“Thermophilous” Trees in the Lateglacial Vegetation of the Eastern Baltic: New Questions for an Old Issue

Abstract

The results of a recent palynological study of the Kulikovo section (southeastern Baltic) allow us to elaborate on issues of the presence of pollen from the “thermophilous” trees (Picea, Alnus, Corylus, Ulmus, Quercus, Tilia, Fraxinus) in Lateglacial sediments. The research shows their continuous presence throughout the interval of 13.9–12.5 ka with a total contribution from 7% to 17%. Comparing the results with regional palynological data revealed certain similarities and patterns, which are not sufficiently explained by contamination by ancient redeposited material. These taxa belonging to the hemiboreal plant group were most probably part of the Lateglacial vegetation along with subpolar and boreal plants. This correlates well with regional paleoclimate reconstructions, assuming that, during the major part of the Lateglacial, July temperatures were comparable to modern average temperatures, which range from +16.5 to +18 °C. Inclusion of hemiboreal tree vegetation in paleoreconstructions will offer an updated picture of the dynamics of the natural environment and increase the accuracy of paleoclimatic reconstructions based on palynological data, allowing us to obtain more accurate temperature values of the climate of the past.

1. Introduction

Assessing how ecosystems will evolve under future climate change requires a thorough understanding of this interconnection in the past. The Lateglacial is one of the epochs providing data on responses of the biota to crucial climate changes during warming periods and cold spells. Paleogeographic studies conducted in recent decades in the Baltic region indicate that the Lateglacial environment in this area underwent numerous changes associated with the melting of the last Weichselian glacier and the formation of the modern appearance of the landscape [1]. The evolution of vegetation has attracted continuous attention as one of the most sensitive indicators of climate change. In the Baltic region, the changes in the vegetation cover were especially dynamic during the Lateglacial [2,3,4,5,6,7,8,9,10]. Based on regional palynological data, there are assumed to be several major stages. Thus, pioneer herbaceous (wormwood, sedge, grasses) and dwarf shrub communities existed in areas freed from ice in the Old Dryas (16–14.7 ka). Then, sparse forests (“park” forests) with pine, birch, and juniper appeared in the Bølling, and true pine–birch forests spread in the Allerød (14.7–12.9 ka). The restoration of tundra communities probably took place during the cooling of the Older Dryas (14.1–14.0 ka), and the reduction in woody vegetation with an increase in open shrub–herbaceous spaces occurred in the Younger Dryas (12.9–11.7 ka) [2,3,4,5,6,7,8,9,10]. Representatives of cold-tolerant flora such as Betula nana, Dryas octopetala, and Selaginella selaginoides are often found in Lateglacial sediments of the Baltics.

At the same time, pollen of species such as spruce, alder, hazel, linden, etc., is regularly found in the sediments of this period [2,3,4,5,6,7,8,9,10]. The term “thermophilous” is often used for these trees in the context of Lateglacial vegetation. The presence of such pollen in Lateglacial sediments is interpreted as contamination by old redeposited material, since researchers assume that the growth of “thermophilous” taxa in the harsh climate of that time was impossible [3,4,8,9]. Most often, the deposits of the previous interglacial period—the Eemian—are considered as the “supplier” of redeposited pollen, and, in an overwhelming majority of cases, it is stated that “the sources of contamination are difficult to establish”. Until now, the in-depth discussion of the phenomenon of contamination is absent, as well as any alternative explanation of the presence of “thermophilous” pollen in Lateglacial sediments.

Four currently investigated Lateglacial archives of the southeastern Baltic (Kaliningrad region, Russia)—Kamyshovoe [8], Kulikovo (this publication), and Sambian and Utinoe Boloto (in progress)—have shown that they also contain pollen of “thermophilous” trees. A comparison of the palynological data highlighted that the percentage of these taxa in four sequences is not random, as might be expected in the case of “contamination”, but shows a certain pattern not only among themselves, but also with other paleoarchives of the Baltic region, from Estonia to Poland (Figure 1). For example, the content of trees such as Carpinus, Fraxinus, Quercus, Tilia, and Ulmus fluctuates between <1% and 3% each, while Alnus in most of the sections varies between 5% and 11%. Furthermore, while the set of “thermophilous” trees may be different, Corylus is present in all sections without exception, with its content reaching 15%.

These facts necessitate studying this phenomenon in more detail. Can contamination explain the observed pollen values and similarities between different sequences? What are the newest data on the climatic conditions of the Lateglacial period in the region? Could these tree species be part of the plant communities that existed back then, and how does this correlate with the simultaneous presence of tundra representatives such as Betula nana, Dryas octopetala, etc.? To elaborate on this topic, we present the results of a palynological study of the Kulikovo section, located in the southeastern part of the Baltics (N 54°56′12″; E 20°21′31″). A spectacular feature of this paleoarchive is that it allows us to zoom in on the vegetation dynamics during 13.9–12.5 ka. The section has a thickness of about 1.92 m. Deposits were studied at the microstratigraphic level every 1–3 cm and have a robust geochronological base. The detailed palynological data on the Kulikovo and comparison with regional palaeobotanical information set against the paleoclimatic frame provide a new look at the issue of “thermophilous” trees in the Lateglacial. In this paper, new questions are posed regarding this issue. These steps are necessary not only to understand the evolution of the Lateglacial environment, but also to refine paleoclimatic reconstructions, which are largely based on paleobotanical data, and to offer thorough insights into climate–ecosystem interconnections in the past to support future investigations.

2. Materials and Methods

2.1. Geographical Setting, Fieldwork, and Sampling

The Kulikovo outcrop is situated in the northern part of the Sambian Peninsula in the southeastern Baltic region (Figure 1). The surface of this territory, represented mainly by undulating marginal moraines, was formed by the retreating Late Weichselian ice [12]. Morain hills formed by till deposits dominate the landscape, with sandy and silty depressions of glaciolacustrine and glaciofluvial origin.

Nowadays, the climate of the southeastern Baltic is transitional from the maritime climate of Western Europe to the moderate continental climate of Eastern Europe. The cold period of the year lasts from 90 to 105 days, with an average long-term temperature in January from −3 °C on the coast to −6.5 °C in the east. The warmest month is July with an average monthly temperature between 16.5 °C and 18 °C. Air humidity is high, ranging from 70% in summer to 80%–90% in winter. The annual amount of precipitation is 600–700 mm, and its distribution is uneven across the territory and across the seasons of the year. The region is located in a zone of excessive moisture. The entire territory of the southeastern Baltic falls in the forest zone and in the mixed coniferous–broad-leaved forests subzone. The typical representatives of the present-day forest vegetation are pine (Pinus), spruce (Picea), birch (Betula), aspen (Populus tremula), maple (Acer), birch (Betula), rowan (Sorbus), alder (Alnus), beech (Fagus), hazel (Corylus), hornbeam (Carpinus), linden (Tilia cordata), ash (Fraxinus excelsior), elm (Ulmus glabra), and oak (Quercus robur). An essential part of the territory is occupied by meadows, bogs, and agricultural landscapes. Typical soils are sod–podzolic and sod–eluvial–gley [12].

Sediment samples were collected in metal boxes of 7 cm diameter and 50 cm length and transported to laboratories for further processing and analysis. Sampling was carried out depending on the visible layering of the sediment (1–3 cm each). In this paper, only the palynological study of the Kulikovo sequence is considered. The methods and results of lithological and geochronological analyses are presented in detail in previous articles [13]. Five samples from the Kulikovo section were subjected to radiocarbon (AMS) dating, which showed that the sequence spans an age range from 14.1 ka to 12.5 ka. From a depth of 192 cm to 14 cm, the sediments are represented by layers of peaty dark brown or clayish grey or brown gittja. A bed of grey dense clay forms the upper part of the sequence (14–0 cm) [13].

2.2. Palynological Analysis

A total of 120 samples of the section were studied. Subsamples of 3 cm³ size were prepared using chemical removal of extraneous matter, such as calcium carbonate by using 10% HCl, and humic acids by 10% KOH, followed by gravity separation of coarse particles with a heavy liquid (CdI2+KI) [14]. Lycopodium spores were added in order to calculate pollen concentrations. Pollen identification was based on [15]. No less than 300 pollen grains were counted in each sample. In most of the samples, the number of counted terrestrial pollen grains exceeded 400.

For calculating and presenting the pollen data, the programs TILIA and TILIA-graph were used [16]. Along with visual inspection, a stratigraphically constrained cluster analysis (CONISS) was used for subdivision of the pollen zones [16].

3. Results

The data obtained can be subdivided into the following palynological zones (Figure 2):

LPAZ 1, 192–184 cm: Samples in the lowest part of the section (192–190 cm) contained an insufficient amount of pollen and spores. Furthermore, in this section, more than 300 microfossils were counted in each sample.

The contribution of AP reaches its maximum value (92%) in this zone. A dominance of Pinus (up to 80%) and Betula sect. Albae (up to 10%) is observed. Corylus reaches a value of 8%, while Juniperus presents up to 5%. Salix appears and shows a constant presence throughout the sequence, fluctuating in the range of 1%–5%. Alnus, Carpinus, Tilia, and Quercus appear at the top of the zone, as does Ericaceae (Figure 3). Among herbs, Artemisia (5%), Chenopodiaceae (5%), and Ranunculaceae (4%) show higher values in the lower part of the zone, while Cyperaceae and Poaceae start to increase at the top. Sphagnum (15%) and Selaginella selaginoides (12%) exhibit the highest values throughout the sequence. Helianthemum and Dryas octopetala both nearly reach 1% in the upper part of the zone. Redeposited pre-Quaternary sporomorphs are also present at the very top of the zone, reaching 1%.

LPAZ 2, 184–167 cm: The contribution of AP drops and fluctuates between 50% and 70%. A sharp decline of Pinus (30%–46%) is observed, as well as a growth of Betula sect. Albae from 15% to 45%, the highest value in the studied sequence. Juniperus also has the highest value (up to 8%) throughout the sequence. Picea appears and constitutes up to 3%. Betula nana (up to 2%) and Alnaster Fruticosus (<1%) are present in the spectrum. Alnus, Corylus, Carpinus, Tilia, Quercus, and Ulmus present and constitute <1% in each taxon. A sharp growth of Cyperaceae (up to 30%) and Poaceae (up to 15%) is observed. Sphagnum nearly disappears, except for a small peak in the upper part of the zone, followed by a peak of Equisetum. Values of Selaginella selaginoides drop to less than 1%. Thalictrum is present. Redeposited pre-Quaternary sporomorphs are also present.

LPAZ 3, 167–102 cm: The proportion of AP slightly grows (to 85%) in the lower part of the zone and decreases to 60%–75% in the upper part. Pinus fluctuates between 22% and 60%, and Betula sect. Albae between 15% and 45%. Picea is present, but pollen values drop. The contribution of Juniperus slightly decreases towards the upper part of the zone, while Betula nana has an opposite trend, reaching the maximum value of 4%–5% at the top of the zone. Alnus, Corylus, Carpinus, Tilia, Quercus, and Ulmus frequently appear in the lower part of the zone and then disappear, except Alnus, Carpinus, Quercus, and Corylus. The latter demonstrates a constant presence (1%–2%). While the pollen percentage of Cyperaceae drops, the values of Poaceae remain comparable with the previous zone, and Artemisia shows a considerable growth (up to 15%) in the upper part of the zone. Ranunculaceae demonstrates a similar trend, reaching 14%. Thalictrum and Rumex are constantly present, while Ephedra and Dryas octopetala, as well as Typha, Myriophyllum, and Urtica, appear in the pollen record sporadically. Single redeposited pre-Quaternary sporomorphs are present in the lower part of the zone, but are not registered further in the sequence.

LPAZ 4, 102–54 cm: After a short-lasting recovery of AP at the bottom of the zone (85%), a gradual drop of the value is observed through the zone, reaching 70% at the top. Pinus varies between 45% and 66%. Betula sect. Albae shows an uneven decline to minimum values of less than 5%. Values of Betula nana considerably drop as well, while the curve of Juniperus shows approximately the same pattern as in the previous zone. The contribution of Corylus increases and reaches 11%. Values of Alnus drop, but in contrast, Carpinus and Quercus, to some degree, are more present. Acer and Fraxinus appear for the first time in the pollen record. Cyperaceae recovers and reaches 30%, and values of Poaceae, in contrast, drop and vary between 2% and 10%. The contribution of Artemisia decreases in the lower part of the zone with a minimum of 2% and grows up to 10% at the top. The values of Ranunculaceae drop as well and do not exceed 6%. Typha, Myriophyllum, and other herbs show increased values compared with the previous zone. Helianthemum and Selaginella selaginoides appear again, mostly in the upper part of the zone. Equisetum has its maximum in the lower part of it (up to 17%) and then sharply declines.

LPAZ 5, 54–18 cm: AP continues to reduce to 50% at the top of the zone. While Betula sect. Albae demonstrates a slight increase (up to 20%), the curve of Pinus pollen shows several short-lasting peaks in the generally decreasing trend, and varies between 30% and 50%. Picea is more present, while Juniperus is less present. Alnaster fruticosus shortly appears in the lower part of the zone. Corylus continues to be present with approximately the same values (up to 11%), and those of Alnus and Carpinus grow (1%–1.5%). Tilia, Quercus, Ulmus, and Fraxinus are sporadically present. Cyperaceae fluctuate between 14% and 30%, and Poaceae between 3% and 7%. Values of Artemisia grow and reach 16%, as well as those of Ranunculaceae, which vary between 4% and 11%. Myriophyllum reaches 1%. Helianthemum and Selaginella selaginoides are present.

LPAZ 6, 18–0 cm: A gradual recovery of AP to 82% is seen. Pinus reaches 70%, and Betula sect. Albae varies between 8% and 16%. Picea constitutes 2%. Alnus and Corylus decrease to <1% and 2%, respectively, while Carpinus and Ulmus reach 1%–1.5%. Tilia, Quercus, and Fraxinus are present, while Salix nearly disappears. Cyperaceae and Poaceae vary between 14% and 25% and between 3% and 4%, respectively. Artemisia sharply decreases to 4%, while a gradual decrease in Ranunculaceae to the same value of 4% is seen. Myriophyllum is sporadically present, reaching up to 1%. Polypodiaceae (up to 2%) and Sphagnum (up to 7%) show increased values in the upper part of the zone, while Selaginella selaginoides reaches up to 1% in the lowest part.

4. Discussion

Firstly, the results of the palynological study of the Kulikovo section have shown that the Lateglacial vegetation in the study area included various types, which were typical for the wider areas of the Baltic region [2,3,4,5,6,7,8,9,10]. The vegetation was characterized by an abundance of steppe and meadow grasses (Poaceae, Artemisia, Asteraceae, Rosaceae, Valeriana, etc.), cold-tolerant plants and/or heliophytes (Betula nana, Ephedra, Helianthemum, Selaginella selaginoides, Dryas octopetala, Thalictrum), as well as plants typical of areas with disturbed soils (Artemisia, Chenopodiaceae). It also included plants typical of swamps, wet meadows, and coastal habitats (Cyperaceae, Ranunculaceae, Urtica, Equisetum), as well as aquatic plants (Myriophyllum, Typha, Nymphaea). The next group of plants, such as Equisetum, mosses, and ferns (Sphagnum, Polypodiophyta), was associated with boreal forest communities. Woody vegetation was represented primarily by trees with a wide ecological amplitude: Pinus and Betula sect. Albae. The palynological data of Kulikovo also include a considerable list of “thermophilous” taxa: Picea, Alnus, Corylus, Tilia, Quercus, Carpinus, and Ulmus. The results of the study show their continuous presence throughout the interval of 13.9–12.5 ka, and their total contribution ranges from 7% to 17% (Figure 3).

As mentioned earlier, pollen of these species with comparable percentage values was also found in dozens of Lateglacial sections of the Baltic region [2,3,4,5,6,7,8,9,10,17,18,19]. Due to the assumption that the severe climate of the Lateglacial prevented the growth of this type of vegetation, contamination by ancient microfossils emerged as a plausible explanation for this phenomenon. However, this conclusion becomes debatable in light of the paleoclimatic data obtained in recent years.

A summary of the latest paleoclimatic research shows that the main trend of the global climate transition from the Last Glaciation to the Holocene was rapid warming, with temperatures close to or possibly exceeding modern ones [1]. Relatively high air temperatures in the Pleniglacial probably triggered intense ice sheet melting, which began before 20 ka. The associated significant input of meltwater into the North Atlantic probably contributed to the pre-HS (Heinrich Stadial) 1 climate event, dated to 20–19 ka [1]. The general warming trend was interrupted by several colder (stadial) epochs, apparently caused by instability of the Atlantic meridional circulation [20]. Temperature estimates of the Lateglacial climate vary to some extent, depending on the data used for the reconstruction (paleobotanical, chironomidae, Cladocera); however, emerging research indicates that the Lateglacial was characterized by higher summer air temperatures than previously assumed [21,22,23].

According to the study by Schenk et al. [22], based on the results of analysis of 15 plant macrofossil records from Scandinavia, July temperatures in the southern part of this zone, in the immediate vicinity of the glacier margin, were around +16 °C in the Late Pleniglacial, Allerod (AL), Younger Dryas (YD), and Preboreal (PB), and slightly lower, +14 °C, in the Bølling (BØ), and Older Dryas (OD). The obtained results correlate well with the chironomid data for the southern and eastern Baltics: based on the study of the Żabieniec and Koźmin Las sections (Poland), July temperatures from +16 to +18 °C in the range of 16,200–10,050 cal. yr BP were reconstructed [21,24]; for Lithuania, in the Lieporiai section, values of +14 °C for BØ and YD and +16 °C for AL were obtained [25]. For July temperatures in the Younger Dryas, a decrease was either not recorded (Żabieniec) or fluctuations were observed within the range of +14.0 °C to +15.8 °C (Lieporiai, Koźmin Las). In this regard, the macrobotanical study of the Kasuciai section (Lithuania) is also remarkable: here, in sediments dating back to the Younger Dryas, numerous macroremains of Alisma plantago-aquatica and Ranunculus flammula were found [2]. These plants are indicators of moderate climatic conditions with summer thermal limits of +14 to +15 °C and +15 to +16 °C, respectively [22]. The sediment interval of the appearance of these plants coincides with the peaks of Alnus, Picea, and Corylus in the section [2]. Thus, the wide spectrum of data shows that the reconstructed July temperatures for the end of Pleniglacial and the Lateglacial, with the possible exception of stadials, were comparable to the modern average temperatures, which range from +16.5 to +18 °C.

Pollen assemblage of the Kulikovo also supports the assumption of mild, rather than severe, summers of the Lateglacial. Thus, the contribution of the woody vegetation does not drop lower than 50%, fluctuating within the range of 60%–70% during a major part of the studied time interval. The constant presence of coniferous trees and their growth as early as 13.9 ka in the vicinity of the studied location were also recorded in the results of the phytolith analysis of the section [26]. Furthermore, the continuous presence of Ranunculaceae and Equisetum, pollen of Typha latifolia, in the record indicates summer temperatures not lower than +14 to +16. Apparently, factors other than summer temperature, such as a change in continentality of climate, or the duration of cold and warm seasons, could have played a determining and limiting role in the dynamics of vegetation cover and the diversity of the ecological groups [22].

In the above-mentioned macrobotanical study in Scandinavia [22], several plant groups were distinguished among 62 species of the Lateglacial vegetation: subpolar (tundra), northern boreal and central boreal (taiga), and hemiboreal (boreal-nemoral or vegetation of mixed coniferous–broad-leaved forest). The study showed the constant presence of hemiboreal plant groups in the immediate vicinity of the southern boundary of the Scandinavian Ice Sheet, starting from the Pleniglacial. Modern hemiboreal vegetation also includes the “thermophilous” tree species under consideration [27]. Thus, typical representatives of hemiboreal woody taxa are spruce (Picea), aspen (Populus tremula), maple (Acer), birch (Betula), rowan (Sorbus), alder (Alnus), beech (Fagus), hazel (Corylus), hornbeam (Carpinus), linden (Tilia cordata), ash (Fraxinus excelsior), elm (Ulmus glabra), and oak (Quercus robur).

Let us look closer at the ecology of tree taxa with the highest pollen values in the Lateglacial sediments: Alnus and Corylus. Alnus is a pioneer plant, with summer thermal limits of +14 to +15 °C, easily occupying new or previously disturbed habitats [28]. The symbiosis with Actinomycetales situated in root nodules allows Alnus to assimilate free nitrogen, enabling the colonization of infertile habitats, including even debris-covered glacier surfaces [29]. Alnus occurs on both mineral and organic soils. It is tolerant to considerable fluctuations of water levels but is less tolerant to drought. Alnus glutinosa prefers places with streaming surface waters: along river and stream valleys, in damp peat meadows and lowland swamps, and other places with a close groundwater level. Corylus, a plant with summer thermal limits of +14.6 to +15 °C, is primarily found in the understory of deciduous, mixed, and coniferous forests [28,30]. However, near the northern limit of its range, where the habitat and climatic conditions are unfavorable for mixed deciduous forests, hazel may become one of the dominant species in forest communities [28]. Corylus prefers sunny sites, temporarily or periodically shaded. It suffers from excessive drought or heat and generally prefers cool climatic conditions. It can grow on a wide variety of soils, including sandy or stony clay with a considerable skeletal fraction. Hazel provides easily decomposing litter for soil formation, thus playing an important role in biocoenoses [28].

As we can see, the ecological tolerance of both tree species is high enough to postulate that these plants might have been among the first to spread in areas freed from the glacier, especially taking into account the above data on summer temperatures of the Lateglacial. Without elaborating on the ecological characteristics of other tree species (Picea, Tilia, Quercus, Carpinus, Ulmus) in this paper, we note that, in terms of climate severity, according to the USDA Plant Hardiness zonation, most of the species fall between the third and fifth zones, where the first value corresponds to the modern climatic conditions of Lapland, and the second to the Baltic states, northern Poland, and central Russia [31]. The modern northern boundaries of hazel and linden, for example, are located beyond the Arctic Circle, up to 66°–68° N [28].

The above paleoclimatic and paleobotanical information leads us to question the thesis that the existence of Picea, Corylus, Alnus, Tilia, etc., in the Lateglacial Baltics is impossible. Examining the phenomenon of “contamination” per se further complicates the discussion.

As already mentioned, in most cases, researchers have referred to pollen “contamination” originating from the Eemian deposits. This interglacial period lasted approximately from 130 to 117 ka [32]. Within the Baltics, the Eemian deposits are found in different stratigraphic settings. For example, in the Kaliningrad region, they are represented by lenses of sand, silt, and clay of up to 34 m thickness on the Sambian Peninsula and up to 6 m thickness on the Vishtynets (Masurian) Upland. The lenses are deposited at levels ranging from +29 m to −6 m and covered by Weichselian deposits (till, limnoglacial) with a thickness ranging from 3 m to 40 m [33]. In Poland, Eemian deposits are found, for example, in paleo-basins outside the Last Glaciation (central Poland) and on the Masurian and Pomeranian Uplands (northern Poland) under a 25–30 m thick layer of moraine [34]. In Lithuania, Latvia, and Estonia, Eemian deposits have been found both in situ and in a redeposited state. In the latter case, their dating is based largely on palynological data [35]. Thus, the depth, thickness, and location of Eemian deposits vary a lot across the Baltic region. No less diverse are the geographic location, geomorphological conditions, altitudes above sea level, distances from Eemian layers, etc., of those sections that are supposedly “contaminated” by Eemian pollen. Meanwhile, a review of published palynological data has shown that it is difficult, if not impossible, to find a Lateglacial section in the region that is not “contaminated”, i.e., does not contain pollen from the “thermophilous” trees. Moreover, if the pollen of Tilia, Quercus, Carpinus, Ulmus, and Fraxinus is maintained at a modest 1%–3% for each tree species, then the content of Alnus and Corylus can reach 15%, and Picea, 28%. There are also synchronous increased values of individual tree species in different sections at certain time intervals. For example, simultaneous peaks of Corylus were recorded during 14,000–13,800 cal. yr BP (Kulikovo, Dukštelis [10]; Haljala [3]); 13,500–13,200 cal. yr BP (Sambian, Utinoe Boloto, Lopaičiai [7], Ūla-2 [36]; Haljala [3], Rapa [5]); and 12,900–12,500 cal. yr BP (Kulikovo, Utinoe Boloto, Lopaičiai [7], Dukštelis [10], Rapa [5], Haljala [3]) (Figure 4). A synchronous peak of Picea was observed in the Allerød layers of the Haljala (up to 11%) and Kunda (up to 20%) sections in Estonia [3]. Taking into account the diverse geography and stratigraphy of the locations, the question following arises: what combination of circumstances could have contaminate all (?) the Lateglacial sections of the region, and in such a way that the pollen curves of some tree taxa are synchronous at certain time intervals and even can demonstrate approximately the same percentage of “contamination”?

There are also a number of other points requiring clarification. The palynological data of the Eemian indicate that the vegetation cover changed radically during the interglacial period, and its evolution included stages of both broad-leaved and mixed forests, as well as coniferous pine–spruce forests [34]. Which stage of vegetation development in each specific case should be considered as a source of contamination? Should the broad spread of pine and birch pollen during the Eemian also be considered “contamination” for the Lateglacial sediments, and, if yes, then to what degree?

The next important issue is whether sediments should be considered contaminated if other signs of ancient material—such as, for example, pre-Quaternary spores and pollen or dinoflagellate cysts—are absent or minimal? In the southeastern Baltic area, the moraines of the Weichselian glaciation incorporate the outliers of the Paleogene age, which, indeed, could be a source of redeposited ancient material [33]. In order to understand the degree of possible contamination of the Lateglacial sediments by the ancient pollen and spores, we must consider the results of a palynological study of Paleogene deposits. Firstly, they are characterized by the presence of ancient gymnosperms and angiosperms, which are representatives of subtropical and tropical flora: Tricolporopollenites exactus–T. retiformis–Quercoidites microhenrici, Platanipollis ipelensis–Castaneoideaepollis oviformis–Tricolpopollenites foraminatus, and Inaperturopollenites–Sciadopityspollenites–Sequoiapollenites assemblages [37]. Secondly, in these deposits, the content of dinoflagellate cysts varies within 30%–60%. Meanwhile, in the Kulikovo section, as well as in other Lateglacial sequences of the Kaliningrad region, dinocysts and other pre-Quaternary palynomorphs do not exceed 1%–2% in total and are present only sporadically (Figure 2). This example demonstrates the scale of the input of ancient material into the Lateglacial sediments in cases where the latter actually occurs. It is difficult to explain why potential “contamination” with Eemian material should have a percentage content that is an order of magnitude higher.

We suppose that the questions posed and the facts considered here cast doubt on the view of the redeposited material as the main source of pollen of Picea, Alnus, Corylus, Tilia, Quercus, Carpinus, Ulmus, etc., in the Lateglacial sections of the Baltic region. If we still admit that the ubiquitous regional presence and close percentage ratios of the indicated species are a result of contamination, then this phenomenon deserves a detailed and targeted study: (1) Its sources and mechanisms should be analyzed and mapped. (2) The probability of input for each territory should be calculated. (3) Coefficients for tree pollen percentage should be introduced for certain depths and/or time intervals. Without these steps, a simple reference to “contamination” and arbitrary deselection of certain palynological data seem unfounded.

We assume that the so-called “thermophilous” trees were part of the hemiboreal plant communities of the Lateglacial period, along with tundra, forest-tundra, and taiga vegetation. They could be part of the patches of woody vegetation during periods with favorable conditions for the latter or be an admixture in the pine–birch forests. Thus, Corylus, for example, can still occur locally even when its pollen values are lower than 2%. Values up to 25% suggest that hazel is present in mixed forests, whereas values above 25% indicate its presence in woodland as either one of the main species or even the dominant species [28]. The coexistence of subpolar, boreal, and hemiboreal plants with very different climatic tolerances indicates a mosaic pattern and a significant role of local conditions and factors such as shading/illumination, soil types, humidity, slope exposure, and relief type. Their simultaneous presence in the Lateglacial is detected by the botanical macroremains not only in the sediments of Scandinavia, but also in the Baltic region [2,3].

Inclusion of hemiboreal tree vegetation in paleoreconstructions will offer an updated picture of the dynamics of the natural environment as a whole and, in particular, increase the accuracy of paleoclimatic reconstructions based on palynological data, allowing us to obtain more accurate temperature values of the climate of the past. Knowledge of the ecology of hemiboreal tree species can provide additional information about the features of the Lateglacial environment. For example, elevated pollen values of Alnus may indicate not only the availability and prevalence of pioneer substrate, occupied by this tree, but also increased waterlogging of the territory and streaming water, which in the Lateglacial could have been a consequence of intensive melting of dead ice. The presence of Corylus in the Lateglacial vegetation may be of particular interest. On the one hand, this fact may be evidence of the existence of mixed forests, where hazel was an undergrowth. On the other hand, another very interesting feature of this tree species must be taken into account: the increasing dominance of hazel occures under unfavorable conditions for other tree species—in a climate that is too cool and humid for other trees (an example is the hazel thicket forests in Norway), or after the death of the forest as a result of natural fires or other natural disasters. The inclusion of hemiboreal species in the picture of the Lateglacial vegetation and the correlation between their ecological characteristics and the existing significant array of other data will therefore help to clarify and, possibly, answer some controversial questions about the evolution of the Lateglacial environment.

The use of the term “hemiboreal” instead of “thermophilous” in relation to such tree species as Picea, Alnus, Corylus, Tilia, Quercus, Carpinus, Ulmus, etc., would help to eliminate the often-unconscious bias in determining their role in the composition of Lateglacial vegetation.

5. Conclusions

The wide spectrum of previously published data (macrobotanical, chironomid), as well as the new results of the Kulikovo palynological study, indicates mild, rather than severe, summers during the Lateglacial in the Baltic region. The reconstructed July temperatures were comparable to modern averages, which range from +16.5 to +18 °C. This could create favorable conditions for the existence of hemiboreal vegetation along with subpolar and boreal plant groups. The hemiboreal vegetation could include the so-called “thermophilous” tree species, which are typical representatives of the present-day hemiboreal woody taxa: spruce (Picea), aspen (Populus tremula), maple (Acer), birch (Betula), rowan (Sorbus), alder (Alnus), beech (Fagus), hazel (Corylus), hornbeam (Carpinus), linden (Tilia cordata), ash (Fraxinus excelsior), elm (Ulmus glabra), and oak (Quercus robur).

The coexistence of subpolar, boreal, and hemiboreal plants with very different climatic tolerances indicates a mosaic pattern and highlights the significant role of local conditions and factors such as shading/illumination, soil types, humidity, slope exposure, relief type, etc. Inclusion of hemiboreal tree vegetation in paleoreconstructions will provide an updated picture of the dynamics of the natural environment as a whole and, in particular, increase the accuracy of paleoclimatic reconstructions based on palynological data.

Attributing the presence of hemiboreal taxa in the pollen records to the contamination of ancient pollen without a targeted study of its sources and mechanisms should be avoided.