Geology Reference
In-Depth Information
15
6
10
A. palmata
All
Massive
4
A. palmata
Massive
2
5
0
25
20
15
10
5
0
0
Palaeo-depth (m)
25
20
15
10
5
0
A. palmata
Massive
All
Fig. 7.
Reef accretion versus palaeowater depth for
branching (red dots) versus massive corals (green squares).
Approximate limits for branching and massive facies
(shaded) refl ect a general overlap in the depth ranges for
the two facies, but with a slight tendency for branching
Acropora palmata
to inhabit shallower water than massive
species. The linear regression lines for branching (red
solid:
y
= 0.02
x
+ 3.98 -
R
2
= 0.001), massive (green dashed:
y
= 0.02
x
+ 3.20;
R
2
= 0.0001) and all corals combined
(black dotted:
y
= 0.03
x
+ 3.52;
R
2
= 0.003) show no
strong depth-related relationship and
R
2
values are very
low, refl ecting the large variance across all depth ranges
(Fig. 8). The difference between the regression lines for
branching (solid) and massive corals (dashed) is not
statistically signifi cant.
Fig. 8.
Reef accretion versus palaeowater depth in 5-m
increments. The depth-related trend is weak (lines shown
here are based on linear regression) regardless of coral
type, and standard deviations are larger than the differ-
ence between mean accretion rates at either depth extreme.
The anomalously high accretion rate for
A. palmata
between 20 and 25 m is based on only one core interval.
mean accretion rate was below that threshold.
Of those fi ve intervals, only one was associated
with a vertical section of signifi cant thickness
(
A. palmata
facies at
d
=
1.4 m).
No strong depth-related pattern emerged
either by facies type or for all corals combined.
The highest
R
2
value for linear regression was
only 0.003. Exponential and polynomial best-fi t
lines yielded similarly poor results. The differ-
ence between means for shallow- and deep-water
accretion (<1.0 m kyr
1
) was lower than standard
deviations within 5-m depth intervals (Fig. 8).
The mean accretion rate for
A. palmata
inter-
vals was faster than that for massive-coral facies
(3.83 vs. 3.07 m kyr
1
), but this difference was not
signifi cant at
DISCUSSION
The higher concentration of measured intervals
in palaeowater depths less than 15 m is consistent
with the presumption that light intensity exerts a
primary control on coral growth, and by extension,
reef accretion. Also, the tendency for massive-
coral intervals to extend into deeper water than
branching-coral sections (compare shaded areas
in Fig. 7) mimics modern reef zonation. However,
neither the species- nor depth-related differences
in average accretion rate that are the foundations
of most modern-reef models clearly emerge from
the data.
Schlager (1981) proposed that shallow reefs
(set at
d
< 5 m) accrete between 1 and 20 m kyr
1
,
compared to less than 2 m kyr
1
at depths between
10 and 20 m (Fig. 2). The depth-related model of
Bosscher (1992; Fig. 1) was built on these assump-
tions, as are most discussions of reef accretion
versus sea-level rise, both past (Macintyre, 1988)
and future (Graus & Macintyre, 1999). Figure 9
compares these presumptions to the results of this
study. The maximum 'shallow-water' accretion
rate (16.6 m kyr
1
) was higher than that for 'deep
reefs' (9.73 m kyr
1
). However, the majority of the
accretion rates for the Caribbean cores analysed
= 0.05 using a Student's
t
-test,
owing to these large variances. Additional data
could raise the signifi cance levels for this com-
parison, but the absolute difference would remain
small, and certainly not approach the relationship
that has been presumed for the two coral types.
Accretion rate was weakly and positively cor-
related with the rate of sea-level rise, but the
R
2
value for the relationship was only 0.015. Only 16
of the measured intervals accreted at rates greater
than 7 m kyr
1
, the upper threshold for Holocene
sea-level rise proposed by Schlager (1981); these
were equally distributed between 'deep' and
'shallow' reefs. In all but fi ve cases, the intervals
of unusually high accretion were small and were
contained within a larger interval over which
