Late Pleistocene palaeoenvironments and the last deglaciation on the Kola Peninsula, Russia
- 63 Downloads
Key geological sections on the southern part of the Kola Peninsula were investigated to reveal the evolution of the Late Pleistocene palaeoenvironments and the history of the last deglaciation of the area. Methods used were conventional sedimentological techniques, including palaeocurrent analysis and geomorphological observations. Optically stimulated luminescence dating was used to determine the age of sand-rich sediment units exposed in sections. The results indicate that the Eemian (MIS 5e) marine environment in the onshore coastal areas gradually changed into a glaciolacustrine environment. The first ice advance into the western part of the study area took place already during the Early Weichselian (MIS 5b?). After deglaciation, glaciolacustrine conditions were re-established in the area. Only one younger until is observed in the area; it cannot be excluded that this stems from a Middle Weichselian (MIS 4) glaciation over the area, though it is more probable that it is from the last Scandinavian ice sheet advance over the area during the Last Glacial Maximum (LGM). The southern and central coastal areas of the Kola Peninsula deglaciated between 16,000 and 12,000 years ago. The most prominent end moraine ridges, the Keiva II and Keiva I end moraines were formed in an interlobate zone between ice mass in the interior of the Kola Peninsula and the White Sea ice lobe during the Younger Dryas.
KeywordsPalaeoenvironments Late Pleistocene Deglaciation Keiva end moraines Kola Peninsula
The Kola Peninsula in northwest Russia (Fig. 1), next to the White Sea and the Barents Sea, situated in a coalescing zone, where the Scandinavian and the Barents Sea ice sheets met during the Weichselian (Valdaian) cold stage . It has been generally accepted that the entire Kola Peninsula was covered by ice during the Last Glacial Maximum (LGM), that locally peaked c. 18–16 ka (kilo-annum, ka) ago [1, 2, 3]. The maximum ice extent of the Scandinavian ice sheet (SIS) situated further east of the Kola Peninsula, on the Kuloi Plateau and on the western side of the Kanin Peninsula. It coalesced with the Barents Sea ice sheet (BSIS) to the north and northeast during the local LGM [1, 3, 4, 5, 6, 7, 8, 9, 10] (Fig. 1).
The Weichselian glacial history for the northeastern sector of the SIS, including the Kola Peninsula, is less well known than its southern and western sectors’ . Reconstruction of palaeoenvironments and ice cover on the Kola Peninsula and adjacent areas is important in understanding palaeoenvironmental evolution and the interplay between the SIS and BSIS in the Late Pleistocene. Although there is a comprehensive amount of stratigraphical and palaeoenvironmental studies carried out in northwestern Russia east of the Kola Peninsula [2, 10], less detailed investigations have been carried out on the Kola Peninsula.
For the past decades, there has been significant controversy on the Quaternary history of the Kola Peninsula, although at present the broad outlines of the Quaternary stratigraphy are relatively well established. However, differing opinions remain on the number of glacial advances, their extent and timings, as well as the palaeoenvironments that existed on the Kola Peninsula during the Weichselian [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]. Most of the previous work suggests that the oldest Quaternary sediments on the Kola Peninsula are Saalian in age (Moscowian in Russian terminology, which is broadly correlative to Marine Isotope Stage (MIS) 6) [cf. 14, 15, 26, 28]. Saalian until is overlain by marine, mollusc-bearing silt and clay which also occurs in similar stratigraphical positions elsewhere in NW Russia [cf. 29, 30, 31, 32]. Mainly based on biostratigraphical evidence, and partially supported by Th230/U234 dates and amino-acid ratios, it has been concluded that these marine beds (the so-called Ponoy Beds) on the Kola Peninsula were deposited during the Eemian interglacial (Mikulinian; MIS 5e), i.e., during the high sea-level stage of the Eemian Sea (the so-called Boreal Transgression phase) [cf. 26, 28, 29, 33, 34, 35, 36]. These Eemian marine deposits (the Ponoy Beds) at different sites on the Kola Peninsula represent only fragments of the entire interglacial [cf. 28]. According to Lavrova , Grave et al. , Gudina and Yevzerov  and Korsakova , the Ponoy Beds accumulated in a relatively warm-water marine basin. In addition, Eemian-aged terrestrial sediments have also been described from a few sites in the interior of the Kola Peninsula (for summary see ).
Events after the Eemian interglacial are more controversial. Several Russian workers consider that there are at least two Weichselian-aged until beds on the Kola Peninsula, suggesting two separate ice advances during the Weichselian. Biske , Grave et al. , Nikonov  and Apukhtin  report two until beds from the central, southwestern, northwestern and southern parts of the Kola Peninsula. They suggest that the lower until bed is Early Weichselian in age, while the upper one was deposited during the Late Weichselian. Armand et al.  report that two Weichselian until beds are separated by fluvial, lacustrine and glaciolacustrine interstadial sediments, suggesting highly variable depositional settings in space and time for the sediments deposited between the two assumed Weichselian ice advances.
There are also different reconstructions of timing and dynamics of the last deglaciation on the Kola Peninsula. At present, it is well established that the Kola Peninsula and the White Sea Basin was covered by ice during the local LGM at around 16–18 ka [cf. 1, 10, 39, 40, 41]. However, an ongoing debate concerns the origins and the age of the most prominent glacigenic landforms on the Kola Peninsula: The Keiva End Moraine I and the Keiva End Moraine II, both running along the southern part of the Kola Peninsula (Fig. 1).
Grosswald [22, 42] and Grosswald and Hughes [24, 43] suggested that the BSIS advanced onshore onto the northern part of the Kola Peninsula from the northeast, and through the neck of the White Sea into the present day White Sea Basin and the southern Kola Peninsula. In conjunction to this, they argued that the Keiva end moraines formed in front of the BSIS ice lobe during the Younger Dryas. However, both the previous and more recent work argues that the BSIS did not advance onto the Kola Peninsula during the Late Weichselian deglaciation [cf. 1, 2, 20].
Many reconstructions, mainly based on geomorphological interpretations, conclude that the Keiva end moraine complex formed in an interlobate zone between an inactive ice dome or ice cap located over the central part of the Kola Peninsula, and an active, warm based White Sea Ice Lobe of the SIS [cf. 1, 11, 20, 44, 45, 46]. There are, however, a number of different interpretations on the more detailed formation history and age of the Keiva end moraine complex and related ice dynamics [see, e.g., 11]. Most authors have suggested that the Keiva end moraine complex formed after the local LGM, i.e., during the following deglaciation (Fig. 2). There are only few absolute dates (e.g., cosmogenic exposure ages) from the Keiva Moraines proper that, unfortunately, scatter over a quite large age range from ca. 14 ka to over 30 ka [11, 47]. Hättestrand et al.  considered, mainly based on the morphological characteristics of the deposits in the Keiva ice marginal zone, that the deposits might be older than the local LGM.
In this paper, we present our lithostratigraphical and sedimentological observations from a key site south of the Varzuga village in the southern part of the Kola Peninsula. This site sheds light on palaeoenvironmental developments during the Late Pleistocene from the Eemian interglacial, over the Weichselian, and into the Holocene. In addition, sites in the southern and eastern parts of the Kola Peninsula were studied and sand sediments dated with optically stimulated luminescence (OSL) to clarify the patterns and timings of the last deglaciation and the age of the Keiva end moraines.
Study area and landform observations
The study area is located on the Kola Peninsula, NW Russia (Fig. 1). The bedrock of the Kola Peninsula forms part of the Fennoscandian shield. Bedrock is composed of a suite of Archaean and Early to Middle Proterozoic metamorphic gneisses, granulites, plutonic rocks, metavulcanites and metasedimentary rocks . Observations on geomorphology and Quaternary stratigraphy were carried out at sites located in the southern and southeastern parts of the peninsula (see Fig. 1 for locations). A survey of major geomorphological features (such as the Keiva end moraines and related esker systems, including their extent and orientation) was performed from a helicopter. A subsequent aerial photographic interpretation of landforms related to the end moraine systems was done to clarify the landform systems around the sites, which were investigated in detail using sedimentological methods.
Section logging and sampling methods
Lithostratigraphical investigations on Quaternary sediments were carried out from exposures along major riverbanks. Sections studied here are located close to, or in the area of, previously described Eemian marine sediments [cf. 15, 29]. A clear identification of the Eemian marker horizon was only done at two sites, at Varzuga and Chapoma (see Fig. 1). The Varzuga South exposure was selected for detailed investigations. In addition, Varzuga East, Varzuga West, Strelna North, Babya and Ponoy exposures (see Fig. 1) were logged in detail to reveal the last deglaciation dynamics in the Keiva end moraine zone.
Twenty sedimentary sections were logged and measured along the riverbank exposures and gravel pits, from which eight representative sections were selected for this work. Thickness, bedding plane attitude, nature of basal contacts, texture, structure and the lateral extent of each unit were determined in the field and paleocurrent directions were measured from current induced bedforms. Unit geometry was also defined where possible. Facies analyses were applied to find out contemporary sedimentary environments into which the sediments were deposited.
Until clast fabric measurements and the orientation of glaciotectonic structures related to diamicton units were measured to clarify ice movement directions. Until clast fabric measurements were performed, measuring orientation and dip direction of a-axis of elongated clasts, as outlined in Andrews . In addition, the dip directions of thrust planes and fold axes from glaciotectonically deformed sediments beneath diamicton units were measured. A facies code scheme modified after Miall [50, 51, 52] and Eyles  was used to illustrate the major textural and structural characteristics of the sedimentary units. A full key for the codes is shown in figure legends.
Organic-bearing fine sediments at Varzuga South exposure were analysed for their pollen content. A laboratory treatment followed conventional analysis methods outlined, for example, in Berglund and Ralska-Jasiewiczowa . A total of 32 pollen samples were extracted; the volume of sediment samples varied between 1 and 2 cm3 and the number of terrestrial pollen counted was more than 1000 pollen per sample. Only selected taxa are presented in the pollen diagram from the Varzuga South section. The results are compared to a previously published pollen diagram, by Armand and Lebedeva , from the same area.
Dating of sand-rich sediments was done using the Optically Stimulated Luminescence (OSL) dating method. Twelve OSL samples were collected from sand-rich sediments from eight sections. The samples were taken by hammering a dark grey PVC tube (diameter 75 mm, length 350 mm, wall thickness 4 mm) into the sediment, both ends were closed with plastic lids and the tubes were carefully wrapped with an aluminum foil and placed into black plastic bags. Age determinations of the OSL samples were carried out in the Nordic Laboratory for Luminescence Dating, Risø National Laboratory Denmark.
All OSL samples were prepared under low-level orange light. The light exposed ends of the samples were retained for water content and dose rate analysis. The remaining portion was wet sieved to recover grains 180–250 µm in diameter. This fraction was etched with HCl, H2O2, HF and again with HCl to obtain a quartz-rich separate. After chemical separation, samples had a detectable IR-stimulated signal, indicating residual feldspar contamination. As a result, during routine measurement, all blue light stimulation was preceded by an infrared stimulation at 125 °C .
OSL measurements were carried out on Risø OSL/TL readers, each equipped with a calibrated 90Sr/90Y beta radiation source and a blue (470 ± 20 nm) light source . Prepared quartz grains were attached in a monolayer to c. 10 mm diameter stainless steel disc (approximately a few thousand grains), using silicone oil. A SAR protocol [58, 59] was used to estimate all equivalent doses, with a 260 °C temperature applied for 10 s, and a cut heat of 160 °C. Dose rates were derived from radionuclide concentrations, measured by high-resolution gamma spectrometry , using conversion factors from Olley et al. . These were modified by attenuation factors, based on the observed sediment water contents of 21–28%. Finally, the internal alpha radiation contribution (assumed to be 0.06 ± 0.03 Gy) and the calculated cosmic ray dose rate  were added to the dose rates.
Stratigraphical observations in the Varzuga area
Sediment exposures along the banks of the River Varzuga were discovered already in 1899–1900, and have been described by, e.g., Lavrova , Gudina and Yevzerov , Apukthin , Korsakova and Kolka  and Korsakova . The Eemian marine sediments (the Ponoy Beds) often form a stratigraphical marker horizon at the base of the sediment exposures, as also elsewhere along the southern coast of the Kola Peninsula.
The most complete sediment successions are located south of Varzuga village, where the ground surface is lying c. 40 m asl. Here three sections were logged (Varzuga South sections 1–3; Figs. 3, 4, 5) along the riverbank over the distance of c. 200 m. In addition, two other exposures east and north of Varzuga village were logged (Figs. 3, 4, 5). The sediments in the Varzuga South section are divided into eight (I–VIII) depositional units. However, not all of them are exposed in all logged sections. The depositional units are described below from the base to the top of the exposures.
Varzuga South: sections S1–S3
Unit I consists of massive clay and silt (mud) with odd unbroken bivalves and shell fragments (Fig. 4). The pollen spectra of the upper 3.5 m of unit I is dominated by tree pollen (78–90%), mostly Pinus and Picea (Fig. 6). In the upper part of the sediment succession, Betula rises abruptly to 20–25% (Fig. 6). Abies is present in small abundances in the lower part of the sediment succession. Unit I represents the Eemian marker horizon for the Varzuga area, as also described by the previous research [for summary see 29].
Unit II, conformably overlaying the Eemian marine mud consists of non-calcareous, stratified fine sand and laminated silt, fining upwards into laminated silt and silt clay (Fig. 4). Outsized clasts, interpreted as ice-rafted debris (drop stones) occur throughout unit II. The average diameter of clasts is around 2 cm although one 50 cm boulder-sized clast was also observed in this unit. Unit II could be traced between sections 1 and 3 (Fig. 4).
Unit III consists in its basal part of a massive, silty sand diamicton, 1.1 m thick in section 1. Boulders up to 60 cm in diameter occur in the diamicton. The unit passes upwards into a 0.4 m thick, parallel- and ripple-bedded medium and fine sand horizon in section 1 (Fig. 4) continuing into faintly stratified silt and sand, which is partially massive, containing abundant clasts. Lower contacts of these subunits are sharp. A clast fabric measurements (Fig. 5a) in the stratified diamicton show a statistically significant clast a-axis distribution (Eigenvector V1 = 121°/27°; S1 = 0.633). In sections 2 and 3, this unit is much thinner but sandy, cobble-rich, massive diamicton occur in both sections at the same stratigraphical position (Fig. 4).
Unit IV is composed of laminated silt at its base, coarsening upwards into parallel-bedded fine sand in sections 1 and 3, and climbing-ripple- and parallel-bedded sand sets in section 2 (Fig. 4). Palaeocurrent measurements on ripples in section 2 suggest a flow direction from the north. The sand in the upper parts of unit IV is deformed, the individual bedding planes tilted, thrust folded and faulted. Measured thrust faults dip 10°–30° towards 280°–310° (Fig. 5b). An OSL sample taken from parallel-bedded sand in section 2, 1 m above the laminated silt and clay (unit III), yields an age of 88 ± 7 ka (Table 1).
OSL dating results from the samples taken on the Kola Peninsula
Dose rate (Gy/ka)
88 ± 7
213 ± 12
2.44 ± 0.13
70 ± 8
165 ± 16
2.31 ± 0.12
6.2 ± 0.8
28 ± 3
2.56 ± 0.17
61 ± 5
133 ± 7
2.13 ± 0.11
20.2 ± 1.4
49.1 ± 2.0
2.40 ± 0.12
15.0 ± 1.0
40.5 ± 2.0
2.70 ± 0.10
13.0 ± 1.1
35 ± 2
2.64 ± 0.15
11.9 ± 0.9
26.9 ± 1.0
2.05 ± 0.12
15.9 ± 1.0
37.1 ± 0.5
2.32 ± 0.12
14.2 ± 1.2
16.2 ± 0.9
1.14 ± 0.06
12.3 ± 1.0
18.5 ± 0.9
1.50 ± 0.09
9.7 ± 1.0
16.7 ± 1.2
1.72 ± 0.13
Unit V is through and planar cross-bedded sand and gravel, 0.9 m thick, and only exposed in section 2 (Fig. 4). It overlies unit IV with an erosional contact and it could only be traced laterally for c. 20 m. Palaeoflow measurements indicate a current direction from 340°. An OSL sample taken from planar cross-bedded sand yields an age of 70 ± 8 ka (Table 1).
Unit VI is a massive, matrix-supported silty–sandy diamicton. The lower contact with unit IV (sections 1 and 3) and unit V (section 2) is sharp and erosional. In a number of places along the c. 200 m long exposure, the sands in the underlying units IV and V are folded and show listric folds and associated thrust planes. Structural measurements suggest that the stress-field during deformation, most likely induced by overriding ice that deposited unit VI diamicton, was applied from the WNW (Fig. 5b).
Unit VII and Unit VIII was only observed around section 2 (Fig. 4). Unit VII is composed of trough cross-bedded gravel and sand with an erosional contact to the underlying diamicton. This is in turn conformably overlain by unit VIII which is well-sorted, massive fine sand. An OSL of the unit VIII sand gave an age 6.2 ± 0.8 ka.
Interpretation of the Varzuga South sediment succession
The sediment succession revealed at Varzuga South might serve as an informal stratotype for the Late Pleistocene history of the Kola Peninsula next to the White Sea Basin as the sediment succession exposed here records the longest spell of its Pleistocene history. In summary, sediments in sections S1–3 in the Varzuga South area show four continuous depositional successions, separated by three erosional events. The first continuous sediment succession includes units I and II; Eemian marine sediments with in situ marine shells, foraminifera, saline diatoms [cf. 29] and pollen with interglacial affinity. Unit I passes conformably into laminated silt and clay with occasional fine sand lamina and ice-rafted debris (IRD) with no in situ shells (unit II). All the sedimentary characteristics found in unit II, and especially the frequent IRD, indicate that deposition took place in a stratified water body, most likely in a glaciolacustrine environment. This implies that an interglacial marine sedimentary environment gradually changed into a basin with glaciolacustrine conditions. This in turn suggests glacial damming.
The interglacial marine and proglacial glaciolacustrine sediments (unit II) were followed by the next depositional succession, composed of diamicton (unit III), conformably overlain by laminated silt and fine sand, coarsening upwards into ripple-bedded and parallel-bedded sand (unit IV). As clast fabric in the upper part of unit III shows a moderate preferred clast orientation not typical for subglacially deposited until [cf. 64], it rather might suggest that the stratified diamicton here represents subglacially deformed sediments (i.e., deformation until) or mass movement diamicton deposited close to an ice grounding line in a sub-aquatic setting. Although the exact genesis of the unit III diamicton suite is not fully known, it is reasonable to conclude that the diamicton suite indicates the presence of glacial ice in the area. Laminated silt and fine sand (unit IV), conformably overlying unit III, further suggests a water-lain origin for the upper part of unit III. These laminated silt as well as parallel-bedded and climbing-ripple fine sands were deposited in a glaciolacustrine basin from the north (350°), the upper part most probably representing sub-aquatic bottomset facies of a glaciofluvial delta.
The trough cross-bedded and planar cross-bedded sands and gravels in section 2 (unit V) deposited from the NNW (340°) suggest sub-aerial deposition in a littoral or fluvial environment indicating water level regression.
Diamicton (unit VI), starting the third depositional succession of the area, is interpreted here as until deposited by glacial ice. Glaciotectonic structures in underlying unit IV sands suggest an ice movement across the area from the west-northwest at the emplacement of the unit VI until. Glaciotectonic large-scale structures have also previously been reported from the area [cf. 19] further south of the site studied here.
The cross-bedded sand and gravel (unit VII, section 2) was most likely deposited in a fluvial or littoral setting, while the well-sorted massive fine sand (unit VIII) most probably has an aeolian origin with a Holocene depositional age.
In summary, the sediment succession in the Varzuga South sections suggest that the Eemian marine basin (unit I) developed into a glaciolacustrine basin (unit II). Subsequently, an ice sheet advanced over the area (unit III), and when it retreated, glaciolacustrine conditions were re-established (unit IV). The glaciolacustrine basin in the area became gradually more shallow and started to be filled with deltaic and/or glaciofluvial sub-aquatic fan bottomset fine sands (unit IV). Littoral or fluvial sand and gravel (unit V) indicate transition to near terrestrial or full terrestrial environment in the area. After this, the area once more became covered by glacial ice (unit VI), followed by a fluvial or littoral environment (unit VII) with an aeolian (unit VIII) activity.
The first glacial event (unit III) is bracketed between Eemian-aged marine sediments and two OSL ages on unit IV sand at 88 and 70 ka. As the younger OSL age is from what is interpreted as subaerially deposited fluvial or littoral sand, it may be assumed that this age on well-bleached sand is more accurate age compared to the older one taken from subaqueous bottomset deposits (i.e., a too high age due to incomplete bleaching). However, regardless of this, the OSL ages suggest an Early Weichselian glaciation, followed by deglaciation. The second glacial event (unit VI) is predated by the OSL age on unit V sand at 70 ka and postdated by an OSL age on aeolian sand (unit VIII). From this we draw the conclusion that, based on these sections, it is an open question if the unit VI until represents a MIS 4 or a MIS 2 glaciation (i.e., Middle or Late Weichselian glaciations).
This site is located c. 6.4 km northwest of the Varzuga South sections (Fig. 3), exposing sediments in a landform interpreted as a sub-aquatic glaciofluvial fan/glaciofluvial delta (c. 60 m asl). The exposure is c. 30 m wide and c. 15–20 m high. The sediment succession logged was divided into three sediment units (Fig. 4).
Unit I is 8 m thick and mainly consists of laminated silt and clay with small amount of fine sand. The bedding planes of clay, silt and fine sand lamina dip 6°–8° towards 140°.
Unit II is 9 m thick. The lower contact with unit I is sharp and the basal 0.5 m consists of medium sand coarsening upwards into sand and gravel with large-scale diagonal planar cross beds (i.e., deltaic-type foresets), each 10–20 cm thick. The dip of the cross beds is c. 15° towards the SSE suggesting a paleoflow direction from between 330° and 340° (Fig. 4). An OSL sample taken from the middle part of the foresets gave an age of 61 ± 5 ka (Table 1).
Unit III is composed of massive-looking pebble-to-cobble gravel capping the exposure having an eriosional contact with unit II sediments.
This site is located c. 2.6 km north-northwest of the Varzuga South sections (Fig. 3), exposing sediments on the western side of a landform interpreted as a glaciofluvial delta/ a sub-aquatic fan (c. 60 m asl). The exposure is c. 100 m wide and 20–26 m high, located on the bank of an eastern tributary to Varzuga River.
Unit I is 17 m thick and composed of laminated silt and clay, interbedded with sand in its upper part (Figs. 3, 4). The lower 7.5 m is grey laminated clay and silt that coarsens upwards and grades into red brown laminated silt with occasional fine to medium sand layers. The bedding planes dip c. 8° towards 140° and some lamina are internally deformed. No diatoms, nor molluscs were found in unit I.
Unit II is 5 m thick, composed of planar and trough cross-bedded sand. Palaeocurrent measurements suggest deposition from the NW. One OSL sample dated from this unit yields an age of 20.2 ± 1.4 ka (Table 1).
Unit III is 5 m thick, composed of inclined large-scale sand beds with internal cross stratification (10–20 cm thick). The inclination of the large-scale cross beds suggests deposition towards an east-southeast sector. An OSL sample dated from this unit yields an age of 15 ± 1 ka (Table 1).
Unit IV is 4 m thick, composed of planar cross-bedded and through cross-bedded sand and gravel. The unit starts with a pebble gravel layer. The lower contact with unit III is sharp and erosional and marked by a clast supported pebble gravel. Palaeoflow measurements on planar cross-bedded sets indicate current from westerly direction. An OSL sample from this unit yields an age of 13 ± 1 ka (Table 1).
Interpretation of Varzuga North and East sections
The Varzuga North and South sections expose sediments in glaciaofluvial landforms that are interpreted as glaciofluvial deltas or sub-aquatic fans. Both sites are located within and in conjunction to the Keiva II end moraine zone that runs across the southern Kola Peninsula (Fig. 1) [cf. 65, 66].
The sediment succession in the Varzuga North section represents a typical Gilbert-type sub-aquatic fan succession including the bottomset and foreset sediment facies architecture . The basal, laminated fine sediments (unit I) with no in situ molluscs or diatoms were deposited into a glaciolacustrine basin, while the sand and gravel foresets (unit II) indicate the prograding deltaic infill in the same basin. The massive pebble-to-cobble gravel beds at the top might represent the topset layer of a Gilbert-type delta. An alternative interpretation of the topmost unit is that it represents beach sediments formed during glacioisostatic uplift.
The Varzuga East section with laminated silt and clay (unit I) is similar to the Varzuga North section and was most likely deposited in the same glaciolacustrine basin. The laminated silt and clay beds in the Varzuga East coarsen upwards into cross-bedded sand (unit II), which both suggest deposition in a pro-delta slope environment and the approach of the geomorphologically defined delta next to the site (Fig. 3). The thick unit of planar cross-bedded sand (unit III), resting with an erosional contact to the underlying glaciolacustrine sediments, might not represent delta foresets proper, but rather a propagation of shoreface sediment during gradual glacioisostatic uplift.
The sand and gravel at the top of the section (unit IV), with a lower contact showing an angular unconformity, is interpreted here as beach-face sediment deposited during final uplift/drainage of the basin.
The OSL ages on the littoral regressive sediment succession (units II–IV) from the Varzuga East section, ranging from 20 to 13 ka, suggest deposition after the last local glacial maximum. We regard the oldest date (20 ka) to give a too high age due incomplete bleaching of these deeper-water sediments, while the upper and younger OSL ages from shore/beach-face sediments are more accurate depositional ages. The 61 ka OSL age on deltaic cross-bedded sand in the Varzuga North section is regarded as too old due to incomplete bleaching. We also regard this sediment succession as deposited in a glaciolacustrine basin at ice retreat after the local last glacial maximum in the area.
Stratigraphical observations between the River Strelna and the River Ponoy
The exposure at Strelna North is situated within the Keiva II end moraine zone (Figs. 1, 7), where the ground surface of the exposed sediments is c. 110 m asl. Exposed sediments in the Strelna river bank were logged in two sections (S1 and S2), and divided into four sedimentary units (I–IV; Fig. 8). The top 2 m was covered by scree material.
Unit I is more than 2 m thick and consists of arched-bedded layers of angular to sub-angular cobbles in a gravel and sand matrix. Long-axis orientation and imbrication of cobble to pebble clasts suggests deposition in a water-flow current from the northwest (330°) (Fig. 8).
Unit II, exposed both in sections 1 and 2 (Fig. 8), is a 7.3 m thick, rhythmically laminated bed, composed of laminated silt, clay and horizontally bedded fine sand. The sediments are conformably resting on Unit I and coarsens upwards.
Unit III is 9.3 m thick, having a sharp, erosional contact with unit I. It is composed of a stacked sequence of inclined beds of sand and gravel. Bed inclinations are between 12° and 20° towards the SE (300°–330°). An OSL sample taken 2.6 m below the top of the section yields an age of 11.9 ± 0.9 ka (Table 1). The section ends with c. 2 m of horizontally bedded, matrix-supported gravel (unit IV).
Interpretation of the Strelna North section
Unit I in the Strelna sediment succession is interpreted as an esker deposit (i.e., a subglacial tunnel or ice crevasse deposit), while the rhythmites above (unit II) were deposited in a glaciolacustrine basin. The inclined sand and gravel beds (unit III) suggest the progradation of deltaic foresets from the c. NW into the same glaciolacustrine basin, followed by topset deposition (unit IV). From the sediment succession, it seems evident that ice existed north of the Keiva II end moraine adjacent to a glacial lake to the south. At present, the area north of the Keiva II end moraine is occupied with numerous lakes (Fig. 7b) formed most probably in glacially eroded basins. These types of basins are typical geomorphological features on the proximal side of major end moraines also suggesting the presence of ice north of the Keiva II end moraine.
The OSL age at 11.9 ka on deltaic foreset-bedded sand (unit III) is regarded as taken from well-bleached sand and thus date a late stage of the glaciolacustrine basin formed at ice retreat after the last local glacial maximum.
The Babya section
In the Babya area (Figs. 1, 9), there are a number of small moraine ridges composed of until and glaciofluvial sediments that belong to Keiva I end moraine zone. The sediment exposure is situated on the left bank of Babya River, c. 20 km from the river mouth at an altitude of c. 112 m asl (Figs. 1, 9). The exposed sediment succession, c. 8 m high, was divided into four (I–IV) sediment units (Fig. 8).
Unit I is 2 m thick and composed of laminated silt and clay with a subordinate amount of parallel- and ripple-bedded fine sand. Palaeocurrent measurements on current ripples indicate deposition from between east and south.
Unit II is parallel- and ripple-bedded sand, which overlies unit I with a sharp but conformable contact. Unit II is 3 m thick but its 0.5 m thick top part is deformed (Figs. 8, 9c). The deformation structures include listric and thrust folds, and also reverse and normal faults. Measured thrust planes dip 10°–15° towards the south and fold axis dip 05°–12° towards the west (Fig. 5c). Palaeocurrent directions measured on current ripples indicate deposition from the southwest (Fig. 8). An OSL sample from OSL age ripple-bedded sand yields 14.2 ± 1.2 ka (Table 1).
Unit III is 0.3–0.4 m thick, massive silty sand diamicton with occasional cobble- and pebble-sized clasts. The diamicton is overlying unit II with an erosional lower contact. Clasts in the diamicton were sparse and clast fabric measurements were not carried out from this unit.
Unit IV is sand that fines into silty sand at the top of the section. The sand bed shows indications of cryoturbation (e.g., convolutions). An OSL sample on the lower part of the sand yields an age of 12.3 ± 1.0 ka and an OSL sample on the upper sand bed yields an age of 9.7 ± 1.1 (Table 1).
Interpretation of the Babya section
The laminated silt and clay (unit I), which coarsen upwards into fine sand (unit II) were deposited in glaciolacustrine basin. The unit III diamicton is interpreted as a subglacial traction until. The thrusts faults and listric folds in the unit II sand are coupled to the deposition of the until (glaciotectonic deformation) and together they suggest ice advance from the south into the area. The sedimentary structures of the unit IV sands are rather undiagnostic to their depositional settings; the lower part might represent fluvial or glaciolacustrine deposition during deglaciation and the upper sand might represent aeolian deposition with the cryoturbation structures indicating a periglacial environment.
According to OSL ages on unit II and IV sands, the ice advance as indicated by the unit III until and underlying glaciotectonics is bracketed between 14.2 ± 1.2 and 12.3 ± 1.0 ka ago.
The Ponoy River area
The Keiva II end moraine and a few feeding eskers from the west and the east are the main glacigenic features in the Ponoy area (Fig. 10). The area has almost no superficial minerogenic sediments and bedrock features dominate the landscape [cf. 65].
One sediment exposure was identified in the bedrock-walled, v-shaped tributary valley to the Ponoy River. One undisturbed section, 50 m wide, was logged in the southern part of the exposure (Fig. 8) The exposed sediment succession, c. 24 m high, was divided into two (I–II) sediment units.
Unit I is 16–20 m thick resting on crystalline bedrock. The unit is composed of a stacked sequence of inclined beds of medium and coarse sand. Bed inclinations are between 15° and 20° dipping towards the NW (60°). An OSL sample, 10 m below the top of the section yields an age of 15.9 ± 1.0 ka (Table 1). The section ends with c. 4 m of stratified pebble and cobble gravel with sand matrix (unit II).
Interpretation of the Ponoy sediment succession
The inclined sand beds (unit I) suggest progradation of deltaic foresets from c. SW into the glaciolacustrine basin, followed by delta topset deposition (unit II). This sediment succession suggests that a waterbody must have existed in the Ponoy River valley proper during the time of deposition, according to the OSL date at around 16 ka ago. Based on the altitude of the delta topset (unit II), the water level in this basin was at around 70 m asl.
Additional stratigraphical information
Pulonga River (Keiva II) and Strelna South (Keiva I)
Additional observations of palaeoflow directions were made at the sites in the Keiva I and Keiva II end moraines. At Pulonga (66°33′N, 39°41.5′E; altitude c. 175 m asl, Fig. 1) observations were made on glaciofluvial sediments from a poorly exposed sediment succession within the Keiva II end moraine. The section, over 15 m high, exposed cross-bedded pebble gravel, suggesting a palaeoflow direction from the NW during deposition.
Close to the entry of the River Strelna in to the White Sea (noted as Strelna South, 66°6.62′N, 38°31.71′E; altitude c. 80 m, Fig. 1) is an exposure of glaciotectonized sediments cutting a landform that is a part of the Keiva I end moraine. Diamicton (interpreted as until) capping glaciotectonized sediments show a girdle-shaped clast fabric with the strongest, though quite weak strength (Eigenvector V1 = 211°/20°; S1 = 0.577), while thrust planes in sediments beneath the until suggest deformation from the SSW (Fig. 5d). Combined together, this suggest an ice advance from the White Sea Basin northwards.
Late Pleistocene palaeoenvironments in the southern part of the Kola Peninsula
The Varzuga South site is a stratigraphical key section to unravel the Late Pleistocene history of the southern Kola Peninsula and the White Sea Basin. The Eemian (MIS 5e) marine sediments at the base of the Varzuga South sites are correlative to marine mollusc-bearing sediments at the earlier described Chapoma and other Eemian marine deposit sites in NW Russia . At Chapoma, the Eemian marine sediments rest on a until, which most likely was deposited during the Saalian glaciation (MIS 6). The Eemian sediments in the Varzuga South exposures were probably deposited in deeper water (depth 40–50 m) than the sediments at Chapoma; the diatom taxa in the latter suggest deposition in a littoral zone influenced by inflow of river water [cf. 29]. Based on the spatial distribution and altitude of Eemian marine deposits in NW Russia (for summary of the sites see ), the geographical outline of the Eemian Sea in the area is reconstructed (Fig. 11). A similar but more general reconstructions of the extent of the Eemian Sea has been previously presented by, e.g., Saarnisto et al.  and Larsen et al. .
The previous sediment investigation along the Varzuga River exposures report that until is lying on top of Eemian marine mud [cf. 19]. However, the continuous sediment succession at our Varzuga South sections 1 and 3 sites indicate a gradual change from a warm Eemian marine to cold(er) glaciolacustrine (ice-rafted debris) basin prior to the first ice advance into the area in the early Weichselian (Fig. 12a–c).
The pollen spectra from the upper part of the marine sediments in the Varzuga North section (Fig. 6) corresponds broadly to that described by Armand and Lebedeva  from the same area. The tree pollen percentage (> 70%), suggests a birch-pine forest with elements of Picea, i.e., a northern boreal forest or a mixed taiga forest. The increase in Betula in the uppermost part of the diagram might indicate a cooling at the end of the Eemian and the transition into the Herning stadial (MIS 5d). This is consistent with what is suggested from Eemian lacustrine sediments at Sokli ([69, 70] for location, see Fig. 1). A speculative alternative would be the MIS 5c Sokli/Brørup interstadial, in which Helmens et al. [69, 70] argue hosted a birch-pine forest around at the Sokli. However, as there are no indications of stadial conditions in the pollen spectra at Varzuga (Fig. 6), this scenario must be considered less likely.
The results presented here and those of Mangerud et al.  suggest that glaciolacustrine conditions existed in the southern coastal area of the Kola Peninsula in the Early Weichselian prior to c. 80–90 ka ago. A change from marine Eemian conditions into fluvial and a lacustrine, cold-water condition is also reported on the east coast of the White Sea. Larsen et al.  record such deposits within and south of the Kanin Peninsula, before and contemporaneously with the Early Weichselian Kara Sea-dominated ice sheet advance into the Pechora lowlands further to the east.
The glacigenic sediment succession at the Varzuga South sites (unit III, Fig. 4) indicate an ice advance into the area immediately after the glaciolacustrine phase (Fig. 12c). This ice advance is chronologically tied to the Early Weichselian. However, since there is an erosional contact between these units, the glaciolacustrine sediments (unit II) and the glacigenic suite (unit III) might not represent a continuum in sedimentation. As the first Weichselian Kara Sea ice sheet glaciation over the Kanin Peninsula did not reach the east coast of the White Sea , this ice sheet seems not to be a good candidate for lowermost glacigenic sequence at the Varzuga South sections. Even though no directional indicators for this glacial advance are at hand, the most probable source is a Scandinavian-based ice sheet. The Sokli stratigraphy [cf. 69, 70] suggest that the SIS did not reach eastern Finnish Lapland during the Herning stadial (MIS 5d) but later rather during the Rederstall stadial (MIS 5b). If the lowermost until in the Varzuga South sections is correctly tied to a SIS advance in the Rederstall Stadial (MIS 5b), then it must be concluded that the ice advanced further to the east than proposed in the ice sheet reconstruction in, e.g., Hättestrand  and Mangerud et al. .
Water-lain sediments above the glacigenic suite (unit III) in the Varzuga area indicate that glaciolacustrine conditions were re-established when the ice retreated from the area, probably in the MIS 5b/5a transition as suggested from OSL datings (Fig. 12d).
The upper until in the Varzuga South sections and glaciotectonic deformation in sands beneath the until suggest a later advance into the area from the west (Fig. 12e). As stated before, it is an open question if this until bed represents a MIS 4 or a MIS 2 glaciation (i.e., Middle or Late Weichselian glaciations). According to Svendsen et al.  and Larsen et al.  the SIS and the BSIS merged during the MIS 4 glaciation during which the Kola Peninsula and the Kanin Peninsula northeast of the White Sea was totally ice covered, and until from this MIS 4 glaciation would presumably be deposited. However, our preferred interpretation is that the until (unit VI) in the Varzuga sections was deposited during the Late Weichselian. Similar until units are observed at several sites [cf. 65, 74], such as along the southern part of the Strelna River bank and at Chapoma, where it occurs on the ground surface.
To our knowledge, there is not yet discovered one single section on the Kola Peninsula, where a complete Weichselian stratigraphy is exposed. However, the Varzuga South sections to date reveal the most continuous sediment section from the Eemian to the Early Weichselian, and then a hiatus to the Late Weichselian glaciation/deglaciation sediment succession related to a SIS expansion over the area.
Deglaciation and the age of the Keiva moraines on the southern Kola Peninsula
The OSL age determination on deltaic sediments at the Ponoy site suggest that this area deglaciated c. 16 ka ago. This age is close to the age of the SIS’s last glacial maximum age, c. 16–17 ka ago, on the western shore of the Kanin Peninsula [2, 10] only c. 150 km NE of the Ponoy site. A glaciolacustrine basin, the water level of which stood at around 70 m asl, existed in the Ponoy River valley at the time when these deltaic sediments were deposited.
Data collected from the Keiva end moraine zones I and II, running along the south and east coasts of the Kola Peninsula (Fig. 1), suggest that ice must have been present on the interior of the Kola Peninsula contemporaneously with ice in the White Sea Basin during last deglaciation, as the previous work also has concluded [cf. 1, 20, 75, 76]. Furthermore, it is also evident that the Keiva I and II moraines were deposited in an interlobate zone between those ice bodies (Fig. 12f).
Palaeocurrent data and geomorphological observations suggest that melt water flow and sediment deposition forming the Keiva II end moraine (as indicated from the Varzuga North, Strelna North and Pulonga sites, all situated within or adjacent to the Keiva II end moraine) was from between northwest and northeast, i.e., from ice situated over the interior of the Kola Peninsula. An OSL age on deltaic sand from the Strelna North section at c. 12 ka (11.9 ± 0.9 ka) suggests that the Keiva II end moraine was formed during the Younger Dryas from a northeasterly direction. An OSL age on littoral sediments formed on the top of the glaciofluvial Keiva II sub-aquatic fan sediments in the Varzuga East section yields an age of c. 13.0 ± 1.1 ka, i.e., over the Alleröd/Younger Dryas transition. All other retrieved OSL ages on deltaic sediments coupled to Keive II moraine, ranging between 20 and 60 ka, are considered too old due to incomplete bleaching. Thus, based on the results presented here, we argue that the Keiva II end moraine, from the Varzuga area and at least as far as the Strelna River area was deposited mainly from meltwater-transported sediments from an ice mass located in the interior of the Kola Peninsula during the Younger Dryas. Furthermore, geomorphological evidence from Niemelä et al. , together with the observation presented here, suggest that the sediments in the eastern part of Keiva II moraine were derived both from an White Sea Ice lobe and ice in the central part of the Kola Peninsula. Our OSL ages also confirm the early assumptions [e.g., 37, 46, 75] that the Keiva II end moraine correlates to the Salpausselkä end moraines in Finland.
The Keiva I end moraine complex is a more discontinuous suite of end moraines south of the Keiva II end moraine  (Fig. 1). The Keiva I moraine is built by glaciofluvial sand and gravel as well as push moraine elements composed of until and deformed sediments. Two sites in this study are located at, or adjacent to, the Keiva I ice marginal formations. Clast fabric measurements and glacitectonic structures from the topmost until at the Strelna South site and the Babya site, respectively, indicate that ice from the south overrode the area. According to OSL dates at Babya, the until unit deposited during the last ice advance over the Babya area took place between 14.2 and 12.3 ka ago. Related glaciofluvial sub-aquatic deposits were, therefore, deposited from the south during the Younger Dryas.
Based on lithostratigraphical, palaeoflow, glaciotectonic and OSL dating results, it can be argued that the Keiva I end moraine, at least in Babya, was deposited by a White Sea Ice lobe from the south at more or less the same time as the Keiva II end moraine, which was deposited from the north. The palaeoflow and dating results presented here thus confirm the assumption by Rainio et al.  and Svendsen et al. , that both Keiva end moraines in the southern part of the Kola Peninsula were formed during the Younger Dryas in an intelobate zone of two ice masses—one in the White Sea and the other over the interior of the Kola Peninsula. Deltas and sub-aquatic fans and silt and clay rhythmites indicate that glaciolacustrine basin(s) were formed in this interlobate zone and later also in the White Sea basin as deglaciation progressed [cf. 39, 40, 41]. Furthermore, the results indicate that the Keiva end moraines correlate to the Younger Dryas end moraine chain running around Fennoscandia as suggested by Lundqvist and Saarnisto .
Based on the results presented here, the following main conclusions can be drawn on the history of the Late Pleistocene and the last deglaciation on the southern part of the Kola Peninsula:
The Eemian interglacial marine basin in the White Sea and the southern Kola Peninsula changed gradually into a glacolacustrine basin during the first part of the Early Weichselian.
Ice entered into the Varzuga area in the Early Weichselian and subsequently retreated, after which glaciolacustrine conditions were re-established.
A second ice advance from the west into the Varzuga area occurred after 70 ka, most probably during the Late Weichselian, though a Middle Weichselian (MIS 4) glaciation in between cannot be excluded.
The last deglaciation of the southern and central coastal areas of the Kola Peninsula took place between c. 16–12 ka ago. The eastern part of the Kola Peninsula around Ponoy was deglaciated around 16 ka and the southern coastal area from Varzuga area to Babya at around 12 ka.
The Keiva end moraines formed in an interlobate zone between the ice masses in the interior of the Kola peninsula and the White Sea ice lobe during the Younger Dryas.
The western part of the Keiva II end moraine (glaciofluvial deltas and sub-aquatic fans) derived its sediments mainly from the north, while the Keiva I end moraine zone (glaciofluvial deltas and sub-aquatic fans) from the south.
The authors thank numerous field assistants and Russian colleagues who worked with MS, JPL and NP for three field seasons on the Kola Peninsula. We also thank Drs. Per Möller and Anders Schomacker for thorough reviews that markedly improved the quality of the manuscript.
- 1.Svendsen JI, Alexanderson H, Astakhov VI, Demidov I, Dowdeswell JA, Funder S, Gataullin V, Henriksen M, Hjort C, Houmark-Nielsen M, Hubberten HW, Ingólfsson Ó, Jakobsson M, Kjær KH, Larsen E, Lokrantz H, Lunkka JP, Lyså A, Mangerud JM, Matiouchkov A, Murray A, Möller P, Niessen F, Nikolskaya O, Polyak L, Saarnisto M, Siegert C, Siegert MJ, Spielhagen R, Stein R (2004) Late Quaternary ice sheet history of northern Eurasia. Quat Sci Rev 23:1229–1271CrossRefGoogle Scholar
- 3.Larsen E, Fredin O, Lyså A, Amantov A, Fjeldskaar W, Ottesen, D (2016) Causes of time-transgressive glacial maxima positions of the last Scandinavian ice sheet. Norw J Geol 96:159–170Google Scholar
- 5.Svendsen JI, Astakhov VI, Bolshiyanov DYu, Demidov I, Dowdeswell JA, Gataullin V, Hjort Ch, Hubberten HW, Larsen E, Mangerud J, Melles M, Saarnisto M, Siegert MJ (1999) Maximum extent of the Eurasian ice sheets in the Barents and Kara Sea region during the Weichselian. Boreas 28:234–242CrossRefGoogle Scholar
- 8.Houmark-Nielsen M, Demidov I, Funder S, Grosfjeld K, Kjaer KH, Larsen E, Lavrova N, Lyså A, Nielsen JK (2001) Early and Middle Valdaian glaciations, ice-dammed lakes and periglacial interstadials in northwest Russia: new evidence from the Pyoza River. Glob Planet Change 31:215–237CrossRefGoogle Scholar
- 10.Demidov IN, Houmark-Nielsen M, Kjaer K, Larsen E, Lyså A, Funder S, Lunkka JP, Saarnisto M (2004) Valdai glacial maxima in the Arkhangelsk district of northwestern Russia. In: Ehlers J, Gibbard PL (eds) Quaternary glaciations—extent and chronology. Part I: Europe. Elsevier: Amsterdam, pp 321–336CrossRefGoogle Scholar
- 12.Ramsay W (1898) Über die Geologische Entwicklung der Halbinsel Kola in der Quartärzeit. Fennia 16:1–151Google Scholar
- 13.Nikonov AA (1966) Stratigrafiya i paleogeografiya antropogena Kolskogo poluostrova I prilezhaschikh oblastei (Quaternary stratigraphy and paleogeography of the Kola Peninsula and adjacent areas). Verhniy Pleistocen. Stratigrafya i absolyutnaya geohronologia (In: Late Pleistocene. Stratigraphy and geochronology). Moscow, pp 92–105 (in Russian) Google Scholar
- 14.Armand AD, Armand NN, Grave MK, Yevzerov VYA, Lebedeva RM (1969) Svodnaya stratigraficheskaya shema chetvertichnykh (antropogenovyh) otlozhenii Kolskogo polustrova v svete noveishikh dannykh. (Generalized scheme of Quaternary deposits of the Kola Peninsula in the light of new data). Osnovnye problemy geomorfologii i stratigrafii antropogena Kolskogo poluostrova (In: Main problems of geomorphology and stratigraphy of the Quaternary at the Kola Peninsula). Leningrad, pp 7–24 (in Russian) Google Scholar
- 15.Grave MK, Gunova VS, Devyatova EI, Lavrova MA, Lebedeva RM, Samsonova LYa, Cheremisinova EA (1969) Mikulino interglacial in southeast of the Kola Peninsula. Osnovnye problemy geomorfologii i stratigrafii antropogena Kolskogo poluostrova (In: Main problems of geomorphology and stratigraphy of the Quaternary at the Kola Peninsula). Leningrad, pp 25–56 (in Russian) Google Scholar
- 16.Devyatova EI (1982) Prirodnaya sreda pozdnego pleistotcena i yego vliyanie na razvitie cheloveka v Severodvinskom basseine i v Karelii (Late Pleistocene environment and its influence on the humans development in the drainage basin of Severnaya Dvina and Karelia). Russian Academy of Sciences, Petrozavodsk (in Russian) Google Scholar
- 17.Gudina VI, Evzerov VYa (1973) Stratigrafiya i foraminifery pleistocena Kolskogo poluostrova (Late Pleistocene stratigraphy and foraminafera of the Kola Peninsula). In: Trudy instituta geologi i geofiziki (Proceedings of the institute of geology and geophysics) 175, Novosibirsk (in Russian) Google Scholar
- 18.Apukthin NI, Klyunin CF, Tkachenko LI 1977: Paleogeography of the Kola Peninsula during the Upper Pleistocene (based on new data obtained in 1972–1974). Stratigrafiya i paleogeografiya chetvertichnogo perioda Severa Evropeiskoi chasti SSSR (In: Stratigraphy and paleogeography of the Quaternary at the north of the European part of the USSR), pp 16–21 (in Russian) Google Scholar
- 19.Apukhtin NI (1978) New data on Quaternary stratigraphy of the southeastern part of the Kola Peninsula. Trudy VSEGEI Leningrad. Chetvertisnaya geologiya i geomorfologiya 297:53–65 (in Russian) Google Scholar
- 20.Ekman I, Iljin V (1991) Deglaciation, the Younger Dryas end moraines and their correlation in the Karelian A.S.S.R. and adjacent areas. In: Rainio H, Saarnisto M (eds) Eastern Fennoscandian Younger Dryas and moraines—Excursion Guide. IGCP Project 253, Termination of Pleistocene. Geological Survey of Finland, Espoo, Finland, pp 73–102Google Scholar
- 29.Ikonen L, Ekman I (2001) Biostratigraphy of the Mikulino interglacial sediments in NW Russia: the Petrozavodsk site and a literature review. Annales Academiae Scientiarum Fennicae. Geologica—Geographica, vol 161. Suomalainen Tiedeakatemia, HelsinkiGoogle Scholar
- 33.Arslanov KhA, Yevzerov VYa, Tertychnyy NI, Gerasimova SA, Lokshin NV (1981) Concerning the age of deposits of the Boreal transgression of Ponoy layers on the Kola Peninsula. In: Velichko AA, Faustova MA (eds) Pleistocene glaciation of the Eastern European Plain. Nauka Press, Moscow, pp 28–38 (in Russian) Google Scholar
- 34.Faustova MA (1984) Late Pleistocene glaciation of European USSR. In: Velichko AA (ed) Late quaternary environments of the Soviet Union. University of Minnesota Press, Minneapolis, pp 3–12Google Scholar
- 37.Lavrova MA (1960) Quaternary geology of the Kola Peninsula. Moskva (in Russian) Google Scholar
- 38.Biske GS (1959) Chetverticnya otlozhyeniya i geomorfologiya Karelii (Quaternary deposits and geomorphology of Karelia). Karelskiy filial Nauk SSSR (Karelian branch of the Academy of Sciences USSR). Petrozavodsk (in Russian) Google Scholar
- 44.Punkari M (1993) Modelling of the dynamics of the Scandinavian ice sheet using remote sensing and GIS methods. In: Aber JS (ed) Glaciotectonics and mapping glacial deposits. University of Regina Press, Regina, pp 232–250Google Scholar
- 48.Koistinen T, Stephens MB, Bogatchev V, Nordgulen Ø, Wennerström M, Korhonen J (2001) Geological map of the Fennoscandian Shield, scale 1:2 000 000. Trondheim: Geological Survey of Norway, Uppsala: Geological Survey of Sweden. Ministry of Natural Resources of Russia. Geological Survey of Finland, EspooGoogle Scholar
- 49.Andrews JT (1971) Techniques of Till Fabric Analysis. Technical Bulletin No. 6. British Geomorphological Research GroupGoogle Scholar
- 52.Miall AD (1996) The geology of fluvial deposits. Springer, BerlinGoogle Scholar
- 53.Eyles N (1983) Glacial geology: a landsystem approach. In: Eyles N (ed) Glacial geology: an introduction for engineers and earth scientists. Pergamon Press, Oxford, pp 1–18Google Scholar
- 54.Berglund BE, Ralska-Jasiewiczowa M (1990) Pollen analysis and pollen diagrams. In: Berglund BE (ed) Handbook of Holocene palaeoecology and palaeohydrology. Wiley, Chichester, pp 455–484Google Scholar
- 55.Armand AD, Lebedeva RM (1966) Sporovo-pyltsevaya kharakteristika opornogo razreza mezhlednikovykh otlozhenii na yuzhnom beregu Kol’skogo poluostrova (Spore-and-pollen characteristic of the key section of Interglacial deposits in the southern coastal area of the Kola Peninsula). In: Formirovanie reliefa i chetvertichnykh otlozhenii Kol’skogo poluostrova (In: The topography and Quaternary deposits formation of the Kola Peninsula). Nauka, Moscow, pp 77–86 (in Russian) Google Scholar
- 63.Korsakova O, Kolka V (2005) Quaternary geology and landforming processes. In: Excursion guide of the international field symposium, Kola Peninsula, NW Russia. Apatity: Print. Kola Science Centre RASGoogle Scholar
- 65.Niemelä J, Lukashov A, Ekman I (1993) Quaternary deposits of Finland and northwestern part of Russian Federation and their resources. Map 1: 1 000 000, Sheet 2. Geological Survey of Finland & Institute of Geology at the Karelian Science Centre of the Russian Academy of Sciences: EspooGoogle Scholar
- 66.Hättestrand C, Clark C (2006) Reconstructing the pattern and style of deglaciation of Kola Peninsula, NE Fennoscandian ice sheet. In: Knight PG (ed) Glaciology and Earth’s changing environment. Blackwell Publishing Ltd, Oxford, pp 199–201Google Scholar
- 67.Nemec W, Steel RJ (1998) Fan Deltas: Sedimentology and Tectonic Settiings. Blackie & Son, London, p 444Google Scholar
- 68.Saarnisto M, Lunkka JP, Yevzerov V, Kolka V, Demidov I (2001) On the deglaciation history of the Kola Peninsula: an interplay between the Scandinavian ice sheet and the Ponoy Ice Cap. In: Abstract of Fourth Quaternary Environment of the Eurasian North (QUEEN) Workshop, Lund, Sweden 7–10 April 2000, p 50Google Scholar
- 71.Mangerud J, Jakobsson M, Alexanderson H, Astakhov V, Clarke G, Henriksen M, Hjort C, Krinner G, Lunkka J-P, Möller P, Murray A, Nikolskaya O, Saarnisto M, Svendsen J (2004) Ice-dammed lakes and rerouting of the drainage of northern Eurasia during the Last Glaciation. Quat Sci Rev 23:1313–1332CrossRefGoogle Scholar
- 72.Hättestrand M (2008) Vegetation and climate during Weichselian ice free intervals in northern Sweden—Interpretation from fossil and modern pollen records. Dissertations from the Department of Physical Geography and Quaternary Geology no 15, pp 5–36Google Scholar
- 74.Krasnov II, Duphorn K, Voges A (1971) International Quaternary Map of Europe, Sheet 3: Nordkapp. INQUA Commission. Bundesanstalt für Bodenforschung, HannoverGoogle Scholar
- 75.Vvedenski LV (1934) Reljef yuzhnoi chasti Kolskogo poluostrova (Relief of the southern part of the Kola Peninsula). In: Izvestiya geographicheskogo obshestva SSSR (Proceedings of the geographical society of the USSR) LXVI, no 6, pp 844–863Google Scholar
- 76.Lunkka JP, Saarnisto M, Demidov I (2001) The last glaciation on the Kola Peninsula, Russia. In: European Unionion of Geosciences—EUG XI Strasbourg, France, 8–12th April, Abstracts; p 219Google Scholar
- 78.Apukhtin NI, Yakovleva SV (1967) History of geological development of NW European part of USSR in Quaternary time. In: Apukhtin NI, Krasnov II (eds) Geology of the North-Western European USSR; with map of Quaternary deposits of the North-Western European part of the USSR, Scale 1: 2 500 000. Nedra, Leningrad, pp 318–326 (in Russian) Google Scholar
- 79.Bakhmutov VG (1993) Paleomagnetism and lithology of Late Weichselian deposits in Ust-Pjalka’s periglacial lake, south-east of the Kola Peninsula. In: Bakhmutov VG, Yevzerov V.Ya., Kolka VV (eds) Geologica Carpatica, vol 44, no 5, pp 315–324Google Scholar