Quaternary stratigraphy of Norden, a proposal for terminology and classification JAN MANGERUD, SVEND T. ANDERSEN, BJÖRN E. BERGLUND AND JOAKIM J. DONNER „ Mangerud, J., Andersen, S. T„ Berglund, B. E. & Donner, J. J. 1974 10 01: Quaternary IJv ^rvTl/vJ stratigraphy of Norden, a proposal for terminology and classification, Boreas, Vol. 3, pp. 109-128. Oslo. ISSN 0300-9483. i, ■'. Principles and terminology for classification of the Quaternary arc discussed, including \ . _ < r> , lithostratigraphy, biostratigraphy, morphostratigraphy, climatostratigraphy and chronostratig- —-•' " raphy. The main conclusion is a proposal for a common chronostratigraphical classification of the Quaternary in Norden (and partly continental NW Europe). The Quaternary is subdivided into the Pleistocene and the Holocenc Scries. The Pleistocene is further subdivided into several provisional stages (Weichselian, Eemian, etc.), based on the sequence of glacials/ interglacials, but with the boundaries preferably defined by stratotypes. The Late Weichselian and the Flandrian (Holocene) are subdivided into chronozones (Boiling, Older Dryas, Allerod, Younger Dryas, Preboreal, Boreal, Atlantic, Subboreal, Subatlantic) with the boundaries defined in conventional radiocarbon years. Jan Mangerud, Department of Quaternary Geology, Geomorphology and Marine Geology, Olaf Ryesvei 19, N-5014 Bergen-Universitetet, Norway; Svend T. Andersen, Geological Survey of Denmark, Thoravej 31, DK-2400 Copenhagen NV, Denmark; Björn E. Berglund, Department of Quaternary Geology, Tomaycigen 13, S-223 63 Lund, Sweden; Joakim J. Donner, Department of Geology and Paleontology. Snellmaninkatu 5, SF-00170 Helsinki 17, Finland, 24th June. 1974. Lithostratigraphy .............................................................. 110 Biostratigraphy ................................................................ 110 Mollusc zones .............................................................. 110 Foraminifcral zones ........................................................ 110 Diatom zones .............................................................. Ill Pollen zones ................................................................ Ill Morphostratigraphy ............................................................ 112 Climatostratigraphy ............................................................ 113 Chronostratigraphy ............................................................ 113 Definition of boundaries .................................................... 114 Time scales ................................................................ 114 The Cenozoic Erathcm ...................................................... 115 The Pleistocene Scries...................................................... 116 The prc-Wcichselian Stages................................................ 116 The Weichselian Stage .................................................... 117 The Late Weichselian Substage ............................................ 117 The Holocene Series ........................................................ 120 The Flandrian Stage ...................................................... 120 Subdivision of the chronozones............................................ 122 References .................................................................... 122 Appendix: Stratigraphic terms for Nordic languages compared to corresponding ones in Ennlish .................................................................. 127 A symposium on stratigraphical terminology and the classification of the Quaternary in Norden, organized by Dr. Risto Aario, was held during the 11th Winter Meeting of Nordic (Danish, Finnish, Swedish, Norwegian and Icelandic) Geologists in Oulu, Finland, in January 1974. The main aim of the symposium was to reach an agreement on a more uniform application of international terminology in Norden. During the symposium a written report was presented by Mangerud (1973), and oral introductions were given by Andersen, Berglund and 110 Jan Mangerud, Svend T. Andersen et al. BOREAS 3 (1974) Mangerud. Scientists representing the principal Quaternary Departments in Norden, and the main disciplines of Quaternary studies, participated. The opinions expressed were therefore assumed to be representative of Nordic Quaternary geologists. At the symposium it was suggested that the present authors should prepare a joint proposal, mainly based on the results from the symposium. The first draft for the chapter on the chronostratigraphy of the Pleistocene has been written by Andersen, for the chronostratigraphy of the Late Weichselian by Berglund, and for the other chapters by Mangerud, based on his earlier report. The proposals are, however, the responsibility of all the authors. The terminology and the principles for strat-igraphical classification during research in the Quaternary of Norden have developed largely independently of international rules. Mangerud (1970b, 1973) has discussed and recommended the application of the results reached by the International Subcommission on Stratigraphic Classification, a commission in the International Union of Geological Sciences. The results of the subcommission have recently been published (Hedberg ed. 1970a, 1970b, 1971a, 1971b, 1972a, 1972b), and we propose that these are applied as far as possible also to the Quaternary of Norden. Lithostratigraphy Formal lithostratigraphical units have been very little used in Norden. Such units were, however, used for lacustrine sediments by Mangerud (1970b), for glacigenic sediments by Mangerud & Skreden (1972), and for marine sediments by Morner (1971). For correlations, lithostratigraphy will probably be of major importance only in a few cases. In our opinion, however, lithostratigraphical units should be used to a much greater extent in the future than they have been in the past, for instance in connection with studies of till beds, marine transgression sediments, Baltic Sea sediments, etc. Biostratigraphy In Norden, biostratigraphical units have been defined on the basis of molluscs, foraminifera, diatoms and pollen. We do not intend to summarize all these classifications, merely put forward some proposals for terminology. Mollusc zones In eastern Norway, Br^gger (1900-1901) described stratigraphical units based on molluscs (summarized by O. Ffoltedahl 1960). He used terms like Yoldia Clay, Area Clay etc., which are combinations of biostratigraphical (Yoldia, Area) and lithostratigraphical (Clay) terms. Correct terms would be Yoldia Zone, Area Zone, etc. However, the names of the taxons have also changed, and therefore, according to Hedberg (op.cit.), the correct term for the Yoldia Clay is Portlandia Zone, or preferably Porllandia arctica Zone. Similar terms were used by A. lessen (1899, summarized by Hansen 1965, pp. 49-50) for northern Denmark, e.g. Older Yoldia Clay, Younger Yoldia Clay, Saxicava Sand and Zir-phaea Beds. On the Swedish west coast Ask-lund (1936) described a complete sequence of Late Weichselian mollusc zones. They were named 'Zones' but with climatological prefixes, such as 'High glacial' and 'Boreal', i.e. a combination of biostratigraphical and climatostrati-graphical terms. In the Baltic area Ancylus Clay and Littorina Clay have been used for sediments deposited during the Ancylus and Littorina phases respectively. There is obviously a biostratigraphic influence on the litho-stratigraphic as well as the chronostratigraphic terminology which should be avoided. There is also a need for revision of the Holoccne mollusc biostratigraphy of the Baltic Sea. Foraminiferal zones Feyling-Hanssen (1964) defined foraminiferal zones in the Oslofjord area, and designated the zones with letters (A-G). He also named some of the zones after the dominating species, i.e. the Bulimina marginata Zone, for which a completely adequate term would possibly be Bulimina marginata Acme-zone. The other zones could also have been named. It is of great importance, however, that these zones are clearly defined biostratigraphical units, and therefore do not cause confusion. From southwestern Norway and northern Denmark Weichselian foraminiferal zones are described as assemblage zones (Feyling-Hanssen et al. 1971). BOREAS 3 (1974) Quaternary stratigraphy of Norden 111 Diatom zones Study of the diatom flora in Holocene Baltic sediments is a classical field of research in Sweden and Finland. Diatom assemblages characterizing sediments deposited in waters of varying salinity have been described. In this way a series of diatom zones were described by Sundelin (1919), Thomasson (1927, 1935), and used by others, i.e. Florin (1944), Berglund (1964), Donner (1964), as combined biostrati-graphic and chronostratigraphic units. Thomas-son (1927, 1935) described the Gyrosigma phase and the Echineis phase as units preceding the Ancylus time, Sundelin (1919) the Masto-gloia time and the Clypeus time as transitional phases between the Ancylus and the Littorina time. The Mastogloia phase (and Mastogloia Sea) is still used as a synonym for the Early Littorina phase. However, diatom as well as mollusc names should be restricted to bio-stratigraphic zones - possibly with a correlation between the two systems. In this respect there is also a lack of information on which to base such a correlation. Pollen zones Earlier the biostratigraphy in a limnic-terrestrial environment was based on macrofossils, mainly fruits, seeds, wood remains etc. As early as 1841 Steenstrup described four Holocene zones on this basis: Aspen-, Pine-, Oak- and Alder-zones. They were based on biostratigraphy but used also for chronostratigraphy. Similarly An-dersson (1896) used the zonation: Dryas-, Birch-, Pine-, Oak- and Spruce-zones, when he described the history of the Swedish vegetation. Lennart von Post (1916, 1967) developed the analysis of pollen into a statistical method extremely useful for biostratigraphic correlation throughout the whole Quaternary. The method was originally used for relative chronology by geologists and later it was developed by botanists into a fine instrument for the interpretation of vegetational changes or ecosystem successions. The most widely used pollen zone systems for the Late Weichselian and the Holocene in Norden was that devised by Knud lessen in Denmark (K. Jessen 1935, 1938). Similar zone systems were described by Nilsson (1935) for southern Sweden and Fasgri (1935, 1940) for southwestern Norway. The zones were described as pollen assemblage zones, which were given numbers easily used in pollen diagrams. probably first applied by von Post (1925). These so-called pollen floristic zone systems were correlated with geologic and archaeologic chronologies and used as independent chronologies (e.g. Nilsson 1935:554, 1965, cf. Hakans-son 1971a:486; Hansen 1965:58; B. G. Andersen 1965:122; Lundqvist 1965:156; Donner 1965:264). According to Nilsson (1948a, b) the pollen zones could be correlated between southern Sweden, Denmark and Germany due to a presumed synchronism within the area. Jessen's zone system was applied to south-central Sweden (Fries 1951), to Finland (Sauramo 1949, 1958; Donner 1951) and to Norway (Fasgri 1954; Hafsten 1956). Meanwhile, Jessen's system was modified and introduced also into central Europe by Firbas (1949). In 1953 (Fasgri 1954:237) Scandinavian pollen analysts agreed to use Jessen's pollen zones for the Late Weichselian and the Holocene throughout the area. This marked an important step forward, as pollen zones were the most precise method for correlation available at the time. In this way the biostratigraphic pollen zone systems developed into synonymous chronostratigraphic systems. However, this led to a false picture due to the lack of an absolute chronology. The radiocarbon method did not provide dates until the 1960's. Zone boundaries based on the spread of immigrating species can of course be expected to be metachronous. This was proved for the Corylus expansion in southern Scandinavia (Berglund 1966a:110; Hafsten 1969), for Picea in Finland (Aario 1965) and for Picea in the whole of Norden (Moe 1970). Recently the extensive use of Jessen's zones has been criticized (Hafsten 1969, 1970; Mangerud 1970b; West 1970; Donner 1971; Birks 1973:280). The main objection is that several of the zones are not recognizable outside limited areas. For instance, Jessen's zone IX is characterized by the immigration of Car-pinus and Fagus. The former species is not found in Norway, and Fagus has a very restricted occurrence. Nevertheless, Jessen's zone IX is widely used also in Norway, the zone being identified on climatostratigraphical or chronostratigraphical criteria. Jessen's zone system is therefore not used exclusively as a bio-stratigraphical classification, but rather as a combination of biostratigraphical, climatostratigraphical and chronostratigraphical classifications. For the present-day need of precision, this is not satisfactory. 112 Jan Mangerud, Svend T. Andersen et al. BOREAS 3 (1974) The international rules (Hedberg ed. 1971a) also apply to pollen zones. In the case of Holo-cene diagrams these rules were probably first applied by Cushing (1967) in Minnesota, U.S.A., and in Norden by Mangerud (1970b), Donner (1971), Hyvarinen (1972, 1973) and Vorren (1972, 1973). For the Isle of Skye in Great Britain, Birks (1973:273-281, 325 ff.) has strictly applied the rules. The main experience and philosophy behind the proposal of the International Subcommis-sion on Stratigraphic Classification (Hedberg ed. 1971a) is the evolution and extinction of taxons, mainly of marine habitats. Biostrati-graphical units based on such criteria will only be found in one stratigraphical position, and usually have a wide geographical extension. Pollen zones in the Late Quaternary are to the contrary mainly based on the geographical migration of terrestrial plants, mainly tree species. These zones, therefore, might be repeated in a sequence, and generally the zones will have a very limited geographical extension. The latter is especially the case in Norden, because of its very differentiated topography and climate. Some modifications of the rules for the pre-Quaternary stratigraphy have to be made. A solution of some of these problems might be the definition of regional pollen assemblage zones, as proposed for Finland by Donner (1971). However, present-day pollen diagrams are generally not satisfactory for such definitions. We need an extensive collaboration for (1) the definitions of biostratigraphic reference areas, (2) thorough investigations of palaeoeco-logical standard profiles (preferably lake sediments within the reference areas), (3) correlation between regional biostratigraphy and the proposed general chronostratigraphy. In pollen studies, the most used type of zones have been assemblage-zones. For ecological interpretations, the use of pollen assemblage zones has clear advantages, as the pollen assemblage will reflect plant communities. From a stratigraphical point of view one might, however, prefer other types of zones, where the boundaries are defined by only one criterion. For instance an interval-zone, the Corylus rise/Alnus rise Interval-zone, defined by the rational limit of Corylus and Alnus, should be unambiguous and recognizable in large parts of Europe. The discussion above refers mainly to the biostratigraphy of the Late Weichselian and the Holocene deposits. However, the same aspects are valid for older Quaternary deposits. In recent investigations of pre-Weichselian and Weichselian sequences of Denmark, S. T. Andersen (1961:17-18, 1965) used numbered zones which correspond to pollen assemblage zones. Morphostratigraphy For many Late Pleistocene and Holocene sediments, the landforms are one of the major characteristics. Frye & Willman (1962) therefore suggested the recognition of morphostrati-graphic units, though Richmond (1959) and Flint (1971:200) are of the opinion that land-forms should not be used as stratigraphic units. No doubt, landforms do differ from the general stratigraphic idea of superposition of strata. Nevertheless, landforms originated at various moments in geologic time. For instance, a succession of shorelines at different levels, or a series of end moraines behind each other, are obvious bases for relative dating of geological events. The differences between morphostrati-graphical and traditional stratigraphical units are, firstly, that landforms are abstract surfaces, while traditional units are real bodies of rock strata, and secondly, that in traditional stratigraphy the younger units are found in vertical sequences above the older ones, while in morphostratigraphy the younger units often occur in lateral sequences at lower levels than the older ones. The important point for us is that in some cases it may be desirable to classify geological events on the basis of their land-forms. We therefore recognize a morphostrati-graphical classification. Moraines, as landforms, are generally associated with a geographical name, for instance the Ra moraines, the Central Swedish moraines and the Salpausselka moraines. The same terms might be used for the morphostratigraphic units, only in capitalized form (Aarseth & Mangerud 1974), for example the Ra Moraines. In Sweden the sequence of Late Weichselian 'moraine lines' could be named the Fjaras Moraines, the Moslatt-Berghem Moraines, the Central Swedish Moraines (cf. Morner 1969). Similarly the Minnesund Delta, the Lavik San-dur, etc. could be used for other morphostrati-graphical units. The term Substage is often used in the same sense. This is, however, a BOREAS 3 (1974) Quaternary stratigraphy of Norden 113 chronostratigraphical term and should not be used in connection with end-moraines. No generally accepted terminology exists for shorelines. Very often they are designated by letters and/or numbers, for instance Marthinus-sen (1960) in northern Norway used Sj-Sjg, P1-P12 and Nj-N9. He also used the term Tapes line, which is generally used in Norway for shorelines connected with the Tapes transgression. The numbered Holocene transgression schemes of the Kattegat (e.g. Morner 1969, with PTM 1-10) and the Baltic Sea (e.g. Berg-lund 1971b, with Littorina I-VI) refer to the morphologically recognized shorelines as well as the transgression sediments. Perhaps also shorelines should be named from type localities. Climatostratigraphy A major characteristic of the Quaternary was the great variations in climate, which have also naturally been a main base for stratigraphic subdivision. Synchronism of climatostratigraphical boundaries has been much discussed and has been briefly mentioned above. We will not discuss this further here, but merely state that in principle such boundaries are not strictly synchronous because climatic changes are complex, and because we do not observe the climatic changes themselves, only the impact of the changes on vegetation, fauna, glaciers, oceans, sediments etc. The question of synchronism is of course also a question of precision. For the pre-Weichselian, time-correlations based on climatic interpretations are certainly more precise than other methods of correlation available today. This, however, will not always be the case for the Holocene and the Late Weichselian. We will therefore maintain the distinction between climatostratigraphical units and chronostratigraphical units (Am. Com. Stratigraphic Nom. 1961; Mangerud 1970b). Flint (1971:373) has recently changed the concept of climatostratigraphy to glacial stratigraphy; in our opinion this is a restriction which is not desirable. We propose the definition that a climatostratigraphical unit is a stratigraphical unit with the boundaries defined by geological indications of climatic changes. Thus, the boundaries may be time-transgressive. Many of the stratigraphical units of the Quaternary - glacials, interglacials, interstadials, etc. - were primarily defined as climatostratigraphical units in the sense outlined above. In case they are changed to chronostratigraphical units, they should preferably be defined by stratotypes in such a way that the boundaries, at least in theory, are synchronous levels. Furthermore, the terms glacials (for instance the Weichsel Glacial) and interglacials should only be used either in a climatostratigraphical or in an informal sense. In formal chronostratigraphical units only the chronostratigraphical term (e.g. the Weichselian Stage) should be used, because a chronostratigraphical unit is defined by certain stratigraphical horizons, and climatic aspects such as glaciation etc., in principle, are not included in the definition. In Norden many stratigraphical units have originally been defined as climatostratigraphical in the sense used in this paper. We recommend that use of climatostratigraphical definitions be generally restricted; chronostratigrahpical units should receive preference. We therefore propose (see Chronostratigraphy) redefinition of the Late Weichselian classification (Boiling, Older Dryas etc.), and of the Holocene Blytt-Sernander classification (Preboreal, Boreal etc.), which by many investigators has been used as a climatostratigraphical classification, so that all these units can be considered chronozones. Chronostratigraphy 'A chronostratigraphic unit is a body of rock strata which is unified by representing the rocks formed during a specific interval of geologic time' (Hedberg ed. 1971b:6). Thus the boundaries of chronostratigraphical units are, in theory, everywhere of the same age. The aim of chronostratigraphical classification is the establishment of a hierarchical sequence of chronostratigraphical units, both of regional and world-wide scope, which can be accepted generally as a standard scale (Hedberg ed. 1971b: 17). The classification scheme for the Cenozoic in Norden given in Table 1 will be discussed in more detail in the following chapters. The units of higher rank, at least erathem, system and series, are of world-wide application, and agreement should be attained through international commissions. Our comments, 114 Jan Mangerud, Svend T. Andersen et al. BOREAS 3 (1974) therefore, are simply arguments for international discourse. Similarly, stages, and partly also chronozones, are valid in Europe outside Norden. As a vast number of classifications of litho-stratigraphical, biostratigraphical, morphostrati-graphical or climatostratigraphical nature have been applied to the Quaternary, we are of the opinion that a hierarchical chronostratigraphi-cal classification is indispensable for the unification of these many independent systems. Definition of boundaries Boundaries for chronostratigraphical units are usually defined by means of stratotypes. We suggest that this practice is applied for units of stage rank or higher. In the Pleistocene, stratotypes may also be preferred for boundaries of units of lower rank. For the Late Weichselian and the Holocene, however, we propose definition of chronozone-boundaries directly in conventional radiocarbon years. The International Subcommission on Strati-graphic Classification (Hedberg ed. 1970b: 15) ascertains that definitions in terms of radiometric ages would indeed accord with the concept of chronostratigraphical units. The commission has, however, two major objections to such definitions: (1) '. . ., these age determinations cannot yet be made with sufficient accuracy or in enough different kinds of rocks to make it practicable to define many of our currently used chrono-stratigraphic terms wholly on the basis of ra-diometrically determined time-spans in years'. (2) 'Likewise, the establishment now of a new scheme of chronostratigraphic units based only on equal intervals of time in millions of years, as some have proposed, would cause us to lose much of the value of the chronostratigraphic work of the past' (Hedberg ed. 1970b: 15). However, none of these objections have any real significance for our proposal, for the following reasons: (1) Nearly all the Late Weichselian and Holocene sediments may directly or indirectly be dated by the radiocarbon method. Actually, the correlations within this time-span are so dependent on radiocarbon datings that if a sediment cannot be correlated to a classification defined in radiocarbon years, then it cannot be correlated to any chronostratigraphical classification defined by stratotypes. (2) In Norden no chronostratigraphical classification of the Late Weichselian and Holocene exists showing satisfactory definitions of the respective boundaries. New definitions are therefore necessary in any case, and instead of proposing a new classification, we propose only precise definitions of already well established terms, leaving the age of the boundaries as close as possible to the tradition of the last few decades. Time scales Absolute ages of Late Weichselian and Holocene sediments can be obtained by several methods: radiocarbon dating, varve chronology, dendrochronology, historical documents, and others. Basically, all dates are related to one of two scales, namely radiocarbon years or calendar years. According to international agreements all radiocarbon dates are calculated on the basis of the Libby C14 half-life of 5568 (or 5570) years. The years in which the resulting time-scale is expressed are usually called conventional radiocarbon-years. Also other variables, such as recent activity, isotope fractionation, etc., have been standardized internationally. For marine shells, however, some varying calculation-procedures are used (Mangerud 1972: 144-146). Some years ago (cf. Olsson et al. 1962; Olsson 1968) new measurements of the half-life indicated that 5730 years is a more correct value than 5570 years and the new value was recommended for use (Godwin 1962). Some conventional radiocarbon ages were therefore recalculated according to the new half-life (Nilsson 1964; Berglund 1966a, b and others). In view of the results on the long-term variations in the length of the radiocarbon years (see below), we propose that a separate correction for the new half-life should no longer be applied. The year A.D. 1950 is used in all cases as the reference year (zero-year) for calculation of radiocarbon years. However, many scientists, in particular archeologists, recalculate dates to refer to the birth of Christ. We find this prac- BOREAS 3 (1974) Quaternary stratigraphy of Norden 115 G co-chronologic Era Period Epoch Age Chrono-stratigraphic Erathcm System Series Stage Holoccne Cenozoic Quaternary Pleistocene Tertiary Pliocene Miocene Oligocene Eocene Paleocene Flandrian Weichselian Eemian + Saalian s.l. + Holsteinian + Elsterian + 'huerglacial III' 'Glacial B' + 'Interglacial 11'= Harreskovian + 'Glacial A' + 'Interglacial 1'= Osterholzian Menapian Several stages in the Netherlands and North Germany Table 1. The generally used subdivision of the Cenozoic Erathem into systems and series. A provisional subdivision in stages for continental NW-Europe is also shown. In this column the horizontal lines indicate recognizable stage boundaries, and crosses stages represented in Denmark. References (for stages): S. T. Andersen 1965; Cepek 1967; Erd 1970; Goedcke ct al. 1966; Griiger 1968; Menke 1968a, 1972; Zagwijn 1960; Zagwijn et al. 1971; Zagwijn 1973. tice illogical, as this date is the zero-year for the calendar year system. Radiocarbon years, however, are not calendar years, as may be demonstrated by the fact that 4000 radiocarbon years B.C. are probably equivalent to almost 5000 calendar-years B.C. Our knowledge of the relationship between radiocarbon years and calendar years has greatly increased during the last few years, mainly through correlations of dendrochronology and radiocarbon dates of the tree-rings (Suess 1970; Damon et el. 1970; Ralph & Michael 1970). Conventional radiocarbon years can be converted directly to dendrochronological years (supposed to be indentical to calendar years) by means of a diagram (e.g. Olsson 1972: Fig. 1). However, the measurements do not cover the whole Holocene, and the results can be further refined. 9 — Boreas 3/74 We propose that only the following two timescales be used for the Late Weichselian and the Holocene, and that different zero-years be used in each case: (1) One scale in conventional radiocarbon years (T,/2=5568 years), using 1950 as zero-year. (2) One scale in calendar years, using the birth of Christ as zero-year. This scale and reference year should preferably always be used for calendar years, whether these are found by historical documents, correction of radiocarbon years, varve chronology or other methods. The Cenozoic Erathem The most common subdivision of the Cenozoic Erathem is shown in Table 1. The Cenozoic is 116 Jan Mangerud, Svend T. Andersen et al. BOREAS 3(1974) subdivided into the Tertiary and Quaternary Systems. The Quaternary is further subdivided into the Pleistocene Series and the Holocene Series. This classification has been proposed by the International Subcommission on Strati-graphic Classification (Hedberg ed. 1971b: 19), and generally accepted in Norden. During the last few years, however, several authors (for instance West 1968:224-225; Flint 1971:384) have proposed that the terms Tertiary and Quaternary should be discarded as stratigraphical terms; that the Pleistocene Series should include all post-Pliocene strata; and consequently that the Holocene Series should be discarded. This latter opinion has been supported in Norden by Hafsten (1969, 1970). Menke (1972) suggested a subdivision of the Quaternary into Cenocene (Kanozan) and Pleistocene, without a Holocene. Menke"s (1972) Cenocene would be a natural and useful unit, but it has not yet received general approval. In future a subdivision of the Quaternary into Cenocene-Pleistocene-Holocene should be seriously considered. There are, no doubt, arguments in favour of the suggested changes. There are, however, also arguments in favour of the established classification (for example Fairbridge 1968:526). Tradition is perhaps the most important argument. Therefore, we recommend that the classification proposed by the International Subcommission on Stratigraphic Classification should be generally accepted and used. The position of the Pleistocene/Holocene boundary has been very much discussed, and placed in a different way in different countries. In Norden, the boundary has generally been put at the Younger Dryas/Preboreal boundary, which is approximately 10,000 radiocarbon years B.P. At the congress of the International Union of Quaternary Research (INQUA) in Paris 1969 (Hageman 1969) this level has also been proposed to be the global boundary, and the search for a stratotype continues. At the meeting of the INQUA Commission for the study of the Holocene 1971, southern Sweden was chosen as the type area (cf. the preliminary report by Morner ed. 1973), but a good type section has not yet been found. The Pleistocene Series The pre-Weichselian stages. - As climatic change is a main feature of the Pleistocene, subdivisions in Northern Europe have been based on indications of major climatic changes. Intervals characterized by evidence of a predominantly cold climate have been called glacials, and predominantly warm intervals were termed interglacials. The Pleistocene has thus been considered a sequence of alternating glacials and interglacials. The glacials were characterized by a cold climate and deforestation of a wide area. However, correlation of these glacials with the Scandinavian ice sheets is not possible in all cases, mainly due to incomplete records in Norden. Continuous sequences are not known in all cases, and their identification and delimitation are sometimes problematic. The interglacials are characterized by evidence of a warm climate and continuous forest successions. Furthermore, eustatically conditioned marine transgressions have been recognized in several cases. They can be correlated by means of pollen analysis or other methods, at least in restricted areas, and can be arranged in a stratigraphic sequence. The glacials contain several geologically recognizable smaller units, which have been called stadials and interstadials. The interglacials and the interstadials are warm phases of two different orders, and their distinctiveness is possibly only a matter of duration. In most cases it has been agreed which units should have the rank of interglacials and which should have the rank of interstadials. A common glacial/interglacial chronology has been used to a wide extent in the Netherlands, northern Germany and Denmark. Elster, Saale and Weichsel, the old German names for glaciations, have been used for glacials, and names such as Eem and Holstein, originally coined for marine deposits, have been adopted for interglacials. Newly discovered units have been named in various ways, in some cases after the locality first described. A formal chronostratigraphy for the Pleistocene Series in the British Isles was established by Mitchell et al. (1973). The traditional names for glacials and interglacials were used for stage names. Stratotypes were selected from various publications, and certain pollen zone boundaries were used for stage boundaries. The British Pleistocene chronostratigraphy cannot be transferred to the continent because of difficulties in correlation. A chronostrati-graphic terminology for Norden, which must BOREAS 3 (1974) Quaternary stratigraphy of Norden ] 17 be common to the adjacent regions of (at least) northern Germany and the Netherlands, has not been established. Hence, we consider it desirable that a formal chronostratigraphy is worked out for continental northwest Europe. Such a chronostratigraphy must be based on the glacial/interglacial chronology, but it will be necessary to define stages and stage boundaries by means of stratotypes. Some difficulties will probably arise due to gaps in the record, and it may be necessary to define local stages in cases of doubt, but in our opinion an ultimate goal should be a common terminology. A provisional chronostratigraphic subdivision of the Pleistocene in continental northwest Europe is shown in Table 1. Deposits from several Pleistocene stages are known in Denmark (Table 1, S. T. Andersen 1965). Deposits from the Eemian have been recognized in Norway (Mangerud 1970a) and Sweden (Lundqvist 1971; Berglund & Lager-lund J 974). The Weichselian Stage. - A proposal for subdivisions of the Weichselian Stage is shown in Table 3. We propose that the Early Weich-selian/Middle Weichselian Substage boundary is placed at the Early Glacial/Pleniglacial boundary of van der Hammen et al. (1971), and that the Middle Weichselian/Late Weichselian Substage boundary is placed at a level corresponding to 13,000 radiocarbon years B.P. (see below). We propose that the stadials and interstadials formerly recognized within the Weichselian should be redefined as chronozones in a formal chronostratigraphy. At present, however, the terminology is incomplete. Within the Early Weichselian Substage the Amcrsfoort and the Brorup interstadials are known from the Netherlands, and Denmark (S. T. Andersen et al. 1960; S. T. Andersen 1961; Zagwijn 1963), Br0rup and Odderade from Germany (Averdick 1967; Menke 1970) and Brorup probably also from Sweden and Finland (Lundqvist 1967, 1971; Berglund & Lager-lund 1974; Korpela 1969). The Amersfoort in-terstadial was characterized by forest in the Netherlands and shrub vegetation in Denmark, and the Br0rup and the Odderade interstadials by forest. During the Br0rup interstadial subarctic forest could even have grown in northern Scandinavia. Within the Middle Weichselian Substage the Moershoofd, Hengelo and Denekamp interstadials with intervening stadials were recognized in the Netherlands and dated at respectively 50,000, 38,000 and 30,000 B.P. (van der Hammen et al. 1971). In the Netherlands none of these warm phases had associated forest. They were preceded and succeeded by intervals with extremely cold conditions (polar desert). All these stadials and interstadials may be redefined as chronozones in a formal chronostratigraphy. The Late Weichselian Substage. - The Middle Weichselian/Late Weichselian Substage boundary is identical to the Pleniglacial/Late-glacial boundary of van der Hammen et al. (1971). This boundary has been dated to about 13,000 B.P. (van der Hammen et al. 1971; Menke 1968b). We propose to accept this date as the definition of the discussed boundary. The Late Weichselian Substage therefore will correspond to the classical Late-glacial in northwestern Europe. It comprises the Oldest Dryas, B0lling, Older Dryas, Allerod and Younger Dryas 'periods' (Hartz & Milthers 1901; Iversen 1942, 1954, 1967, 1973). The term Late-glacial (and also Post-glacial) should, however, not be used for formal stratigraphic units. Van der Hammen (1957) placed the lower boundary of the Boiling interstadial at the lower boundary of the Late-glacial. Recent dating of the Zirphaea beds in northern Denmark now extends their age back to nearly 13,000 B.P. (Krog & Tauber, in press), and it seems natural to extend the Boiling back to that date in accordance with the classification adopted in the Netherlands (van der Hammen et al. 1971). However, it is well known that the biostratigraphic 'birch zone', named Boiling by Iversen (1942), has a lower boundary dated to ca. 12,300 B.P. In Denmark, southern Sweden and southern Norway where the Late Weichselian chronology developed, the old climato-biostratigraphic zones will correspond to the proposed chronozones. The Boiling Chronozone thus comprises the Oldest Dryas and Boiling periods or pollen zones la and lb of Iversen (1954). pollen zones DR1 and BO of Nilsson (1961) and Brondmyr Interstadial and preceding pollen zones (I, II and III) of Fajgri (1940). The Older Dryas Chronozone comprises the Older Dryas period or pollen zone Ic of Iversen and DR2 of Nils-son. The Allerod Chronozone comprises the Allerod period or pollen zone II of K. lessen 118 Jan Mangerud, Svend T. Andersen et al. BOREAS 3 (1974) Table 2. Radiocarbon dates of Late Weichselian boundaries. Only dates of biostratigraphically thoroughly studied profiles through limnic-tcrrcstrial sediments have been selected. The samples refer to levels at, iust below or above the discussed boundaries. Boundary Age No. Site Reference Boiling/Older Dryas 12,070+140 K-542 Usselo, van der Hammen 1951 Netherlands Tauber 1960b - 12,100+140 K-708 Witow, Poland Wasylikowa 1964 Tauber 1962 - 11,900+180 K-706 - - Older Dryas/Allcrod 11,890±140 K-1973 Boiling, Denmark Stockmarr unpubl. _ 11,770±215 St-2508 Klinchogkiinet, Mörncr 1969 Sweden - 11,740±170 St-1423 Losensjon, Berglund 1966a, Sweden Engstrand 1967 - 11,730±150 Lu-210 Trummen, Häkansson 1970 Sweden - 12,070+180 T-672 Blomoy, Mangerud 1970b, Norway Nydal et al. 1970 AllerOd/Younger Dryas 10,970±120 K-101 Ruds Vedby, Krog 1954, Denmark Tauber 1964 10,970±300 K-110 Boiling, Denmark Tauber 1960a - 11,090±130 K-1902 Stockmarr unpubl. - 10,915±230 St-2528 Algare mosse, Monier 1969 Sweden - 11,110±115 Lu-406 Bjorkerods Berglund 1971a, mosse, Sweden Häkansson 1971b - 11,060±100 Lu-209 Trummen, Hakansson 1970 Sweden - 10,800+300 T-152 Brondmyra, Chanda 1965 Norway Nydal 1960 - 10,940+180 T-624 Blomoy, Mangerud 1970b. Norway Nydal et al. 1970 Younger Dryas/Preboi eal 10,400+130 K-1903 Boiling, Denmark Stockmarr unpubl. 9,920+160 K-1604 Gravlev, - Denmark - 9,915+180 St-2165 Algare mosse, Monier 1969 Sweden - 10,120+100 Lu-738 Nackrosdammen, Berglund 1973 10,250+120 Lu-740 Sweden Häkansson in prep. - 10,320+105 Lu-408 Bjorkerods Berglund 1971a, 10,160+105 Lu-409 mosse, Sweden Hakansson 1971b - 9,920+150 St-805 Agerods mosse, Nilsson 1964 Sweden Östlund & Engstrand 1963 _ 9,850+100 Lu-121 Ranviken, Digerfeldt 1973 Sweden Häkansson 1969 - 10,170+230 St-1337 Hallarums Berglund 1966a, 10,000+170 St-1336 mosse, Sweden Engstrand 1967 - 10,170+160 St-1778 Slattmossen, Berglund 1966a, 9,700+150 St-1847 Sweden Engstrand 1967 10,360+105 Lu-208 Trummen, Digerfeldt 1972 10,230+105 Lu-207 Sweden Häkansson 1970 (1935) and AL of Nilsson, and the Younger Dryas Chronozone comprises the Younger Dryas period or pollen zone III of lessen and DR3 of Nilsson. All Late Weichselian chronozone boundaries seem to be climatically conditioned within southern Scandinavia, since there is a close correlation between biostratigraphic changes and deglaciation patterns (cf. Morner 1969:182; Tauber 1970; Mangerud 1970b; Berglund 1971a). As stated above, we propose that the chronozone boundaries should be defined by radiocarbon ages. The definitions are based on dates from non-calcareous lake sediments in southern Scandinavia - the only area of Norden which was ice-free during the greater part BOREAS 3 (1974) Quaternary stratigraphy of Norden 119 Table 3. Provisional chronostratigraphic subdivision of the Weichselian and the Flandrian in continental NW-Europe. Geo-chronologic Age Subage Chron Definitions of boundaries in conventional radiocarbon years B.P. Chronostratigraphic Stage Substage Chronozone Late Flandrian Subatlantic 2500 5000 8000 9000 10,000 11,000 11,800 12,000 13,000 Flandrian Middle Subboreal Flandrian Atlantic Early Boreal Flandrian Preboreal Younger Dryas Late Allerod Weichselian Older Dryas B0lling ? Denekamp Weichselian Middle Weichselian 9 Hengclo 7 Moershoofd 9 Odderade Early Weichselian Brorup 7 Amersfoort 7 of the Late Weichselian. However, for the oldest boundary we must refer to dates from central Europe. Therefore, southern Scandinavia, comprising Denmark, southern Norway and southern Sweden, will be an informal 'type area' for the Late Weichselian chronostratig-raphy. Compilations of dates of the boundaries have been given previously by B. G. Andersen (1968: 76-77), Morner (1969:175-177), Mangerud (1970b: Fig. 11), Berglund (1966a:113-118, 1971a: 13) and others. We have critically selected dates of limnic-terrestrial material related to the boundaries (Table 2) and these form the basis for our proposed definitions of the boundaries (Table 3) as follows: the lower boundary of the B0lling Chronozone is defined at 13,000 conventional radiocarbon years B.P.; the boundary between B0lling and Older Dryas Chronozones is defined at 12,000 B.P.; the Older Dryas/Aller0d Chronozone boundary at 11,800 B.P.; the Aller0d/Younger Dryas Chronozone boundary at 11,000 B.P.; and the Younger Dryas/Preboreal Chronozone boundary is preliminarily defined at 10,000 B.P. Hyvarinen (1973:91) has given a survey of several radiocarbon dates for Early Holocene biostratigraphic zone boundaries in eastern Finland. The boundary of the Artemisia zone and the birch zone in southeastern Finland probably corresponds to the Younger Dryas/Preboreal chronozone boundary and is dated at 10,000 to 10,200 B.P. However, this area is situated close to the Salpausselka moraines from the Younger Dryas, and there may be problems with the correlation of this area with southern Sweden. In our proposal the upper boundary of the Late Weichselian is defined at 10,000 B.P., mainly because the Pleistocene/Holocene boundary will be fixed at that age. However, 120 Jan Mangerud, Svend T. Andersen el al. BOREAS 3(1974) the dates mentioned above indicate that the generally used biostratigraphic boundary between the tundra and birch zones is slightly older in southern Scandinavia. The Holocene Series The Holocene has the rank of a series (Tabic 1), and could thus be subdivided into several stages. During the last few years, however, the most common practice has been to use the term Flandrian Stage, with a lower boundary at the Younger Dryas/Preboreal transition. In this usage the Flandrian Stage is identical to the Holocene Series, and two names are actually superfluous. Morner (1972) argues that the Flandrian (interglacial) should end at the beginning of the next glacial, while the Holocene should include several interglacials and glacials. However, stratigraphy is not concerned with the classification of future events. Nevertheless, for convenience, we propose to accept the present use, with the Holocene and the Flandrian covering the same period, but having different rank (Table 1). We propose that the Flandrian be subdivided into three substages, Early, Middle, and Late Flandrian (Table 3). The boundaries should be at the same level as the boundaries of the chronozones (see below). The Early Flandrian Substage includes the Preboreal and Boreal Chronozones, the Middle Flandrian includes the Atlantic and Subboreal, and the Late Flandrian Substage includes only the Sub-atlantic Chronozone. The Flandrian Stage. - At present, no generally accepted chronostratigraphical classification of the Flandrian exists in Norden. The most used classification is the modified Blytt-Ser-nander scheme in Preboreal, Boreal, Atlantic, Subboreal and Subatlantic. We therefore propose that these well-established terms, with redefined boundaries, be used for formal chronostratigraphical units (chronozones). Since we propose a redefinition of the Blytt-Sernan-der terms, wc will briefly review the original definitions. The terms were introduced by Axel Blytt (1876a, b); he used them, however, exclusively for floral elements of the present-day vegetation of Norway (boreal plants, atlantic plants etc.). The relation to stratigraphy was the following: Blytt (1876a, b) assumed that these floral elements had immigrated during succeeding periods, the arctic plants first, the subatlantic latest. In view of their present distribution, he also assumed that the boreal and subboreal plants had immigrated during periods of continental climate, while the atlantic and subatlantic plants had immigrated during periods of oceanic climate. In Norwegian peatbogs he found alternating layers of peat and stumps, and he related the peat-beds to the periods of oceanic climate and the stump-beds to the periods of continental climate (Blytt 1882a, b). In all these pioneer-papers Blytt consistently restricted the terms boreal, atlantic, etc. to the present-day floral elements of Norway, and avoided use of the terms for stratigraphic units. He published a large number of papers during the 1880's, but seems to have maintained this distinction. The earliest we have discovered the terms used for stratigraphy is in 1889 by Sernander (1899:199), where he uses the concepts of Blytt's 'atlantiska period' and Blytt's 'subboreala period', obviously connecting the terms atlantic, etc. to Blytt's descriptions of peat-bogs and his climatic interpretations. Later Sernander (1890: 17) used all Blytt's terms for succeeding periods, characterized by different climate, still referring to Blytt's investigations. This practice was then adopted by Blytt (1893), who in addition used the terms for the different layers he had described previously in the peat-bogs: the subatlantic peat-bed, the subboreal stump-bed, the atlantic peat-bed, etc. (translated from Blytt 1893:11). So far, the meaning of the terms had changed without any real discussion or new definitions. A fundamental paper in establishing these terms as stratigraphic unit-terms is that of Sernander (1894). He stated that phyto-pale-ontological zones, such as the aspen-birch-horizon, cannot be assumed a priori to be time-synchronous over large areas, due to the migration-time of the species (Sernander 1894:4). Therefore he examined methods for improved determinations of contemporaneity of deposits, and concluded that changes of sea-level and climate were the best methods available. Concerning climatic changes, he refers to Blytt's theories, of which he accepted the main description of the climatic development, though he rejected (Sernander 1894:70) many of Blytt's interpretations and ideas. BOREAS 3 (1974) Quaternary stratigraphy of Norden 121 In the peat-bogs of Gotland, Sernander (1894) found the same stratigraphy as Blytt had described from Norway. Therefore he (1894:71) used Blytt's terms (subatlantic, sub-boreal, etc.) for the different peat-beds, and also for the corresponding period. This publication (Sernander 1894) was later generally recognized as defining the 'Blytt-Sernander units', which soon after were widely used in Norden. The term 'preboreal' was first used in a general way for the time before the boreal period (e.g. Erdtman 1921; Fasgri 1935). Later 'preboreal' became restricted to the time interval between the Younger Dryas and Boreal (e.g. Faegri 1940), which has been the common practice the last decades. During the last few decades the Blytt-Sernander terms have been used with varying meaning, and most researchers (e.g. Nilsson 1961:9) exclude the climatic interpretations. Many authors (e.g. Iversen 1967, 1973; J0rgen-sen 1963; Hafsten 1956, 1960; Danielsen 1970; Fries 1951, 1965; Berglund 1966a, b; Konigsson 1968; Digerfeldt 1972; Donner 1963) also use the terms as synonyms for Jessen's or Nilsson's pollen zones, and therefore the boundaries are as asynchronous as the pollen-zone boundaries. One conclusion is indeed certain; at present no generally accepted definitions of the boundaries of the Blytt-Sernander units exist. Nevertheless, the units have mainly been applied as broadly synchronous zones (Table 4). Most of the boundaries cited in Table 4 are based on compilations from both small and large areas, though the boundaries given by Nilsson (1964) and Berglund (1966b) are partly based on series of dates from a single sediment or peat core also thoroughly biostratigraphical-ly studied. Such series of dates, which are important when dealing with Flandrian radiocarbon ages, are available from the following non-calcareous areas: Kongsmosen, SW Jutland, Denmark, 54 dates (Aaby unpublished); Ager0ds mosse, Scania, Sweden, 33 dates (Nilsson 1964; Ostlund & Engstrand 1963); Ranviken, Scania, Sweden, 25 dates (Digerfeldt 1973; Hakansson 1969); Ffallarums mosse, Blekinge, Sweden, 26 dates (Berglund 1966b; Engstrand 1967; Olsson 1965); Trummen, Sma-land, Sweden, 30 dates (Digerfeldt 1972; Hakansson 1969, 1970). Based mainly on the compilations in Table 4 and the mentioned series of dates, we propose the following definitions (Table 3) of chrono- 52 CO i o _o o- o c _ x ■o .i. o tional r bo n 1970 1965 1964 d 196 1967/ d def of ehr C O c C Conve radioc B.P. Hafste Donne O Ž Berglu [versei Propo; zones 1000 - 2000 - 3000 4000 - 5000 6000 7000 - 8000 9000 - 10000 - 1 55 i i Is! Table 4. The ages of the Blytt-Sernander zones, as given by some authors in Norden during the last decade. The boundaries of Nilsson 1964 and Berglund 1966b are recalculated to conventional radiocarbon years. Proposed definition of chronozones to the right. From Mangerud (1973: Fig. 5). zone boundaries, retaining the Blytt-Sernander terms, and attempting to maintain the essential chronological meaning attached to them during the last few decades: 122 Jan Mangerud, Svend T. Andersen et al. BOREAS 3(1974) Table 5. Proposed subdivision of Flandrian chronc-zones. Chronozone Subatlantic Subborcal Atlantic Boreal Prcboreal Subdivision Late Middle Early Late Middle Early Late Middle Early Late Early Late Early Radiocarbon years B.P. 1000 2000 2500 3000 4000 5000 6000 7000 8000 8500 9000 9500 10,000 (1) The boundary between the Younger Dryas and Preboreal Chronozones is preliminarily defined at 10,000 radiocarbon years B.P. This boundary may later be referred to the strato-type for the Pleistocene/Holocene boundary, if this stratotype should be finally accepted according to the intention discussed above. (2) The Preboreal/Boreal Chronozone boundary is defined at 9000 radiocarbon years B.P. Some dates indicate a shorter Preboreal, an alternative age for the boundary being 9500 B.P. However, for the majority of the area the proposed age is more relevant. (3) The boundary between the Boreal and Atlantic Chronozones is defined at 8000 radiocarbon years B.P. (4) The Atlantic/Subboreal Chronozone boundary is defined at 5000 radiocarbon years B.P. According to Olsson (1972: Fig. 1) this corresponds to 3600-4000 dendrochronological (calendar) years B.C. (5) The Subboreal/Subatlantic Chronozone boundary is defined at 2500 radiocarbon years B.P. According to Olsson (1972) this corresponds to 600-800 dendrochronological (calendar) years B.C. According to these definitions the duration of the Subboreal Chronozone is 2500 radiocarbon years. This is, however, a period of 'long' radiocarbon years, and 2500 radiocarbon years will probably correspond to approximately 3000 calendar years. To avoid confusion with previous usage the combined terms, e.g. Preboreal Chronozone should be used when referring to the present definitions. Subdivision of the chronozones. - With present-day precision of Holocene stratigraphy, there will no doubt be a need for a subdivision of the chronozones. Probably it will be most convenient to define units of that rank locally. There is, however, also a need for a standardized subdivision, during, for instance, the planned registration of palaeobotanical data. Therefore we propose the subdivision shown in Table 5. We emphasize that the definitions of these units have no real geological basis, they are just subdivisions into reasonable time-units, possibly with the rank of Subchronozones. Acknowledgements. - The authors would like to thank Dr. Waldo H. Zagwijn for reading through the manuscript critically, and Dr. Brian Robins for correcting the English language. Wc are also indebted to many Nordic colleagues for stimulating discussions. REFERENCES Aario, R. 1965: Die Fichtcnverhäufigung im Lichte von C"-Bestimmungen und die Altersverhältnisse der finnischen Pollenzonen. Bull. Comm. Geol. Finl. 218, 215-231. Aarscth, I. & Mangerud, J. 1974: Younger Dryas end moraines between Hardangerfjordcn and Sognefjor- den, Western Norway. Boreas 3, 3-22. American Commission on Stratigraphic Nomenclature 1961: Code of Stratigraphic Nomenclature. Bull. Am. Petr. Geol. 45, 645-665. Andersen, B. G. 1965: The Quaternary of Norway. The Geologic Systems. The Quaternary 1. 91-138. New York, London, Sydney. Andersen, B. G. 1968: Glacial geology of western Troms, North Norway. Norges Geol. Unclers. 256. 160 pp. Andersen, S. T. 1961: Vegetation and its environment in Denmark in the Early Weichselian Glacial (Last Glacial). Daum. geol. Unclers., 11. Ra'kke 75. 175 pp. Andersen, S. T. 1965: Interglacialer eg interstadialer i Danmarks kvart