.
Terminology – Definitions and explanations of palaeo-sea-level terminology References – download a spreadsheet of references NOTE: all elevations are quoted relative to LMSL unless otherwise stated.
The relative sea level (RSL) record at Bermuda during sub-stage, interstadial, MIS5a
Albeit scarce, the existence of emergent MIS 5a marine deposits on Bermuda, is no less certain than is the existence of MIS 5e and MIS 7 marine deposits. In all cases, the predominant source of age data is coral fragments found in emergent foreshore clastic deposits at the base of a regressive beach-soil-dune sequence.
The effective type locality for marine 5a deposits, attributed to the Southampton formation, is at the north-eastern extremity of St George’s Island at Fort St. Catherine. Here, 22 Oculina coral fragments collected from cemented emergent beach deposits, have yielded ages between 85 ka and 77 ka with a mean value of 80 ka. (1, 4, 11).
Vacher and Hearty (1989) (16) concluded that the coarse beach deposit with low angle planar foreshore strata from which these corals were collected, record a MIS 5a local relative sea level (LRSL) of approximately +1m (1 m above present local mean sea level).
Others, mindful of widely accepted version of global Pleistocene sea level history in which MIS 5a global mean sea level did not approach present global mean sea level (GMSL), have contended that that St Catherine deposit must have been emplaced during a Holocene storm event (4, 57). Muhs et al. (2002) (11), challenged this assertion, however; claiming that material reworked from the seabed by storm waves would incorporate corals from a variety of ages rather than a small range tightly clustered around 80 ka. It can be argued, additionally, that a storm deposit would not be integrated into conformable beach-soil-dune vertical sequence which likely formed over thousands of years (81).
Potential MIS 5a deposits have been found at only two other locations on Bermuda – at Bird Island on the Great Sound and at Conyers Bay on the south west shoreline. As at St Catherine, these deposits are weakly cemented, coarse, poorly sorted and unquestionable young.
Bird Island is part of an arc of small islands situated in a large inshore body of water known as the Great Sound. For the most part, these islands are formed of sub-tidal and intertidal shelly deposits which rise up to +6 m ASL. This elevation combined with a couple of MIS 5e ages from similar deposits on nearby islands are grounds enough for an MIS 5e attribution. Nonetheless, a single Oculina coral fragment collected from the shelly strata on Bird Island by Harmon et al (1983) (4) yielded a convincing MIS 5a age of 83 ka.
Given the reliability and reproducibility of Pleistocene radiometric ages from Bermuda and the absence of contending ages for the more northerly segment of the arc, which includes Bird Island, a MIS 5a age is arguably credible and certainly is grounds for more dating of coral fragments from this area.
At Conyers Bay a wedge of shelly deposits abuts eolianites of Bermuda’s youngest formation (16). This stratigraphical relationship along with Amino acid racemization (AAR) ratios from molluscs which are comparable to those at Fort St Catherine, make the Conyers Bay marine unit a candidate for potential MIS 5a designation. Unfortunately, no coral fragments have been found in this small unit.
In addition to emergent coastal marine deposits, other evidence of an MIS 5a sea level excursion above the present level on Bermuda has been found in cave speleothems. Wainer et al. (2017) (18) report growth interruptions and a distinctive chemical signature in speleothems which are consistent with submergence by marine waters rising more than 1 m above present local mean sea level (LMSL) at approximately 80 ka.
At Bermuda, still considered by some to be tectonically stable, sea level indicators attributed to MIS 5a support a relative sea level (RSL) in the range of -1m to + 2m (9). In the absence of vertical land motion, these elevations challenge an established consensus opinion that MIS 5a eustatic (global) sea level stood at around the range of – 10 m to – 20 m below present mean sea level. (GMSL). Nonetheless, increasingly there is evidence from relatively stable coastlines of MIS 5a eustatic sea levels ranging up to and above present GMSL (89).
Modeling of the Glacio-hydro isostic (GIA) effect in more recent years has helped to resolve the scale of the GIA contribution to palaeo-relative sea level. The models show that Pleistocene relative sea levels at Bermuda were exaggerated due to GIA-related land subsidence at warm marine isotope stages including MIS 5a. It is inferred that the peak in MIS 5a eustatic sea level would not have exceeded present MSL on its own. Only when combined with GIA- associated land-subsidence of as much as 10 m was it possible for peak RSL to exceed this mark and leave an emergent imprint on Bermuda’s shores (89) .
It turns out that the observed MIS 5a sea level imprint on Bermuda at 0 to + 2m, falls at the lower end of the range of RSL values simulated by GIA modelling. If correct, this contradicts or reverses a long-held belief, by some, that the Bermuda MIS 5a RSL imprint is anomalously high. It may in fact be the opposite.
The global relative sea level (RSL) record during sub-stage, interstadial, MIS5a.
Evidence of peak MIS 5a RSL lower than present mean sea level.
In his studies of planktonic and foraminiferal tests in deep-sea cores Shackleton (1969)(100) first established a correlation between past oxygen isotope levels and ocean water volumes, which, in turn, broadly represent palaeo-sea-level. Largely on the basis of such marine isotope measurements, there has been a long held view that global mean sea level at 80 ka (MIS 5a) stood below the present level.
Measurment of the concentraions of other climate proxies – namely Deuterium, CO2, CH4, dust levels – in the Vostok East Antarctic Ice core (27) corroborate a lower-than-present MIS 5a sea level as did isotope analyses of plankton from a Red Sea core (Rohling et al., 2009)(37).
From uplifting fossil reef tracts of New Guinea it has been calculated that sea level at 80ka could have been as high as -6.6m or as low as -19m (51). These levels are comparable to those derived from extensive high quality data yielded by submerged MIS 5a reefs in tectonically stable southern Florida. Toscano et al. (1999) (57) reported a 5a sea-level peak of -9.0 m in the southeast Florida. While in the Keys, Hsia et al. 2024) reconstructed MIS 5a sea level peaking in the range of -6.5 to -5.1m.
In the Bahamas, the growth history of speleothem deposits in caves reportedly supports an MIS 5a sea level maximum well below current local mean sea (78). Dating of spleothem layers, there, indicates that sea level was higher than -10 to -15m at several stages of the Pleistocene but not between 110 ka and 40 ka when, according to Li et al (1989)(78), there was continuous speleothem growth. If correct this constrains MIS 5a sea level, at ~80ka to less than – 15m. Likewise Richards et al. (1994)(79), citing the research by Li, asserted that there could not have been high sea-stand events near present sea level between 93 ka and 15 ka. These early speleothem studies, however, lacked U–Th dating precision and resolution to reliably identify all growth interruptions in the submerged samples (18). Short-lived marine transgression may have been missed.
Crevling at al. (2017)(89) undertook GIA modelling simulations calibrated against MIS 5a RSL data they had compiled from around the world including the Philippines , Korea , New Guinea, The Caymans, Eastern USA, Bermuda and the Florida Keys. They concluded that eustatic global mean sea level (GMSL) peaked at approximately – 10m ± 5 m during MIS 5a. Noteworthy, here, is their prediction (89) of a MIS 5a RSL at Bermuda of ~+5m, when taking into account glacio-isostatic adjustment (GIA) effects. This, if correct, may help to explain the existence of controversial emergent MIS 5a marine deposits on Bermuda.
Evidence of peak MIS 5a RSL close to present mean sea level.
Sea-level-controlled phreatic overgrowths on speleothems on Mallorca witness a +1.3 to 1.9m high stand at MIS5a (85 to 80 ka)(75), (48), (77), (60). Dorale et al. (2010)(75) conclude from their data that there was “less ice on Earth 81,000 years ago than today”. Evidence from this Mediterranean research indicates a very rapid onset and relatively brief nature of the MIS 5a highstand (possibly missed by the Bahamian speleothem record. See above). It is postulated that this rapidity may have generated lags between the timing of sea-level changes and the timing of coral reef growth, and may provide a partial explanation as to why reefs on Barbados and New Guinea do not record a comparable eustatic response (as Mallorca and Bermuda). (75).
Corroborating the Mallorcan spleothem record are coastal deposits of MIS 5a age (81ka) preserved at 1.5 m above present sea level, also on Mallorca (59). This level may plausibly contain a significant effect of GIA, associated with Northern Hemisphere ice sheet history (48). However, Dorale et al. (2010) (75) propose that the GIA effects have been overestimated for this region, suggesting that sea level nearly follows the eustatic curve.
On the Atlantic Coastal Plain of the USA, emergent marine deposits, a few meters above sea level, have yielded U-series coral ages of ∼ 80 ka. This is unexpected on a passive continental margin given the lower-than-present sea-level estimates from Barbados and New Guinea (50). However, similar results, to those from the Coatal Plain, were obtained on supposedly (51) tectonically stable Bermuda, where the marine facies of the Southampton Formation, at +1m to +2m, yielded U-series ages of coral fragments averaging ∼ 80 ka (51).
On the tectonically active Ryukyu Islands of Japan, reef terraces dating to ∼ 120 ka, ∼ 100 ka, and ∼ 80 ka are all present (50) and witness paleo-sea levels at MIS 5c and 5a that are close to the present level (Ota and Omura, 1992). U-series ages and terrace elevations from the Pacific coast of North America (California and Mexico) also give paleo-sea-level estimates for MIS 5c and 5a that are much closer to present sea level than the (lower) consensus elevation established elsewhere (Muhs et al., 1994).
Evidence of peak MIS 5a RSL significantly higher than present mean sea level.
Marine deposits attributed to MIS 5e (∼125 ka) are rare north of the Florida Keys. Where found, these units are at roughly the same elevation (all at ~+6m to +8m) as those dated at 80 ka (54). The vast majority U-TH ages from these presently emergent corals which fall in the interval between 65 and 85 ka are found at up to +8 m (51). Only by invoking uplift of at least 10 to 15m can the nearly uniform elevation of these ~80 ka units in Virginia, South Carolina, and Georgia be reconciled with the marine isotope curves or dated tropical coral terrace records (67).
On Grand Cayman, in the Caribbean, the presence of MIS 5e wave-cut notch at +6m is posited as evidence of tectonic stability (64). This is because +6 m is consistent with the consensus version of global eustatic sea level during the LIG (64). Emergent marine deposit which has been dated to 74–87 ka on Grand Cayman are, therefore, considered to be a credible record of an MIS 5a GMSL which peaked at +3m to + 6 m (64).
Bermuda Geology 