Geological succession and faunas

The geological framework of the subsurface near borehole B11D2148 is captured by the DGM geological model and the REGIS II hydrogeological model (https://www.dinoloket.nl/en/subsurface-models version 2.2). Both nation-wide models are based on a limited amount of borehole data. For this study, a more detailed cross-section has been assembled using an additional set of regional boreholes as available via https://www.dinoloket.nl/en/subsurface-data. The cross-section (Fig. 1C) shows the lithostratigraphic record down to a depth of 70 m. Near the surface Holocene and Late Pleistocene sediments occur. Marine Eemian sediments – stratigraphically defined near Amersfoort in the central Netherlands and closest occurrence near Sneek to the NW (Fig. 1, Busschers et al., Reference Busschers, Wesselingh, Kars, Versluijs-Helder, Wallinga, Bosch, Timmner, Nierop, Meijer, Bunnik and de Wolf2014) – have not been deposited near Luxwoude (Fig. 1A). Below the Late Pleistocene succession, glacial till, fluvioglacial and periglacial intervals are present that were deposited during MIS6 (Drente and Drachten Formation: detailed descriptions and interpretations can be found at www.dinoloket.nl/stratigrafische-nomenclator). Underlying these glacial units a succession of fine gravelly sands with mollusc remains occurs, which has been attributed to the Urk Formation (cf. Bosch, Reference Bosch1990). The Urk Formation contains an upper gravelly coarse sandy interval (19–29 m b.s.) attributed to the Tynje Member and a lower interval (29–59 m b.s.) with a mostly fine sandy top attributed to the Veenhuizen Member (see below). The Urk Formation is predominantly composed of Rhine-derived sediments which are older than MIS6 (Busschers et al., Reference Busschers, Kasse, Van Balen, Vandenberghe, Cohen, Weerst, Wallinga, Johns, Cleveringa and Bunnik2007; Peeters et al., Reference Peeters, Busschers, Stouthamer, Bosch, Van den Berg, Wallinga, Versendaal, Bunnik and Middelkoop2016). The two members are regionally separated by glacial sediments of the Peelo Formation (formed during the Elsterian glaciation); however, in this borehole the Peelo Formation is absent due to later erosion (Fig. 1C). The mollusc faunas, which are the subject of this paper, occur in both units between 19 and 38 m depth.

The Tynje Member interval consists of coarse-grained gravelly sands with molluscs (Fig. 2). The sandy deposits have a pink-brownish appearance and are rich in carbonate. The basal part of this unit (25–29 m) is particularly coarse-grained and rich in shells and shell fragments. The shells concentrate around the level of deepest erosion where the Peelo Formation was entirely removed by the river Rhine and the Tynje Member filled the incision (Fig. 1C). The mollusc fauna found in the Tynje Member is dominated by warm-temperate and temperate marine species, with mixed preservations styles. Five species occur, mostly in good preservation (fine surface detail preserved, little surface wear), that hitherto have been used as Eemian (MIS5e) indicator species in the North Sea Basin, viz. Bittium reticulatum, B. scabrum, B. jadertinum, Polititapes senescens and Acanthocardia paucicostata (Fig. 2; Van Leeuwen et al., Reference Van Leeuwen, Beets, Bosch, Burger, Cleveringa, van Harten, Herngreen, Kruk, Langereis, Meijer, Pouwer and de Wolf2000; Wesselingh et al., Reference Wesselingh, Rijken, van Nieulande and Pouwer2012; Busschers et al., Reference Busschers, Wesselingh, Kars, Versluijs-Helder, Wallinga, Bosch, Timmner, Nierop, Meijer, Bunnik and de Wolf2014; Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). Some incompatible autecological indicators co-occur: B. cf. jadertinum and Nucula cf. nitidosa suggest polyhaline to euhaline conditions, Cerastoderma glaucum euryhaline conditions and Succinaidae indet. is an amphibious species common in freshwater lowland settings. Littorina saxatilis is an intertidal to shallow subtidal species, whereas Varicorbula gibba does not live above the storm wave base. Most likely this interval of the Tynje Member represents local fluvial reworking of marine deposits with faunas that for the most part lived in warm, clear water and shallow marine environments with seagrass and/or seaweed (Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). The marine species found match those of the underlying 30–38 m interval of the Veenhuizen Member (see next section). In the overlying interval (19–25 m depth) abraded to very abraded fragments of some common marine bivalve species occur, most commonly C. edule in medium coarse sand with gravel. The coarse-grained deposits and severe wear indicate reworking in a fluvial setting.

The lower unit of the Urk Formation is referred to as the Veenhuizen Member, present at 29–59 m b.s. (Fig. 1; TNO-GSN, 2021). This unit is composed of fine- to medium-grained sands that locally contain clay pebbles (likely representing thin clay layers disturbed during drilling of the borehole), mica flakes and some glauconite. Sample 30–31 m depth is dominated by Mytilus fragments. In sample interval 33–34 m, a large fragment of a worn littoral Littorina littorea was found. Two slightly worn occurrences of Bittium and Mytilus were found in sample interval 37–38 m (Fig. 2a, f, g). However, the Bittium has almost perfect preservation of fine surface details and periostracum at the body whorl. The fine grained nature of the sediments at this depth, the presence of clay layers and marine molluscs suggest that at least part of Veenhuizen Member was deposited in a marine setting.

Age constraints

The mollusc fauna is located below MIS6 dated sediments of the Drente Formation (i.e. Busschers et al., Reference Busschers, Van Balen, Cohen, Kasse, Weerts, Wallinga and Bunnik2008; Busschers et al., Reference Busschers, Wesselingh, Kars, Versluijs-Helder, Wallinga, Bosch, Timmner, Nierop, Meijer, Bunnik and de Wolf2014; Kars et al., Reference Kars, Busschers and Wallinga2012) and early MIS6 dated sediments of the Drachten Formation (Kars et al., Reference Kars, Busschers and Wallinga2012; Busschers et al., Reference Busschers, Wesselingh, Kars, Versluijs-Helder, Wallinga, Bosch, Timmner, Nierop, Meijer, Bunnik and de Wolf2014). The faunal assemblage is therefore certainly older than MIS5e. Moreover, since we observe (1) a partial match between molluscs from the basal Tynje interval with those found in the underlying Veenhuizen Member, (2) high mollusc concentrations occurring right at the position of deepest erosion at the base of the Tynje Member and (3) no mollusc-bearing sources in the direct vicinity of the borehole occur, we consider it plausible that the molluscs from the interval 19–29 m b.s. of the Tynje Member were reworked from the older Veenhuizen Member.

The shell-bearing sediments of the Veenhuizen Member are stratigraphically positioned below the Peelo Formation (Figs. 1 and 2), which was deposited during MIS 10 (e.g. Zagwijn, Reference Zagwijn and Turner1996; Meijer & Cleveringa, Reference Meijer and Cleveringa2009) or MIS 12 (Gibbard, Reference Gibbard and Preece1995; Lauer & Weiss, Reference Lauer and Weiss2018; see Lee et al., Reference Lee, Busschers and Sejrup2012 for discussion). Consequently, the lusitanian molluscs present in the 29–38 m depth interval must be older than MIS10. Taking into account the interglacial nature of the molluscs, a MIS11 stage or older is therefore most likely. Additionally, the stratigraphic position of this section is similar to sediments of the Noordbergum Interglacial as defined by Zagwijn (Reference Zagwijn and Turner1996) about 25 km north of the borehole site. It implies these molluscs, including the lusitanian faunas in the basal 4 metres of the Tynje Member, also have an MIS11 or older age.

Implications

The mollusc faunas in the Quaternary intervals of the Luxwoude borehole are dominated by polyhaline, warm-temperate (‘lusitanian’ and ‘celtic’) and shallow marine species (cf. Cleveringa et al., Reference Cleveringa, Meijer, van Leeuwen, de Wolf, Pouwer, Lissenberg and Burger2000; Van Leeuwen et al., Reference Van Leeuwen, Beets, Bosch, Burger, Cleveringa, van Harten, Herngreen, Kruk, Langereis, Meijer, Pouwer and de Wolf2000; Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). Five species hitherto used as indicators for the warm MIS5e (Eemian) intervals are now – for the first time – retrieved from Middle Pleistocene intervals in the southern North Sea Basin, viz. Bittium reticulatum, B. scabrum, B. cf. jadertinum, Acanthocardia paucicostatum and Polititapes senescens. The taxonomy of European Bittium species is not well-established and we follow the distinction proposed by Verduin (Reference Verduin1982); van der Linden & Wagner (Reference Van der Linden and Wagner1990) and Wesselingh et al. (Reference Wesselingh, Rijken, van Nieulande and Pouwer2012). Within the Luxwoude samples, only A. paucicostatum is a thermophile species, but within Eemian assemblages the aforementioned species typically are accompanied by other warm-temperate species (e.g., Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). The low number of specimens in the limited sediment volume that was available implies that larger sediment samples almost certainly will provide further lusitanian species.

Middle Pleistocene interglacial marine mollusc faunas have previously been reported at several – generally isolated – localities in the southern North Sea Basin (Meijer, Reference Meijer2003; Meijer & Cleveringa, Reference Meijer and Cleveringa2009; Barlow et al., Reference Barlow, Long, Gehrels, Saher, Scaife, Davies, Penkman, Bridgland, Sparkes, Smart and Taylor2017; Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). These mostly concern very small and poorly preserved faunas with the exception of the recently reported fauna found in borehole Hollum on Ameland, about 50 km north of Luxwoude, that was assigned a MIS7 age on the basis of amino acid racemisation analyses (Meijer et al., Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021; age model cf. Meijer & Cleveringa, Reference Meijer and Cleveringa2009). The earlier reported faunas are considered to be of the Celtic type (similar as Holocene marine faunas) lacking any of the thermophilous species characteristic of lusitanian/Eemian assemblages. The Middle Pleistocene faunas reported in this study point to submarine conditions very similar to the Eemian (i.e. warmer than the present day) and warmer settings than those of MIS1 and MIS7 faunas reported by Meijer et al. (Reference Meijer, Pouwer, Cleveringa, de Wolf, Busschers and Wesselingh2021). Several lusitanian species are well capable of living in the southern North Sea, but these are unable to survive the cold waters of northern Scotland (Meijer & Preece, Reference Meijer and Preece1995). Their access into the North Sea Basin is therefore unlikely to have been through a northern route. Hence, these species are considered to be good indicators for a southern Channel connection into the North Sea, similar to the conditions existing today.

Even though the new finding shows that lusitanian Bittium-Polititapes dominated faunas cannot be considered a unique Eemian indicator assemblage, marine Middle Pleistocene intervals in the southern North Sea Basin are seldom encountered because of their occurrence at much greater depths only retrievable by boreholes. Conversely, MIS5e intervals are more accessible because of their shallow position, and their age can often be well-constrained through direct dating and their occurrence between distinct Saalian and Weichselian glacial and proglacial sediments. This implies that even though the assemblages are not uniquely Eemian, in the vast majority of cases they will be so, but we urge caution with the use of Bittium-dominated faunas as the sole indicator for MIS 5e.