tract with extensive sandy lagoonal and scattered patch
reefs. The reef-crest facies consists of dense thickets
of Goniopora sp. and Stylophora pistillata and forms
a shallowing-up sequence, indicating seaward
progradation. The lower unit is truncated by an intermittent marine-erosion surface that shows evidence of being
both bioeroded and burrowed suggesting that in some
areas the surface was a firm ground and not completely
cemented. This surface is directly overlain by a second
but more limited fringing-reef tract consisting predominantly of reef-crest and reef-front units with a diverse
assemblage of in-place corals (Bruggemann et al., 2004).
These two superimposed reef tracts therefore have
a back-stepping architecture, which implies that the first
stage of reef development was abruptly terminated and
was followed by the relocation and reinitiation of the second reef-crest further inland over the former lagoon.
Bruggemann et al. (2004) interpreted this reef demise
and relocation to have resulted from a brief eustatic SL
fall, but could not find any evidence of subaerial exposure
associated with the marine-erosion surface. The discontinuous nature of the erosion surface, and the fact that the surface was not completely cemented, is inconsistent with SL
fall, and in fact suggests the opposite scenario whereby
reef back-stepping may have been the result of a rapid rise
of relative SL. Whatever the interpretation, however, the
neotectonic setting of the site means that the possibility
that coseismic displacement was the cause of reef backstepping cannot be discounted.
Along the Egyptian coast, in the northern Red Sea, two
episodes of deposition during the LIG have been
described by Plaziat and others (Plaziat et al., 1998a, b;
Orszag-Sperber et al., 2001). During the first stage,
fringing-reef tracts topped by beach gravels developed
along the coast at elevations between þ5 and 10 m. This
stage was followed by a down-cutting erosive event that
excavated small basins behind breaks in the reef crest.
The first depositional event in these depressions was
a thin sequence of open-marine sediments containing molluscs and in-place corals. This was followed by a thicker
sequence of laminated subaqueous gypsum topped by
tepees that reached an elevation of $3 m below the reefcrest. The final filling unit consisted of massive gypsum
consistent with an emergent sabkha environment.
This unusual sedimentary sequence was interpreted to
result from a rapid SL fall of $10 m which caused the
excavation of the back-reef depressions (Orszag-Sperber
et al., 2001). The subsequent rise of SL back to the same
level as before (within 1–3 m) led to the filling of the
depressions. The abrupt transition from an open marine
to restricted deposits in this fill-sequence was postulated
to result from a brief pluvial episode that increased input of sediment to the coast and occluded the gap in the
fringing-reef unit, thus sealing the depressions off from
the marine environment (Plaziat et al., 1998a; OrszagSperber et al., 2001). However, other interpretations are
also consistent with these stratigraphic data. For example,
down-cutting could result from a rapid SL rise and
reworking of back-reef deposits during storms and sediment export through the reef gaps. These open-marine
conditions would favor the initial deposition of openmarine faunas and the abrupt change to restricted environments would mark the start of SL fall at the end of the LIG.
That no deepening of these depressions occurred during
the ensuing glacial fall in SL, lends support to this latter
interpretation. However, confirmation of either interpretation requires a precise chronological framework which is
still lacking in this region.
Reef development in more stable terrains
Reef development in stable sites can provide a detailed
look into the timing and duration of peak interglacial conditions during the highstand, as well as the absolute elevation and SL behavior during that highstand. Detailed
descriptions of LIG reefs in stable areas have traditionally
concentrated along the southern passive margin of the
North-American Plate, including sites in Florida, the
Bahamas, the Caymans, and the Yucatan. But recent work
has also been reported along the Western Australian passive margin of the Indo-Australian Plate from Cape Range
to Cape Leeuwin. This facilitates a comparison of reef
development within each region and between regions
themselves.
Florida
Pleistocene reef development in Florida’s Key Largo
Limestone was described as early as the late eighteenth
century (Agassiz, 1896), but it was not until the mid1960s that U-series dating proved that those elevated reef
deposits were of LIG age (Broecker and Thurber, 1965;
Osmond et al., 1965). As its name suggests, the reefal
Key Largo Limestone is best developed along Key Largo
and Windley Key where it forms a broad ridge that rises to
a level of þ3 m (Harrison and Coniglio, 1985). Although
widely cited to reach 5.5 m above mean SL, the original
description states that it has an average elevation of
þ1 m “with local prominences which in two places reach
a maximum height of 18 ft” (5.5 m) (Hoffmeister and
Multer, 1964, p. 355). At the type locality on Windley
Key, the reef unit covers an older topographic high at
þ1 m, and forms a 2 m thick layer of coalesced patch reefs
formed of head corals, particularly Montastrea annularis,
interspersed with thickets of A. cervicornis and small
branched Porites (Stanley, 1966). The matrix is a skeletal
sand composed of mostly Halimeda and mollusc fragments. The reef unit gradually dips away from this topographic high and, at its northern and southern outcrop
limits, is overlain by or grades into the oolitic sands of
the Miami Limestone (Multer et al., 2002). These oolitic
units have developed shorelines consistent with a þ6 m SL
highstand at both Key West in the south and Palm Beach
in the north (Hoffmeister and Multer, 1968; Perkins,
1977; Multer et al., 2002). The absence of ooids in the
reef matrix and the fact that the reef unit underlies these
oolitic sand bodies indicates that reef development likely
LAST INTERGLACIAL AND REEF DEVELOPMENT
627
reefs. The reef-crest facies consists of dense thickets
of Goniopora sp. and Stylophora pistillata and forms
a shallowing-up sequence, indicating seaward
progradation. The lower unit is truncated by an intermittent marine-erosion surface that shows evidence of being
both bioeroded and burrowed suggesting that in some
areas the surface was a firm ground and not completely
cemented. This surface is directly overlain by a second
but more limited fringing-reef tract consisting predominantly of reef-crest and reef-front units with a diverse
assemblage of in-place corals (Bruggemann et al., 2004).
These two superimposed reef tracts therefore have
a back-stepping architecture, which implies that the first
stage of reef development was abruptly terminated and
was followed by the relocation and reinitiation of the second reef-crest further inland over the former lagoon.
Bruggemann et al. (2004) interpreted this reef demise
and relocation to have resulted from a brief eustatic SL
fall, but could not find any evidence of subaerial exposure
associated with the marine-erosion surface. The discontinuous nature of the erosion surface, and the fact that the surface was not completely cemented, is inconsistent with SL
fall, and in fact suggests the opposite scenario whereby
reef back-stepping may have been the result of a rapid rise
of relative SL. Whatever the interpretation, however, the
neotectonic setting of the site means that the possibility
that coseismic displacement was the cause of reef backstepping cannot be discounted.
Along the Egyptian coast, in the northern Red Sea, two
episodes of deposition during the LIG have been
described by Plaziat and others (Plaziat et al., 1998a, b;
Orszag-Sperber et al., 2001). During the first stage,
fringing-reef tracts topped by beach gravels developed
along the coast at elevations between þ5 and 10 m. This
stage was followed by a down-cutting erosive event that
excavated small basins behind breaks in the reef crest.
The first depositional event in these depressions was
a thin sequence of open-marine sediments containing molluscs and in-place corals. This was followed by a thicker
sequence of laminated subaqueous gypsum topped by
tepees that reached an elevation of $3 m below the reefcrest. The final filling unit consisted of massive gypsum
consistent with an emergent sabkha environment.
This unusual sedimentary sequence was interpreted to
result from a rapid SL fall of $10 m which caused the
excavation of the back-reef depressions (Orszag-Sperber
et al., 2001). The subsequent rise of SL back to the same
level as before (within 1–3 m) led to the filling of the
depressions. The abrupt transition from an open marine
to restricted deposits in this fill-sequence was postulated
to result from a brief pluvial episode that increased input of sediment to the coast and occluded the gap in the
fringing-reef unit, thus sealing the depressions off from
the marine environment (Plaziat et al., 1998a; OrszagSperber et al., 2001). However, other interpretations are
also consistent with these stratigraphic data. For example,
down-cutting could result from a rapid SL rise and
reworking of back-reef deposits during storms and sediment export through the reef gaps. These open-marine
conditions would favor the initial deposition of openmarine faunas and the abrupt change to restricted environments would mark the start of SL fall at the end of the LIG.
That no deepening of these depressions occurred during
the ensuing glacial fall in SL, lends support to this latter
interpretation. However, confirmation of either interpretation requires a precise chronological framework which is
still lacking in this region.
Reef development in more stable terrains
Reef development in stable sites can provide a detailed
look into the timing and duration of peak interglacial conditions during the highstand, as well as the absolute elevation and SL behavior during that highstand. Detailed
descriptions of LIG reefs in stable areas have traditionally
concentrated along the southern passive margin of the
North-American Plate, including sites in Florida, the
Bahamas, the Caymans, and the Yucatan. But recent work
has also been reported along the Western Australian passive margin of the Indo-Australian Plate from Cape Range
to Cape Leeuwin. This facilitates a comparison of reef
development within each region and between regions
themselves.
Florida
Pleistocene reef development in Florida’s Key Largo
Limestone was described as early as the late eighteenth
century (Agassiz, 1896), but it was not until the mid1960s that U-series dating proved that those elevated reef
deposits were of LIG age (Broecker and Thurber, 1965;
Osmond et al., 1965). As its name suggests, the reefal
Key Largo Limestone is best developed along Key Largo
and Windley Key where it forms a broad ridge that rises to
a level of þ3 m (Harrison and Coniglio, 1985). Although
widely cited to reach 5.5 m above mean SL, the original
description states that it has an average elevation of
þ1 m “with local prominences which in two places reach
a maximum height of 18 ft” (5.5 m) (Hoffmeister and
Multer, 1964, p. 355). At the type locality on Windley
Key, the reef unit covers an older topographic high at
þ1 m, and forms a 2 m thick layer of coalesced patch reefs
formed of head corals, particularly Montastrea annularis,
interspersed with thickets of A. cervicornis and small
branched Porites (Stanley, 1966). The matrix is a skeletal
sand composed of mostly Halimeda and mollusc fragments. The reef unit gradually dips away from this topographic high and, at its northern and southern outcrop
limits, is overlain by or grades into the oolitic sands of
the Miami Limestone (Multer et al., 2002). These oolitic
units have developed shorelines consistent with a þ6 m SL
highstand at both Key West in the south and Palm Beach
in the north (Hoffmeister and Multer, 1968; Perkins,
1977; Multer et al., 2002). The absence of ooids in the
reef matrix and the fact that the reef unit underlies these
oolitic sand bodies indicates that reef development likely
LAST INTERGLACIAL AND REEF DEVELOPMENT
627