Coral Reefs (2001) 20: 95±105

DOI 10.1007/s003380000135

R EP O RT

Ian G. Macintyre á Peter W. Glynn á Robert S. Steneck

A classic Caribbean algal ridge, HolandeÂs Cays, PanamaÂ:

an algal coated storm deposit

Accepted: 27 November 2000 / Published online: 17 March 2001

Ó Springer-Verlag 2001

Abstract Twenty-seven radiocarbon dates of cores re- 1969). It was the ®eld work of Adey and his colleagues
covered from six drill holes indicate that the relief of the (Adey 1975, 1978; Adey and Burke 1976, 1977; Steneck
ridge on the seaward edge of the HolandeÂs Cays, San and Adey 1976; Adey et al. 1977) that focused attention
Blas, o€ the Caribbean coast of PanamaÂ, was formed by on the algal ridges of this area. Although these studies
storm deposits about 2,000 to 2,800 years ago. Although reported the locations of many new algal ridges in the
crustose coralline algae are a dominant component of Caribbean, it was Glynn (1973) who was the ®rst to
the surface cover on this outer ridge, they played a mi- describe the HolandeÂs Cays outer ridge system as an
nor role in the construction of the framework of this algal ridge that ``conclusively con®rms the existence of
bioherm, which therefore cannot be classi®ed as an algal an algal ridge formation in this part of the Caribbean
ridge. The framework of the ridge consists dominantly Sea'' (p. 285). Milliman (1974) even went further to
of Agaricia/Millepora rubble that is extensively lithi®ed suggest that this HolandeÂs ridge complex is an ``algal-
by micritic submarine Mg-calcite cement. The present- ridge system in the coral reefs o€ Panama that appar-
day surface of this area in the HolandeÂs Cays is ently is very similar to those in the Indo-Paci®c'' (p. 166).
primarily one of widespread bioerosion with very little Adey (1978), in a review of the Caribbean±West In-
indication of substrate accumulation over the past dian algal ridges, demonstrated that the distribution of
2,000 years. algal ridges in this area is directly related to what he
refers to as the ``seasonal wind e€ect'' (p. 366), which is
Keywords Algal ridge á Storm deposits á Submarine derived by multiplying the percent of wind constancy by
lithi®cation á Crustose coralline algae the Beaufort wind strength. In his graph of algal ridge
frequency versus seasonal wind e€ects, he noted that
``the strong development of algal ridges in the HolandeÂs
Introduction Cays in Panama is anomalously high for that region''
(p. 361).
Up until the 1970s, it was generally thought that algal In this study we recovered cores from the internal
ridges did not occur in Caribbean reefs (e.g., Stoddart structure of the HolandeÂs Cays outer ridge to study its
late Holocene history and to evaluate the contribution of
crustose coralline algae to the relief of this ridge system.

I. G. Macintyre (&)
Department of Paleobiology, MRC 125,
National Museum of Natural History,
Setting
Smithsonian Institution, Washington, DC 20560, USA
E-mail: macintyre.ian@nmnh.si.edu The HolandeÂs coral reef, a bank barrier system, located
Tel.: +1-202-3572580 about 15 km o€ the northeast coast of Panama (Fig. 1),
Fax: +1-202-7862832 is one of the most exposed areas to wave assault in the
P. W. Glynn San Blas region. With the reef front facing towards the
Division of Marine Biology and Fisheries, north, seaward reef zones are fully exposed to north and
Rosenstiel School of Marine and Atmospheric Science,
University of Miami, 4600 Rickenbacker Causeway, northeast open ocean waves. Northeast trade winds are
Miami, Florida 33149, USA especially well developed during the dry season, bu€et-
R. S. Steneck
ing this coastal area from December through April.
School of Marine Sciences, University of Maine, High seas also develop unpredictably and for shorter
Darling Marine Center, Walpole, Maine 04573, USA periods during the wet season (May to November).
96

Fig. 1 Index map of HolandeÂs Cays in the San Blas Islands. Inset shallow reef ¯at directly leeward of the ridge (Fig. 1). The cores,
of aerial photograph shows details of study site on the outer ridge which have a diameter of 54 mm, were collected with a hand-
with locations of the six core holes. The parapet that marks the operated hydraulic drill (Macintyre 1978) using a tripod for sup-
inner limit of the ridge is the dark line that runs between core holes port (Fig. 2). The hydraulic power unit and the water pump were
5 and 2 placed on a specially constructed six-barrel barge that was towed
out from the research vessel across the shallow back reef and
anchored in the lee of the ridge. The deepest core hole was drilled to
Tropical storms and hurricanes do not often occur at a depth of 4.8 m and core recovery ranged from 100% in
PanamaÂ's low latitudinal position (9±10°N; Stoddart well-lithi®ed sections to no recovery in sandy sections.
Following the drilling, the topography, location of core holes,
1971; Glynn 1973). During the past 120 years, only a and distribution of bottom communities were surveyed along the
single hurricane has moved across the northwest coast of research transect. Surface samples were also collected at core-hole
Panama (Clifton et al. 1997). Although relatively far sites. Bottom cover was determined at core-hole sites 1±3 by
from the coast, the seaward bottom topography slopes counting the organisms and sediment type in contact with chain
links (3.5 cm/link) laid in straight lines of 10 m length on both sides
gradually toward the north with the 200-m isobath and perpendicular to the drilling transect. Dead coral was classi®ed
located about 2.5 km o€shore. The reef ¯at is typically as limestone substrate.
shallow (1±3 m) and quite broad, ranging between 1 and Thin sections for both petrographic and crustose coralline al-
2 km in width. gal studies were prepared from each of the cores at intervals se-
lected to represent the dominant subsurface lithologies. All thin
sections with coralline algae were examined with a microscope and
every identi®able fragment was identi®ed to genus or species. The
Methods taxonomic scheme of Adey (1970) was used. Taxonomic charac-
teristics identi®ed in Adey and Macintyre (1973) were evident in
Fieldwork for this project was carried out in June and July of 1993. most thin sections (Braga et al. 1993). When only generic deter-
Working on the outer ridge of HolandeÂs Cays was a very dicult minations could be made, they were included in the total species
undertaking. Of the 10 days that we spent on the site, there were count if other species of that genus had not been found in the
only 2 days in which the surf was low enough to allow us to set up cores (e.g., Titanoderma and Mesophyllum) or if there was a dis-
our drilling equipment to collect cores. All members of our drill tinguishing feature that indicated that it was a di€erent species
team were swept o€ their feet at one time or another. (e.g., Lithophyllum ``unbranched'' is morphologically and taxo-
Our research transect was located across the outer ridge at the nomically distinct from Lithophyllum congestum). Four other
east end of HolandeÂs Cays (Fig. 1) where we could ®nd safe specimens that could only be identi®ed to genus (Hydrolithon,
anchorage for our research vessel, the M/Y Catyani. A total of six Neogoniolithon, Paragoniolithon, and Lithothamnion) may repre-
core holes were drilled, three on the outer ridge and three on the sent taxa already identi®ed and thus are not counted in the species
 97

Fig. 2 Drilling core hole 6 at

the outer edge of the ridge in
heavy surf. Note the waves
breaking on the outer lip of this
ridge behind the drillers

Table 1 Coralline and associated ¯ora identi®ed from cores subdivided by their family and subfamily anities. Number of each hole
indicates the number of specimens identi®ed

Families and subfamilies Species Taxonomic authority Hole 6 Hole 4 Hole 5 Hole 2 Hole 1

Mastophoroideae Goniolithon improcerum (Foslie et Howe) Foslie 1 1 ± ± ±

Hydrolithon boergesenii (Foslie) Foslie 1 1 ± ± ±
Hydrolithon sp. ± 1 1 ± ± ±
Lithoporella atlantica (Foslie) Foslie ± 1 ± ± ±
Neogoniolithon erosum (Foslie) Adey ± ± 1 ± ±
N. dispalatum (Foslie et Howe) Adey ± 1 ± ± ±
N. mamillare (Harvey) Setchell et Mason 1 ± ± ± ±
N. munitum (Foslie et Howe) 1 ± 2 1 ±
Neogoniolithon sp. ± ± 1 2 ± ±
N. strictum (Foslie) ± ± ± ± 2
Paragoniolithon solubile (Foslie et Howe) 1 ± ± ± 1
Paragoniolithon sp. ± ± ± ± 1 1
Porolithon pachydermum (Foslie) Foslie 2 1 ± ± 1
Lithophylloideae Lithophyllum congestum (Foslie) Foslie 1 2 2 3 2
Lithophyllum (unbranched) ± 4 ± ± ± ±
Titanoderma sp. ± 1 ± ± 1 1
Melobesiodeae Lithothamnion ruptile (Foslie) Foslie ± 1 ± ± ±
Lithothamnion sp. ± 1 ± ± ± ±
Mesophyllum sp. ± 4 2 ± ± 1
Sporolithon dimotum Foslie et Howe ± ± ± 1 ±
Peysonnelliaceae Peysonnellia sp. ± ± ± ± ± 1
Corallinaceae Amphiroa sp. ± 3 2 1 1 ±
Codiaceae Halimeda sp. ± 1 1 ± 1 ±
Total taxa per hole ± ± 15 12 5 6 8

total. In addition to the crustose corallines, one non-coralline and cryptic (i.e., shaded) fore reef habitats. There was surprising
crust (Peysonnellia sp.) and two articulated erect calcareous algae little overlap of species among reef habitats. Often several species
(Amphiroa spp. and Halimeda spp.) were found in the sectioned indicative of particular conditions of disturbance and productivity
material (Table 1). will share anatomical and morphological characteristics (Steneck
Coralline species assemblages are indicative of speci®c habitats, 1986). Based on ®ve published studies (Adey 1975; Adey and
microhabitats, and zones on coral reefs and algal ridges. To Vassar 1975; Steneck and Adey 1976; Adey et al. 1977; Bosence
determine coralline functional groupings (i.e., polyphyletic suites of 1984), only three coralline species dominate exposed algal ridges
species that share ecological characteristics and play equivalent (i.e., Porolithon pachydermum, Lithophyllum congestum, and Neo-
roles in natural communities), we reviewed literature that identi®ed goniolithon mamillare). We have quantitatively examined the cor-
the coralline species and their habitats. The tabulation revealed allines from the six cores to determine if any of the zones show
three to nine species characteristic of algal ridge habitats ``exposed'' strong dominance of the algal ridge-building functional group of
to direct sunlight, as well as some characteristic of both exposed crustose coralline algae.
98

Carbonate mineralogy was determined by a Scintag X-ray dif-

fractometer with Cu Ka radiation of samples that were crushed and
smeared on a zero-background quartz mounting plate. Peak areas
were used to determine the relative proportions of aragonite,
calcite, and Mg-calcite, and the mol% MgCO3 in Mg-calcite was
estimated from the d(211) spacing of the calcite peak (Goldsmith
and Graf 1958). Fluorite peak positions were used as internal
standards to correct the positions of the (211) calcite peaks.
Radiocarbon dates were determined for 27 samples by Beta
Analytic Inc. using standard techniques. Dates are reported as
radiocarbon years before A.D. 1950, conventionally termed ``before
present'' (B.P.), using a Libby half-life of 5,568 years and a modern
standard based on 95% of the activity of the National Bureau of
Standards oxalic acid. No corrections were made for the DeVries
e€ect, reservoir e€ect, or for natural isotopic fractionation. The
quoted errors are from the counting of the modern standard,
background, and sample being analyzed. They represent one
standard deviation statistics (68% probability), based on the Fig. 3 Section through a lip at the outer edge of the ridge showing
random nature of the radioactive disintegration process. extensively bored and micrite-®lled pavement limestone with a thin
The descriptive term ``micrite'' is used in this paper to refer to upper surface cover comprised of a massive single thallus of
microcrystalline carbonate that is generally 4 lm or ®ner and Porolithon pachydermum with enclosed vermetid tubes and the
appears in thin sections as a dense or translucent groundmass, underside coated by thin leafy crusts of predominantly Titan-
following Folk (1974). odermma sp.

coralline algae such as Lithophyllum sp. (unbranched),

Results Neogoniolithon sp., and Porolithon pachydermum (Fos-
lie) Foslie formed a massive smooth pavement, and in
Present-day surface communities sheltered microhabitats commonly Titanoderma sp.
formed thin crusts (Fig. 3). Macroalgae, such as Turbi-
Spurs and grooves dominate the seaward topography of naria turbinata (Linnaeus) Kuntze, Laurencia papillosa
the algal ridge to depths of 8±10 m. The e€ects of (ForsskaÊl) Greville, Sargassum polyceratium Montagne,
scouring are evident, especially on surfaces near sand- and Padina gymnospora (KuÈtzing), were all abundant in
®lled grooves. During previous surveys (Glynn 1973; the shallow (5±10 cm depth) terraced pools. Certain
Dahl et al. 1974), several reef-building coral species were chitons (Chiton marmoratus Gmelin) and snails [Cittar-
present on lithi®ed pavement surfaces but they were ium pica (Linnaeus 1758)] were often found in depres-
dispersed and thus did not form any obvious frame- sions and holes under extreme conditions of exposure.
works. The following scleractinian corals were observed The 2±5 cm high rims of the terraced pools are in large
in the windward algal ridge habitat: Acropora palmata part built up by vermetid [Dendropoma (Novastoa) a€.
(Lamarck), Porites astreoides Lamarck, Siderastrea D. (N.) irregulare (Orbigny)] tubes in a crustose coralline
siderea (Ellis and Solander), Agaricia agaricites (Linna- matrix. Several vagile, invertebrate species are abundant
eus), Favia fragum (Esper), Diploria clivosa (Ellis and on areas of the algal ridge downstream of the breaker
Solander), Diploria strigosa (Dana), and Montastraea line. Some of these groups include grazing mollusks
annularis (Ellis and Solander). Some colonies of Porites (Acanthopleura granulata Gmelin, Fissurella nodosa
astreoides and Agaricia agaricites were found at 1±2 m Born, Nerita spp., and Littorina ziczac Gmelin), gas-
depth in the zone of greatest wave assault. No hydro- tropod predators (Thais haemastoma ¯oridana Conrad
corals (Millepora spp.) were observed on the windward
reef face.
Fig. 4 Live coral cover and other substrate types at three reef-¯at
Areas of the algal ridge that were intermittently core sites: HO Halimeda opuntia; SR Siderastrea radians; DL
exposed were inhabited by numerous algae and inver- Diploria labyrinthiformis; PF Porites furcata; AA Agaricia agari-
tebrate species. On wave-swept surfaces, crustose cites; PA Porites astreoides; LS limestone; SB sand bottom
 99

and Purpura patula LinneÂ), and an echinoid borer Fig. 5 Cross section of the outer edge of HolandeÂs Cays showing
[Echinometra lucunter (Linnaeus)]. present-day surface zonation, position of numbered core holes 1±6,
distribution of reef facies, and location of radiocarbon-dated sam-
Zooxanthellate corals are abundant at shallow depths ples (indicated by dots to the right of the column). Horizontal bars
(1±1.5 m) on the leeward reef ¯at. Sampling of the coral indicate core interval depths in each hole. T.D. Total depth of hole
cover at drilling sites 1, 2, and 3 revealed live cover
values for all species combined that ranged from 32.2% directly shoreward of the HolandeÂs ridge system
(hole 1) to 52.1% (hole 3) with no trend in coral abun- (Newman et al. 1970).
dance from immediately behind the parapet to about 7 m The facies in the Holocene sections are as follows:
to leeward (Fig. 4). The most abundant species were
Agaricia agaricites and Porites astreoides. Other corals · Agaricia/Millepora pavement limestone: This is the
present in the area, but not sampled, included Sideras- dominant facies of the ridge and is also present in two
trea siderea, Favia fragum, Montastraea annularis, and of the leeward reef-¯at core holes (Fig. 5). It consists
the hydrocorals Millepora alcicornis Linnaeus and of an agglomeration of Agaricia sp. and Millepora sp.
Millepora complanata Lamarck.

Internal structure

Five basic reef facies were identi®ed in the six core holes
that extend from the outer edge of the ridge complex
into the leeward reef ¯at (Fig. 5). The pre-Holocene
limestone substrate was only recovered from the bases
of two core holes (1 and 2), which were drilled into the
leeward reef ¯at (Fig. 5). This dense calcirudite/calcar-
enite limestone consists mostly of branching coralline
algae, with some mollusks, Halimeda sp., foraminifers,
coral, and echinoid debris. The matrix consists of silt-
rich peloidal microcrystalline micrite. Much of this
matrix, and some aragonitic grains, especially the
Fig. 6 Photomicrograph of Pleistocene limestone substrate show-
mollusks, have recrystallized to form a ®ne to coarse ing recrystallized blocky calcite mosaic of mollusk fragment
sparry texture (Fig. 6). This limestone is probably (originally aragonite) in comparison to well-preserved crustose
Pleistocene in age, similar to the limestone islands coralline algae (originally Mg-calcite). Hole 1, Core 3±4
100

Fig. 7 Densely cemented Agar-

icia/Millepora pavement lime-
stone. An extensively bored,
in®lled, and lithi®ed agglomer-
ation of predominantly Agari-
cia sp. in a sediment-rich
micrite matrix. A Agaricia sp.;
C crustose coralline algae with
serpulid tubes; S lithi®ed sedi-
ment in®ll (large fragments are
dominantly Halimeda); M sedi-
ment-rich micritic Mg-calcite.
Note the dark brown coating on
an exposed section of a hiatal
corroded surface (H) at the top
of this core. Hole 6, Core 2±2

· Acropora palmata: This facies is almost entirely lim-

ited to the core holes drilled in the leeward reef ¯at
(Fig. 5), with a notable exception located near the
base of the outermost core hole on the ridge (core hole
6). It consists of well-preserved Acropora palmata with
light encrustations of crustose coralline algae and the
encrusting Homotrema rubrum (Lamarck) foraminif-
er, with the exception being a core hole 6 section that
has a thick (up to 4 mm) crust of crustose coralline
algae with associated Homotrema sp. and vermetids
all extensively ®lled by micrite. The corallite growth
patterns of the A. palmata colonies indicate that in all
the cores, with the possible exception of cores in hole
3, the corals have not been overturned and are
Fig. 8 Photomicrograph of Agaricia/Millepora pavement lime- therefore probably in growth position. All submarine
stone showing remnants of original coral skeletal framework lithi®cation, in the form of peloidal micrite, is mostly
``¯oating'' in a dense sediment-rich peloidal micritic Mg-calcite limited to the outer edges of the coral skeletons.
matrix. Note scalloped edges of coral, which is indicative of sponge · Massive corals: This facies is only located in core hole
borings. Hole 6, Core 3±2
1 that was drilled into the leeward reef ¯at (Fig. 5),
and consists of well-preserved colonies of Diploria
commonly encrusted by crustose coralline algae, strigosa and Porites astreoides. Some of these colonies
foraminifers, vermetids, and worm tubes (Fig. 7). have light encrustations of crustose coralline algae
Cavities and borings are ®lled with sediment-rich and Homotrema sp.
micrite that commonly exhibits a peloidal texture (20± · Calcirudite/calcarenite: This is very poorly sorted
60 lm). Multicyclic stages of boring, sediment ®lling, sand (63 lm±2 mm) and gravel (>2 mm) sized skel-
and lithi®cation have commonly resulted in the de- etal grains that are well lithi®ed by a silt-rich micrite,
struction of much of the skeletal framework (Fig. 8). which commonly exhibits a peloidal (20±60 lm) tex-
The poorly sorted sediment in®ll ranges in size from ture. The main skeletal components of this lithi®ed
gravel (>2 mm) to silt (4±63 lm) and consists pre- sediment facies are branching crustose coralline algae
dominantly of branching crustose coralline algae, and Halimeda sp., with some mollusks, corals,
mollusks, corals, Halimeda sp., foraminifers, and foraminifers, and echinoids.
echinoids. Hiatal corrosion surfaces with amorphous · Sand: These sections in the core holes were marked by a
black coatings occur on, and are buried in, this ex- rapid penetration of the drill bit and the pumping up of
tensively lithi®ed pavement limestone (Fig. 7). X-ray sand around the core barrel. We assume that these sec-
microprobe energy dispersive scans of this coating tions of no recovery consist of unlithi®ed sand deposits.
indicated a dominance of manganese with traces of
magnesium, aluminum, calcium, and sulfur. No iron
was found, which contrasted with earlier analyses of Crustose coralline algal ¯ora
similar manganese-rich crusts associated with exten-
sively submarine-cemented substrates (James and Corallines were most abundant at the surface of core
Ginsburg 1979; Land and Moore 1980). hole 6, located at the outer edge of this ridge complex.
 101

Table 2 Coralline ¯ora subdivided by functional groupings com- reported for the ®rst time in this study. The number of specimens
monly found in di€erent shallow reef and algal ridge habitats. identi®ed and the percent recorded for each core hole is given for
Corallines from sun±exposed and shaded cavity (``cryptic'') mi- all habitats. Since several taxa are commonly found in several
crohabitats for algal ridges and fore reefs were determined from the habitats, their abundance is given for each habitat in which they are
literature. Reference numbers are: 1 Adey (1975), 2 Adey and commonly found. For total number of specimens found in each
Vassar (1975), 3 Adey et al. (1977), 4 Steneck and Adey (1976), and core hole (in parentheses), see Table 1
5 Bosence (1984); habitats for coralline species not referenced are

Habitat/refs. Percent functional groups Ridge Ridge Moat Leeward reef ¯at
per core hole
Hole 6 Hole 4 Hole 5 Hole 2 Hole 1

Algal ridge exposed

1, 2, 3, 4, 5 Porolithon pachydermum 2 1 ± ± 1
1, 2, 3, 4, 5 Lithophyllum congestum 1 2 2 3 2
2 Neogoniolithon mamillare 1 ± ± ± ±
Total (total no. specimens) 4 (24) 3 (15) 2 (8) 3 (8) 3 (10)
Functional group per hole (%) 17 20 25 38 30
Algal ridge cryptic
2, 3, 5 Lithothamnion ruptile ± 1 ± ± ±
2, 5 Titanoderma sp. 1 ± ± 1 1
3, 5 Mesophyllum sp. 4 2 ± ± 1
2 Neogoniolithon munitum 1 ± 2 1 ±
2 Hydrolithon boergesenii 1 1 ± ± ±
5 Lithophyllum (‹) 4 ± ± ± ±
2 Neogoniolithon dispalatum ± 1 ± ± ±
2 Peysonnellia sp. ± ± ± ± 1
3 Sporolithon dimotum 1 ± ± ± ±
Total (total no. specimens) 12 (24) 5 (15) 2 (8) 2 (8) 3 (10)
Functional group per hole (%) 50 33 25 25 30
Fore reef exposed
1 Neogoniolithon mamillare 1 ± ± ± ±
1 Porolithon pachydermum 2 1 ± ± 1
1 Lithophyllum congestum 1 2 2 3 2
± Neogoniolithon erosum ± ± 1 ± ±
± Neogoniolithon dispalatum ± 1 ± ± ±
± Neogoniolithon munitum 1 ± 2 1 ±
± Paragoniolithon sp. ± ± ± 1 1
Total (total no. specimens) 5 (24) 4 (15) 5 (8) 5 (8) 4 (10)
Functional group per hole (%) 21 27 63 63 40
Fore reef cryptic
1 Hydrolithon boergesenii 1 1 ± ± ±
1 Neogoniolithon munitum 1 ± 2 1 ±
1 Mesophyllum sp. 4 2 ± ± 1
1 Sporolithon dimotum 1 ± ± ± ±
1 Lithothamnion ruptile ± 1 ± ± ±
1 Paragoniolithon solubile 1 ± ± ± 1
± Hydrolithon sp. 1 1 ± ± ±
± Lithothamnion sp. 1 ± ± ± ±
Total (total no. specimens) 10 (24) 5 (15) 2 (8) 1 (8) 2 (10)
Functional group per hole (%) 42 33 25 13 20
Other exposed habitats
± Amphiroa sp. 3 2 1 1 ±
± Goniolithon improcerum 1 1 ± ± ±
± Halimeda sp. 1 1 ± 1 ±
± Lithoporella atlantica ± 1 ± ± ±
± Neogoniolithon sp. ± 1 2 ± ±
± Neogoniolithon strictum ± ± ± ± 2
Total (total no. specimens) 5 (24) 6 (15) 3 (18) 2 (8) 2 (10)
Functional group per hole (%) 21 40 38 25 20

The coralline ¯ora there was dominated by massive phyllum congestum was found in surface samples taken
thalli of mastophoroid (Porolithon and Neogoniolithon) at core hole 4.
and lithophylloid (Lithophyllum and Titanoderma) spe- A total of 16 species of non-geniculate coralline algae
cies. The algal ridge-building coralline species Litho- (``crustose corallines'') distributed among three sub-
102

Table 3 Radiocarbon dates from HolandeÂs Cays samples, Panama

Sample Material dated Estimated depth Radiocarbon date m.l.s.l. correction Depth below
in core hole (m) mean sea levela

Core hole 1
1 Surface sample Agaricia sp. 0 2610‹70 +1.05 1.16
2 Core 2±6 Porites astreoides 1.75 2660‹70 ± 2.91
3 Core 3±1 Acropora palmata 1.90 2660‹60 ± 3.06
4 Core 3±3 Acropora palmata 2.33 2820‹70 ± 3.49
Core hole 2
5 Surface sample Agaricia sp. 0 2410‹70 +0.95 1.06
6 Core 2±1 Acropora palmata 2.10 2620‹60 ± 3.16
7 Core 2±2 Acropora palmata 2.17 2690‹70 ± 3.23
Core hole 3
8 Core 1±1 Agaricia sp. 0.07 2770‹60 +0.85 1.03
9 Core 1±6 Millepora sp. 0.50 2590‹60 ± 1.46
10 Core 1±7 Agaricia sp. 0.72 3140‹60 ± 1.68
11 Core 3±1 Acropora palmata 1.50 2780‹70 ± 2.46
12 Core 4±1 Acropora palmata 1.75 2830‹70 ± 2.71
Core hole 4
13 Surface sample Millepora sp. 0 1790‹60 ±0.40 +0.29
14 Core 1±8 Agaricia sp. 1.47 2140‹60 ± 1.18
15 Core 2±8 Agaricia sp. 2.20 2420‹60 ± 1.91
16 Core 4±2 Millepora sp. 3.70 1880‹60 ± 3.41
Core hole 5
17 Surface sample Diploria strigosa 0 2090‹50 +0.30 0.41
18 Core 1±2 Millepora sp. 0.15 2650‹60 ± 0.56
19 Core 1±8 Millepora sp. 1.33 2900‹60 ± 1.74
20 Core 2±3 Porites astreoides 1.67 2920‹50 ± 2.08
21 Core 2±7 Millepora sp. 2.16 2910‹80 ± 2.57
Core hole 6
22 Core 1±2 Millepora sp. 0.25 2140‹60 ±0.10 0.26
23 Core 1±8 Millepora sp. 1.20 2190‹80 ± 1.21
24 Core 2±3 Agaricia sp. 1.48 2640‹60 ± 1.49
25 Core 3±1 Agaricia sp. 2.00 2540‹60 ± 2.01
26 Core 4±5 Acropora palmata 2.52 2460‹80 ± 2.53
Parapet
27 Inner wall Montastraea annularis 0 1890‹60 ±0.25 +0.14
a
Mean tidal range 0.21 m. Correction for depth below mean s.l.=0.11 m

families were identi®ed from the thin sections made from The only specimen of Neogoniolithon strictum was
the cores (Table 1). Species diversity was greater in cores found at the base of core hole 1. This species often is
taken in the ridge habitats (core holes 4 and 6) than in found in sediment-dominated habitats (Bosence 1985)
the moat (core 5) or leeward reef ¯at (cores 1, 2, 3). No where it can form algal ridges (Steneck et al. 1997).
corallines were found in any samples from core 3. Corallines characteristic of reef and algal ridge hab-
The coralline ¯ora lacks distinct patterns (Table 1). itats that are exposed to full sunlight are phylogeneti-
No single taxon was dominant among the species of cally and morphologically distinct from those that
corallines or the species of non-coralline crust (Peyson- occupy cryptic habitats that are shaded from direct light.
nellia sp.). Overall, the ¯ora was dominated by species in Exposed corallines are often massive or branched and
the subfamily Mastophoroideae that is indicative of a usually dominated by species of the mastophoroid sub-
relatively shallow tropical marine ¯ora. family or a few species of the lithophylloid subfamily.
Porolithon pachydermum and Lithophyllum conge- These corallines are capable of rapid growth and skeletal
stum, characteristic of exposed algal ridges (Steneck and accretion rates (Adey and Vassar 1975; Steneck and
Adey 1976), were not concentrated in any of the reef or Adey 1976). In contrast, cryptic corallines are charac-
ridge habitats that were drilled (Table 2). However, teristically thin and leafy, and are often dominated by
Lithophyllum congestum was the single most abundant species of the melobisioid subfamily (Fig. 3). They
coralline species found among all cores taken in this usually grow rather slowly because of the low available
bioherm. The high species diversity of corallines found light in that microenvironment (see discussion). There
in the outer ridge core holes 4 and 6 is not characteristic were proportionately more cryptic coralline taxa than
of the usually low diversity of algal ridge corallines those characteristic of exposed and actively accreting
(Bosence 1984; Steneck et al. 1997). surfaces (Table 2).
 103

Fig. 9 Positions of radiocar-

bon-dated HolandeÂs Cays cores
compared with a minimum sea-
level curve for the western
Atlantic (Lighty et al. 1982). See
Table 3 for list of dated
material. Acropora palmata-
dated material is encircled.
Sample number 16 is probably
cave-in material

Mineralogic studies samples plot below this sea-level curve, in marked con-
trast to the remainder of the dates that plot, for the most
X-ray di€raction analyses revealed that one Pleistocene part, well above this curve. Indeed, two samples even
sample is all calcite (2 mol% Mg CO3) and the second plot above mean sea level.
contains 40% calcite along with Mg-calcite and arago-
nite. The presence of calcite in these samples indicates
that this limestone was at one time subaerially exposed Discussion
to freshwater conditions (Tucker and Wright 1990).
In contrast, 21 samples of both dense and chalky The present-day elevated and wave-exposed surface
modern submarine micrite in®llings yielded an average cover of HolandeÂs Cays is almost completely dominated
value of 14.3‹1.5 mol% MgCO3, which is within the by crustose coralline algae, which is in marked contrast
range of values obtained from submarine precipitates to the rather limited evidence of crustose corallines in
of Mg-calcite reported in numerous other coral reef cores from this outer ridge. Encrustations of crustose
frameworks (see Macintyre and Marshall 1988). corallines are common on much of the Agaricia and
Millepora framework but, for the most part, they are
very thin coatings and are made up of species indicative
Radiocarbon dating of cryptic habitats (Table 2). Even the surface sample
from the outer edge of this ridge has only a 1-cm-thick
Twenty-seven radiocarbon dates were obtained from surface cover of crustose coralline algae over an ag-
coral and hydrocoral samples that had all biotic encr- glomeration of encrusters extensively bored and ®lled
usters and submarine-lithi®ed micrite crusts removed. with dense micrite (Fig. 3). Despite the fact that much of
Five samples were collected in place from the surface and the original skeletal material in the multicyclic bored
the remainder from core holes. These dates, which are and micrite-®lled pavement limestone is lost, in better
almost limited to a narrow 1,000-year interval (1,790‹60 preserved sections the crustose corallines form minor
to 3,140‹60 years B.P.), show very little relation to the crusts on the various skeletal components. It is therefore
stratigraphic sequence and reversals are common and apparent that crustose corallines do not form a signi®-
sometimes indicate a major displacement (Fig. 5). Sur- cant part of the framework of this ridge system and that
face samples range in age from 1,790‹ 60 years B.P. on the extensive deposition of inter- and intra-particle mi-
the outer ridge to 2,770‹60 years B.P. at the innermost crite is the dominant cementing agent stabilizing this
leeward reef-¯at drill site. A comparable surface date of framework. Since corallines comprise a relatively small
2,115‹125 years B.P. was reported by Glynn (1973) for a fraction of the bioherm, it cannot be characterized as an
coral that was cemented in the exposed parapet. algal ridge. Further, the high species diversity found in
These dates were subsequently correlated to mean sea the outer ridge sections (core holes 4 and 6, Tables 1 and
level (Table 3) so that they could be plotted against the 2) is not characteristic of algal ridges.
Acropora palmata minimum sea-level curve for the Algal ridges are constructed primarily by crustose
western Atlantic (Lighty et al. 1982). As can be seen in coralline algae (Adey 1978; Steneck et al. 1997). They
Fig. 9, all of the dates obtained from Acropora palmata commonly have very low species diversity (Adey and
104

Macintyre 1973; Steneck and Adey 1976; Bosence 1984; as signi®cant binding agents of coral reef structural el-
Kikuchi and Leao 1997). In the Caribbean, algal ridges ements (Macintyre 1997).
are dominated by two species, Lithophyllum congestum The plot of radiocarbon dates obtained from coral
and Porolithon pachydermum (Adey 1975; Adey and and hydrocoral samples from both cores and surface
Vassar 1975; Steneck and Adey 1976; Adey et al. 1977; samples against the minimum sea-level curve for the
Bosence 1984). The only known exceptions to codomi- western Atlantic (Fig. 9) clearly demonstrates that most
nance by these two species are found in the sediment- of the skeletal material (except Acropora palmata) in the
impacted low energy reefs of the Bahamas that are ridge and leeward reef ¯at are located well above the
dominated by the alga Neogoniolithon strictum (Steneck curve and are thus probably reworked storm deposits.
et al. 1997). Although the single most abundant coralline This is also emphasized by the common reversal in the
species was Lithophyllum congestum, its abundance was position of dated samples in cores and the narrow time
well below that seen in most algal ridges (e.g., Bosence period in which this material was deposited. In fact,
1984). The other algal ridge-building corallines, Poroli- there is little evidence of any accumulation of reef
thon pachydermum and N. strictum, were relatively rare framework in this area over the last 2,000 years. As a
(Table 1). result, this is an area of extensive bioerosion and sub-
Interpreting coralline facies in tropical bioherms is marine lithi®cation. In contrast, there are several dates
complicated by microhabitats. Corallines that grow ex- that plot on, or below, this sea-level curve. This material,
posed to full sunlight and wave action are di€erent from which is limited to the basal sections of core holes, is
those that live in the dimly illuminated cavities on the dominated by Acropora palmata and appears to repre-
underside of platy corals (e.g., Agaricia spp.) or algal sent the in-place coral community over which the storm
understoreys found in and around reefs and algal ridges. debris was deposited. The combination of dates for the
Commonly, corallines indicative of deep water can be buried in-place A. palmata with a maximum date of
found in algal ridges as ``secondary framework'' ele- 2,780‹70 years B.P. and for the transported material
ments (sensu Boscence 1984). Speci®cally, the heavily with a minimum date of 1,790‹60 years B.P., indicates
branched and thick-crusted corallines of the Masto- that this storm ridge was formed by more than one
phoroideae and Lithophylloidea subfamilies that are storm 2,800 to 2,000 years ago.
common on shallow exposed habitats are replaced by Pleistocene limestone was recovered in core holes 1
the thin, leafy deep-water ¯ora (including Melobesioidea and 2 at depths of 3.29 and 3.22 m, respectively, below
subfamily) that are common in cryptic or deep-water mean low sea level (Fig. 5). Depths of penetration for
habitats (Martindale 1992). core holes 3 (4.40 m) and 4 (4.44 m), which are located
The proportion of cryptic to exposed corallines sug- on either side of core holes 1 and 2, were almost a meter
gests the ¯ora came from reef rather than algal ridge lower and yet did not contact the Pleistocene surface
habitats (Table 2). Because carbonate production is much (Fig. 5). This suggests that elongate relief on the Pleis-
greater in the exposed habitats (Adey and Vassar 1975), tocene surface may be responsible for the linear accu-
the percent of an algal ridge comprised of the exposed mulation of storm deposits that formed the HolandeÂs
species far exceeds those comprised of the cryptic or sec- Cays ridge system.
ondary framework species [Bosence (1984) reported 63% Adey (1978) was correct in questioning the existence
of the corallines identi®ed in St. Croix algal ridges were of a well-developed algal ridge system in this area of
Lithophyllum congestum and Porolithon pachydermum.]. lower seasonal minimum wind e€ect. The HolandeÂs
This is in stark contrast with the percent of corallines Cays outer ridge system certainly cannot be classi®ed as
commonly found in coral reefs. Coralline cover on ex- an algal ridge as de®ned by Adey (1978, p. 361), who
posed surfaces of reefs is often relatively low because of the limited this term to ``carbonate frameworks built in large
percent of live coral and algal turf (Adey and Steneck 1985; part by coralline algae.'' By contrast, this ridge system
Steneck 1994). The amount of cryptic habitats is rather has been formed by a series of storm deposits about two
high because the bases of most corals are favorable cryptic to three thousand years ago and has a surface charac-
habitats for corallines. Therefore, the number of cryptic terized by extensive bioerosion and a thin cover of
species commonly associated with algal ridges or coral crustose coralline algae.
reefs would be expected to be higher coming from a reef
than coming from an algal ridge (Table 2).
The peloidal texture and Mg-calcite mineralogy of Conclusions
micrite in®llings, along with the manganese coated hiatal
corrosion surfaces, are all characteristics of submarine The HolandeÂs Cays outer ridge is not an algal ridge. The
lithi®cation that is most pervasive in reef substrates that relief of this topographic feature has been constructed by
are exposed over long time periods (see review by a storm deposit, or series of deposits, that accumulated
Macintyre and Marshall 1988). As a result, shallow reef 2,800 to 2,000 years B.P. Although crustose coralline al-
sites that are exposed to heavy wave action for long gae dominate an extensively bioeroded surface, they
periods of time, such as the HolandeÂs Cays outer ridge, contribute little to the framework of this ridge. In gen-
can be extensively cemented by precipitated micrite in®ll, eral, the crustose coralline ¯ora is a thin veneer over a
which essentially negates the role of crustose corallines coral-reef assemblage and not the massive deposits of low
 105

species diversity dominated by Porolithon pachydermum Dahl AL, Macintyre IG, Antonius A (1974) A comparative survey
and/or Lithophyllum congestum characteristic of algal of coral reef research sites. Atoll Res Bull 172:37±120
Folk RL (1974) The natural history of crystalline calcium car-
ridges. This ridge system, which has been exposed to very bonate: e€ect of magnesium content and salinity. J Sediment
shallow turbulent wave action for 2,000 years, has been Petrol 44:40±53
extensively lithi®ed by submarine-precipitated micritic Glynn PW (1973) Aspects of the ecology of coral reefs in the
Mg-calcite cements that gives it a very dense framework. western Atlantic Region. In: Jones OA, Endean R (eds) Biology
and geology of coral reefs, vol 2. Academic Press, New York,
pp 271±324
Acknowledgements Field assistance provided by David West, Goldsmith JR, Graf DL (1958) Relation between lattice constants
James E. McRae, and the crew of the R/V Catyani is gratefully and composition of the Ca±Mg carbonates. Am Mineral 43:84±
acknowledged. In addition, we are indebted to the following people 101
for their contributions to this study: Donald A. Dean (thin sec- James NP, Ginsburg RN (1979) The seaward margin of Belize
tions); William T. Boykins (X-ray di€raction and graphics): James barrier and atoll reefs. Int Assoc Sedimentol Spec Publ No 3
E. McRae (graphics); and Mary E. Parrish (drafting). I. Rubino€, Kikuchi RKP, Leao ZMAN (1997) Rocas (southwestern equatorial
Director of the Smithsonian Tropical Research Institute, provided Atlantic, Brazil): an atoll built primarily by coralline algae. Proc
facilities and other logistic support for this project. Financial 8th Int Coral Reef Symp 1:731±736
support for this study was provided by the Smithsonian Institution Land LS, Moore CH (1980) Lithi®cation, micritization and syn-
Scholarly Studies Program. depositional diagenesis of biolithites on the Jamaican island
slope. J Sediment Petrol 50:357±370
Lighty RG, Macintyre IG, Stuckenrath R (1982) Acropora palmata
reef framework: a reliable indicator of sea level in the western
References Atlantic for the past 10,000 years. Coral Reefs 1:125±130
Macintyre IG (1978) A hand-operated submersible drill for coring
Adey WH (1970) A revision of the Foslie crustose coralline reef substrata. In: Stoddart DR, Johannes RE (eds) Coral reefs:
herbarium. K Nor Vidensk Selsk Skr 1970:1±46 research methods. UNESCO Monographs on Oceanographic
Adey WH (1975) The algal ridges and coral reefs of St. Croix: their Methodology 5, Paris, pp 75±80
structure and Holocene development. Atoll Res Bull 187:1±67 Macintyre IG (1997) Re-evaluating the role of crustose coralline
Adey WH (1978) Algal ridges of the Caribbean sea and West algae in the construction of coral reefs. Proc 8th Int Coral Reef
Indies. Phycologia 17:361±367 Symp 1:725±730
Adey WH, Burke RB (1976) Holocene bioherms (algal ridges and Macintyre IG, Marshall JF (1988) Submarine lithi®cation in coral
bank barrier reefs) of the eastern Caribbean. Geol Soc Am Bull reefs: some facts and misconceptions. Proc 6th Int Coral Reef
87:95±109 Symp 1:263±272
Adey WH, Burke RB (1977) Holocene bioherms of Lesser Antilles Martindale W (1992) Calci®ed epibionts as palaeoecological tools:
± geologic control of development. Stud Geol 4:67±81 examples from Recent and Pleistocene reefs of Barbados. Coral
Adey WH, Macintyre IG (1973) Crustose coralline algae: a Reefs 11:167±177
re-evaluation in the geological sciences. Geol Soc Am Bull Milliman JD (1974) Marine carbonates. Springer, Berlin Heidel-
84:883±904 berg New York
Adey WH, Steneck RS (1985) Highly productive eastern Caribbean Newman WA, Curray JR, Dana TF, Crampton PJS (1970) AL-
reefs: synergistic e€ects of biological, chemical, physical and PHA HELIX Research Program: 1969±1970. Univ California
geological factors. In: Reaka ML (ed) The ecology of deep and Press, San Diego
shallow coral reefs. Symposia Series for Undersea Research, 2. Steneck RS (1986) The ecology of coralline algal crusts: con-
Oce of Undersea Research, NOAA, Rockville, Maryland vergent patterns and adaptive strategies. Annu Rev Ecol Syst
Adey WH, Vassar JM (1975) Colonization, succession, and growth 7:273±303
rates of tropical crustose coralline algae (Rhodophyta, Cryp- Steneck RS (1994) Is herbivore loss more damaging to reefs than
tonemiales). Phycologia 14:55±69 hurricanes? Case studies from two Caribbean reef systems
Adey WH, Adey PJ, Burke RB, Kaufman L (1977) The Holocene (1978±1988). In: Proc Colloq on Global Aspects of Coral Reefs:
reef systems of eastern Martinique, French West Indies. Atoll Health, Hazards and History. 1993. University of Miami,
Res Bull 218:1±40 Florida, pp 220±226
Bosence DWJ (1984) Construction and preservation of two modern Steneck RS, Adey WH (1976) The role of environment in control of
coralline algal reefs, St. Croix, Caribbean. Paleontology 27:549± morphology in Lithophyllum compressum, a Caribbean algal
574 ridge builder. Bot Mar 19:197±215
Bosence DWJ (1985) The morphology and ecology of a mound Steneck RS, Macintyre IG, Reid RP (1997) Unique algal ridge
building coralline alga (Neogoniolithon strictum) from the system of Exuma Cays, Bahamas. Coral Reefs 16:27±35
Florida keys. Paleontology 28:189±206 Stoddart DR (1969) Ecology and morphology of Recent coral
Braga JC, Bosence DWJ, Steneck RS (1993) New anatomical reefs. Biol Rev 44:433±498
characters in fossil coralline algae and their taxonomic impli- Stoddart DR (1971) Coral reefs and islands and catastrophic
cations. Paleontology 36:536±547 storms. In: Steers JA (ed) Applied coastal geomorphology. MIT
Clifton KE, Kim K, Wul€ JL (1997) A ®eld guide to the reefs of Press, Cambridge, MA, pp 155±197
Caribbean Panama with an emphasis on western San Blas. Proc Tucker ME, Wright VP (1990) Carbonate sedimentology. Black-
8th Int Coral Reef Symp 1:167±184 well Scienti®c Publications. Cambridge, MA