Elsterian ice sheet dy nam ics in a top o graph i cally var ied area (south ern part of the Racibórz–Oœwiêcim Ba sin and its vi cin ity, south ern Po land)
Tomasz SALAMON1, *
1 Uni ver sity of Silesia in Katowice, Fac ulty of Earth Sci ence, Bêdziñska 60, 41-200 Sosnowiec, Po land
Salamon, T., 2017. Elsterian ice sheet dy nam ics in a top o graph i cally var ied area (south ern part of the Racibórz–Oœwiêcim Ba sin and its vi cin ity, south ern Po land). Geo log i cal Quar terly, 61 (2): 465–479, doi: 10.7306/gq.1350
Dur ing the Elsterian Gla ci ation, the Scan di na vian Ice Sheet, be fore reach ing its max i mum ex tent, ad vanced over the moun - tain ous fore land of South ern Po land, which was char ac ter ized by a sub strate of in creas ingly var ied re lief. One such area was the hilly Rybnik Pla teau of the south ern part of the Racibórz–Oœwiêcim Ba sin, lo cated di rectly north of the Ostrava Ba sin and Moravian Gate, where the ice sheet ad vanced far south and formed a large lobe. In this study, till from the £aziska site, lo - cated in south ern part of the Rybnik Pla teau, was ana lysed and in ter preted. The till was de pos ited at the top of an W–E ori - ented ridge com posed of Neo gene clays. The study in di cates that a high basal wa ter pres sure oc curred at the base of the ice sheet, fa vour ing a basal mech a nism of ice sheet move ment. Till fab ric and ki ne matic struc tures in di cate that ice flowed eas - ily from the west over the ridge, par al lel to its axis. Re con structed pat terns of ice flow di rec tion re flect an ice sheet that was char ac ter ized by a strong spa tially var ied dy nam ics. The Odra Val ley was the main cor ri dor of fast flow ing ice to the Ostrava Ba sin, from where ice was dis trib uted ra di ally in dif fer ent di rec tions. The sec ond cor ri dor of ac tively flow ing ice was prob a bly lo cated in the Ruda palaeovalley to the north of the Rybnik Pla teau. In the cen tral part of the Rybnik Pla teau, ice flow was re - stricted. This less dy namic ice sheet be hav iour re sulted mostly from the much higher lo ca tion of the area. This study in di - cates that the large mor pho log i cal and lithological vari a tions of the ice sheet sub strate in duced var ied fric tional re sis tance, pro duc ing a spa tially var ied stress field within the ice sheet. This gen er ated ac tive zones of stream ing ice ad ja cent to more pas sive zones. Po ten tial switch ing be tween neigh bour ing ac tive zones trans port ing ice to wards the ice sheet mar gin is pos - tu lated.
Key words: ice sheet dy nam ics, fore-moun tain area, subglacial till, Elsterian Gla ci ation, Ostrava Ba sin.
INTRODUCTION
Al though the Scan di na vian Ice Sheet reached its great est ex tent in Cen tral Eu rope dur ing the Elsterian (Sanian) Gla ci ation (Mojski, 2005), rel a tively lit tle is known about its glaciological char ac ter is tics and dy nam ics. This is due to the fact that de pos its of the Elsterian Gla ci ation are ob scured by over ly ing youn ger de - pos its, and ap pear on the sur face only lo cally, mainly in the fore - land of the Sudetes and Carpathian moun tains of south ern Po - land. Two Elsterian till units were de scribed from the Sudetes fore land (Krzyszowski and Ibek, 1996; Badura and Przybylski, 1998; Krzyszkowski and Karanter, 2001). The Saalian till is un - der stood to have a sim i lar ex tent. There fore, dif fer en ti at ing tills of dif fer ent glaciations from each other is at times prob lem atic (cf.
Czubla, 2013). This prob lem does not ex ist at the Carpathian fore land, where tills are known from a few sites of the Sandomierz Ba sin, the Racibórz–Oœwiêcim Ba sin and the Carpathian Foot hills (Klimaszewski, 1952; Jahn, 1952;
Wojtanowicz, 1985; Butrym et al., 1988; Lewandowski, 1988;
Nitychoruk, 1991; £anczont, 1997; Wójcik, 2003; Wójcik et al., 2004; Salamon and Wójcik, 2010; Salamon, 2014b, 2016a).
These tills are gen er ally cor re lated with one gla ci ation, al though dif fer ent opin ions ex ist about their age, i.e. Sanian II (Ma rine Iso - tope Stage 12) or Sanian I (MIS 16) have been pro posed (cf.
Lindner, 2001; Marks, 2005; Mojski, 2005). All of these stud ies are mainly con cerned with is sues of stra tig ra phy and of re gional palaeo ge ogra phy. De tailed sedimentological anal y ses of the till and in ter pre ta tion of subglacial con di tions have only rarely been un der taken (e.g., Salamon, 2016a).
This pa per deals with the £aziska site, which pro vides in for - ma tion on the dy nam ics of the mar ginal part of the Scan di na - vian Ice Sheet dur ing the Elsterian Gla ci ation, at the wes tern - most part of the Carpathian fore land (Fig. 1). The site is lo cated at the south ern most sec tor of the Racibórz Ba sin, in the re gion of the Rybnik Pla teau. The lo ca tion of the site is spe cific, ly ing di rectly north from the Moravian Gate, i.e. the mor pho log i cal de pres sion sep a rat ing the Sudetes and Carpathian Moun tains, where a large top o graph i cally-con trolled lobe formed (cf.
Macoun and Králík, 1995; Rùžièka, 2004; Mojski, 2005). There - fore, the study site can pro vide im por tant in for ma tion about the ice be hav iour and its spa tial dis tri bu tion within the en tire Moravian Gate, where the ice sheet was char ac ter ized by many os cil la tions (Macoun and Králík, 1995; Rùžièka, 2004; Nyvlt et al., 2011). Prior to reach ing the mar gin of the moun tains, the ice sheet ad vanced over a sub strate of in creas ingly var ied re lief
* E-mail: tomasz.salamon@us.edu.pl
Received: December 16, 2016; accepted: February 7, 2017; first published online: March 14, 2017
and li thol ogy, which must have sig nif i cantly af fected the ice flow (cf. Salamon, 2016a). At the £aziska site the till was de pos ited over Mio cene clay. This type of ice sheet sub strate was typ i cal of a large part of the Rybnik Pla teau.
The pur pose of this ar ti cle is: (a) to de fine and in ter pret the sed i men tary and struc tural char ac ter is tics of the till within the
£aziska site, (b) to iden tify subglacial con di tions in the area of the south ern part of the Rybnik Pla teau, (c) to re con struct ice flow di - rec tions within the ice sheet, (d) to de ter mine the re la tion ship be - tween ice flow and the to pog ra phy of the sub strate. All these el e - ments help un der stand the pat tern of ice dis tri bu tion and ice sheet dy nam ics in the study area. They also pro vide a broader per spec tive on the dis cus sion of Pleis to cene ice sheet be hav - iour, par tic u larly in ar eas with strongly var ied sub strate re lief.
REGIONAL SETTING
The study area is lo cated in the south ern part of the Rybnik Pla teau, which con sti tutes the high est part of the Raci -
bórz–Oœwiêcim Ba sin. This area is also no tice ably el e vated above the ad ja cent Ostrava Ba sin to the south, which con tin ues to wards the SW into the Moravian Gate (Fig. 1). The Rybnik Pla teau is a hilly, up land area reach ing over 300 m a.s.l. and it is dis sected by river val leys up to 50 m deep. To wards the SE, the Racibórz–Oœwiêcim Ba sin passes into the Silesian Foot hills of in creas ingly var ied to pog ra phy, and these form the fore land of the Silesian Beskid Moun tains (Fig. 1), i.e. one of the ranges of the West ern Carpathian Moun tains. The ax ial part of the study re gion is the Odra Val ley, which sep a rates the Rybnik Pla teau from the G³ubczyce Pla teau, the hilly up land of the East ern Sudetes fore land (Fig. 1A).
The £aziska site is lo cated in the mar ginal part of the Rybnik Pla teau, a few kilo metres to east of the Odra Val ley (Fig. 1B).
The Rybnik Pla teau is bor dered to the south by the Olza River Val ley, which is a right-bank trib u tary of the Odra River. Both rivers form the vast al lu vial plain of the north ern part of the Ostrava Ba sin. The south ern part of the Rybnik Pla teau is dis - sected by the Leœnica River, which flows south wards from the cen tral part of the pla teau, and joins with the Olza River
466 Tomasz Salamon
Fig. 1A – lo ca tion of the study area and ex tent of the Scan di na vian Ice Sheet dur ing the Elsterian Gla ci ation (white dashed line) based on Mojski (2005); B – re lief map of the im me di ate area of the £aziska site; C – sim pli fied geo log i cal cross-sec tion through the Odra Val ley and south ern part of the Rybnik Pla teau (based on Drozd and Trzepla, 2006); D – mor pho log i cal pro file of the ridge stud ied
(Fig. 1B). The south ern, W–E ori ented sec tion of the interfluve ridge be tween the Olza and Leœnica val leys (Fig. 1B) reaches a height of 250–260 m a.s.l. and rises above the Olza val ley at a height of ap prox i mately 40–50 m (Fig. 1D). The study site is lo - cated on this interfluve ridge. It is sit u ated just north of the axis of the ridge near its top and reaches a height of 255 m a.s.l.
The Racibórz–Oœwiêcim Ba sin is part of a foredeep ba sin mainly filled with Neo gene clay de pos its. In the Rybnik Pla teau re gion, these de pos its reach a thick ness of about 200–300 m and over lie Car bon if er ous horsts, which lo cally reach the sur - face (Kotlicka and Kotlicki, 1979). The interfluve ridge stud ied is com posed of Mio cene clay. The hill is cov ered by a thin layer of Qua ter nary de pos its, mainly com pris ing gla cial till ly ing di rectly on the clayey sub strate. The till is lo cally over lain by glaciofluvial sands up to a few metres thick (Fig. 1C).
Only one unit of Elsterian till was rec og nized in the cen tral and east ern parts of the Rybnik Pla teau (Karaœ-Brzozowska, 1963; Klimek, 1972; Lewandowski, 1988). Glaciotectonically de formed mo raine ridges, which oc cur in the ad ja cent south ern part of the G³ubczyce Pla teau (Opava Hilly Land in the Czech Re pub lic) to the west, re cord a few ice-mar ginal os cil la tions (Macoun and Králík, 1995; Nyvlt et al., 2011). In the East ern Sudeten Moun tains, Elsterian till was ob served at a height of over 380 m (Hanáèek and Nývlt, 2009). This till is found at sim i - lar heights in the Silesian Foot hills (Salamon, 2014b). The Odra Val ley and G³ubczyce Pla teau to gether with the north ern part of the Ostrava Ba sin were also cov ered by an ice sheet dur ing the Saalian gla ci ation (Lewandowski 1988; Macoun and Králík, 1995; Badura and Przybylski, 2001; Rùžièka, 2004; Nyvlt et al., 2011; Tyráèek, 2011; Salamon 2014a, 2015a, 2016b).
METHODS
The £aziska site was ex posed for a short time as a re sult of work re lated to the con struc tion of the A1 motor way. The sedimentological and struc tural fea tures of the till and un der ly - ing clayey de pos its of the sub strate were in ves ti gated in de tail.
The in ves ti ga tions in cluded anal y ses of lithofacies prop er ties, till fab rics, ori en ta tion of ki ne matic struc tures, and till pe trog ra - phy. Clast macrofabric data were col lected by mea sur ing the ori en ta tions (strike and dip di rec tions) of the a-axes of at least 30 elon gated peb bles with a min i mum a-axis of 1 cm and an a-axis/b-axis ra tio of at least 1.5. The sam ples were mea sured at ver ti cal in ter vals of 20 to 30 cm. The re sults are shown as rose, con tour and point di a grams, which are based on the Schmidt equal-area grid de vel oped with the StereoNet pro - gram. The eigenvalues (S1 and S3) and eigenvector (v) were cal cu lated for each sam ple (Mark, 1973, 1974). Anal y ses of till pe trog ra phy were per formed on sam ples of at least 300 par ti - cles rang ing from 5 to 10 mm in di am e ter.
RESULTS
£AZISKA SITE
The de pos its were ex posed in the two op po site walls of the ex ca va tion for the motor way, with heights of 6–8 m and lengths of 150 m. Two sep a rate units of Qua ter nary de pos its rest ing di - rectly on Neo gene clays were dis tin guished (Figs. 2 and 3A), the lower of which, a till, is over lain by glaciofluvial sands. The lat ter are lo cally in ter ca lated or over lain by diamictic de pos its (Fig. 2).
DEPOSITS OF THE SUBSTRATE
The gla cial de pos its are un der lain by Neo gene clay – silty clay or rarely, clayey silt and silt (Figs. 3 and 4). The ex po sure dis plays a 1.5 m thick up per sec tion of the sub strate de pos its.
These fine-grained de pos its vary in col our from blue to grey.
Their pri mary lam i na tion is poorly pre served. This is be cause the sed i men tary struc ture is usu ally oblit er ated by a de for ma - tion struc ture of a brec cia type. The de pos its are bro ken into subangular frag ments rang ing in size from a few to sev eral milli - metres, lo cally a few centi metres (Fig. 3C, D). Due to the gen er - ally poor ex po sure of the clays, the thick ness of the layer with de for ma tions is not known. In some places the brec cia struc - ture was ob served to a depth of at least 1 m. The brec cia struc - ture oc curs es pe cially within the clay, and is less com mon within the silty sed i ments, which are mostly ho mo ge neous (Fig. 3E). In di vid ual shear planes, or sets of small, usu ally subhorizontal or slightly ir reg u lar shear planes, a few centi - metres to a few decimetres in length, are su per im posed on the brec cia struc ture (Fig. 3C). Lon ger and more reg u lar subhorizontal shear planes spaced a few centi metres apart are also pres ent lo cally (Fig. 4B). This is not a sed i men tary fea ture, though some shear planes may have ini ti ated along the con - tacts of poorly de vel oped laminae. Large shear planes sev eral metres long were also ob served in a few places. Some of these struc tures are in clined and in con tact with the lower sur face of the diamicton (Figs. 3D and 4A). The in cli na tion of these large shear planes di min ishes pro gres sively to a subhorizontal po si - tion in a W/NW di rec tion.
Small con cen tra tions of dis persed gravel clasts, and more rarely, small lenses of sandy grav els oc cur lo cally within the de - pos its un der ly ing the till (Fig. 4C). The gravel clasts are usu ally 2 to 6 cm (the larg est one >15 cm) in di am e ter and are usu ally well rounded. Quartz clasts are dom i nant with mi nor amounts of lo cal sand stones and mudstones. Scan di na vian rocks were not ob served.
DIAMICTON
The till, 2.5–4 m thick, rests on a clayey sub strate (Fig. 3A, B). This is a silty-sandy diamicton with a small gravel con tent.
The diamicton is mostly dark grey in col our, ex cept its up per - most part, which is rusty in col our. The lower bound ary of the diamicton is slightly un du lat ing or flat (Fig. 3A). Con tact of the diamicton with the un der ly ing de pos its is gen er ally sharp but lo - cally it is gradational, es pe cially where the diamicton over lies silty de pos its (Fig. 4C, D). Three units were dis tin guished in the diamicton. Bound aries be tween suc ces sive units are gradational (Fig. 2).
The lower unit 1 of the diamicton is up to 1 m thick and oc - curs lo cally. The most im por tant fea ture of this unit is nu mer ous in clu sions of silt or silty clay. They form subhorizontal or slightly in clined laminae, mostly from 1 to 20 cm thick and from a few decimetres to >30 m long (Fig. 4C). Lo cally, laminae are very thin and reach only a few milli metres in thick ness. The in fill ing sed i ments are mac ro scop i cally mas sive and ho mo ge neous.
These laminae form strongly at ten u ated folds of sed i ment in - cor po rated into the diamicton from the sub strate. Many of them rep re sent folds com pletely de tached from the sub strate, but in sev eral places rooted folds also oc cur (Fig. 4C, D). In sev eral places unit 1 dis plays large in clu sions of poorly sorted sands and grav elly sands (Fig. 4F). These oc cur in both walls of the ex ca va tion, and are spaced be tween 30–60 m apart. In clu sions
oc cur within the basal part of the diamicton or at the con tact of the diamicton with un der ly ing clay. The in clu sions form ir reg u lar sed i ment bod ies with a lat eral ex tent from 3 to 6 m and a thick - ness from 0.3 to 0.8 m. The in fill ing sed i ments are usu ally de - formed, al though the scale of de for ma tion var ies. As a re sult, their orig i nal sed i men tary struc ture is poorly pre served. The in - fill ing sed i ments are usu ally mas sive, sandy grav els in the lower part, and sands in the up per part. Fold de for ma tions with su per - im posed brit tle de for ma tions, i.e. nu mer ous subhorizontal or slightly in clined shear planes, are pres ent within the sed i ments of the in clu sions (Fig. 4F). Above these in clu sions, thin in ter ca - la tions of silty clays from the sub strate oc cur lo cally within the diamicton (Fig. 4F). The diamicton of unit 1 lo cally re veals a lam i nar struc ture, es pe cially near de for ma tion struc tures, but most of ten the diamicton is mas sive. Nu mer ous frag ments of Neo gene wood oc cur in the diamicton. These are usu ally small, but reach up to 30 cm long in the case of the larg est elon gated frag ment which oc curs in the lower part of unit 1, and is in clined to wards the west at a an gle of ~30°. Its lo cally pol ished sur face dis plays a few sets of small cross-cut ting grooves (Fig. 4E).
Unit 2 is 2–3 m thick and com prises most of the till thick ness (Fig. 2). This is mostly a mas sive, ho mo ge neous diamicton. It con tains in fre quent lam i nar in clu sions of sand (Fig. 4G). The
lon gest laminae con tinue for a dis tance of at least 20 m. Sandy laminae only lo cally re veal small de for ma tion struc tures in the form of small in clined folds or faults (Fig. 4H). In di vid ual subhorizontal string ers of sand a few milli metres thick and up to 1 m long were also ob served in the diamicton in a few other places (Fig. 3E).
Unit 3 is 0.5–1 m thick and com prises the up per part of the diamicton. Lo cally, the lower part of the diamicton of unit 3 is lam i nar; silty-clayey laminae, a few milli metres to 3 cm in thick - ness, oc cur al ter nately with diamictic laminae (Fig. 5A, B). The silty clay is brecciated and its pri mary struc ture is not pre served.
The unit con tains also nu mer ous, var i ously ori ented sets of shear planes and fis sures. The most com mon are subhorizontal shear planes. How ever, lo cally these are ac com - pa nied by in clined struc tures. Some of the shear planes are at least a few metres in length. The up per sur face of the diamicton is eroded and lo cally cov ered with a thin gravel lag.
The gravel frac tion of the till is petro graphi cally dom i nated by quartz (>40%). The sec ond most com mon group con sti tutes dif fer ent types of lo cal sand stones (~25%). Also im por tant are lo cal wood frag ments (8%). North ern (Scan di na vian and the Bal tic de pres sion) rocks oc cur in small fre quen cies (crys tal line rocks 8%, sand stones 10%, lime stones 2%).
468 Tomasz Salamon
Fig. 2. Schematic sedimentary logs of the Quaternary succession at the £aziska site
Lithofacies codes based on Eyles et al. (1983): Dmm — diamicton, matrix-supported, massive; Sh, SGh — sand, gravelly sand, horizontal stratified; Fh — clay, horizontally laminated; Fm — clay, massive
Fig. 3A – overall view of the SE escarpment of the £aziska site; B, C – diamicton resting on Miocene, mostly clayey substrate at the NW escarpment of the site; D – closer view of the clays with distinct breccia structure and superimposed inclined large shear plane; E – closer view of massive silts directly underlying the diamicton B – in the cen tral part a 0.5 m thick layer of mas sive silt oc curs above the clay, the thick ness of silts suc ces sively di min - ished to wards the right, silts con tinue and die out within the lower part of the diamicton; C – mul ti ple gen er a tions of de for - ma tion are vis i ble within the de pos its un der ly ing the diamicton, es pe cially within the clay, a set of shear planes are lo cally su per im posed on the brec cia struc ture; E – small diamictic wedges are vis i ble within the silts, they are in clined to wards the NW and frag mented, the in di vid ual frag ments are dis placed rel a tive to each other along a few small faults
470 Tomasz Salamon
MACROFABRIC
Macrofabric anal y ses were car ried out in the three pro files (Fig. 6). Eight and nine sam ples were mea sured re spec tively in pro files A and B, and 4 in pro file C. The clast align ments gen er - ally in di cate high fab ric strength shown by high eigenvalues S1, which usu ally reach val ues in the range of 0.7–0.8 and, in some cases, even greater than 0.9. Ex cep tions were found in two sam ples of the lower part of pro file A, in which the S1
eigenvalues are dis tinctly lower and range from 0.65–0.67. The S3 eigenvalues are low, rang ing from 0.01–0.07. Clast align - ment is mostly bi modal, rarely unimodal. It can be as sumed that the S1 and S3 eigenvalues are al most con stant through out the whole pro file of the diamicton. The same ap plies to the az i - muths of the mean vec tors, which vary within a very small range and in di cate a main stress di rec tion from W/WNW to E/ESE.
KINEMATIC STRUCTURE ORIENTATION
Plas tic and brit tle ki ne matic struc tures, which oc cur mainly within the lower part of the diamicton, at the diamicton/clay con - tact and within the clays of the sub strate, were mea sured. These are mostly strongly at ten u ated, in clined or re cum bent folds. All the struc tures ana lysed are char ac ter ized by a E/ESE con ver -
gence, so their ori en ta tions are con sis tent with the clast fab ric (Fig. 6). The in cli na tions of large shear planes within the clays un der ly ing the till in di cate a dip to wards the W/NW (Fig. 3C).
Thus, they show also the same di rec tions of prin ci pal stress.
INTERPRETATION
The tab u lar shape, a large lat eral con ti nu ity, the pres ence of Scan di na vian ma te rial and nu mer ous struc tural fea tures, such as re cum bent folds and shear planes, in di cate that the diamicton rest ing on the Mio cene clays is a basal till. The largely oblit er ated pri mary struc ture of the un der ly ing clays sug - gests that the de pos its of the sub strate were de formed at the ice sheet base. The var ied types of de for ma tion and mul ti ple gen er a tions of su per im posed struc tures re flect a multi-stage pro cess of de for ma tion with subglacial con di tions vary ing over time. Ho mo ge neous and mas sive silts or clayey silts, which lo - cally oc cur di rectly be low the diamicton and are en gaged in fold struc tures (Figs. 3B and 4C, D), dis play a plas tic style of de for - ma tion. The most com mon brec cia struc tures, oc cur ring es pe - cially within the clays, rep re sent more brit tle-plas tic sed i ment be hav iour. The strain in this case was dis trib uted among the net work of nu mer ous mul ti ple cracks sep a rat ing the sed i ments into small subangular frag ments, which could ro tate rel a tive to
Fig. 4A – the till over ly ing the clay sub strate at the SE es carp ment of the site. Note the large shear plane within the clay; B – dis tinct subhorizontal shear planes su per im posed on brec cia struc tures; C, D – tran si tional lower con tact of the diamicton. Nu mer ous folds of clayey silts in cor po rated into the diamicton are vis i ble; E – large frag ment of wood ex ca vated from the lower part of the diamicton. Note nu mer ous cross-cut ting grooves on its sur face; F – de formed in clu sion of grav elly sandy de pos its at the diamicton/clay con tact. Duc tile and brit tle de for ma tion struc tures are vis i ble within the in clu sion, which is com posed of two bod ies of coarse-grained sed i ment sep a rated by a silty clay fold in cor po rated from the sub stra tum. This iso cli nal fold in di cates vergence to wards the E/SE. The smaller re cum bent or com pletely de tached folds of silty clay oc cur lo cally within the lower part of the in clu - sion. Clay laminae are pres ent within the diamicton above the in clu sion; G – the two most prom i nent, closely spaced, subhorizontal sandy laminae in the lower part of unit 2; H – rare de for ma tion struc tures of sandy laminae
Fig. 5A, B – Up per till unit. Note thin al ter nat ing laminae of clay and diamicton within the lower part and nu mer ous brit tle de for ma - tion struc tures with dom i nant subhorizontal shear planes; C–E – deposits over ly ing the diamicton. These are mainly poorly sorted sands and grav elly sands with hor i zon tal strat i fi ca tion (C, D). At (E) a pack age of diamictic de pos its, which lo cally over lies the sands, is vis i ble
472 Tomasz Salamon
Fig. 6. Till macrofabric of the £aziska site in vertical sections (A–C)
Location of samples in Figure 2. Note the uniform fabric within the sections. The mean vectors indicate that ice was flowing from the northern part of the Ostrava Basin towards the east, parallel to the axis of the ridge
each other. This is a typ i cal ef fect of clay de for ma tion, which has been ob served in many places in the study area where the ice sheet over rode fine-grained Neo gene strata (Salamon, 2014a, 2015b). The brec cia struc ture in di cates that shear stresses were quite eas ily trans ferred from the ice sheet base into the sub strate, al though in fine-grained sed i ments such a pro cess is not com mon (cf. Tulaczyk, 1999). This sug gests that, at least for a short pe riod, ice/bed cou pling was quite strong.
From this it fol lows that the amount of wa ter in the ice sheet base dur ing this pe riod may not have been large, de spite the low per me abil ity of the sub strate, oth er wise wa ter of pro gres - sively in creas ing pres sure would have quickly iso lated the ice from the bed and re stricted shear trans fer to the bed (cf. Iverson et al., 1999, 2003; Fischer and Clarke, 2001). The small wa ter con tent in the ice sheet base may sug gest that the brec cia struc ture could have de vel oped dur ing the ini tial phase of ice sheet over rid ing in the study area. How ever, such a struc ture could de velop within the clay quite quickly, and so the pre cise de ter mi na tion of the time of its or i gin is dif fi cult. For the same rea sons, it is not pos si ble to pre cisely de ter mine the scale of the fi nite strain of the clayey de pos its. The brec cia struc ture may be a re sult of short-term de for ma tion, and may re flect lim ited sed i - ment dis place ment, es pe cially in the case of the dis tinctly an gu - lar char ac ter of the ag gre gates. How ever, it may also ef fec tively mask many ear lier de for ma tion ep i sodes, if these oc curred.
Shear planes su per im posed on brecciated clays in di cate a more brit tle re sponse of sed i ments to ap plied stress. This may sug gest an in crease in the shear strength of the sed i ments. The de vel op ment of shear planes could have been pro moted lo cally by a greater con cen tra tion of clay in some pri mary laminae.
This con clu sion re fers es pe cially to the reg u lar, subhorizontal shear planes (Fig. 4B). Their cor re la tion with the pri mary lam i - na tion sug gest a rather re stricted in ten sity of clay de for ma tion.
The tim ing of shear plane de vel op ment is also dif fi cult to es tab - lish. It may be re lated to a much later phase when the clays were over lain by till. How ever, the only con fi dent as ser tion is that they were formed later than the brec cia struc ture. Par tic u - larly in ter est ing are the larg est shear planes lo cated be low the till (Figs. 3C and 4A) which re flect a com pletely non-plas tic be - hav iour of the clay. The fail ure of the sed i ments along a dis tinct decollement, lo cated at some depth be low the diamicton, may in di cate a pro gres sive stiff ness of clays, which ex pe ri enced stron ger com pac tion.
The source of the gravel clasts oc cur ring within the clays be low the till is prob lem atic in in ter pre ta tion. Salamon (2015b) in ter preted in di vid ual peb bles in sim i lar po si tions to be the re sult of clasts sink ing into the sub strate af ter pre vi ously be ing lodged at the till/sub strate con tact. In this case, such an in ter pre ta tion seems un likely be cause the grav els form dis tinct con cen tra - tions or even small in clu sions. More over, al though a gla cial or i - gin of grav els can not be ex cluded, the lack of north ern ma te rial within these de pos its does not fa vour such an in ter pre ta tion. In turn, the highly-rounded clasts and dom i nant quartz and lo cal sand stones may sug gest that the gravel is as so ci ated with a re - sid ual Plio cene or Early Pleis to cene al lu vial cover de pos ited in front of the moun tains. Rem nants of such de pos its are known from the fore-moun tain area of South ern Po land and also from the Rybnik Pla teau (Makowski, 1936; Klimek, 1972). In this case, the grav els may oc cur close to their orig i nal po si tion or were trans ported for some dis tance in a subglacial de for ma tion ho ri zon.
The gradational lower con tact of unit 1, and nu mer ous de - for ma tion struc tures such as re cum bent and in clined folds, and the laminae of un der ly ing sed i ments in cor po rated from the sub - strate to the till, re flect the pro cess of subglacial shear ing (cf.
Hart and Boulton, 1991; Hart and Rob erts, 1994; Rob erts and
Hart, 2005; Lee and Phillips, 2008). The laminae are the re sult of a strong at ten u a tion of folds ini ti ated at the diamicton/clay con tact as a re sult of lo cal com pres sion (cf. Boulton, 1996;
Boulton et al., 2001; Van der Wateren, 2002). The lay ered struc ture of the diamicton oc cur ring above is a tec tonic fo li a tion and re flects more in tense de for ma tion in the up per part of the subglacial shear zone (Van der Wateren et al., 2000). These de for ma tion struc tures re flect a ver ti cal suc ces sion of strain typ i cal of clas sic de for ma tion till (cf. Benn and Ev ans, 1996;
Van der Wateren et al., 2000; Van der Wateren, 2002).
Large in clu sions of sands and grav elly sands that oc cur at the diamicton/clay con tact or in the lower part of the diamicton most prob a bly re flect de po si tion from subglacial flows. The in - clu sion infills prob a bly re cord a dis trib uted-ca nal drain age sys - tem, which de vel oped at the ice sheet base (cf. Clark and Walder, 1994; Walder and Fowler, 1994; Ng, 2000; Salamon, 2015b). The ca nal sed i ments are de formed as a re sult of subglacial shear ing. Folds with su per im posed sets of nu mer - ous shear planes within the sands in di cate the brit tle-plas tic na - ture of de for ma tion. Quite well-pre served sandy in clu sions sug - gest, in gen eral, a re stricted in ten sity of de for ma tion and fi nite strain. How ever, silty laminae within the diamicton just above the in clu sions are the re sult of much stron ger fold ing and fur ther stretch ing of silt folds, which were in cor po rated into the diamicton in the vi cin ity of ca nal lithosomes.
The de formed na ture of the diamicton is con sis tent with the wood com po nent within the diamicton (Fig. 4E). This ma te rial, with low re sis tance to gla cial abra sion, is un likely to be pre - served in such good con di tion fol low ing very high basal fric tion.
The in clined po si tion of the clast also does not in di cate a lodge - ment pro cess from the slid ing ice base. Fur ther more, sets of cross-cut ting grooves at the clast sur face in di cate at least a few ep i sodes of clast ro ta tion, such as is likely to have taken place within a de for ma tion ho ri zon. The till fab ric from two lower sam - ples of pro file A may sup port shear ing pro cesses oc cur ring within the till. The S1 eigenvalues of 0.65–0.67 in di cate mod er - ate clast ar range ment, which is typ i cal of de for ma tion till (Dowdeswell and Sharp, 1986; Hart, 1994; Benn and Ev ans, 1996; Hicock et al., 1996). How ever, in pro files B and C fab ric strength is much higher (S1 approx. 0.8). This may in di cate that de for ma tion was spa tially re stricted, an in ter pre ta tion that is sup ported by the lat eral ex tent of unit 1.
The mas sive struc ture of unit 2 is dif fi cult to in ter pret. A lack com mon de for ma tion struc tures and a gen er ally high fab ric strength do not in di cate in tense shear ing in a thick de form ing bed. High S1 eigenvalues which ad di tion ally are rel a tively con - stant and vary within a small range through out the till, may re - flect a pro cess of suc ces sive ac cre tion of de pos its merely from the thin de for ma tion ho ri zon (Larsen et al., 2004). This may be con firmed by a high fab ric strength, which is not typ i cal of a thick de for ma tion till (cf. Dowdeswell and Sharp, 1986; Hart, 1994;
Clark, 1997). A strong fab ric would ac tu ally in di cate re stricted pos si bil i ties for clast ro ta tion (cf. Hart et al., 2009). Lo cally, a very high fab ric strength (S1 > 0.9) sug gests very high strain within the ice sheet base and may in di cate that sed i ments were lodged from a slid ing ice base. Sand string ers, es pe cially the long subhorizontal laminae, prob a bly re cord subglacial wa ter flow in a thin film along the ice/bed con tact sur face (Brown et al., 1987; Piotrowski and Tulaczyk, 1999; Munro-Stasiuk, 2000;
Wysota, 2002; Piotrowski et al., 2006; Salamon, 2015b). This re flects very high subglacial wa ter pres sure, which was equal to or even higher than the over bur den ice pres sure. These con di - tions could have fa voured basal slid ing (cf. Brown et al., 1987;
Piotrowski and Tulaczyk, 1999; Piotrowski et al., 2001; Narloch et al., 2012, 2013). The rar ity of sandy laminae may in di cate that subglacial flow through the wa ter film oc curred ep i sod i cally,
or that most of the de pos its con nected with such flow events are not pre served be cause of later shear ep i sodes. Based on the de scribed fea tures, unit 2 is in ter preted as trac tion till (Ev ans et al., 2006).
Unit 3 is also in ter preted as a basal till. In this case, the rare silty clay laminae are par tic u larly in ter est ing. Most prob a bly they are the re sult of de po si tion from sus pen sion, from a thin col umn of wa ter in subglacial cav i ties. They in di cate tem po ral lo cal sep a ra tion of the ice from the bed. This pro cess re flects a very high basal wa ter pres sure in the ice sheet base bal anc ing over bur den pres sure. Con versely to the pre vi ous pe riod, wa ter was lo cally stored at the ice sheet base. Ac cu mu la tion of silty clay from sus pen sion was in ter mit tent, with the de po si tion of diamictic laminae re leased from the ice sheet base. This type of suc ces sion, es pe cially the well-pre served silty clay laminae, may in di cate de po si tion dur ing tem po ral ice sheet stag na tion.
The brecciated struc ture and nu mer ous subhorizontal shear planes re flect the brit tle de for ma tion of pre vi ously de pos ited sed i ments and in di cate a change of ice sheet be hav iour. How - ever, when the ice sheet was again ac tive, its dy nam ics was rather re stricted. Brit tle be hav iour of the sed i ments in re sponse to the ap plied stress may in di cate a stron ger com pac tion of the diamicton.
DEPOSITS ABOVE THE DIAMICTON
The re main der of the suc ces sion within the £aziska site is poorly ex posed. The up per sur face of the diamicton is over lain by 3–6 m of poorly-sorted sands and grav elly sands (Fig. 5C).
These de pos its are mostly hor i zon tally strat i fied or mas sive (Fig. 5D). In a few places, the sands are in ter ca lated with thin beds of diamictic sand or sandy silt. The diamictic sands or sandy diamictons up to 1.5 m thick oc cur lo cally above the sands (Fig. 5E). These rep re sent glaciofluvial sed i ments de - pos ited from shal low sheet flows most prob a bly at the sur face of a terminoglacial fan. The diamictic sed i ments were in ter - preted as flow tills. The whole unit was de pos ited af ter ice sheet re ces sion and may re cord later ice mar gin os cil la tion or a youn - ger (Saalian) gla ci ation.
DIRECTIONAL DATA FROM OTHER SITES
Di rec tional data mea sured within Elsterian till are also known from a few other sites (Salamon, 2001, 2014b, 2016a). The near - est is the Koñczyce site, lo cated about 20 km to wards the SE, a
few kilo metres in front of the mar gin of the Silesian Foot hills. The ice sheet ad vanced over an open area of fore-moun tain al lu vial fan. The till from the site is char ac ter ized by very high fab ric strength; S1 is usu ally higher than 0.8 (Salamon, 2016a). Mean az i muths in di cate ice flow di rec tion from NW to SE (Fig. 7). Sim i - lar ori en ta tions were re corded from other ki ne matic struc tures such as folds. The Wieszczêta and Œwiêtoszówka sites are lo - cated far ther SE in the area of the Silesian Foot hills (Fig. 7). Two till units re ported at the Wieszczêta site are cor re lated with ice os - cil la tion in the zone of the foot hills mar gin (Salamon, 2001).
Mean az i muths of clasts to wards the east, within the lower till, in - di cate that ice flowed through the nar row val ley from the larger Vistula Val ley. In the up per till the clast fab ric in di cates a gen er - ally sim i lar di rec tion of ice flow, i.e. to wards the SE, prob a bly over the mar ginal hill of the foot hills. At the Œwiêtoszówka site, which is lo cated much far ther south, the fab ric was not mea sured, how - ever, mesoscale fold struc tures are gen er ally ori ented to wards the south.
DISCUSSION
ICE FLOW DIRECTIONS
It is com monly ac cepted that clast ori en ta tion re flects the di - rec tion of the main stress ex erted on the sed i ments by mov ing gla ciers (cf. Dowdeswell and Sharp, 1986; Hicock and Dreimanis, 1992; Hart, 1994; Ev ans et al., 2006). A-axes of clasts are usu ally ori ented par al lel to the stress field di rec tion (cf. Hooyer and Iverson, 2000). How ever, Tay lor’s model (1923) of clast ro ta tion de scribed also trans verse ori en ta tion.
This type of clast ar range ment may oc cur spe cif i cally within till de pos ited within a compressional stress field re gime (e.g., Boulton, 1970; Mark, 1974; Allen, 1984). This shows that sed i - ment re sponse to the ap plied stress may be com plex (cf. Carr and Rose, 2003). More un equiv o cal data, which may di rectly in - di cate the di rec tion of ice flow, is pro vided by ki ne matic struc - tures with a dis tinct dip and ori en ta tion, such as re cum bent folds (Van der Wateren et al., 2000). Based on the uni form ori en ta - tion of till fab ric and nu mer ous ki ne matic struc tures at the
£aziska site, it can be con cluded that the mean vec tors of clast a-axes cor re spond with the main di rec tion of the act ing stress, and thus re flect real ice flow di rec tion. All di rec tional data from the £aziska site in di cate that ice flowed gen er ally from the west, from the north ern part of the Ostrava Ba sin. Such a pat tern of
474 Tomasz Salamon
Fig. 7. Ice flow direction at the study area based on directional data from: the £aziska site; the Koñczyce site (Salamon, 2016a);
the Wieszczêta site (Salamon, 2001); and the Œwiêtoszówka site (Salamon, 2014b)
flow di rec tion is par tic u larly im por tant be cause the till is lo cated at the top of a ridge. This in di cates that ice was flow ing above this elon gated subglacial ob sta cle, par al lel to its axis. The val - ues of mean vec tors vary in a very nar row range through out the en tire ver ti cal pro file of the till and in di cate an al most con stant di rec tion of ice flow dur ing pro gres sive till ac cre tion.
SUBGLACIAL CONDITIONS AT THE CLAY RIDGE STUDIED
The lack of large glacitectonic de for ma tion struc tures and a basal till de pos ited gen er ally un der high basal wa ter pres sure con di tions in di cate that the stud ied ridge of the south ern part of the Rybnik Pla teau did not pro vide sig nif i cant re sis tance for the ad vanc ing ice sheet, de spite the fact that it con sti tuted a dis tinct top o graphic ob sta cle. The char ac ter is tics of the ice/bed re la - tion ship were con di tioned by a clayey sub strate with very low per me abil ity, which, due to a con tin u ously high basal wa ter pres sure, re stricted the trans fer of shear stress to the sub strate and even oc ca sion ally fa voured the ice de coup ling from the bed (cf. Iverson et al., 1994, 1995, 2003; Fuller and Murray, 2000).
The brec cia struc ture of the sub-till clay is some what ex cep - tional, and in di cates re stricted wa ter con tent at the ice sheet base dur ing the first phase of ice sheet over rid ing the ridge stud ied. How ever, it is most likely that this ep i sode was very short-lived. The de vel op ment of the dis trib uted-ca nal sys tem evac u at ing ex cess wa ter from the ice sheet base in di cates the quick in crease and re lease of basal wa ter pres sure. This fol - lows from the lo ca tion of sandy gravely in clu sions at the diamicton/clay con tact or within the lower part of the diamicton.
The de vel op ment of the subglacial drain age sys tem, which ef fi - ciently re duced subglacial pore-wa ter pres sure, could have ini ti - ated the pro cess of clay com pac tion as well as stiff ness and sub se quently brit tle de for ma tion. Shear planes within the clay re flect oc ca sion ally stron ger cou pling of the ice with the bed.
How ever, rare sandy laminae and string ers within unit 2, which are cor re lated with flow as so ci ated with the wa ter film, in di cate that gen er ally high basal wa ter pres sure con di tions pre vailed at the ice sheet base. The study case is sim i lar to that de scribed from an other part of the Racibórz–Oœwiêcim Ba sin for the Saalian Gla ci ation, where a subglacial ob sta cle may even have fa voured ice sheet slid ing (Salamon, 2015b). How ever, in this case, ep i sodes of ice/bed sep a ra tion were less prom i nent.
ICE DISTRIBUTION WITHIN THE MORAVIAN GATE LOBE
Con sid er ing the fact that dur ing the Elsterian Gla ci ation the ice sheet reached the mar gins of the Sudetes and Carpathian Moun tains, it seems that the flow of the ice from north to south was gen er ally un im peded. How ever, this study in di cates that af - ter reach ing a more var ied sub strate to pog ra phy, the ice flow be came more com plex. The di rec tional data col lected al lows a broader con clu sion about the dis tri bu tion of ice at the study area and in its wider vi cin ity. The data from the £aziska site are par - tic u larly im por tant. The east ern di rec tion of ice flow in di cates that ice was sup plied to the study area not di rectly from the north, across the Rybnik Pla teau, but most prob a bly from the neigh bour ing Odra Val ley (Figs. 7 and 8B). Such an in ferred ice flow pat tern sug gests that this rel a tively nar row cor ri dor con sti - tutes the main tran sit zone of ice to the whole area of the Moravian Gate. Ice flowed through the Odra Val ley prob a bly as a nar row ice stream to wards the north ern area of the Ostrava Ba sin, from where it was fur ther ra di ally dis trib uted in dif fer ent di rec tions, suc ces sively fill ing the mor pho log i cal de pres sion of the Ostrava Ba sin and later the en tire Moravian Gate. The ice dis tri bu tion prob a bly was sim i lar to that of mod ern piedmont
gla ciers, al though the ice did not flow down but ad vanced gen - er ally over flat or a slightly re versely in clined sur face. Di rec - tional data from the Koñczyce site (Salamon, 2016a) in di cate that ice was spread ing far to wards the SE (Fig. 7). It is also pos - si ble that the ice that reached the Wieszczêta site, lo cated in the more east ern part of the Silesian Foot hills (Fig. 7), also flowed over the same path way and was sup plied from the north ern part of the Ostrava Ba sin. A di ver gent style of ice dis tri - bu tion re flects a large ice sup ply and sug gests that ice flow through the Odra Val ley could be rel a tively fast.
THE INFLUENCE OF THE HILLY AREA OF THE RYBNIK PLATEAU ON ICE SHEET BEHAVIOUR
The east ern di rec tion of ice flow es tab lished at the £aziska site shows some re la tion ship be tween the to pog ra phy and the pat tern of ice dis tri bu tion. Ice flow di rec tion, par al lel to the axis of the elon gated study ridge, in di cates that it did not con sti tute a sub stan tial bar rier for the ice sheet, which moved over the hill quite eas ily. Al most all the di rec tional data in di cate that ice was con stantly flow ing over the hill from the same di rec tion. This spe cific ice flow pat tern re corded at the south ern mar gin of the Rybnik Pla teau al lows one to in di rectly in fer ice sheet dy nam ics for the rest of the Rybnik Pla teau. This is be cause the subglacial con di tions in ter preted from the £aziska site the o ret i - cally could also be ex pected at other parts of the hilly area of the Rybnik Pla teau, where the till was mostly de pos ited over a clayey sub strate. How ever, de spite the Neo gene clays of the sub strate, which cre ated high basal pres sure con di tions fa vour - ing rapid basal move ment, the flow of the ice in this area was rather char ac ter ized by low dy nam ics. In this case, the ice would eas ily and quickly flow across the whole Rybnik Pla teau from north to south. In stead, the con cen tra tion of ice flow within the Odra Val ley sug gests that this low land cor ri dor, filled with coarser grained al lu vial sed i ments, gen er ally pro duced lower flow re sis tance than the el e vated hilly pla teau com posed of clays. In ferred low ice sheet mo bil ity in much of the Rybnik Pla - teau area prob a bly re sulted from the fact that the more north ern part of the pla teau was lo cated at a higher level (~40–50 m in com par i son to the £aziska site). The till within the neigh bour ing north palaeovalley of the Ruda River rests >100 m lower in com par i son to the clay sub strate of the high est part of the Rybnik Pla teau (cf. Kleczkowski et al. 1972). Such large vari a - tions in height caused the ice to be con cen trated within the sur - round ing val leys. The Ruda palaeovalley prob a bly con sti tuted the ax ial part of the sec ond cor ri dor of quickly flow ing ice (Fig. 7), through which ice was dis trib uted from the north ern part of the Racibórz Ba sin to wards the Oœwiêcim Ba sin and far - ther into the Carpathian Foot hills.
FACTORS CONTROLLING ICE DISTRIBUTION IN THE STUDY AREA
The in ferred ice flow pat tern within the mar ginal part of the Elsterian ice sheet prob a bly evolved over time. The het er o ge - ne ity of the ice flow re sulted from stress pat terns within the ice sheet, a con se quence of spa tially var ied fric tional re sis tance be tween the ice and the bed. Al though the most im por tant con - trol ling fac tor was to pog ra phy, a var ied sub strate li thol ogy and the ice mar gin/per ma frost re la tion ship, which de ter mined the strength of ice/bed cou pling, may also have played a sig nif i cant role (cf. Salamon, 2014a, 2015a, 2016a). If the zones that in - duced dis tinctly var ied stress fields ex isted close to each other, then such in ter nal ice streams could de velop within the ice sheet. This hy po thet i cal pro cess is pre sented as a model in Fig - ure 8. It shows that, when the ice sheet reached high top o -
476 Tomasz Salamon
Fig. 8. Hypothetical model of ice flow distribution within the marginal part of the ice sheet in the area of the Racibórz–Oœwiêcim Basin and its vicinity during the Elsterian Glaciation
graphic bar ri ers, its flow over them to wards the south was stopped for some time and be came pos si ble only af ter the ice reached suf fi cient thick ness. Ob struc tions to ice flow be hind the lo cal bar ri ers caused in creased ac tiv ity else where, where the sub strate gen er ated smaller fric tional re sis tance. The nar row zones of con cen trated flow were de vel oped in this way. This pro cess could have oc curred, for ex am ple, within the north ern fore land of the Nisky Jesenik Hills (Fig. 8A, B). When the ice sheet reached this high moun tain edge, fur ther move ment of ice to wards the south was not pos si ble. In creas ing mar ginal com pres sion most prob a bly caused ice flow re or ga ni za tion in its hin ter land, and con se quently, ice flow con cen tra tion within the neigh bour ing Ostrava Ba sin. This could have caused the ice sheet mar gin sup ply in the Nisky Jesenik sec tor to di min ish, while an in tense and con cen trated flow of ice through the Odra Val ley to wards the Ostrava Ba sin caused the de vel op ment of the Moravian Gate lobe (Fig. 8B). Sim i lar phe nom ena could have oc curred be tween the neigh bour ing val leys of the Odra and Ruda rivers, which were sup plied by sep a rate ice streams.
The top o graph i cally re stricted Moravian Gate lobe, af ter reach - ing the edge of the sur round ing hills, had lim ited pos si bil ity of fur ther flow to wards the south. As a re sult, the sup ply ing ice stream could at the same time have ceased caus ing the en - hanced ac ti va tion of the Ruda ice stream, which might have be - come the main ar tery through which ice was flow ing to wards the large and open area of the Oœwiêcim Ba sin, and the Silesian Foot hills lo cated far ther south (Fig. 8C). More over, a tem po ral lack of sup ply to the in ac tive part of the ice sheet could have in - duced lo cal ice mar gin re ces sion. As a re sult, the ice/bed re la - tion ship had to change in this area over time. This could have sig nif i cantly changed the stress field within the ice sheet and thus the in flu ence on ice flow be hav iour, and even could have led to switch ing be tween ad ja cent streams. It is pos si ble that the switch ing be tween neigh bour ing streams might have been re peated (Fig. 8C, E). There is no clear ev i dence for this pro - cess, though a phase of pas sive melt ing and later phase of ice sheet re ac ti va tion re corded within the up per unit of the till in
£aziska site may be con sis tent with this phe nom e non. Mul ti ple ice sheet re advan ces within the Ostrava Ba sin and Moravian Gate, and es pe cially ice sheet mar gin os cil la tion within glaciotectonically de formed mo raine ridges of the south ern part of the G³ubczyce Pla teau (Opava Hilly Land; Macoun and Králík, 1995; Rùžièka, 2004; Nyvlt et al., 2011), may also be re -
lated to this spe cific ice sheet be hav iour gov erned by a com plex ice/sub strate re la tion ship.
CONCLUSIONS
Study of till at the £aziska site in the south ern part of the Racibórz–Oœwiêcim Ba sin pro vides in for ma tion about ice dis tri - bu tion, subglacial con di tions and the dy nam ics of the Scan di - na vian ice sheet in the di rect fore land of the moun tains dur ing the Elsterian Gla ci ation.
The sedimentological and struc tural fea tures of the till in di - cate that rel a tively high basal pres sure con di tions oc curred at the ice sheet base over the clayey ridge stud ied, on the south - ern part of the Rybnik Pla teau. These con di tions fa voured basal mech a nisms of ice sheet move ment and al lowed for easy ice flow over the ridge. How ever, the ice sheet be hav iour over the rest of the Rybnik Pla teau was much less dy namic.
Re con struc tion of palaeo-ice flow di rec tions in di cates that the move ment of the ice sheet in the area stud ied was com plex.
This re sulted mainly from the fact that, be fore reach ing the mar - gin of the moun tains of south ern Po land, the ice sheet was ad - vanc ing over a sub strate of in creas ingly var ied re lief. Large mor pho log i cal and lithological vari a tions in the ice sheet sub - strate in duced dif fer en tial fric tional re sis tance pro duc ing a spa - tially var ied stress field within the mar ginal part of the ice sheet.
This al lowed the for ma tion of pas sive zones of ice be hind sig nif - i cantly large mor pho log i cal bar ri ers and ac tive zones of stream - ing ice within the area, pro duc ing a dis tinctly lower flow re sis - tance.
In the top o graph i cally var ied area, where more than one pref er en tial ac tive zone of flow ing ice (ice stream) oc curred, lo - cal os cil la tion of the ice sheet mar gin did not have to be the re - sult of ice sheet mass bal ance, but may have been the con se - quence of switch ing be tween neigh bour ing ice streams, al ter - nately con vey ing ice to wards the ice sheet mar gin. Melt ing of the pas sive ice zone and its in ter mit tent lack of sup ply might have in duced lo cal re ces sion of the ice sheet mar gin.
Ac knowl edge ments. I would like to thank J. Rose and an anon y mous re viewer for their help ful and con struc tive com - ments. Many thanks are ad dressed to T. Zieliñski for dis cus sion dur ing prep a ra tion of the manu script.
REFERENCES
Allen, P., 1984. Field Guide to the Gipping and Waveney Val leys.
Qua ter nary Re search As so ci a tion. Cam bridge.
Badura, J., Przybylski, B., 1998. Ex tent of the Pleis to cene ice sheets and deglaciation be tween the Sudeten and the Silesian Ram part (in Pol ish with Eng lish sum mary). Biuletyn Pañstwowego Instytutu Geologicznego, 385: 9–28.
Badura, J., Przybylski, B., 2001. Prob lem korelacji stratygraficznych kenozoiku po³udniowej Opolszczyzny (in Pol - ish). VIII Konferencja: Stratygrafia plejstocenu Polski. Serie rzeczne i lodowcowe po³udniowej Opolszczyzny. Jarno³tówek:
9–27.
Benn, D.I., Ev ans, D.J.A., 1996. The in ter pre ta tion and clas si fi ca - tion of subglacially-de formed ma te ri als. Qua ter nary Sci ence Re views, 15: 23–52.
Boulton, G.S., 1970. Till gen e sis and fab ric in Svalbard, Spitsbergen. In: Till: a Sym po sium (ed. R.P. Goldthwait): 41–72.
Ohio State Uni ver sity Press, Co lum bus, OH.
Boulton, G.S., 1996. The or i gin of till se quences by subglacial sed i - ment de for ma tion be neath mid-lat i tude ice sheets. An nals of Gla ci ol ogy, 22:75–84.
Boulton, G.S., Dobbie, K.E., Zatsepin, S., 2001. Sed i ment de for - ma tion be neath gla ciers and its cou pling to the subglacial hy - drau lic sys tem. Qua ter nary In ter na tional, 86: 3–28.
Brown, N.E., Hallet, B., Booth, D.B., 1987. Rapid soft bed slid ing of the Puget Gla cial Lobe. Jour nal of Geo phys i cal Re search, 92:
9059–9082.
Butrym, J., Maruszczak, H., Wojtanowicz, J., 1988.
Thermoluminescence chro nol ogy of the Sanian (=Elsterian) in - land-ice de pos its in the San and up per Dniester River bas ins (in
Pol ish with Eng lish sum mary). Annales Societatis Geologorum Poloniae, 58: 191–205.
Carr, S.J., Rose, J., 2003. Till fab ric pat terns and sig nif i cance: par ti - cle re sponse to subglacial stress. Qua ter nary Sci ence Re views, 22: 1415–1426.
Clark, P.U., 1997. Sed i ment de for ma tion be neath the Laurentide Ice Sheet. In: Late Gla cial and Postglacial En vi ron men tal changes: Qua ter nary, Car bon if er ous–Perm ian, Pro tero zoic (ed.
I.P. Mar tini): 81–97. Ox ford Uni ver sity Press, New York.
Clark, P.U., Walder, J.S., 1994. Subglacial drain age, eskers and de form ing beds be neath the Laurentide and Eur asia ice sheets.
GSA Bul le tin, 106: 304–314.
Czubla, P., 2013. Analiza frakcji >20 mm w glinie ze ¿wirowni w Mokrzeszowie (in Pol ish). XX Konferencja Stratygrafia Plejstocenu Polski, Plejstocen przedpola Sudetów Œrodkowych, Lasocin: 171–174.
Dowdeswell, J.A., Sharp, M.J., 1986. Char ac ter iza tion of peb ble fab rics in mod ern ter res trial glacigenic sed i ments.
Sedimentology, 33: 699–710.
Drozd, M., Trzepla, M., 2006. Szczegó³owa mapa geologiczna Polski w skali 1: 50 000, ark. Zabe³ków (in Pol ish). Pañstwowy Instytut Geologiczny, Warszawa.
Ev ans, D.J.A., Phillips, E.R., Hiemstra, J.F., Auton, C.A., 2006.
Subglacial till: for ma tion, sed i men tary char ac ter is tics and clas - si fi ca tion. Earth-Sci ence Re views, 78: 115–176.
Eyles, N., Eyles, C.H., Miall, A.D., 1983. Lithofacies types and ver - ti cal pro file mod els; an al ter na tive ap proach to the de scrip tion and en vi ron men tal in ter pre ta tion of gla cial diamict and diamictite se quences. Sedimentology, 30: 393–410.
Fischer, U.H., Clarke, G.K.C., 2001. Re view of subglacial hy - dro-me chan i cal cou pling: Trapridge Gla cier, Yu kon Ter ri tory, Can ada. Qua ter nary In ter na tional, 86: 29–43.
Fuller, S., Murray, T., 2000. Ev i dence against per va sive bed de for - ma tion dur ing the surge of an Ice lan dic gla cier. Geo log i cal So ci - ety Spe cial Pub li ca tions, 176: 203–216.
Hanáèek, M., Nývlt, D., 2009. Subglacial gravely tills from Jindøichov in the Osoblaha Spur. Èasopis Slezského Zemského Muzea, Opava (A), 58: 193–214.
Hart, J.K., 1994. Till fab ric as so ci ated with de form able beds. Earth Sur face Pro cesses and Land forms, 19: 15–32.
Hart, J.K., Boulton, G.S., 1991. The in ter re la tion of glaciotectonic and glaciodepositional pro cesses within the gla cial en vi ron - ment. Qua ter nary Sci ence Re views, 10: 335–350.
Hart, J.K., Rob erts, D.H., 1994. Cri te ria to dis tin guish be tween subglacial glaciotectonic and glaciomarine sed i men ta tion: I.
De for ma tion styles and sedimentology. Sed i men tary Ge ol ogy, 91: 191–213.
Hart, J.K., Rose, K.C., Mar ti nez, K., Ong, R., 2009. Subglacial clast be hav iour and its im pli ca tion for till fab ric de vel op ment:
new re sults de rived from wire less subglacial probe ex per i - ments. Qua ter nary Sci ence Re views, 28: 597–607.
Hicock, S.R., Dreimanis, A., 1992. De for ma tion till in the Great Lakes re gions: Im pli ca tions for rapid flow along the south-cen - tral mar gin of the Laurentide Ice Sheet. Ca na dian Jour nal of Earth Sci ences, 29: 1565–1579.
Hicock, S.R., Goff, J.R., Lian, O.B., Lit tle, E.C., 1996. On the in ter - pre ta tion of subglacial till fab ric. Jour nal of Sed i men tary Re - search, 66: 928–934.
Hooyer, T.S., Iverson, N.R., 2000. Clast-fab ric de vel op ment in a shear ing gran u lar ma te rial: im pli ca tions for subglacial till and fault gouge. GSA Bul le tin, 112: 683–692.
Iverson, N.R., Jansson, P., Hooke, R., Le, B., 1994. In situ mea - sure ment of the strength of de form ing subglacial sed i ment.
Jour nal of Gla ci ol ogy, 40: 497–503.
Iverson, N.R., Hanson, B., Hooke, R.L., Jansson, P., 1995. Flow mech a nism of gla ciers on soft beds. Sci ence, 267: 80–81.
Iverson, N.R., Baker, R.W., Hooke, R.LeB., Hanson, B., Jansson, P., 1999. Cou pling be tween a gla cier and a soft bed: I. A re la tion be tween ef fec tive pres sure and lo cal shear stress de ter mined from till elas tic ity. Jour nal of Gla ci ol ogy, 45: 31–40.
Iverson, N.R., Co hen, D., Hooyer, T.S., Fischer, U.H., Jack son, M., Moore, P.L., Lappegard, G., Kohler, J., 2003. Ef fects of basal de bris on gla cier flow. Sci ence, 301: 81–84.
Jahn, A., 1952. The pro file of the Pleis to cene de pos its in Góry Kêckie near Kety (Carpathians) (in Pol ish with Eng lish sum - mary). Biuletyn Pañstwowego Instytutu Geologicznego, 65:
467–477.
Karaœ-Brzozowska, C., 1963. Limit of the Mid dle-Pol ish (Saale) gla ci ation in the Racibórz Ba sin (in Pol ish with Eng lish sum - mary). Przegl¹d Geograficzny, 35: 431–446.
Kleczkowski, A.S., Dendewicz, S., Dyjor, J., Kowalski, J., Mi³kowski, M., 1972. Plio cene-Qua ter nary Rybnik-KoŸle Trough and its hydrogeological prop er ties. Bul le tin de l’Academie Po lo naise des Sci ences, Série des Sci ences de la Terre, 20: 71–83.
Klimaszewski, M., 1952. The prob lems of the Pleis to cene in South - ern Po land (in Pol ish with Eng lish sum mary). Biuletyn Pañstwowego Instytutu Geologicznego, 65: 137–268.
Klimek, K., 1972. Kotlina Raciborsko-Oœwiêcimska. In:
Geomorfologia Polski, 1 (ed. M. Klimaszewski): 116–138. PWN, Warszawa.
Kotlicka, G.N., Kotlicki, S., 1979. Mapa geologiczna Polski w skali 1: 200 000, ark. Gliwice. Wyd. Geol., Warszawa.
Krzyszkowski, D., Ibek, M., 1996. Mid dle Pleis to cene sed i men ta - tion and palaeo ge ogra phy of the Dzier¿oniów Ba sin, Sudetic Fore land, South west ern Po land. Annales Societatis Geologorum Poloniae, 66: 35–58.
Krzyszkowski, D., Karanter, R., 2001. Gla cial sed i ments and palaeo ge ogra phy of the north west ern fore land of the Mount Œlê¿a mas sif, South west ern Po land. In: Late Caino zo ic Stra tig - ra phy and Palaeo ge ogra phy of the Sudetic Fore land (ed. D.
Krzyszkowski): 109–145. WIND, Wroc³aw.
Larsen, N.K., Piotrowski, J.A., Kronborg, C., 2004. A multiproxy study of a basal till: a time-transgressive ac cre tion and de for ma - tion hy poth e sis. Jour nal of Qua ter nary Sci ence, 19: 9–21.
Lee, J.R., Phillips, E.R., 2008. Pro gres sive soft sed i ment de for ma - tion within a subglacial shear zone - a hy brid mo saic-per va sive de for ma tion model for Mid dle Pleis to cene glaciotectonised sed - i ments from East ern Eng land. Qua ter nary Sci ence Re views, 27:
1350–1362.
Lewandowski, J., 1988. Mid dle Pleis to cene in the up stream Odra val ley re gion: Moravian Gate – Racibórz Ba sin (syn thet i cal ap - proach) (in Pol ish with Eng lish sum mary). Przegl¹d Geologiczny, 36: 465–474.
Lindner, L., 2001. Prob lems of the age and ex tent of the Scan di na - vian glaciations at the mar gin of the Pol ish Carpathians (south - ern Po land) (in Pol ish with Eng lish sum mary). Przegl¹d Geologiczny, 49: 819–821.
£anczont, M., 1997. Pro file of the Qua ter nary de pos its at Siedliska (Fore land of the Przemyœl Carpathians, south-east ern Po land) and its paleogeographic as pect. Studia Geomorphologica Carpatho-Balcanica, 30: 5–29.
Macoun, J., Králík, F., 1995. Gla cial his tory of the Czech Re pub lic.
In: Gla cial De pos its in North-East Eu rope (eds. J. Ehlers, S.
Kozarski and P.L. Gibbard), 389–405. A.A. Balkema, Rot ter - dam-Brookfield.
Makowski, A., 1936. Sprawozdanie z badañ geologicznych w 1935 r. w zachodniej czêœci arkusza Gorzyce mapy Polskiego Zag³êbia Wêglowego w skali 1:25 000. Posiedzenia Naukowe Pañstwowego Instytutu Geologicznego, 45: 1–4.
Mark, D.M., 1973. Anal y sis of ax ial ori en ta tion data, in clud ing till fab rics. GSA Bul le tin, 84: 1369–1374.
Mark, D.M., 1974. On the in ter pre ta tion of till fab rics. Ge ol ogy, 2:
101–104.
Marks, L., 2005. Pleis to cene gla cial lim its in the ter ri tory of Po land.
Przegl¹d Geologiczny, 53: 988–993.
Mojski, J.E., 2005. Ziemie polskie w czwartorzêdzie (in Pol ish).
Zarys morfogenezy. Pañstwowy Instytut Geologiczny, Warszawa.
478 Tomasz Salamon
Munro-Stasiuk, M.J., 2000. Rhyth mic till sed i men ta tion: ev i dence for re peated hy drau lic lift ing of a stag nant ice mass. Jour nal of Sed i men tary Re search, 70: 94–106.
Narloch, W., Piotrowski, J.A., Wysota, W., Larsen, N.K., Men - zies, J., 2012. The sig na ture of strain mag ni tude in tills as so ci - ated with the Vistula Ice Stream of the Scan di na vian Ice Sheet, cen tral Po land. Qua ter nary Sci ence Re views, 57: 105–120.
Narloch, W., Wysota, W., Piotrowski, J.A., 2013.
Sedimentological re cord of subglacial con di tions and ice sheet dy nam ics of the Vistula Ice Stream (north-cen tral Po land) dur ing the Last Gla ci ation. Sed i men tary Ge ol ogy, 293: 30–44.
Ng, F.S.L., 2000. Ca nals un der sed i ment-based ice sheets. An nals of Gla ci ol ogy, 30: 146–152.
Nitychoruk, J., 1991. Gla cial de pos its of the Tarnów Pla teau (S Po - land) and the prob lem of Dunajec out flow dur ing the San II (=
Elsterian II) gla ci ation. Annales Universitatis Mariae Cu - rie-Sk³odowska Sectio B Geographia, Geologia, Mineralogia et Petrographia, 46: 187–200.
Nývlt, D., Engel, Z., Tyráèek, J., 2011. Pleis to cene glaciations of Czechia. De vel op ment in Qua ter nary Sci ence, 15: 37–46.
Piotrowski, J.A., Tulaczyk, S., 1999. Subglacial con di tions un der the last ice sheet in north west Ger many: ice bed sep a ra tion and en hanced basal slid ing. Qua ter nary Sci ence Re views, 18:
737–751.
Piotrowski, J.A., Mickelson, D.M., Tulaczyk, S., Krzyszkowski, D., Junge, F., 2001. Were subglacial de form ing beds be neath past ice sheets re ally wide spread? Qua ter nary In ter na tional, 86:
139–150.
Piotrowski, J.A., Larsen, N.K., Men zies, J., Wysota, W., 2006.
For ma tion of subglacial till un der tran sient bed con di tions: de po - si tion, de for ma tion, and basal de coup ling un der a Weichselian ice sheet lobe, cen tral Po land. Sedimentology, 53: 83–106.
Rob erts, D.H., Hart, J.K., 2005. The de form ing bed char ac ter is tics of a strat i fied till as sem blage in north East Anglia, UK: in ves ti - gat ing con trols on sed i ment rhe ol ogy and strain sig na tures.
Qua ter nary Sci ence Re views, 24: 123–140.
Rùžièka, M., 2004. The Pleis to cene gla ci ation of Czechia. De vel op - ment in Qua ter nary Sci ence, 2: 27–34.
Salamon, T., 2001. Sed i men tary suc ces sion and morphologic con - trol of glaciomarginal paleogeography in the foot hill area, an ex - am ple from the Silesian Beskid (Po land). Studia Geomorphologica Carpatho-Balcanica, 35: 39–61.
Salamon, T., 2014a. Basal till and subglacial con di tions at the base of the Up per Odra ice lobe (south ern Po land) dur ing the Odranian (Saalian) Gla ci ation. Geo log i cal Quar terly, 58 (4):
779–794.
Salamon, T., 2014b. Evo lu tion of the glaciomarginal zone in the Silesian Beskid foot hill (Bielsko di vi sion of the Silesian Foot hills) (in Pol ish with Eng lish sum mary). Przegl¹d Geologiczny, 62:
103–110.
Salamon, T., 2015a. De vel op ment of the to pog ra phy-con trolled Up - per Odra ice lobe (Scan di na vian Ice Sheet) in the fore-moun tain
area of south ern Po land dur ing the Saalian gla ci ation. Qua ter - nary Sci ence Re views, 123: 1–15.
Salamon, T., 2015b. Sed i men tary re cord of a Scan di na vian Ice Sheet drain age sys tem and till de po si tion over subglacial ob sta - cles pro mot ing basal slid ing (an ex am ple from south ern Po - land). Sed i men tary Ge ol ogy, 330: 108–121.
Salamon, T., 2016a. Subglacial con di tions and Scan di na vian Ice Sheet dy nam ics at the coarse-grained sub stra tum of the fore-moun tain area of south ern Po land. Qua ter nary Sci ence Re views, 151: 72–87.
Salamon, T., 2016b. Re ply to com ment by D. Nývlt on “De vel op - ment of the to pog ra phy-con trolled Up per Odra ice lobe (Scan di - na vian Ice Sheet) in the fore-moun tain area of south ern Po land dur ing the Saalian gla ci ation”. Qua ter nary Sci ence Re views, 133: 185–187.
Salamon, T., Wójcik, A., 2010. De vel op ment of a fore-moun tain al - lu vial fan of the Olza River (south ern Po land) dur ing the Pleis to - cene. Annales Societatis Geologorum Poloniae, 80: 39–52.
Tay lor, G.I., 1923. The mo tion of el lip soi dal par ti cles in a vis cous fluid. Pro ceed ings of the Royal So ci ety of Lon don, Se ries A, 103: 58–61.
Tulaczyk, S., 1999. Ice slid ing over weak, fine-grained tills: de pend - ence of ice-till in ter ac tions on till granulometry. GSA Spe cial Pa - per, 337: 159–177.
Tyráèek, J., 2011. Con ti nen tal gla ci ation of the Moravian Gate (Czech Re pub lic). Jour nal of Geo log i cal Sci ences, Anthropozoic, 27: 39–49.
Van der Wateren, F.M., 2002. Pro cesses of glaciotectonism. In:
Mod ern and Past En vi ron ments (ed. J. Men zies). Butterworth and Heinemann, Ox ford: 417–443.
Van der Wateren, F.M., Kluving, S.J., Bartek, L.R., 2000. Ki ne - matic in di ca tor of subglacial shear ing. Geo log i cal So ci ety Spe - cial Pub li ca tions, 176: 259–278.
Walder, J.S., Fowler, A., 1994. Channelised subglacial drain age over a de form able bed. Jour nal of Gla ci ol ogy, 40: 3–15.
Wojtanowicz, J., 1985. The TL dated pro file of the Qua ter nary de - pos its at Giedlarowa (Sandomierz Ba sin) and its paleogeographic im por tance (in Pol ish with Eng lish sum mary).
Studia Geomorphologica Carpatho-Balcanica, 19: 37–44.
Wójcik, A., 2003. Qua ter nary of the west ern part of the Jas³o-Sanok De pres sion (Pol ish Outer Carpathians) (in Pol ish with Eng lish sum mary). Prace Pañstwowego Instytutu Geologicznego, 178:
1–148.
Wójcik, A., Nawrocki, J., Nita, M., 2004. Pleis to cene in the Koñczyce pro file (Oœwiêcim Ba sin) - sed i ment gen e sis and age anal y sis at the back ground of strati graphic schemes of the Qua - ter nary (in Pol ish with Eng lish sum mary). Biuletyn Pañstwowego Instytutu Geologicznego, 409: 5–50.
Wysota, W., 2002. Stra tig ra phy and sed i men tary en vi ron ments of the Weichselian Gla ci ation in the south ern part of the Lower Vistula River (in Pol ish with Eng lish sum mary). Wydawnictwo Uniwersytetu Miko³aja Kopernika, Toruñ.
