Planktonic foraminifera document palaeoceanographic changes across the middle Cenomanian carbon-isotope excursion MCE 1: new evidence from the UK chalk

Abstract Planktonic foraminifera were studied at Lydden Spout, near Folkestone (southeast England, UK), the reference section of the middle Cenomanian Event 1 (MCE 1) characterized by a prominent double-peak δ13C excursion of 1 ‰ identified in different ocean basins and considered a global event. Biostratigraphic and quantitative analysis of planktonic foraminifera are correlated to the δ13C perturbation, to the positive δ18O shifts identified within MCE 1 and to the occurrence of Boreal macrofossils (the bivalves Chlamys arlesiensis and Oxytoma seminudum, and the belemnite Praectinocamax primus). Variations in abundance and species richness of planktonic foraminifera and the inferred palaeoecological preferences of taxa permit the identification of distinct palaeoenvironmental settings across MCE 1. The stratigraphic interval corresponding to MCE 1 is characterized by the absence of oligotrophic rotaliporids, and by the evolutionary appearance of meso-eutrophic dicarinellids and of Muricohedbergella portsdownensis, a cold-water species that occurs at the same level as the Boreal macrofossils. These observations indicate a palaeoceanographic scenario characterized by reduced stratification of surface waters and absence/disruption of the thermocline in a dominantly eutrophic regime during MCE 1. Evidence provided by planktonic foraminifera, Boreal macrofossils and δ18O records documented for the late Cenomanian Plenus Cold Event (PCE) at Eastbourne (UK) reveal similarities that confirm the periodic inflow of cold Boreal seawater originating in the Norwegian Sea as previously postulated to explain the occurrence of Boreal fauna in the Anglo-Paris Basin. The southerly extension of this water mass may be related to the reorganization of circulation driven by the long eccentricity cycle.


Introduction
This study presents a detailed analysis of the middle Cenomanian planktonic foraminifera from the Lydden Spout section near Folkestone (Kent, UK) ( Fig. 1). This section is particularly important because it records the middle Cenomanian Event 1 (MCE 1), a carbon isotope excursion first described by Paul et al. (1994) in northwest European sections from the Anglo-Paris Basin and Cleveland Basin, that is characterized by a prominent double-peak δ 13 C carb positive excursion of up to 1 ‰ named MCE 1 by Mitchell et al. (1996) in a study of the δ 13 C record at Folkestone, Speeton in the Cleveland Basin (North Yorkshire, UK), and Wünstorf in the Lower Saxony Basin. In these sections, Mitchell et al. (1996) described the lower isotopic peak of the bipartite excursion as MCE 1a and the upper one as MCE 1b, and observed that the early portion above MCE 1b has a relatively flat δ 13 C signature followed by increasing δ 13 C values that at Speeton and Folkestone are interrupted by a small negative δ 13 C excursion that was named MCE 2.
The MCE 1 was proposed as a precursor to the late Cenomanian early Turonian Oceanic Anoxic Event 2 (OAE 2, e.g. Jenkyns, 2010), although of lower amplitude (e.g. Coccioni & Galeotti, 2003;Friedrich et al. 2009;Zheng et al. 2016) and currently much less well understood than OAE 2. The location of the carbon-burial sink responsible for the MCE 1 positive carbon-isotope excursion is poorly known, but   fig. 11) noted the presence of laminated, organic-rich sediments at one level in the Vocontian Basin (southeast France). Laminated black shales corresponding to the δ 13 C positive excursion also occur in the tropical Atlantic Ocean (Demerara Rise, total organic content values from 5 to 18 %, Friedrich et al. 2009).
The MCE 1 and OAE 2 differ in the amplitude and shape of the δ 13 C excursion and in the rates of organic carbon burial. The size of the MCE 1 excursion is smaller than the 2-3 ‰ positive excursion recorded during OAE 2 and is of shorter duration (Jarvis et al. 2006;Joo & Sageman, 2014;Eldrett et al. 2015;Gambacorta et al. 2015). Astronomical age models from the US Western Interior Seaway indicate a duration of c. 0.21 Ma for the MCE 1 compared to the estimation of OAE 2 that varies between 0.52 and 0.92 Ma (Sageman et al. 2006;Eldrett et al. 2015;Batenburg et al. 2016).
The carbon isotope record during the onset of the MCE 1 and OAE 2 suggests comparable triggering mechanisms and climatecarbon cycle feedbacks during the two events, although there is no undeniable evidence for a similar initiation mechanism. Volcanic activity and emplacement of Large Igneous Provinces (LIPs: Caribbean, High Arctic, Madagascar and Ontong-Java) were suggested to be responsible for the initiation of OAE 2. For instance, evidence of volcanism based on the osmium-isotope record suggests that a major pulse of volcanism and associated sea-floor hydrothermal weathering began prior to the onset of the carbon isotope excursion at the base of OAE 2 (e.g. Turgeon & Creaser, 2008;Du Vivier et al. 2014Jenkyns et al. 2017). Neodynium-isotope positive excursion during OAE 2 in the European epicontinental sea (Zheng et al. 2013) and in the tropical Atlantic Ocean Martin et al. 2012) were interpreted to reflect transport of radiogenic neodynium released from LIP volcanism under anoxic conditions. By contrast, no neodynium-isotope increase was observed across the MCE 1 (Jiménez Berrocoso et al. 2010;Zheng et al. 2016), although its record supports reorganization of ocean circulation during the MCE 1 and OAE 2 and high productivity in surface waters, with nutrients probably deriving from deep waters and/or LIPs sources (Zheng et al. 2016). According to Scaife et al. (2017), sedimentary mercury concentration data from the US Western Interior Seaway and North Atlantic Ocean are consistent with an initial magmatic pulse at the time of the MCE 1 followed by a second greater pulse at the onset of OAE 2, possibly related to the emplacement of LIPs in the Pacific Ocean and/or the High Arctic.
Cenomanian times (100.5-93.9 Ma) register one of the bestdocumented episodes of eustatic rise in sea-level in Earth history and the beginning of the Late Cretaceous Thermal Maximum, driving the global expansion of epicontinental seas and the onset of widespread pelagic and hemipelagic carbonate deposition (e.g. Giorgioni et al. 2015). It was also a time when the evolutionary diversification of planktonic foraminifera took place at a fast rate. After the major turnover across the Aptian-Albian boundary interval (Kennedy et al. 2000;Leckie et al. 2002;Huber & Leckie, 2011;Petrizzo et al. 2012Petrizzo et al. , 2013Kennedy et al. 2014), the progressive appearance of new lineages is observed throughout the Albian and Cenomanian.
A marked phase of high evolutionary rates is registered in the Cenomanian within the polyphyletic Rotalipora group (Rotalipora Pseudothamanninella Wonders, 1978;Parathalmanninella Lipson-Benitah, 2008), that includes trochospiral taxa with sutural or umbilical supplementary aperture and a peripheral keel (Wonders, 1980;González-Donoso et al. 2007). Rotaliporids are the major component of the assemblages until the end of the Cenomanian when the last representative (Rotalipora cushmani) disappears. The morphological plasticity and species variation within the Rotalipora lineages (e.g. Robaszynski et al. 1994;Petrizzo & Huber, 2006;González-Donoso et al. 2007; has influenced the planktonic foraminiferal biozonation (e.g. Ando et al. 2010;Petrizzo et al. 2015;Falzoni et al. 2018;Petrizzo & Gilardoni, 2020). Specifically, the classic Tethyan biozonations made by Robaszynski & Caron (1995) and Premoli Silva & Sliter (1995), that were incorporated in the Geological Time Scales (Gradstein et al. 2004(Gradstein et al. , 2012Ogg et al. 2016;Gale et al. 2020, have recently been revised by Petrizzo and Gilardoni (2020) who introduced the Thalmanninella greenhornensis Zone. This is defined as the stratigraphic interval between the lowest occurrence (LO) of T. greenhornensis and the LO of R. cushmani, to overcome the difficulty in identifying the Thalmanninella reicheli Zone of previous authors because of the rarity and absence of the species in many western and eastern Tethyan localities and in the US Western Interior Basin record. The appearance of T. greenhornensis is demonstrated to be an easy identifiable bioevent as the species is characterized by a wider geographic distribution and is often reported to appear in the same stratigraphic level as T. reicheli when the latter species is present (see discussion in Petrizzo & Gilardoni, 2020). Moreover, the middle to late Cenomanian planktonic foraminiferal record is also characterized by the evolutionary appearance of the unkeeled trochospiral and heavily muricate Whiteinella and of the doublekeeled genera Dicarinella and Marginotruncana (Premoli Silva & Sliter, 1995;Fraass et al. 2015;Falzoni et al. 2018;Petrizzo & Gilardoni, 2020).
In this study, the planktonic foraminifera stratigraphic distribution, abundance, species richness and population dynamics are coupled with the inferred palaeoecological preferences of taxa to interpret the palaeoceanographic changes and the trophic features of the surface waters across the MCE 1 positive δ 13 C excursion, which is also characterized by a series of positive δ 18 O shifts indicating episodes of remarkable cooling.

374
MR Petrizzo and AS Gale

2.a. Location and stratigraphy
The sea-cliff at Lydden Spout, 500 m west of Samphire Hoe, on the coast between Folkestone and Dover, Kent ( Fig. 1), provides 25 m of accessible exposure of lower and middle Cenomanian chalks of the Westbury and Zigzag Chalk Formations (Fig. 2). The chalks are rhythmically bedded on a scale of several decimetres, darker marly levels alternating with paler chalky beds, interpreted as precession cycles driven by dilution from the input of clay . The stratigraphy and ammonite zonation of the succession between Dover and Folkestone was described by Kennedy (1969). The detailed stratigraphy of the middle Cenomanian succession at Lydden Spout was illustrated by Gale (1989), who provided a detailed log and identified a succession of distinctive beds which could be correlated across the Anglo-Paris Basin (Gale, 1990. These papers also demonstrated the presence of macrofossil taxa with Boreal affinities at this level, previously recorded from the upper part of the late Cenomanian Plenus Marl Member (Jefferies, 1962(Jefferies, , 1963. These include two species of bivalve, Chlamys arlesiensis and Oxytoma seminudum, and a nektonic belemnite, Praectinocamax. In the middle Cenomanian, Boreal species are concentrated in a lower C. arlesiensis Bed and a higher P. primus Bed ( Fig. 2;  the Cast Bed of Gale, 1989).

2.b. The stable carbon-isotope and oxygen-isotope record
A positive carbon isotope excursion containing two peaks was identified in the lower part of the middle Cenomanian of the Lydden Spout succession by Paul et al. (1994) and was correlated with detailed records of macrofossils and foraminifera. This carbon isotope excursion was subsequently named MCE 1 by Mitchell et al. (1996; see also Jarvis et al. 2006) who described the lower peak of the bipartite excursion as MCE 1a and the upper peak as MCE 1b. The MCE 1a begins in the Cunningtoniceras inerme ammonite Zone (Jarvis et al. 2001(Jarvis et al. , 2006, and is associated with a lowstand systems tract Gale & Kennedy, 2021); the MCE 1b falls in the lowermost part of the ammonite Acanthoceras rhotomagense Zone (Jarvis et al. 2001(Jarvis et al. , 2006 and is associated with a transgressive systems tract Mitchell et al. 1996). The carbon isotope record at Folkestone in the studied stratigraphic interval records the MCE 1 excursion (Fig. 2) similar to that reported by Paul et al. (1994) and Zheng et al. (2016). In this study, bulk-chalk δ 13 C was measured every~5 cm and registered peak values of~2.5 ‰ in the C. arlesiensis Bed at 31.7 m which are followed by a trough of~2.2 ‰ and by a subsequent δ 13 C rise tõ 2.5 ‰ toward the top of the P. primus Bed before decreasing again at~35.2 m (Fig. 2). The stratigraphic position of the double-peak δ 13 C positive excursion is in agreement with that reported by Paul et al. (1994), although δ 13 C values are slightly different between the two studies. Here we define the MCE 1 event as the positive carbon excursion extending from the rise in values at the base of the C. arlesiensis Bed (31.5 m), to the fall in values at 36 m (Fig. 2).
The δ 18 O bulk carbonate record at Folkestone documents cyclic fluctuations within MCE 1 related to individual couplets with heavier values in marls and lighter values in chalks (Fig. 2;Paul et al. 1994;Voigt et al. 2004) that indicate a primary temperature change in sea-surface waters (Ditchfield & Marshall, 1989). The interval from 31.5 to 34 m registers the heaviest values, with a positive peak of 1.68 ‰ in the C. arlesiensis Bed at 31.75 m followed by subsequent positive peaks of 1.95 ‰ at 32.55 m, 2.0 ‰ at 33.3 m, and 2.07 ‰ at 33.85 m; the latter peak is registered in the P. primus  Table S1). The observed δ 18 O positive peaks in the C. arlesiensis Bed and in the P. primus Bed were also previously documented by measurements on brachiopod shells collected from Lydden Spout (Voigt et al. 2004;Gale & Kennedy, 2021).

2.c. Sequence stratigraphy
There are marked changes in sea-level during the MCE 1 in the Anglo-Paris Basin, interpreted from sequence-stratigraphical analysis (Robaszynski et al. 1998) as representing the boundary between their Cenomanian sequences 3 and 4 (see summary in Gale & Kennedy, 2021). The sequence boundary at the top of sequence Ce3 has been identified within the upper Mantelliceras dixoni Zone, and over the London Platform, to the north, this surface erodes down into older Cenomanian chalks. Subsequent sealevel rise, commencing at the time of deposition of the C. arlesiensis Bed and Cast Bed, resulted in onlap of the condensed Totternhoe Stone (Acanthoceras rhotomagense Zone) over lower Cenomanian chalks across the Transitional Province of southern England (Gale & Kennedy, 2021). The Lydden Spout succession is uncondensed and was deposited in shallow waters (<200 m) in a pelagic epicontinental environment at the junction of the Boreal Sea, Tethys and proto-North Atlantic. The Ce3-Ce4 sequence boundary has been identified very widely, from the US Western Interior , Texas (Eldrett et al. 2015), Germany (Wilmsen, 2003(Wilmsen, , 2007, north Africa (Kuhnt et al. 2009), Ukraine and Kazakhstan ) and southeast India (Gale et al. 2019b). The events involved sea-level fall and subsequent rise of c. 30 m over a relatively short timespan (200 kyr) based on observation of precession cycles .
A sustained cooling phase that may even have resulted in ephemeral glaciations in Antarctica was also suggested for the MCE 1 based on a bulk carbonate δ 18 O increase in combination with sequence-stratigraphic evidence for a high-amplitude sealevel fall from the Contessa Quarry (Gubbio, Umbria-Marche Basin, Italy) and the New Jersey Coastal Plain (Stoll & Schrag, 2000;Miller et al. 2003Miller et al. , 2005. Significant cooling pulses during the MCE 1 were determined from brachiopod shells from the mid-latitudinal shelf of Europe, including Folkestone, by Voigt et al. (2004) that, to the contrary, related it to major reorganization of ocean circulation in the North Atlantic that probably favoured increased oceanic ventilation and upwelling of nutrient-rich waters. Subsequently, the middle Cenomanian glaciation hypothesis was challenged by isotope records of glassy planktonic and benthic foraminifera that did not indicate a measurable ice volume signal in δ 18 O records across the MCE 1 (Moriya et al. 2007;Ando et al. 2009).

2.d. Planktonic foraminifera
The Lydden Spout section was sampled for planktonic foraminiferal analyses using a sampling resolution of c. 50 cm from 26.70 to 42.75 m (Fig. 2). A total of 32 samples were processed following the standard procedure using hydrogen peroxide to obtain washed residues for micropaleontological analyses. Biostratigraphic and quantitative analyses were carried out for all the washed residues of the >125 μm size-fraction. In addition, the 125-38 μm size-fraction was scanned for marker species. Abundance counts of planktonic vs benthic foraminifera were made by microsplitting samples to obtain aliquots of equal volume for all samples, from which the foraminifera were counted (Supplementary Material Table S2).   (Gale & Kennedy, 2021), lithostratigraphy (Gale, 1989) and ammonite biostratigraphy (Wright et al. 2017). The carbon-isotope and oxygen-isotope curves are based on 5 cm spaced samples, analysed at Cambridge University in 1997. The data show the double carbon-isotope positive excursion of the MCE 1. The Chlamys arlesiensis Bed, Praectinocamax primus Bed and Cast Bed are according to . Foraminiferal abundances, planktonic foraminiferal species distribution and species richness and per cent planktonic foraminifera (% P) according to this study. Planktonic foraminifera zonation after Petrizzo & Gilardoni (2020). To the right are records of important macrofaunal elements, including benthic macrofossils with Boreal affinities, ammonites and an inoceramid bivalve (Gale & Kennedy, 2021). Light orange band = middle Cenomanian Event 1 (MCE 1); light blue rectangles = δ 18 O positive peaks and occurrence of Boreal fauna; PF = planktonic foraminifera.

Composition and distribution of the planktonic foraminiferal assemblage
The foraminiferal population in the size-fraction >125 μm is composed of planktonic and benthic foraminifera that are almost equally abundant, although the proportion of planktonic foraminifera as a percentage of the total foraminiferal assemblage (% P) shows values between 30 and 50 % in one-third of the samples and only in eight samples are planktonic foraminifera less than 30 % (% P: Fig. 2; Supplementary Material Table S2). However, planktonic foraminifera are always rare in the >250 μm size-fractions and they are totally absent at the base of the section from 26.70 to 29.30 m and within the upper part of the MCE 1 interval from 34.20 to 35.70 m (Fig. 2). An increase in total abundance of both benthic and planktonic foraminifera is observed from above 38 m, and the latter group exceeds 50 % of the foraminiferal assemblage in almost all samples ( Fig. 2; Supplementary Material Table S2).
The planktonic foraminiferal assemblage shows poor to moderate preservation that improves toward the top of the studied section. Diversity ranges from 6 species near the base to 15 species at the top except for a minimum value of 5 species recorded in the upper part of the MCE 1 interval above the P. primus Bed ( Fig. 2; Supplementary Material Table S2).
The most common and abundant species that consistently occur throughout the section are Muricohedbergella delrioensis, Pseudoclavihedbergella simplicissima, Praeglobotruncana stephani, Praeglobotruncana delrioensis, Praeglobotruncana cf. compressa and Whiteinella brittonensis (Table 1). Keeled rotaliporids show a scattered occurrence, are limited to the lower and the upper part of the studied section and are absent in the stratigraphic interval from 32.80 to 35.15 m within the MCE 1 interval (from the peak values of~2.5 ‰ in the C. arlesiensis Bed to the second peak value of~2.5 ‰ above the P. primus Bed). Thalmanninella reicheli and Thalmanninella micheli are only observed at 30.20 and 30.80 m, respectively. One specimen of Thalmanninella deckeei was present at 31.75 m (

Planktonic foraminifera biozonation
According to the composition of the assemblages and the position of the marker events, the stratigraphic interval from the base of the studied section at 26.70 m to 38.80 m is assigned to the Thalmanninella greenhornensis Zone, whereas the interval above is assigned to the Rotalipora cushmani Zone (Fig. 2). No specimens of T. greenhornensis were found in the assemblages; however, the occurrence of Whiteinella aprica from the base of the studied section (Table 1) demonstrates that the stratigraphic interval corresponds to the upper part of the T. greenhornensis zone following the biozonation by Petrizzo and Gilardoni (2020). This assignment is supported by the presence of a rare specimen of T. reicheli at 30.20 m, of R. montsalvensis that occurs from 27.80 m upwards, of Thalmanninella gandolfii first observed at 28.80, and of T. deckeei identified at 31.75 m ( Fig. 2; Table 1).

Comparison with previous studies of the planktonic foraminiferal assemblages at Folkestone
Comparison of the data acquired in this study with those obtained in previous studies on the same stratigraphic sequence (Paul et al. 1994;Mitchell & Carr, 1998;Moghadam & Paul, 2000) highlights some differences. In the present study, the appearance of R. cushmani is recorded at 38.80 m whereas it was first observed by Paul et al. (1994) above this level at c. 40.5 m. The difference is minor and is probably related to the different sampling resolution (1 m in Paul et al. 1994 compared to 50 cm in this study), to the preservation of the specimens and to the rarity of the first representatives of R. cushmani. In fact, R. cushmani evolved from R. montsalvensis Petrizzo & Gilardoni, 2020) and occurs together with common specimens that are morphologically transitional with its ancestor, making the identification of the first R. cushmani sometimes difficult to detect (Ando et al. 2015;Erbacher et al. 2020).
A discrepancy is also observed in the stratigraphic range of T. reicheli which was found in only one sample below the MCE 1 level at 30.20 m in this study. Moghadam and Paul (2000) recorded T. reicheli at c. 30.5 m. Paul et al. (1994) reported T. reicheli from two samples below these levels at c. 28-29 m. By contrast, only Mitchell and Carr (1998) observed specimens related to T. reicheli (identified as Rotalipora ex gr. reicheli) discontinuously occurring from the base of the section to within the MCE 1 level at c. 34 m. These differences in the identification of T. reicheli are not surprising considering the known discontinuous occurrences and rarity of this species in many Tethyan localities (northern Israel: Lipson-Benitah et al. 1997; southern Spain, central Tunisia and northern Italy: Wonders, 1980;Poland: Peryt, 1983;Dubicka & Machalski, 2017) and in the western North Atlantic Ocean (Blake Nose: Bellier et al. 2000), and its absence in the US Western Interior Seaway (Pessagno, 1969;Eicher & Worstell, 1970;Denne et al. 2014;Eldrett et al. 2015) and northern California (Douglas, 1969), that limit its reliability for intra-and inter-basin correlation. Favusella washitensis, observed by Paul et al. (1994) and Mitchell and Carr (1998) within the MCE 1 level at c. 34-35 m, has not been observed in this study.
A significant increase in the proportion of planktonic foraminifera was first recognized by Carter and Hart (1977) in the middle Cenomanian of southern England and northern France and was suggested to record a significant break in the succession; the so-called 'mid-Cenomanian non-sequence'. Subsequently, the abrupt increase in abundance of planktonic foraminifera proved to be a useful marker horizon, named the P/B break (a level at which the proportion of planktonics increases from c. 5 % to as much as 50 %) by Paul et al. (1994), and also identified in other stratigraphic sections of the Anglo-Paris Basin (Mitchell et al. 1996;Moghadam & Paul, 2000), coinciding with the first occurrence of R. cushmani or falling close to the base of its stratigraphic range.
In the Cenomanian of the Anglo-Paris Basin, values of % P in excess of 40 % have not been recorded below the appearance level of R. cushmani (Hart et al. 1989;Paul et al. 1994) and only Mitchell and Carr (1998) reported % P higher than 40 % in the C. arliesiensis Bed at c. 32 m in the Folkestone section. By contrast, our data reveal that the abundance of planktonic foraminifera exceeds 50 % at only a few levels throughout the stratigraphic section studied (Fig. 2). The P/B break is placed at c. 40.5 m at Folkestone by Paul et al. (1994) and Moghadam and Paul (2000); we record a systematic increase of planktonic foraminifera with values higher than 50 % from above the termination of the MCE 1 interval at 37.80 to the top of the section, although in this interval some samples show lower % P values (23 to 34 %: Fig. 2; Supplementary Material Table S2). Therefore, we continue to use the term P/B break as employed in previous publications as we observe a cyclic alternation in abundance of the foraminiferal assemblages in the upper part of the section. Paul et al. (1994) documented a decrease in size of rotaliporids and muricohedbergellids (=Hedbergella in Paul et al. 1994) in the lower part of the MCE 1 interval from 31 to 34 m that has not been observed in the current study. The reasons for this discrepancy are difficult to explain in the absence of detailed information on how the decrease in size was measured by Paul et al. (1994) who analysed samples with 1 m resolution, examined the >250 μm sizefraction and only scanned the smaller size-fractions for smaller species. In the present study, samples were collected at 50 cm intervals and the size-fractions >125 μm were studied. Interestingly, we observed planktonic foraminifera larger than 250 μm only in the interval where Paul et al. (1994) observed the decrease in size (31 to 34 m); by contrast, the increase in size near the top of the section from 40 to 42 m (at the P/B break) shown by Paul et al. (1994) correlates with our data documenting an increase in abundance of planktonic foraminifera in the >250 μm size-fraction (Fig. 2).

Palaeoceanographic inferences
The planktonic foraminiferal assemblage in the middle Cenomanian at Folkestone is composed of common muricohedbergellids and praeglobotruncanids followed in abundance by whiteinellids, rotaliporids and dicarinellids. The palaeoenvironmental setting can be determined by considering the palaeoecological preferences of Cenomanian planktonic foraminiferal taxa which are mainly based on their stable isotope (δ 18 O and δ 13 C) composition, and on their biogeographic distributions across latitudes, with additional information provided by their abundances. Studies focused in near-coastal, hemipelagic and pelagic settings indicate that planktonic foraminifera may have tolerated variations in salinity levels and are capable of completing their life cycles in both shallow-and deeper-water environments (e.g. Leckie, 1987;Hart, 1999;Huber et al. 1999;Premoli Silva & Sliter, 1999;Bornemann & Norris, 2007;Petrizzo et al. 2008Ando et al. 2010).
According to the literature (see reviews in , Rotalipora and Thalmanninella inhabited cold/deep layers of the water column close to the thermocline, although with some differences observed among species and through time. Rotaliporids found in pelagic settings of the Tethyan Realm have been interpreted as oligotrophic, and thus required a relatively thick and stratified upper water column to complete their life cycle. Muricohedbergellids are adapted to relatively cold/deep layers of the water column in open ocean settings at low latitudes (Norris & Wilson, 1998;Wilson et al. 2002;Petrizzo et al. 2008;Ando et al. 2010), but they may have inhabited shallower layers at the higher latitudes of the Southern Hemisphere, probably because sea-surface waters were cooler Petrizzo et al. , 2021Petrizzo et al. , 2022a. Moreover, muricohedbergellids are commonly found in low-salinity coastal environments to normal-salinity open-ocean settings, indicating they had an opportunist meso-eutrophic life strategy (e.g. Leckie, 1987Leckie, , 1998Hart, 1999;Premoli Silva & Sliter, 1999). Praeglobotruncana species have been interpreted as intermediate-or winter mixed-layer dwellers Falzoni et al. 2016), and adapted to cool and poorly stratified surface waters (Ando et al. 2010). Dicarinella were adapted to cold waters and lived close to the permanent thermocline in different localities at low to middle latitudes (Huber et al. 1999;Wendler et al. 2013;Falzoni et al. 2016;. Data from the Southern Hemisphere indicate a shallower habitat at higher latitudes (Petrizzo et al. , 2021(Petrizzo et al. , 2022a. Therefore, water depth, ocean circulation patterns (e.g. currents) and boundaries between different water masses (Schiebel & Hemleben, 2017) are the main factors controlling the presence or absence of taxa in ecological niches. Other ecological factors are the thickness of the mixed layer, the position and stability of the thermocline, trophic conditions, and salinity (e.g. Hart, 1999;Abramovich et al. 2003;Ando et al. 2009Ando et al. , 2010Falzoni et al. 2013Falzoni et al. , 2016Petrizzo et al. 2017; and discussions therein).
Our knowledge of the palaeoecological preferences of planktonic foraminifera coupled with their abundance and distribution is consistent with a palaeoceanographic scenario at Folkestone that implies changes in circulation patterns. Specifically, in the MCE 1 interval the absence of the thermocline-dwelling oligotrophic rotaliporids is balanced by the appearance of the intermediate-depth and cold-water adapted dicarinellids. Moreover, the assemblage is characterized by the relatively high abundance of the eutrophic-mesotrophic Praeglobotruncana and Muricohedbergella. Interestingly, M. portsdownensis, a species also documented as occurring in the middle Cenomanian record of the North German Basin at Wünstorf (Erbacher et al. 2020), shows a stratigraphic distribution limited to the carbon isotope excursion identifying the MCE 1 at Folkestone. The species first appears in the C. arlesiensis Bed at the same level of the occurrence of Boreal macrofossils species and coincident with a positive δ 18 O shift of 1 ‰ and disappears slightly after the termination of the MCE 1 at the level (37.80 m) that records the increase in % P and the appearance of the intermediate forms between R. montsalvensis and R. cushmani (Fig. 2). According to the observed record we infer that M. portsdownensis had a meso-eutrophic life strategy, proliferated in cold water and is probably a Boreal species. However, further studies are needed to confirm the stratigraphic and geographical distribution of this species which has rarely been mentioned in the literature, and, as discussed in the Taxonomic remarks section further below, additional studies are required to investigate its possible synonymy with Muricohedbergella kyphoma, a species coinciding with the coldest episode  of the late Cenomanian Plenus Cold Event (PCE).

Remarks on the cooling pulses within the MCE 1 and comparison with the Plenus Cold Event (PCE)
At Lydden Spout, and in other sections in the Anglo-Paris Basin, the occurrence of Boreal benthic macrofossils (C. arlesiensis, O. seminudum and P. primus) is recorded within the MCE 1 δ 13 C excursion (Paul et al. 1994; and coincides with a positive excursion in δ 18 O (Voigt et al. 2004;Gale & Kennedy, 2021) and isotopically light neodymium values (Zheng et al. 2016). The shelf-sea cooling pulses observed within the MCE 1 at the C. arlesiensis Bed and P. primus Bed by Voigt et al. (2004) were interpreted as coinciding with transgressive phases during a major third-order sea-level lowstand and reflecting a major reorganization of the North Atlantic circulation.
Our new oxygen-isotope data (bulk chalk) through MCE 1 at Lydden Spout ( Fig. 2; Supplementary Material Table S1) show a heavy shift of c. 1 ‰ in the C. arlesiensis Bed. This level is precisely coincident with the base of the MCE 1 carbon-isotope excursion, the occurrence of Boreal macrofossils, the disappearance of rotaliporids and the appearance of M. portsdownensis. This change translates into a cooling of c. 4°C (temperature equation of Anderson & Arthur, 1983), probably in sea-surface waters as the values are derived from calcareous nannofossils in the bulk sediment. Brachiopod shells through the C. arlesiensis Bed (Gale & Kennedy, 2021, fig. 6) also show a 1 ‰ heavy shift in δ 18 O, but to values which are consistently 1 ‰ heavier than those from the bulk sediment record. These data thus provide evidence of a significant temperature gradient from the surface to the sea floor (c. 4°C) in the middle Cenomanian Chalk Sea. The fluctuating values of δ 18 O coincide precisely with the boundaries of chalk:marl couplets, and heavier values are consistently present in the marls, lighter values in chalks. The absolute heavy values of δ 18 O decrease progressively in the three couplets succeeding the C. arlesiensis Bed (Fig. 2). The couplets have been identified as precession cycles  and it can therefore be argued that orbital forcing was a primary control on the fluctuating temperature values.
The cold-water planktonic foraminiferal assemblages occurring in the MCE 1 interval indicate the presence of poorly thermally stratified water masses characterized by a reduced/collapsed thermocline that disrupted the ecological niches of the oligotrophic rotaliporids which, thus, are totally absent in the MCE 1 interval at Folkestone (Fig. 2). Keeled planktonic foraminifera are also absent across the MCE 1 interval identified in the Boreal North German Basin at Wünstorf, and their absence was controlled by sea-level changes in an epicontinental sea where environmental conditions were unsuitable for keeled planktonic foraminiferal species (Erbacher et al. 2020).
Together, these observations indicate that the enhanced influence of Boreal seawater in the European epicontinental sea resulted in a sensitive response of ocean circulation to climate fluctuations and support the hypothesis of the inflow into the Anglo-Paris seaway of cold Boreal waters originating in the Norwegian Sea (Fig. 3). A similar interpretation was proposed to explain the planktonic foraminiferal assemblage changes and the occurrence of the Boreal macrofauna C. arlesiensis, O. seminudum and P. plenus in the PCE interval at Eastbourne (UK) (O'Connor et al. 2019;. Comparison between the planktonic foraminiferal assemblages during the cooling episodes within the MCE 1 and those occurring at the PCE reveals a similar composition, with dominance of taxa more adapted to cooler and relatively poorly stratified water masses (Muricohedbergella, Praeglobotruncana, Dicarinella). Additional evidence for a Boreal origin of these assemblages is provided by the occurrence of common M. portsdownensis at Folkestone and M. kyphoma at Eastbourne . Regardless of the taxonomic assignments of these species (as explained in the Taxonomic remarks section further below), they are only observed in the levels containing Boreal macrofossils and are therefore interpreted as mesotrophic-eutrophic species that did not require a thermally stratified water column. The absence of oligotrophic rotaliporids in the MCE 1 interval at Folkestone and the extinction in the PCE interval of R. cushmani (the last representative of the rotaliporid group) are probably related to the ingress of cold Boreal waters that temporally disrupted the stratified oligotrophic water column. Therefore, the equatorward expansion of the Boreal macrofossils and the dominance of cold-water and meso-eutrophic planktonic foraminifera observed at the MCE 1 and PCE may reflect cyclic latitudinal shifts of the proto-Arctic Front of c. 10-20°that forced a major reorganization of the surface and deep ocean circulation patterns in the middle and late Cenomanian (Zheng et al. 2016;. Using an orbitally tuned dataset, Batenburg et al. (2016) suggested that the MCE 1 and OAE 2 carbon isotope events were driven by the long eccentricity cycle (mode at 2.4 Ma) with positive excursions coincident with minima. In parallel, the southerly extensions of cold Boreal water during the MCE 1 and the PCE may also have been caused by long eccentricity-driven reorganization of ocean circulation.

Conclusions
Planktonic foraminifera at Folkestone were studied across the MCE 1 positive δ 13 C excursion and are correlated with the occurrence of Boreal macrofossils and of heavy δ 18 O excursions in bulk carbonate values, indicating a cooling of c. 4°C. In general, planktonic foraminifera stratigraphically below the MCE 1 level indicate a pelagic epicontinental environment characterized by a stratified water column with a well-defined mixed layer and a thin thermocline, as confirmed by the occurrence of both surface-and deepwater dwelling taxa. This palaeoenviromental setting is interrupted at the onset of the δ 13 C excursion of the MCE 1 as revealed by changes in the composition of the planktonic foraminiferal assemblage and in the stratigraphic distribution of species. The doublekeeled Dicarinella first evolved within the MCE 1, indicating the availability of favourable ecological niches in a thick mixed layer. By contrast, the total absence of the single-keeled rotaliporids within the MCE 1 interval is interpreted to reflect a substantial disruption of water mass stratification and of the ecological niches occupied by the thermocline-dwelling oligotrophic rotaliporids. Cooling phases within the MCE 1 are also confirmed by the occurrence of the Boreal macrofossils C. arlesiensis, O. seminudum and P. primus, the occurrences of which identify the C. arlesiensis Bed and P. primus Bed. Remarkable is the appearance of the species M. portsdownensis at the base of the MCE 1 interval at the same level of the occurrence of Boreal benthic macrofossils (C. arlesiensis Bed). This observation suggests that M. portsdownensis was a surface and cold-water dweller and it is probably a useful middle Cenomanian marker species that permits identification of the MCE 1 interval. After the termination of the perturbation associated with the MCE 1, planktonic foraminifera increase in abundance and species richness as testified by the occurrence of keeled and deep-water species, indicating the return of a stratified water column with a mixed layer and well-defined thermocline.
Overall, the planktonic foraminiferal assemblages recorded in the MCE 1 interval reflect conditions of eutrophy associated with cold surface waters. In addition, the occurrence of Boreal benthic fauna associated with heavy oxygen isotope excursions provides evidence of cooling episodes which might coincide with transgressive mixing over the epicontinental Anglo-Paris Basin. These falls in temperature possibly resulted from the southerly extension of cold Boreal low-salinity and poorly stratified water masses which originated in the Norwegian Sea, as previously postulated to explain the occurrence of Boreal macrofossils within the positive δ 13 C excursion that identifies MCE 1. Finally, similarities in the composition of benthic macrofossils and planktonic foraminiferal assemblages, and in the heavy oxygen isotope excursions which occurred during the MCE 1 and the PCE, suggest that two latitudinal shifts of the proto-Arctic Front of c. 10-20°took place in these intervals. The changes may be related to orbital configurations driven by long eccentricity maxima (cf. Batenburg et al. 2016).

Taxonomic remarks
Foraminifera genera and species with authors and years identified in the studied sediments (Table 1) are listed in alphabetical order and comments are included for some species so as to clarify the taxonomic concepts followed in this study. Remarks  Remarks. Previously included in Globigerinelloides, see taxonomic revision in Huber et al. (2022).
Muricohedbergella portsdownensis is very similar to Muricohedbergella kyphoma, a species described from Japan (Hasegawa, 1999) as occurring discontinuously from below the first appearance of Rotalipora cushmani to above the extinction of R. cushmani. The species has been recognized within the Plenus Cold Event (PCE) at Eastbourne and Clot Chevalier (Gale et al. 2019a) by  in the interval recording the extinction of R. cushmani and the occurrence of the Boreal macrofossils (C. arlesiensis and P. plenus), and thus interpreted as a marker species indicative of the coolest episodes of the PCE.
Muricohedbegella portsdownensis and M. kyphoma are here regarded as separate species although the possibility that they are morphotypes of the same species cannot be excluded since they show a similar muricate and pustulose wall texture and only differ by the sutures on the ultimate chambers of the spiral side that are curved in M. kyphoma and straight in M. portsdownensis. However, for the time being, we separate the two species until detailed stratophenetic studies of the planktonic foraminiferal assemblages are performed in stratigraphic sections containing a continuous record from the interval below the MCE 1 to the interval above the PCE in order to investigate the timing and the mode in the evolution and distribution of the two species.
Remarks. The specimens identified in this study as P. cf. compressa closely resemble the holotype of P. compressa and only differ by having a slightly high trochospire and a biconvex lateral profile.
Supplementary material. To view supplementary material for this article, please visit https://doi.org/10.1017/S0016756822000991