boreas - Wiley Online Library

0 downloads 0 Views 2MB Size Report
Karl-Ernst Behre (e-mail: [email protected]), Niedersд chsisches Institut fь r historische ...... hennetz und das Haupthц hennetz 1985 der Bundesrepublik Deutsch-.

A new Holocene sea-level curve for the southern North Sea KARL-ERNST BEHRE


Behre, K.-E. 2007 (January): A new Holocene sea-level curve for the southern North Sea. Boreas, Vol. 36, pp. 82 102. Oslo. ISSN 0300-9483. A new sea-level curve (MHW, mean high water level) for the southern North Sea is presented, spanning the last 10 000 years and based on new data recently obtained along the German coast. The 118 dates were selected from basal as well as intercalated peats of the Holocene sequence and archaeological dates from the last 3000 years. Because of different MHW levels along the German North Sea coast, all data were corrected to the standard tide gauge at Wilhelmshaven to make them comparable. Special advantages of this area for sea-level reconstructions are negligible tectonic and isostatic subsidence and the absence of coastal barrier systems that might have mitigated or masked sea-level changes. Changes of water level had therefore immediate consequences for the facies and could be dated exactly. The chronostratigraphic Calais Dunkirk system has been improved and adapted to the new data. Altogether seven regressions (R 1 R 7) have been identified, each of them characterized by a distinct decline in sea level. These fluctuations are in accord with the evidence from other parts of the North Sea region. A draft of former North Sea shorelines is presented on the basis of this sea-level curve. Karl-Ernst Behre (e-mail: [email protected]), Niedersa ¨ chsisches Institut fu ¨ r historische Ku ¨ stenforschung, Viktoriastr. 26/28, D-26382 Wilhelmshaven, Germany; received 29th August 2005, accepted 29th May 2006.

The rise in sea level as a consequence of global warming is a topical issue for both specialists and the wider public. In areas where shallow seas predominate, the consequences of changing sea levels can be enormous, and shallow coasts such as in the southern North Sea region provide excellent sources for the documentation of geological and historical sea-level changes. After several predecessors such as Nilsson (1948), the first sea-level curve to rely only on radiocarbon ages was published by Jelgersma (1961). It became a benchmark work not just for The Netherlands but also adjoining areas. As in the case of the North Sea, sealevel curves were also constructed for other parts of the world (Pirazzoli 1991). They looked rather different and gave rise to the highly debatable question whether eustatic changes are best reflected by a smooth or an oscillating curve. The area around the North Sea offers the highest density of sea-level data in the world. In Germany, in particular, both the amount and the quality of data have greatly increased in recent years. Much information has been provided by geologists working in the Clay District, but many new data also derive from archaeological investigations in the coastal region. The latter provide a good time resolution for the younger part of the curve. Two important features of the southeastern North Sea make this area particularly suitable for evaluation of regional eustatic sea-level changes. First, the southeastern margin is the most stable one in the North Sea area with respect to tectonics and isostatic movements and, second, the coastline is open and without beach barriers, so it is

particularly sensitive to sea-level changes, which are readily registered. The article presents a carefully evaluated body of critically selected, available and unpublished data resulting in a new sea-level curve. For further details as to the data used, as well as the methods employed, the reader is referred to Behre (2003).

The CalaisDunkirk system Although the appropriateness of the Calais Dunkirk system has been discussed by some authors (e.g. Baeteman 1999), it is the only one that can be used for supraregional contexts because it is based on dated transgressive and regressive trends, and it has the only terminology that is applicable throughout the southern North Sea. In this contribution, the definition of these terms used hitherto has been changed in some details and  most importantly  regression phases have been inserted. The terms used in the Calais Dunkirk system have their origin in the Belgian French coastal region. Modifications of the system, which was first introduced in 1924, have been compiled by Roeleveld (1974); see also Baeteman (1981) and Ervynck et al. (1999). Initially, the system was based solely on lithostratigraphy. In particular, Dubois (1924), Tavernier (1948) and Tesch (1930) described the basic outline for a stretch of coast extending from NE France through Belgium to the SW Netherlands. At the base is the so-called ‘basis peat’, overlain by the mainly clastic Calais series which is separated by a peat layer from the following Dunkirk series which is also DOI 10.1080/03009480600923386 # 2007 Taylor & Francis

BOREAS 36 (2007)

predominantly clastic. This basic lithological succession corresponds with the stratigraphy as known from northern Germany. When this system was applied over a larger region, it gradually changed from a lithostratigraphic to a chronostratigraphic scheme, based on the assumption that the transgressions and regressions were regional in nature. This modification took place mainly in The Netherlands in the 1960s (Hageman 1969; Brand et al. 1965) and was finalized during the 1970s (de Jong 1971; Roeleveld 1974, see also Zagwijn 1986). It was also accepted and used in Germany (e.g. Menke 1988), such that the Calais Dunkirk system found wide acceptance and use as a chronostratigraphical framework. In most cases, peat layers provided the dates for the chronological subdivision of the Calais and Dunkirk series. However, not all peats are suitable for this purpose; dates from peats that are part of a widespread peat-forming event were preferred to dates derived from peats that represent local developments.

Methods and data The basis peat, which occurs at the base of the Holocene sequence, provided reliable sea-level data as it is situated directly above Pleistocene sediments. Compaction can therefore be excluded. Intercalated peats have generally been excluded from consideration in earlier articles because of the effects of possible compaction of the underlying sediments. Such peats, however, have yielded numerous important dates with respect to the occurrence as well as the time and duration of regression phases. Along open coasts, as in northern Germany, the formation of peats on top of marine or brackish deposits requires a change to freshwater conditions, and, as these intercalated peats often occur synchronously across a wide area, they signify with certainty a fall in sea level. If such peats lie on coarse deposits such as sand, or if it can be assumed that compaction came to an end before peat formation started, they have also been used to estimate the former height of MHW (mean high water level). A more precise estimation of the MHW level was also obtained by the composition of peat. In modern environments, the optimum elevation for the formation of Phragmites peat is between /0.5 m MHW and about MHW, while peat from tall sedges and tall herb communities is formed at approximately MHW level and fen-wood peat forms at about /0.5 m MHW (Jelgersma 1961; Menke 1988). The maximum heights of dated minerogenic deposits provided at least minimum sea-level heights that were often unavailable from other sources. An example is given by the top of the Calais IV deposits; these indicate a minimal MHW level of c. 0 m NN (/a.s.l., German Ordnance Datum) at about 1500 cal. yr BC.

Holocene sea-level curve, S North Sea


Deposits from above the MHW, such as the elevated levees along the lower reaches of rivers, have also been used as indicators of former MHW level. Here dates are derived from the wood remains of river bank forests, the composition of which is adapted to the frequency of floodings and indicates vertical distance to the MHW. Willow (Salix) is indicative of MHW, while oak (Quercus) and elm (Ulmus), which occupy the uppermost river-bank forest, suggest a water level of c. 0.8 m above MHW. Humic horizons with preserved plant remains from fossil salt marshes on the sandy barrier islands have recently been investigated and dated by Streif (1986) and Freund & Streif (2000). The great advantage of these sites is their stability with respect to compaction, as the islands are built of coarse and medium sand. Dates from archaeological contexts have been used intensively, especially for the last 3000 years. The height of occupation layers in former settlements was adapted to the maximum storm surge level. Therefore it was necessary to estimate the difference between MHW and storm surge level in the period in question. This strongly differs from modern times, when dikes reduce the areas over which flooding occurs, resulting in a considerable increase in the height of storm flood levels. Reliable estimates of the difference between MHW and storm surge level under natural conditions were deduced from archaeological excavations in the Clay District. At two sites, Hatzum-Boomborg on the lower river Ems (c. AD 0; Haarnagel 1980) and Feddersen Wierde on the coast north of Bremerhaven (in the 1st century AD; Haarnagel 1979), gangplanks (piers) were excavated which indicate MHW level and which could be linked with the contemporary occupation layers. The estimated difference of not more than 1 m between MHW and occupation layers in both places indicates surprisingly low storm floods at that time. For the period about AD 700, Meier (2001a) gives an estimate of c. 1 m for northern Dithmarschen in Schleswig-Holstein. Extreme storm surges with considerably higher flood levels can be excluded since these would have led to the destruction of the settlements in the whole area. Therefore the estimation of 1 m difference has been used for the reconstruction of MHW from storm flood data.

Reference levels and time-spans The original data of the publications used in this article are published in NN (German Ordnance Datum) or NAP (Dutch Ordnance Datum), which means about mean water or mean sea level (MSL). For coastal areas, in particular along flat shores, the use of MSL is not practical because its line runs in front of the coast; therefore, the mean high water level (MHW) was chosen for construction of the sea-level curve.


Karl-Ernst Behre

According to the different tidal amplitude (Behre 2003), the MHW differs along the coast and from the coastal zone to inland where it is also influenced by additional morphological and hydrological factors, so the data from the coast itself are the most reliable. In compiling data from the various coastal areas, all depths had to be corrected to a standard tide gauge, for which Wilhelmshaven was chosen. In the curves presented in Figs 1, 3 and 7, the symbols always show the local MHW as given by the original authors, while

BOREAS 36 (2007)

the curve itself is based on corrected data that are indicated by crosses. As the tidal range is dependent on coastline configuration and bathymetry, there must have been a considerable deviation from the modern tidal range during the early Holocene, in particular before the connection of the North Sea with the English Channel, and during the existence of the Dogger Bank. Recent considerations on former tidal ranges have been made by Austin (1991) and Shennan et al. (2000). According to their models, the difference

Fig. 1. New curve showing sea-level changes (MHW) in the southern North Sea. Horizontal lines with short vertical lines at the ends show calibrated 14C dates and the 1s range of the calibrated date, which is often asymmetrical after calibration. Where there are no vertical lines, age as estimated by pollen analytical or archaeological dating is indicated. The curve is based mainly on the position of the / symbols. Possible compaction effects and transgression and regression trends that can be fixed only in time have also been considered. For numbering of the points in the later part of the curve, see Fig. 3.

Holocene sea-level curve, S North Sea

BOREAS 36 (2007)

between mean tide level and high tide level east of Yorkshire increased from 1.6 m at 8000 uncal. yr BP to 2.1 m at 6000 uncal. yr BP. This deviation from the modern curve, however, affects only the earliest part of the curve, and this influence is only small, due to the steepness of the curve in this period. Once the Dogger Bank had disappeared (after 6000 cal. yr BC), the differences in tidal amplitude compared with the modern ones became small. In order to establish a curve with a chronology based on calendar years, all 14C dates were calibrated using CALIB 4.3 (Stuiver et al. 1998). In the sea-level curve, the symbol is placed at the most probable point within the 1s age range of the calibrated date. The accuracy of other dates, such as from pollen analysis and archaeology, was estimated, too, and indicated by horizontal lines in Figs 1 and 3.

Table 1. New ages and MHW levels of transgressions and regressions along the southern North Sea. Age cal. BC/AD

Age uncal. BP (before 1950)





























/3900 /4150

5100 5300



( /6650)



Regression 7


Dunkirk IIIb


Regression 6


Dunkirk IIIa


Regression 5


Dunkirk II


Regression 4


Dunkirk Ib


Regression 3


Dunkirk Ia


Regression 2


Calais IV


Regression 1


Calais III


Calais II

The Calais series, including Regression 1 Only a few contacts are known from the North Sea itself, which can be used for construction of the sealevel curve. The oldest peat profiles used here are from the southern Dogger Bank (46 m water depth, point 1 in Table 3; cf. Behre & Menke 1969), the Oyster Ground (42.5 m water depth, pt. 2; cf. Behre et al. 1984) and the White Bank (c. 38 m water depth, pt. 3; cf. Ludwig et al. 1979). There are two even deeper contacts (no peats) at 54 and 72 m water depth (Konradi 2000) that were not included here because they are located north of the area covered by the curve. They lie within the expected range though somewhat higher, which is probably due to the stronger isostatic uplift in the north. Another index point from the Dogger Bank area, situated about 50 km north of the region of the curve (Fig. 2) is no. 55//02/213VE (Shennan et al. 2000). With a depth of /3l.06 m OD (a.s.l.) and an age of 81409/55 yr BP (7105 [7291 7062] cal. yr BC, AA 22662) it fits into the curve very well. The curve (Figs 1, 7) shows a very rapid sea-level rise during the Calais I transgression which averages 1.25 m per century. However, a clear distinction between the

m NN (a.s.l.) /1.70

Dunkirk IV

Results: construction of the new sea-level curve The main focus of this article is the identification of temporary falls in sea level (Fig. 1). Previous researchers have already recognized regressions (Louwe Kooijmans 1974; Menke 1988) or regressive phases (Roeleveld 1974; van de Plassche 1985); however, these regressions were regarded as periods during which the otherwise continuous sea-level rise was interrupted rather than as periods of sea-level decline. The exact meaning of the term ‘regression’ is the horizontal retreat of the sea, not a vertical decline. This term is often used in both senses, however, and this is the case in this article, too.



Calais I

deposits of C I and C II is not yet possible. Occurrences of wood near Emden (Barckhausen 1984) may indicate a slowing or stagnation of the sea-level rise around 5350 cal. yr BC, which could correspond to the separation of C I from C II. Afterwards, the curve steepens again and the C II phase represents the first extensive demonstrable transgression. During this period the basal peats around Wilhelmshaven (pts 21 and 22) were also covered by marine deposits. Sea-level rise, however, slowed, although without an interruption, at c. 4000 cal. yr BC, between C II and C III. During this interval, the first intercalated peats were formed in the area of Emden (Hv 6320: 52409/60 BP, 4018 [4218 3977] cal. yr BC, Barckhausen & Streif 1978) and at the lower river Weser (Hv 102: 53509/ 130 BP, 4021 [4338 3992] cal. yr BC; Streif 1993). These continued to grow for c. 200 years before they were covered by sediments. It is unlikely, however, that a fall in sea level occurred during this time.


Karl-Ernst Behre

BOREAS 36 (2007)

Fig. 2. Map showing location of all points used for the sea-level curve.

In the course of the Calais III transgression from c. 3900 cal. yr BC, the MHW level rose again considerably. Marine and brackish deposits from this phase are found far inland in Lower Saxony and SchleswigHolstein. During this period, MHW must already have reached /2.00 m NN (b.s.l.). A key point is pt. 35 in Fig. 1, the base of an intercalated reed peat that lies at /2.22 m NN. This indicates the lowest possible position of the MHW of that time, but if allowance is

made for probable compaction of the underlying Holocene material, c. /2.00 m NN is more realistic. This is supported by pts 32 34, which derive from surfaces of intercalated peats, and a single basal peat sample. At these sites MHW may have reached the same level, if strong compaction of the underlying peat is taken into account. Menke (1988) described the surface of C III deposits at even /1.90 m NN in the Eider area, which is in agreement with Louwe

Holocene sea-level curve, S North Sea

BOREAS 36 (2007)


Table 2. Radiocarbon ages from peats of Regression 1; if available, the date relates to the base of the peat layer (upper sites close to the coast, downwards with increasing distance from the coast). Locality

Laboratory no.


G 24 GK Hooksiel G 69 GK Wangerland G 82 GK Wangerland G 72 GK Wangerland G 84 GK Wangerland G 71 GK Wangerland 2/77 Sahlenburg/Cuxhaven Ge 103 GK Emden-West G 124 GK Emden Sehestedter Moor/Jade bay II/4-2 Land Wu¨rden/Lower Weser G 98 GK Nordenham G 119 GK Nordenham B 287 GK Nordenham G 76 GK Jade G 206 GK Brake G 203 GK Brake G 18 GK Brake

Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv Hv

41759/65 42109/45 42609/95 42959/95 44259/45 44509/95 42709/130 43759/55 43159/70 41609/45 43809/140 44259/130 43259/105 43809/120 42509/70 42709/145 41659/250 44009/160

2771 2877 2884 2900 3066 3096 2888 2987 2909 2753 2956 3066 2914 2956 2883 2888 2757 3021

44659/75 41009/85 41059/95

3245 (3345 2936) BC 2622 (2870 2496) BC 2652 (2876 2494) BC

9943 4153 7680 4159 6688 4156 8256 6319 10452 8019 7203 13654 13663 100 13667 14836 14246 14824

Schleswig-Holstein: start of autogenic series in peats Delve/Eider Hv 625 Braaken/Miele Hv 1445 Meldorfer Moor Hv 6193

C years BP

Kooijmans (1976) who, for the Rhine/Meuse delta, also assumed a MHW of /2.00 m NAP. The increase in sea level during C III is followed by the first general fall of the MHW which can be traced in the German coastal area and is called Regression 1 (R 1). This regression is connected with a far-reaching change to freshwater conditions and the subsequent formation of peat. The peat, often called ‘middle peat layer’, is such a common and synchronous phenomenon along the German coast that it cannot be regarded as caused by local coastal changes alone. The available data indicate that peat formation extended, at least in parts, beyond the present coastline. Recent mapping by the Geological Survey (Table 2) provided numerous 14C dates from various parts of Lower Saxony that lie consistently between 3000 and 2800 cal. yr BC (c. 44004200 14C yr BP). In Schleswig-Holstein, Menke (1988) was able to show indirectly the onset of this regression based on the start of an autogenic series in the pollen spectra of peat deposits of the Eider/Miele district. In The Netherlands, some authors suggest a shortlived transgression within this period of regression and have referred to it as C IVa to distinguish it from the next transgression (cf. Louwe Kooijmans 1974; Griede 1978). The weak nature of the evidence, however, led Roeleveld (1974) to include C IVa for the northern Netherlands in C III, and van de Plassche (1985), too, regarded it as uncertain. The numerous dates from peat relating to this period from northern Germany do not support the idea of a C IVa transgression (Table 3). After 2400 cal. yr BC (3900 yr BP), R 1 was followed by another considerable rise in sea level, the Calais IV

Calibrated years (2882 2623) BC (2884 2703) BC (2920 2703) BC (3017 2876) BC (3259 2925) BC (3347 2920) BC (3021 2680) BC (3085 2911) BC (3016 2882) BC (2877 2624 )BC (3335 2881) BC (3350 2899) BC (3084 2879) BC (3328 2884) BC (2912 2710) BC (3080 2637) BC (3079 2458) BC (3352 2881) BC

Source Streif (1985) J. Barckhausen, pers. comm. J. Barckhausen, pers. comm. J. Barckhausen, pers. comm. J. Barckhausen, pers. comm. J. Barckhausen, pers. comm. Linke (1979) Barckhausen & Streif (1978) Barckhausen (1984) Streif (1984) Preuss (1979) Streif (1993) Streif (1993) Streif (1993) Streif (1993) Barckhausen (1995) Barckhausen (1995) Barckhausen (1995)

2002 2002 2002 2002 2002

Menke (1988) Menke (1988) Menke (1988)

transgression. In some inland areas, peat formation continued, but closer to the coast it ceased and clastic deposits followed. Soon, /1.0 m NN was reached, as is shown by the base of an intercalated peat near Wilhelmshaven (pt. 36 in Figs 1, 2). After a short slowdown (but without interruption), the increase in MHW continued. It should be noted that along the German coast a distinction between the C IV and Dunkirk 0 transgression is not possible. At this time there was neither a regression nor a stagnation. The clastic deposits that hitherto were assumed to represent the D 0 transgression should be included in C IV, which was followed by R 2 (see below). According to the scheme proposed for The Netherlands D 0 covers the period between 3500 and 3000 yr BP (Zagwijn & van Staalduinen 1975; Vos & van Heeringen 1997: p. 12). Doubt remains, however, as to whether this was really a transgression. According to Louwe Kooijmans (1974), extensive peat formation took place in the western Netherlands during this time; in the Dutch scheme of Jelgersma (1979: fig. IV-36), D 0 starts with peat formation which corresponds to the Upper Peat in Germany, while Roeleveld (1974: p. 101) notes a decreasing rate of sea-level rise during this period which often resulted in a change from fen peat to raised bog peat in the backswamp areas, and the formation of an extended layer of intercalated peat closer to the coast (Roeleveld 1974: plates I, II). In Dutch sites where, as well as peat, also clastic deposits have been assigned to D 0, it should be checked whether they might be more correctly assigned to C IV. In summary, the available evidence suggests that the D 0 transgression in The Netherlands is mainly


Karl-Ernst Behre

BOREAS 36 (2007)

Table 3. Main data of the points used for the sea-level curve construction. For location of points, see Fig. 2 (for more details, see Behre 2003). A/Number of the point; B/locality of the point (Fig. 2); C /situation and material (OL/occupation layer, tf/tidal flat); D /depth given by the original author (all dates refer to NN (Normal Null, German ordnance datum /a.s.l.), except for nos. 1 and 2, which refer to mean water level and 3, 4, 6 8, which refer to Marine Chart datum); E /correction of the material or the depth to the local MHW (m NN); F/correction of the local MHW to the MHW at the standard tide gauge Wilhelmshaven; G /type of dating and 14C-laboratory numbers; H/calibrated 14C years BC (AD in italics); I/references (p.c./personal communication). A 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61








Dogger Bank Oyster Ground White Bank Eastern Dogger Bank B172 Marne B 35 NE of Wangerooge A 10 NE Wangerooge A 10 NE Wangerooge A 10 Scharho¨rn 56/67 Scharho¨rn 58/67 Neuwerk 60/67 GK Hooksiel G 57 Land Wu¨rden II/1 GK Wilhelmshaven G 738 GK Emden-West G 320 Land Wursten I/4 GK Emden-West G 103 GK Emden-West G 81 GK Wilhelmshaven G 479 GK Wilhelmshaven G 743 GK Wilhelmshaven G 743 GK Hooksiel G 24 GK Wilhelmshaven G 749 Wilhelmshaven Kanalweg GK Emden G 170 GK Hooksiel G 19 Land Wursten III/7 GK Hooksiel G 26 Fo¨hr, F 3/1 GK Wilhelmshaven G 743 GK Wilhelmshaven G 743 Fo¨hr, F 2/1 GK Hooksiel G 24 Sehestedter Moor indike GK Wilhelmshaven G 744 GK Wilhelmshaven G 749 GK Wilhelmshaven G 743 GK Hooksiel G 44 Wilhelmshaven Kanalweg GK Hooksiel G 19 Below Feddersen Wierde Sehestedter Moor outdike Rodenkirchen Esing-Haferacker Jemgum I PRIA Hatzum-Boomborg OL 1A Jemgumkloster PRIA Hatzum-Boomborg Rheiderland Hatzum-Boomborg Boomborg Westdorf Jemgumkloster OL 3 Barward OL 1 Feddersen Wierde OL 1 a Bentumersiel OL 1 Eppingawehr Fo¨rriesdorf/Zissenhausen Seeverns Hodorf OL 1 Boomborg RIA

5 cm fen peat with Salix 19 cm reed peat/gyttja 13 fen peat 3 cm peat Contact on top of gyttja Lagoon with Phragmites Lagoon with Phragmites Lagoon with Phragmites 7 cm gyttja on Pleistoc. Gyttja/reed peat Wood on top of Pleist. Top of basal reed peat Base of basal peat Top of basal fen peat Wood from river bank Top of basal reed peat Wood from carr peat Wood from river bank Base of basal peat Base of basal fen peat Top of basal fen peat Top of basal reed peat Top of basal reed peat Top of reed peat Wood from river bank Top of basal peat Base of intercalated peat Top of basal reed peat Top of basal peat Basis intercalated peat Top intercalated peat Top basal peat Top basal peat Top intercalated peat Base of Upper Peat Base of Upper Peat Base of Upper Peat 6 cm Upper Peat Base of Upper Peat Total of 8 cm Upper Peat Base of Upper Peat Base reed peat on clay Lowermost OL Peat in beach barrier Lowermost OL Lowermost OL Lowermost OL House post OL 1A Oak stump on river bank House post OL 2 Base OL 5 Base Flachsiedlung Base OL 3 on top of DIb Base OL 1 Base Flachsiedlung 1a Base OL 1 on river bank Base OL on river bank Base OL Base OL Base OL 1 Pier on a gully

/46.00 /42.50 /37.99 /37.28 /29.50 /22.58 /21.95 /21.51 /24.90 /24.34 /21.31 /14.32 /12.78 /12.37 /11.90 /11.50 /10.16 /9.42 /7.42 /6.73 /6.55 /5.92 /7.05 /7.10 /5.14 /7.18 /5.97 /5.74 /5.13 /4.65 /4.43 /3.93 /3.60 /3.20 /2.22 /0.92 /2.32 /1.19 /0.74 /0.18 /0.65 /0.22 /1.00 /0.10 /0.56 /0.90 /0.86 /0.90 /0.40 /0.40 9/0.00 9/0.00 /0.20 /0.25 /0.30 /0.40 9/0.0 /0.15 /0.40 9/0.0 /0.60

/46.00 /42.50 /38.28 /37.58 /29.00 /24.23 /23.60 /23.16 /24.40 /24.34 /21.31 /14.32 /13.28 /12.37 /12.40 /11.50 /10.66 /9.92 /7.92 /6.73 /6.55 /5.92 /7.05 /7.10 /5.94 /7.18 /5.97 /5.74 /5.13 /4.65 /4.43 /3.93 /3.60 /3.20 /2.22 /0.92 /2.32 /1.19 /0.74 /0.18 /0.65 /0.22 /1.80 /0.90 /1.36 /1.70 /1.66 /1.70 /1.20 /1.20 /0.80 /1.00 /1.00 /0.75 /0.70 /0.80 /0.80 /0.65 /0.60 /0.80 /0.70

/44.80 /41.60 /36.88 /36.08 /28.74 /23.83 /23.20 /22.76 /24.04 /23.98 /21.04 /14.08 /13.28 /12.37 /12.08 /11.43 /10.20 /9.46 /7.92 /6.73 /6.55 /5.72 /7.05 /7.10 /5.62 /6.98 /5.90 /5.54 /4.43 /4.65 /4.43 /3.23 /3.40 /3.23 /2.22 /0.92 /2.32 /0.99 /0.74 /0.02 /0.58 /0.25 /1.80 /0.65 /1.04 /1.38 /1.34 /1.38 /0.88 /0.88 /0.48 /0.49 /0.68 /0.68 /0.63 /0.48 /0.48 /0.41 /0.46 /0.60 /0.38

pollen Hv 12092 Hv 7095 pollen Hv 6189 Hv 8602 Hv 8601 Hv 8600 Hv 2575 Hv 2143 Hv 2242 Hv 11607 Hv 7195 Hv 8928 Hv 6306 Hv 7133 Hv 6321 Hv 4755 Hv 8944 Hv 8936 Hv 8935 Hv 9944 Hv 8943 pollen Hv 10453 Hv 9940 Hv 7434 Hv 9946 KI 1050.01 Hv 8934 Hv 8933 KI 1049.02 Hv 9943 Hv 8019 Hv 8937 Hv 8941 Hv 8932 Hv 10674 Hv 6904 Hv 9939 pollen KI 4159.09 arch. /14C Hv 6753 archaeol. arch. /14C archaeol. GRN4669 Hv 2051 GRN4668 archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol.

7755 7113 7234 7499 6925 6948 6919 6417 6618 6525 6159 5817 5938 5946 5317 5726 5343 5303 4514 5013 4665 4396 4151 3964 3965 3466 3518 3398 3431 3355 2917 2662 2771 2755 2516 2156 1932 1664 1520 1519 1300 1248 900 801 650 550 550 534 397 398 350 50 100 25 50 25 50 50 50 150 0

I (8240  7560) (7281  7080) (7466  7056) (7597  7144) (7047  6838) (7176  6534) (7055  6706) (6456  6264) (6688  6480) (6638  6462) (6397  5916) (5985  5727) (6022  5911) (6064  5785) (5471  5149) (5837  5642) (5470  5301) (5478  5044) (4900  4238) (5207  4856) (4772  4461) (4499  4335) (4306  3984) (3982  3944) (3985  3944) (3618  3357) (3637  3370) (3628  3376) (3633  3357) (3497  3101) (3078  2887) (2875  2503) (2882  2623) (2877  2624) (2834  2466) (2295  2042) (2029  1783) (1745  1525) (1680  1415) (1645  1430) (1400  1200) (1259  1130) (950 850) (829 791) (700 600) (600 500) (600 500) (584 484) (404 383) (403 392) (400 300) (0 100) (130 70) (50 0) (100 0) (50 0) (0 100) (0 100) (0 100) (100 200) (30 30)

Behre & Menke (1969) Behre et al. (1984) Ludwig et al. (1981) Ludwig et al. (1981) Behre et al. (1979) Ludwig et al. (1981) Ludwig et al. (1981) Ludwig et al. (1981) Linke (1982) Linke (1982) Linke (1982) Streif (1985) Preuss (1979) Streif (1981) Barckhausen (1984) Preuss (1979) Barckhausen (1984) Barckhausen (1984) Streif (1981) Streif (1981) Streif (1981) Streif (1985) Streif (1981) Behre et al. (1975) Barckhausen (1984) Streif (1985) Preuss (1979) Streif (1985) Hoffmann (1980) Streif (1981) Streif (1981) Hoffmann (1980) Streif (1985) Streif (1984) Streif (1981) Streif (1981) Streif (1981) Streif (1985) Behre et al. (1975) Streif (1985) Ko¨rber-Grohne (1967) Behre & Kucˇan (1999) Strahl (2002a, b) Menke (1988) Haarnagel (1957) Haarnagel (1969) Brandt (1980) Haarnagel (1969) Behre (1970) Haarnagel (1969) Haarnagel (1969) Haarnagel (1980) Brandt (1980) Brandt (1980) Haarnagel (1979) Brandt (1980) Brandt (1980) Brandt (1980) Brandt (1980) Brandt (1980) Haarnagel (1980)

Holocene sea-level curve, S North Sea

BOREAS 36 (2007)


Table 3. (Continued) A







62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94

Feddersen Wierde OL 1c Lu¨dingworth Osterende Tholendorf/Eiderstedt Ostermoor Feddersen Wierde OL 1 c Einswarden, Flachsiedlung Ritsch Tiebensee (Dithmarschen) Borkum BO 1 b Barnkrug Feddersen Wierde OL 2 Juist Su¨derbusenwurth Borkum BO 2 Tofting Feddersen Wierde OL 3 Langwarden Feddersen Wierde OL 4 Feddersen Wierde OL 5 Feddersen Wierde OL 6 Feddersen Wierde OL 7 Feddersen Wierde OL 8 Wangerooge Wangerooge Wellinghusen Elisenhof Elisenhof Elisenhof Groothusen Niens Oldorf Juist Neuwarfen

Base of OL 1c Base of OL Peat in beach barrier Base of OL Pier on a gully Base OL on river bank Base OL Base OL of settlement 1 Hydrobia ulvae Base OL Lowermost house OL 2 Soil of salt marsh House of lowermost OL Cerastoderma edulis Base of lowermost OL Lowermost house OL 3 Base of OL Lowermost house OL 4 Lowermost house OL 5 Lowermost house OL 6 Lowermost house OL 7 Lowermost house OL 8 Scrobicularia in upper tf Scrobicularia in upper tf OL 1 Bridge over gully Plough land Base Wurt (elev. levee) Base OL II Base OL I (elev. levee) Base of Wurt Soil on salt marsh Base of Wurt (field)

/0.70 /0.65 /0.70 /0.70 /0.15 /0.80 /0.80 /1.00 /0.26 /0.85 /0.85 /0.48 /1.50 /0.40 /1.45 /1.35 /1.40 /1.50 /1.65 /1.80 /1.40 /1.40 /0.10 9/0.00 /2.00 /0.90 /1.30 /2.00 /1.80 /1.50 /1.00 /0.38 /0.85

/0.30 /0.35 /0.30 /0.30 /0.25 /0.20 /0.20 9/0.00 /0.11 /0.15 /0.15 /0.08 /0.50 /0.10 /0.45 /0.35 /0.40 /0.50 /0.65 /0.80 /0.40 /0.40 /0.10 /0.20 /1.00 /0.60 /0.60 /0.60 /0.20 /0.50 /0.40 /0.02 /0.05

/0.63 /0.10 /0.05 /0.10 /0.18 /0.20 9/0.00 /0.10 /0.60 /0.05 /0.08 /0.48 /0.57 /0.61 /0.66 /0.42 /0.54 /0.57 /0.72 /0.87 /0.47 /0.47 /0.46 /0.56 /1.16 /0.81 /0.81 /0.81 /0.66 /0.64 /0.64 /0.38 /0.29

95 96 97 98 99

Hatzum-Burg Alte Boomborg Klunderborg Pellworm Juist P 7 b

Base OL on river bank Base OL on river bank Base OL on river bank Base Flachsiedlung Soil on salt marsh

/0.70 /0.45 /0.30 /0.50 /0.60

/0.10 /0.35 /0.50 /0.50 /0.20


Juist P 6 a

Soil on salt marsh



Juist P 6 b

Soil on salt marsh

102 103 104

Juist Borkum BO 1 a Juist P 8


H 35 50 77 75 35 90 0 50 94 0 100 39 149 167 125 150 200 200 250 300 350 400 538 634 691 750 750 725 725 675 625 786 838


/0.22 /0.03 /0.18 /0.18 /0.60

archaeol. archaeol. Hv 3371 archaeol. archaeol. archaeol. archaeol. archaeol. Hv 22942 archaeol. archaeol. Hv 13131 dendrochr. Hv 22943 archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. archaeol. Hv 9257 Hv 300 dendrochr. archaeol. archaeol. archaeol. archaeol. archaeol. dendro. Hv 13130 Hv 20508 Hv 20509 archaeol. archaeol. archaeol. archaeol. Hv 22575

(25 45) (0 100) (4 131) (50 100) (25 45) (130 50) (50 50) (0 100) (351 49) (50 50) (75 125) (109 213) (146 150) (358 2) (100 150) (125 175) (175 225) (175 225) (225 275) (275 325) (325 375) (375 428) (426 616) (421 768) (680 700) (700 800) (700 800) (700 750) (700 750) (650 700) (620 630) (686 992) (682 982)

Haarnagel (1979) Brandt (1980) Menke (1988) Bantelmann (1960) Haarnagel (1979) Brandt (1980) Brandt (1980) Meier (2001a) Freund & Streif (2000) Brandt (1980) Haarnagel (1979) Streif (1986) Meier (2001b) Freund & Streif (2000) Bantelmann et al. (1984) Haarnagel (1979) Brandt (1980) Haarnagel (1979) Haarnagel (1979) Haarnagel (1979) Haarnagel (1979) Haarnagel (1979) Hanisch (1980) Sindowski (1969a) Meier (2001a) Bantelmann (1966) Bantelmann (1966) Bantelmann (1975) Reinhardt (1965) Brandt (1991) Schmid (1994) Streif (1986) Ey (1995)

850 925 900 850 1123

(800 900) (875 975) (850 950) (800 900) (1019 1220)

Brandt (1980) Brandt (1980) Brandt (1980) Higelke et al. (1984) H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 Freund & Streif (2000) Freund & Streif (2000) H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 H.Freund, H.Streif, p.c. 2001 Barckhausen (1969) Streif (1986) Schu¨tte (1939) Rohde (1975) Rohde (1975) Rohde (1975) Rohde (1975) Siefert & Lassen (1985) Gezeitenkalender (2002)



Hv 22570

1372 (1261 1419)




Hv 22568

1297 (1223 1413)

9/Brackish reed peat Gavia stellata in upper tf Soil on salt marsh

/0.40 /0.12 /0.94

/0.40 /0.27 /0.54

/0.80 /0.98 /0.94

Hv 22941 Hv 22576

1367 (1282 1413) 1375 (1284 1393) 1452 (1414 1639)

Juist P 7 a

Soil on salt marsh




Hv 22572

1443 (1412 1616)


Juist P 3 b

Soil on salt marsh




Hv 22563

1418 (1325 1447)


Juist P 3 a

Soil on salt marsh




Hv 22561

1336 (1238 1450)


Juist P 2

Soil on salt marsh




Hv 22559

1352 (1293 1409)


Juist P 1

Soil on salt marsh




Hv 22554

1404 (1305 1433)

110 111 112 113 114 115 116 117 118

Langeoog Wangerooge Sande Cuxhaven Cuxhaven Cuxhaven Cuxhaven Cuxhaven Cuxhaven

Vegetation horizon Roots in salt marsh Soil on salt marsh 1644 Tide gauge datum 1784 Tide gauge datum 1852 Tide gauge datum 1900 Tide gauge datum 1950 Tide gauge datum 1975 Tide gauge datum 2000

/1.56 /1.55 /0.70 /1.04 /1.15 /1.30 /1.40 /1.45 /1.50

/0.86 /1.15 /0.40 /1.04 /1.25 /1.30 /1.37 /1.45 /1.50

/1.27 /1.47 /0.40 /1.29 /1.50 /1.55 /1.62 /1.70 /1.75

Hv 1998 Hv 12303 historical hydrolog. hydrolog. hydrolog. hydrolog. hydrolog. hydrolog.

1293 (1255 1398) 1407 (1305 1438) 1644 1784 1852 1900 1950 1975 2000


Karl-Ernst Behre

characterized by peat formation which is more or less equivalent to the Upper Peat in the adjoining German region. The interpretation offered by Dutch authors (e.g. van de Plassche 1982) is that this peat was formed during a slowing but still rising sea level. If this is correct, then it is appropriately denoted as a transgression. However, the overall evidence points to a period of sea-level lowering, so it is proposed that the D 0 transgression be replaced by R 2. The end of the C IV transgression is fixed in time by dates from the base of the Upper Peat which is widespread in Germany and began to form between 1500 and 1300 cal. yr BC (3200 3000 yr BP; Table 3). Before this, an important sea-level maximum was reached that was not recognized earlier. The base of the Upper Peat indicates a minimum level for MHW at that time, but as a compaction of the underlying beds cannot be totally excluded, the original level must have been even higher. A key area is the Wangerland, northwest of Wilhelmshaven, where the Upper Peat is extensive and was encountered in 77 cores (J. Barckhausen, pers. comm. 2002). The regressive contact of the Upper Peat is undulating, due to varying compaction of the underlying Holocene deposits. The highest position of the C IV deposits lies above /0.10 m NN in seven instances, with a maximum of /0.02 m NN. In fitting this to the sea-level curve, 0.20 m has to be added to correct to the MHW standard of Wilhelmshaven, so that the maximum is /0.22 m NN. Another profile to the east and close to Hooksiel indicates that the top of C IV lies at /0.18 m NN, which when corrected gives /0.02 m NN (Streif 1985; pt. 40 in Fig. 1). Further to the south in Sehestedter Moor, at the southeastern margin of the Jade Bay, Behre & Kucˇan (1999) recorded the surface of C IV deposits at /0.22 m NN (pt. 42). Here, greater compaction is probable because of the comparatively soft underlying clay. Also, in other coastal areas, C IV occurs at relatively high elevations, but not as high as those mentioned above. In summary, the use of basal dates from the Upper Peat indicates that the surface of the C IV deposits which date to c. 1500 cal. yr BC was at c. 0 m NN or slightly higher, where no compaction has occurred. This is considerably higher than previously assumed for the southern North Sea region (e.g. Jelgersma 1979; Linke 1982; van de Plassche 1982). The Upper Peat (Regression 2) The Upper Peat is the best known phenomenon in the Holocene sequence along the southern North Sea. Although it was eroded in some parts during later periods, it still covers very large areas and generally rests on C IV deposits. Many 14C dates are available for the base and top of this peat, in particular from Lower Saxony (Table 3). They show that the Upper Peat started to grow between 1550 and 1300 cal. yr BC

BOREAS 36 (2007)

(3300 3000 yr BP). This suggests a widespread change from marine to freshwater conditions, which can only be explained by a fall in sea level. Numerous cores in the Wadden Sea, between the mainland and the East Frisian Islands (Sindowski 1957; Barckhausen 1970) and in the area northwest of Cuxhaven (Linke 1979) showed that the Upper Peat stretches considerably beyond the present or historical coastline. In Schleswig-Holstein, the Upper Peat is quite common in the North Frisian Wadden Sea. Also to the west, in the Dutch marine Clay District, another open coast, the Upper Peat (‘Holland V regressive interval’ after Roeleveld 1974) is quite common and covers large areas. There are many 14C dates from the peat as well as from vegetation horizons which date his ‘Holland V’ to 32252975 yr BP (Roeleveld 1974: plates I III, Table 2; i.e. c. 1500 1200 cal. yr BC); this fits well with the evidence from the adjoining German area. Further to the southwest, the regression is no longer expressed by a sharply delimited intercalated peat layer. Due to protection by the coastal barrier system, the so-called hollandveen (Dutch peat) developed over a long timespan and, in several areas, was itself interrupted by intercalated clay layers that represent transgressive phases (Jelgersma et al. 1979; Zagwijn 1986). In many places, the composition of the Upper Peat shows a hydrosere succession starting with reed peat which is followed by sedge peat and ends with fen wood. This means a decline in the ecologically effective groundwater table in the area, which is related to the MHW. In large areas this peat formation is capped by several decimetres of raised bog peat. The cessation of fen peat formation, which is dependent on availability of groundwater, and the change to ombrotrophic raised bogs indicate a prolonged lowering in groundwater level and indirectly the related MHW. For The Netherlands, the maps of Zagwijn (1986) show that, in extensive areas from the north to Zeeland in the south, the composition of the hollandveen had changed around 3000 yr BP from fen peat to raised bog. Zagwijn (1986: p. 17) emphasizes that this must have happened very rapidly, since raised bog peat rests without transition on top of fen peat. This widespread event points to a strong decrease in groundwater level and hence a considerable fall in MHW. The 14C dates for the end of the Upper Peat in Germany show considerable variation and may be affected by erosion as well as contamination by roots. Acceptable dates from the area close to the coast include 28509/60 yr BP (1002 [1125 920] cal. yr BC, Hv 21652) and an identical date from a neighbouring profile (Hv 21658; J. Barckhausen, pers. comm. 2002). Four dates from south of Cuxhaven gave an average age of 870 cal. yr BC (Sindowski 1969b). From the area northwest of Emden there are several dates, starting with 29159/55 yr BP (1126 [1254 1004] cal. yr BC, Hv 4739) at the coast and ending with 26659/60 yr BP (815 [891 798] cal. yr BC, Hv 4740) further inland

BOREAS 36 (2007)

(Barckhausen & Streif 1978). In some areas away from the coast, this peat continues throughout the following D Ia transgression until the strong D Ib transgression or even further. The significantly lower sea level allowed occupation of the former salt marshes. Several settlements have been excavated in the West Frisian part of The Netherlands close to the coast at Hoogkarspel, Andijk and Bovenkarspel. They date to 1600 1000 yr BC and it has been shown by botanical investigations that the former marine and brackish area was increasingly influenced by fresh water. A considerable decrease in groundwater level took place which led to compaction and even to the formation of inversion ridges, on top of which the first houses were erected. In the settlement area of Bovenkarspel*Watertoren, the top of the ridge lies between /1.40 m and /2.40 m NAP (Bakker et al. 1977: p. 192). Taking into account the younger houses on the flanks of the ridges, a fall of the local MHW level to below /3.0 m NAP during R 2 is possible. MHW correction to the standard Wilhelmshaven gauge is about 1.0 m and suggests a comparable MHW of /2.00 m NN. Later occupation at these sites between 1000 and 800 cal. yr BC had to contend once again with a rising water table (Ijzereef & van Regteren Altena 1991). In Germany, the MHW decline during R 2, compared with the maximum of slightly above 0 m NN of the preceding transgression phase, is estimated to be in the range 1.60 2.00 m. This can be shown by the

Holocene sea-level curve, S North Sea


height of the first settlements in the German Clay District at Rodenkirchen on the lower banks of the river Weser (Strahl 2005). They start, however, only at the beginning of the subsequent sea-level rise. Pt. 43 (Fig. 3) shows the position of the oldest house in Rodenkirchen, which rests at /1.00 m NN, indicating a MHW level of /1.80 m NN. However, a possible compaction of the underlying peat of up to 0.2 cm has to be taken into account, such that a MHW level of /1.60 m NN is more realistic. That means good agreement between Lower Saxony and Dutch Westfriesland of a substantial sea-level decline during R 2 and also the ending of R 2 at c. 1000 cal. yr BC, as already deduced from peat dates and supported by archaeology. It has to be emphasized that the Upper Peat (sensu stricto) ends with the start of the D Ia transgression (see below). This transgression is not very strong (Fig. 3), so in some areas peat development continued until the end of R 2, and in these cases the whole peat layer may be called Upper Peat (sensu lato). The Dunkirk series with intermittent regressions The strong Regression 2 which led to the formation of the Upper Peat was followed by a new unit  Dunkirk  which in the coastal zone is dominated by clastic deposits. From this point in time onwards archaeological evidence provides much data with respect to transgressions and regressions (Behre 2004).

Fig. 3. Younger part of the sea-level curve. Symbols as in Fig. 1. C /Calais; D/Dunkirk; R/Regression. For the exact identification of the symbols, see coordinates in Table 3. P-R IA/Pre-Roman Iron Age; RIA/MP/Roman Iron Age and Migration Period; EMA /Early Middle Ages; HMA/LMA/High and Late Middle Ages; MT/Modern Times.


Karl-Ernst Behre

From c. 1000 cal. yr BC the sea level rose again during the Dunkirk I transgression, which can be subdivided into D Ia and D Ib. The Dunkirk Ia phase shows only a slight increase of the MHW, interrupted by calm periods. This is indicated by the existence of prehistoric settlements. The first house at Rodenkirchen/Weser, already mentioned, was erected around 900 cal. yr BC on top of the earliest D Ia deposits, while the house of a second occupation phase was built around 800 cal. yr BC on a thin clay layer from subsequent floodings, but was also protected by platforms against the rising groundwater (Strahl 2005). The first flooding phase between 1000 and 900 cal. yr BC can be correlated with the interruption of the occupation at Bovenkarspel/NL at the same time (see above). The following Regression 3 starts around 800 cal. yr BC and marks the end of the D Ia phase. In Germany at that time peat formation began again in some places (cf. Table 3 in Behre 2003). Based on the subsequent settlements, the MHW at D Ia can be estimated as c. /1.40 m NN. For the period 650 400 cal. yr BC, several settlements have been excavated which were established on non-elevated surfaces (pts 4551 in Fig. 3), the most important being Hatzum Boomborg on the banks of the lower river Ems (Haarnagel 1969). It started around 550 cal. yr BC in an environment of river-bank forests (Behre 1970, 1985) and was abandoned in the 4th century BC due to flooding. This regression has also been described from the northern Netherlands, where it was considerably stronger. The occupation of the Frisian Clay District generally started at about 600 cal. yr BC. As was shown by the important excavations at Ezinge (Waterbolk 1994), the first settlements were erected on level ground (in Ezinge around 500 BC). The oldest radiocarbon date for such a settlement, however, dates to 25559/ 35 yr BP (787 [796 672] cal. yr BC, GRN 7902) for Middelstum (Boersma 2005). In response to the rising storm flood level of D Ib, the settlements in the area of Friesland and Groningen were raised to the first dwelling mounds (Wurten) from about 400 cal. yr BC. The height of the MHW during R 3 is cautiously estimated at /1.60 m NN. The MHW was probably even lower, which would be supported by the Dutch evidence: the settlements from around 600 yr BC were established there on former salt marshes, which indicates a considerable lowering of the MHW. The culmination of D I was reached during the Dunkirk Ib phase, when sea level rose strongly and rapidly, with severe consequences for the coastal zone. This took place between 400 and 150 BC. In Germany, all settlements in the Clay District were abandoned and a new coastline developed. In particular at the mouths of small rivers, the sea advanced into the hinterland and new bays were formed (Behre 1999). The deposits of this transgression cover large areas. Elevated levees were a characteristic feature of the coastline and these show the new coastal configuration. The ridges nor-

BOREAS 36 (2007)

mally attain a height of about /1.40 m NN; exceptionally they reach /2.10 m NN in Lower Saxony and /2.20 m NN (adjusted value /2.41 m NN) in Schleswig-Holstein (Behre 2003). The associated MHW level is estimated to be /0.60 m NN (adjusted value). This extraordinary event was followed by another very distinct fall in sea level, Regression 4, which is traceable everywhere around the southern North Sea (Behre 1986 and Table 3). The decrease in MHW was rapid and rather sudden, as is shown by the very limited peat formation at this time. Instead, there was a widespread soil formation, which suggests relatively dry and non-saline conditions. Where mires had remained in the backswamp areas, fen peat often gave way to raised bog formation. Indicative dates include 199 (396 125) and 196 (367 107) cal. yr BC, and the somewhat younger dates of AD 41 and 80, from the Sehestedter Außendeichsmoor on the eastern shore of Jade Bay (Behre & Kucˇan 1999; Behre 2005). During this regression, the first extensive colonization of the Clay District along the southern North Sea took place. These settlements were again erected on non-elevated surfaces (Flachsiedlungen), because they were not threatened by storm floods. The earliest date back to c. 130 cal. yr BC and they rapidly increased in number during the last century BC. The minimum of this regression can be estimated from the position of settlements on a solid substratum. Three key sites (pts 53 55; Fig. 3), Jemgumkloster near the Ems estuary and Barward and Feddersen Wierde on the outer Weser bank, gave heights between /0.68 and /0.63 m NN (corrected to the Wilhelmshaven standard gauge) and all date from the last century BC. Soon after the Birth of Christ the sea level rose again and gangplanks (a pier), dated to c. AD 35 (pt. 66), suggest a MHW of around /0.18 m NN (Haarnagel 1979). The end of R 4 can be placed at AD 50 and marks the onset of the Dunkirk II transgression. From this time, all settlements in the Clay District were raised to dwelling mounds (Wurten) to protect the inhabitants. In response to the increasing storm flood level,

Fig. 4. Schematic cross-section of the Wurt (dwelling mound) Feddersen Wierde, north of Bremerhaven. Thin lines/settlement layers; thick lines/occurrence of houses.

BOREAS 36 (2007)

Holocene sea-level curve, S North Sea


Fig. 5. The sequence of transgressions and regressions and their consequences for landscape and settlement. C /Calais; D /Dunkirk; R/Regression.


Karl-Ernst Behre

these dwelling mounds were continuously raised. The tops of the raised settlement layers were often raised considerably above the storm flood level, and so they do not necessarily indicate the maximum storm flood level in a particular period. The lowermost buildings of a settlement layer are best used for sea-level reconstruction so that more or less complete excavation is essential as of the Wurt Feddersen Wierde, north of Bremerhaven (Haarnagel 1979). Figure 4, constructed from the plans and dates presented by Haarnagel, shows that the storm flood level increased to /1.80 m NN until occupation layer 6 (at c. AD 300) and then decreased to /1.40 m NN (Fig. 3: pts 55, 62, 72, 77, 7983). Given the estimated difference of c. 1 m between MHW and general storm flood level, MHW towards the end of the D II transgression was c. /0.87 m NN (adjusted value). There are only a few proxies to determine the exact beginning of Regression 5 that followed the Dunkirk II transgression. It is clear from the settlement history that there was a general lowering in storm flood level, and also a decrease in MHW. After the gap in habitation during the Migration period, a new occupation of the Clay District took place again in the form of numerous settlements on non-elevated ground (Flachsiedlungen, Schmid 1988; Behre 2002). Occupation layer 7 in the Feddersen Wierde indicates a decline in MHW at about AD 350 to /0.40 m NN (corrected /0.47 m NN). This is independently supported by geological points 84 and 85 from the island of Wangerooge (Sindowski 1969a; Hanisch 1980). The retreat of the North Sea can also be shown by the seaward growth of the salt marshes in many places in Lower Saxony and Schleswig-Holstein. The following Dunkirk III transgression has been subdivided in The Netherlands into D IIIa and D IIIb, which agrees with Germany. Around AD 700, the sea level rose again and forced the settlers in the Clay District to raise their dwelling places again, establishing a new generation of Wurten. There are only a few reliable indicators for the maximum of the Dunkirk IIIa phase, and all are from settlements (Table 3). They indicate that the minimum MHW was above /0.80 m NN. By c. AD 850, D III was interrupted by Regression 6. During this time new Flachsiedlungen were erected on flat areas previously at risk of flooding. These include three settlements in the lower Ems area, one north of Wilhelmshaven and another on the island of Pellworm in Schleswig-Holstein (pts 94 98). This lowering of sea level has so far largely been unrecorded because most settlements at that time were on existing dwelling mounds. This regression is supported by evidence from the southern Netherlands, where a compilation by Vos & van Heeringen (1997) for Zeeland showed four Flachsiedlungen and six circular fortresses that had been erected in the Clay District at the end of the 9th and in the 9/10th centuries, respectively. The authors concluded that storm flood

BOREAS 36 (2007)

levels declined there at that time. For the German coast, a MHW of 9/0 m NN is tentatively assumed. Subsequent to R 6, the archaeological sources largely lose their value as indicators of sea-level change. With the erection of the first dikes in the late 11th century, the area influenced by storm floods was reduced and this led to a widening of the difference between MHW and storm flood level, the estimation of which is not possible. The lack of archaeological data is compensated, however, by dates from fossil vegetation horizons from the East Frisian island of Juist (Freund & Streif 2000; H. Freund, pers. comm. 2001). R 6 probably lasted until about AD 1100 and was followed in the 12th century by the D IIIb phase. While the dates for pts 99109 (Fig. 3) scatter to some extent, the dates for pts 106 111 are fairly tightly clustered and suggest that the rise in MHW culminated at c. /1.40 m NN in the 14th century AD. This is only 30 cm below the present level. This significant rise of the MHW should at least partially be ascribed to human activity in that the continuous dike line was closed in the course of the 13th century and with this many drainage systems were cut off, which also had consequences for the MHW. The 14C dates from the Frisian Islands suggest that the end of D IIIb was at c. AD 1450, which was followed by Regression 7 (Fig. 3). The MHW decreased rapidly and reached a minimum at AD 1644 (pt. 112). For this, a well-dated old dike on Pleistocene ground at Sande, south of Wilhelmshaven, was used as a key point (Schu¨tte 1939). Both global and regional factors played a role in the marked decline in MHW at that time. At the global level, the cooling during the Little Ice Age led to a general fall in sea level, while at the regional level the newly created medieval bays of Dollart, Ley and Jade (13th to 15th century) reached their maximum extension, a multiple of their modern size at c. AD 1500 and the same is true for the north Frisian Wadden Sea. The daily tides could once again extend over a wider area and, within the rather confined German Bight, this led to a decline in MHW. The final rise of the MHW, the Dunkirk IV transgression, which began around AD 1700, also had a global as well as a regional component. The global factor is connected with the warming at the end of the Little Ice Age which led to a rise in MHW (Fig. 3), while the regional factor is connected with the extensive re-diking during the last three centuries which has also contributed to an increase in MHW in the southeastern North Sea.

Tectonic and isostatic influences In reconstructing a sea-level curve for the southern North Sea, both long-term tectonic as well as isostatic crustal movements are concerned. The North Sea basin is characterized by long-term and still ongoing tectonic

BOREAS 36 (2007)

subsidence that extends back to the Tertiary. The amount of this subsidence can be estimated with the help of marine Eemian deposits which have been encountered at several places along the German coast. Their upper surface rises from /6.41 m NN (b.s.l.) in the west at Juist to /5.00 m NN in the east in the centre of Schleswig-Holstein (Behre et al. 1979). This means that the subsidence of this area between the maximum level of the Eemian Sea around 120 000 years ago and today amounts to between 0.64 cm and 0.54 cm per century. These modest long-term differences suggest that tectonic movements can be neglected for the reconstruction of Holocene sea-level curves for this region. The isostatic component is more difficult to ascertain, especially since it may have changed during the Holocene. According to Donner (1995), the zero line of the present Scandinavian uplift runs through Jutland in Denmark. Little is known about the isostasy of the southern North Sea region during the early Holocene. However, because of the steepness of the curve during this time the influence of isostasy on it must have been small. There has been much discussion regarding a possible subsidence during the Holocene in the southeastern North Sea area. To obtain more reliable information, three precision levelling surveys were carried out along the German coast in the years 1928 1931, 1949 1955 and 1980 1985. Comparison of the results from the first and second surveys suggested an average subsidence of less than 1 cm per century, which was not regarded as significant (Gronwald 1960). In contrast, the third series of measurements shows considerable deviations in both directions (Augath 1993; Wu¨bbelmann 1993), including an extreme surface subsidence northwest of Emden that is estimated, on a century basis, to be as much as 14 cm. This, however, has to be connected with the local exploitation of a large gas field. Further to the west, in the area of the Groningen gas field, surface subsidence of up to 18 cm was noted in the 26-year interval between 1964 and 1990 (Doornhof 1992). Other irregularities in the German coastal area may also have anthropogenic origins, such as water extraction. Therefore the third precision levelling cannot be used to solve questions relating to sea-level changes. In summary, the available measurements and estimates suggest a subsidence of B/1.0 cm/century at present and probably also for the Holocene as a whole. This includes tectonic as well as isostatic movements.

Comparison with other regions in NW Europe Many sea-level curves are now available (see for example Pirazzoli 1991). There has been a long discussion as to whether the eustatic changes occurred gradually and continuously during the Holocene or

Holocene sea-level curve, S North Sea


whether there were clear oscillations of a eustatic character. In this article, it is shown that there have been several obvious falls in sea level during the last 5000 years in the German North Sea area (Fig. 7). Short comparisons are now made with other regions. The Netherlands and Belgium The coast of the northern Netherlands is geologically a continuation of the German North Sea coast. Since the benchmark article by Jelgersma (1961), it has generally been accepted in The Netherlands that there was a continuously rising sea level without any declines. Nowadays, this curve is often referred to as a ‘trend curve’, which provides an opportunity for discussion and refinement. Well-defined sea-level falls have been demonstrated, particularly in archaeological publications, but these have been explained in most cases as local phenomena, while the general course of sea-level change should have been upwards. There is some evidence that already Regression 1 can be traced in The Netherlands. A fall of MHW around 4200 yr BP was accepted by van de Plassche (1982) on the basis of a peat layer from north of the Old Rhine estuary, but he considered it as local only. However, the archaeological evidence is much stronger. During the late Neolithic Vlaardingen culture occupation advanced in several places into the freshwater tidal area. In particular, the most important site, Hekelingen in the Rhine delta, shows settlements between 2900 and 2600 cal. yr BC. The excavator (Louwe Kooijmans 1974: fig. 13; 1985) postulated a distinct lowering in MHW at that time which he interpreted as being local. Other settlements from the Single Grave culture in North Holland were excavated by Hogestijn (1997) and dated to between 2600 and 2400 cal. BC. They rest on marine clay, but the investigations showed a decreasing marine influence and a change to freshwater conditions. Clear evidence for R 2 can also be found in many places in The Netherlands. Particularly convincing is the upper peat layer in the Groningen Clay District (Roeleveld 1974) which corresponds to the Upper Peat in the adjoining German region. The main part of the so-called ‘surface peat’ which extends over large areas in the Belgian Clay District was also formed in the same period (Ervynck et al. 1999). The archaeological evidence, in particular from West Friesland, provides strong support for a considerable fall in MHW and the mean sea level as well. This was indicated in the sealevel curve of Louwe Kooijmans (1974: fig. 13) but was interpreted by the author to be only of local significance. As regards R 3 between D Ia and D Ib, the marked occupation period of the Pre-Roman Iron Age in the western (Dutch) part of Friesland corresponds very well with the occupation phase in German East Friesland.


Karl-Ernst Behre

R 4 can also be demonstrated in The Netherlands, for example by the expansion of settlements with streepband pottery (200 BC to AD 100) in the Clay District of Friesland and Groningen (Waterbolk 1966). It is further supported by extensive soil formation on top of D Ib deposits, which are known from several parts of The Netherlands (Schoute 1984). For the Belgian coast, the archaeological evidence shows quite a number of new settlements that were established in the Clay District on former marine sediments around the Birth of Christ. This is in agreement with soil formation on top of D Ib deposits (Thoen 1978, 1987). R 5 is not as distinct in The Netherlands and Belgium as it is along the German coast. The relatively small R 6, however, is clearly visible in the archaeological record from the southwestern Netherlands (Vos & van Heeringen 1997). Based on considerable differences between the sealevel curves from Belgium and the western Netherlands, Kiden et al. (2002) assumed major crustal movements in this region during the early Holocene. This region is well away from that of the present study, and the early Holocene sea-level data are still too few to support this idea. Great Britain For Great Britain, the evidence of sea-level changes is complicated because of strong isostatic uplift in the north and tectonic subsidence in the southeast (Shennan 1989). However, if these movements are taken into consideration, at least qualitative trends of sea-level fluctuations can be demonstrated. A suitable region for comparisons and a key for England seems to be East Anglia, where the general trends resemble those along the southern North Sea. In the Fenlands, there is extensive intercalated peat, likewise called Upper Peat, which is similar in age to the German Upper Peat horizon (Waller 1994). The sequence of transgressions and regressions defined by Shennan (1986) is surprisingly similar to that described in this article (latest updates according to Waller 1994: p. 14). In particular, the sea-level decline Fenland IV (c. 4500 to c. 4200/3900 yr BP) is equivalent to R 1, Fenland V (c. 3300 to c. 3000 yr BP) is equivalent to R 2, Fenland VI starts at c. AD 50, mid-way in R 4, and Fenland VII (c. AD 800 to c. AD 1000) may be equivalent to R 6. All these regressions are regarded as sea-level lowerings (see also Shennan 1994). In particular, the strong decline around 1200 cal. yr BC, which corresponds to R 2, was recently confirmed for the Norfolk area when the foraminifera-based transfer function was applied (Horton & Edwards 2005). It should be mentioned, however, that Kidson (1977) proposed a smooth curve without regressions for the Somerset Levels in southwestern England; he regards the intercalated peats in his sections ‘as responses to minor variations in

BOREAS 36 (2007)

the relative rate of sedimentation and sea-level rise’ (Kidson 1977: p. 288). The comparison with sea-level curves from northwest England (Tooley 1978, 1982) is more difficult because of the isostatic movements there. However, steep relative declines of MHW, such as R 4 in Germany, are also clearly reflected here. Western Scandinavia Owing to strong isostatic uplift during the Holocene, sea-level curves from Norway are very steep and possible declines in sea level are hidden by the crustal movements. Further to the south, Mo¨rner (revised 1976) produced a detailed curve with a calendar-year timescale for the Kattegat area, western Sweden. This curve was obtained using data of shore-level displacement from which the estimated amount of isostasy was subtracted. Although Mo¨rner’s methodology differs from that employed here, his curve is similar to that from the southern North Sea and shows clear oscillations. Furthermore, there is good correlation between the declines in the younger part of Mo¨rner’s curve and R 1 to 6. In western Sweden, however, R 3 appears to have been much stronger than has so far been demonstrated in Germany. The agreement between the curves from the southern North Sea and western Sweden, which are based on different methodologies, enhances the validity of the described sequence of transgressions and regressions and extends its spatial application. The southwestern Baltic The Baltic also experienced strong isostatic uplift, particularly in the north. The southern Baltic coast is less affected and there are several curves that cover the period from the Litorina Sea onwards after full connection with the North Sea was established. Correlations between the Baltic and the more general events in the North Sea are difficult. Klug (1980) published a curve for the southwestern Baltic that starts with a steep rise and continues from 5800 yr BP onwards in stepwise oscillations. Though the database is small, two clear relative falls in sea level can be recognized which correlate with R 2 and R 6. Another curve for West Pomerania (Kliewe & Janke 1982) shows the same trends, but with a much smoother course. Also the curve of Lampe & Janke (2004) indicates sea-level falls which correspond with the strong declines R 3, R 6 and R 7 in the North Sea. These curves are still tentative, but the available data (see also Yu et al. 2004) indicate probable regressions also in the Baltic that may be due to sealevel changes in the AtlanticNorth Sea region.

BOREAS 36 (2007)

Holocene sea-level curve, S North Sea


Other regions

Former shorelines in the North Sea area

Although many sea-level curves have been published for other parts of the world, comparisons are difficult. The time resolution is often poor, particularly for the last 5000 years, such that changes in the curves are often fairly hypothetical. Another difficulty arises from instability due to crustal movements. But even unstable areas can provide qualitative results for the direction and relative degree of sea-level change, as has been shown by, for example, Mo¨rner (1976). Since the discovery of deformation of the geoid relief and variations in the surface levels of the oceans, quantitative long-distance correlations of sea level are unsuccessful. The general picture that comes from the numerous curves compiled by Pirazzoli (1991), however, is a fairly steep postglacial eustatic sea-level rise until about 5000 to 4000 cal. yr BC. This is followed by a deceleration, often connected with oscillations, the number and range of which depends largely on the amount of data available for the particular area.

There are only a few reliable dates of transgressive contacts from the bottom of the North Sea. The reconstruction of former shorelines is therefore based on modern depth contours linked to the sea-level curve. For the oldest period in particular, this means interpolation from a few data points or extrapolation backwards. Nevertheless, this approach is reasonably acceptable, since the rate of sea-level rise is more or less constant at this time. The coastlines, however, must be regarded as hypothetical, especially because the isostatic factor cannot be estimated with certainty. Jelgersma (1961, 1979) was one of the first authors to outline North Sea shorelines in this way, while Shennan et al. (2000) presented model predictions of shoreline positions there. In Fig. 6, a new approach is proposed based on the new sea-level curve and additional fixed points. The modern depth contours are used as a basis and dates are given in calibrated

Fig. 6. Postulated North Sea shorelines between 12 000 and 6000 cal. yr BC.


Karl-Ernst Behre

BOREAS 36 (2007)

Fig. 7. Simplified sea-level curve for the southern North Sea. The time scale is in calibrated/calendar years; subdivisions of the Calais/Dunkirk system and pollen-analytical-based subdivisions are also indicated. Roman numbers/transgressions; Arabic numbers/regressions.

years BC, as in the new sea-level curve. In contrast to this MHW curve, however, the shorelines show the mean water level; /7200 years represents the level of /36 m, 6800 that of /30 m and /6000 that of

/20 m. For the time before 8000 cal. yr BC the model of Shennan et al. (2000; 9600 cal. yr BC, /65 m) was used, and the hypothetical shoreline of Jelgersma (1979; 12 000 cal. yr BC, representing

BOREAS 36 (2007)

/90 m) was added. The oldest shoreline is the most speculative, but with decreasing age the certainty of the reconstructions increases. Archaeological studies of Mesolithic populations in the North Sea area indicate land bridges as well as open water bodies within the region (cf. Schwabedissen 1951; Louwe Kooijmans 1971). Recently, Coles (1999) published speculative maps of a so-called Doggerland in the Stone Age. The most important morphological developments in the history of the North Sea were the isolation of the Dogger Bank and the establishment of a connection between the North Sea and the English Channel. After 9000 cal. yr BC the North Sea inundated a ridge in the west and extended around the southern part of the Dogger Bank. The Dogger Bank became an island when, more than a thousand years later, the sea invaded the palaeo valley of the river Elbe in the east and joined the waters coming from the west. An important artefact has been found recently on the Dogger Bank, dating to around 6050 cal. yr BC when the island was already small (Coles 1999). The Dogger Bank disappeared within the following 1000 years. The connection between the main part of the North Sea and the English Channel via the Southern Bight took place around 7000 cal. yr BC. However, full marine conditions in the Southern Bight were reached only after 6000 cal. yr BC (Eisma et al. 1981), the substantial input of fresh water from the rivers Rhine, Meuse and Thames being largely responsible for this delay. From this time onwards, the North Sea was influenced by two tidal waves from the Atlantic, the interference of which led to the formation of a new amphidromic point with a tidal amplitude of zero between Ijmuiden and Great Yarmouth, while the position of the main amphidromic point in the centre of the North Sea shifted until the Dogger Bank was submerged. At about 5500 cal. yr BC the sea reached the present coastal area. In the western part of The Netherlands, an extensive barrier system developed which continued to the North as barrier islands, the size of which decreases to the east while their mobility increases. Behind these islands in the northern Netherlands and Germany, a tidal flat area was formed representing an open coast which directly reacted to sea-level changes. From the river Elbe to the north this succession is more complicated because of large medieval land losses; island size again increases and in the north the outer coast consists of Pleistocene deposits with beach barriers. For Germany, maps giving detailed outlines of the coast for the Birth of Christ, AD 800 and AD 1500 have recently been published by Behre (1999, 2003).

Summary and conclusions A new sea-level curve is presented for the southern North Sea which deviates in several respects from previous

Holocene sea-level curve, S North Sea


curves. The main difference is the clear record of seven sea-level falls (R 1R 7) during the younger Holocene. . Out of several hundred available data sets, 118 points were selected to construct the new sea-level curve. For each of these points the exact position of the MHW was estimated. In order to use the data that were collected from a wide area with different tidal ranges for a single curve of the MHW, all altitudinal data were corrected to a standard tide gauge for which Wilhelmshaven was chosen. Here the MHW is at present 1.70 m above NN (/mean sea level). Therefore, to obtain a MSL curve, 1.70 m must be subtracted from the values presented here. . The curve shows a steep sea-level rise until about 5000 cal. years BC, and then slows down so that it becomes closer to the horizontal with several oscillations. This is roughly in accordance with the available evidence from Britain as well as from southwestern Scandinavia. . An important fact about the area around the German Bight is the existence of an open coast, where former sea-level changes had immediate consequences for the environment and sediment facies and therefore can be dated exactly. There are no beach barrier systems mitigating or masking the effects of sea-level changes in the hinterland. . Intercalated peats, formed between marine or brackish deposits, have been used here as an important source for indicating sea-level changes. Their formation requires an environmental change from salt to freshwater conditions. Several of these peats developed synchronously across a large area and their bases provided reliable dates for sea-level declines, while their tops dated the following transgressions. Their absolute height, however, was often affected by compaction, but at least minimum sea-level heights could be estimated from such profiles. . Many archaeological dates were used for the younger part of the sea-level curve. The most important ones came from Flachsiedlungen, i.e. settlements in the flat Clay District that were established on nonelevated ground. The synchronous and widespread occupation of former salt marshes gave strong indications of regressions. In some areas this is supported by contemporary well-developed soils. Evidence of former storm-flood heights was provided by dwelling mounds or Wurten. The altitude of the lowermost building in a settlement layer indicates the former storm flood level. To use these archaeological data, the difference between MHW and storm flood level in the pre-diking period was estimated to have been only 1 m. . The Calais Dunkirk system, which provides the only terminology that is applicable throughout the southern North Sea region, has been employed here with some modifications, in particular the omission of Dunkirk 0 in favour of Regression 2. It is used as


Karl-Ernst Behre

a chronostratigraphic system and apart from the transgressions it includes the newly defined regressions. . In the study area, tectonic movements are estimated as less than 1 cm/century subsidence and the isostatic depression is even less. This relative stability means that the region is most suitable for the construction of a sea-level curve. Acknowledgements.  I have to thank all authors who produced the data evaluated in this contribution. Special thanks go to M. Spohr, who prepared the figures, and to M. O’Connell, who improved the English. I sincerely appreciate the help of C. Murray-Wallace and an anonymous referee and, in particular, J. A. Piotrowski for significantly improving the original text.

References Augath, W. 1993: Stand und Weiterentwicklung der Ho¨henu¨berwachung der niedersa¨chsischen Nordseeku¨ste. Nachrichten aus Niedersachsens Vermessungs- und Katasterverwaltung 43, 78  92. Austin, R. M. 1991: Modelling Holocene tides on the NW European continental shelf. Terra Nova 3, 276 288. Baeteman, C. 1981: An alternative classification and profile type map applied to the Holocene deposits of the Belgian coastal plain. Bulletin Socie´te´ Belge de Ge´ologie 90, 257 280. Baeteman, C. 1999: The Holocene depositional history of the Ijzer palaeo-valley (western Belgian coastal plain) with reference to the factors controlling the formation of intercalated peat beds. Geologica Belgica 2, 39 72. Bakker, J. A., Brandt, R. W., Geel, B. van, Jansma, M. J., Kuijper, W. J., Mensch, P. J. A. van, Pals, J. P. & Ijzereef, G. F. 1977: Hoogkarspel Watertoren. Towards a reconstruction of ecology and archaeology of an agrarian settlement of 1000 BC. In Beek, B. L. van, Brandt, R. W. & Groenman-van Waateringe, W. (eds.): Ex horreo. Cingula 4, 187 225. Bantelmann, A. 1960: Die kaiserzeitliche Marschensiedlung von Ostermoor bei Brunsbu¨ttelkoog. Offa 16, 53 79. Bantelmann, A. 1966: Die Landschaftsentwicklung an der schleswigholsteinischen Westku¨ste, dargestellt am Beispiel Nordfriesland. Die Ku ¨ ste 14, 5 99. Bantelmann, A. 1975: Die fru¨hgeschichtliche Marschensiedlung beim Elisenhof in Eiderstedt. Landschaftsgeschichte und Baubefunde. Studien zur Ku ¨ stenarcha ¨ ologie Schleswig-Holsteins, Series A, 1, 190 pp. Bantelmann, A., Hoffmann, D. & Menke, B. 1984: Vera¨nderungen des Ku¨stenverlaufs: Schleswig-Holstein. In Kossack, G., Behre, K.-E. & Schmid, P. (eds.): Archa ¨ ologische und naturwissenschaftliche Untersuchungen an la ¨ ndlichen und sta ¨ dtischen Siedlungen im deutschen Ku ¨ stengebiet vom 5. Jh. v. Chr. bis zum 11. Jh. n. Chr., Band 1, 54 68. Barckhausen, J. 1969: Entstehung und Entwicklung der Insel Langeoog. Oldenburger Jahrbuch 68, 239 281. Barckhausen, J. 1970: Geologische Karte von Niedersachsen 1:25000, Blatt 2210 Baltrum, Blatt 2211 Ostende-Langeoog. NLfB Hannover. Barckhausen, J. 1984: Geologische Karte von Niedersachsen 1:25000, Blatt 2609 Emden. NLfB Hannover. Barckhausen, J. 1995: Geologische Karte von Niedersachsen 1:25000, Blatt 2616 Brake. NLfB Hannover. Barckhausen, J. & Streif, H. 1978: Geologische Karte von Niedersachsen 1:25000, Blatt 2608 Emden West. NLfB Hannover. Behre, K.-E. 1970: Die Entwicklungsgeschichte der natu¨rlichen Vegetation im Gebiet der unteren Ems und ihre Abha¨ngigkeit

BOREAS 36 (2007)

von den Bewegungen des Meeresspiegels. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 9, 13 48. Behre, K.-E. 1985: Die urspru¨ngliche Vegetation in den deutschen Marschgebieten und deren Vera¨nderung durch pra¨historische Besiedlung und Meeresspiegelbewegungen. Verhandlungen der ¨ kologie 13, 85 96. Gesellschaft fu ¨r O Behre, K.-E. 1986: Meeresspiegelverhalten und Besiedlung wa¨hrend der Zeit um Christi Geburt in den Nordseemarschen. Offa 43, 45  53. Behre, K.-E. 1999: Die Vera¨nderungen der niedersa¨chsischen Ku¨stenlinien in den letzten 3000 Jahren und deren Ursachen. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 26, 9 34. Behre, K.-E. 2002: Landscape development and occupation history along the southern North Sea coast. In Wefer, G., Berger, W., Behre, K.-E. & Jansen, E. (eds.): Climate Development and History of the North Atlantic Realm, 299 312. Springer, Berlin and Heidelberg. Behre, K.-E. 2003: Eine neue Meeresspiegelkurve fu¨r die su¨dliche Nordsee. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 28, 9 63. Behre, K.-E. 2004: Coastal development, sea-level change and settlement history during the later Holocene in the Clay District of Lower Saxony (Niedersachsen), northern Germany. Quaternary International 112, 37 53. Behre, K.-E. 2005: Meeresspiegelanstieg  Marschenentwicklung  Ku¨stenlinien. Die letzten 10.000 Jahre an der deutschen Nordseeku¨ste im Zeitraffer. Schriftenreihe des Landesmuseums Natur und Mensch, Oldenburg 33, 25 36. Behre, K.-E., Barckhausen, J., Brandt, K. & Streif, H. 1975: East Frisia/Germany. Excursion Guide INQUA Holocene Commission and Subcommission Shorelines of NW Europe. Wilhelmshaven, Hannover. Behre, K.-E., Do¨rjes, J. & Irion, G. 1984: Ein datierter Sedimentkern aus dem Holoza¨n der su¨dlichen Nordsee. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 15, 135 148. Behre, K.-E. & Kucˇan, D. 1999: Neue Untersuchungen am Außendeichsmoor bei Sehestedt am Jadebusen. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 26, 35 64. Behre, K.-E. & Menke, B. 1969: Pollenanalytische Untersuchungen an einem Bohrkern der su¨dlichen Doggerbank. Beitra ¨ ge zur Meereskunde 24/25, 122 129. Behre, K.-E., Menke, B. & Streif, H. 1979: The Quaternary geological development of the German part of the North Sea. In Oele, E., Schu¨ttenhelm, R. T. E. & Wiggers, A. J. (eds.): The Quaternary History of the North Sea. Acta Universitatis Upsaliensis, Symposia Universitatis Upsaliensis Annum Quingentesimum Celebrantis 2, 85 113. Boersma, J. 2005: Colonists on the clay. The occupation of the northern coastal region. In Louwe Kooijmans et al. (eds.): The Prehistory of the Netherlands 2, 561 576. Brand, G., Hageman, B. P., Jelgersma, S. & Sindowski, K. H. 1965: Die lithostratigraphische Unterteilung des marinen Holoza¨ns an der Nordseeku¨ste. Geologisches Jahrbuch 82, 365 384. Brandt, K. 1980: Die Ho¨henlage ur- und fru¨hgeschichtlicher Wohnniveaus in nordwestdeutschen Marschengebieten als Ho¨henmarken ehemaliger Wassersta¨nde. Eiszeitalter und Gegenwart 30, 161 170. Brandt, K. 1991: Die Ergebnisse in den Grabungen der Wurten Niens und Sievertsborch (Kreis Wesermarsch). Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 18, 89 140. Coles, B. 1999: Doggerland’s loss and the Neolithic. WARP Occasional Paper 12, 21 57. Donner, J. 1995: The Quaternary History of Scandinavia. 200 pp. Cambridge University Press, Cambridge. Doornhof, D. 1992: Surface subsidence in The Netherlands. The Groningen gas field. Geologie en Mijnbouw 71, 119 130. Dubois, G. 1924: Recherches sur les terrains quaternaires du nord de la France. Me´moires de la Socie´te´ Ge´ologique du Nord 8, 355 pp.

BOREAS 36 (2007)

Eisma, D., Mook, W. G. & Laban, C. 1981: An early Holocene tidal flat in the Southern Bight. In Nio, S.-D., Schu¨ttenhelm, R. T. E. & Weering, T. C. E. van (eds.): Holocene Marine Sedimentation in the North Sea Basin. International Association of Sedimentologists, Special Publication 5, 229 237. Ervynck, A., Baeteman, C., Demiddele, H., Hollevoet, Y., Pieters, M., Schelvis, J., Tys, D., Strydonck, M. van & Verhaeghe, F. 1999: Human occupation because of a regression, or the cause of a transgression? A critical review of the interaction between geological events and human occupation in the Belgian coastal plain during the first millennium AD. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 26, 97 121. Ey, J. 1995: Die mittelalterliche Wurt Neuwarfen, Gde. Wangerland, Ldkr. Friesland. Die Ergebnisse der Grabungen 1991 und 1992. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 23, 265  315. Freund, H. & Streif, H. 2000: Natu¨rliche Pegelmarken fu¨r Meeresspiegelschwankungen der letzten 2000 Jahre im Bereich der Insel Juist. Petermanns Geographische Mitteilungen 143, 34 45. Gezeitenkalender 2002: Hoch- und Niedrigwasserzeiten fu ¨ r die Deutsche Bucht und deren Flussgebiete. 127 pp. Bundesamt fu¨r Seeschiffahrt und Hydrographie, Hamburg. Griede, J. W. 1978: Het ontstaan van Frieslands Noordhoek. Ph.D. dissertation, Universiteit Amsterdam, 186 pp. Gronwald, W. 1960: Welche Erkenntnisse zur Frage der vermuteten neuzeitlichen Nordseeku¨stensenkung hat die Wiederholung des deutschen Nordseeku¨stennivellements gebracht? Die Ku ¨ ste 8, 66 82. Haarnagel, W. 1957: Die spa¨tbronze-, fru¨heisenzeitliche Geho¨ftsiedlung Jemgum bei Leer auf dem linken Ufer der Ems. Die Kunde Neue Folge 8, 2 44. Haarnagel, W. 1969: Die Ergebnisse der Grabung auf der a¨ltereisenzeitlichen Siedlung Boomborg/Hatzum, Kr. Leer, in den Jahren von 1965 bis 1967. Neue Ausgrabungen und Forschungen in Niedersachsen 4, 58 97. Haarnagel, W. 1979: Die Grabung Feddersen Wierde. Methode, Hausbau, Siedlungs- und Wirtschaftsform sowie Sozialstruktur. Feddersen Wierde 2. 364 pp. Steiner, Wiesbaden. Haarnagel, W. 1980: Die Besiedlung im nordwestdeutschen Ku¨stengebiet in ihrer Abha¨ngigkeit von Meeresspiegelschwankungen und Sturmfluten. In Verhulst, A. & Gottschalk, M. K. E. (eds.): Transgression en occupatiegeschiedenis in de kustgebieden van Nederland en Belgie, 209 239. Belgisch Centrum voor Landelijke Geschiedenis, Gent. Hageman, B. P. 1969: Development of the western part of the Netherlands during the Holocene. Geologie en Mijnbouw 48, 373  388. Hanisch, J. 1980: Neue Meeresspiegeldaten aus dem Raum Wangerooge. Eiszeitalter und Gegenwart 30, 221 228. Higelke, B. 1988: Topographie, Morphodynamik und Hydrographie der su¨dlichen nordfriesischen Watten. Offa-Bu ¨ cher Neue Folge 66, 39 50. Higelke, B., Ku¨hn, H.-J. & Mu¨ller-Wille, M. 1984: Siedelra¨ume: Schleswig-Holstein. In Kossack, G., Behre, K.-E. & Schmid, P. (eds.): Archa ¨ ologische und naturwissenschaftliche Untersuchungen an la ¨ ndlichen und sta ¨ dtischen Siedlungen im deutschen Ku ¨ stengebiet vom 5. Jh. v. Chr. bis zum 11. Jh. n. Chr. 1, La ¨ ndliche Siedlungen, 132 142. Verlag Chemie, Weinheim. Hoffmann, D. 1980: Ku¨stenholoza¨n zwischen Sylt und Fo¨hr. Ro ¨ misch-Germanische Forschungen 39, 85 130. Hogestijn, J. W. H. 1997: Enkele resultaten van het archeologische onderzoek op twee woonplaatsen van de Enkelgrafcultuur bij Winkel in de Groetpolder. In Hallewas, D. P., Scheepstra, G. H. & Woltering, P. J. (eds.): Dynamisch Landschap, 27 44. van Gorcum, Amersfoort. Horton, B. P. & Edwards, R. J. 2005: The application of local and regional transfer functions to the reconstruction of Holocene sea levels, north Norfolk, England. The Holocene 15, 216 228.

Holocene sea-level curve, S North Sea


Ijzereef, G. F. & van Regteren Altena, J. F. 1991: Nederzettingen uit de midden- en late bronstijd bij Andijk en Bovenkarspel. Nederlandse Archeologische Rapporten 13, 61 81. Jelgersma, S. 1961: Holocene sea level changes in the Netherlands. Mededelingen van de Geologische Stichting C 6:7, 1 100. Jelgersma, S. 1979: Sea-level changes in the North Sea basin. Acta Universitatis Upsaliensis, Symposia Universitatis Upsaliensis Annum Quingentesimum Celebrantis 2, 233 248. Jelgersma, S., Oele, E. & Wiggers, A. J. 1979: Depositional history and coastal development in the Netherlands and the adjacent North Sea since the Eemian. Acta Universitatis Upsaliensis, Symposia Universitatis Upsaliensis Annum Quingentesimum Celebrantis 2, 115 142. Jong, J. D. de 1971: The scenery of the Netherlands against the background of Holocene geology. Revue de Ge´ographie Physique et Ge´ologie Dynamique 2, 143 162. Kiden, P., Denys, L. & Johnston, P. 2002: Late Quaternary sea-level change and isostatic and tectonic land movements along the Belgian Dutch North Sea coast: geological data and model results. Journal of Quaternary Science 17, 535 546. Kidson, C. 1977: The coast of south west England. In Kidson, C. & Tooley, M. J. (eds.): The Quaternary History of the Irish Sea, 257  298. Seel House Press, Liverpool. Kliewe, H. & Janke, W. 1982: Der holoza¨ne Wasserspiegelanstieg der Ostsee im nordo¨stlichen Ku¨stengebiet der DDR. Petermanns Geographische Mitteilungen 2/82, 65 74. Klug, H. 1980: Der Anstieg des Ostseespiegels im deutschen Ku¨stenraum seit dem Mittelatlantikum. Eiszeitalter und Gegenwart 30, 237 252. Ko¨rber-Grohne, U. 1967: Geobotanische Untersuchungen auf der Feddersen Wierde. Feddersen Wierde 1. Steiner, Wiesbaden, 358 pp. Konradi, P. B. 2000: Biostratigraphy and environment of the Holocene marine transgression in the Heligoland Channel, North Sea. Bulletin of the Geological Society of Denmark 47, 71 79. Lampe, R. & Janke, W. 2004: The Holocene sea level rise in the southern Baltic as reflected in coastal peat sequences. Polish Geological Institute Special Papers 11, 19 30. Linke, G. 1979: Ergebnisse geologischer Untersuchungen im Ku¨stenbereich su¨dlich Cuxhaven. Ein Beitrag zur Diskussion holoza¨ner Fragen. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 13, 39 83. Linke, G. 1982: Der Ablauf der holoza¨nen Transgression der Nordsee aufgrund von Ergebnissen aus dem Gebiet Neuwerk/Scharho¨rn. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 14, 123  157. Louwe Kooijmans, L. P. 1971: Mesolithic bone and antler implements from the North Sea and from the Netherlands. Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 20/21, 27 73. Louwe Kooijmans, L. P. 1974: The Rhine/Meuse Delta. Four Studies on its Prehistoric Ooccupation and Holocene Geology. Ph.D. dissertation, Universiteit Leiden, 421 pp. Louwe Kooijmans, L. P. 1976: Pra¨historische Besiedlung im RheinMaas-Deltagebiet und die Bestimmung ehemaliger Wasserho¨hen. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 11, 119  143. Louwe Kooijmans, L. P. 1985: Sporen in het land. De Nederlandse delta in de prehistorie. 160 pp. Meulenhoff, Amsterdam. Ludwig, G., Mu¨ller, H. & Streif, H. 1979: Neuere Daten zum holoza¨nen Meeresspiegelanstieg im Bereich der Deutschen Bucht. Geologisches Jahrbuch D 32, 3 22. Ludwig, G., Mu¨ller, H. & Streif, H. 1981: New dates on Holocene sea-level changes in the German Bight. International Association of Sedimentologists, Special Publications 5, 211 219. Meier, D. 2001a: Landschaftsentwicklung und Siedlungsgeschichte des Eiderstedter und Dithmarscher Ku¨stengebietes als Teilregionen des Nordseeku¨stenraumes. Universita ¨ tsforschungen zur pra ¨ historischen Archa ¨ ologie 79, 175 pp.


Karl-Ernst Behre

Meier, D. 2001b: Su¨derbusenwurth  eine Wurt der ro¨mischen Kaiserzeit. Die Heimat 108, 38 46. Menke, B. 1988: Die holoza¨ne Nordseetransgression im Ku¨stenbereich der su¨do¨stlichen Deutschen Bucht. Offa-Bu ¨ cher Neue Folge 66, 117 137. Mo¨rner, N. A. 1976: Eustatic changes during the last 8000 years in view of radiocarbon calibration and new information from the Kattegatt region and other northwestern European coastal areas. Palaeogeography, Palaeoclimatology, Palaeoecology 19, 63 85. Nilsson, T. 1948: Versuch einer Anknu¨pfung der postglazialen Entwicklung des nordwestdeutschen und niederla¨ndischen Flachlandes an die pollenfloristische Zonengliederung Su¨dskandinaviens. Meddelingen Lunds Geologisch-Mineralogisch Institut 112, 1 79. Pirazzoli, P. A. 1991: World Atlas of Holocene Sea-Level Changes. 300 pp. Elsevier, Amsterdam. Plassche, O. van de, 1982: Sea-level change and water-level movements in the Netherlands during the Holocene. Mededelingen Rijks Geologische Dienst 36, 93 pp. Plassche, O. van de, 1985: Time-limit assessment of some Holocene transgressive and regressive periods in the Northern Netherlands. Eiszeitalter und Gegenwart 35, 43 48. Preuss, H. 1979: Die holoza¨ne Entwicklung der Nordseeku¨ste im Gebiet der o¨stlichen Wesermarsch. Geologisches Jahrbuch A 53, 89 pp. Reinhardt, W. 1965: Studien zur Entwicklung des la¨ndlichen Siedlungsbildes in den Seemarschen der ostfriesischen Westku¨ste. Probleme der Ku ¨ stenforschung im su ¨ dlichen Nordseegebiet 8, 73 148. Roeleveld, W. 1974: The Holocene Evolution of the Groningen Marine-Clay District. Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek, Supplementum to 24, 132 pp. Rohde, H. 1975: Wasserstandsbeobachtungen im Bereich der deutschen Nordseeku¨ste vor der Mitte des 19. Jahrhunderts. Die Ku ¨ ste 28, 1 96. Schmid, P. 1988: Die mittelalterliche Neubesiedlung der niedersa¨chsischen Marsch. Archeologie en landschap. Festschrift H. T. Waterbolk, 133 164. Publisher Universiteit Groningen. Schmid, P. 1994: Oldorf  eine fru¨hmittelalterliche friesische Wurtsiedlung. Germania 72, 231 267. Schoute, J. F. Th. 1984: Vegetation horizons and related phenomena. Dissertationes Botanicae 81, 270 pp. Schu¨tte, H. 1939: Sinkendes Land an der Nordsee? Zur Ku¨stengeschichte Nordwestdeutschlands. Schriften des Deutschen Naturkundevereins Neue Folge 9, 144 pp. Schwabedissen, H. 1951: Zur Besiedlung des Nordseeraumes in der a¨lteren und mittleren Steinzeit. Festschrift G. Schwantes, 59 77. Wachholtz, Neumu¨nster. Shennan, I. 1986: Flandrian sea-level changes in the Fenland. II: Tendencies of sea-level movement, altitudinal changes, and local and regional factors. Journal of Quaternary Science 1, 155 179. Shennan, I. 1989: Holocene crustal movements and sea-level changes in Great Britain. Journal of Quaternary Science 4, 77 89. Shennan, I. 1994: Coastal evolution. In Waller, M. (ed.): The Fenland Project Number 9: Flandrian Environmental Change in Fenland. East Anglian Archaeology 70, 47 84. Shennan, I., Lambeck, K., Flather, R., Horton, B., McArthur, J., Innes, J., Lloyd, J., Rutherford, M. & Wingfield, R. 2000: Modelling western North Sea palaeogeographies and tidal changes during the Holocene. Geological Society London, Special Publications 166, 299 319. Siefert, W. & Lassen, H. 1985: Gesamtdarstellung der Wasserstandsverha¨ltnisse im Ku¨stenvorfeld der Deutschen Bucht nach neuen Pegelauswertungen. Die Ku ¨ ste 42, 1 77. Sindowski, K.-H. 1957: Die geologische Entwicklung des Wattengebietes su¨dlich der Inseln Baltrum und Langeoog. Jahresbericht der Forschungsstelle Norderney 8, 1956, 11 36. Sindowski, K.-H. 1969a: Geologische Karte von Niedersachsen 1:25000, Blatt 2213 Wangerooge, Hannover.

BOREAS 36 (2007)

Sindowski, K.-H. 1969b: Geologische Karte von Niedersachsen 1:25000, Blatt 2118 Cuxhaven, Hannover. Strahl, E. 2002a: Erste Bauern in der deutschen Marsch  Die jungbronzezeitliche Siedlung Rodenkirchen-Hahnenknooper Mu¨hle, Ldkr. Wesermarsch. Berichte zur Denkmalpflege in Niedersachsen 22, 79 82. Strahl, E. 2002b: Erste Bauern in der Marsch  Die Siedlung Hahnenknooper Mu¨hle bei Rodenkirchen, Ldkr. Wesermarsch. Wilhelmshavener Tage 8, 76 83. Strahl, E. 2005: Die jungbronzezeitliche Siedlung RodenkirchenHahnenknooper Mu¨hle, Ldkr. Wesermarsch  Erste Bauern in der deutschen Marsch. Schriftenreihe des Landesmuseums fu ¨ r Natur und Mensch Oldenburg 33, 52 59. Streif, H. 1981: Geologische Karte von Niedersachsen 1:25000, Blatt 2414 Wilhelmshaven, NLfB Hannover. Streif, H. 1984: The coastal Holocene on the eastern side of Jade Bay. INQUA-Excursion Guide, Subcommission on Shorelines of NW Europe, Field Conference 1984, 13 18. Streif, H. 1985: Geologische Karte von Niedersachsen 1:25 000, Blatt 2314 Hooksiel. NLfB Hannover. Streif, H. 1986: Zur Altersstellung und Entwicklung der Ostfriesischen Inseln. Offa 43, 29 44. Streif, H. 1993: Geologische Karte von Niedersachsen 1:25000, Blatt 2516 Nordenham. NLfB Hannover. Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormack, G., van der Plicht, J. & Spurk, M. 1998: Calibration issue. Radiocarbon 40, 1041 1083. Tavernier, R. 1948: De jongste geologische geschiedenis der Vlaamse kustvlakte. Handelingen der Maatschappij voor Geschiedenis en Oudheidkunde te Gent 3, 107 115. Tesch, P. 1930: Einige toelichting bij de Geologische Kaart van Nederland 1:50000. T.K.N.A.G. 47, 692 701. Thoen, H. 1978: De Belgische kustvlakte in de Romeinse tijd. Verhandelingen van de Koninklijke Academie voor Wetenschappen, Letteren en Schone Kunsten van Belgie XL, 88 pp. Thoen, H. (ed.) 1987: De Romeinen langs de Vlaamse kust. 181 pp. Gemeentkrediet, Brussels. Tooley, M. J. 1978: Sea-Level Changes. North-West England during the Flandrian Stage. 232 pp. Clarendon, Oxford. Tooley, M. J. 1982: Sea-level changes in northern England. Proceedings Geological Association 93, 43 51. Vos, P. C. & Heeringen, R. M. van 1997: Holocene geology and occupation history of the province of Zeeland. Mededelingen Nederlands Instituut voor Toegepaste Geowetenschappen (TNO) 59, 5 109. Waller, M. 1994: Geological history. In Waller, M. (ed.): The Fenland Project Number 9: Flandrian Environmental Change in Fenland. East Anglian Archaeology 70, 7 17. Waterbolk, H. T. 1966: The occupation of Friesland in the prehistoric period. Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 15/16, 13 35. Waterbolk, H. T. 1994: Ezinge. In J. Hoops (ed.): Reallexikon der germanischen Altertumskunde 8, 60 67. De Gruyter, Berlin. Wu¨bbelmann, H. 1993: Vergleich zwischen Ho¨hen im DHHN 85 und im Nivellementnetz 1960. In Arbeitsgemeinschaft der Vermessungsverwaltungen der La¨nder der Bundesrepublik Deutschland, Arbeitskreis Ho¨henfestpunktfeld und Schwerefestpunktfeld (ed.): Die Wiederholungsmessungen 1980 und 1985 im deutschen Hauptho ¨hennetz und das Hauptho ¨ hennetz 1985 der Bundesrepublik Deutschland, 155 219. Mu¨nchen. Yu, S.-Y., Berglund, B. E., Andre´n, E. & Sandgren, P. 2004: MidHolocene Baltic Sea transgression along the coast of SE Sweden  ancient lagoons correlated with beach ridges. Geologiska Fo ¨ reningens i Stockholm Fo ¨ rhandlinger 126, 257 272. Zagwijn, W. H. 1986: Nederland in het Holoceen. Geologie van Nederland. 46 pp. RGD, Haarlem. Zagwijn, W. H. & Staalduinen, C. J. van 1975: Toelichting bij geologische overzichtskaarten van Nederland. 134 pp. RGD, Haarlem.