Whitecliff Bay, at the eastern end of the Isle of Wight (IoW), is one of the most complete Paleogene successions in Europe and has been studied by a range of geoscientists for over 200 years (Fig. 1). The first reported description of “Tertiary” geology in England was the monograph by Solander (in Brander, 1766), which described a collection of fossils from the cliffs near Barton-on-Sea and which were subsequently deposited in the collections of the newly opened British Museum. The first account of the English “Lower Tertiaries” was provided by Webster (1814) using the IoW successions. This was followed by Bowerbank (1842), who focussed on the London Clay. The first relatively complete account of the Whitecliff Bay succession was given by Prestwich (1846), although he modified the names of the various formations in the following years (Prestwich, 1847a, b, 1850, 1852, 1853, 1854a, b, 1855); some of the beds were also assigned Roman numerals. It was during this time that Carpenter (1850) described the larger foraminifera of the Tertiary succession, including Nummulina. In the same year, Dixon (1850) described both the larger foraminifera (including Alveolina and Nummularia) and the smaller foraminifera (including Quinqueloculina, Triloculina, and Biloculina) of the Bracklesham Beds in Sussex. The Bracklesham Beds were later described by Fisher (1862), who applied Roman numerals to his various subdivisions (both lithological and palaeontological). These Roman numerals (i–xix) were still in use during the 20th century (Curry, 1965, 1967; Curry et al., 1968; Murray and Wright, 1974) and are often used up to the present day. In the same paper, Fisher (1862) used Arabic numbers for the Bracklesham Beds at Bracklesham and Selsey Bill, and while he defined the base of his Bracklesham Beds, he did not define the base of the overlying Barton Beds, possibly because of poor exposure at the time. The base of the Barton Beds was eventually defined by Keeping (1887), using the first appearance of Nummulites prestwichianus, with Gardner et al. (1888) providing further comment on the possible sub-division of the Barton Beds. The problems of the global stratigraphic section and point (GSSP) for the Bartonian stage have recently been discussed by Hooker and King (2019). The overlying succession of the Headon Beds was described by Bristow et al. (1889) with further clarifications being provided by Stamp (1921), White (1921), Bhatia (1955, 1957), and Stinton (1971). The seminal publications of Dennis Curry (1965, 1967) were the result of his two presidential addresses to the Geologists' Association and remain important references to this day. The most comprehensive investigation of the foraminifera was provided by Murray and Wright (1974), with much of the information in that account being the foundation for the later Stratigraphical Index of Fossil Foraminifera (Murray et al., 1981, 1989). In their research on the Paleogene succession, John Murray and Chris Wright shared the workload, with the former investigating the younger strata and the latter having more responsibility for the London Clay Formation and the Bracklesham Group sediments. Ahead of the Murray and Wright (1974) publication, Wright (1972) described a distinctive influx of planktic foraminifera within the London Clay Formation, describing it as a “planktonic datum”. Wright (1972) demonstrated the value of this “event” and used it in his correlation of the London Clay Formation across the Hampshire Basin. The London Clay Formation was also the subject of a major study by Chris King (1981) and the recently published update (King, 2016) on the Cenozoic.

Whitecliff Bay, along with other locations on the Isle of Wight, has also been a popular destination for geological field excursions (e.g., Herries and Monckton, 1895; Curry, 1954, 1966; Stinton, 1971; Lord and Bown, 1987). Some of these excursion guides provide additional stratigraphical information and documentation of the fossil assemblages. The Whitecliff Bay succession has also featured in many sedimentological and palaeontological papers in recent years and remains, therefore, an important reference section for much work on the UK Cenozoic (e.g., Gale et al., 1999).

At the eastern extremity of the IoW, the chalk promontory of Culver Cliff is a distinctive feature, marking the boundary between the upper Cretaceous succession and that of the Cenozoic. This headland, and the contact between the chalk succession and the overlying Reading Beds, forms the southern end of the bay (Fig. 2a). The northern end of the bay is marked by Foreland, near Bembridge. Within Whitecliff Bay, the greater part of the Paleocene–Eocene succession is vertically dipping, with the Barton Sands and Lower Headon Beds gradually becoming less inclined. The Bembridge Limestone is almost horizontal with, at low tide, the closure of the Hampshire Basin syncline being visible on the foreshore.

The cliff succession is unstable, with silt-rich and sand-rich parts of the succession often standing up against the slumps and mudflows within the more clay-rich units (Fig. 2b). One can never, therefore, guarantee access to the complete succession at the time of any visit. It is also known that weathering profiles, affecting the vertical strata, may selectively remove some of the fossil assemblages, as was reported by De Jonghe et al. (2011) in an investigation of the Marsh Farm Formation. After some storm events, the bay can become stripped of sediment and the foreshore fully exposed, providing unparalleled access to the vertically dipping strata (Fig. 2a, b). It is also our experience that samples collected from the foreshore often contain better preserved, and probably more complete, microfossil assemblages. During investigations of the foreshore (Fig. 2b), erosion surfaces, often overlain by mudstone clasts and small phosphatic nodules, were located, and these indicate potential hiatuses and non-sequences. Assessing the scale of these gaps in the overall succession is, therefore, quite challenging and indicates that the London Clay succession in Whitecliff Bay (and Alum Bay) may be incomplete; such features are probably the norm, though rarely described.

During the last 20 years, the Paleogene has been recognised as containing a number of significant hyperthermal events, including the Palaeocene/Eocene Thermal Maximum (PETM); see Zachos et al. (2001, 2008), Bijl et al. (2009), Westerhold and Röhl (2009), Lauretano et al. (2015), and Westerhold et al. (2020) for important summaries of both the events themselves and the methodologies employed in their documentation. Westerhold (2020, fig. 1) clearly identifies (1) the PETM at the Thanetian/Ypresian boundary in the lower part of magnetic event Chron C24; (2) the Early Eocene Climatic Optimum (EECO) – which includes a number of minor hyperthermals – between mid-Chron C24 and lowermost Chron C21; (3) the Late Lutetian Thermal Event; and (4) the Middle Eocene Climatic Optimum (MECO), which is a short-duration event (ca. 500 kyr; Giorgioni et al., 2019) within the lower part of the Barton Clay Formation, low in Chron C18. The MECO event is clearly preceded by the Late Lutetian Thermal Maximum, and the presence of all these events raises questions about their impact on the following:

the possibility of glacio-eustatic sea level changes (associated with raised global temperatures) being recorded in the Isle of Wight succession;

In this work, samples (ca. 2 kg) collected from Whitecliff Bay have been recorded accurately against the lithological succession and processed using the white spirit method (see Brasier, 1980). This method, though requiring careful preparation within a fume hood, was found to be extremely effective in the 100 % breakdown of claystone and siltstone samples from Whitecliff Bay and other Hampshire Basin localities with no damage to the foraminifera or other microfossils being detected. All samples, after breakdown, were washed on a deep 63 µm sieve; dried in a cool oven (< 40 °C); and investigated in the >500, 500–250, 250–150, and 150–63 µm size fractions.

Foraminifera are protists (single-celled organisms) that secrete a carbonate shell (test) or create a structure from sedimentary particles using organic fibres or cement that may be carbonate based or siliceous (rare). Micropalaeontologists classify foraminifera using wall structure and the organisation of the chambers but also use an “informal” set of groupings based on a planktic mode of life (PL) or benthic taxa that are described as “larger” (LBF) or “smaller” (SBF). The LBF are grouped, not by wall structure but by their relatively large size and complex internal structures. Many of the LBF can only be identified at the species level by the use of oriented thin sections, while the SBF are normally identified by external, microscopic examination or scanning electron microscopy. In the Eocene succession of Whitecliff Bay, SBF are most commonly recorded, being used in both biostratigraphy and palaeoecological analysis (e.g., Murray and Wright, 1974). In the mid-Eocene, however, LBF are found to be associated with assemblages of SBF, and this is quite unusual for such clay-rich sediments at relatively northern palaeolatitudes. LBF are typical of shallow-water carbonate-rich shelves and are usually recorded in tropical or sub-tropical environments (Langer and Hottinger, 2000; Simmons, 2020). The internally complex LBF often contain algal or diatom symbionts (Hallock, 2003; Lee, 2011; Eder et al., 2016), and this limits their occurrence to > 120–130 m in clear waters and relatively shallower depths in more turbid environments.

The LBF often display complex life histories with the megalospheric (A) forms and the microspheric (B) forms. The (A) forms produce gametes which merge sexually to create the microspheric (B) forms, which, in turn, undergo cellular division (asexually) to generate the larger (A) forms: see Hallock (1985) and Hohenegger et al. (2019). SBF may also display A and B forms, though descriptions of such life histories are relatively rare, especially in the geological record. In the Eocene sediments of the Isle of Wight, the LBF are represented by a limited number of genera, including the nummulitids (large discoidal foraminifera) and the alveolinids (globular to fusiform miliolids with a porcelaneous wall structure that appears distinctly white in colour). The alveolinids appear in the Thanetian, reaching a maximum size in the mid-Eocene and disappearing in the Bartonian together with the large nummulitids. In the following discussion, it is the occurrence of the alveolinids and nummulitids in the mid-Eocene (Lutetian and lowermost Bartonian) that will be important in our palaeoenvironmental interpretations.

The Reading Formation (Lambeth Group) is thought to be Palaeocene in age (Hopson, 2011, fig. 8), though the stable isotope excursion that marks the PETM (e.g., Westerhold et al., 2020) has not yet been located within the Isle of Wight succession. The boundary with the overlying London Clay Formation (Thames Group) in Whitecliff Bay is invariably exposed, either in the cliffs or on the foreshore (Fig. 2a, b). The outcrop of the London Clay Formation is often poorly exposed, with the cliffs often intersected by several mudslides. In order to both measure and collect a complete succession of the London Clay Formation, it has been necessary to make return visits to the area in order to benefit from times when the foreshore has been stripped of sediment and the complete succession is exposed.

In their major investigation of the London Clay Formation, Murray and Wright (1974, p. 8) indicate that they collected their samples in January 1969 when the Reading Formation was obscured by landslides and mudflows and significant parts of the London Clay Formation succession were also obscured. They record a thickness of 89.0 m, which is only slightly different from the measurements of other authors and probably reflects the problems of dealing with the incomplete exposure of the succession. They also record that significant parts of the succession were barren of foraminifera, especially in the lowest 38.0 m. In our work, both in the London Clay Formation and in the Bracklesham Group, it has been noted that foreshore samples often yielded microfossil assemblages when comparable samples from the cliff were either barren or contained foraminifera with significantly poorer preservation. It seems certain that the vertical nature of the beds has allowed much greater water penetration and that this has resulted in significant modern or sub-recent in situ weathering.

Westerhold et al. (2020, fig. 1) indicate the presence of a warming event in the late Lutetian (ca. 43 Ma) – slightly earlier than the Middle Eocene Climatic Optimum (MECO) at ca. 40 Ma. These two “events”, which straddle the Bracklesham Group–Barton Group transition, are characterised by the presence of both LBF and SBF. Aside from an isolated occurrence of Nummulites planulatus within glauconite-rich sand in the Wittering Formation, the LBF are only present in the Whitecliff Bay succession in the Earnley Sand Formation to the Barton Clay Formation: an interval that embraces the uppermost Lutetian and the lowermost Bartonian. This occurrence of LBF has been known for some time, with Keeping (1887) using the first occurrence of N. prestwichianus to define the base of the Barton succession. The first significant occurrence of LBF is in the N. laevigatus Bed of the Earnley Sand Formation. Between these two horizons there is a complex of inter-related Nummulites species – including N. variolarius (e.g., Curry, 1962) – with a number of morphotypes being described by Norvick (1969, fig. 13, p. 62). The horizon with abundant N. variolarius is the most widespread, being found across a large area of the Hampshire Basin, including the well-known succession on Hengistbury Head. Just above the N. prestwichianus Bed there is a distinctive occurrence of N. rectus, from the highest stratigraphical level at which Nummulites are recorded in the UK.

Associated with Nummulites are other genera and species of LBF – a feature noted by Curry (1967, fig. 3) and, to a lesser degree, Dixon (1850). While a number of genera (and species) were noted by Curry (1967), the most important is the occurrence of alveolinids. These internally complex miliolids are cigar-shaped and visible in the field with a hand lens or even the naked eye. Alveolinids are known only from three locations in England and some marine sampling sites offshore of Jersey (see later).

The best-known localities are Fisher's Beds 21 and 22 (Fisher, 1862, p. 74), with Bed 22 being known locally as the Clibs, which is on the foreshore at Selsey (Bracklesham Bay), Sussex (Fig. 10). This is the locality mentioned by Dixon (1850) and is accessible only at low tide on the west side of Selsey Bill (Curry et al., 1969). In an offshore reef (Mixon Rocks) about one mile (1609 m) south of Selsey Bill is an outcrop that is only available at low spring tides (Bone and Bone, 2014, figs. 1, 2). This is a miliolid/alveolinid limestone that was described by Dixon (1850, p. 13) as follows:

calcareous rocks that are situated opposite Selsey and Bracklesham are composed almost entirely of these minute shells being analogous to the Milliolite limestone of the Paris basin

These three locations are, collectively, the northernmost occurrences of Alveolina in northwestern Europe, and this is important to interpretations of palaeogeography. In a lengthy taxonomic discussion, Adams (1962) indicated that the two species present in the UK are Alveolina fusiformis and Alveolina sp. cf. A. elongata, based on a detailed assessment of the internal structure and, in particular, the dimensions of the proloculus and the first, post-embryonic chamber. Adams (1962) noted that the external shape was not 100 % diagnostic of the species (A. fusiformis and A. elongata) despite the names applied to these taxa. Hottinger (1980, p. 149) records that A. bosci is present in the lower and middle Lutetian and is recorded in the Lutetian across the Paris Basin. The species of Alveolina recorded in areas around Jersey and in core O-VC18, may, therefore, be any of the three species, and only detailed thin-section or micro-CT analysis can confirm the actual species present.

The other species listed by Curry (1967, fig. 3) as occurring in the Hampshire Basin include Halkyardia minima, Linderina brugesi, Fabularia bella, Fabularia ovata, Orbitolites sp., Cuvillierina sp., and Asterocyclina sp. These are smaller members of the LBF, except Asterocyclina, which belongs to a large group of discocyclinids and is typical of warm, shallow-water environments. Many of these taxa, as noted by Curry (1967, pp. 444–446), are also known from the Sables de Cuise and Calcaire Grossier of the Paris Basin, and all are typical of mid-Eocene successions.

A zonation for the London Clay Formation of the IoW based on foraminifera has been generated using well-known, identifiable taxa. Using palaeoecological information based on these species, a marked sea level change during the EECO event has been detected, though there are problems associated with the identification of this as being of glacio-eustatic origin. The occurrence of larger benthic foraminifera (LBF) in the LLTM and MECO events could also be attributed to sea level changes, especially as the same assemblages can be described from the sea area adjacent to Jersey and the adjacent English Channel.