Geology of the
Amber-Bearing Deposits of the Greater Antilles
Caribbean Journal of Science, Vol. 00, No. 0, 141-167, 2001
Copyright 2001 College of Arts and Sciences
University of Puerto Rico, Mayaguez
Geology of the Amber-Bearing Deposits of the Greater Antilles
Manuel A. ITURRALDE-Vennet Museo Nacional de Historia Natural, Obispo
no. 61, Plaza de Armas,
La Habana Vieja 10100, Cuba.
iturra@mnhnc.inf.cu
ABSTRACT
Amber and associated lignitic rocks are known from Cuba (Miocene
lignite), Haiti (Miocene lignite and traces of amber), the Dominican
Republic (Miocene lignite and amber in exploitable quantities), Puerto
Rico (Oligocene and Miocene lignite and traces of amber), and Jamaica
(Maastrichtian Paleocene amber).
However, there is no modern review of the geology of the amber-bearing
deposits and the data available is dispersed in many
contributions. This paper fills this gap and presents the results
of five years of original research on the subject.
Greater Antillean amber probably derived from the resin of Hymenaea
protera, an extinct leguminous tree that probably grew in evergreen
forests. Amber is the consequence of diagenetic changes that
operate in the resin after burial in the sedimentary pile, sometimes
over 1000 m deep, where it is subjected to higher temperature and
pressure over millions of years. The origin of unusually large Miocene
deposits of amber in the Dominican Republic can be explained by the
fortunate combination of adequate conditions of relief and soil for the
development of a large populations of resin producing trees during a
constrained warm and humid climate optimum that occurred about 16 my
ago.
INTRODUCTION
Dominican amber was known by the natives of Hispaniola and
was taken to Spain by Christopher Columbus as one of the treasures of
the West Indies. During the 20th century, Dominican amber became
famous for the quality of its fossils, which include extremely
well-preserved fungi, algae, plant remains, land invertebrates
(arthropods, nematodes, gastropods) and land vertebrates (amphibians,
reptiles, remains of mammals and birds) (Perez-Gelabert, 1999). Amber
has also been reported from Haiti (Maurrasse, 1982), Puerto Rico
(Iturralde-Vinent and Hartstein, 1998) and Jamaica (G. Draper, pers.
comm.), but none of these occurrences has economic value. Despite
widespread interest in amber, there is no modem review of the geology
of the amber-bearing deposits of the Greater Antilles. This paper
fill a gap by (1) reviewing the occurrences, age, stratigraphic
position, and environment of deposition of the amberiferous beds; (2)
presenting new information about the age and origin of the amber as a
fossil resin; (3) discussing the paleogeographic scenario when amber
bearing deposits were formed, and (4) describing the fossil assemblage
found not in the amber itself but in the associated deposits
(Iturralde-Vinent and MacPhee, 1996; Iturralde-Vinent and Harstein,
1998).
WHAT is RESINITE, Resin, COPAL AND AMBER?
Resin-producing trees are widely distributed in the
tropics, but amber in tropical America most probably derived from the
resin of the extinct leguminous tree Hymenaea protera. Although
several extant species of this genus occur from the Amazon Basin to
Mexico, only two species occur in the Greater Antilles: the widespread
H. courbaril L. (commonly known as courbarit in Cuba and as algarrobo
in the Dominican Republic and Puerto Rico), and the northeastern Cuban
endemic H. torrei Leon. The extinct amber producing tree of the
Dominican Republic was named Hymenaea protera by Poinar (1991). This
taxon is related to H. verrucosa of western Africa and probably to H.
torrei of eastern Cuba (Lee and Langenheim, 1975).
Resinite
This is an all-encompassing term for all types of plant-derived resins,
regardless of age and physical or chemical characteristics.
Resin
This term applies to material recently exuded from the tree and which
has not been buried. Its physical and chemical characteristics
depend on the species of tree. Resin can be sticky or dry and
fragile. It can have a whitish or darker coating, but internally
is transparent, yellowish, reddish, or brown. Resin is
non-volatile, relatively inert, hydrophobic, amorphous, and strongly
resistant to decay. Its decay (as well as that of copal and
amber) is due to atmospheric exposure, and the material is best
preserved when deposited in subaqueous or waterlogged
environments. Atmospheric weathering often produces opaque
surface crusts and oxidation rims, and may cause an overall darkening
of the resin grains (Tyson, 1995). Naturally exuded resin seems
to be a physical surface barrier against infection by plant pathogens
(Lee and Langenheim, 1975; Grimaldi, 1996).
Copal
This is an older resin that can be found in the leaf litter layer or
buried in the soil below the tree. It is usually solid, fragile,
transparent with a dark surface coating and internally yellowish,
reddish, or brown. Copal of H. courbaril usually sinks in fresh
water and floats in salt water (personal observation). Resin in
the form of copal accumulates in forest soils and ombrogenous peat
deposits, whose high (domed) water tables aid preservation.
Amber
This is the fossil equivalent of copal. It is
usually indurate, massive, and resistant to organic solvents.
Amber can be transparent, but more frequently it is translucent with
yellow, reddish, brown or blue brown color. These characteristic
are the consequence of diagenetic changes that operate in copal after
burial in the sedimentary pile, sometimes at depths over 1000 m, where
it is subjected to elevated temperature and pressure. Under these
conditions and several millions of years, copal is naturally cooked and
transformed into amber. Dominican amber usually sinks in fresh
water and floats in salt water (personal observation).
After erosional exposure, copal and amber are frequently reworked and
redistributed by flash floods and the normal fluvial and runoff
processes. Amber can also be reworked by coastal erosion, as in
the Baltic region (Schlee, 1990; Grimaldi, 1996). Amber clearly
has significant potential for redeposition if protected from prolonged
exposure to atmospheric oxygen. Langenheim (1990) indicates that
near-flotation allows some amber particles to be readily moved by weak
currents and tends to protect them from abrasion and fracture, even
during prolonged transport. Amber is typically deposited in low
energy bays, deltas, river mouths, estuaries, and other coastal
areas. Large amounts may be carried further offshore as
inconspicuous microscopic particles (Grimaldi, 1996).

The Amberiferous Deposits of the
Greater Antilles
Amber and associated lignitic rocks are known from the
Greater Antilles (Fig. 1) in Cuba (Miocene lignite), Hispaniola
(Miocene lignite and amber), Puerto Rico (Oligocene and Miocene lignite
and traces of amber), and Jamaica (Maastrichtian/Paleocene traces of
amber). The following paragraphs describe the basins and
sediments known to contain amber, regardless of their commercial
importance.
Hispaniola
The Late Tertiary rocks of the Dominican Republic occur in a variety of
geological contexts, predominantly of sedimentary origin, formed in
subaerial to deep-marine depositional environments (see references in
Mann et al., 1991). Amber in commercial quantities is well known
from areas north of Santiago de los Caballeros and northeast of Santo
Domingo (Fig. 1; Vaughan et al., 1922). These mining districts
are known respectively as the Northern Area and the Eastern Area.
Minor occurrences of amber have been reported from the Plateau
Central-San Juan Area (Lemoine in Sander-son and Farr, 1960; Garcia and
Harms, 1988; Harms 1990). All amber occurrences are of late Early to
early Middle Miocene age and are associated with lignitic material, but
each represents different depositional environments, from
terrestrial-marine transition to deep marine (Iturralde-Vinent and
MacPhee, 1996).
Eastern Mining District
The Eastern Mining District is on the northern margin of the Cordillera
Oriental, between Sabana la Mar, Hato Mayor, Bayaguana, and Cotui (Fig.
2). The sedimentary rocks that form
the substrate are mostly Neogene; along the west, south and east
margins of the basins they directly overlay Cretaceous sedimentary,
volcanic, and igneous rocks. On the northern side of the area, a
W-NW trending fault is present along the edge of the Neogene
limestone. Pliocene overlies the limestone on the northern block
of the fault to Recent deposits. In general, Miocene rocks dip
gently toward the north-northeast, so the thickness of the Neogene
section varies from less than 100 m in the south to several hundreds of
meters in the north (Brouwer and Brouwer, 1982; Toloczyki and Ramirez,
1991; Lebron and Mann in Mann et al., 1991).
Brouwer and Brouwer (1982) recognized four stratigraphic units in the
Eastern Mining District: the basal conglomerates, the amber-bearing
Yanigua Formation, the "caliza Cevicos", and the “caliza de los
Haitises." According to some authors, there is only one limestone or
caliza unit (Bowin, 1966; Toloczyki and Ramirez, 1991; Lebron and Mann
in Mann et al., 1991). My field observations suggest the
limestone that Vaughan et al. (1922) named Cevicos are those found as
intercalations in the upper part of the Yanigua Formation, and are not
worthy of being differentiated as an independent unit. The basal
conglomerates, as they occur intercalated with the other lithologies of
the Yanigua Formation, are also incorporated to this unit.
Therefore, only two formations are recognized in the Miocene basin:
Yanigua formation (with the basal conglomerates) and Los Haitises
formation (Figs. 2 and 3).

YANIGUA FORMATION
The Yanigua Formation (Brouwer and Brouwer, 1982) is the amber-bearing
unit of the eastern Area. Outcrops are located mostly around the
present-day margins of the Neogene basin, at the localities of Anton
Sanchez, Bayaguana, Comatillo, Colonia San Rafael, Yanigua, El Pont6n,
Sabana la Mar, and in several falls along the Yanigua and Comatillo
rivers (Fig. 4). Its thickness is calculated from the map as
around 100 m.
The basal conglomerates have limited outcrops (Champetier
et al., 1980) but may have an extensive distribution, as they were
found in several boreholes drilled in the basin (Brouwer and Brouwer,
1982).
The conglomerates are composed of poorly sorted subangular to
subrounded elements contained in a fine-grained matrix. Since
fossils have not been reported, their age assignment remains
problematic, but according to stratigraphic position it can be basal
Miocene-Oligocene. Champetier et al. (1982) suggest these
conglomerates are frequent intercalations in the base of the Yanigua
Formation in the Samana area. They do not report fossils from
these localities, but the association of conglomerates and ignite
suggest that the basal conglomerates are a lateral, lowermost facies of
the Yanigua Formation. Champetier et al (op cit) reported no
amber from this conglomerates. The sedimentary characteristics of
the conglomerates (oblique lamination, formation of channels, erosional
surfaces below the coarse grained beds) suggest a fluvial environment
of deposition.
In general, the rest of the Yanigua section shows minor lateral
differences (Fig. 4). Dark clays and laminated sandy clays
containing fresh-water mollusks are the most common lithologies, along
with lignite and carbonaceous clays and sandstones. Sandy clays
may contain authigenic pyrite and are composed of 80-90 % mud and 10-20
% grains of calcite, calcareous earthy particles, detrital quartz, and
igneous rock fragments. Laminae in the sandy clays are normally
parallel, but can be slightly disturbed by micro-ripples and by
isolated gravel inclusions that indicate local turbulent
currents. These beds contain flattened and irregular inclusions
of amber, usually as pockets or lenses ranging from a few millimeters
to several centimeters in size, and may also contain fresh- to brackish
water ostracods and mollusks (see below).
The clays are typically dark gray-black, carbonaceous, with rare
authigenic pyrite, almost without detritus, and with only some grains
of calcite and quartz. They are usually fossiliferous and contain
fresh- to brackish-water mollusks, ostracods, foraminifera, bryozoa,
fish teeth, etc. (Tables 1 and 2). Mollusk shells have been
usually fractured and flattened during diagenesis. Fossils of
crocodiles, turtles and dugongs have been found in this context
(Iturralde-Vinent and MacPhee, 1996). These amber bearing beds
contain abundant fragments of carbonaceous material that still shows
textures of a vegetal nature. Although Brouwer and Brouwer (1982)
report that all 'the amber from this source shows erosional features,
all the specimens found by the present author have sub-rounded, oval or
stalactite-like shapes representing the original form and not one
caused by secondary erosion.
144
The lignite is found intercalated with the carbonaceous
clays and sandy-clays. Lignite sometimes includes needles and
crystalline aggregates of gypsum, probably originated during
diagenesis. The lignite occurs usually as layers a few
millimeters thick, but also occur as beds over 1 m thick. The
lignite is hard, compact, and deep black with a metallic shine.
No fossil inclusions, other than plant impressions on the surface of
the beds, have been found. Bioturbation due to animal burrowing
and plant root penetration is common in the lignite and related
sediments. The burrows are vertical, sometimes with horizontal
bifurcations, and they are filled with material from the layer above.
Locally in the area of Sierra del Agua is a well-developed 1 to 2 m
thick, reddish paleosol clay layer, located near the top of the
lignite. Coarse gravel and sand on top of the paleosol is also
found in this region (Fig. 4). Dugong and crocodile bones have
been recovered from the clays that underlie the lignite at this
locality.
Environment
According to Champetier et al. (1982) the section at Yanigua was formed
in a low energy environment, which is the common type of section in the
basin. Van den Bold (1988) studied the ostracods of two samples
of the Yanigua Formation obtained from amber mines near Laguana and
Bayaguana, and suggested that the association indicated a low salinity
environment. Lithology, sedimentological features, and fossil
composition clearly corroborate the opinion of Brouwer and Brouwer
(1982) concerning the lagoonal to coastal marine depositional
environment for most of the Yanigua Formation. This opinion was
followed by Toloczyki and Ramirez (1991) when they described these
rocks as “marls of littoral facies".
Toward the top of the Yanigua Formation are beds of
biocalcarenites, first intercalated with the clays (Cevicos limestone),
but then dominating the section until the overlaying Los Haitises
Formation. These biocalcarenites are very fossiliferous, with
large foraminifers, algae, fragmentary marine mollusk shells,
echinoids, bryozoa, corals, etc. They also show evidence of
bioturbation due to animal burrowing and oxidized plant roots, which
suggests a marine shallow to coastal marsh environment. These
successions of strata record the transition from lagoonal deposition
into carbonate shelf sedimentation (Yanigua through Los Haitises
Formations).
Age
The Yanigua Formation was dated as Miocene by Brouwer and Brouwer
(1982), who report the fossils Ammonia beccarii, Elphidium advenum, and
Parocnus sp. from Sierra del Agua. [Note: Since Parocnus is a
Quaternary sloth, a photograph of the fossil, kindly provided by S.
Brouwer, was examined along with additional material recovered from the
same site. The study suggests that the fossil bones correspond to
a Tertiary sirenian, probably Methaxytherium? sp. (Iturralde-Vinent and
MacPhee, 1996)].
Toloczyki and Ramirez (1991) ascribed an Upper Miocene to Lower
Pliocene age to the Yanigua Formation, while Lebron and Mann. (In Mann
et al., 1991) ascribed to it an Upper Miocene age (however, neither
presented supporting paleontological data). Several samples
collected by the author in different parts of the basin were examined
for foraminifera, ostracods, and nanno fossils. The samples
96RD-10A, 96RD-10D and 96RD-11A were collected at the Yanigua River and
in mines, from limestone beds intercalated near the top of the Yanigua
Formation. The foraminiferal assemblages suggest a late Early
Miocene age (Table 1, Fig. 4). The microfossils in samples
collected 2-3 m below these limestones (96RD-10E, 11B, 11C; Table 1)
and in Colonia San Rafael mines (96RD-5A, 6A; Table 1) are not as
distinctive and suggest a late Early to Middle Miocene age. Van
den Bold (1988) reported late Early to early Middle Miocene ostracods
for two samples from Laguana and Bayaguana amber mines and corroborated
this date with samples supplied by the author (Table 1).
Therefore, the Yanigua Formation can be dated as late Early to early
Middle Miocene (Miogypsina-Soritiidae zone of Iturralde-Vinent (1969),
or slightly younger, about 15 to 20 million years old according to the
scale of Berggren et al. (1995).
Los Haitises FORMATION
The Haitises Formation (Brouwer and Brouwer, 1982) represents a shallow
water shelf limestone overlying the Yanigua Formation. These
limestones have been named Los Haitises or Cevicos in previous
literature but Los Haitises is used herein. They are
fossiliferous, containing mollusks (Strombus sp., Kuphus sp. and many
others), corals (Porites sp., Acropom spp.), echinoderms, and
algae. The age has been assigned without paleontological
justification as- Pliocene-Quaternary (Toloczyki and Ramirez, 1991),
but Middle to Late Miocene is more likely because it rests conformably
above the Yanigua Formation (Figs. 3 and 4). The thickness of the Los
Haitises limestone can be estimated as 300 m minimum 'considering that
it rests nearly horizontal and gives rise to steep karstic hills, some
of them over 250 m high.
148
Northern Mining District
The Northern Mining District is located in the Cordillera
Septentrional, north of Santiago de los Caballeros (Figs. 1, 2).
This fault-bounded unit was designated as the La Toca Block (Dolan et
al., 1991; de Zoeten and Mann, 1991), and part of the El Mamey Belt,
represented by deformed Mesozoic and Cenozoic rocks. The La Toca
block is limited today by the Camu, Septentrional, and Rio Grande fault
zones (de Zoeten and Mann, 1991) (Fig. 5).
Originally, the La Toca Block was a part of a larger basin filled with
Oligocene to Pliocene sedimentary rocks that unconformably overlies
deformed and partially metamorphosed Lower Eocene and older igneous and
sedimentary rocks (Toloczyki and Ramirez, 1991, Dolan et al., 1991; de
Zoeten and Mann, 1991). The Late Tertiary deposits have been
recently described as La Toca and Villa Trina Formations (Fig. 3; Dolan
et al., 1991; de Zoeten and Mann, 1991). The amber-bearing
deposits in the area are probably in the upper part of the La Toca
Formation.
LA TOCA FORMATION
The La Toca Formation (de Zoeten, 1988) was originally described as a
clastic Oligocene to lower Middle Miocene unit (de Zoeten, 1988; de
Zoeten and Mann, 1991). It is equivalent to the Altamira facies
of the El Mamey Formation of Eberle et al. (1980) (Fig. 5). The
La Toca Formation has many similarities in composition, sedimentology,
and age to the Oligocene to Early Miocene Maquey Formation of the
Guantanamo Basin of eastern Cuba (Nagy et al., 1983; Calais et al.,
1992; Iturralde-Vinent, 1994) and to the Thomonde Formation of the
western part of the Plateau Central-San Juan area, but neither of these
formations is amberiferous (Nagy et al., 1983; Butterlin, 1960;
Maurrasse, 1982). The La Toca Formation was subdivided into three
units that are characterized below according to de Zoeten (1988), with
additional observations by the present author.
149
The lower conglomerate unit is about 300 m thick and
corresponds with amalgamated conglomeratic facies and minor interbedded
sandstones. Disorganized, matrix supported facies and thinner,
clast supported facies, comprise the lower 200 m of the unit measured
by de Zoeten (1988) along the Rio Grande section. Most beds are
tabular, medium-bedded to massive, with planar basal contacts that
rarely drape over clasts protruding from the underlying bed. The
internal organization of the unit seems to increase upward in the
section. In the later part, inversely graded facies are
interbedded with clast supported, parallel, stratified
conglomerates. Clasts in the base range from granule to boulder
in size. They are equi dimensional to oblate in shape, and are
subangular to rounded. The composition of the clasts near the
outcrops of the underlying Pedro Garcia Formation are about 70 %
volcanic rocks (tuffs, andesites, and lavas), 20 % intrusives
(tonalites) and 10 % sandstones, argillites, recrystalized limestone,
quartz veins, serpentinite, and coral rudstones. This suggests
that the major source for the clastics were uplifted lands with rocks
like those of the Pedro Garcia Formation. Such rock types are
found not only in Cordillera Septentrional but also in Cordillera
Central and Cordillera Oriental (Toloczyki and Ramirez, 1991).
The flysch unit consists of alternating, thin- to medium-bedded,
thinning-upward sandstones and shale couplets. A section measured
by de Zoeten (1988) consists of 500 m of thick very thin- to medium
bedded sandstones and shales. The beds are tabular, laterally
continuous, and exhibit sharp basal contacts. Basal sands are
relatively coarse-grained, graded or structure less, and have mud-rich,
massive, bioturbated upper divisions. A few thin- to
medium-bedded, tens-shaped calcareous sandstone beds are interbedded
with the sandstone and shale couplets. Higher in the section,
lithic-rich calciturbidites and lithic conglomerates are interspersed
between thin-bedded silicidastic couplets. Together with the 1-2
m-thick conglomerate beds, calciturbidites form small (1-5 m thick)
coarsening- and thickening-up cycles. The unit is laterally
interbedded with the lower conglomerates. These deposits carry
abundant planktonic foraminifera and nanno-fossils, as well as small
benthic foraminifera typical of marine bathyal depth (500-1500 m)
(Dolan et al., 1991).
The upper unit, 300 m thick according to de Zoeten (1988), is composed
of thick-bedded sandstones with lesser amounts of conglomerate
beds. Sandstone beds have high sand to silt ratios (2:1 to 10:1)
and are commonly tabular, laterally continuous, and thick- to very
thick-bedded (0.5 to 3.5 m). Basal sands range from gravel to medium
sand in size. The lower divisions of the beds are coarse-tail graded,
or massive, and grade up into parallel laminated sands rarely
containing ripple marks. Load casts are the dominant type of sole
marks. The upper division in many beds contains concentrations of
lignite and amber fragments that define parallel laminae in the
siltstones. The finer fraction of the beds is mud-poor and
parallel-laminated or massive. Considerable amounts of small,
carbonized plant fragments, including seeds, have been observed in
these beds. The flysch and upper limit grade laterally into each
other and become almost indistinguishable in the field.
Occurrence of amber
Eberle et al. (1982) indicate that amber is always contained in
lignite-rich sandstone beds or in lignite seams, and that considerable
amounts have been mined from a few tens of meters above prominent
conglomeratic horizons. However, according to Redmond (1982)
amber is very rarely associated with the conglomerates and occurs in
laminated, blue-gray, biotite-bearing siltstone with a rich organic
content, sometimes concentrated in lignite stringers within the
siltstone. Lignite beds are several inches thick at least in one
location. Redmond also pointed out that amber-bearing siltstone
is overlain almost always by thick, or more usually massive
sandstones. This suggests that the amber was deposited in a
mid-fan below the end of the channels, and that the massive sand from
the channel was laid down later, prograding over the amberiferous
bed. For Redmond, the amber seemed to occur in three to five
horizons, suggesting that it was deposited during a short period of
time and not during all the time lapse of deposition of La Toca.
150
Several mining areas were visited by the present author to verify the
position of amber in the La Toca Formation. In the Palo Alto Mine
(Fig. 5), a section about 30 m thick was deposited under moderate
energy and is composed of a set of complete or incomplete cycles of
fine sand, carbonaceous silt, supra-carbonaceous shale with amber, and
infra-carbonaceous shale and lignite (Champetier et al., 1982).
Large fragments of amber, some weighing up to 13.6 kg, have been
recovered from this mine. In La Toca Mine (Fig. 6), the beds
overlying the amberiferous horizons are a set of 3-5 cm thick fine to
coarse-grained sands without gradation or coarsening upward. Two
or three beds of conglomerates, up to 60 cm-thick, are intercalated in
the section; they have a peculiar gradation from fine grains at the
base to coarse grain in the middle and fine grains on the top.
The amberiferous section is about 30 m thick and is composed of
interbedded black, laminar, medium-to-fine grained sands and
ligniferous silt. There are some beds of actual lignite measuring
up to 5 cm thick. The section contains many carbonaceous plant
remains and very small mollusks, both typical of a brackish water
environment. The clastic materials in the rocks are composed of
grains of igneous rocks and detrital minerals, including some quartz
and calcite. The amber appears as flattened lenses, stalactites,
and also as irregularly shaped, typically detrital fragments.
El Cacao Mine is located very close to La Toca, on the
southern side of the same mountain range (Fig. 6). This site is
particularly known for its unique blue amber. The amber-hearing
section is about 20 m thick and is represented by fine-grained, well
bedded (20-40 cm thick), black sandstone with some thinning upward
gradation. The sandstone contains abundant small
gastropods. Palo Quemado Mine (Fig. 6) is a few kilometers south
of the previous mine. The section observed is about 30-40 m thick
and is represented by a well bedded, 10-20 cm thick stratum of
fine-grained, laminated, black sandstone. Coarse-grained
sandstone with carbonate cement is interspersed in the section.
This sandstone became locally conglomeratic, as it contains isolated
large pebbles within the matrix.
151
The section at Los Higos Mine is about 30 m thick and is represented by
well-bedded, laminated, black, fine-grained sandstone and silt,
interspersed with coarse grained, 40-50 cm thick layers of sandstone
and conglomerate. These beds are similar to those seen in La
Toca, with a peculiar gradation from fine grains at the base to coarse
grains in the middle and finer grains on the top, sometimes laminated,
with intercalated lignite layers up to 2 cm thick. As in Palo
Quemado, isolated pebbles are found on the very top of these
beds. Amber occurs as irregular fragments, 20 to 40 cm long,
embedded in the fine-grained rocks. Scattered fragments of
lignite, approximately 6 cm long and 3 cm wide, are also present.
In all the mining areas visited, the amber is always associated with
ligniferous material, and occurs in laminated sands and silts of deep
marine environments. The amberiferous horizons occur within rock
sections up to 40 m thick. Samples collected from these mines did
not produce a precise date (up to the nanno-stratigraphic zone), so it
was impossible to accurately correlate the sections from mine to
mine. However, all the samples lie within the late Lower
Miocene-early Middle Miocene interval, about 15 to 20 my
(Iturralde-Vinent and MacPhee, 1996). Therefore, it is possible
to accept the conclusion of Redmond (1982) that all the amber mines are
within the same stratigraphic level in the upper part of La Toca
Formation.
Environment
Composition and sedimentary features of the lower conglomerate unit
suggest that they were derived from a local source, and probably laid
down as rapid, cohesive debris flows in terrestrial to near shore
environments; probably related to the active uplift of a highland
area. However, the deposition of the flysch and upper
amberiferous units took place in marine environments close to a fluvial
plain with deltaic development.
According to Eberle et al. (1982), the lithological association of the
flysch and upper unit of La Toca Formation point to a fan delta
depositional environment, the amberiferous beds being deposited in
interchannel flats, while marly units would represent the distal fan
base environment. According to de Zoeten (1988), because vertical
organization is lacking in the thick, tubular sandstone beds, it is
unclear if the beds were deposited as lobes, channel-lobe transitional
facies, or as delta front sands. Moreover, the microfossil
assemblage and sedimentary features of the flysch unit suggested to
Redmond and to de Zoeten and Mann that the beds were deposited as
turbidity flows in a deep-water basin. Northwest paleocurrent
directions in La Toca and the isochronous Las Lavas Formations suggest
sediment source was located toward the southeast (de Zoeten and Mann,
1991).
Lithological observations and paleontological evidence indicate that
the flysch and upper units were deposited in a deep basinal environment
located dose to a forested mountain range. All the samples
recovered from the previously mentioned mining areas contain planktonic
and small benthic forams in associations suggesting a deposition depth
over 500 m, probably 1000-1500 m (de Zoeten, 1988). The common
presence of nannofossils agrees with this interpretation, although some
of these fossils are redeposited from older beds.
Age
No fossils have been found in the lower conglomerate, but its age is
probably Oligocene-Early Miocene because the unit is typically at the
base of the sections or partially interfingers with the flysch.
The conglomerate probably represents a lateral equivalent of the basal
conglomerate of the Yanigua Formation (Fig. 4).
Samples with microfossils reported by de Zoeten (1988) and
Dolan et al. (1991), collected from different levels in the flysch,
yielded ages of latest Oligocene and Early Miocene (7787, 10487, 13887;
Table 3) and Early to Middle Miocene (P-14, D-229, 11587, 13187, 13487,
14087; Table 3). In the Miocene samples there is some reworking
from late Oligocene, which may suggest auto cannibalism of the basin
sediments. From the same localities (La Toca, Palo Alto, and Rio
Grande mines), Cepek (in Schlee, 1990) dated the rocks as Middle Eocene
through Middle Miocene based on unlisted nannofossils (see also
Grimaldi, 1995). The microfossils older than latest Oligocene
found by Cepek are probably redeposited, since this process is common
in the basin and samples taken from the same mines were dated as
Miocene (see below).
152
Although microfossils are rare in the actual amber-bearing
beds, several samples collected by the present author contain
identifiable fossils. Sample 96RD-15, collected in the flysch
several hundred meters stratigraphically below the La Toca mine area,
yielded a poorly constrained Oligocene or Early Miocene assemblage
(Table 4). The La Toca mine amber-bearing sandstone sample 96RD-12B
carries a poorly preserved Oligocene or Early Miocene assemblage (Table
4). From just over the section at La Toca, sample 96RD-12C from
the Villa Trina, limestone yielded a very late Miocene or Pliocene
assemblage (Table 5). Additional samples of different Miocene
ages were identified using foraminifers from the Palo Quemado, El
Cacao, and Los Higos mines (Figs. 5, 6; Table 4). This age range
was corroborated by nannofossils (Table 4) from La Toca (96RD-12), Los
Cacaos (96RD-18), and Los Higos (96RD-19 and 20). Only the
results from Palo Quemado (96RD-17) samples are contradictory, -since
according to its nannofossils the sample is probably Late Oligocene and
according to the foraminifers it is Miocene. However, both
samples are poorly fossiliferous and the age assignments are not well
constrained, being mostly based on absence of some index taxa.
Therefore, the entire range of the age of the La Toca Formation can be
identified as Late Oligocene through Middle Miocene, as pointed out by
de Zoeten (1988) and de Zoeten and Mann (1991); but the amberiferous
beds are more probably restricted to the upper half of the formation,
which is late Early to early Middle Miocene in age (Iturralde-Vinent
and MacPhee, 1996).
VILLA TRINA FORMATION
The Villa Trina Formation (Vaughan et al., 1922) is
composed of shallow marine marls and calcarenites resting unconformably
above La Toca Formation (Figs. 3, 5). The base of the unit is
coincident with a basal conglomerate (Eberle et al., 1982). The
age of the limestone has been assigned as Middle Miocene to Pliocene
(de Zoeten and Mann, 1991), but samples of the Villa Trina limestone
collected by the author (Table 5) yield assemblages of Late Miocene and
Pliocene age. The thickness of the limestone unit can
reach several hundred meters as can be inferred from present day
topography.
Plateau Central-San Juan area
This basin, unlike those previously described, is located south of the
Cordillera Central in southwestern Hispaniola (Fig. 2). The
depression was filled with Late Tertiary clastic and carbonate
sediments, many hundred meters thick, which show important differences
in environment of deposition (Fig. 3). Only the Sombrerito and
Maissade Formations, which contain amber and lignite, are described
here. Garcia and Harms (1988) and Maurrasse (1982) provide more
information about this area.
154
Sombrerito FORMATION
The Sombrerito Formation is mostly developed in the eastern part of the
Plateau Central-San Juan Basin (Fig. 2). It is represented by
about 500 m of pelagic limestone, marls, and calcarenites. The
marls are rich in pelagic and small benthic microfossils. Calcarenites
are coarse to fine grained, graded, with sedimentary features
indicating turbidity current origin. Sedimentary features and
nannofossils suggest that the formation was deposited in a lower slope,
middle bathyal setting open to the sea. The formation has been
paleontologically dated as late Early to early Middle Miocene (N8 -
N12) (Garcia and Harms, 1988).
It should be emphasized that this unit generally lacks detritus from
older rocks, suggesting that little erosion of the catchment area was
taking place at the time of the Sombrerito Formation deposition (Garcia
and Harms, 1988). Isochronous shallow water organic fragments
derived from a carbonate shelf and coral reef environment dominate the
detritus material in the calcarenites. Nevertheless, the
calcarenites that crop out east of Presa Sabaneta contain isolated
fragments of amber (1-3 cm) associated with small remains of carbonized
plants and detrital grains of quartz, plagioclase, and probably
glauconite (Garcia and Harms, 1988; Harms, 1990). This clearly
indicates that local erosion was taking place in the catchment area of
the basin, but the size and type of the detritus suggests that the
source area was a low land mass. This sedimentary environment
contrasts with that of the amber-bearing La Toca Formation, which
corresponds to a clastic dominated section representing extensive
erosion of a highland source.
MAISSADE FORMATION
The Maissade Formation (Jones, 1918) occurs near the town of Maissade
in central Haiti, and over the southwestern part of the Plateau
Central-San Juan Basin. Lemoine (in Sanderson and Farr, 1960)
reported traces of amber in a sample recovered from a well drilled in
the Maissade area (Figs. 2, 3). The formation consists of 200 m-thick
clay, shale, marl, gypsum, some sandstone, and characteristically
lignite beds. The lower part of the unit is blue clay containing
shallow water marine mollusks (Turritella sp., Arca sp., and Ostrea
sp.), overlain by black carbonaceous shale, hard shaly sandstone,
lignite, lignitic gray days, gray sandy shales, gray argillaceous
marts, blue clays, and some gypsiferous horizons. These beds
usually contain the brackish water mollusks Potamides sp., Hemisinus
sp., Hydrobia sp., Nerita sp., Scapharca sp., and Mytylopsis sp.
(Woodring et al., 1924; Butterlin, 1960; Maurrasse, 1982). The Maissade
Formation overlies with a local unconformity the latest Oligocene
lowermost Miocene Madame Joi Formation, and is overlain in turn by the
Pliocene Hinche Formation (Butterlin, 1960; Maurrasse, 1982). The
Maissade is generally considered a lateral facies of the Thomonde and
Las Cahobas Formations (Bermudez, 1949). The age of the Maissade
is therefore late Early Miocene to Middle Miocene (Butterlin, 1960),
but some beds similar to Maissade can probably reach the Pliocene
(Maurrasse, 1982). The Maissade represents an initial shallow marine
landward invasion followed by a coastal-swamp environment (Woodring,
1922; Woodring et al., 1924; Maurrasse, 1982). In many respects,
the Maissade Formation resembles the amberiferous Yanigua Formation of
the Dominican Republic, the Cibao Formation of Puerto Rico, and the Los
Arabos, Lagunitas, and Magantilla Formations of Cuba, although amber
has not been reported from these last units.
PUERTO Rico
155
Amber has been discovered in two localities in Puerto
Rico, one of Oligocene and the other of Miocene age. The
Oligocene amber is a tiny droplet recovered by the author in 1999 from
ligniferous clays of the San Sebastian Formation near Lares (Figs. 1,
7). Miocene amber was collected from a well south of the
Cementerio Nacional in San Juan (Fig. 7). At the latter locality
the amber occurred in gray-blue marl and sandstone of the upper part of
the late Early to early Middle Miocene Cibao Formation
(Iturralde-Vinent and Hartstein, 1998). Two small nieces of Miocene
amber have been collected from the same horizon. One piece
(originally 20x30x15 mm) is a somewhat flattened, deep red, transparent
fragment, with subangular edges suggesting that it experienced some
transportation. The piece is brittle, has several fractures, and
resembles low quality amber from the eastern mines (Yanigua and Colonia
San Rafael) of the Dominican Republic. The fragility and low
density of the Puerto Rican amber suggest that it suffered slight
diagenetic transformation. This is confirmed by the presence of
noncrystalline lignitic beds in the same stratigraphic position as the
amber, The Miocene lignitic and amberiferous beds are within the
transitional Early to Middle Miocene upper part of the Cibao
Formation. Features common to all these lignitic occurrences are
the deep black color of the sediment, the presence of fresh water
mollusks, and the position within the uppermost part of the Cibao
Formation near the contact with Aguada limestones (Iturralde-Vinent and
Hartstein, 1998). In these beds there are shark, ray, and
barracuda teeth, other unidentified fish elements, crocodilian teeth
and vertebrae, sirenian ribs and vertebrae, and many turtle scuts.
CUBA
Amber has not been discovered in Cuba,
Los Arabos and equivalent Formations (Iturralde-Vinent, 1969).
The best occurrence has been reported in the Las Tunas basin of eastern
Cuba (Figs. 1, 3), where the Neogene strata unconformably overlie
Eocene and Cretaceous rocks within a low relief plain. In this
basin, the Early to Middle Miocene Los Arabos Formation is represented
by more than 130 m of marls, calcareous clays, and marly limestones,
with thin lenses of gypsum. Some isolated beds of lignite and
plant-bearing clays are found. Clays are more common toward the
bottom of the unit, while carbonate material is more abundant in the
upper region. Pyrite and microfossils with pyrite-filled chambers
are common like in the Yanigua Formation. The biocoenosis
includes fresh and brackish water forams, ostracods, mollusks, and
echinoids. The Los Arabos Formation is conformably overlaid by
more than 80 m of the Guines/Vasquez Formations, represented by Middle
Miocene shallow-water marine limestone (Iturralde-Vinent, 1969).
The resin-producing tree Hymenaea torrei, the closest relative of H.
protera and H. courbaril, occurs today in eastern Cuba (Bisse, 1988).
156
Jamaica
The oldest occurrence of amber in the but Miocene ligniferous
rocks occur in the Greater Antilles, from the Maastrichtian-Paleocene
Cross Pass Formation of Jamaica was reported recently to the author by
Grenville Draper (pers. comm.). According to Draper, during a field
trip led by Geoffrey Wedge in 1971, they found some pieces of dark
brown amber with many dark inclusions, within a bedding plane of the
well-bedded shales of the Cross Pass Formation. The amber was
found at a small waterfall, close to the Indian Cony River near Bath in
the Blue Mountains. According to Wadge and Draper (1978), the
Cross Pass Formation is a deep marine well bedded with arenaceous to
argillaceous cyclic turbidites, with many intercalated beds of coarse
volcaniclastic sandstones. The unit was dated as
Maastrichtian-Paleocene according to its stratigraphic position and the
occurrence of reworked latest Cretaceous microfossils (Pseudorbitoides?
rutteni rutteni and Globotruncana sp.) (Krijnen and Lee Chin, 1978).
This isolated occurrence of amber in the Blue Mountains is important
because it is the oldest from the Caribbean realm. However, the
specimens are not available for study and their origin from a
leguminous or other tree cannot be determined. Additional search
should be conducted within the outcrops of the Cross Pass Formation to
collect more amber.
AGE of THE GREATER ANITLLEAN
AMBER
The precise determination of the age of the amber deposits is very
important because current data suggests that Dominican amber has the
same age as the sediments where it is found, and because the age
generally coincides with a climate maxima dated elsewhere as 16 my
(Tsuchi, 1990), which may have enhanced resin production.
Several authors have proposed that Dominican amber may be older than
the strata in which it occurs, from Cretaceous (Brouwer and Brouwer,
1982) to pre-Lower Miocene (Baroni-Urbani and Saunders, 1982).
This has been discussed in papers by Grimaldi (1995) and
Iturralde-Vinent and MacPhee (1966) who concluded that there is no
solid data to indicate that amber has been redeposited from older
deposits.
However, lack of redeposition from older rocks does not rule out that
contemporaneous relocation may have taken place during the
transportation of copal from the forest litter to the final repository.
Arguing against redeposition from older deposits is the fact that
Dominican amber has not been reported, in large amounts, from rocks of
indisputable Cretaceous, Paleocene or Eocene age. The only
Oligocene report is a tiny droplet found by the author in ligniferous
clays of the San Sebastian Formation in Puerto Rico, and this is no
proof that there were amberiferous Oligocene sediments in the Greater
Antilles. Furthermore, since Eocene and Oligocene rocks are not
exposed in the catchment areas of the eastern mining district (Brouwer
and Brouwer, 1982; Toloczyki and Ramirez, 1991) they cannot be the
source of older amber, if such amber deposits ever existed.
Paleogene rocks occur in the Cordillera Central and Cordillera
Septentrional of Hispaniola, west and south of the northern district,
but they do not contain amber (Fig. 3). Also, the mean paleocurrent
directions measured in the La Toca Formation (Dolan et al., 1991)
suggest that those Paleogene rocks were not the source of the clastic
material in the amberiferous deposits. These measurements suggest
that sediments in the northern district were derived from a Source
located southeast, which points to the same catchment basin that
produced the eastern district sediments (see also Iturralde-Vinent and
MacPhee, 1996).
In connection with the age of Dominican amber, it is important that
organisms preserved in amber can almost always be allocated to extant
groups at low hierarchical levels and many seem to differ little from
relatives living in Hispaniola today (cf. Baroni-Urbani and Saunders,
1982; Grimaldi, 1995, 1996; Poinar and Poinar, 1999). If
Dominican amber is of late Mesozoic or Paleogene age, more primitive
faunal elements should be well represented (Grimaldi, 1996). For
all these reasons, it is logical to conclude that Dominican amber is of
the same age as the sediments that contain it. That is, that
redeposition of amber of different ages did not occur (Iturralde-Vinent
and MacPhee, 1996).
157
Paleocene Cross Pass Formation of Jamaica was reported recently to the
author by Grenville Draper (pers. comm.). According to Draper, during a
field trip led by Geoffrey Wedge in 1971, they found some pieces of
dark brown amber with many dark inclusions, within a bedding plane of
the well-bedded shales of the Cross Pass Formation. The amber was
found at a small waterfall, close to the Indian Cony River near Bath in
the Blue Mountains. According to Wadge and Draper (1978), the
Cross Pass Formation is a deep marine well bedded with arenaceous to
argillaceous cyclic turbidites, with many intercalated beds of coarse
volcaniclastic sandstones. The unit was dated as
Maastrichtian-Paleocene according to its stratigraphic position and the
occurrence of reworked latest Cretaceous microfossils (Pseudorbitoides?
rutteni rutteni and Globotruncana sp.) (Krijnen and Lee Chin, 1978).
This isolated occurrence of amber in the Blue Mountains is important
because it is the oldest from the Caribbean realm. However, the
specimens are not available for study and their origin from a
leguminous or other tree cannot be determined. Additional search
should be conducted within the outcrops of the Cross Pass Formation to
collect more amber.
AGE OF THE GREATER Antillean
AMBER
The precise determination of the age of the amber deposits is very
important because current data suggests that Dominican amber has the
same age as the sediments where it is found, and because the age
generally coincides with a climate maxima dated elsewhere as 16 my
(Tsuchi, 1990) which may have enhanced resin production.
Several authors have proposed that Dominican amber may be older than
the strata in which it occurs, from Cretaceous (Brouwer and Brouwer,
1982) to pre-Lower Miocene (Baroni-Urbani and Saunders, 1982).
This has been discussed in papers by Grimaldi (1995) and
Iturralde-Vinent and MacPhee (1966), who concluded that there is no
solid data to indicate that amber has been redeposited from older
deposits.
However, lack of redeposition from older rocks does not rule out that
contemporaneous relocation may have taken place during the
transportation of copal from the forest fitter to the final repository.
Arguing against redeposition from older deposits is the fact that
Dominican amber has not been reported, in large amounts, from rocks of
indisputable Cretaceous, Paleocene or Eocene age. The only
Oligocene report is a tiny droplet found by the author in ligniferous
clays of the San Sebastian Formation in Puerto Rico, and this is no
proof that there were amberiferous Oligocene sediments in the Greater
Antilles. Furthermore, since Eocene and Oligocene rocks are not
exposed in the catchment areas of the eastern mining district (Brouwer
and Brouwer, 1982; Toloczyki and Ramirez, 1991) they cannot be the
source of older amber, if such amber deposits ever existed.
Paleogene rocks occur in the Cordillera Central and Cordillera
Septentrional of Hispaniola, west and south of the northern district,
but they do not contain amber (Fig. 3). Also, the mean paleocurrent
directions measured in the La Toca Formation (Dolan et al., 1991)
suggest that those Paleogene rocks were not the source of the clastic
material in the amberiferous deposits. These measurements suggest
that sediments in the northern district were derived from a Source
located southeast, which points to the same catchment basin that
produced the eastern district sediments (see also Iturralde-Vinent and
MacPhee, 1996).
In connection with the age of Dominican amber, it is important that
organisms preserved in amber can almost always be allocated to extant
groups at low hierarchical levels and many seem to differ little from
relatives living in Hispaniola today (cf. Baroni-Urbani and Saunders,
1982; Grimaldi, 1995, 1996; Poinar and Poinar, 1999). If Dominican
amber is of late Mesozoic or Paleogene age, more primitive faunal
elements should be well represented (Grimaldi, 1996). For all
these reasons, it is logical to conclude that Dominican amber is of the
same age as the sediments that contain I; That is that redeposition of
amber of different ages did not occur (Iturralde-Vinent and MacPhee,
1966).
This conclusion does not, however, agree with efforts to date amber
using exomethylene resonance signatures visualized by nuclear magnetic
resonance spectroscopy (NMRS) (Lambert et al., 1985). To derive
an age assessment by this method, resonance intensity must first be
calibrated against NMRS results for specimens of known age. The
only calibration curve (Lambert et al., op. cit.) relevant to the
dating of Dominican amber is based on two data points: amber from Palo
Alto mine (northern district), accepted as Early Miocene because
sediments yield microfossils of that age (Baroni-Urbani and Saunders,
1982); and a resin sample from Hymenaea courbaril. Age estimates
based on this curve (Lambert et al., 1985) indicate a Late Eocene age
for amber recovered from mines at La Toca and Tamboril in the northern
district, and a Middle Miocene age for specimens from Bayaguana and
Cotuf in the eastern area. Microfossil evidence supports a
Miocene age for amber-bearing sediments in the mines at Bayaguana (Van
den Bold, 1988), but microfossils also establish that the La Toca mines
are Miocene, which contradicts the NMRS results. If amberiferous
sediments at La Toca, Palo Alto, and Bayaguana are paleontologically
equivalent in age, then the exomethylene decay curve does not produce
meaningful results (Grimaldi, 1996; Iturralde-Vinent and MacPhee,
(1996).
The most interesting aspect of the Miocene amber found in Puerto Rico
is that it was found in the uppermost part of the Cibao Formation,
dated as Early to Middle Miocene (with large Sorites marginalis and the
absence of Miogypsina sp. and Lepidocyclina sp., which occur
stratigraphically below). The Cibao Formation underlies the
Middle Miocene Aguada Formation (with Archaias angulatus, S.
marginalis, and Peneroplis sp.) (Monroe, 1980; Iturralde-Vinent and
Hartstein, 1998). These data place the Miocene amber-bearing bed
of Puerto Rico within the same time frame as those of Hispaniola (Fig.
3).
PALEOGEOGRAPHY
Having established the age of the amber in Hispaniola and Puerto Rico
(Cibao Formation) as late Early to early Middle Miocene (20-15 my,
probably close to 16 my), the next step is to produce a paleogeographic
map to determine the origin of the amber. Several attempts have
been made to produce a paleogeographic reconstruction and historical
geologic interpretation of the Greater Antilles during the Late
Tertiary, but although authors were well aware of plate tectonics, they
produced essentially fixist reconstructions (Maurrasse, 1982).
Plate tectonic reconstructions such as those of Calais et al. (1992) or
Pindell (1994) are directed to solving geometric problems of terrane
location and not of physical geography (relieve). Only recently
has there been an attempt to presented maps combining tectonic
reconstruction and physical paleogeography (Iturralde-Vinent and
MacPhee, 1996, 1999). Developing this theme, three maps are
presented to illustrate the paleogeographic scenario of the Greater
Antilles before, during, and after formation of the Dominican amber
(Fig. 8). The data to construct these maps were compiled by
Iturralde-Vinent and MacPhee (1 996, 1999) and Iturralde-Vinent (1969).
Paleogeography
during the amber epoch
The paleogeographic
reconstruction presented here displays the Greater Antilles Ridge as
subdivided into several groups of islands separated due to the presence
of a shallow water shelf surrounded by moderate to deep-water channels
or basins. The first scenario for the Late Oligocene (27-25 m.y.)
displays three main archipelagos as Western Cuba, Central Cuba and
eastern Cuba-Cordillera Central-Puerto Rico Virgin Islands (Fig.
8). These insular groups changed little until the Early-Middle
Miocene (Fig. 8:16-14 m.y.), when separation between eastern Cuba and
Hispaniola P.R.-Virgin Islands became apparent due to activity along
the Oriente fault (MacPhee and Iturralde-Vinent, 1995). It is at
this time when resin was produced in large quantities in northern
Hispaniola and in small amounts in southern Hispaniola and northern
Puerto Rico. During the- Middle Miocene (Fig. 8: 12-10 my) the
subdivision of eastern Cuba and Hispaniola was completed and resin
production had declined and stopped within the interval; a probable
cause being a climate change induced by the new oceanographic scenario
that arose from the opening of a marine channel between eastern Cuba
and the Cordillera Central of Hispaniola.
In
the early Neogene, terranes of northern Hispaniola were located west of
their current positions,
closer to present day southeastern Cuba.
Then, the eastern and northern mining districts were part of the same
structural depression, located north of the ancestral Cordillera
Central. Their present (Holocene) position is due to
post-Oligocene left-lateral displacement along the northern Caribbean
plate boundary. It is proposed here that the paleogeographic
basin represented by both the northern district and the eastern
district during the Neogene be called the Comatillo Basin (Fig. 8)
because there is no such feature today in the geography of the island.
The
relief and configuration of paleo Hispaniola was very peculiar in the
Early to Middle Miocene (16-14 m.y.). A large uplifted ridge that
includes present-day Montagnes Noires, Cordillera Central, Cordillera
Oriental, and continued into Puerto Rico and the Virgin Islands was at
that time the largest landmass in the Greater Antilles.
Several small islands formed an archipelago south of the Plateau
Central-San Juan Basin, and westward another large island was present
which is represented today by southeastern Cuba (Fig. 8). The
relief of these islands was differentiated with low and high
topography. Some mountains were present west of the Comatillo
Basin embayment and were the source for the huge amounts of clastics
deposited in the structural depression known today as La Toca block and
the Cibao basin (La Toca and Baitoa Formations). Highlands were
also present in southeastern Cuba and provided the clastic deposits of
the Maquey and Imias Formations (Nagy et al., 1983; Fig. 3).
The mountains in both source areas were dissected by small rivers that
discharged into the littoral plains, giving rise to siliciclastic ramps
were alluvial fans and deltas typically developed, dissected the
mountains in both source areas. Massive coarse-grained debris
flows that deposited thick conglomeratic beds in the littoral and basin
areas occurred locally. The rest of the uplifted territory was
probably lowland, as the seashore was characterized by lagoons and
mangrove swamps. These fresh to brackish-water lagoons were
filled with very fine-grained sediments, resulting from slope-wash
drainage of the surroundings areas. There were locally small
rivers that deposited some gravel and coarser clastics.
The plant-derived organic debris in the amber and in the clastic rocks
north of the Hispaniolan paleo-island suggest that the highlands were
covered by humid tropical forest, like that which exists today on the
windward side of higher elevations throughout the Greater
Antilles. Humid forests also occurred in the lowlands facing the
lagoonal-mangrove swamp coastal areas, as suggested by the common
occurrence of lignite and the high organic content of the sediments.
The ecological interpretation of the amber biota (Grimaldi, 1996;
Poinar and Poinar, 1999) and the presence of amber in the sediments,
indicate that Hymenaea trees were abundant in the humid forest
surrounding the Comatillo Basin (Fig. 9) on the northern slope of the
Hispaniolan paleo-island. The presence of some amber in the
Sombrerito, Maissade, and Cibao Formations suggest that some Hymenaea
trees were also present in the southwestern slope of the Hispaniolan
area and in the northeastern slope of the Puerto Rican area.
ORIGIN OF THE MIOCENE AMBER
160
Miocene Hispaniolan and Puerto Rican lignitic and amber deposits occur
within the same time interval as Miocene lignitic deposits of Cuba
(Fig. 1). This suggests that the unusual accumulation of Early to
Middle Miocene amber and lignitic deposits was correlated with a
nonsedimentological event, probably a warm climatic optimum (Tsuchi,
1990). However, a single factor is not sufficient to explain the
production and accumulation of large amber deposits of the Dominican
Republic (Schlee, 1990; Grimaldi, 1996).
There are important differences between the Miocene lignitic lagoonal
deposits of the Dominican Republic and those of Cuba, Haiti, and Puerto
Rico. These differences are the state of diagenesis and the
thickness of the lignitic beds. Lignite in the Yanigua Formation
of the Dominican Republic is more crystalline and thicker than in the
other deposits (Los Arabos, Magantilla, and Cibao Formations).
This suggests a larger primary accumulation of plant remains in the
original basin and deeper post-depositional burial. The lignite
found in Cuba, Haiti, and Puerto Rico is not crystalline and occurs in
thin beds generally represented by ligniferous clays. The author
believes that the occurrence of important concentrations of amber in
the mining districts of the Dominican Republic is due to the
combination of several factors that are considered in the following
paragraphs (see also Figs. 9 and 10).
The botanical factor is the unusual abundance of a peculiar species of
Hymenaea, presumably H. protera, in the Early Neogene of the Greater
Antilles. This species produced large amounts of resin resistant
to weathering and biological degradation. After segregation, the
resin was polymerized, transformed into copal (Langenheim, 1990), and
accumulated in the ground with other plant debris as part of the
organic rich litter (Fig. 10A, Grimaldi, 1996). The occurrence of
abundant plant and solid inclusions in the amber, as well as present
day accumulation of copal in the soil around H. courbaril trees,
demonstrate the importance of this factor.
It is generally accepted that the production of resin by H. courbaril
is the result of mechanical and/or biotic wounds because resin protects
against infection (Langenheim, 1990; Grimaldi, 1996). However,
during 1996 and 1997 the author observed in the Dominican Republic
several H. courbaril trees that naturally exude very small amounts of
resin (also noted by Wu, 1997). Resin was not produced when
branches were artificially broken or when the cortex was damaged.
The author observed no resin production by H. courbaril during November
1998, several weeks after Hurricane Georges severely damaged many trees
in the Dominican Republic The only place where H. courbaril contained
resin was under the bark, in places where termites, ants, and other
insects built their nests. Unexpectedly, the resin was clear and
contained no trapped animals. Botanists working with species of
Hymenaea in Peru and Bolivia have not noticed massive exudates of resin
(Robin Foster, pers. comm.).
162
According to Sergio Leiva, an old amber miner and nature observer from
El Valle, Dominican Republic, H. courbaril produces small amounts of
resin about two weeks after its cortex is damaged. He indicates
that large amounts of resin are produced only if the tree is struck by
lighting. Sergio showed the author that digging deeply around the
trunk of an old H. courbaril copal might be found, sometimes in large
amounts and with biotic inclusions. It is well known in the
Dominican Republic that copal is sometimes collected and sold as amber.
Brouwer and Brouwer (1982) considered that forest fires could force
trees to produce larger amounts of resin. High temperature
produced by fire could work as efficiently as lighting, but significant
amounts of ash or burned animals and plants have not been reported in
Dominican amber.
The climate-geography factor is related to the position of the
Hymenaea-rich forest on the northern slope of the Hispaniolan
highlands, south of the Comatillo lagoonal marine basin (Fig. 9).
Today, a humid rain forest (more than 2000 mm/year rainfall) exists on
the northern slopes of the Greater Antilles because winds bring
moisture that produces rainfall. The northern slope of the
Miocene mountains probably received much rainfall and the forest was
frequently exposed to rainstorms and thunderstorms. The high
relative humidity would have enhanced bacterial, fungal and in-sect
attacks on the trees (Langenheim, 1990), while the action of the wind
produced mechanical damage.
The hydrodynamic factor is related to the rapid burial of the
copal. This was very important because weather destroys exposed
copal. Frequent rains washed the organic rich soil of the forest
and transported copal down-slope. In the Comatillo basin, organic
litter was efficiently transported and deposited in the lower reaches
of the slope, at the lagoonal and coastal areas, creating a growing bed
of turf (peat) intercalated with organic-rich clays containing large
amounts of copal. This process was favored by the fact that the
Hymenaea-rich forest faced a well-protected marine embayment (Figs. 8,
9). At least three main deposition environments were present in
this basin: lagoonal, deltaic, and deep marine.
In river and lagoons, resin and copal will float in fast moving water
but sink as soon as the current is negligible (Fig. 10: B). In
the sea, some copal will float and disperse in high-energy environments
and some will sink. Sinking may be enhanced by high density
material adhered to the copal, such as soil and decayed wood, and by
the occurrence of density flows which carry the copal mixed with mud
(Fig. 10: C). The occurrence of isolated fragments of amber in
well-documented turbidites has been reported (Garcia and Harms, 1988;
Harms, 1990), so it is likely that copal can sink into deep-water
deposits. Theoretically, a high energy, heavy mudflow can carry
large amounts of copal down-slope if they are well incorporated in the
high-density mud. When the mudflow reached the seafloor and
spread laterally in layers a few centimeters thick, the copal probably
concentrated along with the sand. The occurrence of thin beds of
lignite within this deepwater section is probably due to secondary
accumulation of the lighter vegetal debris embedded in water and
carried into the sea by the density flow.
The Yanigua Formation represents the depositional scenario in the
low-energy lagoon and coastal swamp environments. The reduced
amount of rock detritus in Yanigua clearly indicates that erosion in
the forest facing the lagoon was not very extensive, and that the soil
and organic litter were mostly removed during rainstorms. Some
isolated turbulent flows that deposited lenticular beds of gravel
within the clay are so limited in area and scale, that they probably
represent the action of local high-energy water conducts, such as local
creeks.
The amber deposited in the scenario of deltaic and. open marine
environments (La Toca Formation) followed a different process (Fig. 10:
C). Rivers that cut across the Cordillera Central carried large
amounts of silt, sand, and gravel that became intermingled with the
slope-wash organic-rich input. As is common for sediment-rich
rivers, large deltaic-fan deposits 'probably developed in the
silicidastic ramp facing the marine basin (Fig. 9). The copal
bearing alluvium was deposited in the river delta, mostly along the
low-energy environments of the interchannel flat. Frequent
sinking of plant debris probably concentrated as turf layers which
diagenesis transformed into lignite. As is normal in deltas,
contemporaneous erosion of already deposited beds will carry fragments
of turf (future lignite) and copal (future amber) into deeper areas of
deposition in the basin sea floor (Fig. 10: C). Therefore, in the
sedimentary basin the hydrodynamic factor was responsible for rapid
burial and resedimentation (recycling) of the copal in the different
depositional environments. This guarantees that the copal was
additionally protected from the action of erosion and surface
biodegradation, which is the first and most important requisite for
fossilization.
163
Amber is usually found as local concentrations in particular horizons
(Brouwer and Brouwer, 1982; Redmond, 1982; Eberle et al., 1982).
This implies that the factor controlling the concentration of copal
fragments operates during sedimentations combination of paleo-relief
and transport. In the lagoonal and coastal swamp environments,
where slope-wash caused by rain is the main way of transporting copal
to the basin, it can be suggested that the paleorelief of the basin
floor and source areas determined the areas of accumulation and
dispersion. Where a mountain ridge faced the basin, little copal
accumulated in front of the ridge, whereas increased accumulation took
place in places where a local amphitheater faced the basin. The
irregularities of the water table in the lagoon and coastal swamps
created additional local barriers where more copal accumulated; this
was probably most important in the deltaic flat, where slow-moving
waters or local sandy bars and vegetal barriers probably enhanced the
accumulation of organic rich material, including copal. The way
copal was concentrated in deep-water sediments in the La Toca Formation
can also result from secondary concentration due to density segregation
within the mud, and the shape and extension of the turbiditic lobes.
Diagenesis was another key factor in the formation of amber (Fig.
10:D).
Diagenesis includes the chemical and physical modifications that
occurred in the sediments after deposition. Authigenic pyrite is
found in the lagoonal deposits of the Yanigua Formation, suggesting
that diagenesis took place in an anoxic, sulfur-rich environment
(Brouwer and Brouwer, 1982). In the Yanigua Formation, the lignitic
beds are thick, very hard, brittle, and the intercalated clays and
silts have gone through compaction strong enough to break and flatten
mollusk shells. This phenomenon was produced by confined pressure
due to deep burial of the copal-bearing sediments, probably up to one
kilometer below the earth's surface, since the Yanigua Formation is
covered by a few hundred meters of limestone (Los Haitises
Formation). This event, which lasted several million years,
placed the sedimentary pile under conditions of increased temperature
and pressure, ultimately producing sediment compaction, diagenesis, and
lithification. In the La Toca Formation, diagenesis also took
place in anoxic conditions, as demonstrated by the dark color of the
organic-rich amberiferous beds. As in the previous example, the
sediments were deeply buried, more than one kilometer below earth's
surface, during several millions of years. Copal matured under
this condition until it became amber.
Still, it is a puzzle why the Maissade and Cibao Formations do not
contain large accumulations of amber, since they have the same
lithology and age as the Yanigua Formation (Fig. 2). Another
puzzle is why amber is unknown from Cuba and South America, where
Hymenaea trees are common today. No less problematic is the rare
occurrence of small amounts of amber in the Sombrerito Formation of
south central Hispaniola (Garcia and Harms, 1988). The
explanation is probably related to some missing factor among those
previously listed for the amber-bearing region of northern
Hispaniola. For example, it is evident in the paleogeographic map
(Fig. 8) that a large slope forest, providing more area for the
development of the Hymenaea protera forest, surrounded the Comatillo
basin. Land areas near the Plateau Central-San Juan basin of
Hispaniola and the Las Tunas basin of Cuba were probably smaller.
Perhaps H. protera was not widely distributed in the Greater Antilles,
or maybe the tree was more abundant in the northern slope of
paleo-Hispaniola due to particularly favorable conditions of altitude
and humidity.
164
A secondary factor that may be involved is modem topography (Fig. 10:
E). For example, the dissected mountain ranges efficiently expose
the amber-bearing deposits, in such a way that many years ago amber was
even mined along rivers. The Maissade and Los Arabos Formations
outcrop in areas of less dissected relief, and natural exposures are
sparse and heavily weathered. Geological exploration is required
before a conclusion can be reached about the amber prospect of these
formations. Nevertheless, the Sombrerito (south central
Hispaniola), San Sebastian, and Cibao Formations (Puerto Rico) are well
exposed in deeply dissected areas, but only small isolated fragments of
amber have been reported (Garcia and Harms, 1988; Iturralde-Vinent and
Hartstein, 1998).
In conclusion, the origin of the unusually large Miocene deposits of
amber in Dominican Republic remain a puzzle, but can probably be
explained as the fortunate combination of:
i) Adequate conditions of relief and soil for the development of a
large local population of resin-producing trees;
ii) Abundant occurrence of Hymenaea protera, with peculiar
characteristics for the production of resin;
iii) A timely constrained warm and humid climatic optimum nearly 16
m.y. ago, which enhanced the production of resin, maybe as a
consequence of larger populations of tree-dwelling bacteria, insects
and fungi, but especially due to frequent thunderstorms;
iv) Special conditions for the rapid deposition of copal in marine
basins located near the source;
v) Adequate environment of diagenesis due to deep burial of copal and
turf bearing sediments.
165
Acknowledgements
I thank Ross MacPhee and David Grimaldi (American Museum of Natural
History, N.Y.), Salvador Brouwer (Sociedad Dominicana de Ge6logos,
D.R.). Ivin Tabares (Direcci6n General de Mineria, D.R.), Victor
GonzAlez (San JuAn, - Puerto Rico), Jake Brodzinski (Santo Domingo,
D.R.), Jorge Caridad (Museo Mundo de Ambar, D.R.), Dyoris Perez (Museo
Nacional de Historia Natural, D.R.), and Joss A. C)ttenwalder (Oficina
PNUD, D.R.) for assistance during field work.
Consuelo Diaz and Rafaela Perez (histituto de Geologia y Paleontologia,
La Habana), Gena Fernandez (New York Botanical Garden, N.Y.), W.A. van
den Bold (Holland), Paula M. Mikkelsen (American Museum of Natural
History, N.Y.), and Timothy Bralower (University of North Carolina at
Chapel Hill) identified the fossils from the samples collected by the
author.
Grenville Draper (Florida International University) kindly provided
unpublished data about the occurrence of amber in Jamaica.
Grenville Draper, R. A. Davis, Jr. (University of Southern Florida),
and Robin Foster (Field Museum of Natural History, Chicago) reviewed
the manuscript and made many useful suggestions. This work was
supported by grants from the Office of Grant and Fellowships of the
American Museum of Natural History, the former RARE Center for Tropical
Conservation, and the National Geographic Society (600997).
Literature CITED
Baroni-Urbani, C. and J. B. Saunders, 1982. The fauna of the
Dominican Republic amber: the present status of knowledge. Trans.
9h. Carib. Geol. Conf., Santo Domingo 1:213-223.
Berggren, W. A., D. V. Kent, C. C. Swisher, and M.-P. Aubry.
1995. A revised Cenozoic geochronology and
chronostratigraphy. Spec. Publ. Soc. EcorL
Paleontol. Mineral. 54:129-21?
Bermudez, P.J. 1949. Tertiary smaller foraminifera of the
Dominican Republic. Cushman @b. Foram. Res.,
Spec. Publ. 25:1-39?
Bisse, J. 1988. Arboles de Cuba. Edit.
Cient.-Tecnica, La Habana, 384 pp.
Bowin, C.O., 1966. Geology of Central Dominican Republic (A case
history of part of an island arc). GSA Mem. 98:11-84
Brouwer, S. B. and P. A. Brouwer, 198Z Geologia de la region ambarifera
oriental de la Repdblica Dominicana. Trans. 9th Caribbean
Geol. Conf., Santo Domingo 1: 305-322.
Butterlin, J., 1960. Geologie G6n6rale et R6gionale de la
R6publique D'Haiti. Universit6 de Paris, Trav. et Mem.de L'Inst.
des Hautes Etudes de I'Amerique Latine-VI: 194 pp).
Tsuchi, R- (ed.). 1990. Pacific Neogene events; their timing,
nature and interrelationship. University of Tokyo Press, 206 pp.
Tyso@ R.V. 1995. Sediinentary organic matter. Organic
facies and palynefacies. Chapman and Hall, London, 615 pp. Van
den ]3old, W. 1988. Neogene paleontology in the northern
Dominican Republic: 7. The subclass 0stracoda (Arthropoda:
Crustacea). Bull. Amer. Paleontol. 94(329):1-79.
Vaughan, T.W., W. Cooke, D. D. Condit, W.P. Woodrin& and F.C. Ca@s.
1922. A geological reconnaissance of Dominican Republic.
Servicio Geo16,gico Rep. Dominicana 1, 302 pp.
Wadge, G. and G. Draper. 1978. Structural geology of the
southeastern Blue Mountains, Jamaica. Geologie en heinbouw
57(2):347-352.
Woodrin& W. P. 1922. Stratigraphy, structure, and possible
oil resources of the Miocene rocks of the Central Plain. Republic
of Haiti, Departinent of Public Works-Geological Survey of the Republic
of Haiti, 19 pp.
Woodring, W. P., J. S. Brown, and W. S. Burbank. 1924. Geology of
the Republic of Haiti, Lord Baltimore Press, Baltimore, 631 pp. Wu, F-
J. C. 1997. @ets of the lost world: Dominican amber and its
inclusi@. Printed in the Don-dnican Republic, 222 pp.
NOTE ADDED IN PROOF
During a recent trip to Haiti (September 15-30, 2001) supported by an
NGS research grant to the author, a crew of the Museo Nadonal de
Historia Natural (Reinaldo Rojas, Stephen Diaz, and the author)
discovered amber in two localities of the Plateau Centrdl, near
Hinche. Preliminary field observations indicate that both sites
are in Miocene sediments strongly resembling the Cibao Formation of
Puerto Rico and the Lagunitas Formation of Cuba. The main
dffference with the amberiferous Yariigua Formation in Dominican
Republic is that the sections in Haiti contain noncrystalline
ligniferous horizons. One of the amberiferous horizons, located
north of Thomonde, is represented by indurate gray calcareous
claystones, with isolated corals, mollusks, and teeth of bony fish,
shark and rays, as well as turtle fragments. Also present are
large specimens of Sorites sp. like those from the Lagunitas, Cibao,
and Yanigua Formations. The amber was found within strongly
oxidized subrounded inclusions of vegetal origin (seeds, fragments of
roots, and other plant parts), some of which are dark reddish pieces of
indurate resinite, with diameters less than 10 cm. The second
locality, northwest of Hinche, is an aluviodeltaic section of gray to
light brown sandy clays with intercalations of ligniferous beds,
conglomerates, gravels, and isolated nodular limestone lenses.
The ligniferous beds are poorly indurate, few centimeters thick clays
and sandy clays with lamellar levels of black plant fragments.
These horizons contained tiny pieces of light colored yellowish
indurate resinite. No macrofossils other than vegetal fragments
were observed. These new amberiferous sites discovered in Haiti
are particularly interesting because they are of the same general age
as other miocene amber deposits in Hispaniola and Puerto Rico.
|