Lithostratigraphical studies in the outcrop at Ujscie , Toruft-Eberswalde Pradolina , western Poland

[ L i t h o s t r a t i g r a p h i s c h e U n t e r s u c h u n g e n i n d e m A u f s c h l u ß b e i U s c h , T h o r n E b e r s w a l d e r U r s t r o m t a l , W e s t P o l e n ] K u r z f a s s u n g : Der Aufschluß in Usch liegt am Südrand des Thorn-F 'berswalder Urstromtales zwischen de r Frankfurter Staffel und d e m P o m m e r s c h e n Stadium. D i e rund 50 m mächt ige Sedimentfo lge wird aus vier G e s c h i e b e m e r g e l ­ b ä n k e n mit zwischengescha l te ten glazifluvialen Sanclen gebi ldet . N e b e n G e s c h i e b e e i n r e g e l u n g s m e s s u n g e n , granu lomet r i schen Analysen und der Bes t immung des Karbo­ natgehal tes liegt der Schwerpunkt der Un te r suchungen in den pe t rographischen Analysen der Kiesante i le der G e ­ sch iebemerge l in der Fraktion 4 12,5 mm und in der Leit*) Anschriften der Verfasser: Prof. Dr. M A R G O T B Ö S E , Freie Universität Berlin, Institut für Geograph i sche Wissenschaf ­ ten, Phys i sche Geograph ie , Grunewaldstr . 3 5 , D 1 2 1 6 5 Berlin. Mgr MARIA G O R S K A , A. Mickiewäcz LIniversity, Q u a ­ ternary Research Institute, Department o f G e o m o r p h o l o g y . S c i e n c e Centre PAS, Wien iawsk iego 1 7 / 1 9 , 61 -712 Poznan, Poland . gesch iebebes t immung . B e i d e Methoden g e b e n Erkennt­ nisse übe r Eisfl ießrichtungen im Inlandeis. D e r obers te G e s c h i e b e m e r g e l (Kolmarer Staffel) enthält viel ostbal t isches Material mit Aland-Graniten, vielen pa läozo i schen Kalksteinen und relativ vie len devonischen Dolomi ten . Der zwei te G e c h i e b e m e r g e l (Brandenburge r Frankfurter Stadium) enthäl t im Vergleich dazu weniger os tbal t i sches Material und wen ige r pa l äozo i sche Kalke, da­ g e g e n nimmt der Feuerstein-Anteil e twas zu, was ebenfalls a u f e in weiter west l ich ge l egenes 1 lerkunftsgebiet hin­ weist . D e r dritte G e s c h i e b e m e r g e l hat saalezei t l iches Alter; die zahlreichen Dolomi te s o w i e die sons t ige Gesch iebezu ­ sammense tzung weisen wieder auf ein os tbal t i sches Her­ kunftsgebiet hin, das typisch für warthezei t l iche Glazialab­ lagerungen ist. Der unterste G e s c h i e b e m e r g e l mit Gesch ie ­ b e n aus Dalarna und Smäland läßt auf e inen Eistransport aus Norden schl ießen. D i e s e Moräne ist wahrscheinl ich ebenfa l l s .saalezeitlich, e in elsterzeitl iches Alter kann je­ d o c h nicht völlig ausgesch los sen werden.


Introduction
The investigated outcrop is situated in western Po land, at the southern margin of the Toruh-Eberswal de Pradolina (Fig. 1).The Pradolina consists of three (KRYGOWSKI 1961) or four (KOZARSKI 1962) parts among which the Ujscie Basin and the Czarnkow part are the central area of this huge ice marginal spillway in the young morainic area (Fig. 2).The morphology of the study area has already attracted the attention of KEILHACK (1897;1898); MAAS (1904); KORN (1917) and WOLDSTEDT (1932) to the very southern edge of the val ley so that all terraces have been des troyed and the morainic plateau forms a cliff to the flood plain.The outcrop has existed at least since the beginning of this century and be longs to a glass factory (Fig. 3).The outcrop was investigated in 1990, 1991, 1992 and 1994 fourth till was revealed in some parts as the pit was excavated at lower levels than before (Fig. 4).
The uppermost till (till A) is only partly preserved; it is calcareous and medium brown in colour.Till A is separated from the underlying till B by a thin sandy layer only a few cen timetres thick, characterized by a mylonitization zone (KOZARSKI & KA SPRZAK 1992) (Fig. 5).This till is of a slightly darker brown colour and shows a sandy facies also including more gravels.
In the eastern part of the investigated site this till is directly underlain by a light greyish till (till C) with a friable stmcture due to a great number of small joinLs.-Grain size distribution was established by sieve analysis and CASAGRANDE'S gravitational silt (< 0.1 mm) analysis in PRÖSZYNSKI'S modification (LITYNSKI 1972).
-The calcium carbonate content was determined ac cording to SCHEIBLER.
-Petrographic gravel analysis of the 4 -12.5 mm par ticle size was undertaken to facilitate stratigraphical interpretation.Ten components were distinguished megascopically or by using magnifying lenses: Kcrystalline rocks, S-sandstone and quartzite, TU-Palaeozoic shale and siltstone, F-flint, KK-Cretaceous chalk, PK-Palaeozoic limestone, D-Devonian dolo mite, L-lydite, Q-quartz, WQ-white quartz; the sum for a palaeogeographical interpretation of the study area than for stratigraphical purposes.

Till A
The uppermost till was investigated using two sam ples with very different granulometric compositions.Sample 1 contains an unusually high quantity of clay, whereas sample 2 represents the "normal" san dy till (Fig. 7).The fine-grained material probably corresponds to a sediment called "lehmiger bis toniger Sand mit undurchlässigem Tonmergel -und tie ferem Geschiebemergel-Untergrund" mapped as top layer on parts of the till plain by KORN (1907) on the adjacent geological map sheet "Scharnikau".The calcium carbonate content varies from 26% in the clay-rich part, which is definitely the highest value of all samples, to about 10% in the "normal" till (Fig. 8).Till fabric measurements by KASPRZAK & KOZARSKI (1985) show an ice flow from ENE for the upper till which is mainly due to the fact that the outcrop is lo cated in the western part of the ice tongue which shaped the end moraine of the Chodziez subphase (Fig. 4).

^ CaC03
• PK 14.0-12.5iconsists almost entirely of grey components, whereas red ones are rare.Both samples contain dolo mite and relatively few sandstones and very few flints (Table 1).The composition of the coarser 12.5-20 mm fraction is similar.
The boulders in the >20 mm frac tion, used for indicator boulder countings, also contain more Palae ozoic limestone (43%) than crystal line (39%).This distribution in all investigated fractions already indi cates a general ice advance from NE. Thirty-one boulders were iden tified as indicator boulders with a high percentage of Aland granites.
The TGZ (Fig. 9a, 10) indicates an East Baltic influence which is sup ported by the high quantity of Palae ozoic limestone.

TUIB
Till B (Fig. 4) contains more sand (Fig. 7), and silt and sand show a bimodal distribution.The gravel content in the three samples taken for gravel analy sis is also higher in comparison with till A. Most of the gravels are well rounded, indicating that prob ably a high percentage of fluvial material has been incorporated which would also explain the high sand content.The CaC0 3 content is <12% (Fig. 8).Till fabric measurement for this till gave a general di rection of ice flow from NNE (KASPRZAK & KOZARSKI 1985) (Fig. 4).
In samples 3 and 5 the proportion of crystalline to Pa laeozoic limestone is higher than in till A (Table 1); two samples (3 and 4) contain a low quantity of do lomite, in one sample (5) dolomite is absent.Sample 4 is probably not representative for the whole till as it derives from a basal part and includes more quartz and flint (Q/K >2) (Table 1 and 2).As those "hard" components are normally enriched in fluvial materi al, it gives a further indication of the incorporation of fluvial and/or fluvioglacial material (BÖSE 1989).
With increasing particle size the proportion of crystalline increases; in the indicator boulder frac tion (>20 mm) crystalline amounts to 50% whereas limestone totals 26%.The flint content is also higher (6%) than in till A. This material is more influenced by material from the western Baltic basin and Swed ish rocks mainly from the Växjö area and some from southern Sweden; the TGZ value is situated much farther south and more to the west than that of the till above (Fig. 9b, Fig. 10).

Till C
Till C has a different appearance: iLs colour is greyish and its friable structure evidences hea\y stress and pres sure.The granulometric analyses show a poorly sorted sandy silty matrix (Fig. 7).The two samples whose cal cium carbonate content has been detemtined are very similar to those of the overlying till (Fig. 8).Till fabric measurements indicate the local ice flow direction which shifts from ENE in the lower part of the till layer towards NNE in the upper part (Fig. 4).

57
The almost easterly direction in the basal part may be due to a very local flow direction as the till dips down from east to west in the outcrop into a very shallow palaeo-depression.Between the till and the underlying sands is a transitional zone.Hence these sands are of proglacial origin, from the same ice ad vance as till C.
The four samples (6, 7, 8 and 9) reveal a much more homogeneous petrographical composition than those of the tills described above.Till C contains on average more crystalline, and the sandstone content is more consistent.Sample 7, taken in the lowest part of the till layer, again shows an increasing amount of quartz (Q/K >2) and flint (Table 1 and 2).
In the indicator boulder fraction (>20 mm) numer ous boulders show weathering features at their sur faces.The crystalline content increases with the size of the particles.Among the 50 indicator boulders Aland rocks are quite frequent, but Hardeberga sandstone from southern Sweden and boulders from Smäland are also included (Fig. 9c).The east Baltic influence is represented by the Palaeozoic limestone as well as dolomite (4%).The TGZ-value is situated nevertheless somewhat farther to the south and west than that of the uppermost till A (Fig. 10).

Till D
The lowest till layer (Fig. 4), unknown until now, is sandysilty and poorly sort ed, too.The bottom part is partly brown ish and more sandy (Fig. 11).The five samples -three from the black and two from the brownish part -show a some what greater variety in their calcium car bonate content, es pecially sample 14 has an extremely low CaCOj content in the fine material (Fig. 8).
Till fabric measure ments in the lower part indicate north to eastnortheasterly directions; in the upper part of the till layer the western part of the outcrop indicates an ice flow from the east, whereas about 150 m to 200 m to the east northerly directions are indicated (Fig. 4).This may also be due to a very local effect of ice flow into a shallow depression or directed by a slope in the eastern part of the outcrop, as has already been described for till C. Seven samples for gravel analysis have been taken, two of them (15 and 16) from the brownish bottom part (Table 1).The crystalline content is always high er than the Palaeozoic limestone content.The flint occurs in varying amounts, but is generally higher than in the other three tills, whilst sandstone and crystalline are less frequent.Samples 11 and 13 have a higher quartz content, which is also clearly shown in the quartz/crystalline ratio (>2).Besides the Nor dic components, some probably Tertiary lignite and pyrite as well as secondary iron concretions have been found.
The indicator boulders were from central and.more often, southern Sweden (Fig. 9d); consequently the TGZ value is clearly different (Fig. 10).
Between this till and the Tertiary surface is a se quence consisting of about 40 m of Quaternary sedi ments of unknown composition.The surface of the Tertiary sediments forms a northwest-southeasterly oriented trench where the pre-Quaternary surface is situated between -20 and -50 m below sea level.A small tributary branch in an eastnortheast-westnorthvvest direction is located just below the outcrop at -10 m, filled with fine sand containing charcoal (cf.Mapa Geologiczna Polski 1:200 000, Pila, 1977).

Discussion
The interpretation of the sediments in this study is mainly related to stratigraphical questions.Lithostratigraphical investigations of tills have been conducted by several Polish authors (cf. NUNBERG 1971;RZECHOWSKI 1980: CHOMA-MORYL et al. 1991;KRZYSZKOWSKI 1988KRZYSZKOWSKI , 1990KRZYSZKOWSKI , 1994)).Problems arise in comparing the results of different studies as the methods differ to a certain extent.NUNBERG (1971) bases her interpretation on two coefficients: -O/K" and -A/B O are sedimentary rocks: Palaeozoic limestone, do lomite, sandstone, quartzite, and brown coal, i. e. both Nordic material and more local, Tertiary mate rial are included.AT are crystalline and quartz.
A are "non-resistant rocks" such as Palaeozoic lime stone, dolomite, and brown coal; B are "resistant rocks'' stich as crystalline, quartzite, quartz and flint.
Cretaceous limestone (KK) as well as Palaeozoic shale and siltstone (TU) have not been distinguished.RZECHOWSKI (1971; 1980) uses three coefficients, two of which are similar to those of NUNBERG.His sedi mentary rock components (O) are Palaeozoic sand stone, quartzite and Palaeozoic limestone; as in NUN BERG (1971), if represents crystalline and quartz.
His "non-resistant" components consist of Palaeo zoic limestone, Palaeozoic shale and siltstone, and dolomite, so brown coal is absent but Palaeozoic shale and siltstone is included; his "resistant" rocks are crystalline, sandstone, quartzite and quartz.Flint is not taken into consideration in the coefficient.K/Wis his third coefficient consisting of crystalline to calcareous material.The tills are related to Saalian deposits, but his study area is situated in Middle Po land, and therefore too far away for comparison with the results from Ujscie.
CHOMA-MORYL et al. (1991) give much more general indications about the types of rock material of their coefficients.O/K represents sedimentary rock to crystalline, K/Wcrystalline to calcareous material in general.The components of the "non-resistant to re sistant" coefficent A/B are not specified, so it is diffi cult to compare any other results with them.KRZYSZKOWSKI (1988. 1990. 1994) also uses these three coefficients.For the sedimentary rocks he dis tinguishes Palaeozoic limestone, dolomite, sand stone, quartzite, Palaeozoic shale and siltstone; for K, he distinguishes crystalline and blue quartz in his paper of 1990, but only crystalline in his other paper published in 1994.As this paper mentions only milk quartz and not the other, more frequent quartzes (blue and shiny) they are presumably considered to be part of the crystalline.Unlike KRZYSZKOWSKI and RZECHOWSKI, we have not included quartz in K, because the amount of quartz in a sample is not only related to the crystalline con tent (cf.changing Q/K-coefficient in Table 2) but de pends also on the amount of locally reworked fluvi al or glaciofluvial material as described in the sam ples 4, 11 and 13 (Table 1), and can therefore easily be enriched.Hence only a limited comparison is possible be tween the diagrams (Fig. 12 a-d) and the results of  Kap. 5).
;1 coefficients Koeffizienten of four tills in Ujscie (explanation cf.chap.5).der Kiesfraktion aus den vier Geschiebemergeln (Erklärung vgl. C as belonging to the Wartha phase of the Saa lian Glaciation.The lowermost till of the outcrop, till D, is charac terised by more flint in most of the samples (Ta ble 1), so five of seven samples show an F/K co efficient of >0.04 (  (1991) is much lower than in the other tills owing to weathering which is not the case in the till in Ujscie (Fig. 8), where only sample 14 is significantly reduced in calcium carbonate but not in PK.The lowermost till is the only one in which the content of cal cium carbonate does not parallel that of the Palae ozoic limestone.(1991) describe their val ues as typical in western Poland, but a similar till has not been found at Ujscie.Consequently the composition of till D is not typical for the Elsterian till of the area.
The petrological compo-sition of the Drenthe(-Odranian) till has not been studied in detail in western Poland, so no compari son is possible.Like the gravel analyses, the indicator boulders sam pled in the outcrop show that the ice stream flowed almost N-S (cf.BÖSE 1990: Fig. 5).This result sup ports the interpretation that this till is related to an early phase of a glaciation in the sequence of a gla ciation cycle.It may be interpreted as an Odra till which has not undergone excessive weathering.But the possibility that it is an Elsterian till which has not been strongly weathered cannot be excluded.Drillings in the Quaternary sediments below the out crop could give more information, especially if fur ther till layers were found and investigated.

Fig. 5 ; 200 SWEDENFig. 6 :
Fig. 5; Contact between the Chodziez (till A) and the Leszno-Poznan (till B) tills with horizontal wedge, develop ed in the stoss of a stone (Photo: BOSK, 21/6/90).Abb.5: Kontakt zwischen Kolmarer Geschiebemergel (A) und Brandenburg-Frankfurter Geschiebemergel (B) mit horizonta ler Sandeinschaltung an der Luvseite eines Steines (Foto: BÖSE, 21.6.90).ofall particles of one sample equals 100%, other components such as concretions or Tertiary material were counted separately.The source areas of Palae ozoic, Devonian and Cretaceous sedimentary rocks are marked in Fig.6.To facilitate comparison, com positional properties were then characterized by co efficients.Lydite and white quartz are considered to be southern fluvial components incorporated in the tills.-Indicatorboulder countings and evaluation by the TGZ (Theoretisches Geschiebezentrum) method were carried out according to LÜTTIG (1958).-Asthey represent local ice-flow directions, till fab ric measurements are considered to be more useful

Fig. 9 :
Fig. 9: Distribution of the source areas of the indicator boulders based on their percentage content; A = till A, B = till B, C = till C, D = till D. Abb.9: Die Herkunftsgebiete der Leitgesehiebe gemäß der prozentualen Verteilung; A = Geschiebemergel A, B = Geschiebemergel B, C = Geschiebemergel C. D = Geschiebemergel D.
KRZYSZKOWSKI's A/B coefficient consists of the "nonresistant" rocks Palaeozoic limestone, dolomite, Pa laeozoic shale and siltstone.The "resistant" ones are crystalline, sandstone and quartzite.The coefficient K/W corresponds to K/PK+D in Ta ble 2, Cretaceous limestone (KK) is not taken into consideration.Since various authors use the same coefficient names to denote different compositions, even dia grams which look very similar at first glance are not wholly compatible(CHOMA-MORYL et al. 1991; KRZY  SZKOWSKI 1994).Nevertheless, in order to compare a general trend in the diagrams the values of the coefficients O/K, K/W, and the A/B have been defined for this study as fol lows: O = S (sandstone and quartzite) + TU + F + KK + PK + D K= K W= PK + D A=S (sandstone and quartzite) + TU + KK + PK + D ß=K + F + L + Q + WQ Fig. 12: GravAhl).12:
(BÖSE 1989;ike the upper till than the samples of the same layer.Sample 3 fits in with the description by KRZYSZKOWSKI (1994) of the Bytyn till as the uppermost Weichselian till layer south of the study area but in the hinterland of the Poznan end moraine.Taken in the upper part of the till, it al ready indicates the more easterly influence, al though the dolomite content is slightly lower than in the overlying till and some Paleozoic shale and siltstones are still present (Table1); similar characteris tics have already been described from an area more than 200 km to the west(BÖSE 1989; 1990).
This is probably due to the fact that the proportion of carbonatic rocks in the Ujscie till has been reduced by weathering processes, clearly showing the east Baltic influence by the relation PK and D, although they have both been reduced.Dolomite is also represented in a considerable quantity among the indicator boulders and the TGZ is situated farther northeast.The results allow us to interpret this till