FORMAL STRATIGRAPHY IN THE OF FINLAND

PHILIP L. GIBBARD

GIBBARD, PHILIP L. 1992. Formal stratigraphy in the Pleistocene of Finland. Bull. Geol. Soc. Finland 64, Part 2, 125—134. The importance of the application of formal lithostratigraphical, biostratigraphical and chronostratigraphical procedures to the subdivision of Finnish Pleistocene sedi- ment sequences and the need to define both regional and local stratotypes are dis- cussed. The particular problems relating to the definition of stratotypes in areas that have been subjected to glaciation are addressed. Problems regarding nomenclature and correlation of Pleistocene sequences, largely arising from non-definition or unsatis- factory definition of stratotypes are illustrated with reference to the Saalian of northern Continental Europe and some Finnish examples. Recommendations are pre- sented to encourage and improve the selection and definition of stratotypes for Qua- ternary sediment sequences in Finland.

Key words: glacial geology, stratigraphy, stratigraphic units, stratotypes, type localities, Pleistocene, Finland.

Philip Gibbard, Subdepartment of Research, Botany School, University of Cambridge, Downing Street, Cambridge CB2 3EA, England.

Introduction ed for inter-regional correlation and international communication. Research during the last twenty in Finland With these views in mind, this paper presents has increasingly demonstrated that the Quaternary some thoughts on the need to apply international- sequence apparently represents considerably more ly accepted, standardised stratigraphical principles of Pleistocene time than previously envisioned. to Finnish Quaternary sediment sequences. These This development has arisen from the identifica- principles represent a concensus agreed and estab- tion of multiple till sequences, and in- lished by the International Commission on Stratig- terstadial organic sediments as well as non-glacial raphy. The International Stratigraphic Guide (ISG) cold-climate sediment sequences and soils. This (Hedberg 1976) provides the basic reference for increasing complexity is accompanied inevitably these principles in the form of a code. This guide by an expansion in terms required to refer to indi- clearly states that lithostratigraphical units (forma- vidual units. However, the terms that are becom- tions and members), chronostratigraphical units ing established in the new literature, together with (stages and substages) and likewise biostratigraph- the concepts adopted to subdivide the sequences, ical units (biozones) need to be clearly differenti- are in some cases not ideal. The nomenclature cur- ated and moreover, should be firmly based on and rently in use has evolved from a combination of specifically defined from actual stratotypes at type long-established practice for those units and fea- localities. tures that have been known for many years, sup- The objectives of these internationally-accept- plemented by field terms and section logging prac- ed stratigraphical procedures can be summarised tices for the new units recognised. These terms, as follows: although they may be implicitly understood and 1) To subdivide and classify rock sequences and clear to individual practitioners familiar with the geological time; concepts represented, are not necessarily well-suit- 2) To provide an unambiguous basis for such sub- 126 Philip L. Gibbard

divisions; cial period or stage are still persistently in use, in 3) To facilitate correlation within and between dif- spite of repeated pleas by several authorities that ferent geographical areas; they should be abandoned. 4) To facilitate communication. These views have arisen because the terms gla- In this paper the advantage in increased rigour cial, interglacial, stadial and interstadial were orig- and scientific precision in adopting the ISG rec- inally and often still are used in place of the terms ommendations is presented. stage and substage respectively. There is & funda- First, the classical subdivisions of the Quater- mental difference between these terms. Glacial and nary in northern Europe are considered, to exam- interglacial etc. are geological-climate units, as ine how these subdivisions, particularly those used defined by the American Commission on Strati- for chronostratigraphical purposes, were erected graphic Nomenclature (1961) and are based on cli- and to what extent they now conform to the rec- matic events inferred from geological sequences. ommendations of the International Stratigraphic As such they are used almost exclusively in the Guide. Pleistocene. In practice, they can only be truly val- Secondly, the particular problem of the subdi- id in glacierised areas, for which they were origi- vision of the Saalian Stage in northwest Europe to nally conceived. An additionally significant prob- illustrate the need for precision in the definition of lem is that the boundaries of such climatically-de- unit and boundary stratotypes is presented. termined units will almost certainly not be syn- Thirdly, the problems of the Finnish Quaternary chronous across large areas, especially in continen- sequence are discussed to emphasise the need to tal sequences. This is because climatic changes the define that sequence in accordance with the must very often be strongly-influenced by local ISG recommendations. climate and geographical factors. In contrast, stage Finally, some recommendations for the selec- and substage are chronostratigraphical units, the tion and definition of national stage and substage function of which is clearly defined by the Inter- stratotypes, their placement as far as possible in national Stratigraphic Guide. A stage is a chro- non-glacial sediment sequences and the need to nostratigraphical unit of relatively minor rank, con- compile a reference document of such sites are sidered to be the most suitable for interregional presented correlation. The boundaries of stages are ideally isochronous surfaces between fixed points both within and beyond the type area. (cf. Hedberg Stage-status stratigraphical units in northern 1976). In deep-ocean cores stages have been de- Europe fined on the basis of oxygen isotope curves, which are also linked to climatic changes. Though it The Pleistocene Period is characterised by large seems likely that the boundaries of these isotope amplitude climatic fluctuations, one of the most stages are not precisely synchronous throughout striking consequences of which are intervals of ex- the world's oceans, they nevertheless may be more pansion of ice sheets across the whole of Finland nearly so than the onset of the 'glacial' and 'inter- into lowland present-day temperate regions. Be- glacial' units recognised in continental sequences. cause these massive expansions of glacial ice pro- The use of geological-climatic subdivisions duced such a major impact on the land surface, 'glacial', 'glacial stage', 'glaciation' and equiva- they inspired geologists in the 19th and early 20th lents have been repeatedly discouraged by West centuries to attempt a subdivision of 'Ice Age' time (1963, 1984, 1989) in favour of the term 'cold using these events. This led pioneers of Quater- stage' (Gibbard & Turner 1990). Liittig (1965) nary stratigraphy, such as Geikie (1874), to be al- went further by suggesting abandonment of the ready recognising 'glacial' and 'interglacial' events these and the term interglacial etc., and replacing in the late 1800's. The terms glacial and intergla- them with the words cryomere and thermomere re- Formal stratigraphy in the Pleistocene of Finland 127 spectively. Ltittig's proposals have not, however, temporary glaciation. found general acceptance. Nevertheless, they are In keeping with a tradition begun by Penck & still applied occasionally. For example, they were Brückner (1901—1909) in the Alpenvorland, cold recently advocated by Zubakov & Borzenkova or 'glacial' stages in northern and central Europe (1990) who adopted these terms as the basis of a have nevertheless been named from the occurrence highly theoretical complex hierarchy of climate- of glacial, glaciofluvial and glaciolacustrine depos- based stratigraphical units that are of little practi- its. The earliest, and still widely-used, terms for cal utility. cold stage stratigraphical units were those proposed A further complication to this discussion arises by Penck & Brückner; Würm, Riss, Mindel, Günz. from recent proposals by the Norwegian Commit- Ever since these names have been applied in many tee on Stratigraphy (Nystuen 1989) for a new se- parts of the world far from their original geograph- ries of stratigraphical subdivisions termed dia- ical area, not as proper chronostratigraphical terms chronous units (time-transgressive event stratigra- but simply as concepts with a broadly understood phy). They are divided into episodes and events. but undefined meaning (i.e. last, penultimate, an- These new units are intended to represent the time tepenultimate etc.). Even in their span of a geological event characterised by a spe- type region, the terms were attached in a general- cific geological process. For use in the Pleistocene ised way to suites of glacial and glacifluvial de- such units may represent a glaciation, a marine posits in several valleys. Only in 1983 was the trangression or a period of aeolian sand deposition, chronostratigraphical term Würmian formally de- for example. They are clearly then very similar to fined by Chaline & Jerz and assigned to a meas- the geological-climate terms glacial and intergla- ured and well-studied stratotype at Samerberg in cial. However, as Bouchard et al. (1990) have em- southern Germany. The remaining terms have yet phasised, if diachronous units are applied, it is vi- to be formal defined, and, perhaps more important- tally important that an unambiguous distinction is ly, there no longer seems to be a concensus of opin- made between any particular event (e.g. glacial ion that the ages and stratigraphical position of the advance or retreat) or episode (a glaciation) and glacial and glacifluvial deposits associated with the the time (chronostratigraphical stage or geochrono- terms Riss and Mindel by Penck & Brückner in logical age) during which these events or episodes the Alpenvorland type region necessarily corre- occurred. As an illustration, this means that the spond chronologically with the concepts assumed main (Weichselian Glacial and used by Quaternary geologists and particular- Episode), that begun in Finland at c.75ka and end- ly archaeologists in other parts of the world. ed by 8.9ka , began in the Late Weichselian Sub- This practice of naming major episodes of gla- stage and ended in the early Flandrian Stage ciation, usually after the names of local rivers, was ( ). extended by Keilhack in the 1920's to the area of It has been argued that glacial deposits must be the North European Plain affected by the Scandi- known from any particular period in order to de- navian ice sheet on maps produced by the Preus- fine a particular 'glacial' stage or event. However, sische Geologische Landesanstalt (cf. Woldstedt it is not always possible to unequivocally demon- 1929 p. 179). These terms, Weichsel, Saale and strate glaciation, particularly in the Early and ear- Elster, were only informally published and were ly Middle Pleistocene. Moreover, the 'glacial' time not clearly associated with any defined type local- intervals themselves are virtually never represent- ities. Regrettably, this is still the case today. ed in the geological record exclusively by glacial As Gibbard & Turner (1990) have emphasised, deposits. All recognised cold stages include sedi- such terms as Saale and Elster have been applied ments laid down under periglacial, non-glacial and indiscriminately for the purposes of both lithos- indeed even short-lived (interstadial) equitable cli- tratigraphy and chronostratigraphy; for example mates, particularly outside the actual area of con- the expressions Saale I, II, III Grundmoräne, Saale- 128 Philip L. Gibbard

Hauptterrasse, Saalevereisung, as well as Saalee- Table 1. Middle to chronostratigraphical stag- iszeit. This was recognised as a problem by Dutch es of western central Europe and Finland (sources: Donner, 1976; Grube, 1981). Quaternary stratigraphers early in the 1950's (Van der Vlerk 1953, Van der Vlerk & Florschütz 1953, English / Dutch German Finnish Van der Vlerk 1957, Van der Heide & Zagwijn Weichselian* Weichsel {Kaltzeit Veiksel-jääkausi 1967). As a result they subdivided the Netherlands {Glazial sequence into lithostratigraphical formations and began to use the terms Saalian (or Drenthian) and * Eem {Warmzeit Eem-interglasiaali Elsterian to denote the cold stages as chronostrati- {Interglazial Saalian* Saale {Kaltzeit Saale-jääkausi graphical units. Likewise, terms such as Eemian {Glazial and Holsteinian (Needian) were used to denote Holsteinian* Holstein* {Warmzeit Holstein- temperate (interglacial) stages. {Interglazial interglasiaali In spite of this, in non-English speaking coun- Elsterian* Elster {Kaltzeit ?Elster-jääkausi {Glazial tries outside the Netherlands, the subdivision of the Cromerian*/ 'Cromer Komplex' ? Pleistocene remains largely based upon the geo- 'Cromerian Complex'* logical-climate units, contrary to the ISG recom- *Stage with formally-defined stratotype(s) mendations (Table 1). In Germany, for example, the terms Glazial or Eiszeit and Interglazial or Warmzeit still predominate (e.g. Grube 1986). Likewise, the corresponding Finnish terms are mon with the rest of Phanerozoic time. Neverthe- jääkausi and interglaasiaali (Donner 1976). In less, Quaternary chronostratigraphic subdivision common with the English terms 'glacial' or 'inter- has become almost inextricably linked with geo- glacial', these quasi-stage type subdivisions carry logical-climate units. To justify this it has been an implied climatic significance, whilst also in- claimed that major climatic changes during the cluding the chronostratigraphical implication of the Quaternary, such as those recorded in the deep term stage. In the light of modern knowledge that ocean oxygen isotope curves, are globally synchro- all the later Pleistocene cold periods include shorter nous or virtually so compared to the dating preci- warm climate events it is apparent that the contin- sion achievable for chronostratigraphical bounda- ued use of climate-based terms is unfortunate. The ries defined for earlier geological periods. fact that these terms are so strongly embedded in However, as Zagwijn (1985) and De Jong the literature means that it is unlikely that they will (1988) have emphasised, it is no longer practica- be discarded in favour of the more precise chron- ble in Pleistocene stratigraphy to assign a new ostratigraphical terms. However, the long term aim stage to each climatic event. This is because the should be to replace geological-climate terms. This scale of individual events varies considerably, and aim is more likely to be achieved in Finland, where because the number of events now known is so this matter is currently being actively debated, than high. These authors have therefore stressed the elsewhere in the non-English speaking world need for complex stages, i.e. intervals of time con- where geologists seem unconcerned about or una- taining several climatic oscillations at the current- ware of these problems. ly detectable scale. The cold stages of the later Pleistocene are typical examples of this principle, particularly the early part of the Weichselian Chronostratigraphy and climatic change (Woillard 1975, 1978; Mangerud 1991) and the Saalian (Zagwjn 1986, Gibbard & Turner, 1990). In the geological community a consensus is now The classification and subdivision of the Saal- accepted that Pleistocene sequences should be di- ian in northern Europe illustrates problems that vided chronostratigraphically into stages in corn- derive directly from the lack of a formal and agreed Formal stratigraphy in the Pleistocene of Finland 129 stratotype for this stage (Fig. 1). Although origi- Netherlands Germany nally recognised in eastern Germany, much of the detail of events during this stage has been estab- Eemian lished in the Netherlands and neighbouring north Germany (Grube 1981, Ter Wee 1983, Ehlers et Warthe Stadial al. 1984, Urban et al. 1989). As already men- ?'main Interstadial' tioned, it is now fully accepted that cold stages (?Rügen) include not only prolonged intervals of cold cli- Drenthe(Drenthian) Drenthe mate accompanied on occasions by glaciation, but also shorter intervals of relatively warm climate, often referred to as interstadials. Such stadial and Bantega Intst interstadial intervals have frequently been desig- Hoogeveen Intst Wacken/ nated, usually informally, as substages, but the Dömnitzian problems of definition and synchroneity of sub- Stadial I Fuhne Stadial stage boundaries are essentially the same as for stages. In the Netherlands two such intervals or Holsteinian episodes, the Hoogeveen Interstadial and the Ban- tega Interstadial are recognised in the early part of Fig. 1 Saalian Stage chronostratigraphy in the Netherlands and Germany, based on sources quoted in the text. the Saalian (Zagwijn 1973; cf. also the 'Belvédére Interglacial' of Vandenberghe, personal commu- nication). The onset of the Saalian is taken as the fined. Fortunately, the 1992 meeting of the INQUA point at which cold-climate conditions, indicated Subcommission of European Quaternary Stratig- by renewed dominance of open ground plant taxa raphy agreed to define the base of the Saalian at in pollen diagrams, mark the end of the Holsteini- the end of the Holtenian, as in the Netherlands. an temperate Stage. However, in northern Germa- ny, particularly in the former German Democratic Republic, in what is classically thought of as the Stratotypes undefined type area of the 'Saale' of Keilhack, one of these intervals of warming, probably equivalent When Pleistocene sequences are divided chron- to the Hoogeveen Interstadial, is regarded as of full ostratigraphically into stages, it is vital that type interglacial status and has been defined as the localities be presented from which these units can Dömnitz or Dömnitzian 'Stage' (Cepek 1967, be defined. The ISG requires that all stratigraphi- 1986). The Dömnitzian is separated from the Hol- cal units be formally-defined from type localities steinian by a short interval termed the Fuhne Sta- or stratotypes. Stratoypes may be of three basic dial . A similar stratigraphy has been recognised kinds: at Wacken in Schleswig-Holstein (Menke 1968) (a) Unit stratotypes, i.e. at localities where indi- and at Helmstedt (Urban et al. 1989) in western vidual units are defined; Germany. Cepek (1986) includes the 'Dömnitz (b) Boundary stratotypes, i.e. at localities where a interglacial', the 'Fuhne glacial period (?)', togeth- particular boundary is defined; er with the 'Holstein interglacial (sensu stricto )' (c) Composite stratotypes, i.e. at localities where within a 'Holstein Complex'. The base of the Saal- both boundary and unit stratotypes are recog- ian Stage, as recognised in eastern Germany (ex- nised. GDR), is therefore younger than in the Nether- By far the most consistent and detailed defini- lands. This state of affairs is clearly not satisfacto- tion of stratotypes in Europe has been in the Neth- ry and would not have arisen if a stratotype for the erlands. However, both the Dutch, and to a lesser Saalian Stage had been properly and formally de- extent the British, have defined cold stages and 130 Philip L. Gibbard substages by reference to their position intermedi- significant boundaries in rock sequences has been ate between temperate (interglacial) stages or sub- much discussed by stratigraphers in connection stages. The actual boundaries between these stag- with marking the actual boundary line either no- es have predominantly been defined using paly- tionally or actually with a 'golden spike' (see dis- nology at the point where 'interglacial' type vege- cussion in Holland 1986). Ideally such boundaries tation changes to that of a cold stage type. The ends should be located in continuous sediment sequenc- of cold stages are automatically defined by the base es, but in practice it may be difficult. There is how- defined for the immediately following temperate ever, ample potential for the definition of such stage. boundary points. At many sites in western Europe Although they are abundant in Finland, glacial stage boundaries occur in lacustrine and fluvial deposits are not well suited as national or interna- sediments, e.g. the Eemian/Weichselian boundary tional cold stage or substage stratotypes for a vari- is recorded at Samerberg (Chaline & Jerz 1983) ety of reasons. Most importantly, it is very diffi- and at La Grande Pile (Woillard 1978). Similarly cult to determine the amount of time any particu- in Britain, the Hoxnian / Wolstonian boundary is lar glacial deposit might represent. Secondly, it can identified in the pollen analyses of Turner (1970) be difficult to correlate glacial deposits over large and the Wolstonian / Ipswichian boundary is rec- areas (particularly extra-regionally), since they sel- ognised by West (1957) at the Ipswichian Stage dom contain material that can be used biostrati- stratotype at Bobbitshole. Comparable sites that graphically or for geochronology. Thirdly, glacial could be used for this purpose also exist on the deposits frequently rest on an eroded or glacially- Continent for other stage boundaries. It should be modified surface, and, therefore, inevitably there emphasised that there is no formal procedure for will have been a loss of information, at best a short defining the top of a stage. The time-period of a time gap and in the worst cases an hiatus of thou- stage is deemed to continue until the onset of the sands or even millions of years. For these reasons next-youngest stage defined. it is apparent that cold stages and indeed substag- es should ideally be defined in non-glacial se- quences, where additional detail of stratigraphical Stratigraphical problems in the Finnish Qua- significance is potentially available. Clearly this ternary is not possible in Finland. However, it is possible to define boundaries and stratotypes in the most In this section five examples of problems aris- complete sedimentary sequences, where the suc- ing from a lack of application of formal strati- cession contains unambiguous sedimentary rela- graphical approaches to Pleistocene stratigraphy in tionships (e.g. Bouchard et al. 1990). Finland are presented. The order of these exam- Throughout much of the later Middle and Up- ples is not significant. Each example is selected to per Pleistocene there are few formally defined illustrate a particular area of difficulty or ambigu- boundary stratotypes in north and western Europe ity. outside the southern North Sea basin, for cold stag- es and substages. This is a serious lack of preci- Salpausselkä Ridges sion in formal chronostratigraphy and is contrary to the strong recommendations of the ISG and the The Salpausselkä End Moraines are possibly the formal International Commission on Stratigraphy best known moraine ridges in the world and are (ICS) guidelines (Cowie et al., 1986). central to any understanding of ice recession dur- A boundary stratotype is a locality at which the ing the Late Pleistocene in Finland (Gliickert base of a stage (and thus by inference the top of 1986). In general they are recognised by their ex- the previous stage) is defined in a sediment ternal form, i.e. as morphological units, and have sequence.The identification of stage and other time been assigned ages on the basis of their relation- Formal stratigraphy in the Pleistocene of Finland 131

ship to the varve chronology and to sediments fill- of the sediment beneath the till. In many cases this ing kettle hole lakes in their uppermost surface may be correct, but the assumption clearly ob- (Donner 1971). Excavations in the Salpausselkäs scures objective description of the sedimentary expose a complex lithostratigraphy within the unit. It once again mixes morphology and litholo- ridge-like landforms at most localities. The Sal- gy- pausselkä Moraines are therefore both morphos- tratigraphical and lithostratigraphical units, al- and interstadials though the two types of evidence provide differ- ent stratigraphical information. For this reason it The recognition of relatively warm climate is important to differentiate between the landform events in the Finnish Pleistocene is a relatively new and its internal depositional sequence. phenomenon. The basis for the identification of these events is the fossil assemblages present in Late Pleistocene Baltic Sea sediments the sediments (Korpela 1969, Aalto et al. 1983, Donner 1983, Forsström 1982, Hirvas & Kujan- During the Late Weichselian and Flandrian suu 1979, Gibbard et ai. 1989, Grönlund 1991). Stages much of southern and western Finland was Considerable discussion has taken place in recent submerged beneath the Baltic Sea or its lacustrine years as to the climatic significance of individual equivalents (e.g. Donner 1965). The sedimentary sequences. However, generally attempts have been sequences that accumulated beneath the waters of made to equate these sequences with the relatively the Baltic have provided the evidence to recognise well-established events in central western Europe four major phases in the Late- and Post-glacial (e.g. Grönlund 1991). Arguments based upon bio- Baltic: respectively, the Baltic Ice Lake, the Yol- logical assemblages are clearly restricted by the dia Sea, the Ancylus Lake and the Littorina Sea. biotal changes at ecotonal boundaries. This is es- These stages are based on the interpretation of pecially important in the interglacial sequences in microfossil assemblages from the sediments (al- which assemblages in temperate western Europe most exclusively pollen and diatom assemblages). differ markedly from the predominantly as- Radiocarbon dating of events associated with the semblages in Finland. changes in the assemblages has allowed the Nevertheless, interglacial sequences have been changes to be used chronologically within the ba- correlated mostly with the last or Eemian Stage sin (e.g Eronen 1974). interglacial event using palaeontological assem- The terminology used to subdivide the Baltic blages. These correlations have been supported by sediment sequences has evolved from the stages absolute dates and other evidence. The need to recognised e.g. the Littorina clays etc. The result define one or more local stratotype sequences is is not a correct lithotratigraphical subdivision , but however, vital in order to provide an unambigu- a combination of litho-, bio- and chronostratigra- ous reference sequence for this event and its pal- phy that could potentially lead to confusion. aeontological succession in the region. This is par- ticularly important for older deposits where doubt over the age and equivalence must be considera- Till-covered eskers ble in view of the complexity of events known It is now apparent that in many parts of the from other parts of the world. country glaciofluvial sediment bodies are overlain by glacial diamicton. Although a landform cannot Multiple till sequences always be identified, these sequences are very fre- quently referred to as 'till-covered eskers" (e.g. The extensive investigations by the Geological Niemelä & Tynni 1979). Clearly, this descriptive Survey throughout the country, but particularly in terminology implies both genesis and morphology Lapland by Hirvas and co-workers (Hirvas & 132 Philip L. Gibbard

Nenonen 1987, Nenonen 1986, Hirvas 1991) and lems in related glaciated regions in Europe and similar investigations by Saarnisto & Peltoniemi beyond (cf. Gibbard & Turner 1990). (1984) in eastern Finland and Aario & Forsström 1. It is vital that stratotypes are defined for (1979) in northern Finland have resulted in the rec- all stages and substages. Priority should be giv- ognition of multiple till sequences in many areas. en to locating stage stratotypes equivalent to the The sequences have been interpreted as indicated Weichselian and Saalian of central Europe. Stra- repeated glacial events. In some areas it has been totypes for interglacial stages such as the Tep- possible to relate these events to non-glacial fos- sankumpu event (Hirvas 1991) are urgently siliferous sediments and thereby develop a relative needed. Haapavesi (Forsström et al. 1988) or chronology. Evijärvi (Eriksson et ai. 1980) might be appro- These important discoveries have revolutionised priate as parastratotypes for southern Finland concepts of stratigraphy in Finland and precipitat- last interglacial (Eemian Stage equivalent) se- ed the need for consideration of a more formal quences. The sediment sequences in Pohjanmaa stratigraphical framework. This framework is re- offer possible pre-Eemian localities for strato- quired because, at the time of writing, the till units types for cold stage deposits. lack stratotypes and have not been assigned unit 2. Wherever possible, boundary stratotypes names in keeping with the ISG recommendations. should be placed in continuous sediment se- quences, or at minor erosional hiatuses. In gen- As these short summaries illustrate there seems eral, lacustrine sequences appear to provide the to be much confusion over the nomenclature adopt- most unambiguous continuous sedimentary ed and the separation of stratigraphical evidence records on the continents and have the added into its component parts for the sake of greater advantage of generally yielding good biostrati- clarity and precision in stratigraphical correlation. graphical evidence. The adoption of formal stratigraphical classifica- 3. Sound lithostratigraphy is required for tion provides a solidity, strength and resilience to many of the sedimentary successions, particu- the stratigraphical sequence in any area by allow- larly where they are buried by younger depos- ing modifications to be made to the structure or its, as stressed by Bouchard et al. 1990. Of par- order, yet obviates the need to change the entire ticular importance however, is the need to fully structure if changes occur. Furthermore, the sys- understand the sedimentology of a sediment unit tematic application of internationally standardised before its full stratigraphical significance can be nomenclatoral systems simplifies and clarifies assessed. communication both within the country and be- yond. Clearly any terms adopted will require trans- 4. It is preferable that any national or re- lation and some modification from the English for gional stratotype localities that are selected be, use in the Finnish or Swedish languages . What is as far as possible, in non-glacial sediment se- important however, is the recognition of the need quences since they include potentially a wider to adopt a stricter, more systematic approach to range of information that can be used for inter- stratigraphical concepts regional and possibly land/ocean correlations. 5. A reference document or catalogue of all stratotypes should be compiled, in which sites could be described, subdivided and defined in a Recommendations uniform way.

Acknowledgements. I am grateful to Dr C. Turner for permission The following recommendations arise from the to reproduce material from our joint paper (Gibbard & Turner, examples of confusion quoted above and from gen- 1990). I would also like to thank Dr V.-P. Salonen for his eral experience in Quaternary stratigraphical prob- encouragement and discussions. Formal stratigraphy in the Pleistocene of Finland 133

References Forsström, L., Aalto. M., Eronen, M. & Grönlund, T., 1988. Stratigraphic evidence for Eemian crustal movements and Aalto, M.. Donner, J J., Niemelä, J. & Tynni, R.. 1983. An eroded relative sea level changes in eastern Fennoscandia. interglacial deposit at Vimpeli, south Bothnia, Finland. Geo- Palaeogeography, Palaeoclimatology Palaeoecology 68,317— logical Survey of Finland Bulletin 324, 1—41. 335. Aarit), R.& Forsström, L., 1979. Glacial stratigraphy of Koillismaa Geikie, J., 1874. The great ice age and its relation to the antiquity and North Kainuu, Finland. Fennia 157, 1—49. of Man. W. Isbister & Co.: London 575p. American Commission on Stratigraphic Nomenclature, 1961. Gibbard, P L. & Turner, C., 1990. Cold stage type sections; some Code of stratigraphic nomenclature. Bulletin of the American thoughts on a difficult problem. Quatemaire 1, 33—40. Association of Petroleum Geologists 45, 645—660. Gibbard, P., Forman, S., Salomaa, R.. Alhonen, P., Jungner, H . Bouchard, M.A., Gibbard, P.. Salonen, V.-P., 1990. Litho- Peglar, S., Suksi, J. & Vuorinen, A., 1989. Late Pleistocene stratotypes for Weichselian and pre-Weichselian sediments in stratigraphy at Harrinkangas, Kauhajoki, western Finland. southern and western Finland. Bulletin of the Geological Annales Academiae Scientiarum Fennicae A III Society of Finland 62, 79—95. Geologica-Geographica 150, 1—36. Cepek, A.G., 1967. Stand und Probleme der Quartärstratigraphie Glückert, G., 1986. The First Salpausselkä at Lohja, southern im Nordteil der DDR. Berichte der Deutschen Gesellschaft für Finland. Bulletin of the Geological Society of Finland 58, geologische Wissenschaft A, Geologie. Paläontologie 12, 45—55. 375—404. Grube, F., 1981. The subdivision of the Saalian in the Hamburg Cepek, A.G., 1986. Quaternary stratigraphy of the German Dem- region. Mededelingen Rijks Geologische Dienst 34, 15—25. ocratic Republic. Quaternary Science Reviews 5, 359—364. Grönlund, T., 1991. The diatom stratigraphy of the Eemian Baltic Choline. J. & Jerz, H., 1983. Proposition de creation d'une étage Sea on the basis of sediment discoveries in Ostrobothnia, würmienne par la sous-commission de stratigraphie du Finland. Geological Survey of Finland Report of investigation Quatemaire européen de 1'INQUA. Bulletin de l'Association 102, 26pp. franfaise pour l'Etude du Quatemaire (N.S.) 16, 149—152. Hedberg, HD., 1976. International stratigraphic guide. Wiley Cowie,J.W..Zeigler, W..Boucot,AJ.,Bassett,M. & Remane.J., Interscience. J. Wiley & Sons: New York 200p. 1986. Guidelines and statutes of the International Commis- Hirvas, H., 1991. Pleistocene stratigraphy of Finnish lapland. sion on Stratigraphy (ICS). Cour. Forschung -Inst. Bulletin of the Geological Survey of Finland 354, 123pp. Senckenberg. 83, 1—4. Hirvas, H. & Kujansuu, R.. 1979. On glacial, interstadial and De Jong, J.. 1988. Climatic variability during the past three interglacial deposits in northern Finland. IGCP, project 73/1/ million years, as indicated by vegetational evolution in north- 24. Quaternary glaciations in the northern hemisphere. Report west Europe and with emphasis on data from The Netherlands. 5, 146—164. Philosophical Transactions of the Royal Society of London B Holland, C.H., 1986. Does the golden spike still glitter? Journal of 318,603—617. the Geological Society of London 143, 3—21. Donner, J.J., 1965. The Quaternary of Finland. In: Rankama, Korpela, K.. 1969. Die Weichsel-Eiszeit und ihre Interstadial in K.(ed.) The Quaternary. 1 J.Wiley & Sons: London. 199— Peräpohjola (nördliches Nordfinnland) im Licht von 272. submöranen Sedimenten. Annales Academiae Scientiarium Donner, J.J., 1971. Towards a stratigraphical division of the Fennicae Series A III 99, 108pp. Finnish Quaternary. Commentationes Physico-Mathematicae Lüttig, G , 1965. Interglacial and interstadial periods. Journal of 41,281—305. Geology 73,579—591. Donner,J .J., 1976. Suomenkvartäärigeologia. Helsingin Yliopisto. Mangerud.J., 1991. The Last interglacial/glacial cycle in northern Moniste N:o 1 Europe. In: Shawe, L.C.K. & Cushing, E.J. (eds.) Quaternary Donner, J.J.. 1983. The identification of Eemian interglacial and landscapes. Weichselian interstadial deposits in Finland. Annales Menke,B., 1968. Beiträge zurBiostratigraphiedesMittelpleistozäns Academiae Scientiarum Fennicae Series A III Geologica- in Norddeutschland. Meyniana 18, 35—42. Geographica 136, 1—38. Nenonen, K.. 1986. Orgaanisen aineksen merkitys moreenistrati- Ehlers, J., Meyer, K.-D., & Stephan, H.-].. 1984. The pre- grafiassa. Geologi 38, 41—44. Weichselian glaciations of Northwest Europe. Quaternary Niemelä, J. & Tynni, R.. 1979. Interglacial and interstadial Science Reviews 3, 1—40. sediments in the Pohjanmaa region, Finland. Geological Sur- Eriksson, B.. Grönlund, T., & Kujansuu, R., 1980. vey of Finland Bulletin 302, 1—48. Interglasiaalikerrostuma Evijärvellä, Pohjanmaalla. Geologi Nystuen.J.P. (ed.), 1989. Rules and recommendations for naming 32, 65—71. geological units in Norway. Norsk Geologisk Tidsskrift 69 Eronen, M., 1974. The history of the Littorina Sea and associated (Supplement 1)111pp. Holocene events. Commentationes Physico-Mathematicae Penck, A. & Brückner, E., 1901—9. Die Alpen im Eiszeitalter. 44, 1—79. Leipzig 1199pp. Forsström, L., 1982. The Oulainen interglacial in Ostrobothnia, Saarnisto, M. & Peltoniemi, H., 1984. Glacial stratigraphy in western Finland. Acta Universitatis Ouluensis Series A Kainuu, eastern Finland. Fennia 162, 163—199. Scientiae Rerum Naturalium 136, 1—123. Ter Wee, W.M., 1983. The Saalian Glaciation in the Netherlands. 134 Philip L. Gibbard

In: Ehlers, J. (ed.) Glacial deposits of north-west Europe. J. & Schluchter, C. (eds.) Type sections. 3-10 Balkema: 405—412. Balkema: Rotterdam. Rotterdam. Turner, C., 1970. The Middle Pleistocene deposits at Marks Tey, Woldstedt, P.. 1929. Das Eiszeitalter. Ferdinand Ecke Verlag: Essex. Philosophical Transactions of the Royal Society of Stuttgart 406p. London B 257, 373—440. Woillard, G., 1975. Recherches palynologique sur le Pléistocéne Van der Heide, S. & Zagwijn, W.H., 1967. Stratigraphical nomen- dans Test de la Belgique et dans les Vosges lorraines. Acta clature of the Quaternary deposits of the Netherlands. geographia lovaniensia 14, 118p. Mededelingen van de Geologische Stichting Nieuwe Serie 18, Woillard, G., 1978. Grande Pile peat bog: a continuous pollen 23—29. record for the last 140,000 years. Quaternary Research 9,1— Van der Vlerk, I.M., 1953. The stratigraphy of the Pleistocene of 21. the Netherlands. Koninklijke Nederlandse Akademie van Zagwijn, W.H.. 1960. Aspects of the and Early Pleistocene Wetenschappen. Proceedings B 56, 34—44. vegetation in the Netherlands. Geologische Stichting Van der Vlerk, IM.. 1957. Pleistocene correlations between the Mededelingen (C-III-1) 5, 78p. Netherlands and adjacent areas: a symposium. Conclusion. Zagwijn, W.H.. 1973. Pollenanalytic studies of Holsteinian and Geologie en Mijnbouw (Nieuwe Serie) 19, 310—312. Saalian beds in the northern Netherlands. Mededelingen van Van der Vlerk, l.M. & Florschürz, F., 1953. The palaeontological de Rijks Geologische Dienst 24, 139—156. base of the subdivision of the Pleistocene in the Netherlands. Zagwijn, W.H., 1985. An outline of the Quaternary stratigraphy of Verhandelingen der Koninklijke Nederlandse Akademie van the Netherlands. Geologie en Mijnbouw 64, 17—24. Wetenschappen. 20, 2, 1—58. Zagwijn, W.H., Montfrans, van H.M. & Zandstra, J.G., 1971. West, R.G.. 1957. Interglacial deposits at Bobbitshole, Ipswich. Subdivision of the 'Cromerian' in the Netherlands: pollen Philosophical Transactions of the Royal Society of London B analysis, palaeomagnetism and sedimentary petrology. 241, 1—31. Geologie en Mijnbouw 50, 41—58. West, R.G.. 1963. Problems of the British Quaternary. Proceed- Zuhakov. V.A. & Borzenkova, 1.1., 1990. Global palaeoclimate of ings of the Geologists Association 74. 147—186. the Late Cenozoic. Developments in palaeontology and West, R.G., 1984. Interglacial, interstadial and oxygen isotope stratigraphy 12. Elsevier: Amsterdam. 456 pp. stages. Dissertationes Botaniceae 72, 345—357. West, R.G.. 1989. The use of type localities and type sections in the Received and accepted October 7.1992 Quaternary with especial reference to East Anglia. In: Rose,