nature of the reef surfaces, they argued that the fall from
the brief highstand must have been as rapid as the preceding rise, and that the reefs therefore had little time for modification in the marine environment. The brief duration of
the þ6 m highstand was contested by Carew (1997), who
suggested that a slower rate for notch erosion was more
representative and gave a duration of 3,000 years as
a better estimate. He also questioned the significance of
reef preservation indicating that the well-preserved reefs
were more likely to have been rapidly entombed by
mobile subtidal sand bodies before SL regression.
Grand Cayman
An almost identical sequence of reef development has
been reported from Grand Cayman (Jones and Hunter,
1990; Hunter and Jones, 1996; Coyne et al., 2007), where
a lower lagoonal patch-reef unit developed over a wide
area but never exceeded an elevation of þ3 m (Jones
and Hunter [1990] reported patch reefs up to 3.8 m, but
these were later corrected to 2.8 m in Jones [1994, p. 32]
and Coyne et al. [2007, p. 550]). The patch reefs in this
unit were protected to the northwest by a reef tract, also
with a þ3 m elevation. In most areas, the lower patch-reef
unit is overlain by cross-bedded oolitic sands that develop
up to an elevation of þ6 m. The base of cross-bedded unit
is marked by a discontinuous marine-erosion surface
which is consistent with an abrupt switch to a high-energy
subtidal to intertidal setting. Although Jones and Hunter
(1990) interpreted this sequence to represent a simple
shallowing-up succession, they omitted to explain why
reef development was restricted to þ3 m. They also
suggested that the erosion surface represented scour during storms, but did not explain why storm waves and
shifting sands were suddenly able to invade large areas
of an extensive and apparently well-protected lagoon
where patch reefs with delicate branching corals had previously flourished.
Later Coyne et al. (2007) suggested that cross-bedded
units were younger based on
230 Th ages of marine
gastropods. Unfortunately, these ages failed any reasonable measure of isotopic reliability due to open-system
diagenesis and were given priority over basic stratigraphic
relations, which showed conformable transitions between
the patch reef and cross-bedded units in several areas
(including newly reported cores from George Town Harbor). As a result, the claim that these units are significantly
younger is unsupported by reliable data.
Along the northeast coast of Grand Cayman, Vèzina
et al. (1999) also reported a þ3 m patch-reef sequence of
LIG age adjacent to a relict sea-cliff with a notch at
þ6 m (Figure 3). Using closely spaced cores, they demonstrated that reef development onlapped the cliff base and
reached an elevation of þ2.6 m. In the outer part of the
section, patch reefs aggraded vertically and then prograded
horizontally seaward, as if constrained by a stillstand of
sea level (Vèzina et al., 1999). The lack of reef development or any sediment above þ3 m was interpreted as
being due to submergence in a lagoonal setting with a
þ6 m SL. However, in the absence of a preserved reefcrest unit (due to modern coastal erosion) this interpretation is equivocal and ignores the progradation of patch
reefs and the fact that, in modern lagoons, patch reefs
commonly reach mean low water.
From the foregoing, it is clear that reef sequences in
Florida, the Bahamas and the Caymans flourished up to
an elevation of þ3 m but no higher. The fundamental
question is why were reefs restricted to this elevation
when there is abundant evidence that SL peaked at þ6
m? The claim of Neumann and Hearty (1996) that sea
level jumped to þ6 m is a possible explanation, but falls
short in accounting for why reefs did not recover. Even
if the rise was transient, and lasted 600 years as claimed,
reefs could have accreted 6 m in that time assuming
a conservative accretion rate of 10 mm/years. Clearly
something happened during the rise from þ3 to þ6 m that
inhibited reef development in these areas.
Northeast Yucatan Peninsula
A possible answer to this problem has recently been provided by Blanchon et al. (2009), who reported two
superimposed fossil reef-crest units from the northeast
Yucatan, one at þ3 m and the other at þ6 m (Figure 3).
Reliable radiometric ages confirmed that both crests were
of LIG age, between 125 and 117 ka, but could not differentiate between them due to subtle diagenetic alteration of
corals in the lower unit. The relative-age of the two units,
however, could be determined from their facies differences and contact relationships. The lower-reef tract at
þ3 m consisted of an A. palmata crest flanked by an
A. cervicornis-dominated reef-front and a large lagoonal
patch-reef complex that extended 200 m inland. Framework in the lower-tract had a cap of coralline algae and,
in places, was leveled by a discontinuous marine-erosion
surface. The upper-reef tract at þ6 m initiated atop the seaward edge of the lagoonal patch-reef complex. But in this
new location reef development was significantly different.
In the lagoon, frameworks were dominated by a sedimenttolerant assemblage of branching coralline algae and
Porites. They also lacked a cap of corallines, and were
instead infiltrated by beach-derived shell-gravels from
regressive shoreface deposits above. This infiltration pattern shows that the upper reef must be younger and was
active shortly before SL fell from its highstand.
The cap of corallines on the older þ3 m crest attests to
its rapid demise and the discontinuous marine-erosion surface indicates that a patchy demise of corals also took
place in the lagoon. Importantly, however, areas of continuous accretion between the lower patch-reef complex and
the overlying tract require that lower-reef demise was ecologically synchronous with upper-reef initiation. Furthermore, in the upper tract, reef-flat development above
þ5 m and large in-place colonies of A. palmata at the base
of the upper-crest unit indicated that SL must have jumped
from þ3 m to more than þ5 m to drown the lower reef and
630
LAST INTERGLACIAL AND REEF DEVELOPMENT
the brief highstand must have been as rapid as the preceding rise, and that the reefs therefore had little time for modification in the marine environment. The brief duration of
the þ6 m highstand was contested by Carew (1997), who
suggested that a slower rate for notch erosion was more
representative and gave a duration of 3,000 years as
a better estimate. He also questioned the significance of
reef preservation indicating that the well-preserved reefs
were more likely to have been rapidly entombed by
mobile subtidal sand bodies before SL regression.
Grand Cayman
An almost identical sequence of reef development has
been reported from Grand Cayman (Jones and Hunter,
1990; Hunter and Jones, 1996; Coyne et al., 2007), where
a lower lagoonal patch-reef unit developed over a wide
area but never exceeded an elevation of þ3 m (Jones
and Hunter [1990] reported patch reefs up to 3.8 m, but
these were later corrected to 2.8 m in Jones [1994, p. 32]
and Coyne et al. [2007, p. 550]). The patch reefs in this
unit were protected to the northwest by a reef tract, also
with a þ3 m elevation. In most areas, the lower patch-reef
unit is overlain by cross-bedded oolitic sands that develop
up to an elevation of þ6 m. The base of cross-bedded unit
is marked by a discontinuous marine-erosion surface
which is consistent with an abrupt switch to a high-energy
subtidal to intertidal setting. Although Jones and Hunter
(1990) interpreted this sequence to represent a simple
shallowing-up succession, they omitted to explain why
reef development was restricted to þ3 m. They also
suggested that the erosion surface represented scour during storms, but did not explain why storm waves and
shifting sands were suddenly able to invade large areas
of an extensive and apparently well-protected lagoon
where patch reefs with delicate branching corals had previously flourished.
Later Coyne et al. (2007) suggested that cross-bedded
units were younger based on
230 Th ages of marine
gastropods. Unfortunately, these ages failed any reasonable measure of isotopic reliability due to open-system
diagenesis and were given priority over basic stratigraphic
relations, which showed conformable transitions between
the patch reef and cross-bedded units in several areas
(including newly reported cores from George Town Harbor). As a result, the claim that these units are significantly
younger is unsupported by reliable data.
Along the northeast coast of Grand Cayman, Vèzina
et al. (1999) also reported a þ3 m patch-reef sequence of
LIG age adjacent to a relict sea-cliff with a notch at
þ6 m (Figure 3). Using closely spaced cores, they demonstrated that reef development onlapped the cliff base and
reached an elevation of þ2.6 m. In the outer part of the
section, patch reefs aggraded vertically and then prograded
horizontally seaward, as if constrained by a stillstand of
sea level (Vèzina et al., 1999). The lack of reef development or any sediment above þ3 m was interpreted as
being due to submergence in a lagoonal setting with a
þ6 m SL. However, in the absence of a preserved reefcrest unit (due to modern coastal erosion) this interpretation is equivocal and ignores the progradation of patch
reefs and the fact that, in modern lagoons, patch reefs
commonly reach mean low water.
From the foregoing, it is clear that reef sequences in
Florida, the Bahamas and the Caymans flourished up to
an elevation of þ3 m but no higher. The fundamental
question is why were reefs restricted to this elevation
when there is abundant evidence that SL peaked at þ6
m? The claim of Neumann and Hearty (1996) that sea
level jumped to þ6 m is a possible explanation, but falls
short in accounting for why reefs did not recover. Even
if the rise was transient, and lasted 600 years as claimed,
reefs could have accreted 6 m in that time assuming
a conservative accretion rate of 10 mm/years. Clearly
something happened during the rise from þ3 to þ6 m that
inhibited reef development in these areas.
Northeast Yucatan Peninsula
A possible answer to this problem has recently been provided by Blanchon et al. (2009), who reported two
superimposed fossil reef-crest units from the northeast
Yucatan, one at þ3 m and the other at þ6 m (Figure 3).
Reliable radiometric ages confirmed that both crests were
of LIG age, between 125 and 117 ka, but could not differentiate between them due to subtle diagenetic alteration of
corals in the lower unit. The relative-age of the two units,
however, could be determined from their facies differences and contact relationships. The lower-reef tract at
þ3 m consisted of an A. palmata crest flanked by an
A. cervicornis-dominated reef-front and a large lagoonal
patch-reef complex that extended 200 m inland. Framework in the lower-tract had a cap of coralline algae and,
in places, was leveled by a discontinuous marine-erosion
surface. The upper-reef tract at þ6 m initiated atop the seaward edge of the lagoonal patch-reef complex. But in this
new location reef development was significantly different.
In the lagoon, frameworks were dominated by a sedimenttolerant assemblage of branching coralline algae and
Porites. They also lacked a cap of corallines, and were
instead infiltrated by beach-derived shell-gravels from
regressive shoreface deposits above. This infiltration pattern shows that the upper reef must be younger and was
active shortly before SL fell from its highstand.
The cap of corallines on the older þ3 m crest attests to
its rapid demise and the discontinuous marine-erosion surface indicates that a patchy demise of corals also took
place in the lagoon. Importantly, however, areas of continuous accretion between the lower patch-reef complex and
the overlying tract require that lower-reef demise was ecologically synchronous with upper-reef initiation. Furthermore, in the upper tract, reef-flat development above
þ5 m and large in-place colonies of A. palmata at the base
of the upper-crest unit indicated that SL must have jumped
from þ3 m to more than þ5 m to drown the lower reef and
630
LAST INTERGLACIAL AND REEF DEVELOPMENT