Past hydrological events reflected in Holocene history of Polish rivers

Past hydrological events reflected in Holocene history of Polish rivers

Catena 66 (2006) 24 – 33 www.elsevier.com/locate/catena Past hydrological events reflected in Holocene history of Polish rivers Leszek Starkel a,*, R...

732KB Sizes 0 Downloads 3 Views

Catena 66 (2006) 24 – 33 www.elsevier.com/locate/catena

Past hydrological events reflected in Holocene history of Polish rivers Leszek Starkel a,*, Roman Soja a, Danuta J. Michczyn´ska b a

Institute of Geography and Spatial Organisation Polish Ac.Sc.-Krako´w, Poland Radiocarbon Laboratory, Silesian University of Technology, Gliwice, Poland

b

Accepted 22 July 2005

Abstract A database of 335 radiocarbon dates from Holocene fluvial sediments in Poland has been compiled. Each sample is characterized by 25 parameters. All samples were grouped into one of 7 regions characterized by hydrological regime, and one of 8 depositional environments. Statistical analysis was used to enhance the palaeoenvironmental data and 17 phases may be distinguished, 10 of which have a distinct sharp peak. Abandoned paleochannels (149 dates) clearly reflect the major wetter phases of the Holocene. An opposite picture is presented by overbank aggradation (86 dates) connected mainly with expansion of agriculture. Preliminary correlations are made on peat bog records, changes in vegetation, lake water level, rates calcareous tufa and speleothems growth, landslides phases and debris flow phases. These show many convergences, explained by climatic fluctuations during the Holocene. D 2005 Published by Elsevier B.V. Keywords: Radiocarbon dating; Holocene; Fluvial sediment; Poland

1. Introduction The presence of Holocene alluvial fills with organic remains was first recognized in the river valleys of the Carpathian foreland 100 years ago (Friedberg, 1903; ‘omnicki, 1895 –1903). However, it took about 50 years to dislodge the dominant view, adopted from Germany that the coarser channel facies is inherited from the last cold stage and the upper overbank loams reflect accelerated erosion after forest clearances during last millennia (Klimaszewski, 1948). In 1960, Starkel described 2 –3 Holocene alluvial fills in the upper Vistula catchment and, on the basis of pollen spectra, connected them with two wetter phases during the Atlantic and sub-Atlantic (Starkel, 1960). The discoveries of Late glacial age abandoned paleomeanders – of much larger dimensions than several generation of Holocene palaeochannels (Falkowski, 1975; Szuman´ski, 1983; Kozarski

* Corresponding author. Tel./fax: +48 12 422 40 85. E-mail address: [email protected] (L. Starkel). 0341-8162/$ - see front matter D 2005 Published by Elsevier B.V. doi:10.1016/j.catena.2005.07.008

and Rotnicki, 1977) – and the mapping a of sequence of floods from a humid phase 8400 – 7800 14C years BP (Niedziaakowska et al., 1977), combined with dozens of radiocarbon datings, created the background to the concept of alternate wetter and drier phases, albeit with more frequent and more rare extreme events, characterized by the dominance of river aggradation or erosion (Starkel, 1983). The number of alluvial fills that have been recognized has increased and in last decade it has been possible to distinguish phases of increased fluvial activity (Starkel, 1990; Kalicki, 1991; Starkel et al., 1996) at c. 12.1– 11.2 cal ka BP (Younger Dryas), 9.5 – 8.5, 7.4– 6.9, 6.2– 5.6, 5.2 – 4.6, 3.5 – 3.2, 2, 9– 2.7, 2.2– 1.8, 1.4– 1.3, 10– 11th c., and 16 –19th c. AD. Each of these phases lasted between 1 –3 and more centuries and was characterized by higher frequency of floods, often grouped in clusters (Starkel, 1998, 2003a). The coincidence of fluvial records with higher lake water levels, speleothems, landslides, alpine debris flows, glacier advances and other phenomena indicates that those periods were characterized not only by frequent floods (continuous rains) but also by other extremes like heavy downpours, extended rainy seasons,

L. Starkel et al. / Catena 66 (2006) 24 – 33

25

Fig. 1. Types of depositional environment (A – H described in text): (1) channel facies, (2) overbank facies, (3) peat, (4) buried soil, (5) wood or organic remains, (6) horizon dated by 14C, (7) event or change older/younger/synchronous with the date.

and severe winters (Starkel, 1997; Magny, 1993). The extension of these phases over the whole of Europe and the fact that they were significant on the global scale suggested that they were linked to fluctuations in solar radiation and volcanic activity (Starkel, 1999, 2002, cf. Stuiver, 1995; Bryson and Bryson, 1998). During the two last millennia a

distinct tendency to aggradation was explained by anthropogenic accelerated soil erosion. The concept of phases with frequent extremes seems to be very prognostic. In addition, in this paper, a quantitative approach using a radiocarbon database of fluvial environment is used to test this model.

Fig. 2. Dated localities in Poland (number after databank).

26

L. Starkel et al. / Catena 66 (2006) 24 – 33

2. Identifying changes in fluvial environment The presence of organic material (wood, peat, humus horizon, etc.) in a sequence of alluvial sediments may date the depositional events with varying degrees of precision. This organic matter can be found inside an alluvial member and directly dated, if it has not been re-worked, can postdate it if the organic dated layer is above mineral flood deposits, or predate it if the organic layer is buried by flood deposits. But the fundamental question of how to define an event is often problematic and rarely can we distinguish a flood cluster or even a particular flood. Usually we may only equate an event with a time of increased fluvial activity. It is mainly a sequence of events deposited as a separate member coarser than the neighbouring ones or separated from these by erosional surface. If the concept of phases with more frequent extreme events is correct, we expect that the particular events of shorter duration, including floods, are incorporated in it. But in dating the beginning or end of a phase with more frequent floods several centuries long, we may expect that the probability curve would show two peaks marking the start and the end of phase with intensive erosion and deposition, separated by a depression.

We can distinguish the following types representing various depositional environments in Polish river valleys (Fig. 1): A. Organic remains preserved in channel deposits in the form of lenses and smaller organic fragments, which are synchronous with mineral deposits and probably were not redeposited (tree trunks are generally excluded) B. Organic remains at the base of abandoned paleochannels that postdate the avulsion or cut-off of the river channel C. Organic matter (peat, organic mud) from the top of overbank facies, buried by channel deposit (event younger than date), assuming that organic layer was not truncated D. Peat or fossil soil buried by a younger overbank flood event E. Channel facies buried by overbank deposit with organic remains (event older than the date) F. Mineral intercalation in peat sequence (dating a short flood event on a peaty floodplain) G. Base of peat overlying overbank facies (floods older than date)

Fig. 3. Number of dates in different regions.

L. Starkel et al. / Catena 66 (2006) 24 – 33

27

Fig. 4. Number of dates after size of river catchment. Fig. 6. Distribution of dated samples in 500-year-long phases.

H. Organic intercalation in overbank deposits (date should be synchronous with the phase of deposition).

3. Structure and nature of past hydrological events database The present database on past hydrological events for territory of Poland was constructed following the slightly modified methodology developed by Macklin and Lewin (2003). The database comprises about 700 fluvial radiocarbon dates, primarily from the Gliwice Radiocarbon Laboratory although only 335 (Fig. 2), which fall into sample categories A to H (described above) have been used in the analysis. Each sample is characterized by 25 parameters. The first columns indicate the author who collected the organic material and the year of sampling. Next follows the site location (locality, name of river and catchment area, elevation asl, geographical coordinates). Samples have been grouped into 7 regions differing in physiography and type of floods (the number of samples in each area is shown in brackets, Fig. 3). Especially regions A, C and E with headwaters in the mountains are influenced by summer floods connected with continuous heavy rains. A. Upper course of the Vistula valley (83), B. Middle and lower course of the Vistula valley (39), C. Carpathians tributaries in the upper Vistula catchment (122), D. Upland and lowland tributary valleys of the Vistula (37), E. Upper Oder valley and its Sudetan tributaries (6), F. Other upland and lowland valleys in the Oder catchment (37), G. Other catchments in the Baltic perimarine area (11).

Fig. 5. Type of dated material.

The next parameter is the size of the catchment (Fig. 4). In our databank there are only 10 samples representing catchments below 10 km2 and 43 of samples 10 –100 km2. The largest number of 150 samples represents a basin of 1000 – 10 000 km2. Only 32 samples come from larger basins, principally the Vistula and Oder. There is also information about the depth of sampled layers (below the surface) and type of dated material (Fig. 5), which includes peats (144), mud (74), wood in peats (48), tree trunk (only for 27 samples probably not rebedded), organic detritus (18), gyttya (16) and others (16). The next columns contain the laboratory number, uncalibrated date BP with standard deviation and finally age calibrated using the Oxcal 1.3 programme, presented both in years BP and BC/AD. The oldest considered date is 15 200 BP, and the youngest is 250 BP The distribution of dates between 500 and 12 000 BP (in 14C years) shows two peaks: at 1000– 2000 (42 samples) and 9000 –10 000 (41 samples). Only 20 samples represent the period 6000 –7000 BP (Fig. 6). The depositional environment (see Fig. 1) was identified for each samples. Only three groups of them are more representative from statistical point of view: B – 152 samples, D – 87 samples, and A – 48 samples — see Figs. 7 and 9. In the last three columns, additional information is included which dates are either synchronous, predate or postdate past hydrological event: column ‘‘change after’’ (172 samples), ‘‘change before’’ (107 samples) and ‘‘change at’’ (56 samples).

4. Statistical analysis Statistical analysis of large sets of datings results may is used as a source of palaeoenvironmental reconstructions (e.g. Hercman, 2000; Macklin and Lewin, 2003; Michczyn´ska et al., 2003; Singhvi et al., 2001). We attempted to

Fig. 7. Distribution of samples after depositional environment (A – H).

28

L. Starkel et al. / Catena 66 (2006) 24 – 33

reconstruct the Holocene history of Polish rivers on the base of frequency distributions of 331 14C-dated samples. Almost all analysed samples were investigated in the Gliwice Radiocarbon Laboratory. 14C activity measurements were carried out by gas proportional counting techniques (Pazdur et al., 2000). The result of radiocarbon dating is given as measured radiocarbon age T and its uncertainty DT (T T DT) and is described by Gaussian probability distribution. According to commonly accepted convention (Stuiver and Polach, 1977), DT value is calculated only on the base of statistical analysis of the measurements and properties of the apparatus, without taking into account any extra-laboratory factors (Fig. 8). For different sets of radiocarbon dates (different regions and depositional environments) cumulative probability density functions (CPDFs) were created by superposition of individual distributions of dates. We assume that fluctuations of the height of constructed cumulative curve generally reflect changes in environmental conditions. Unfortunately, there are some problems, which may impede the interpretation (Michczyn´ska et al., 2003). However, it should be stressed here that: 1. Variations in atmospheric concentration of 14C in the past require the use of probabilistic calibration in order to overcome difficulties in conversion of radiocarbon age into calendar age. We have used updated version of the Gliwice Radiocarbon Laboratory Calibration Programme GdCALIB (Pazdur and Michczyn´ska, 1989; Michczyn´ska et al., 1990). 2. The number of available dates must be sufficiently large. Otherwise, the constructed cumulative curve would have been influenced not only by environmental changes but also by sample selection. Gaps of the curve would reflect the periods from which the samples have not been collected. Monte Carlo simulations (Michczyn´ska and Pazdur, 2004) show that the minimum number of dates for analysed time interval is

ca. 120, when the mean uncertainty of radiocarbon dates equals DT = 125 years. 3. The critical selection of dates used in such statistical evaluations plays the key part. The general rule is to take samples from places where changes of sedimentation are visible (e.g. from the top and the bottom of layer) but that may lead to the presence of artificial, unrealistic bimodality of the constructed CPDF curve for some time periods (beginning and end of phase with frequent floods). 4. Younger deposits are generally found more frequently than the older ones and preservation bias can have a significant influence on the alluvial record (cf. Lewin and Macklin, 2003). This leads to an obvious tendency whereby the older the samples the lower the height of the CPDF curve. In the analysis of peat samples for Poland, the role of the calibration curve was considered particularly in relation to the formation of narrow peaks of the cumulative probability density function (Michczyn´ski and Michczyn´ska, in press). This coincidence is very spectacular and the steep slope sections of the calibration curve seem to work as an amplifier and increase the height of CPDF. But in our case of CPDF for fluvial samples, the wider higher portions of the curve seem to reflect the periods of increased fluvial activity, which have been documented by distinct sedimentological and geomorphological records at key localities. The extent to which the CPDF narrow peaks indicate the extremes in fluvial activity and how to separate the impact of the 14C calibration curve, should be the theme of future analysis.

5. Temporal distribution of dated Holocene fluvial units Seventeen flooding phases may be distinguished (Fig. 8) each of which is between 200 and 1000 years long: 13.8–

Fig. 8. Probability density curve of all 331 dates indicating hydrological events in all Polish rivers.

L. Starkel et al. / Catena 66 (2006) 24 – 33

29

Fig. 9. Probability density curves for 3 selected depositional environments.

13.5, 13.1– 12.7, 11.3 –10.3, 9.7– 9.4, 8.6 – 8.4, 7.7 –7.6, 7.0 –6.7, 6.4 –6.2, 6.0– 5.7, 5.3– 5.1, 4.8– 4.5, 3.9, 3.6 –3.4, 2.9 –2.8, 2.1 – 1.7, 1.5– 1.3 and 1.0– 0.6 cal BP. But only 10 of them include a distinct sharp peak concentrated in one century: 11.2 –11.1, 9.6 –9.5, 8.4, 7.7 –7.6, 5.8– 5.7, 4.8 – 4.7, 2.9 – 2.8, 1.9– 1.8, 1.4 – 1.3, and 1.0 –0.8 cal BP. The longest phase 11.3– 10.3 ka represents the very beginning of the Holocene. The well known late Boreal humid phase (Starkel, 1999) comes between two peaks at the very beginning and at the end. A very sharp culmination is

located at the Atlantic – Subboreal transition followed by a relatively wetter early Subboreal. Higher frequency is recorded again in the late Roman period followed by the second post-Roman sharp peak. The highest peak in the Middle Ages coincides with the warm phase of the 10– 14th c. AD. Abandoned paleochannels (147 dates, Fig. 9) accurately reflect the major climatic wetter phases at the beginning of the Holocene, at the Boreal – Atlantic transition, Atlantic – Subboreal transition and again in 5 –6th century AD. An

Fig. 10. Probability density curves for 2 selected groups of regions (with mountain and lowland hydrological regime).

30

L. Starkel et al. / Catena 66 (2006) 24 – 33

Fig. 11. Comparison of curve of hydrological events based on

opposite picture is presented by the overbank aggradation above peat horizons (86 dates), with highest peaks during the expansion of agriculture in the late Roman period and in the 11 – 15th c. AD (Starkel, 2003b). A different distribution is represented by the extension of peat bogs (only 19 dates) culminating at the Younger Dryas –Preboreal transition and during wetter phases (early Atlantic and early Subboreal). Comparing the upper catchment of the Vistula (region A + C) controlled by floods originating from the Carpathians with the extensive Polish lowland (regions B, D, F, G) some differences are evident (Fig. 10). All the phases visible across the whole of Poland are much better expressed in the Upper Vistula, especially during the Allero¨d, at c. 7.6, 2.8 and 1.5 ka cal BP peaks. This is due better expression of extreme floods as well as due to greater number of dated localities. In the Polish lowland the long Preboreal phase (11.2– 10.7) of gradual decline of flood activity is especially clearly visible as well as wetter phase of early Subboreal (5.8 –5.6 + 5.3– 5.0) and human impact during Mediaeval period (1.0 – 0.8). The comparison with previously distinguished phases (Starkel, 1983; Starkel, 1990; Starkel et al., 1996; Kalicki, 1991) shows many similarities. But our present database indicates a much more extensive phase of early Holocene transformation well known from the adaptation of river channels to the new regime (Starkel, 2003a). However, the Boreal – Atlantic transition so well documented in the exposures is split in our database into two, the initial and the closing phase (Figs. 8 and 9). More marked are also the mid-Atlantic and the 5 –6th c. AD flooding episodes. The latter phase, as well as most of older ones, coincide very

14

C data bank with other records referred in the text.

clearly with the aggradation of subfossil oak trunks (Kra˛piec, 1998; Fig. 11). In general, our attempt to distinguish hydrological active phases using the radiocarbon database has been very successful.

6. Preliminary correlation with other palaeoclimatic and palaeohydrological records from Poland The distinguished phases of increased fluvial activity elaborated on the base of 14C data generally repeat the earlier concept based on dozens of key localities concentrated in the upper Vistula Basin. To understand better the climatic and anthropogenic causes of changes in fluvial activity and environment discussed here, a preliminary correlation with other paleoclimatic and paleohydrological records has been done. These include the following groups of records (Fig. 11): peat bogs reflecting fluctuations in ground water level, phases of vegetation change, changes in lake water level, rates of calcareous tufa growth, rates of speleothem growth, landslide phases in the Carpathians, debris flow phases in the Tatra Mts. For peat bogs, two approaches have been taken into consideration. The older one (Ralska-Jasiewiczowa and Starkel, 1988; Ralska-Jasiewiczowa, 1989) is based on localities included in the IGCP-158 programme. The

L. Starkel et al. / Catena 66 (2006) 24 – 33

following wetter phases are distinct: transition Preboreal– Boreal (10.6 – 10.0 ka BP), transition Boreal –Atlantic (9.2 – 8.6 ka BP), transition Atlantic – Subboreal (5.8 – 5.0) and two younger ones (3.5 – 2.9 and 2.2 –1.7 ka BP). Later some modifications have been made based mainly on new records from NE Poland (Z˙urek et al., 2002). These phases generally coincide with phases of flooding. In the Holocene forest development, distinct phases of rapid transformation are visible (Ralska-Jasiewiczowa, 1989). Earlier ones coincide with warmings and invasions of new species (11.9– 11.3 ka and 10.3– 9.8), while the younger ones correspond very well with coolings and flood phases (9.2 – 8.4, 8.0– 7.7, 5.9 –5.6, 5.1 – 4.7, 2.4– 2.1) with just one exception (4.0 – 3.8 ka BP). The lake water level shows various oscillations depending on the local conditions of ice melting, lake drainage and overgrowth (cf. Ralska-Jasiewiczowa, 1989; Starkel, 2003a). Taking Gos´cia˛z˙ Lake (Ralska-Jasiewiczowa et al., 1998) and Biskupin (Niewiarowski, 1995) as the examples, we find at least four high water level phases: 11.2– 10.5 ka (melting phase), 9.4 –8.5, 5.0 –4.1 and 2.5 –2.0 ka BP (all younger ones coincide well with flood phases). Calcareous tufa growth shows varying patterns of change in various parts of karstic areas. On the Little Poland

31

Upland, most of the fast growth phases are synchronous with wetter periods (9.5 – 9.0, 7.6– 6.8, 6.3 –5.8, 2.2– 1.9) although one is extremely long (5.2 –3.2), which is probably connected with a wetter phase and Subboreal warming in between (Pazdur et al., 1988). Also, faster growth of spring tufa in E-Poland (Dobrowolski, 1998) probably reflects higher temperature rather than humidity. A less complicated pattern is reflected in speleothems found in the caves of the Little Poland Upland (Pazdur et al., 1999). These records clearly characterise a succession of wetter phases: 8.3– 7.6, 7.4 –6.8, 5.9– 4.8, 3.4 –3.2, 2.8– 2.2, 1.0 –0.8 ka BP. A very distinct correlation between flood records and landslides has been established in the last years (Alexandrowicz, 1997; Starkel, 1997; Margielewski, 2003). A number of dated large rocky landslides makes it possible to distinguish about 10 phases, characterized by more frequent rainy seasons: 9.5– 8.5, 8.0 –7.7, 7.3 –6.8, 5.9– 5.5, 5.1– 4.2, 3.5 –3.2, 2.4– 1.9, 1.3 – 1.1, 0.75 –0.6, 0.4 – 0.1 cal. ka BP. It must be added that the earlier phase 12.5 –11.3 ka BP is connected with the melting of permafrost and the formation of new groundwater drainage. A similar pattern is reflected in the frequency of debris flows in the alpine belt of the Tatra Mts., registered in lake

Fig. 12. Comparison of probability density curve with fluvial phases distinguished earlier in Vistula basin by L. Starkel (1990) with phases of advances of Alpine glaciers (by various authors) and 14C productivity curve (by Stuiver, 1995).

32

L. Starkel et al. / Catena 66 (2006) 24 – 33

sediments (the youngest debris flows are dated by lichenometric methods). The following phases have been recorded: > 11.5, 9.4 –8.9, 7.6 –6.7, 5.6 – 5.1, 4.8 – 4.5, 2.7 – 2.2, 0.75– 0.6 and 0.4 – 0.1 cal. ka BP (Kotarba, 1995; Kotarba and Baumgart-Kotarba, 1997). The debris flows are connected mainly with heavy downpours. All these data correspond very well with hydrological fluvial events registered in the database, documenting the existence during the Holocene in Central Europe of relatively wetter and cooler phases, characterized first of all by high frequency of various kinds of extreme rainfall: heavy downpours, continuous rains and rainy season (Starkel, 2002, 2003a) as well as by advances of Alpine glaciers (Fig. 12). These phases coincide also with the periods of reduced solar activity, generally reflected in the peaks of 14C productivity curve (Stuiver, 1995) and superimposed phases of high volcanic activity (Bryson and Bryson, 1998).

Acknowledgements The authors express cordial thanks to Professor Anna Pazdur, Head of the Radiocarbon Laboratory at the Silesian Technical University in Gliwice for her permission to use their databank. This study was supported by the ICSU grant ‘‘Past hydrological events related to understanding of Global Change’’ led by Professor K. J. Gregory, president of the INQUA Commission on Global Continental Paleohydrology and the authors thank him as well as Professor Mark Macklin and Dr. Eric Johnstone for discussion in constructing and analyzing the radiocarbon database and editorial help.

References Alexandrowicz, S.W., 1997. Holocene dated landslides in the Polish Carpathians. Pala¨oklimaforschung – Palaeoclimate Research 19, 75 – 83. Bryson, R.U., Bryson, R.A., 1998. Application of a global volcanicity timeseries on high-resolution paleoclimatic modelling of the eastern Mediterranean. In: Issar, A.S., Brown, N. (Eds.), Water Environment and Society in Times of Climatic Change. Kluver Ac. Publ., pp. 1 – 19. Dobrowolski, R., 1998. Strukturalne uwarunkowania rozwoju wspo´aczesnej rzez´by krasowej na mie˛dzyrzeczu s´rodkowego Wieprza i Bugu. Wydawnictwo UMCS, Lublin. Falkowski, E., 1975. Variability of channel processes of lowland rivers in Poland and changes of the valley floors during the Holocene. Biuletyn Geologiczny UW 19, 45 – 78 (Warszawa). Friedberg, W., 1903. Atlas geologiczny Galicji, 16, arkusze: De˛bica, Rzeszo´w, Rudnik. PAU, Krako´w. Hercman, H., 2000. Reconstruction of paleoclimatic changes in Central Europe between 10 and 200 thousand years BP, based on analysis of growth frequency of speleothems. Studia Quaternaria 17, 35 – 70. Kalicki, T., 1991. The evolution of the Vistula river valley between Cracow and Niepoaomice in Late Vistulian and Holocene times. Geographical Studies, vol. 6, pp. 11 – 37. Warszawa. Klimaszewski, M., 1948. Polskie Karpaty Zachodnie w okresie dyluwialnym. Proc. Wroc. Tow. Nauk. ser.B, vol. 7.

Kotarba, A., 1995. Rapid mass wasting over the last 500 years in the High Tatra Mountains. Questiones Geographicae 4, 177 – 183 (spec. issue). Kotarba, A., Baumgart-Kotarba, M., 1997. Holocene debris-flow activity in the light of lacustrine sediment studies in the High Tatra Mountains, Poland. Pala¨oklimaforschung – Palaeoclimate Research 19, 147 – 158. Kozarski, S., Rotnicki, K., 1977. Valley floors and changes of river channel pattern in the North Polish Plain during the Late-Wurm and Holocene. Questiones Geographicae 4, 51 – 93. Kra˛piec, M., 1998. Oak dendrochronology of the Neoholocene in Poland. Folia Quaternaria 69, 5 – 134. Lewin, J., Macklin, M.G., 2003. Preservation potential for late Quaternary river alluvium. Journal of Quaternary Science 18, 107 – 120. ‘omnicki, A.M., 1895 – 1903. Atlas geologiczny Galicji 10, 12, 15. PAU, Krako´w. Macklin, M.G., Lewin, J., 2003. River sediments, great flood and centennial-scale Holocene climate change. Journal of Quaternary Science 18, 101 – 105. Magny, M., 1993. Holocene fluctuations of lake levels in the French Jura and sub-Alpine ranges and their implications for past general circulation pattern. The Holocene 3 (4), 306 – 313. Margielewski, W., 2003. Late glacial – Holocene paleoenvironmental changes in the Western Carpathians: case studies of landslides forms and deposits. Folia Quaternaria 74, 1 – 96. Michczyn´ski , A., Michczyn´ska, D.J., in press. The effect of CDPF peaks height increase during calibration of radiocarbon dates sets. Geochronometria 25. Michczynska, D.J., Pazdur, A., 2004. Shape analysis of cumulative probability density function of radiocarbon dates set in the study of climate change in late glacial and Holocene. Radiocarbon 46 (2), 733 – 744. Michczyn´ska, D.J., Pazdur, M.F., Walanus, A., 1990. Bayesian approach to probabilistic calibration of radiocarbon ages. In: Mook, W.G., Waterbolk, H.T. (Eds.), Proceedings of the 2nd International Symposium 14C and Archaeology, vol. 29. PACT, Strasbourg, pp. 69 – 79. Michczyn´ska, D.J., Michczyn´ski, A., Pazdur, A., Z˙urek, S., 2003. 14C dates of peat for reconstruction of environmental changes in the past. Geochronometria 22, 47 – 54. Niedziaakowska, E., Skubisz, A., Starkel, L., 1977. Lithology of the Eoand Mesoholocene alluvia in Podgrodzie upon Wisaoka river. Studia Geomorphologica Carpatho-Balcanica 11, 89 – 100. Niewiarowski, W. (Ed.), 1995. Zarys zmian s´rodowiska geograficznego okolic Biskupina pod wpaywem czynniko´w naturalnych i antropogenicznych w po´z´nym glacjale i holocenie. Turpress, Torun´. Pazdur, M.F., Michczyn´ska, D.J., 1989. Improvement of the procedure for probabilistic calibration of radiocarbon dates. In: Long, A., Kra, R.S., Srdoc´, D. (Eds.), Radiocarbon, 31, pp. 824 – 832. Pazdur, A., Pazdur, M., Starkel, L., Szulc, J., 1988. Stable isotopes of the Holocene calcareous tufa in Southern Poland as paleoclimatic indicators. Quaternary Research 30 (2), 177 – 189. Pazdur, A., Goslar, T., Gradzin´ski, M., Hercman, H., 1999. Zapis zmian hydrologicznych i klimatycznych w obszarach krasowych Polski Poaudniowej na podstawie badan´ izotopowych. In: Pazdur, A., Bluszcz, A., Stankowski, W., Starkel, L. (Eds.), Geochronologia go´rnego czwartorze˛du Polski. Inst. Fizyki Politechniki S´la˛skiej Gliwice, pp. 157 – 177. Pazdur, A., Michczyn´ski, A., Pawlyta, J., Spahiu, P., 2000. Comparison of the radiocarbon dating methods used in the Gliwice Radiocarbon Laboratory. Geochronometria 18, 9 – 14. Ralska-Jasiewiczowa, M. (Ed.), 1989. Environmental changes recorded in lakes and mires of Poland during the last 130 000 years: Part III. Acta Paleobotanica, vol. 29, 2, pp. 1 – 20. Warszawa-Krako´w. Ralska-Jasiewiczowa, M., Starkel, L., 1988. Record of the hydrological changes during the Holocene in the lake, mire and fluvial deposits of Poland. Folia Quaternaria 57, 91 – 127. Ralska-Jasiewiczowa, M., Goslar, T., Madeyska, T., Starkel, L. (Eds.), 1998. Lake Gos´cia˛z˙, Central Poland, A Monographic Study. W. Szafer Institute of Botany, Pol. Ac. Sc., Krako´w.

L. Starkel et al. / Catena 66 (2006) 24 – 33 Singhvi, A.K., Bluszcz, A., Bateman, M.D., Someshwar Rao, M., 2001. Luminescence dating of loess – palaeosol sequences and coversands: methodological aspects and palaeoclimatic implications. Earth Science Reviews 54, 193 – 211. Starkel, L., 1960. Rozwo´j rzez´by Polskich Karpat fliszowych w holocenie. Prace Geograficzne Inst. Geogr. PAN, vol. 22, pp. 1 – 239. Starkel, L., 1983. The reflection of hydrologic changes in the fluvial environment of the temperate zone during the last 15 000 years. In: Gregory, K.J. (Ed.), Background to Paleohydrology. Wiley, pp. 213 – 237. Starkel, L. (Ed.), 1990. Evolution of the Vistula river valley during the last 15 000 years, vol. 3, Geogr. Studies, spec. issue 5. Starkel, L., 1997. Mass movements during the Holocene: the Carpathian example and the European perspective. Pala¨oklimaforschung – Palaeoclimate Research 19, 385 – 400. Starkel, L., 1998. Frequency of extreme hydroclimatically-induced events as a key to understanding environmental changes in the Holocene. In: Issar, A.S., Brown, N. (Eds.), Water, Environment and Society in times of Climatic Change. Kluver A. P., pp. 273 – 288. Starkel, L., 1999. 8500 – 8000 yrs BP humid phase – global or regional? Science Reports of Tohoku University, 7th Series. Geography 49 (2), 105 – 133. Starkel, L., 2002. Change in the frequency of extreme events as the indicator of climatic change in the Holocene (in fluvial systems). Quaternary International 91, 25 – 32.

33

Starkel, L., 2003a. Short-term hydrological changes. In: Gregory, K.J., Benito, G. (Eds.), Paleohydrology, Understanding Global Change, vol. 20. Wiley, pp. 337 – 356. Starkel, L., 2003b. Clustering of flood events in the context of increased human activity. In: Thorndycraft, V.B., Benito, G., Barriendos, M., Llasat, M.C. (Eds.), Paleofloods, Historical Data and Climatic Variability; Applications in Flood Risk Assessment, Proceedings of PHEFRA Workshop in Barcelona, October 2002, pp. 79 – 84. Starkel, L., Kalicki, T., Kra˛piec, M., Soja, R., Ge˛bica, P., Czyz˙owska, E., 1996. Hydrological changes of valley floors in upper Vistula basin during late Vistulian and Holocene. In: Starkel, L. (Ed.), Evolution of the Vistula River Valley During the Last 15 000 Years, Geographical Studies, vol. 9, pp. 7 – 128. Wrocaaw. Stuiver, M., 1995. Solar and climatic components of the atmospheric 14C record. Pala¨oklimaforschung – Palaeoclimate Research 16, 51 – 59. Stuiver, M., Polach, H.A., 1977. Reporting of 14C data. Radiocarbon 19, 355 – 363. Szuman´ski, A., 1983. Paleochannels of large meanders in the river valleys of the Polish Lowland. Quaternary Studies in Poland 4, 207 – 216. Z˙urek, S., Michczyn´ska, D.J., Pazdur, A., 2002. Time record of palaeohydrologic changes in the development of mires during the late glacial and Holocene, North Podlasie Lowland and Holy Cross Mts. Geochronometria 21, 109 – 118.