Palaeoenvironmental Analysis of the Southern Part of the Danube–Tisza Interfluve (Hungary): The Northern Loess Wall of Katymár and the Hay Meadows and Loess Banks of Hajós

Abstract

This study presents a comprehensive palaeoenvironmental reconstruction of the southern part of the Danube–Tisza interfluve in the Carpathian Basin from the Late Pleistocene to Early Holocene, addressing the region’s response to global climate forcings and local environmental factors based on multi-proxy analyses of two key protected areas: the Katymár brickyard and the hay meadows and loess banks of Hajós. By integrating radiocarbon-dated malacological, macrobotanical, pollen, phytolith, geochemical, and sedimentological analyses, it was possible to provide a picture of past climate–ecosystem interactions. The Katymár North sequence provides an extended chronology from Marine Isotope Stage 3 (MIS3, ca. 36,000 cal BP) through the Last Glacial Maximum (LGM) and into the Early Holocene, while the Hajós sequence offers high-resolution data for the LGM–Holocene transition. By the late Ice Age, humidity and surface moisture superseded temperature as primary palaeoecological limiting factors, promoting ecotone-like forest–steppe environments during cooling periods.

1. Introduction

The Duna–Tisza interfluve, a region characterised by diverse wetland and loess plateau habitats, is undergoing significant environmental changes due to predicted temperature increases in this century. These changes are particularly pronounced in the southern part of the interfluve, encompassing the Katymár brickyard and the Hajós–Kaszálók grasslands within the Kiskunság National Park. These sites, crucial for understanding the impact of climate change on both wetland and terrestrial environments, have experienced a complex history of human intervention and natural restoration efforts.

Previous research [1] has established the significance of the Katymár brickyard as housing the thickest loess sequence in Europe, providing a valuable record of palaeoenvironmental changes over the last 35,000 years. The Hajós hay meadows and loess banks, on the other hand, have been recognised for their unique wetland habitat and peat-forming environment. Building upon earlier investigations of the Katymár West loess–palaeosol sequence [1], this study expands the scope of analysis to encompass the northern section of Katymár, introducing radiocarbon-dated sedimentological, geochemical, phytolith, and malacological analyses. On one hand, a difficult situation was encountered when comparing the analyses conducted in the region, as the basic approach for palaeoecological studies is considered to be conducting multiproxy analysis on the same samples [2]. On the other hand, the basis for comparison between profiles can be provided by sampling with adequate geochronological data and appropriate temporal resolution in fine stratigraphy [2,3,4]. However, despite the large number of processed profiles in the region [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21], comparison problems were posed by the significantly greater geological development thickness of the Katymár profile compared to previously published profiles, or the sporadic occurrence [5,6,7] or complete lack of geochronological data. As a result, the results of samples collected from the northern part of the Katymár brickyard could only be sequentially compared with those of studies conducted in Vojvodina in the Crvenka brickyard [15,19,20] and in the Madaras brickyard [21]. However, due to the extraordinary thickness (1120 cm) of the Katymár profile and the 4 cm sampling, changes could be reconstructed on a century scale, while in the Crvenka profile [19,20], only a millennial-scale approach was allowed by the sampling.

This study aimed to fill the gaps by conducting an integrated environmental historical analysis of the Hajós hay meadows and loess banks and Katymár brickyard. Our approach allowed us to gain specific insights into the impacts of the palaeotemperature changes in the diverse habitats within the Katymár brickyard site. Additionally, this research provides more details on the palaeoenvironmental context of the Neolithic transition in Central Europe [3], based on the Hajós study area, which is a critical period in human history marked by the shift from hunter–gatherer lifestyles to settled agriculture. With these new analyses, this study will provide a detailed understanding of the region’s palaeoenvironmental history, contributing to both climate change research and conservation efforts.

2. Materials and Methods

2.1. Study Areas and Sampling

The initial phase of this study was conducted on a 1120 cm vertical loess profile exposed in a north-facing sequence at the Katymár brickyard (Figure 1). Scaffolds and digging were used during sampling to reach the higher and lower sections, respectively. The wall section was cleaned before samples were collected in 4 cm intervals. From each interval, 1 dm³ (cca. 2.7 kg) sediment was collected for malacological examination and an additional 0.5 kg for other analyses. The material of anthracological examination [22] was extracted during the wet sieving of the malacological sediment. Troels-Smith classification [23] was used to describe the lithostratigraphy during the sampling process.

Katymár is located on the northern edge of the Bácska Loess area in southern Hungary, in which an area of significant sand movement commenced at relatively high altitudes during the drier climatic periods of the Late Pleistocene [24,25]. This aeolian activity contributed to the formation of a varied landscape by the early stage of the last glacial period. During the Middle and Upper Weichselian phases, these sandy areas were subsequently covered by loess deposits, forming horizons that reached a thickness of 1000 to 1100 cm [25,26]. The exposition settings of a sequence, such as the direction and intensity of insolation, play a crucial role in managing local environmental conditions, which in turn influence the species composition of Pleistocene snail fauna [27,28]. This north-facing sequence receives less insolation compared to the west-facing wall [1].

Four cores were extracted from the hay meadows and loess banks of Hajós [3] by a Russian peat corer [29,30,31]. The longest core (430 cm), taken from the deepest point beneath the high bluff, was selected for processing [3]. Substantial lacustrine sediment accumulation was indicated by the core in the middle of the oxbow lake. Toward the edge, the sediment narrowed and transitioned into peat layers due to biogenic infilling.

The Hajós study area is a former Danube River basin located on the eastern edge of the Sárköz region (Figure 2). It is a classic horseshoe-shaped lakebed that formed when a meander of the river was cut off from the main channel. This unique formation is only connected to the main river system during significant floods, functioning as a standalone lake during normal conditions.

The nearby 400–600 cm loess banks played a crucial role in maintaining the lake’s water level between flood events and also created a distinct microenvironment that influenced the vegetation growing on the bluff and near the lake [3]. At the foot of the loess bank, seeping groundwater supplies, tall sedge meadows, swamp meadows, wet hay meadows, fen meadows, and reed beds can be found. Unlike other lowland landscapes, this freely accessible reserve has an outstandingly diverse topography with picturesque landscape architecture [32].

2.2. Walter-Lieth Diagram

A Walter–Lieth diagram [33] was created in climatol 4.1.0 [34] in RStudio 2024.04.1 [35] using data provided by the National Meteorological Service [36,37] for the city of Katymár (110 m above sea level) between January 2015 and December 2023 (9 years) and for the city of Hajós (92 m above sea level) between January 2006 and December 2023 (18 years).

2.3. Radiocarbon Analysis

For the Katymár sequence, dating was conducted on one Holocene organic soil sample, two charcoal samples, and ten Pupilla muscorum samples. The preparation and pre-treatment of the samples followed Sümegi et al. [1]. These measurements were performed at the Hertelendi Laboratory of Environmental Studies (HEKAL), part of the Institute for Nuclear Research of the Hungarian Academy of Sciences in Debrecen, Hungary. Conventional radiocarbon ages for Hajós were derived from previous studies conducted on the same core [3,38,39]. Age-depth modelling (Bacon) and the estimation of the sedimentation rate (accrate.depth) were conducted using RBacon 2.5.8 [40] in RStudio 2024.04.1 [35] and an IntCal20 calibration curve [41,42]. The BP and BC/AD dates used in the text and figures are the calibrated median ages.

2.4. Magnetic Susceptibility, Grain Size, Loss on Ignition, and Geochemical Analysis

Bulk samples underwent magnetic susceptibility analysis [43]. Samples were taken at 4 cm intervals and were first crushed in a glass mortar after weighing. The samples were then placed in plastic boxes and air-dried in an oven at 40 °C for 24 h. The magnetic susceptibilities of the samples were then measured at a frequency of 2 kHz using a Bartington MS2 magnetic susceptibility meter with an MS2E high-resolution sensor [44]. Each sample was measured three times, and the average values of magnetic susceptibility were calculated and reported.

The grain size data of the pre-treated sediment samples [1] were obtained at 4 cm intervals for 42 grain-size classes ranging from 0.1 μm to 500 μm, respectively, using laser diffraction with an OMEC Easysizer20 laser particle size analyser [45]. The Easysizer20 performs all measuring steps automatically after the initial parameter setup, which provides easy operation and accurate results. The U-ratio (16–44 μm/5.5–16 μm) was calculated to determine the dominant and subordinate phases of wind transport during cold–warm and dry–wet periods [46,47,48,49,50]. High U-ratio values indicate more intense sediment transport, which is characteristic of dry periods, while low values may suggest wet periods. Additionally, the grain size index (GSI—20–50 μm/<20 μm) was calculated, which considers the clay fraction and is suitable for determining the formation, transportation, and accumulation of loess [51,52]. The GSI, like the U-ratio, is a measure of climatic conditions during sediment accumulation, but post-accumulation processes such as pedogenesis also play a role. Low GSI values indicate a wet climate or soil formation, while high GSI values indicate dry climatic conditions.

The organic matter, inorganic matter, and carbonate content of the samples were determined using the loss-on-ignition method [53] with a 4 cm sample interval. Initially, the empty ceramic containers were weighed, and then the samples were placed in the crucibles and dried at 105 °C for 24 h in a laboratory furnace. Subsequently, the dried samples in the crucibles were weighed and placed in a furnace again for 1 h at 550 °C to ignite the organic carbon (OM or LOI550). The samples were weighed again after cooling to room temperature. Third, the samples were put into the furnace for 1 h at 900 °C and, after cooling to room temperature, the weight lost was equal to the carbonate content (CC or LOI900). The residuum (IM or LOIres) represents the inorganic matter content [53,54].

The content of the water-soluble elements barium (Ba), aluminium (Al), iron (Fe), manganese (Mn), strontium (Sr), calcium (Ca), magnesium (Mg), sodium (Na), and potassium (K) in the samples was determined using Dr Péter Dániel’s sequential extraction geochemical analysis method with a 4 cm sample interval [55,56].

2.5. Malacological Analysis

All samples from the Katymár site were washed and sieved using a 0.5 mm mesh sieve to remove fine wind-blown sand, loess, and soil material. After sieving, mollusc shells were dried, sorted, and identified under a stereo dissecting microscope at magnifications 6–50×. The malacofauna were divided into different palaeoecological groups following the palaeoecological classifications in [28,57,58,59,60]. The malacological record was also classified according to the recent geographical distribution of the species [61,62,63,64,65,66,67,68] and based on palaeoclimatological indicator roles [58,59,69]. July palaeotemperature values were obtained with the help of a malaco-thermometer [70].

2.6. Anthracological Analysis

The charcoal material from the Katymár site was obtained from uniformly 2.7 kg of 4 cm interval samples [22,71] according to German standards [72]. Charcoal remains were found on three levels on the loess wall. The double flotation method of Gyulai [73] was adopted, using 0.5 and 0.25 mm Ø mesh size sieves. Material was initially passed through the 0.5 mm mesh size sieve. The residue was then passed through the next sieve of a smaller mesh size to retrieve smaller-sized remains as well. Charcoal samples were selected and counted. The charcoal fragment counts are presented following the methodology outlined by Chabal et al. [74]. In the case of floated samples, it made no difference if we counted or weighed the amount of charred wood remains [75]. Charcoals were identified using a polarisation microscope with magnifications of 100×, 200×, and 500×. Taxon identification was carried out using reference books [76,77,78].

2.7. Phytolith Analysis

A personally modified version of the heavy-liquid extraction was adopted in the analysis of the soil and loess samples, developed at the Department of Geology and Paleontology, University of Szeged [79]. In total, 5 g of the samples were air-dried and shaken with the addition of a Calgon solution to remove the organic matter and the carbonates from the samples. This was followed by the removal of the clay fraction and those with a grain size higher than 250 μm. A floatation with a heavy liquid of 2.3 g/cm³ enabled the separation of plant opalites from other non-vegetal quartz grains. The retrieved phytoliths were stored in an Eppendorf tube in glycerine for further study. For the determination process, individual slides were prepared, and opalites were counted at a magnification of 500× under a biological stereomicroscope-type Nikon Eclipse E600 line by line. Counts were double-checked preceding the final quantification of the results. The secondary features of the identified phytoliths were also documented [80].

3. Results

3.1. Walter–Lieth Diagram

The modern climate data provide a baseline for interpreting the past temperature reconstructions. This helps in understanding how current conditions differ from those inferred for various points in the past. The Walter–Lieth diagrams [33] (Figure 3) of the two study areas are almost identical with only small differences in, for example, an average yearly rainfall of only 34 mm, and around ±1 °C differences regarding average temperatures. Katymár has a meteorological monthly data log history dating back to 2014 and is located approximately 3 km away from the study site; meanwhile, Hajós has a meteorological monthly data log history dating back to 2005 and is located about 500 m away from the study site, making them the most suitable meteorological locations for the analysis. The mean annual temperature was 12–13 °C with an annual precipitation of 559–593 mm, respectively. The most probable frost months are from October to April (Katymár) and from September to May (Hajós), when the absolute monthly minimums were equal to or lower than 0 °C. The precipitation graph shows that July is an arid period, with the highest chance for a draught event in both cases, as indicated by the dotted red vertical lines. The period with the highest precipitation is in May (76 mm at Katymár) and May–June (77–85 mm at Hajós). The average minimum temperature was recorded in January (−10.3 °C Katymár and −11.8 °C Hajós), while the average maximum temperature was recorded in August for Katymár (36.4 °C) and July for Hajós (36 °C).

3.2. Radiocarbon Results

3.2.1. Katymár North Loess Wall

The sequence spans approximately 36,000 to 16,000 cal BP (

Table 1. Results of ¹⁴C AMS analysis of the Katymár samples.

Depth (cm)	Lab Code	Material	Uncal BP + σ	(95.4%) cal BP + σ	Modelled cal BP + σ	Modelled cal BP Median
2	DeA–22498	Holocene organic soil	5390 ± 26	6210 ± 60	14,450 ± 600	14,510
82	DeA–22499	Pupilla muscorum shell	13,309 ± 45	16,000 ± 90	16,220 ± 370	16,080
182	DeA–22500	Pupilla muscorum shell	17,178 ± 59	20,730 ± 90	19,230 ± 830	19,430
282	DeA–22501	Pupilla muscorum shell	17,846 ± 71	21,670 ± 150	21,380 ± 300	21,440
296	DeA–23495	Pinus charcoal	17,881 ± 69	21,720 ± 160	21,630 ± 250	21,610
382	DeA–22502	Pupilla muscorum shell	18,655 ± 67	22,590 ± 130	22,660 ± 240	22,610
482	DeA–22503	Pupilla muscorum shell	19,952 ± 78	23,980 ± 100	23,830 ± 290	23,890
582	DeA–22504	Pupilla muscorum shell	19,925 ± 74	23,960 ± 100	24,930 ± 360	24,940
682	DeA–22505	Pupilla muscorum shell	22,119 ± 91	26,300 ± 190	26,220 ± 250	26,210
782	DeA–23497	Pinus charcoal	23,046 ± 94	27,340 ± 100	27,440 ± 250	27,340
882	DeA–22506	Pupilla muscorum shell	24,070 ± 110	28,220 ± 200	28,720 ± 490	28,560
982	DeA–22507	Pupilla muscorum shell	24,917 ± 121	29,120 ± 150	30,610 ± 1040	30,020
1120	DeA–22508	Pupilla muscorum shell	31,598 ± 259	35,920 ± 270	33,890 ± 1780	33,510

, Figure 4), encompassing the late Marine Isotope Stage (MIS) 3, the transition from MIS 2 to MIS 3, the MIS 2 stage, and the Last Glacial Maximum (LGM). This age interval is broadly similar but approximately 3000 years younger at the base compared to the western section of the Katymár brickyard, which dates to around 39,000 cal BP [1]. The MIS 3/MIS 2 transition is recorded between 26,000 and 24,000 cal BP [1]; in contrast, we propose that this transition at Katymár North occurred between approximately 28,200 and 25,000 cal BP. The LGM is defined as occurring between approximately 24,800 and 19,000 cal BP, during which the average July palaeotemperature is estimated to have dropped to nearly 10 °C. Within this period, Greenland Interstadials GI-2.1 and GI-2.2 are characterised by relatively warmer temperatures, with average July palaeotemperatures of 15.5 °C and 14 °C, respectively. Greenland Stadial GS-2.2, which occurs between these two warmer interstadials, is characterised by a lower average July palaeotemperature of 11.5 °C. The unmodelled cal BP and modelled cal BP radiocarbon dates (Table 1) show similar values and are correlated with GISP2 stadials and interstadials [81] with Heinrich events and Dansgaard–Oeschger cycles [82,83,84] (Figure 4). The calibrated ages of Sample DeA–22498 (2 cm) at 6210 ± 60 years and Sample DeA–22499 (82 cm) at 16,000 ± 90 years are accurate; however, the calibrated ages between these two depths exhibit reduced accuracy and are not usable. Furthermore, the ages at 982 cm (DeA–22507) and 1120 cm (DeA–22508) are accurate, although the modelled, calibrated, and correlated ages between these depths exhibit a lower degree of precision; however, they remain within an acceptable range for use.

The sediment accumulation rate initially progresses slowly for approximately 6000 years. This period is succeeded by a phase of slightly increased accumulation from approximately 29,000 to 28,000 cal BP. Subsequently, the accumulation rate accelerates until around 21,700 cal BP, which represents the interval of greatest sediment deposition (approximately 600 cm). Following this peak, the accumulation rate decreases over a thousand years. This trend of reduced accumulation continues until around 16,000 cal BP, after which the rate of accumulation increases once more, continuing to the surface.

3.2.2. Hajós Hay Meadows and Loess Banks

The record spans from the Late Pleistocene (approximately 25,500 cal BP) to the present day, providing an almost continuous archive of environmental change through the Late Pleistocene and Holocene (

Table 2. Results of ¹⁴C AMS analysis of the Hajós samples.

Depth (cm)	Lab Code	Material	Uncal BP + σ	(95.4%) cal BP + σ	Modelled cal BP + σ	Modelled cal BP Median
100	Deb–9329	Peat	1650 ± 80	1540 ± 90	1640 ± 256	1586
165	Deb–9326	Peat	1770 ± 80	1660 ± 90	2989 ± 661	2906
220	Deb–9328	Peat	3960 ± 80	4410 ± 130	4858 ± 578	4686
236	ETH–41278	Typha	5560 ± 40	6350 ± 40	6306 ± 144	6332
270	Deb–9325	Peat	7310 ± 80	8130 ± 90	8132 ± 231	8127
296	Deb–9327	Peat	9130 ± 130	10,310 ± 180	10,070 ± 481	10,179
344	ETH–41275	Peat	10,540 ± 45	12,570 ± 70	12,445 ± 244	12,550
420	GdA–554	Shell	21,190 ± 140	25,520± 160	17,290 ± 2003	16,953

, Figure 5). The bottom of the Hajós sequence is around the MIS3/MIS2 transition or the start of MIS2 (430–420 cm). The sedimentological characteristics of the section between 430 and 420 cm differ significantly from the subsequent layer (420–296 cm). This distinct change suggests that the ages corresponding to the 420–297 cm section are likely closer to the modelled ages starting from 296 cm, rather than being a continuation of the age at 420 cm. Immediately above 420 cm, there is an abrupt jump of approximately 8000 years in the chronology. This would place the age from around 419 cm at ca. 17,000 cal BP. The sequence has 10 cm of MIS3/MIS2 or early MIS2 sediment (430–420 cm) and 20 cm of late-MIS2 sediment (420–400 cm) as well.

From around ca. 17,000 cal BP at 400 cm, date-wise, the sequence continues where the Katymár North loess wall sequence left off (around 16,000 cal BP at 82 cm), but ca. 40 km away in an oxbow lake environment, and right next to a loess wall. The Bølling–Allerød Interstadial [81,85], a major warm and moist phase, is represented approximately between 380 and 350 cm. The 11.4-kiloyear event [81,85], a brief cold spell at the onset of the Holocene, can be located around a depth of ca. 324–320 cm in the core. The 9.3-kiloyear cooling event [81,85,86,87] can be approximately located at around 284 cm in depth. The 8.2-kiloyear event [81,85,86,88,89,90], a widely studied cooling episode, can be identified at ca. 272–268 cm. The 4.2-kiloyear drought event [81,85,86,91,92] can be identified between ca. 204 and 220 cm.

3.3. Sedimentological and Geochemical Results

3.3.1. Katymár North Loess Wall

The data (

Table 3. Statistics of LOI, MS, grain size, and geochemical results of the Katymár North sequence.

Factors	Units	Minimum	Average	Median	Maximum	Standard Deviation
Organic material	%	0.54	1.68	1.51	7.29	0.75
Carbonate content	%	4.96	12.05	12.89	14.81	2.38
Inorganic material	%	83.61	86.27	85.72	93.18	2.09
Magnetic susceptibility	10−6 m3kg−1	14.9	24.18	24.2	67.9	6.69
Clay	%	1.18	18.39	18.71	23.89	2.92
Silt	%	49.8	70.32	70.23	82.87	4.52
Sand	%	2.15	11.28	10.84	38.03	5.34
Calcium (Ca)	ppm	12.8	391.87	345.88	1355.3	235.71
Magnesium (Mg)	ppm	1.15	111.43	103.23	1173.13	87.98
Sodium (Na)	ppm	0.01	1.22	0.77	10.4	1.25
Potassium (K)	ppm	0.04	1.24	1.15	4.83	0.81
Iron (Fe)	ppm	0.3	17.52	15.86	96.55	12.68
Manganese (Mn)	ppm	0.01	1.34	0.71	146	8.7
Aluminium (Al)	ppm	0.07	10.52	8.86	42.1	7.71
Barium (Ba)	ppm	0	0.15	0.1	1.77	0.22
Strontium (Sr)	ppm	0	0.5	0.41	2.41	0.4

, Figure 4) suggest an inverse relationship between carbonate and inorganic material content, as evidenced by their complementary ranges, with the only exception at around the surface from ca. 100 to 0 cm. The organic matter content starts off with an increasing trend, only to remain consistently low, up until the modern-day topsoil layer where it reaches its maximum of 7.29%. Carbonate content fluctuates within a small range as well, generally remaining around 12–13%. The dominance of inorganic matter highlights the significant role of physical processes, especially aeolian transport, in shaping the depositional environment (Figure 5 and Figure 6). From 1120 cm to ca. 548 cm, the analysed elements show rapid changes and fluctuations (Figure 7). Some peak around palaeosol levels (296 cm, 780–788 cm, 880–888 cm) where charcoal remains were found, like magnetic susceptibility: Ba, Al, Ca, Sr, Na, and K. After, up until 142 cm, more steady geochemical element values with slow and extended changes can be observed. Mn stays at a minimum up until 142 cm, where it peaks, after which Ca, Mg, Sr, Na, and K elements start to increase and fluctuate more again. The U-ratio shows an increasing trend, along with the appearance of very fine and medium sand grain sizes (ca. 700–220 cm) (Figure 4 and Figure 6), with a lot of peaks from ca. 900 to 300 cm, up until the middle of the LGM, from where the U-ratio decreases toward the surface. In the top 20–30 cm, the U-ratio, MS value, and organic matter content peak, while carbonates and Mn, Sr, and Na elements decrease. Longer, finer cycles and shorter loess cycles are observed at the bottom. From ca. 24,000 cal BP onward, the loess cycles become longer, the finer-grained cycles become shorter.

3.3.2. Hajós Hay Meadows and Loess Banks

The data (

Table 4. Statistics of LOI, MS, grain size, geochemical, and pollen [3] results of the Hajós sequence.

Factors	Units	Minimum	Average	Median	Maximum	Standard Deviation
Organic material	%	0	50.85	70.9	97.5	35.1
Carbonate content	%	0	2.97	1.5	8.4	2.43
Inorganic material	%	0.8	46.18	25.6	100	36.15
Magnetic susceptibility	10−6 m3kg−1	7	37.29	38	71	20.02
Clay	%	0	10.61	13.15	27.6	6.42
Silt	%	51.5	87.77	86.45	100	7.21
Sand	%	0	1.63	0	48.5	2.36
Calcium (Ca)	ppm	11	3086.43	1905.5	8514	2415.27
Magnesium (Mg)	ppm	8	1027.02	642	2855	808.62
Sodium (Na)	ppm	3	328.27	205.44	913.6	257.14
Potassium (K)	ppm	3	241.49	112.29	721.74	213.87
Iron (Fe)	ppm	7	366.53	386.93	852	173.54
Manganese (Mn)	ppm	0	134.52	142.59	315.24	65.63
Arbour pollen (AP)	pieces	166	303.6	275	567	115.71
Non-arbour pollen (NAP)	pieces	9	238.26	267	467	117.26
Aqua + Spore	pieces	0	32.16	35	60	13.29
Total pollen count	pieces	477	574.02	572	670	45.65

, Figure 5 and Figure 8) suggest an inverse relationship between organic and inorganic material content, as evidenced by their complementary ranges, although only from 300 cm, from which the organic matter content increases from 1.12% to 49.8%. This also marks the second interval between 298 and 280 cm, where carbonate content reaches 0%. This exact interval is where Ca, Mg, and Na contents decrease, while Fe and Mn contents peak (Figure 9). After 280 cm, carbonate content is positively correlated with Ca, Mg, Na, and K content fluctuations.

The Polish bottom lake classification system (Figure 8) [93,94], which is based on loss-on-ignition data (Figure 5, Table 4) [3], was used to differentiate between the phases and provide a clearer understanding of the evolution and development of the oxbow lake-mire system. With this system, it is possible to interpret the LOI data to describe the sediment types, especially if paired with the Troels-Smith classification [23], grain size, magnetic susceptibility, and geochemical analysis (Figure 9) [95,96].

SMZ-1 (430–420 cm, around 25,500 cal BP): The bottom layer is characterised by cross-bedded fluvial fine sand layers, indicating a riverine environment. No pollen or macrobotanical remains, but it contains shell and snail fragments, suggesting a well-oxygenated, shallow-water habitat [3]. Sand and coarse silt grain sizes dominate with high inorganic matter and low carbonate content. The lowest magnetic susceptibility value in the sequence. No organic matter or carbonate content, 100% inorganic matter.

SMZ-2 (420–296 cm, ca. 17,000–10,200 cal BP): A shift to a lacustrine environment, as evidenced by the fine silt, clay content, and lack of riverine sand. This suggests that the oxbow lake was cut off from the main river channel, leading to calmer waters and the deposition of finer sediments (fine, medium, and coarse silt). During this period, the magnetic susceptibility values, as well as the carbonate content, stay relatively stable and show a stable lacustrine environment. From ca. 15,200 cal BP (388 cm), all the water-soluble geochemical elements increased noticeably.

SMZ-3 (296–275 cm, ca. 10,200–9000 cal BP): Between 296 and 280 cm, a carbonate-free zone with reduced Ca, Mg, and Na content, along with increased organic matter and peak Fe and Mn content. This shows the start of the biogen infilling of the oxbow lake, where it became a marshy wetland habitat with “brown moss carpet” [97], and where mainly local and extra-local pollen grains accumulated [3]. An increase in clay and very fine silt grain sizes and a decrease in medium silt grain sizes.

SMZ-4 (275–170 cm, ca. 9000–1700 cal BP): The evolution of the catchment basin changed into a peatland, rich in pollen and macrobotanical remains. The sediment type also changed to coarse detrital gyttja and algal gyttja, indicating the continuous dominance of the organic matter with at least 70% organic matter content. Altogether, eight peaks of organic matter content can be described (more than 80% of organic matter: 268–266 cm, 260–256 cm, 234–230 cm, 224 cm, 218–212 cm, 208–200 cm, 194–190 cm, 186–184 cm), which might have had better conditions during vegetation periods. Low geochemical element values along with low clay and very fine silt grain size values. Increased magnetic susceptibility values. A decrease in the NAP count throughout the zone.

SMZ-5 (170–14 cm, ca. 1700–310 cal BP): An increase in the pollen count, carbonate content, and Ca, Mg, Na, K, Fe, and Mn element contents. Peatland phase with high organic matter content (algal and coarse detrital gyttja), especially between 104 and 78 cm (79.6–95.6%) and 20–16 cm (80.3–97.5%). A negative peak in carbonate content along with Ca, Mg, Na, and K content, and an increase in organic matter content between 98 and 90 cm. Low fluctuation in clay and very fine silt grain size values. Another increase in magnetic susceptibility values with small changes and a drop at the end of the zone along with carbonate and pollen count (477 pieces at 14 cm).

SMZ-6 (14–0 cm, ca. 310 cal BP to present): Fast transition from algal gyttja to coarse, and then from fine detrital gyttja (4) to clay detrital gyttja (5), which means a decrease in organic content, and an increase in inorganic and carbonate content. An increase in clay, very fine silt, and very fine, fine, and medium sand grain sizes near the surface.

3.4. Malacological Results

3.4.1. Hajós Hay Meadows and Loess Banks

The living water of Hajós lacked macrophytes, was well oxygenated, and was relatively rich in nutrients as indicated by the number of rheophilic molluscs (Valvata piscinalis, Valvata naticina, Lymnaea stagnalis, Planorbis cf. carinatus, Unio cf. crassus, Pisidium amnicum) from the base sand-rich layer between 430 and 420 cm [3].

3.4.2. Katymár North Loess Wall

The malacological analysis (Figure 10 and Figure 11) of the Katymár North loess wall site reveals distinct ecological shifts over the past 36,000 years, reflecting changes in climate, vegetation, and biogeography. This sequence shows similarities to the Madaras sequence [21] as well as the Katymár West profile [3]. The malacological record provides high-resolution insights into local climatic conditions [3,21], complementing regional environmental reconstructions [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Based on the comprehensive analysis of mollusc assemblages, temperature preferences, humidity requirements, vegetation affinities, and biogeographical distributions [57,58,59,60,61,62,63,64,65,66,67,68,69,70], malacological zones (MLZ) were set (

Table 5. Malacological zones (MLZ) for the Katymár sequence.

Zone	Depth (cm)	Age (cal BP)	Environment	Dominant Groups	Dominant Species
MLZ-1	1116–856	36,000–28,400	Warm steppe, forest–steppe	Mesoph., Xeroph. SSE EU, Holarctic	Granaria frumentum, Pupilla triplicate, Chondrula tridens
MLZ-2	856–624	28,400–25,400	Cold steppe	Mesoph., Xeroph. SSE EU, W EU, EU, Holarctic	Pupilla muscorum, Pupilla triplicata, Chondrula tridens, Vallonia costata
MLZ-3	624–408	25,400–23,000	Cold steppe	Mesoph., Hygroph. EU, W EU, Holarctic	Vitrea crystallina, Punctum pygmaeum
MLZ-4	408–12	23,000–6200	Cold open habitat	Most diverse groups by humidity and biogeography	Vallonia tenuilabris, Pupilla sterri
MLZ-5	12–0	6200–	Warm steppe, forest–steppe	Mesoph., Xeroph. SSE European, Holarctic	Granaria frumentum, Pupilla triplicata

and

Table 6. Average July palaeotemperature [70] statistics for the Katymár sequence.

Zone	Units	Minimum	Average	Median	Maximum	Standard Deviation
MLZ-1	°C	16.14	18.51	18.55	20.03	0.98
MLZ-2	°C	12.20	16.69	16.75	18.62	0.81
MLZ-3	°C	11.71	15.08	15.25	16.76	0.88
MLZ-4	°C	10.46	12.62	12.46	16.04	1.29
MLZ-5	°C	19.52	19.78	19.85	19.92	0.18

).

An increase in species diversity from around 998 cm can be observed (Figure 10 and Figure 11). In the MLZ-2, the appearance and peak of Fruticicola fruticum and Perpolita hammonis, alongside Vallonia costata and Chondrula tridens, suggest a more humid environment (Table 5 and Table 6). The disappearance of Succinella oblonga, Cochlicopa lubrica, and Granaria frumentum indicates a significant ecological shift, possibly related to the onset of colder conditions preceding the LGM. The dominance of mesophilous species increase up to 96% in the MLZ-3, and warm-loving species slowly withdraw and give place to clod-resistant species around the start of the LGM (572 cm) (Table 5 and Table 6). The transformation is indicated by the afforestation that occurred during the LGM, with shade-loving snail species (Vitrea crystallina, Punctum pygmaeum, Vitrina pellucida) reaching a maximum of 14% and ecoton species taking over the dominance (at 624 cm) with an average of 75% and maximum of 93%. The disappearance of several species (e.g., Pupilla muscorum, Pupilla triplicata, Vallonia costata, Chondrula tridens, Fruticicola fruticum, and Helicopsis striata) and the emergence of others (e.g., Vitrea crystallina, Orcula dolium, Punctum pygmaeum, and Vitrina pellucida) reflect the harsh conditions of the LGM. The peak in pollen count at 600 cm suggests a brief period of increased vegetation cover, as well as starts a transition. The MLZ-4 encompasses the second half of the LGM, the Late Glacial period and the Bølling–Allerød interstadial and Younger Dryas stadial. The malacofauna show rapid fluctuations in species composition, reflecting the climatic instability of this period. Succinella oblonga, Columella columella, Vallonia tenuilabris, and Clausilia dubia appear again and stay until the end of the zone. The reappearance of thermophilous species like Chondrula tridens, alongside cold-tolerant taxa, suggests a dynamic environment responding to abrupt climate changes and suggests increasing habitat diversity (Pupilla sterri). The pollen count decreases and reaches a count of ca. <200 from 140 cm onwards, from which point the temperature starts to increase again. Granaria frumentum (max. 18%), Pupilla triplicate (82%), Vallonia pulchella (2%), and Chondrula tridens (12%) peak at the surface (16–0 cm), but the MLZ-5 is species and individual poor. The Holocene malacofauna indicate a return to steppe-like conditions, raising questions about the relative roles of climate and human activity in shaping the modern landscape (Figure 10 and Figure 11, Table 5 and Table 6).

3.5. Phytolith and Anthracological Results

Phytoliths from the northern wall of the Katymár brickyard indicate a simple, almost homogeneous, species-poor steppe environment, with the soil levels primarily suggesting an open parkland or forest–steppe vegetation cover (Figure 12,

Table 7. Results of the phytolith and anthracological results with modelled cal BP ages.

Zone	Depth (cm)	Age (cal BP)	Main Characteristics	Plant Remains
PTZ-1	1120–1100	36,000–34,800	Elongate long cells, elongate long-cell skeletons, non-grass larger cells, a small number of parallelepipedal bulliforms and cuneiform bulliforms	Pinus, Betula, burnt Pinus
PTZ-2	1100–892	34,800–28,700	Abundant only in elongate long cells and sponge spicules	–
PTZ-3	892–870	28,700–28,550	A decrease in elongate long cells, sponge spicules drop to 0, and an increase in other forms	Pinus, burnt Pinus
PTZ-4	870–792	28,550–27,450	Abundant only in elongate long cells and sponge spicules	–
PTZ-5	792–776	27,450–27,250	A decrease in elongate long cells, sponge spicules drop to 0, and an increase in other forms	Pinus, burnt Pinus
PTZ-6	776–300	27,250–21,650	Abundant only in elongate long cells and sponge spicules	–
PTZ-7	300–296	21,650–21,550	A decrease in elongate long cells, sponge spicules drop to 0, and an increase in other forms	burnt Pinus
PTZ-8	296–20	21,550–6200	Abundant only in elongate long cells and sponge spicules, declining from 100 cm	–
PTZ-9	20–0	6200–	Sponge spicules drop to 0, and various phytolith forms increase	Quercus at 0 cm

). This interpretation is supported by the pollen profile from Hajós [3], which reveals the presence of species-poor steppes and open parklands at the end of the glacial period. The data (Figure 12, Table 7) show a cyclical pattern alternating between periods dominated by elongate long cell and sponge spicules (PTZ-2, 4, 6, 8) and periods with more diverse phytolith forms (PTZ-1, 3, 5, 7, 9) in the palaeosol layers and modern-day soil. Although, like in the western part, Katymár lacks well-developed palaeosol layers [1]. Pinus and Betula remains are present in the palaeosol layers; meanwhile, Quercus remains were found at the surface (Table 7). The presence of burnt Pinus remains in several zones indicates natural fires. The core can be separated into three main compositions: one that consists of elongate long cells, elongate long-cell skeletons, non-grass larger cells, parallelepiped bulliforms, and cuneiform bulliforms with greatly reduced sponge spicules, which correlates with the magnetic susceptibility peaks and the palaeosol levels (PTZ-1, 3, 5, 7); another one with only elongate long cells and sponge spicules (PTZ-2, 4, 6, 8), for which the composition is dominant throughout the sequence; and the last one near the surface, which is similar to the first composition (PTZ-1, 3, 5, 7), but the elongate long cells are minimal and with new phytolith forms—rectangle scorbiculate, orbicular scorbiculate, and trichome (PTZ-9).

4. Discussion

The comparison of the Hajós and Katymár profiles was made possible by radiocarbon measurements. However, even before these measurements, the Danube Valley was considered to be between 25,000 and 35,000 years old [98,99,100,101,102]. This dating was based on the understanding that during this period, the Danube River slid off the alluvial fan that had developed in the Danube–Tisza interfluve and incised into its current location. The channel exposed at Hajós, which became marshy during the Holocene, belonged to the oldest generation of Danube riverbeds. This channel dates back to the MIS3/MIS2 boundary. The multi-proxy analyses of the Katymár North loess–palaeosol sequence and the Hajós hay meadow and loess bank cores provide compelling evidence for the climatic and environmental evolution of the Carpathian Basin over the Late Pleistocene and Early Holocene. These two sequences offer complementary insights into the region’s palaeoenvironmental history, demonstrating both shared sensitivities to global climate forcings and site-specific responses to local environmental conditions.

The Katymár North sequence presents an extended chronology, spanning from the latter part of Marine Isotope Stage 3 (MIS3) through the Last Glacial Maximum (LGM), Late Glacial period, and into the early Holocene. This comprehensive record provides a broader context for understanding the long-term climate variability and environmental dynamics of the region. In contrast, the Hajós sequence focuses on a shorter timeframe, capturing the transition from the LGM through the Late Glacial interstadial and into the Holocene, including its early climatic instabilities. For Hajós, the transition from fluvial deposits to lacustrine sediments and finally to peat layers reflects not only local [3,38,39] geomorphological processes but also broader regional changes in hydrology [95,98,99,100,101,102].

Both sequences clearly demonstrate the high responsiveness of the Carpathian Basin to global climate forcings, particularly during periods of rapid change. The distinct signal of the Bølling–Allerød Interstadial, evident in both records, exemplifies the synchronous regional response to large-scale climate oscillations. This correlation between the two sequences not only validates their individual findings but also underscores the coherent nature of climate-driven environmental changes across the Carpathian Basin. The comparison of these two sequences highlights the value of multi-site studies in palaeoclimatological research. While the overall trends align, subtle differences between the sites offer insights into local factors influencing environmental responses.

Due to the malacofauna strongly tied to the local environment in the area, it can primarily be compared to the Serbian Vojvodina [15,19,20] and Croatian Baranja [103,104], which can be considered continuations of the loess area. However, a significant problem with this comparison is that our analysis, based on finer sampling, has a century-scale time resolution, while the Serbian and Croatian profiles have millennial-scale malacological analyses due to coarser sampling. A global comparison could be made with the malacological materials of Chinese loess profiles [105,106]. The results clearly show that in the southern part of the Carpathian Basin, at the end of the Pleistocene, a different faunal development took place compared to the rest of the basin. Based on the continuous presence of thermophilic steppe-dwelling faunal elements, it can be assumed that the temperate grassland habitat [60,66,67,68,107,108,109], even if in patches, persisted in the studied region during the most significant glaciations (cooling periods).

The same changes observed in quartermalacological data were also detected in the southern Great Hungarian Plain profiles [45,58,59]. A key characteristic is that during significant warming periods (MIS2/MIS3 transition, early Holocene), soil formation processes were accompanied by the appearance of Cepaea vindobonensis, a characteristic taxon of Pannonian forest–steppe environments [110,111,112,113,114,115,116]. Károly Bába’s works [114,115,116] highlighted the recent relationship between the temperate forest–steppe vegetation and snail fauna in the southern Great Hungarian Plain. Consequently, Cepaea vindobonensis is the characteristic element of the final malaco-successional community of the Pannonian –steppe, which developed under the influence of the regional climate resulting from the basin situation [117], rather than Drobacea banatica (Helicigona banatica [20]). The regional connection is established by the cyclical appearances of Granaria frumentum, which is also characteristic of temperate forest–steppe environments, during warming periods. This taxon only appears in the southern part of the Carpathian Basin during the MIS2 time horizon [58,59,118], primarily during warming phases [58,59]. In the LGM horizon, temperate Pannonian forest–steppe elements appear in subordinate proportions during the so-called micro-interstadial phases [119,120,121]. This characteristic is unique to the central and southern parts of the Great Hungarian Plain [1,45,69]. Thus, both geochemical and malacological data from the Katymár loess profile indicate the development of an independent regional role for malacological changes observable only within this region, signalling the formation of an Ice Age ecoregion in the central and southern parts of the basin. This is further supported by malacological changes observed only in the region, such as the last spatial occurrence of cold-loving, xeromontane Asian elements in the basin, which do not appear further south [58]. Moreover, the finer temporal changes in the malacofauna already differ from those in the southern part of Transdanubia. However, alongside the spatial pattern, the temporal pattern of changes and faunal composition is characteristic of this ecoregion. A prominent feature of the ecoregion, as revealed by local data [1,45,69], is that during the LGM horizon, in the Upper Palaeolithic and then Mesolithic levels, an extremely species-rich fauna characteristic of mesophilic boreal and temperate forest–steppe developed. This faunal horizon then evolved into a hygrophilous forest–steppe fauna rich in forest-dwelling elements at the end of the LGM and the beginning of the post-LGM period. A notable characteristic of the profile is that following the faunal horizon associated with the micro-interstadial in the LGM level, a merging temperature minimum can be reconstructed at the end of and following the LGM [25]. Similar changes were detected in the region based on pollen, anthracological, and phytolith remains [1,45,69,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133].

During the late glacial cooling, various species-rich boreal forest–steppes developed, with the main differences being the proportion of tree-covered areas (boreal open parkland, parkland, boreal forest–steppe) and the composition of woody vegetation [130]. During warming periods, a cascade-type change [134] led to a decrease in species numbers in both vegetation and malacofauna. Based on malacofaunal changes, the number of temperate forest–steppe taxa is about one-third of the number of taxa detectable in the late glacial mixture of boreal steppe/forest–steppe/forest patches (ecotone [117,135]). As a result, in the southern and central parts of the Carpathian Basin, temperature was no longer the most significant palaeoecological limiting factor for biota at the end of the Ice Age, but rather humidity and surface moisture. Thus, during Quaternary cooling periods, humidity increased, favouring the species and individual numbers of ecotone-like forest–steppe intermittently mixed with tundra and temperate forest–steppe elements. During warming periods, the ecotone structure remained, but Boreo-Alpine, Asian xeromontane, and Central European montane elements were displaced, leading to a decrease in species numbers, while trees and herbaceous plants characteristic of temperate forest–steppe spread across this ecotone vegetation structure. The most prominent feature of the late glacial ecoregion in the studied area, including the examined profiles, was that in contrast to the northern and western parts of the basin where Picea (spruce) was the most important woody taxon, in this region, Pinus sylvestris, Betula, and Abies taxa appeared. Thus, on the southern edge of the extremely heterogeneous Eurasian loess zone, the central and southern parts of the Carpathian Basin form an independent mosaic (ecoregion) due to the basin’s climatic influence and the Danube alluvial fan background. This can be imagined as if the Eurasian loess belt (zone) were a parquet floor (parquet–mosaic, fragment, fractal element from an environmental history perspective [136]) composed of fragments (fractal elements) with independent developmental histories covering several million km², and one independently developing part of about 40–50 thousand km² is located in the central and southern part of the Carpathian Basin. The developmental history of this fractal element bears characteristic differences even within the loess belt of the Carpathian Basin and is markedly distinct from the SE European and other European loess belt regions in terms of its developmental moments.

4.1. MIS3 Level (ca. 36,000–29,000 cal BP)

The malacological assemblage during Marine Isotope Stage 3 (MIS3) at the Katymár North loess wall site is characterised by a diverse fauna dominated by xerophilous and thermophilous species. The presence of Granaria frumentum, Pupilla triplicata, and Chondrula tridens suggests a predominantly open, steppe-like environment with patches of shrubland or sparse woodland (Figure 10 and Figure 11). This assemblage is indicative of relatively warm and dry conditions, with estimated July palaeotemperatures ranging from 16 to 20 °C. The occurrence of Vallonia costata in the accompanying fauna further supports the interpretation of a mosaic landscape, where open grasslands were interspersed with areas of denser vegetation. The gradual increase in species diversity observed from around 998 cm in depth (corresponding to the latter part of MIS3) indicates a shift toward more favourable environmental conditions. This diversification may reflect an increase in habitat heterogeneity, possibly due to localised increases in moisture availability or the development of more complex vegetation structures. Geochemically, this level is characterised by elevated levels of Ba, Al, and Fe, suggesting enhanced weathering and soil formation processes, which is corroborated by the lower carbonate content (Ca and Mg) compared to later, colder periods [137,138,139,140,141,142,143,144,145,146] (Figure 7). The organic content, while generally low throughout the sequence, shows slight increases during this period, further supporting the interpretation of a more productive ecosystem (Figure 4). The U-ratio is lower, and magnetic susceptibility is higher, indicating finer particle deposition and potential pedogenesis, consistent with more stable, warmer conditions (Figure 4 and Figure 6).

4.2. MIS3/MIS2 Transition (ca. 29,000–25,000 cal BP)

The transition from MIS3 to MIS2 is marked by gradual changes in multiple proxies, reflecting the shift toward colder and drier conditions. This period is characterised by the dominance of the Pupilla muscorum palaeoassociation, indicating a general cooling trend with July palaeotemperatures dropping to approximately 15–17 °C (Figure 10 and Figure 11). The appearance and peak abundance of Fruticicola fruticum and Perpolita hammonis, alongside the persistence of Vallonia costata and Chondrula tridens, suggest an increase in environmental humidity and the development of more mesic habitats. A notable feature of this transitional period is the disappearance of several species that were prominent during MIS3, including Succinella oblonga, Cochlicopa lubrica, and Granaria frumentum. This faunal turnover reflects a significant ecological shift, likely driven by the increasingly cold conditions preceding the LGM. The persistence of some thermophilous elements, such as Chondrula tridens, alongside the emerging cold-tolerant fauna, indicates a complex environmental mosaic during this transition. This pattern suggests that while the regional climate was cooling, local topographic and microclimatic factors may have allowed for the maintenance of diverse habitats, supporting a mixed malacofauna adapted to varying environmental conditions. Geochemically, this transition is characterised by decreasing trends in Ba, Al, and Fe, indicating reduced weathering intensity. Conversely, Ca and Mg concentrations begin to increase, signalling enhanced carbonate content in the sediments (Figure 7). The U-ratio starts to increase while magnetic susceptibility decreases, reflecting a shift toward coarser particle deposition typical of loess accumulation under colder, windier conditions (Figure 4 and Figure 6). These changes collectively point to the onset of more severe glacial conditions, with reduced biological productivity and increased aeolian activity.

4.3. MIS2 Level (ca. 25,000–14,000 cal BP)

The most significant cooling events in the Carpathian Basin occurred during the LGM, with the snail fauna (Figure 10 and Figure 11) indicating temperatures of 10–12 °C. However, we must acknowledge that we collected the fauna on the northern side, which did not receive sunlight, unlike the already published western side [1]. We hypothesise that the cooling was accompanied by an increase in humidity, leading to the development of greater vegetation cover (maximum dominance of shade-loving, Holarctic snail species). The MIS2 level displays clear signatures of a cold and dry environment and reflects the harshest climatic conditions of the sequence. The early part of this period sees a marked increase in mesophilous species, while warm-loving taxa gradually withdrew (Figure 11). The disappearance of species such as Pupilla muscorum, Pupilla triplicata, and Vallonia costata, coupled with the emergence of cold-tolerant taxa like Columella columella and Vallonia tenuilabris, clearly demarcates the onset of full glacial conditions. As the LGM progresses, the malacofauna shows signs of adaptation to extreme cold and dry conditions (Figure 10). The appearance and increasing abundance of species such as Vitrea crystallina, Orcula dolium, Punctum pygmaeum, and Vitrina pellucida suggest the development of patchy, humid microhabitats, possibly associated with sparse vegetation cover or sheltered locations. The transformation from 600 cm is indicated by the afforestation, with shade-loving species reaching a maximum dominance of over 20%. The fauna composition at the end of the Ice Age was blurred, with no marked difference between the LGM and post-LGM snail fauna. The maximum dominance of the Boreo-Alpine Columella columella, the South and North Asian Vallonia tenuilabris, the Central European mountainous Clausilai dubia, the Palearctic Succinella oblonga, and the Western European Trocholus hispidus developed. However, in levels younger than 18,000 years (cal BP), the appearance of SE European thermophilic species began, along with a parallel decrease in the combined proportion of cold-loving species. This process culminated in the Holocene soil level. Geochemically, this period is marked by low concentrations of Ba, Al, and Fe, reflecting minimal weathering and soil formation [137,138,139,140,141,142,143,144,145,146] (Figure 7). In contrast, Ca and Mg levels reach their peak, indicating maximum carbonate content in the loess deposits. The U-ratio is at its highest, while magnetic susceptibility is at its lowest, clearly indicating an intense loess deposition of coarser particles under cold, arid, and windy conditions (Figure 4 and Figure 6). Organic content is minimal, suggesting very low biological productivity. The accumulation rate shows a slight increase during this period, consistent with enhanced loess deposition under glacial conditions (Figure 4).

During the MIS2 level at the Hajós site, corresponding to the Last Glacial Maximum and early deglaciation, the sedimentary record indicates a predominantly cold and dry climate. A dominance of coarse silt and fine sand is observed, suggesting aeolian deposition under periglacial conditions. The environment shifts to lacustrine, as evidenced by the fine silt, clay content, and lack of riverine sand (Figure 5, Figure 8 and Figure 9). This suggests that the oxbow lake was cut off from the main river channel, leading to calmer waters and the deposition of finer sediments. The macrobotanical record [3] for this period is notably sparse, dominated by Monocot. Undiff., U.O.M, a low amount of Phragmites, and sporadic occurrences of Equisetum cf. fluitantis. The pollen assemblage [3] during this time was overwhelmingly dominated by the Pinus subgenus Diploxylon pollen type (Pinus sylvestris), but other cold-tolerant taxa were also present like Betula, or steppe elements like Artemisia and Poaceae, indicating that the landscape was not uniformly forested but rather a complex mosaic of boreal forest and cold steppe vegetation. A significant decrease in Pinus pollen and a concurrent increase in deciduous trees indicate the slow replacement of a boreal forest–steppe mosaic with the mixed deciduous forest consisting of Betula, Ulmus, Quercus, and Fraxinus [3]. This mosaic structure likely provided a variety of habitats for the Pleistocene fauna and potentially for human populations of the Upper Palaeolithic.

4.4. Late Glacial Level (14,000–11,700 cal BP)

The Late Glacial period in Katymár is characterised by rapid and significant changes across all proxies, reflecting the transition out of full glacial conditions. The reappearance and increasing abundance of species such as Succinella oblonga, Columella columella, and Vallonia tenuilabris indicate a gradual amelioration of climatic conditions (Figure 10). This period is marked by fluctuations between warmer interstadials (e.g., Bølling–Allerød) and cooler stadials (e.g., Younger Dryas), which are reflected in the alternating dominance of more thermophilous and cryophilous species, respectively (Figure 11). The emergence of Pupilla sterri and the return of Chondrula tridens during this period suggest an increase in habitat diversity and the development of a mosaic landscape. The decrease in pollen count observed from around 140 cm depth, coupled with rising temperatures, may indicate a shift toward more open vegetation structures, possibly due to increased seasonality or changes in precipitation patterns. The malacological record during this period provides valuable insights into ecosystem responses to rapid climate change, demonstrating the sensitivity of terrestrial mollusc communities to environmental fluctuations. Geochemically, this period sees a reversal of MIS2 trends, with increasing Ba, Al, and Fe concentrations indicating renewed weathering and soil formation processes [137,138,139,140,141,142,143,144,145,146] (Figure 7). Carbonates reach another high period, followed by a decreasing trend. The U-ratio decreases while magnetic susceptibility increases, showing a shift toward finer particle deposition and potential pedogenesis (Figure 4 and Figure 6). Organic content begins to increase, albeit slowly, suggesting gradually improving conditions for biological productivity. These changes collectively indicate the onset of warmer and more humid conditions, leading to the reestablishment of more diverse and productive ecosystems as the region transitions toward Holocene conditions.

The Late Glacial period marks a significant shift in the vegetation dynamics at the Hajós site (Figure 5). As the climate began to warm, there was a gradual expansion of boreal forest–steppe vegetation, with Pinus sylvestris and Betula becoming more prevalent [3]. This transition reflects the amelioration of climatic conditions, with rising temperatures and increased moisture availability supporting the re-establishment of tree species in the region. The pollen data [3] indicate a mosaic landscape, where patches of boreal forest were interspersed with open areas, reflecting the ongoing but incomplete recovery of forest cover following the LGM.

4.5. Early Holocene (11,700–8200 cal BP)

The Early Holocene at the Hajós site is characterised by significant environmental changes, reflecting the peak of post-glacial warming: the transition from pine-dominated landscape to temperate forest ecosystems, an increase in forest diversity, the development of wetland habitats with “brown moss carpet” [97], and a vegetation response to Early Holocene climate fluctuations. The pollen assemblage [3] suggests the development of a mixed deciduous forest (Quercus, Ulmus, Tilia, and Corylus) in the region, with a mosaic of different forest types depending on local conditions. The rapid expansion of Corylus is especially noteworthy from ca. 10,500 cal BP, showing a distinct peak that may be associated with early Mesolithic human activities or climate-driven vegetation changes, evidenced by local (Great Hungarian Plain) [147,148,149,150,151,152] and regional (Carpathian Basin) [153,154,155,156] studies. Sedimentological and geochemical data support this transition (Figure 5, Figure 8 and Figure 9). The early part of this period shows a shift from riverine to lacustrine conditions, with fine silt and clay dominating the sediment (Figure 5). This suggests that the oxbow lake was cut off from the main river channel, leading to calmer waters. The later part indicates the start of the biogenic infilling of the oxbow lake, with increased organic matter content and peak Fe and Mn values. The macrobotanical record [3] shows increased plant diversity, with species like Carex elata, Menyanthes trifoliata, and various mosses indicating rich fen ecosystems. The emergence of aquatic plants such as Chara oogonia, Phragmites australis, and Glyceria cf. maxima suggests the formation of shallow water habitats, reflecting more favourable growing conditions.

4.6. Mid-Holocene (8200–4200 cal BP)

The Holocene malacofauna of Katymár North indicate a return to steppe-like conditions, but with lower diversity compared to the bottom of the sequence (MLZ-1) (Figure 10 and Figure 11). This raises questions about the relative roles of climate and human activity in shaping the modern landscape. The reduced diversity could be a result of anthropogenic habitat modification or a sign of ongoing environmental change. Phytolith diversity increases with a peak in magnetic susceptibility (Figure 12).

During the mid-Holocene, the Hajós site witnessed the establishment of mature, diverse forest communities, as well as the first clear evidence of human impact on the landscape [3,38,39]. The pollen record [3] shows the appearance of cereal pollen, indicating the onset of agriculture in the region. Concurrently, there is an increase in pollen from plants associated with human activities, such as Plantago lanceolata and Rumex. The forest composition continues to change, with an increase in Fagus and Carpinus pollen toward the upper part of this interval. This shift might reflect both climatic changes (toward slightly cooler and more humid conditions) and potentially the human management of forest resources. NAP percentages increase gradually through this interval, suggesting a progressive opening of the landscape, likely due to agricultural activities and forest clearance. The catchment basin transforms into a reed-dominated peatland (Thelypteris palustris) [3], rich in pollen and macrobotanical remains. The sediment type changed to coarse detrital gyttja and algal gyttja, indicating the continuous dominance of organic matter with at least 70% organic matter content (Figure 5, Figure 8 and Figure 9). This period is characterised by low geochemical element values along with low clay and very fine silt grain size values but increased magnetic susceptibility values (Figure 5 and Figure 9). The pollen count shows a steady decrease toward the end of the zone, possibly reflecting changing local conditions or human impact (Figure 5).

4.7. Late Holocene (4200 cal BP to Present)

The modern July temperature (Figure 3) in Katymár (around 23–24 °C) is higher than most of the reconstructed July palaeotemperatures, which generally range between 14 and 20 °C (Figure 4). This suggests overall warmer summer conditions in the present compared to much of the past record. The modern climate diagram (Figure 3) shows July as potentially arid, with high temperatures and lower precipitation. This could be compared to past periods with lower July temperatures, which might have experienced less evaporation and potentially different moisture regimes. The generally lower past July temperatures compared to the present suggest a long-term warming trend in the region, which aligns with broader patterns of climate change.

The Late Holocene at the Hajós site is marked by significant fluctuations in vegetation cover, driven by a combination of climatic variability and increasing human activity. Pollen data indicate periods of forest contraction, likely associated with cooler and drier climatic episodes, as well as the expansion of open landscapes, which may have been influenced by anthropogenic activities such as agriculture and deforestation. The gradual increase in human impact indicators, particularly in the upper part of the sequence, reflects the intensification of agricultural practises and the growing influence of human populations on the landscape [3,38,39]. Local variations in species composition and the timing of vegetation shifts at Hajós suggest that site-specific factors, such as microclimate and soil conditions, played a significant role in modulating the response of vegetation to these broader environmental drivers. The overall pollen assemblage points to a landscape that was increasingly open and managed, with a mosaic of cultivated fields, pastures, settlements, and remnant woodland. This transformation reflects the intensification of human impact on the environment, likely associated with population growth, technological advancements, and changing socio-economic structures. A peak in plant diversity around 4300–1600 cal BP occurs due to favourable climatic conditions first, with the abundance of species such as Carex, Schoenoplectus, Cyperus, Drepanocladus, and Calliergonella, along with a rise in Equisetum cf. fluitans, which is followed by a small decrease in the abundance of certain species, yet overall diversity remains high [3]. From ca. 1540 cal BP, partial reforestation, shift in forest composition, potential changes in land use intensity, and possible climatic fluctuations affecting vegetation patterns occur. In the last thousand years, agriculture has intensified and diversified, the area has maintained an open landscape with managed woodlands, and recent alterations due to modern land management practises can be observed. Fluctuations in organic matter content, carbonate content, and various elemental compositions reflect changing local conditions and possibly human influences on the landscape. The upper parts of the core suggest the presence of a relatively stable wetland ecosystem with minor fluctuations, which could be attributed to local hydrological changes or minor climate variations. The water level rises with the abundance of Carex elata, a critical peat-forming species within the mire; also, Phragmites australis is present, contributing to the peat composition [3]. The deterioration of the diverse tussock–hollow vegetation complex also reflects the changes in the ecosystem’s structure and function. The most recent part of the record (SMZ-6) (Figure 5) shows a rapid transition in sediment composition, with decreasing organic content and increasing inorganic and carbonate content, possibly reflecting modern land use changes and drainage activities.

5. Conclusions

This study, integrating multi-proxy analyses from the Katymár North loess–palaeosol sequence and the Hajós hay meadow and loess bank cores, provides a comprehensive reconstruction of the palaeoenvironmental evolution in the Carpathian Basin from the Late Pleistocene to the Early Holocene. Our findings reveal the region’s high responsiveness to global climate forcings and highlight the development of a unique Ice Age ecoregion in the central and southern parts of the basin.

The data from both sites demonstrate synchronous responses to large-scale climate oscillations, such as the Bølling–Allerød Interstadial, while also capturing site-specific environmental changes. The persistence of thermophilic steppe-dwelling faunal elements throughout significant glaciations suggests the continuous presence of temperate grassland habitats in the region, even during cooling periods. This characteristic distinguishes the southern part of the Carpathian Basin from other areas within the Eurasian loess belt, indicating the formation of an independent mosaic or ecoregion influenced by the basin’s unique climatic conditions and the Danube alluvial fan.

Our results indicate that by the end of the Ice Age, humidity and surface moisture, rather than temperature, became the most significant palaeoecological limiting factors for biota in the southern and central parts of the Carpathian Basin. This shift led to the development of ecotone-like forest–steppe environments during cooling periods, characterised by increased species richness and individual numbers. Conversely, warming periods saw a reduction in species diversity but an expansion of temperate forest–steppe elements.

The transition from the Last Glacial Maximum to the Holocene is marked by significant changes in vegetation dynamics, with a shift from boreal forest–steppe mosaics to more diverse, temperate forest ecosystems. The Early Holocene witnessed rapid environmental changes, including increased forest diversity and the development of wetland habitats, reflecting the peak of post-glacial warming.

Human impact becomes increasingly evident in the Mid to Late Holocene, with clear indications of agricultural activities and landscape modification. This anthropogenic influence, combined with ongoing climatic changes, led to a progressive opening of the landscape and alterations in forest composition.

In conclusion, this study underscores the complex interplay between global climate forcings, regional environmental conditions, and human activities in shaping the palaeoenvironmental history of the Southern part of the Carpathian Basin, more specifically the southern part of the Danube–Tisza Interfluve. The identification of a distinct Ice Age ecoregion in the central and southern parts of the basin contributes to our understanding of the spatial heterogeneity within the Eurasian loess belt. These findings have important implications for palaeoclimatic reconstructions, ecological modelling, and our understanding of human–environment interactions in Central Europe during the Late Quaternary.