatchuk andNegadaev-Nikonov, 1963). They also overlie the Balta Series in a smallarea west of the South Bukh.

The Upper Kimmerian fluvial Kuchurgan Suite occurs in interfluvesbetween 120 and 260 m a.s.l., rising to the Carpathian piedmontwhere it correlates with the Krasna Level at 300–500 m a.s.l. (Bilinkis,1992). This deposit is inset several tens of metres below the level ofthe Stolnichen Series, such that the Stolnichen Series and KuchurganSuite form the two highest terraces of the River Dniester (Bukatchuket al., 1983; Matoshko et al., 2004). The lower 8–11 m of theKuchurgan Suite consists of medium sands with interbeds of poorlysorted sands; its upper 4–5 m consists of poorly sorted sand, gravel,and conglomerate, which unconformably overlie the eroded top of thelower strata. The Kuchurgan Suite occurs close to the outcrop of theAnaniev Suite, but the stratigraphical relation between these depositsis not established (Hubka, 1981).

An analogue of the Kuchurgan Suite farther west in the Prutcatchment (Fig. 1) is provided by the upper part of the Karboliia Series(Karboliia Beds), which consists of fluvial and deltaic deposits. Thebase of this series slopes southward from 140 m a.s.l. to 20 m belowsea level (b.s.l.), its average thickness being 70–80 m. The KarboliiaSeries includes up to eight stacked suites, each 3–8 m thick. It overliesPontian deposits and, in some places, it is above red-brown residualdeposits (waste mantle formed on the Pontian deposits); in somesections it is overlain by red buried soils of the Late Pliocene and EarlyPleistocene (Hubka et al., 1983).

Rhythmically bedded fluvial, deltaic and lacustrine deposits occurnear the mouth of the Prut and along the Lower Danube. Theirlowermost part has been referred to theupper Pontianand theoverlyingbeds (sands, silts and clays), forming between seven and nine suites(each between 5 and 13m thick), to the Kimmerian (Hubka et al., 1983;Bilinkis,1992). This part is regarded by the present authors as the fluvialNear-Danube Series; the Upper Porat Series (see below) overlies theNear-Danube Series. Channel deposits of theNear-Danube Series consist

Fig. 6. Cross-section through the buriedMiocene to Pliocene marine and fluvial deposits in the vicinity of the modern mouth of the River Dnieper (profile 3 in Fig. 1). Marine depositsare unornamented. Ages of rocks are abbreviated thus: N1, Miocene; N1s, Sarmatian; N1m, Meotian; N1p, Pontian; N2, Pliocene; N2k, Kimmerian; N2t, Tendra Suite; N2ku1, LowerKuialnyk Suite; N2ku2, Upper Kuialnyk Suite; N2-Pl1, Pliocene–Lower Pleistocene; Pl1ka, Lower Pleistocene, Kairy Suite; Pl-H, Pleistocene–Holocene; Pl2, Middle Pleistocene; Pl2eu,Euxinian Suite; Pl2pg, Periglacial Series; Pl2–3al, Aleshki Suite; Pl2–3, Middle–Upper Pleistocene; Pl2–3st, Middle–Upper Pleistocene, fluvial Suite of the Second Terrace; Pl2–3-H, Lower–Middle Pleistocene–Holocene; Pl3-H, Upper Pleistocene–Holocene.


of light grey sands with lenses of greenish grey clays containingfreshwater molluscs (Gozhik and Chirka, 1973). In their upper part thechannel sands are gravelly inplaces and overlain by overbank clays up to4m thick. These deposits are revealed by boreholes in the lowest part ofthe deep fluvial incision between the towns of Reni and Izmail in theUkraine–Moldova–Romania border region (Fig. 1). The valley of thisKimmerian River coincides here (at 84–140 m b.s.l) with the deep-seated Saint George fault zone (Bala et al., 2003), being incised in faultbreccia and Upper Proterozoic schist (Fig. 5).

Within the ancient Dnieper delta, sand lenses (forming the TendraSuite) are recorded near the floor of a depression that is infilled by theKimmerian marine clay (Fig. 6). The thalweg of the depressiondescends towards the modern coast from 40 to 120 m b.s.l. It isinferred that these deposits of the Tendra Suite are also fluvial(deltaic), having been deposited in the lower reaches of the river, closeto the low-level sea that existed in the lower Kimmerian (followingthe transgression of the same name).

The mammal fossils from the Stolnichen Series (Dicerorhinus aff.megarhinus De Christol, Antilope sp., Hipparion sp., Gazella cf. borbo-nica Deperet and Bravard., G. deperdita (Gervais), and Homotherium cf.crenatidens Fabrini; Matoshko et al., 2004) are insufficient fordistinguishing any particular biozone but suggest a Kimmerian ageand can thus be expected to be associated with the early part of theMoldova faunal complex. According to the magnetostratigraphic dataof Vangengeim et al. (1995), the oldest deposits of the Stolnichen Suite

are Early Kimmerian. The Karboliia Series has a clearer stratigraphicposition, as the Moldova faunal complex has been identified (at sitesincluding Baimakliia, Teteresht and Yetuliia Nuoe; localities 1, 2 and 3in Fig. 1; Hubka, 1982). The characteristic taxa of this complex includeMastodon. borsoni Hays, Hipparion gracile Kaup, D. megarhinus DeChristol, Promimomys moldavicus Kormos, and Pliomys kowalskiiSсhevtschenko (Matoshko et al., 2004). The unionids Eolymniumsturdzae (Cobălcescu), Potomida saratae (Teisseyre), P. sibiensis(Penecke) and Psilunio craiovensis (Tournouer), typical of the UpperDacian of Romania, have been reported in the lower part of theKarboliia Series (Hubka, 1982). Other molluscs, such as Plicatibaphiaflabellatiformis (Grigorovich-Berezovski), P. bogatschevi (Grigorovich-Berezovski), P. haneri (Horn), P. stolizkai (Neumayer) and Ruguniolenticularis Sabba, are also common in the Kuchurgan Suite and upperpart of Karboliia Beds, as well as in the subsequent Runkashiv Suite(Matoshko et al., 2004).

Reversed geomagnetic polarities have been determined for most ofthe Karboliia deposits, normal polarities being found only in the lowerpart and in the uppermost part of the succession (Hubka et al., 1983).The former are interpreted as representing the Cochiti subchron of theGilbert chron and the latter as the Gauss chron (Fig. 2).

Basedon itsmolluscan fauna, theLowerPorat alluviumwasconsideredby Tchepalyga (1967) to be an analogue of the upper part of the KarboliiaBeds. According to Gozhik and Chirka (1973) the Near-Danube Series ischaracterized by a mollusc faunawith Prosodacna rumana, Viviparus ovidii,

Fig. 7. (a) Cross-section through the buried Kuialnyk Series in central-southern Ukraine, between the town of Kahovka and Sivash bay of the Sea of Azov (profile 4 in Fig. 1). Marinedeposits are unornamented. Deposits assigned to the Kuialnyk Series are shaded; note their southward dip and southward thickening. Ages of rocks are abbreviated thus: N1p,Pliocene, Pontian; N2ku, Pliocene; Kuialnyk Series; Pl1, Lower Pleistocene; Pl2–3, Middle–Upper Pleistocene; Pl2–3-H, Lower–Middle Pleistocene to Holocene. (b) Locations ofboreholes used to compile part (a).


and Viviparus neumayri. The similarity in stratigraphy suggests that theKarboliia and Near-Danube Series are equivalent deposits.

The Kuchurgan Suite fluvial deposits also contain fossils of theMoldova faunal complex (Matoshko et al., 2004, see above). The smallmammals from the Kuchurgan Suite, which are younger than thosefrom the lower-middle Karboliia Beds, are placed in the LateKimmerian (Shevchenko, 1965). This evidence, the presence of thesame Upper Karboliia unionids (see above), and a geomagneticpolarity reversal (thought to be the Gilbert–Gauss; Hubka et al., 1983),allows correlation of the Kuchurgan Suite with the upper part of theKarboliia Series (Gozhik, 1992).

Korotkevich (1988) identified the ‘Kuchurgan faunal complex’,thought to pre-date the Moldova complex, in the Kuchurgan deposits.Vangengeim et al. (1995) dated the Kuchurgan gravel to the late EarlyKimmerian and Middle Kimmerian (i.e., the mid Gilbert chron) andthe Karboliia Beds to the late Kimmerian, using geomagnetism andsmall mammal data. These points of view do not take into account thelower hypsometric position of the Kuchurgan Suite in comparisonwith the Stolnichen deposits and its higher disposition comparedwiththe Runkashiv Suite (Bukatchuk et al., 1983; Matoshko et al., 2004).Another problem is the apparently mixed composition of faunalremains in the Kuchurgan deposits, which raises questions about theirinterpretation. From the same geological and faunal evidence, Bilinkis(1992) considered the Nizhneporat (in our definition, the KarboliiaSeries), Stolnichen and Kuchurgan deposits as equivalent (middlePliocene) deposits, uplifted to different altitudinal positions bydifferential post-depositional vertical movements. In the presentauthors' opinion this view oversimplifies the fluvial development

spanning ~2 Ma and overestimates the amplitude of differentialcrustal movements.

4. The Late Pliocene

There is abundant evidence of the Upper Pliocene (Akchagyl Stage)fluvial deposits of the Kuialnyk, Dniester Liman and Upper Poratseries, which form a buried alluvial plain along the north coast of theSea of Azov and Black Sea (Fig. 1). The normal geomagnetic polaritiesin the Early Kuialnyk fluvial deposits at sites on the Sea of Azov coast(from oriented borehole cores) are referred to the Gauss chron and thereversed polarities of the Kuialnyk deposits of the Odessa region arecorrelated with the beginning of the Matuyama chron (Semenenko,1987). Several subdivisions of these deposits can be distinguishedacross the study region, providing a basis for regional correlation.

These Upper Pliocene fluvial deposits typically overlie the lime-stones, clays and near-shore sands of the Upper Pontian NovorossiiskBeds. Red-brown clays characterize the adjacent uplands, farthernorth, and serve as an important index ‘cover bed’ for the Pliocene–Lower Pleistocene of these regions (Matoshko et al., 2004); however,these deposits are scarce in this coastal area. There are uncertaintiesabout the detailed distinction of the various Late Pliocene terraces,depending on methodology (classification according to the number ofburied soils within overburden deposits, as opposed to conventionalterrace stratigraphy; cf. Matoshko et al., 2004).

Two fluvial suites of this age occur in the eastern part of the Sea ofAzov coastal plain and in the Dnieper–Molochna interfluve. The LowerKuialnyk Suite occurs in relatively narrow incisions into Kimmerian

Fig. 8. Cross-section through the buried Pliocene–Pleistocene fluvial incisions along the Black Sea coast near the Dniester Liman (profile 5 in Fig. 1), modified after Shnyukov et al.(1984). Marine deposits are unornamented. Ages of rocks are abbreviated thus: N1, Miocene; N1s, Sarmatian; N1m, Meotian; N1, Miocene; N1p, Pontian; N2, Pliocene; N2dl1, DniesterLiman Series (lower member); N2dl2, Dniester Liman Series (upper member); N2-Pl1, Pliocene–Lower Pleistocene; Pl1-H, Lower Pleistocene–Holocene; Pl1ro, Roksolany Suite; Pl3,Upper Pleistocene, Pl3och-ant, Ochakiv–Ant suites (undivided); Pl-H, Pleistocene–Holocene; Pl3neu, New Euxinian horizon; Hch, Holocene, Chornomor Horizon. Maximum incisionpreceded deposition of the deposits named in the present study as the Dniester Liman Series, during the Akchagyl stage of the Pliocene. Deposits of the Roksolany Suite are shaded toemphasise the extreme (~50 km) width of the Dniester Valley in this locality during the mid Early Pleistocene.


marine rocks within the Molochna valley and its tributaries (Matsui,1982) and is also present around the delta of the modern RiverDnieper (Figs. 1 and 6; see also below). The more widespread UpperKuialnyk Suite (Fig. 9) consists of channel and overbank deposits(total thickness 2–14 m); between the mouths of the rivers Don andMolochna it is almost always found overlying marine deposits ofKuialnyk age and covered by 20–30 m thick terrestrial deposits, withLower Pleistocene fluvial deposits overlying it around river mouths. Inthe eastern part of the area, near the Ukrainian Shield, the UpperKuialnyk Suite has a characteristic basal layer of crystalline rockdebris, often cemented by carbonates to form sandstones andconglomerates. Quartz sands are predominant along the coast ofTaganrog Bay, with polymict and arkosic sands typical farther west.The overbank facies consists of hard grey plastic clays, with carbonateand ferruginous/manganese concretions, in places passing upwards tosilts with sand interbeds (Matsui et al., 1981).

Between the Molochna valley and the Dnieper this Series ofdeposits (without resolution into suites) has a simple disposition(Fig. 7). At all localities except the ancient Dnieper delta, the KuialnykSeries overlies Pontian limestones. At the northern boundary of thisarea (~45–50m above present-day sea-level) the base of the KuialnykSeries lies at 12–15 m a.s.l., sloping southwards to 35–40 m b.s.l.

beneath the modern shoreline, as the thickness increases from a fewmetres to 40–45 m; the average thickness of the deposit is 15–20 m.The predominant fine, well-sorted and horizontally laminatedchannel sands with rare lenses of fine gravel, overlying a basallimestone gravel, reach a typical thickness of 14–16 m. Overbank siltsand clays, which typically form b20% of the total thickness, do notexceed 4 m thickness. In some exposures sets of poorly sorted sands,with carbonate concretions, 30–70 cm thick and dipping at up to 20°,are revealed at the base of the deposit. They cover white sands withabundant marine molluscs (? Pontian), forming cemented thin platesat the contact. Close to the modern coastline, interbeds of clays appearin the channel sands; their number and thickness increase southwardand the fluvial deposits grade into marine, probably lagoonal,sediments of the Kuialnyk Stage (Semenenko, 1975).

The well-known site of Kairy, in the Lower Dnieper (e.g., Matoshkoet al., 2002) (locality 13 in Fig. 1), is situated 6 km northeast ofKahovka, beyond the northern end of the cross-section in Fig. 7. Thestratigraphy at Kairy is reportedly very similar to that in the northernpart of Fig. 7a, although the deposits at Kairy (discussed in more detailbelow) are considered to be much younger than those at Kahovka, asthey represent the late Early Pleistocene Taman mammal biozone(e.g., Rekovets, 1994). Nonetheless, it remains uncertain how many

Fig. 9. Field photograph of a section in the Upper Kuialnyk Suite near the village of Kulikovskoe (Urzuf) on the Sea of Azov coast (locality 15 in Fig. 1). See text for discussion.


distinct fluvial suites occur at Kairy (cf. Goretsky, 1970) as most of thesite is now flooded by the Kakhovka Reservoir.

Within the ancient Dnieper delta the Upper Pliocene fluvial depositshave hitherto been assigned to two fluvial suites (Zamorii, 1940;Goretsky, 1970). Goretsky resolved an older Kimmerian–Kuialnyk Suiteand a younger Kairy Suite. Work by the present authors, includingreanalysis of Goretsky's sections, indicates three Pliocene suites in thislocality, representing the Lower Kimmerian (see above), LowerKuialnykand Upper Kuialnyk (Fig. 6).

The Lower Kuialnyk Suite, inset within the Kimmerian marinedeposits, is represented by sand, replaced southwards by interbeddedsands and clays. The level of the base of the Lower Kuialnyk Suite slopestowards themodern coastline from40 to 100mb.s.l. (it is ~70m b.s.l. inFig. 6); the thickness of thedeposit shows a corresponding increase from30 m (at Kahovka) to 60 m (at the modern Black Sea coastline). In thepresent authors' view the probable analogue of this deposit farther upthe Lower Dnieper is the sand distinguished by Goretsky (1970) in twocross-sections near Kahovka (Fig. 1) at 35–45 m b.s.l. and referred byhim to the Lower Pleistocene Kairy Suite. Hitherto, other authors(including Matoshko et al., 2002) have designated other parts of thedeposit beneath the modern Dnieper Delta, now assigned as the LowerKuialnyk Suite, as the Parafiivka Series of the Early Kimmerian.

Given its extent (Fig. 1) and composition, the present interpreta-tion of the Lower Kuialnyk Suite (based on considerably more datathan those of previous authors) indicates the development of a LatePliocene Dnieper Delta of comparable dimensions to that whichsubsequently developed during the Middle Pleistocene (see below).The Lower Kuialnyk Suite is overlain by the Upper Kuialnyk Suite(Fig. 6), which correlates with the suites of the same age in theMolochna–Dnieper and South Bukh–Dniester interfluves. Fig. 6 showsa modified interpretation, for comparisonwith Matoshko et al. (2002,their Fig. 5); much of what is now regarded as the Late PlioceneKuialnyk Suite was formerly classified either as the older ParafiivkaSeries or the younger Kairy Series.

The deposits of the Upper Kuialnyk Suite in the South Bukh–Tiligulinterfluve (Figs. 1 and 4) extend much farther inland from the BlackSea than elsewhere (it probably persists beyond the inferred limits tothe northeast and east, illustrated in Fig. 1). Its base is highest (110–120 m a.s.l.) adjoining the upper reaches of the Kuialnyk and Tiligulrivers. From this area it gradually slopes southeastward to between+5m and−5 m a.s.l. at the Black Sea coastline, as illustrated in Fig. 4.This fluvial deposit overlies Miocene and Lower Pliocene marinesediments and is overlain by 6–10 m of silt and loess. Its compositionis spatially variable; overbank clays and silts pass along strike intosandy channel facies deposits. The archives of the Geological Surveyshow this fluvial deposit to be resolvable into two or three suites thatcorresponded to fluvial terraces, although no evidence in support ofthis divisionwas found in the present study. Instead the correlation ofborehole data demonstrates that it is a single deposit sloping towardsthe Black Sea (see Fig. 4). These conclusions from the borehole dataare supported by observations of key outcrop sections. Thus, theheterogeneous overbank facies is represented by interstratified thinlaminated sands and silts (e.g., at Polovinka and Parutino; localities 11and 12 in Fig. 1) and massive clays (e.g., at Cape Karabush; locality 10in Fig. 1) with irregular contacts. The channel sands are poorly sorted,cross- or horizontally-laminated, in places gravelly with lenses of siltand clay and with a distinct coarser basal horizon.

Thick fluvial sediments are also reported in boreholes around themouth of the River Dniester (Fig. 1). Here designated as the DniesterLiman Series (Fig. 2), these deposits (up to 65 m thick) include twomembers (Fig. 8), with bases at 58–73m and 48m b.s.l. Bothmembersare represented by fine–medium gravelly sand of channel faciesoverlain by overbank facies clay. The Dniester Liman Series is inset intoMiocene marine rocks and overlain by Pleistocene deposits of diverseage and origin. On the basis of altitude and sedimentary facies thesemembers are here correlated with the Upper Kuialnyk Suite.

In localities farther west, similar deposits with similar dispositions(Fig. 1) can also be presumed contemporaneous. Their base declines


southeastward from 180–200 m a.s.l. to sea level, whereas farthersouth (in the extreme SW part of Ukraine; Fig. 1) it declinessouthward and southwestward from 80–100 m a.s.l. to 70 m b.s.l.This disposition is documented by a cross-section (Palatnaya, 1991)through the Sasyk Liman (locality 5 in Fig. 1), where the base of thedeposit is 52–55m b.s.l., inset into Meotian clay. The lower member inthe cited cross-section consists of quartz sand with clasts of igneousrocks and interbeds of silt and clay contain shells of freshwater mol-luscs, including Viviparus. sp., Unio sp., Fagotia acicularis (Ferussac),F. esperi (Ferussac), Valvata piscinalis (Müller), Theodoxus sp., andCaspia sp. (Palatnaya, 1991). This fauna confirms correlation with theKuialnyk Suite elsewhere. The overlying upper fluvial member (locally27–46 m b.s.l.) can thus also be presumed part of the Kuialnyk Suite,although Palatnaya (1991) regarded it as Pliocene–Lower Pleistocene.

The Upper Porat Series (Upper Pliocene) in part overlies the Near-Danube Series, cutting out the latter deposit (its gradient is steeperthan that of the Near-Danube Series), and in part overlies Miocenemarine deposits. Near Reni (Figs.1 and 5) the Upper Porat Series formstwo ‘shelves’ in the bedrock, at 31–54 and 85–90 m b.s.l., withcorresponding fluvial suites. In many geological profiles the uppersuite can be followed northward, beyond the limits of ancient valley,where its base rises to 10 m b.s.l., and eastward, to the coast, where itstop is at 74 m b.s.l. (Nikiforov and Diyakonu, 1963). The lower suiteconsists mainly of poorly sorted sand with inclusions of coarsematerial, the overbank facies being absent. A basal layer (gravel andgravelly sand) and laterally variable lithology (sand, clay and silt)characterize the upper suite. According to Bilinkis (1992) most of thecoarse material consists of clayey-carbonate balls and other poorlysorted local rock debris, as well as characteristic far-transportedjaspers. Beyond the limans and Danube Delta, the Upper Porat Series isoverlain by Pleistocene–Holocene silt and loess; within the limans, theDanube flood plain and the Danube Delta, it is overlain by Middle–Upper Pleistocene and Holocene marine and fluvial deposits. In manyplaces the base of the reddish Upper Pliocene–Lower Pleistoceneresidual clays (‘waste mantle’) lies directly on the Upper Porat Seriesdeposits (Hubka, 1969; Chirka, 1974; Bilinkis, 1992).

The Kuialnyk and Upper Porat Series deposits contain largemammalfossils of the Khapry faunal complex (Gromov, 1948; Konstantinova,1967; Lebedeva, 1972). A species list (including Archidiskodon. gromoviGarutt and Alexeeva, A. meridionalis meridionalis (Nesti), Equus robustusPоmel, D. etruscus (Falconer) and Elasmotherium caucasicum Воrissiak)for the site at Khapry (locality 17 in Fig.1) and for the Rashkiv Suite (i.e.,for the deposits of the Rashkiv terrace) of the Middle Dniester wasprovided by Matoshko et al. (2004). Large mammal and rodent faunashave also been described by Matsui and Moskina (1976) at sites alongthe Sea of Azov coast (i.e., Obitochnaya, Urzuf, Shirokino, and Khapry;localities 14–17 in Fig. 1). Voles (predominantly cementless forms,including Villanyia petenyii Mehely, rarely the rhizodont cementedMimomys cf. reidi) are themost common element (Matsui andMoskina,1976). The presence of primitive cementless voles of the generaDolomysand Pliomys, aswell as thebeaver (Stenefiber), is also characteristic of theKuialnyk deposits (Matsui andMoskina,1976). Conversely, the presenceof Mimomys is characteristic of the Upper Porat Series (Konstantinova,1967), enabling it to be distinguished from the Kuialnyk Series (atKuialnyk and Kryzhanivka; localities 8 and 9 in Fig. 1) (Shevchenko,1982).

Unionid molluscs (Rytia bielzi. Czekelius and P. lenticularis Sabba)and limnic Cardiidae have been collected in the Upper Porat Series(Semenenko, 1987). Bogatschevia tamanensis (Eberzin) occurs atKhapry (Rodzyanko, 1984) and also in the Rashkiv Suite of theDniester (Matoshko et al., 2004), but it is not found in youngerdeposits and thus probably became extinct at the end of the Pliocene.The characteristic unionids of the Upper Porat Series (including Cu-neopsidea doljensis, R. bielzi, Crassiana procumbens, and C. davilai) alsooccur in the low-salinity facies of the marine Late Kuialnyk of the Seaof Azov coast (Moliavko, 1960; Semenenko, 1987).

5. The Early Pleistocene

Most of the Pleistocene part of the fluvial archive in the studyregion is deeply buried beneath subaerial and marine—estuarinecover and thus relatively poorly documented. Nevertheless fragmentsof fluvial bodies are referred to the different stages of the Pleistocene.These deposits occur in three main regions: on the left bank of theLower Danube, near the mouth of the Dniester, and near the mouth ofthe Dnieper. The proposed correlation of these Pleistocene fluvialsuites is summarized in Fig. 3.

Lower Pleistocene fluvial deposits, overlying the Upper KuialnykSuite and with total thickness increasing westward to 33 m, are re-cognized in the vicinity of the Dnieper Liman (Fig. 6). They consist ofthree units of fine channel sands, the upper two separated by over-bank or subaerial reddish clays, capped by Middle Pleistocene toHolocene aeolian sands and loess. The fluvial deposits are correlatedwith the Kairy Suite of the Lower Dnieper (Matoshko et al., 2002). TheKairy (Zapadnye Kairy) site, 70 km north-east of the modern Dnieperdelta (locality 13 in Fig. 1), is characterized by the Taman mammalfauna, including Elephas (Mammuthus) meridionalis, Equus stenonis,Spermorphilus nogaici, Sicista vinogradovi, Cricetus nannnus, Villanyiafeirvaryi, Lagurodon arankae, Prolagurus pannonicus, Mimomys inter-medius, and Allophaiomys pliocaenicus (Rekovets, 1994).

In the vicinity of the Dniester Liman the Dniester terrace depositsconvergewith themodern valley floor (Matoshko et al., 2004). Probablythe oldest of the Pleistocene deposits in the vicinity of the moderncoastline (designated in this paper as the Roksolany Suite) is exposednear themouthof theDniester atRoksolany (locality 7 in Fig.1) andfixedstratigraphically by boreholes near the Black Sea coast (Fig. 8). This suite(thickness up to 22m) is represented by channel sandswith ‘Carpathian’jaspers andoverbank silty sands, sandy loamsand clays. AtRoksolany thetop of the fluvial deposits occurs 3 m above the level of the DniesterLiman (VeklichandSirenko,1972); it is overlainby2.4mof clayand loam(reddish ‘waste mantle’) and by loess with palaeosols (total thickness33–36 m). The Roksolany Suite overlies the marine Pontian rocks andfluvial deposits of the Lower Dniester Suite and is inset against thickreddish ‘wastemantle’ residual clays (Fig. 8). Remains of largemammalssuch as the elephant Archidiskodonmeridionalis tamanensisDubrovo andthe rhinoceros Dicerorhinus etruscus (Falconer) are found in this fluvialdeposit (Veklich and Sirenko, 1976). It also contains small mammals,including Mimomys ex gr. reidi-pusillus, Lagurodon arankae, Prolaguruspannonicus, and Allophaiomys pliocaenicuswith enamel of more evolvedstructure than at the Khadzhimus site in the Lower Dniester (Mihailescuand Markova, 1992). Khadzhimus, within the Boshernitsa Suite of theDniester, was thought by Matoshko et al. (2004) to be early EarlyPleistocene.

The fauna at Roksolany is assigned to the Nogaisk (Taman) complexand correlated with that from the site of Zapadnye Kairy (Lower Dnieper,theKairy Suite in this paper; locality 13 in Fig.1) (Markova,1998). Shells ofthe freshwater molluscs Bogatschevia colorata Bogatschev, B. scutumBogatschev, Potomida sublitoralis Tchepalyga and Crassiana crassoidesTchepalyga are known from the Roksolany Suite (Mihailescu andMarkova, 1992). Dodonov et al. (2005) recognized the normal-polarityJaramillo subchron in the basal part of the Roksolany loess-palaeosolsequence (i.e., in the upper fluvial deposits and lower reddish clays andloams). The Roksolany Suite can thus be placed in the Early Pleistocene.However, it has previously been correlated with the Khadzhimus orBoshernitsa Suite (Chirka,1974),with theMihailovkaSuite (Bilinkis,1992)(contemporaneous with the Matuyama–Brunhes geomagnetic reversal,according to Matoshko et al., 2004), and with the late Early Pleistocene(Markova, 1998). The second and third of these interpretations conflictwith the palaeomagnetic data, cited above, that is now available.

The Roksolany Suite is thus now placed in the mid EarlyPleistocene; it is younger than the Boshernitsa Suites and older thanthe Kairy Suite (Fig. 3). The disposition of the Roksolany Suiteindicates that the width of the Early Pleistocene Dniester valley near


the modern coastline was ~50 km, much wider than the modernvalley in this area (Fig. 8). The presence of thick clays and muddysands in the upper part of the suite indicates proximity of the ancientriver mouth.

6. The Middle Pleistocene

The Middle Pleistocene river terraces, analogous to those in thelower and middle reaches of the valleys, occur in localities adjoiningthe northwest coast of the Black Sea, notably in the region betweenthemouths of the rivers Prut and Dniester. The reconstruction of theselowest-reach terraces has developed over a lengthy period. Descrip-tions and schematic maps by Konstantinova (1967), Chirka (1974),and Mihailescu and Markova (1992) indicate a staircase of manyPleistocene terraces in a wide coastal region, it being suggested thatmost of this expanse of terraces was formed by the rivers Danube, Prutand Dniester. These ideas were based on traditional concepts of fluvialgeomorphology and the study of field exposures, although thenumber of localities studied in the field seems to have been ratherlimited; for instance, almost all studied sections occur alongside thelimans, there being no geological profiles crossing the interfluvesbetween the limans.

Reinterpretation here of the data in these publications, combinedwith analysis of the landscape and borehole data, indicates, on thecontrary, that the landscape consists of a single surface representing aformer coastal plain that is gently inclined towards the moderncoastline, declining from 150–180 m to 20 m a.s.l. It is dissected bydeep gulleys (known locally as ‘balkas’) and river valleys, which flowinto the enclosed limans. There are no breaks of slope or flats asimplied by the idealized stepped profiles drawn previously (e.g.,Mihailescu and Markova, 1992). In some of these schematic profileswhat are here regarded as different parts of the same deposit at thesame altitude are assigned to different ‘terraces’ (up to three on theprofile to the east of the Kakhul liman; locality 1 in Fig.1). True ‘terracedeposits’ are buried under loess, which can reach a thickness of 30 m.Only some of the deposits assigned to the ‘4th–7th terraces’ byMihailescu and Markova (1992) are fluvial or deltaic, even accordingto their own interpretation; the others are lacustrine, estuarine ormarine. According to Mihailescu and Markova (1992) these Pleisto-cene deposits overlie, with erosional contact, the Upper Pliocenefluvial deposits (the Kuialnyk Series of this paper). They, as well asPleistocene estuarine and lacustrine deposits, are absent within theinterfluves of between the Danube tributaries (Fig. 5).

The fluvial deposits, which are 0.5–7 m. thick, consist of channelfacies (fine and medium horizontally- and cross-bedded sands) andbasal facies (coarse sand, gravelly sand, and gravel), with manyfreshwater mollusc shells (Mihailescu and Markova, 1992). Thedeposits are capped by estuarine and lacustrine deposits with shellsof molluscs indicative of stagnant brackish water or freshwater(Mihailescu and Markova, 1992). The gravel is formed mainly of clastsof local Neogene marine rocks (Bilinkis, 1992). Fluvial gravel ofCarpathian provenance has been noted only in sections north of Reni(in deposits of the Prut). These deposits also differ from those of themodern Danube and its delta, which consist of fine and very fine sandsand silts (Nikiforov and Diyakonu, 1963).

In contrast with the many terraces of previous workers, just threedistinct fluvial deposits are here recognized in the studied part ofwesternmost Ukraine, east of Reni, each occuring in a narrow zonealong the flanks of left-bank Danube tributaries. These deposits aretermed the Nagorne, Suvorovo and Ozerne suites, each based on a typelocality of the same name. Data from Mihailescu and Markova (1992)indicate that sections in the deposits that are now called the Nagorneand Suvorovo suites are very similar. Both have subhorizontal or verygentle inclination of the basal surface, steepening 8–10 km beyond theseaward edge of the limans; both rest against subaerial strata includingPliocene–early Early Pleistocene reddish residual deposits. However,

the Nagorne Suite ismuch thicker and its base is 4–7mhigher (5–10mabove the level of the limans in their southern parts).

The fluvial deposits of the Nagorne Suite (originally described asliman sandswith lenses of gravel at the base;Mihailescu andMarkova,1992) are characterized by the Early Tiraspol small mammal fauna,including M. savini, M. pusillus, P. episcopalis, Microtus (Allophaiomys)pliocaenicus, M. (Terricola) arvalidens. In the estuarine facies M. (A.)pliocaenicus is not found. The fauna of these facies is considered asadvanced Tiraspol fauna. Bone fragments of Archidiskodon cf. wusti(Pavlov), Equus sp. and Cervidae gen. are also found in these fluvialdeposits (Mihailescu and Markova, 1992). Molluscs in the overlyingestuarine deposits include Tschaudia tschaudae Andrusov and Didacnapseudocrassa Pavlov (Mihailescu and Markova, 1992). In contrast, theestuarine beds that overlie the fluvial deposits of the Suvorovo Suiteare characterized by a Late Tiraspol fauna. This is distinguished fromthe fauna in the Nagorne Suite by the predominance of Lagurustransiens, the diversity of the Microtus genus, and the appearance ofthe stable morphological type of M. (Stenocranius) gregalis (Mihai-lescu and Markova, 1992). The fluvial deposits of the Suvorovo Suiteand the overlying estuarine deposits also contain shells of molluscs,including Viviparus tiraspolitanus Pavlov, Didacna baericrassa Pavlov,D. portatica Mihailescu, D. tschepalygae Mihailescu, Dreissena tschau-dae Andrusov, Dr. rostriformis (Deshayes) and Corbicula fluminalisMüller (Mihailescu and Markova, 1992). Overall, this biostratigraphicevidence indicates an early Middle Pleistocene age for the NagorneSuite and a mid-Middle Pleistocene age for the Suvorovo Suite(Mihailescu and Markova, 1992; Fig. 3).

In the Danube Delta area, the base of the Ozerne Suite lies belowthe modern liman level. This suite is represented by two thin fluvialdeposits separated by thick estuarine and lacustrine deposits with amolluscan fauna indicative of the Palaeo-Euxinian stage (Mihailescuand Markova, 1992). The small mammal fauna in the first estuarinedeposit overlying the Ozerne Suite belongs to the Singil complex. Thiscomplex lacks ancient arhizodont voles of the genera Mimomys andBorsodia, as well as the voles Microtus (Stenocranius) gregaloidesand M. (Terricola) arvalidens, which are replaced by M. (S.) gregalisand M. (M.) arvalis. Yellow steppe lemmings are represented by thesubspecies Eolagurus luteus volgensis. The vole M. agrestis is alsopresent in this fauna, which is characterized by the predominance ofArvicola cantiana. The Ozerne Suite is thus mid-Middle Pleistoceneand younger than the Suvorovo Suite (Fig. 3).

The Euxinian Suite of the Dnieper–Bukh Liman area (Figs. 1, 3and 6) is regarded here as a possible analogue of the Ozerne Suite. TheEuxinian Suite (base at 30–50 m b.s.l.) overlies Kimmerian marinedeposits and is inset against Upper Kuialnyk Suite fluvial deposits,represented by fine uniform sands 25–30m thick (Fig. 6). According toGoretsky (1970), it consists of two members, the thermophilousmolluscs Fagotia esper (Ferussac) and C. fluminalis being found in theupper member. This suite is overlain by Uzunlar–Karangat marinedeposits within the liman and by Periglacial Suite fluvial deposits onthe left bank of the liman (Fig. 6).

The Periglacial Suite (up to 35m thick) consists of fine sands (withinclusions of loam and buriedmarsh soils), overlain by silts up to 11mthick. These sands lack any basal horizon and any paleontologicalremains. The irregular surface of the Periglacial Suite varies from 25to 48 m a.s.l. in the left bank of the Dnieper Liman; it is overlain byUpper Pleistocene and Holocene aeolian deposits. The PeriglacialSuite is correlated with the Dnieper Glaciation, thought to equatewith the Early Saalian glaciation of western and central Europe.According to the interpretation of Goretsky (1970), in some bore-holes the Periglacial Suite overlies marine deposits with Uzunlar–Karangat and Karangat molluscan faunas. All these boreholes arelocated within lowermost marine terrace (2–5 m a.s.l.), which in thepresent authors' opinion is much younger than the fluvial terraceof the Periglacial Suite. Thus they cannot include Periglacial Suitealluvial components.


AMiddle Pleistocene fluvial deposit younger than the Periglacial Suiteis also recognized in theeasternpartof theDnieper-moutharea. This is theAleshki Suite (Fig. 6), which can be traced from Aleshki Sands (a modernarea of aeolian activity) to the south as a narrow (17–20 km) strip,widening near the modern coastline; its subhorizontal surface is at 15–20 m a.s.l. and its base near modern sea level. This deposit consists ofchannel sands, cemented to sandstones at the base, and overbank silts.

7. The Late Pleistocene

Two buried fluvial deposits (their bases respectively at 14–15 and22–26 m b.s.l.) have been identified in cross-sections across theDniester Liman (Gozhik, 1982); according to that author, both are LatePleistocene. However there is also a deeper incision that reaches 28 mb.s.l. in the vicinity of the Dniester mouth and 45–48 b.s.l. at the coast(Fig. 8), which is infilled by fluvial deposits of the Ant and (in somelocalities) Ochakiv suites (Gozhik and Novoselsky, 1989). Theseincisions are recognized in most limans and in many offshoreboreholes opposite the limans and mouths of rivers and balkas(Gozhik, 1982) between the Dnieper and Danube. In most localitiesthese deposits are overlain bymarine-estuarine deposits from theNewEuxinian and Chenomor stages of the latest Pleistocene and Holocene.

The infills of these incisions consist of basal, channnel and overbankfacies (Gozhik, 1982). The basal deposits consist of sandy gravel,including boulders of igneous rocks. The channel sands, with differentgranulometric composition, are overlain by sandy loams, loams, claysenriched byplant detritus, andpeats. Organic remains in these channelsands are scarce, there being a notable absence of any thermophilousfreshwater fauna. Shells from the channel deposits have yieldedradiocarbon ages of 24–16 ka and the overbank deposits date from 17–13 ka (Gozhik, 1982), whereas the overlying New Euxinian marinemud dates from 13.1–9.8 ka according to Semenenko and Kovaliukh(1973) or 12.7–11.2 ka according toGozhik andNovoselsky (1989). Thedeepest Pleistocene incisions in the region thus date from the lastglaciation.

8. Palaeoenvironments

The processes of fluvial erosion and accumulation within the Sea ofAzov and Black Sea coastal plains developed in response to changes incrustal deformation (‘neotectonic activity’), sea level and climate. Theinformation from different independent sources summarized below hasallowed the importanceof eachof these influencesonfluvialdevelopmentto be determined and has thus led to the revised interpretations of thefluvial data.

8.1. Crustal deformation

The area under review is in a marginal southern position withinthe East European Platform (Kruglov and Tsypko, 1988), whichincludes part of the Ukrainian Shield and the Scythian Platform. TheEast European Platform adjoins structural regions transitional to theAlpine folded zone, including the Carpathian Foreland and Alpidestructures, as well as the Black Sea basin, which is infilled byMesozoicand Cenozoic sediments. The basement rocks of the East EuropeanPlatform decline southwards to 1500–1900 m depth at the northerncoastline of the Black Sea. According to Palienko (1992), subsidenceprevailed in this area between the Late Oligocene and the MiddlePliocene, interrupted by uplift events. Since the Middle Pliocene thisarea has been experiencing predominant uplift.

Mainly terrigenous Miocene–Pliocene–Pleistocene sediments,with numerous hiatuses, form a monocline sloping towards thecentral part of the Black Sea Depression (Gozhik et al., 1999). ThePleistocene–Holocene strata on the Ukrainian part of the shelf formsuperimposed unconformity-separated wedges and display offlapwith seaward thickening (Ryan et al., 2003). These facts confirm the

conclusion of Malovitsky et al. (1975) that the Black Sea Depressionhas experienced general continued downwarping during the Plio-cene–Pleistocene. The presence on the shelf of loess deposits withburied soils of Late Pleistocene age, together with instrumental dataindicating sea level rise for the past 150 years (Zelinskii, 1993), duringwhich period there has been global sea-level stability, indicates thatthe process continues now. The downwarping involved the folding ofPontian and older rocks in the Pliocene as well as vertical blockmovements along Late Miocene faults (Zelinskii, 1993) continuing inthe Pliocene–Pleistocene (Palienko, 1992; Bilinkis, 1992). Data fromthe fluvial archive is eminently suitable for supplementing therelatively meagre direct evidence for crustal movements in the LateCenozoic (see Section 9).

8.2. Oscillations of sea level

After the Late Sarmatian transgression a general regressional trendprevailed in the studied part of the Paratethys basin. Its interruptions bythe Early Meotian, Early Pontian, Early Kimmerian and Early Kuialnyktransgressions demonstrate the successive reduction of the basin(compare the positions of coastlines in Fig. 1). The latest two of thesetransgressions were associated with coastlines close to those of themodern Sea of Azov and Black Sea. There are various theories about thenumber and scale of the Pleistocene sea-level oscillations. According toMihailescu and Markova (1992), limans similar to the modern analoguesappeared in the Early Pleistocene and existed with interruptions duringsubsequent regressive phases. The estuarine deposits of this age areuplifted to 33 or more m a.s.l. in the limans near the Lower Danube(Mihailescu and Markova, 1992). Comparable deposits of MiddlePleistocene age (Palaeo-Euxinian) occur at 16–21 m a.s.l. maximum,whereas the Upper Pleistocene Uzunlar and Karangat deposits occur at10–25 m a.s.l. maximum. Rare fragments of the low Karangat marineterraces are represented on the Crimean Peninsula (Fedorov, 1978).Between Reni and Kiliia, two young terraces (Late Pleistocene–Holocene),represented by marine and estuarine deposits, are expressed in thevicinity of the Danube channel or itsmodern delta, aswell as on the limancoasts in the altitudinal range 7–13 m a.s.l. The New Euxiniantransgression penetrated into the lower reaches of the rivers up to severaltens of kilometres from their modern mouths, reaching 17 m a.s.l. in theDnieper valley (Gozhik, 1982; Matoshko et al., 2004).

The ‘regressive’ part of the paleoenvironmental record is less cleardue to the problems of Paratethys stratigraphy (see Section 1), as wellas the lack of direct age determinations in deposits on the shelf and onthe slope of the Black Sea Depression.

Drilling at DSDP sites 380 and 381 in the central Black Sea revealedterrigenous deposits and thin evaporitic carbonates, placed at theMiocene–Pliocene boundary (Ross, 1978). This evidence led to theidea that the Black Sea basin became desiccated at the end of thePontian, possibly synchronous with the Late Messinian salinity crisisin the Mediterranean (e.g., Hsü and Giavanoli, 1979). The validity ofthis idea has continued to be debated; for instance, according to Jonesand Simmons (1997) these evaporites (Unit IYd at DSDP Site 380) arein part Pontian and in part Kimmerian and the major regressive phasewas thus during the Kimmerian (which started at 5.5 Ma in theirchronology). However, the more recent BLaSON survey (Gillet et al.,2007) concluded that the wide intra-Pontian erosional surfacerevealed in the Romanian shelf of the Black Sea can be traced to thetop of the Upper Miocene shallow-water deposits at DSDP site 381 inthe central Black Sea, consistent with the view that the Black Sea basinbecame desiccated during the Pontian.

The fall in sea level during this Pontian desiccation has been correlatedwith the development of an incised drainage network in the MiddleDnieper,whichwas infilledby thefluvial ParafiivkaSeries, and thedeepestincision in the Lower Dnieper (Matoshko et al., 2002). However, thepresent analysis has concluded that the deposits hitherto regarded as theParafiivka Seris are in fact part of the Tendra Suite (Lower Pliocene, Lower


Kimmerian) and part of the Kuialnyk Series (i.e., they are Pliocene, seeabove). Fig. 6 shows that the Tendra Suite infills the incision in themarineKimmerian sediments, which indicates that it was formed after theKimmerian transgression. There are many outcrops of the Upper Pontianmarine Novorossiisk beds in coastal cliffs of the Black Sea and its limans,with nowashouts or unconformities indicating a contemporaneous fall inwater level. These facts indicate that no significant fall in sea-leveloccurred during or after deposition of the Novorossiisk beds until the endof the Kimmerian transgression. This does not contradict the theory ofdesiccation of the Black Sea corresponding with the Messinian salinitycrisis in theMediterranian. Presumably, the regionalmoisturedeficit led toa simultaneously lowered Black Sea or even to the termination of thefluvial processes in the surrounding landscape.

The fluvial deposits within the Black Sea shelf and the Sea of Azovfloor (providing traces of the regressive phases) are revealed by arelatively sparse network of cores, grab samples and shallow boreholes(Scherbakov et al., 1976, 1978; Semenenko and Sidenko, 1979; Ryanet al., 1997; Gozhik et al., 2006). A network of buried valleys (reachingas low as 85 m b.s.l.), thought to date from the late Late Pleistocene,was thus established throughout the northwestern Black Sea shelf outto its outer margin (Fig. 1). Up to five submerged marine terraces havebeen reported in different parts of the Black Sea shelf at depthsbetween20 and121mb.s.l. (Shimkus et al.,1980). As Ryan et al. (2003)noted, shells from these terraces, indicative of past littoral environ-ments, date from 19–9 ka. There is no reliable evidence for earlierPleistocene regressions,which some researchers havehypothesized onthe basis of palaeohydrological considerations.

TheDanube Fan (e.g., Kazantsevand Shainurov,1978; YevsyukovandKara,1990) is located beyond the shelf within the slope of the Black SeaDepression (~150 km southeast of the present Danube mouth, at 120–500mdepth). One view (Kazantsev and Shainurov,1978) is that this fanhas formed as a result of turbidity currents; another (Yevsyukov andKara, 1990) is that it indicates direct fluvial deposition at times ofregression of the Black Sea. Near the shelf margin several submergedpalaeoshorelines are documented between 108 m and 145 m b.s.l.(Yevsyukov and Kara, 1990). The 1998 BLaSON survey revealed thicksediment at the shelf margin west of Crimea. This sediment bodyappears to have been fed from a ~10 km wide submarine canyon, thebase of which lies deeper than 600 m b.s.l., fed by the Dnieper andDniester (Lericolais et al., 1998; Ryan et al., 2003). Regional seismicsurvey has also revealed a body of Pleistocene or Pliocene–Pleistocenesediment that is several thousandsof square kilometres in area andup to600–700 m thick, within the shelf south of the Kerch Strait (Malovitskyet al., 1975). This sediment can be presumed to have been deposited bythe palaeo-river that drained the Sea of Azov through this strait.

8.3. Palaeoclimatic variations

Information about palaeoclimatic variations in the Late Cenozoichas mainly been obtained from interpretation of palaeobotanical andpalaeozoological data. According to Schekina (1966), in the Pontianthe vegetation of southern Ukraine developed from a predominance ofsalt semi-desert and steppe to the expansion of forest and scrub.Conditions were close to subtropical on the south coast of Crimea:much warmer than at present. Such conditions persisted during theKimmerian, when pine forests (predominant subgenus Diploxylon)developed in the river valleys, where there is also evidence of a greatvariety of freshwater aquatic plants (Schekina, 1966). The contem-poraneous fauna of large mammals thus lived in subtropical, humidforests and on floodplains, during the deposition of the sediments ofthe Kuchurgan Suite (Korotkevitch, 1988).

The beginning of the Akchagyl stage marked a decrease in vegetationdiversityanda reduction in thepercentageof subtropicalplants (Schekina,1966). Later, the deciduous forests of the tugai (flood plain) landscapespread and the areaof pine forestwas reduced. By theendof theAkchagyl,halophytic semi-desert and steppe prevailed (Schekina,1966). The spores

and pollen from this time are similar to those from younger deposits thatare assigned to the Early Pleistocene (Schekina, 1966).

During the Pleistocene the climate of the study region tended to bedrier and colder than in the Pliocene, with steppe conditions generallyprevailing. Cold azonal tundra-steppe spread up to the modern coastof the Sea of Azov and the Black Sea during the continental glaciations(Turlo, 1982). At the maximum of the Dnieper Glaciation (MiddlePleistocene=MIS 8, according to Matoshko et al., 2004), theScandinavian ice sheet reached within a few hundred kilometres ofthe Black Sea, with the Dnieper valley carrying meltwater outwash(Matoshko and Chugunny, 1993). The latter situation also prevailedduring the later glaciations, when ice covered the upper reaches of theDnieper and its major tributaries. Apart from this there was no directinfluence of glaciation on the region (Matoshko and Chugunny, 1993).

Thus during the time span under consideration there was a changefrom subtropical conditions (Pontian–Kimmerian) to semi-desert andsteppe (Ackchagyl and Pleistocene–Holocene) as well as azonaltundra-steppe (periods of continental glaciation). The decrease inprecipitation since the Ackchagyl and the corresponding reduction oflocal runoff (river discharge) could be a further reflection of thischange. The presence of the red (red-brown) clays of residual soil(Late Pliocene–Early Pleistocene) is interpreted as a result of relativelywarm and humid climatic conditions and provides further evidencefor later runoff reduction. The climatic influence was different forlocal, minor rivers and the larger rivers that flow across the area fromfurther afield, locally termed ‘transit’ rivers (e.g. Dniester, Dnieper,Don). While most of the catchment areas of the Dnieper and Donexperienced moderate-continental climate during the Late Plioceneand in Pleistocene interglacial periods, the Dniester was subjected togreater variability of precipitation in the Carpathians. During theperiods of glaciation all the major rivers were at times subjected to aglacial regime with meltwater input. Caution should be exercised,however, in applying a simple correlation between warm–coldPleistocene climatic alternations and changes in fluvial activity inthis region (cf. Matoshko et al., 2004); it should be noted that, as arule, palaeoclimatic data do not include parameters of precipitationand it is that parameter that is arguably most important in theinterpretation of fluvial development.

9. Fluvial development during the Late Cenozoic

The history of fluvial sedimentation and fluvial relief developmentin the study area can be determined on the basis of the data cited inthe previous sections. It is demonstrated in terms of the position ofdepocentres, the disposition of fluvial sediments, sedimentationdynamics (cycles, phases) and landform types. The timescale hasbeen established in Sections 2–7.

9.1. The Late Miocene

The Balta Series depositional basin, which developed in the LateMiocenewithin the area of modern catchments of the South Bukh andDniester (Fig. 1), is bounded by the uplifted part of the UkrainianShield to the east and north, by the Toltry ridges (huge reefconstructions of the Late Tortonian) to the northwest and by thecontemporaneous coastline of the Paratethys Sea to the south(Zosimovich et al., 1975). Apparently, large rivers fed this basin andformed extensive channel facies sediments, but their geometry is notestablished. The depocentre developed in the highly stable Ukrainianshield (the most stable part of the East European Platform) andpersisted there for more than two million years. The development ofthis depocentre evidently brought about subsidence (under thesedimentary load) and subsequent uplift, around the end of theMiocene. This great cycle of erosion–accumulation resulted in theformation of a vast coastal alluvial plain surrounded by uplands.


Currently this area is within the Pridniprovska Upland, with nomodern expression of the ancient intra-shield depression.

9.2. The Late Pontian and Kimmerian stages

At the beginning of the Pliocene the environment changed substan-tially. After the short-term regression of the Pontian sea (this phase is notexpressed in geological sections; see above) the Early Kimmeriantransgression of the Black Sea resulted in a repeated rise in base level.This rise is associatedwith the extensive deposition of the Stolnichen andAnaniev suites (Fig. 2) and formation of the corresponding alluvial plain.The depocentre extended from the Carpathian Mountains to the modernDniester–Bukh interfluve, thus partly occupying the area of the previousBalta Series depositional basin; its distal limit of preservation is ~50 kminland of the inferred position of the contemporaneous palaeocoastline(Fig. 1). This era of stable conditions was followed by the maximal LateCenozoic incisionof lowest reach (near-mouth) river channels throughoutthe study region. This incisionphase is especiallyclear in the lower reachesof the modern Dnieper (Fig. 6) and Danube (Fig. 5). As was suggestedbefore for the Middle Dnieper area (Matoshko et al., 2004), this incisionwas characterized by extremely narrow and deep valleys with limitednumbers of tributaries. This could perhaps be explained by a paradoxicalcombination of a high velocity of incision and small levels of runoff. Suchincision could have been triggered by movements along the hinge zonebetween the subsiding Black Sea Depression and the uplifting peripheralpart of the East European Platform. It is possible that this incision into theEarly Kimmerian alluvial plain gave rise to the desiccation of this plain andthe separation the valleys of theDniester, Prut and Siret,much earlier thanwas supposed by Bilinkis (1992). Since that time themain depocentre hasmigrated to the south, within the limits of the Black Sea.

Later in the Kimmerian, according to the present authors, theseincisionswere infilled by thedeposition of theNear-Danube Series (LowerDanube; Fig. 5) and Tendra Suite (Lower Dnieper; Fig. 6); the KarboliiaSeries was also deposited, north of the Lower Danube, around the sametime. Minor further incision in the late Kimmerian provided theaccommodation space for the deposition of the Kuchurgan Suite at alevel rather lower than the previous Stolnichen Suite. From the scantytraces of the Kuchurgan Suite the reconstruction of the contemporaneousvalley is not possible.

9.3. The Akchagyl stage

Further incision is evident in the Late Pliocene, in the vicinity of themodernBlack Sea coastline, given thedispositionof theKuialnyk,DniesterLiman and Porat series that subsequently filled this incision. Thus, duringtheAkchagyl stage, incisionaround themodernmouthof theDniesterwasfollowed by aggradation of the ~60 m thick stacked succession of theDniester Liman Suite (Fig. 6). In contrast, during the same span of time inthe Middle Dniester, three inset fluvial deposits accumulated, known asthe Runkashiv, Vadul-lui-Vode and Rashkiv suites (Matoshko et al., 2004),indicating net incision by up to ~40 m. The stratigraphic relationshipbetween these Middle Dniester terrace deposits and the Dniester LimanSuite remains unclear; however, it is apparent that at this time a transitionexisted between uplift in the Middle Dniester and subsidence near themodern Dniester mouth. Matoshko et al. (2004) traced the Runkashiv,Vadul-lui-Vode andRashkiv terraces downstream into the LowerDniestertowithin ~80 km of themodern Dniester mouth; for instance, they occur~130, ~120, and ~110ma.s.l. aroundParkan (locality 18 in Fig.1), ~100 kminland. The contemporaneous ‘hinge zone’ between uplift and subsidencecan thus have been no more than a few tens of kilometres inland of themodern Dniester mouth. This hinge zone cannot be located precisely,because the Pliocene fluvial terrace deposits that can be traced fartherupstream in the Dniester (Matoshko et al., 2004) have been removed inthe vicinity of the Dniester Liman, possibly as a result of Early Pleistoceneerosion.

A dramatic change in environmental conditions marked the start ofdeposition of theUpper Kuialnyk Suite (after short epizode of stabilizationorminor incision). As Figs.1, 4, and 7 indicate, this deposit coversmuch ofthe land surface north of the modern Black Sea; it is not confined to rivervalleys. The deposit slopes southward towards the Black Sea and alsotypically thickens in this direction (as in Fig. 7), before passing southwardinto contemporaneous marine deposits. Sedimentary characteristics, inparticular the great extent of overbank facies, indicates that there wasnegligible relief at that time. Modern river valleys, notably those of theLower Dniester and Lower South Bukh, have since become entrenchedbelow this depositional level. The altitude of the Upper Kuialnyk Suitedeposits flanking the Lower Dniester matches that of the Runkashiv Suitefarther upstream, suggesting that these two deposits were synchronous.The formation of this vast coastal plain, with its thin sheet of fluvialdeposits, could reflect stable crustal conditions following the long periodofprevious subsidence. It is inferred that in the lateLatePliocene the ‘hingezone’ shifted 100–150 km to the north.

9.4. The Early Pleistocene

In the present authors' view the amount of Early Pleistocene fluvialsediment preserved in the study region is very limited; it is much lessthan for the Late Pliocene. This is largely due to the reinterpretation ofsediments previously considered to be Early Pleistocene as older, asdiscussed above. The two principal deposits in the coastal region,recognized by the present authors as Early Pleistocene, are theRoksolany Suite around the mouth of the Dniester (Fig. 8) and theKairy Series in the Lower Dnieper (Fig. 6), as already discussed.

Farther inland, Matoshko et al. (2004) recognized three EarlyPleistocene terraces of the Middle and Lower Dniester, formed by theearly Early Pleistocene Boshernitsa Suite, the mid Early PleistoceneKisnytsia Suite, and the Mihailovka Suite, which dates from near theEarly–Middle Pleistocene boundary. The disposition of these deposits,which differ in height by up to ~40 m in the Middle Dniester, ledMatoshko et al. (2004) and Bridgland and Westaway (2007) to inferan increase in uplift rate around the start of the Early Pleistocene, itsproposed cause being an increase in erosion. There might be a linkwith an acceleration of the crustal movements at this time (Bilinkis,1992), increasing the contrasting elevations either side of the ‘hingezone’. A significant increase in the relief in inland parts of the studyregion thus occurred during the Early Pleistocene, as rivers becamemore entrenched into their valleys, which were typically quitenarrow; the great width of the Dniester valley near its contempora-neous estuary, indicated in Fig. 8 from the disposition of the RoksolanySuite, is atypical. The valleys with staircases of terraces that typify themodern landscape began to form after the Early Pleistocene. Duringthe Pleistocene local fluvial activity reduced substantially under drierclimatic conditions (Zelinskii, 1993).

9.5. The Middle and Late Pleistocene

The disposition of Middle and Late Pleistocene sediments varieslaterally across the study region such that each of the major rivers hasproduced a different record. Thus, inland in the Dniester a staircase of7 terraces developed, with the latest Early Pleistocene MihailovkaSuite up to ~80 m above the modern river (Matoshko et al., 2004).This progressive fluvial entrenchmentwas associated with an increasein uplift rates in the early Middle Pleistocene, attributed to an increasein erosion rates. In contrast, around the modernmouth of the Dniester(Fig. 8) there are no Middle Pleistocene fluvial sediments.

Uplift on this timescale is also evident in the Lower Danube, giventhe disposition of the Nagorne, Suvorovo and Ozerne suites (seeabove). In contrast, instead of progressive uplift, in inland localitiesthe fluvial successions of the rivers Dnieper and Don recordalternations of incision and aggradation, or uplift and subsidence(Matoshko et al., 2004). In the Lower Don and the Don Delta, detailed


cross-sections (similar to those in Figs. 4 to 8) demonstrateprogressive fluvial aggradation from the beginning of the MiddlePleistocene to the present, interrupted by marine-estuarine transgres-sions (Goretsky, 1982). Also in contrast with the Dniester, MiddlePleistocene fluvial sediments arewidespread around the mouth of theDnieper (Fig. 6). The mid Middle Pleistocene Euxinian Suite is ~30 mbelow the Early Pleistocene Kairy Suite, suggesting uplift on thistimescale. Conversely, the late Middle Pleistocene Periglacial Suiteoverlies the Euxinian Suite, its top ~30 m above that of the olderdeposit. As in the Middle Dnieper and Middle Don areas (Goretsky,1970,1982; Matoshko et al., 2002, 2004) the unusually great thicknessof the Periglacial Suite and the high position of the correspondingterrace surface, as well as the uniform composition of the deposits, areexplained by the extremely high values of bed and suspended load inconditions of substantial discharge and seasonally undifferentiatedglacial runoff during the timewhen the Dnieper glaciation covered thecatchment area of the Dnieper River.

In contrast, during the Last Glaciation the Scandinavian ice sheetonly reached the uppermost part of the Dnieper catchment and itsdirect influence on the river was thus limited. Nevertheless at thattime the lower reaches of all the Black Sea rivers (glacially-fed,mountain, local and ‘transit’ rivers) and balkas were deepened andextended to the shelf margin and probably to the upper part of theBlack Sea continental slope. This can be linked to global decline in sea-level at that time.

It can be summarized that in the coastal area of the Black Sea andin its western part a general tendency for net incision during theMiddle and Late Pleistocene was broken only during aggradation ofthe Palaeo-Euxinian and Periglacial suites. It can be suggested thatsteady uplift with short periods of stability or subsidence occurred inthis area in the Pleistocene. Conversely, in the area of Lower Don andthe Don Delta, detailed cross-sections demonstrate that progressiveaggradation of fluvial deposits since the beginning of the MiddlePleistocene was interrupted by minor marine-estuarine transgres-sions (Goretsky, 1982).

There are no traces of glacial influence (the basin of Don River wasglaciated once or twice during the Middle Pleistocene) nor is therereliable evidence of Late Pleistocene fluvial incision within the Sea ofAzov and the Kerch Strait. Therefore the eastern part of the Sea of Azovcoast can be presumed to have experienced predominant subsidence.However, the modern Don alluvium in the delta area lies on thedeposits of the New Euxinian and Chernomor horizons (Fig. 2). In thisrespect the modern deltas of Sea of Azov and Black Sea are similar. Allof them are very young. Taking into account the rates of deltaprogradation in the last two centuries (0.6 km/yr−1 for the DnieperDelta: Kostyanitsyn, 1969; 0.7 km/yr−1 for the Kiliia subdelta of theDanube Delta: Mikhailov and Mikhailova, 1991), their formation hasoccurred over about 1100 years. Thus we can surmise that theNimphian transgression finished before 1.1 ka.

10. Conclusions

Late Cenozoic buried fluvial deposits are widespread in thenorthern coastal regions of the Black Sea and the Sea of Azov. Theyhave various modes of occurrence and lithological features, repre-senting sediment bodies of diverse origins: marginal basin deposits(Balta Series), piedmont sheets (Stolnichen and Ananiiv suites,Karboliia and Near-Danube aggradational series, Kuchurgan Suite),coastal sheets (Upper Kuialnyk and Upper Porat suites), prograda-tional deltas (Middle Pleistocene Dnieper delta, modern Danube Deltaand probably deep-sea Danube Fan), infills of palaeo-incisions,including deltaic infills (Tendra Suite, Lower Porat Suite, LowerKuialnyk Suite, Euxinian Dnieper Delta, modern Dniester and Dnieperdeltas), cut-and-fill suites of the lower reaches of the majorPleistocene valleys (Dniester, South Bukh, Ingul, Molochna andothers) and their Late Pleistocene continuations on the shelf. The

sheets are expressed in the landscape as flat or terraced plains, withincisions giving rise to terraced valleys.

There have been two main fluvial erosion–accumulation cycles inthe area: during the Late Sarmatian–Early Kimmerian and during theEarly Kimmerian–Akchagyl. The erosion phase of a third cycle beganin the Pleistocene and has been continuing to the present. Themaximum river incision occurred in the Early Kimmerian. Except fornotable events such as the desiccation of the shelf and almost theentire basin during the Messinian, fluvial processes have beenrestricted to a relatively narrow belt approximately 100 km wide,with its axis close to the modern coast.

During erosion phases the coastal plain has experienced valleydeepening and an increased intensity of horizontal and verticaldissection. Depending on crustal movements the plain could eithermaintain the same elevation or become elevated. In contrast, theaccumulative phases have been characterized by a tendencytowards low-level planation, with flattening because of burial byaccumulating sediment. Each cycle has included smaller-scaleerosion–accumulation subcycles, but only the erosional phases canbe distinctly recognized in the landscape in the form of riverterraces. The main phases have individual characteristics in differentregions of the area, some of them demonstrating mismatch withupstream areas. Thus the Lower Dniester is characterized by onecycle that occupies a lengthy interval from the Early Kimmerian tothe present (Matoshko et al., 2004), whereas near the DniesterLiman two cycles are observed. There is the same mismatchbetween the subcycles; the Middle–Late Pleistocene fluvial historiesof the areas around the Dniester and Dnieper limans are quitedifferent from those of the Don Delta.

As concluded previously (Matoshko et al., 2004), much of thefluvial record, including erosion–aggradation cycles and stages,terrace formation, and the localized accumulation of abnormallythick alluvial suites, can be linked to crustal movements. In thereviewed area a key factor was differential movement across a ‘hingezone’ separating the East European Platform and the Black SeaDepression. As a result of this the southern part of the coastal plain hasbeen relatively and progressively sinking in comparison with itsnorthern elevated part. It is considered that shifts in the position ofthis ‘hinge zone’ have determined the location of the shoreline andthat the amplitude of vertical movements essentially determinedfluvial development on the coastal plains during the Pliocene–Pleistocene–Holocene. This general ‘hinge-zone’ regime was compli-cated by differential movement of local structures.

During the considered time span the shoreline shifted by morethan 600 km and the landscape changed between shallow sea floorand lowland terrestrial plain, with commensurate disruption of riversystems. The transgression of sea water into lower fluvial reaches hasgiven rise to repeated estuarine episodes since the Early Pleistocene,although unravelling the effects of regional and local crustal move-ments from the influence of global marine transgressions is proble-matic. Only episodes of higher-level oscillation, such as the Messiniansalinity crisis in the Black Sea and the drainage of the shelf during theLast Glaciation, can be confidently linked to global climatic andhydrological events.

Concerning the role of climatic in fluvial development, the authorsadhere to the position (cf. Matoshko et al., 2004) that a directinfluence can come only from changes in precipitation. The correlationbetween the structure of fluvial sediments and landforms and thedegree of climatic moisture is clear: large rivers existed in the humidclimate of the end of the Pontian and Kimmerian, whereas small riversexisted in the drier climate of the Pleistocene–Holocene. However theexistence of widespread alluvial coastal plains in the second half of theAckchagyl contradicts the palaeobotanical indications of increasingaridity at this time (Schekina, 1966). The Pleistocene glaciations wereshort but notable events that had profound and differing effects onfluvial processes: a large Middle Pleistocene Dnieper delta formed a


nearshore upland during the Dnieper Glaciation, whereas the LastGlaciation brought about river deepening and shelf dissection.

Acknowledgements

We thank Rob Westaway and David Bridgland for considerableeditorial work on the manuscript and English improvement, MykolaiBarschevskyi for providing access to geological data and AntonMatoshko for preparing the figures.

References

Ananova, Y.N., Volkova, N.S., Zubakov, V.A., Pavlovskaya, V.I., Remizovsky, V.N., 1985.Novye dannye ob opornom razreze Taman mio-pliotsena rayona Chornogo morya[New data on Taman key section of the Mio-Pliocene Black Sea region]. DokladyAkademii Nauk SSSR 284 (4), 925–928 (in Russian).

Bala, A., Radulian, M., Popescu, E., 2003. Earthquakes distribution and their focalmechanism in correlation with the active tectonic zones of Romania. Journal ofGeodynamics 36, 129–145.

Bilinkis, G.M., 1992. Geodinamika krainego yugo-zapada Vostochno-yevropeiskoiplatformy v epochu morphogeneza [Geodynamics of the south-western part ofthe East European platform during the epoch of its morphogenesis]. Shtiintsa,Chishinau [Kishinev], Moldova (in Russian).

Bridgland, D.R., Westaway, R., 2007. Preservation patterns of Late Cenozoic fluvialdeposits and their implications: results from IGCP 449. Quaternary International189, 5–38.

Bukatchuk, P.D., Negadaev-Nikonov, K.N., 1963. Alyuvialnye pliotsenovye obrazovaniyaMoldavskoi SSR [Alluvial Pliocene deposits of the Moldaviya SSR]. IzvestiyaAcademii nauk Moldavskoi SSR. Seriya Biologicheskikh i Khimicheskikh 3, 81–85.

Bukatchuk, P.D., Gozhik, P.F., Bilinkis, G.M.,1983. Okorelyatsii aluvialnychotlozheniiDnestra,Pruta i nizhnego Dunaya [On correlation of alluvial deposits of the Dniester, Prut andLower Danube]. In: Geologiya chetvertichnych otlozhenii Moldavii [Geology of Quater-nary deposits of Moldavia]. Shtiintsa, Chishinau [Kishinev], pp. 35–70 (in Russian).

Cande, S.C., Kent, D.V., 1995. Revised calibration of the geomagnetic polarity time scale forthe Late Cretaceous and Cenozoic. Journal of Geophysical Research 100, 6093–6095.

Chirka, V.G., 1974. Terasy dolin Dnestra, Pruta i Dunaya [The terraces of the Dniester,Prut and Danube valleys]. In: Bondarchuk, V.G. (Ed.), Materialy po chetvertichnomuperiodu Ukrainy. Naukova Dumka, Kiev, pp. 279–295 (in Russian).

Chumakov, I.S., 2000. K probleme granitsy miotsena-pliotsena v evksine [On theMiocene–Pliocene boundary in Euxin]. Stratigraphija i Geologicheskaya korelyatsija8 (4), 84–89 (in Russian).

Didkovskii, V.Y., Nosovskii, M.F., 1975. Prychornomorska zapadyna. Miotsen [Prychorno-morskadepression.Miocene]. In: Bondarchuk, V.G. (Ed.), UkrSSR.Neogen [Stratigraphyof the Ukrainian SSR. Neogene]. Naukova Dumka, Kiev, pp. 33–70 (in Ukrainian).

Dodonov, A.E., Sadchikova, T.A., Tesakov, A.S., Titov, V.V., Trubikhin, V.M., 2005. Voprosystratigraphii pliotsen-chetvertichnykh pokrovnykh obrazovanii severo-zapadnogoPrichernomorya [Issues of stratigraphy of Pliocene–Quaternary formations of thenorth-western coast of the Black Sea]. In: Matishov, G.G. (Ed.), Problemypaleontologii i arkheologii yuga Rossii i sopredelnykh teritory. Rostov-na-Donu,Izdatelstvo ‘CVVR’, pp. 27–28 (in Russian).

Fedorov, P.V., 1978. Pleistotsen Ponto-Caspiia [The Pleistocene of the Ponto-CaspianBasin]. Nauka, Moscow (in Russian).

Gillet, H., Lericolais, G., Renault, J.-P., 2007. Messinian event in the Black Sea: evidence ofa Messinian erosional surface. Marine Geology 244, 142–165.

Gozhik, P.F., 1982. Stroenie i usloviya formirovaniya pozdnepleystotsenovogo allyuviya vnizovyakh rek Prichernomorya [Structure and conditions of formation of the LatePleistocene alluvium of lower reaches of rivers in the Black Sea coastal plain]. In:Bondarchuk, V.G. (Ed.), Materialy po izucheniyu chetvertichnogo perioda naterritorii Ukrainy. Naukova Dumka, Kiev, pp. 132–138 (in Russian).

Gozhik, P.F., 1992. Presnovodnye mollusci i korreliatsia verkhnekajnozojskih alluvialnyhotlozhenii yuga Vostochno-Yevropeiskoi platformy [Freshwater molluscs andcorrelation of the Upper Cenozoic alluvial deposits of the southern part of theEast European Platform]. Ph.D. thesis, Institute of Geological Sciences, NationalAcademy of Sciences of Ukraine, Kiev (in Russian).

Gozhik, P.F., 2002. Pontychni presnovodni mollusci pivdnya Ukrainy i Moldovy [Pontianfreshwater molluscs of the southern part of the Ukraine and Moldova]. Kiev (inUkrainian).

Gozhik, P.F., Chirka, V.G., 1973. Novyie danye o pliotsenovyh otlozheniiah nizovii Pruta iDunaia i voprosy ikh koreliatsii [Newdata on Pliocene deposits of the lower reaches ofthe Prut and Danube rivers and problems of their correlation]. In: Bondarchuk, V.G.(Ed.), O nizhnei granitse chetvertichnogo perioda. Naukova Dumka, Kiev, pp. 66–72(in Russian).

Gozhik, P.F., Novoselsky, F.A., 1989. Geologicheskiye usloviya stroitelstva Dneprovsko-Bukhskogo gidrouzla [Geological structure of the Dnieper-Bukh river system].Naukova Dumka, Kiev [Kyiv] (in Russian).

Gozhik, P.F., Mitropolskyi, A.Y., Maslun, N.V., Tsihotska, N.N., 1999. Osoblyvosti sedymen-togenezu v Chornomorskii zapadyni v kainozoi [Features of the Cenozoic sedimento-genesis in the Black sea depression]. Geologiya i poleznyye iskopaiemye Chornogomorya [Geology and minerals of the Black Sea]. Karbon Ltd’, Kiev [Kyiv], pp. 210–214(in Ukrainian).

Gozhik, P.F., Maslun, N.V., Plotnikova, L.F., Ivanik, M.M., Yakushin, L.M., Ischenko, I.I.,2006. Stratygraphiya mezokainozoiskyh vidkladiv pivnichno-zahidnogo shelfu

Chornogo morya [Stratigraphy of the Mesozoic–Cenozoic sediments of the BlackSea north-western shelf]. National Academy of Sciences of Ukraine, Institute ofGeological Sciences, Kiev [Kyiv] (in Ukrainian).

Goretsky, G.I., 1970. Allyuvialnaia letopis velikogo pra-Dnepra [Alluvial chronicle of thegreat ancient Dnieper]. Nauka, Moscow (in Russian).

Goretsky, G.I., 1982. Paleopotamologicheskiye eskizy Palaeo-Dona i Pra-Dona [Palaeo-potamogical sketches of the palaeo-Don and ancient Don]. Nauka i Technika, Minsk,Belarus (in Russian).

Gradstein, F.M., Ogg, J.G., Smith, A.G., Bleeker,W., Lourens, l.J., 2004. Anewgeologic time scale,with special reference to the Precambrian and Neogene. Episodes 27 (2), 84–100.

Gromov, V.I., 1948. Paleontologicheskoye i archeologicheskoye obosnovanie stratigra-phii kontinentalnyh otlozheniy chetvertichnogo perioda na teritorii SSSR [Palaeon-tological and archaeological basis of the stratigraphy of the Quaternary continentaldeposits within the territory of the USSR]. Trudy Geologicheskogo InstitutaAkademii Nauk SSSR. Seriya Geologicheskaya 17 (64) 521 pp. (in Russian).

Hsü, K.J., Giavanoli, F., 1979. Messinian event in the Black Sea. Palaeogeography,Palaeoclimatology, Palaeoecology 29, 75–93.

Hubka, A.N., 1969. Litologo-facialnaia harakteristika i stratigrafiia kontinentalnyh otlozheniiverhnego miotsena MSSR i smezhnyh raionov UkrSSR [Litho-facies characteristics andstratigraphy of the Upper Miocene continental deposits of the Moldavian SSR andadjacent regionsof theUkrainianSSR]. In:StratigraphiianeogenaMoldavii i yugaUkrainy.Shtiintsa, Chishinau [Kishinev], Moldova, pp. 56–72 (in Russian).

Hubka, A.N., 1981. Stratigraphicheskoe polozhenie stolnichensko-ananiivskikh sloev[Stratigraphic position of the Stolnichen and Ananiiv beds]. Biostratigraphiya antropo-gena i neogena yugo-zapada SSSR. Shtiintsa, Chishinau [Kishinev], Moldova, pp. 67–71.

Hubka, A.N., 1982. Kharakteristica carboliyskih sloev i ikh stratotipicheskogo rayona[Characteristics of theCarboliyaBedsand their areaofoccurrence]. In:Adamenko,O.M.(Ed.), Problemy antropogena Moldavii. Shtiintsa, Chishinau [Kishinev], Moldova,pp. 36–75 (in Russian).

Hubka, A.N., Tretyak, A.N., Volok, Z.Y., 1983. Stratigraficheskoe polozhenie karboliiskikhsloiev [Stratigraphical position of the Karboliia Beds]. In: Zelenin, I.V. (Ed.), Geologiiachetvertichnykh otlozhenii Moldavii. Shtiintsa, Chishinau [Kishinev], Moldova,pp. 82–89 (in Russian).

Jones, R.W., Simmons, M.D., 1997. A review of the stratigraphy of Eastern Paratethys(Oligocene–Holocene), with particular emphasis on the Black Sea. In: Robinson, A.G.(Ed.), Regional and petroleum geology of the Black sea and surrounding region: AAPGMemoir, vol. 68, pp. 39–52.

Karmishina, G.I., 1975. Pliotsenovye ostracody yuga Evropeiskoi chasti SSSR [PlioceneOstracoda from the southern European part of the USSR]. Izdatelstvo SaratovskogoUniversiteta, Saratov (in Russian).

Kazantsev, R.A., Shainurov, R.V., 1978. Konus vynosa mytievyh potokov Dunaiskogopodvodnogo kaniona [Turbidity fan of the Danube submarine canyon]. Geomorfo-logia 3, 79–82 (in Russian).

Konstantinova, N.A., 1967. Antropogen Yuzhnoi Moldavii i yugozapadnoi Ukrainy[Anthropogene of South Moldova and the south-western part of the Ukraine].Trudy Geologicheskogo Instituta Akademii Nauk SSSR. Seriya Geologicheskaya 173,1–140 (in Russian).

Korotkevich, Ye.L., 1988. Istoriya formirovaniya hiparionovoy fauny vostochnoy Evropy[History of the Hipparion fauna of Eastern Europe]. Naukova Dumka, Kiev (inRussian).

Kostyanitsyn, M.N., 1969. Gidrologiya ustievykh oblastey Dnepra i Yuzhnogo Bukha[Hydrology of mouth areas of Dnieper and South Bukh]. Gidometeoizdat, Moscow(in Russian).

Krakhmalnaya, T.V., Svetlitskaya, T.V., Chepalyga, A.L., 1993. New data on stratigraphy,magnetostratigraphy and mammal faunas of the late Miocene locality of NovayaEmetovka (Ukraine). Newsletters on Stratigraphy 29, 77–89.

Kruglov, S.S., Tsypko, A.K., 1988. Tektonika Ukrainy [Tectonics of Ukraine]. TrudyUkrNIGRI [Ukrainian Institute of Geological Exploration] (in Russian).

Lebedeva, N.A., 1972. Antropogen priazovia [Anthropogene of the Sea of Azov coast].Trudy geologicheskogo instituta Akademii Nauk SSSR, Seriya Geologeskaya 215,1–106 (in Russian).

Lericolais, G., Panin, N., Ion, G., Bernu, S., Popescu, I., 1998. Danube and Dniepr paleovalleys:new discoveries during BLASON survey on the North-western Black Sea shelf. Geo-Eco-Marina 3, 34–35.

Malovitsky, Y.P., Moskalenko, V.N., Udintsev, G.B., Shehvatov, B.V., 1975. Stroenie pliotsen-chetvertichnyh otlozheniy Chernomorskoi vpadiny [Structure of Pliocene–Quaternarysediments of the Black Sea basin]. Oceanologia 15 (5), 844–849.

Markova,A.K.,1998. EarlyPleistocene smallmammal faunasof EasternEurope.MededelingenNederlands institut voor Toegepaste Geowetenschappen TNO 60, 313–326.

Mankinen, E.A., Dalrymple, G.B., 1979. Revised geomagnetic polarity time scale for theinterval 0–5 m.y. B.P. Journal of Geophysical Research 84, 615–626.

Matoshko, A.V., Chugynny, Y.G., 1993. Dneprovskoye oledeneniye territorii Ukrainy [TheDnieper Glaciation in the territory of Ukraine]. Naukova Dumka, Kiev [Kyiv],Ukraine, 191 pp.(in Russian).

Matoshko, A.V., Gozhik, P.F., Ivchenko, A.S., 2002. The fluvial archive of the Middle andLower Dnieper (a review). Netherlands Journal of Geosciences 81, 339–355.

Matoshko, A.V., Gozhik, P.F., Danukalova, G., 2004. Key Late Cenozoic fluvial archives ofeastern Europe: the Dniester, Dnieper, Don and Volga. Proceedings of theGeologists' Association 115, 141–173.

Matsui, V.M., 1982. O vozraste pozdnepliotsenovyh pereuglublenii Severnogo priazoviia[On the age of the overdeepenings of the Sea of Azov coast]. In: Bondarchuk, V.G.(Ed.), Materialy po izucheniiu chetvertichnogo perioda na teritorii Ukrainy.Naukova Dumka, Kiev [Kyiv], pp. 59–65 (in Russian).

Matsui, V.M., Khristoforova, T.F., Shelkoplias, V.M., 1981. Subaeralinyie otlozheniyasevernjgo Priazoviia [Subaerial deposits of the Sea of Azov coast]. Naukova Dumka,Kiev [Kyiv] (in Russian).


Matsui, V.M., Moskina, O.D., 1976. Gryzuny (Rodentia, Microtidae) iz verhnepliotseno-vyh kontinentalnyh i nizhnechetvertichnyh otlozhenii severnogo priazovia[Rodents (Rodentia, Microtidae) from the Upper Pliocene and Lower Quarternarydeposits of the northern part of the Sea of Azov coast]. Chetvertichnyi period[Quaternary] 16, 86–91 (in Russian).

Melnik, A.P., 1970. Litologiia i zakonomernosti formirovaniia neogenovyh otlozheniiPobuzhiia [Lithology and regularities of the formation of the Neogene deposits ofthe Bukh area]. Naukova Dumka, Kiev [Kyiv] (in Russian).

Mihailescu, K.D., Markova, A.K., 1992. Paleogeographicheskie etapy razvitiya faunyMoldovy v antropogene [Palaeogeographic stages of the Moldavian fauna in theQuaternary]. Shtiintsa, Chishinau [Kishinev], Moldova (in Russian).

Mikhailov, V.N., Mikhailova, N.V., 1991. Zakonomernosti formirovaniya delt vydvizhe-niya na otkrytom morskom poberezhii [Regularities of formation of progradationaldeltas in opened sea coast]. Vestnik Moskovskogo universiteta, Seriya Geografis-kaya 5, 36–44 (in Russian).

Moliavko, G.I., 1960. Neogen pivdnia Ukrainy [Neogene of the southern part of Ukraine].Vydavnytstvo Akademii Nauk Ukrainskoi RSR, Kiev [Kyiv] (in Ukrainian).

Mulika, A.M., Vyrvyklenko, M.K., 1966. Davni pokhovani richkovi dolyny na mezhyrichiDnipro–Molochna [Ancient buried river valleys in the Dnieper–Molochna interfluve]. In:Veklich, M.F., Moliavko, G.I. (Eds.), Paleogeografichni umovy terytorii Ukrainy v pliotsenita antropogeni. Naukova Dumka, Kiev [Kyiv], pp. 119–125 (in Ukrainian).

NSCU [National Stratigraphic Committee of Ukraine], 1997. In: Teslenko, J.V. (Ed.), The Strati-graphic Code of Ukraine. NTP Naftogaz-prognoz Press, Kiev [Kyiv]. 35pp. (in Ukrainian).

Nevesskaya, L.A., 1965. Pozdnepleistotsenovye dvustvorchatye moluski Chornogomorya, ih sistematikai ekologiya [Late Quaternary bevale molluscs of the BlackSea, their systematics and ecology]. Trudy Paleontologicheskogo instituta Akademiinauk SSSR 105, 1–391 (in Russian).

Nikiforov, Y.D., Diyakonu, N., 1963. Gidrologiya ustevoi oblasti Dunaya [Hydrology of near-mouth area of Danube]. Gidrometeorologicheskoe izdatelstvo, Moscow (in Russian).

Palatnaya, N.N., 1991. Geologicheskoe stroenie donykh otlozheny limana Sasyk[Geological structure of floor deposits of Sayk Liman]. In: Shelkoplyas, V.N. (Ed.),Antropogenovye (chetvertichnye) formatsii Ukrainy. Institut geologicheskikh naukAkademii nauk UkrSSR, Kiev [Kyiv], pp. 126–130 (in Russian).

Palienko, V.P., 1992. Neotektonika i ee otrazhenie v reliefe Ukrainy [Neotectonics and itsmanifestation in relief of Ukraine]. Nakova Dumka, Kiev [Kyiv] (in Russian).

Paramonova, N.P., Ananova, Ye.N., Andreeva-Grigorovich, A.S., Belokrys, L.S., Gabunia, L.K.,Grushinskaja, K.F., Kondkarian, S.O., Karmishina, G.I., Kozirenko, T.F., Majsuradze, L.S.,Movlazade-Ateava, E.Z., Nevesskaya, L.A., Ponomareva, L.D., Roshka, V.N., Chapalyga, A.L.,1979. Paleontological characteristics of the Sarmatian sensu lato, and Meotian of thePonto-Caspian area and possibilities of correlation to the Sarmatian sensu stricto andPannonian of Central Paratethys. Annales Géologiques des pays Helléniques 1979,961–971.

Pevzner, M.A., Semenenko, V.N., Vangeigeim, E.A., 2003. Polozhenie ponta vostochnogoParatetisa v magnitohronologicheskoi shkale. [Position of the Pontian within themagnetostratigraphcal scale]. Stratigraphiia i geologicheskaia korelyatsia 11 (5),72–81 (in Russian).

Popov, S.V., Shcherba, I.G., Ilyina, L.B., Nevesskaya, L.A., Paramonova, N.P., Khondkarian,S.O., Magyar, I., 2006. Late Miocene to Pliocene palaeogeography of the Paratethysand its relation to the Mediterranean. Palaeogeography, Palaeoclimatology,Palaeoecology 29, 91–106.

Rekovets, L.I., 1994. Melkiye mlekopitayushchiye antropogena yuga VostochnoyYevropy [Small mammals of the Quaternary of the southern part of EasternEurope]. Naukova Dumka, Kiev [Kyiv] (in Russian).

Rodzyanko, G.N., 1984. Istoriya razvitiya Dona v pliotsene [History of development ofthe Don in the Pliocene]. In Alekseev, A.A. (Ed.), Vozrast i genezis pereuglubleniy nashel'fah i istoriya rechnyh dolin [Age and genesis of overdeepenings within shelvesand the history of river valleys]. Nauka, Moscow, pp. 64–76 (in Russian).

Ross, D.A., 1978. Summary of results of Black Sea drilling. In: Ross, D.A., Neprochnov, Y.P.(Eds.), Initial Reports of the Deep Sea Drilling Project, 42 (part 2). U.S. GovernmentPrinting Office, Washington, DC, pp. 1149–1178.

Ryan, W.B.F., Pitman III, C.W., Major, C.O., Shimkus, K., Moscalenko, V., Jones, G.A.,Dimitrov, P., Gorur, N., Sakinc, M., Yuce, H., 1997. An abrupt drowning of the BlackSea shelf at 7.5 kyr BP. Geo–Eco-Marina 2, 115–126.

Ryan, W.B.F., Major, C.O., Lericolais, G., Goldstein, 1.S.L., 2003. Catastrophic flooding ofthe Black Sea. Annual Reviews of Earth and Planetary Sciences 31, 525–554.

Sacchi, M., Horvath, F., 2002. Towards a new time scale for the Upper Miocenecontinental series of the Pannonian basin (Central Paratethys). EGU StephanMueller Special Publication Series 3, 79–94.

Schekina, N.O., 1966. Materialy do istorii rozvytku roslynnogo pokryvu pivdnia UkrainskoiRSR v pliotseni [Materials on the history of development of the vegetation of thesouthern part of the Ukrainian SSR in the Pliocene]. In: Veklich, M.F., Moliavko, G.I.(Eds.), Paleogeografichni umovy terytorii Ukrainy v pliotseni ta antropogeni. NaukovaDumka, Kiev [Kyiv], pp. 26–37 (in Ukrainian).

Scherbakov, F.A., Kuprin, P.N., Polyakov, A.S., Balandin, Y.G., Ivanov, G.I., Rotar, M.F., 1976.Shelf severo-zapadnoi chasti Chernogo morya v pozdnem pleystotsene-golotsene[The north-western Black Sea shelf in the Late Pleistocene–Holocene]. Chetver-tichnyi Period 16, 141–152 (in Russian).

Scherbakov, F.A., Kuprin, P.N., Potapova, L.I., Polyakov, A.S., Zabelina, E.K., Sorokin, V.M.,1978. Sedimentatsia na kontinentalnom shelfe Chernogo morya [Sedimentation onthe Continental Shelf of the Black Sea]. Nauka, Moscow (in Russian).

Semenenko, V.N., 1975. Prychornomorska zapadyna. Pliotsen [Northern Black Seadepression. Pliocene]. In: Bondarchuk, V.G. (Ed.), Stratigrafiia SSSR. Neogen[Stratigraphy of the USSR. Neogene]. Naukova Dumka, Kiev [Kyiv], pp. 70–81 (inUkrainian).

Semenenko, V.N., 1987. Stratigraficheskaia koreliatsiia verhnego miotsena i pliotsenaVostochnogo Paratetisa [Stratigraphical correlation of the Upper Miocene andPliocene of the Eastern Paratethys]. Naukova Dumka, Kiev [Kyiv] (in Russian).

Semenenko, V.N., 1989. Problems of the Pliocene correlation of East Paratethys andTethys. Geologica Carpathica 40, 75–79.

Semenenko, V.N., Kovaliukh, N.N., 1973. Absolyutny vozrast verchnechetvertichnykhotlozheny Azovo–Chernomorskogo baseina po danym radiouglerodnogo analiza[Absolute age of the Upper Quaternary deposits of the Azov–Black seas basin byradiocarbon data]. Geologichesky Zhurnal 33 (6), 91–97 (in Russian).

Semenenko, V.N., Sidenko, O.G., 1979. Otrazhenie glubinnyh structur v morskihchetvertichnyh otlozheniyah tsentralnoi chasti Azovskogo morya [Reflection ofthe deepseated structures in the marine Quarternary deposits of the central part ofthe Sea of Azov]. Pozdnechetvertichnaya istoriya i sedimentogenez okrainykh ivnutrenikh morei. Nauka, Moscow, pp. 87–99 (in Russian).

Shackleton, N.J., Berger, A., Peltier, W.R., 1990. An alternative astronomical calibration ofthe lower Pleistocene timescale based on ODP site 677. Transactions of the RoyalSociety of Edinburgh, Earth Sciences 81, 251–261.

Shevchenko, A.I., 1965. Opornyie kompleksy pliotsena i nizhnego antropogena yugo-zapadnoi chasti Ruskoi ravniny [Key sites of the pliocene and Lower Antropogene ofthe south-western part of the Russian Plain]. Stratigraficheskoie znachenieantropogenovoi fauny melkih mlekopitayuschih. Nauka, Moscow, pp. 7–59 (inRussian).

Shevchenko, A.I., 1982. Nizhnyaya granitsa kvartera na teritorii Ukrainskoi SSR i voprosyglobalnoi koreliatsii [Base of Quarternary in the Ukraine territory and problems ofthe global correlation]. In: Bondarchuk, V.G. (Ed.), Materialy po izucheniyuchetvertichnogo perioda na territorii Ukrainy. Naukova Dumka, Kiev [Kyiv],pp. 5–18 (in Russian).

Shimkus, K.M., Evsyukov, Y.D., Solovjeva, R.N., 1980. Podvodnye terassy nizhnei zonyshelfa i ikh priroda [Submarine terraces of the lower shelf zone and their nature].In: Shimkus, K.M. (Ed.), Geologicheskoe i geophizicheskoe izuchenie predokea-nicheskoi zony. P.P. Shirshov Institut Oceanologii, Moscow, pp. 81–92 (in Russian).

Shnyukov, Y.F., Lutsiv, Y.K., Inozemtsev, Y.I., 1984. Geologicheskoe stroenie chetvertich-nyh otlozhenii Odessko–Dnestrovskogo raiona Chernogo moria [Geologicalstructure of the Quarternary deposits in the Odessa–Dniester region of the Blacksea]. In: Shnyukov, Y.F. (Ed.), Izuchenie geologicheskoi istorii i protsesovsovremenogo osadkoobrazovania Chernogo i Baltiiskogo morei, vol. 2. NaukovaDumka, Kiev [Kyiv], pp. 54–64 (in Russian).

Shnyukov, Y.F., Inozemtsev, Y.I., Maslakov, N.A., 1999. Geologicheskaia istoria rechnoiseti na severozapadnom shelfe Chernogo moria [Geological history of the riversystem development in the north-western shelf of the Black Sea]. Geologiia ipoleznyye iskopaiemyie Chornogo morya. Karbon Ltd, Kiev [Kyiv], pp. 180–184 (inRussian).

Steininger, F.F., 1999. Chronostratigraphy, geochronology and biochronology of theMiocene ‘European Land Mammal Mega Zones’ (ELMMZ) and the Miocene‘Mammal Zones’ (MN-Zones). In: Rossner, G.E., Hessig, K. (Eds.), The MioceneLand Mammals of Europe. Pfeil, Munich, pp. 9–24.

Tchepalyga, A.L., 1967. Antropogenovye presnovodnye molluski yuga Russkoi ravniny iikh stratigraphicheskoye znachenie. [Quaternary freshwater molluscs of thesouthern part of the Russian Plain and their stratigraphic importance]. TrudyGeologicheskogo Instituta Akademii Nauk SSSR, Seriya Geologeskaya 166,1–222 (inRussian).

Turlo, S.I., 1982. Osnovnye etapy razvitiia rastitelnosti Ukrainy v pleistotsene [Themain stages of the vegetation development in Ukraine in the Pleistocene]. In:Bondarchuk, V.G. (Ed.), Vateraly po izucheniu chetvertichnogo perioda na teritoriiUkrainy [On the Quaternary of Ukraine]. Naukova Dumka, Kiev [Kyiv], pp. 153–161(in Russian).

Vangengeim, E.A., Pevzner, M.A., Tesakov, A.S., 1995. Chronological relationships ofPliocene deposits in fluviatile plains between Prut and Southern Bukh Rivers.Stratigraphy and Geological Correlation 3 (1), 54–64.

Vasiliev, I., Krijgsman, W., Langereis, C.G., Panaiotu, C.E., Matenco, L., Bertotti, G., 2004.Towards an astrochronological framework for the eastern Paratethys Mio-Pliocenesedimentary sequenses of the Focsani basin (Romania). Earth and Planetary ScienceLetters 227, 231–247.

Veklich, M.F., Sirenko, N.O., 1972. Opornye geologicheskie razrezy antropogena Ukrainy[The key sections of the Ukrainian Quaternary], vol. 3. Naukova Dumka, Kiev [Kyiv](in Russian).

Veklich, M.F., Sirenko, N.O., 1976. Pliotsen i pleistosen levoberezhya Nizhnego Dnepra iravninnogo Kryma [The Pliocene and Pleistocene of the left bank of the LowerDnieper and Crimea]. Naukova Dumka, Kiev [Kyiv], 185 pp. (in Russian).

Yevsyukov, Y.D., Kara, V.I., 1990. Relief tsentralnoi chasti dunaiskogo konusa vynosa[Relief of central part of the Danube alluvial fan]. Geomorfologia 2, 71–75 (inRussian).

Zamorii, P.K., 1940. Geomorfologichnyi narys livoberezhia nyzhnogo Dnipra [Geomor-phological sketch of the left bank area of the Lower Dnieper]. Geologichnyi Zhurnal7 (4), 5–31 (in Ukrainian).

Zelinskii, I.P., 1993. Opolzni severo-zapadnogo poberezhiya Chornogo morya. Ikhizuchenie i prognoz. [Landslides of the north-western part of the Black sea coast.Issues of study and prognosis]. Naukova Dumka, Kiev [Kyiv] (in Russian).

Zosimovich, V.Yu., Kulichenko, V.G., Molyavko, G.I., Savron, E.B., 1975. The UkrainianShield. In: Bondarchuk, V.G. (Ed.), Stratigraphy of Ukraine, tom X, Neogen. NaukovaDumka, Kiev [Kyiv], pp. 111–117 (in Ukrainian).

Zubakov, V.A., 2000. Pliotsen Ponto-Kaspiya i yego korelyatsiya [The Pont-CaspianPliocene and its correlation]. Stratigraphija i Geologicheskaya Korelyatsija 8 (1),66–82 (in Russian).

Late Cenozoic fluvial development within the Sea of Azov and Black Sea coastal plains

Documents

Transcript of Late Cenozoic fluvial development within the Sea of Azov and Black Sea coastal plains