|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
FLORAL COMPOSITION OF A LOWER CRETACEOUS PALEOTROPICAL ECOSYSTEM
INFERRED FROM QUANTITATIVE PALYNOLOGY
PAULA JENIFER MEJIA VELASQUEZ
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
2007 Paula Jenifer Mejia Velasquez
To my Mom, my Daddy, Kata, Jeisson and Lucas.
I would like to thank my advisor and committee chair Dr. David Dilcher for general advice
in my research, review of my chapters, patience with all my questions, support of all my ideas,
and in general, everything.
I would like to thank my committee members for their useful feedback and comments, the
Smithsonian Institute for supporting my internship in Panama, which was an important step in
the analyses of my samples, to the Colombian Institute of Petroleum and Petrobras for allowing
the sampling of the cores for this study and the Evolving Earth Foundation for the funding they
gave for this study.
I also would like to thank my fiancee, Lucas, who helped review the text, but most
important, who gave me very good ideas through discussion, helped me with the statistics, gave
me anti-stress treatment with his enthusiasm, and gave all the love and support I needed.
Finally I want to thank my family in Colombia, my mother for all her love and support, to
my daddy for all his love, to my sister Kata and my brother Jeisson for their support.
TABLE OF CONTENTS
A CK N O W LED G M EN T S ................................................................. ........... ............. .....
LIST OF FIGURES ................................. .. ..... ..... ................. .7
1 INTRODUCTION ............... ................. ........... .............................. 10
2 M A TER IA L S A N D M ETH O D S ........................................ .............................................17
S a m p lin g ................... ...................1...................7..........
Laboratory Procedures ........................................................ .......... .......... .... 18
A nalyses...................................................................... 19
Statistical M methods ................................................. 20
R arefaction .............................................................................20
D distribution and V ariance T est.................................................................................. 20
A b u n d a n c e ................................................................................................................. 2 1
Species R richness ................................................. ... ... .. .. ... ............ 21
Cluster Analysis: Grouping Samples with Similar Composition.............. .....................22
Relationship between Species Distribution and Lithology ..........................................22
Comparison between Paleotropical and North American Samples..............................23
3 R E SU L T S ...........................................................................................2 6
A b u n d an ce ................... ...................2...................6..........
S p e cie s R ich n e ss ......................................................................................................2 7
Hierarchical Cluster Analyses: Sample Associations........................... ...........29
Multi-response Permutation Procedure (MRPP): Species Distribution and
L ithology R relationship ....................................................... ... .. ........ ........... 30
Comparison of Abundance and Number of Species between the Paleotropical Site
Studied with Middle and High Paleolatitude Sites of North America.............. .............. 31
R elativ e A b u n d an ce ................................................................................................ 3 1
Species R richness ....................................................... 32
4 D IS C U S S IO N ........................................................................................................4 2
Floral Composition of the Tropical Site Analyzed...................... .... ... .................. 42
Differences in Floristic Composition between the Paleotropical Site Analyzed and
P aleotem operate L attitudes .......................................................................... ....................48
A SPECIES COUNTS PER SAMPLE.......... ........ ................. ... ............... 52
B PH O TO G R A PH IC PL A TE S ........................................................................ ...................61
L IST O F R E F E R E N C E S ............................................................................... ...........................79
B IO G R A PH IC A L SK E T C H ............................................................................... .....................85
LIST OF FIGURES
1-1. Upper Magdalena Valley (UVM) in Colombia showing the geographical location of
L os M angos field. ......................................................... ................. 16
2-1. Lithological column of Los Mangos 31 core and sample locations. ...................................25
3-1. Absolute abundance of angiosperm pollen, gymnosperm pollen and spores represented
as the total number of individuals found in each one of the samples (all samples
rarefied to 200 counts). .......................... ...... ....................... .... .. ..... ........ 34
3-2. Absolute richness of angiosperm pollen, gymnosperm pollen and spores represented as
the total number of species found in each one of the samples (all samples rarefied to
200 counts) ......... .................. ....................................... ............................35
3-3. Dendrogram showing the different lithological associations based upon their species
composition for the analyzed core. ..... ........................... .......................................36
3-4. Dendrogram showing the different associations of depositional environments based
upon their species associations for the analyzed core................................ ............... 37
3-5. Relative abundances of palynomorphs found in each kind of lithology. The number in
parenthesis indicates the number of samples......................................................................38
3-6. Relative species richness of palynomorphs found in each lithology. The number in
parenthesis indicates the number of samples......................................................................38
3-7. Comparison of the relative abundances of angiosperm pollen for the Aptian-Albian
interval between low (site studied), mid and high paleolatitudes ......................................39
3-8. Comparison of the relative abundances of gymnosperm pollen for the Aptian-Albian
interval between low, mid and high paleolatitudes..................... ...................... 39
3-9. Comparison of the relative abundance of spores for the Aptian-Albian interval between
low m id and high paleolatitudes. ............................................... .............................. 40
3-10. Comparison of the relative species richness of angiosperms for the Aptian-Albian
interval between low, mid and high paleolatitudes..................... ...................... 40
3-11. Comparison of the relative species richness of gymnosperms for the Aptian-Albian
interval between low, mid and high paleolatitudes..................... ...................... 41
3-12. Comparison of the relative species richness of spores for the Aptian-Albian interval
between low, mid and high paleolatitudes ...................................... ............... 41
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
FLORAL COMPOSITION OF A LOWER CRETACEOUS PALEOTROPICAL ECOSYSTEM
INFERRED FROM QUANTITATIVE PALYNOLOGY
Paula Jenifer Mejia Velasquez
Chair: David Dilcher
Angiosperms are the most important floral components of modern ecosystems. It has been
hypothesized that angiosperms originated in low paleolatitudes during the Lower Cretaceous, but
the fossil record of low latitude areas is mainly composed of qualitative data, making it difficult
to make an accurate floral reconstruction of tropical ecosystems of that age. The main objective
of this study was to reconstruct the floral composition of a low paleolatitude ecosystem (Upper
Magdalena Valley, Colombia) in the Lower Cretaceous through quantitative analyses of
palynological samples. The results show that angiosperms were a minor component of the
ecosystem, with medians of 5 individuals and 4 species in the core, followed by gymnosperms
with an average per sample of 65 individuals and 7 species, and finally by spores, which were
the dominant component of the low latitude ecosystem analyzed, with averages per sample of
122 individuals and 21 species. These results differ from the composition of an eastern
paleotropical site, where gymnosperms were the dominant component followed by angiosperms
and spores. These floristic differences may reflect different environmental conditions between
east and west South America during the Aptian-Albian interval. Differences in abundance and
species richness were found between the lower and upper portion of the core analyzed. Higher
angiosperm richness and abundance found in the upper portion of the core are evidence that
angiosperm diversification took place during the Albian. Concurrent with the angiosperm
increase, there was an increase the number of spore species indicating that they were also
diversifying during this age. Furthermore, a comparison between the paleotropical site studied
and published literature from middle and high paleolatitudes shows that paleo tropical
angiosperm abundance and species richness were similar to that of mid paleolatitudes, but higher
in comparison to high latitudes for this age. These results partially support the hypothesis that
angiosperms originated in low latitude areas and later radiated to middle and higher latitudes, in
the sense that they were more abundant and diverse in low latitudes than in high latitudes.
However, although a gradient in angiosperm richness and abundance was observed from the
tropics to high latitudes, differences between low and mid paleo latitudes failed to meet
statistical significance. With these results it is not possible to determine whether angiosperms
originated in low latitudes and radiated to mid latitudes or vice versa. Future analyses of
multiple paleotropical sites will help determine if the patterns observed in this study are
consistent throughout paleotropical ecosystems of Lower Cretaceous age.
Angiosperms are the dominant floristic component of most modern terrestrial ecosystems
(Burnham and Johnson 2004, Friis et al. 1987, Friis et al. 2006, Lupia et al. 1999, Wing and
Boucher 1998). Because of their crucial role in present ecosystems, angiosperm radiation is
considered one of the most significant evolutionary events in the history of the planet (Lidgard
and Crane 1990). Angiosperms originated during the Lower Cretaceous (approx. 135My; Sun
and Dilcher 2002) and subsequently radiated and expanded to become the dominant group of
plants in almost every terrestrial ecosystem by the Upper Cretaceous (Crane and Lidgard 1990).
The earliest palynological records that contains definitive angiosperm pollen come from different
geographic sites, which range from tropical paleolatitudes in Israel (Brenner 1996) to high
paleolatitudes in China (Sun and Dilcher 2002).
It is widely believed that angiosperms appeared first in tropical areas (Brenner 1976,
Crane 1987, Crane and Lidgard 1989, Friis et al. 1987, Lupia 1999, Lupia et al. 2000, Retallack
and Dilcher 1981, Taylor and Hickey 1996, Wing and Boucher 1998) and then radiated to higher
paleolatitudes (Axelrod 1959, Crane and Lidgard 1989). If this hypothesis is true, then it is
expected that during the initial stage of angiosperm radiation the abundance and species richness
of angiosperms would be higher in tropical areas than in higher paleolatitudes. Unfortunately,
there is a lack of quantitative data needed to reconstruct accurately the composition of the flora
from tropical Lower Cretaceous ecosystems. Quantitative data are crucial to determine potential
differences between the floral composition of tropical areas and high and middle paleolatitudes
at that time. Also, the use of quantitative data analyses, such as multivariate techniques, allows
summarization of large datasets and determination of the patterns present in them in a simple
graphical manner (Kovach 1993).
Species richness and abundance are the two main quantitative measures that characterize
the floral composition of an ecosystem. Species richness is simply the number of species present
and abundance is the number of individuals per species (Magurran 2003). It is often difficult to
estimate these variables accurately from the fossil record (palynomorphs) due to the limitations
arising from preservation and representation of individuals in the ecosystems (Lidgard and Crane
1990). Fossil palynomorphs have been used extensively as a data source for population studies
because they are produced in abundance by plants, are extremely durable, and easily dispersed,
deposited, and preserved in sediments (Traverse 1988). Another very important characteristic of
palynomorphs is that a small sediment sample can contain thousands or even millions of
individuals (Traverse 1988). These high numbers increase the probability of finding a good
representation of the population by analyzing the palynological contents of the sediments.
Palynological studies are more numerous in medium and high paleolatitudes than in
paleotropical areas (Crane and Lidgard 1990). High and mid latitudes have numerous
quantitative studies of Cretaceous floras (Lupia et al. 1999). On the other hand, most of the
palynological publications on Lower Cretaceous sites consist of mainly qualitative or descriptive
data and/or taxonomic work (e.g. de Lima 1978, 1979, 1980, 1987, 1989, Dino et al. 1999,
Muller 1966, Regali and Viana 1989). Most of those publications contain biostratigraphic work
made for oil companies. In those studies the objective was to find key species that can be
correlated to age or to specific stratigraphic units. Consequently, most of the past tropical
biostratigraphic publications are focused on marker species, paying little or no attention to other
species in the samples. Of the remaining studies that do not focus on specific species, many
present only the data for selected palynomorphs (such as angiosperms), which creates inaccurate
reconstructions of community composition (e.g. Schrank 1994). Additionally, these studies have
limited use in reconstruction of Lower Cretaceous ecosystems because most do not present
abundance and species richness data. In fact, most publications do not present counts (e.g.
Brenner 1968) and the few that do lack standardized palynomorph counts for each sample (e.g.
Herngreen 1975), which makes comparisons between samples problematic. As an example, a
study by Herngreen (1975) includes palynomorph counts per sample that range from 25 to 326
palynomorphs. Samples with low counts probably do not reflect the real composition of the
ecosystem. Only samples with higher counts would be useful for inferring statistically the
floristic composition of the site.
One of the greatest limitations of the few studies that have published quantitative data
(e.g. Brenner 1974, Ibrahim 1996) is that they present the abundance and diversity of
palynomorphs in grouped intervals (e.g. single, rare, occasional, common and abundant). These
intervals are defined arbitrarily and are consequently different in each study. As an example,
Ibrahim (1996) and Schrank (2002) assign the category "abundant" to values of 11 30% and >
50% of grains in the sample, respectively. Although the total number of individuals in a sample
is known in these studies, this categorization is not accurate enough to infer sample or ecosystem
richness and abundance because with the percentages given it is not possible to make precise
Due to this lack of quantitative data in tropical areas, many questions related to the
appearance and early radiation of angiosperms are still unanswered. Given the limitations of the
data currently available, the main objectives of this study were (1) to provide quantitative
palynological data of a paleotropical ecosystem through the quantitative analysis of a Lower
Cretaceous section from northern South America (Colombia) to infer its floral composition
(abundance and species richness of angiosperm and gymnosperm pollen, and spores of ferns and
allies) and (2) to compare these results with data from higher paleolatitudes, specifically from
North America, to determine whether angiosperms were more abundant and diverse in the
tropical site studied compared to higher paleolatitudes.
The specific questions to be addressed in this paper are: (1) What is the abundance and
species richness of angiosperm pollen, gymnosperm pollen and spores for the low latitude
paleotropical site studied? (2) Were angiosperms more abundant and diverse in the low latitude
site studied compared to higher paleolatitudes during the Aptian-Albian interval? Additionally,
taking advantage of the quantitative dataset obtained, two questions involving specific
characteristics of the analyzed core are also addressed here. Multivariate statistical analyses
provide a valuable tool to determine patterns present in the data that otherwise would remain
hidden or would be less clear. With the use of multivariate techniques I want to answer: (3) Are
there any relationships among the species present in the analyzed samples based on their
distribution though the core? and (4) Does lithology determine the composition of
palynomorphs present in the samples?
I hypothesize that the paleotropical site studied will show a similar floristic composition
to other low paleolatitude flora of similar age (Herngreen 1975) (Hypothesis 1).
Additionally, based on the hypothesis of early angiosperm radiation from the tropics, I
hypothesize that angiosperm pollen was more abundant and more diverse in the tropical site
studied than in North America during the Aptian-Albian interval (Hypothesis 2).
Samples for this study were taken from the Caballos Formation in the Upper Magdalena
Valley (UMV) in SW central Colombia. (Fig. 1-1). All samples come from Los Mangos field,
one of the numerous oil fields located in the UMV. This field is composed of numerous wells
for oil exploration purposes. Some of the wells have been cored for different kinds of studies
(stratigraphy, lithology, etc.). A Los Mangos 31 core was selected for this study because it has a
very complete rock core that contains most of the Caballos Formation.
The name Caballos Formation was first used by Olsson (1956) in the region of Prado
Dolores, Tolima (cited in (Barrio and Coffield 1992, Blau et al. 1992). The Caballos Formation
is 100 to 400m thick (Vergara et al. 1995) and is defined as the top of the first sands under the
marls, mudstones, and calcareous rocks of the Villeta Formation (Ramon and Fajardo 2004) and
is either lying conformably on top of the Yavi Formation, where this formation is present
(Vergara 1992) or unconformably overlapping pre-Cretaceous rocks (Corrigan 1967). The
Upper Magdalena Valley (UMV), an intra-montane basin located between the Central and
Eastern Cordilleras of Colombia (Barrio and Coffield 1992, Prossl and Vergara 1993). The
basin's sediments are Mesozoic and Cenozoic in age (Blau et al. 1992). The Caballos
Formation's age of deposition is estimated to lie within the middle Aptian middle Albian
(Beltran and Gallo 1968, Corrigan 1967, Florez and Carrillo 1994). Support for the assignment
of this age comes from dinoflagellates found at the base of the Villeta Formation (on top of the
Caballos Formation) that suggest an age of middle Albian (Prossl 1992) and ammonites studied
by Etayo (1993).
The Caballos Formation is composed mainly of sandstones (80 90%) and shales (10 -
20%) (Florez and Carrillo 1994). The sandstones of the formation are the reservoir for most of
the petroleum produced in the Upper Magdalena Valley (Blau et al. 1992). Informally the
Formation has been divided into three lithologic sequences (Beltran and Gallo 1968, Corrigan
1967, Florez and Carrillo 1994). The three sequences are Lower, Middle and Upper Caballos,
which are not always present in the different localities of the basin, generating stratigraphic
confusion among authors (Vergara et al. 1995). The Lower Caballos sequence is mainly sandy,
being interpreted as floodplains (Ramon and Fajardo 2004) and littoral environments (Vergara et
al. 1995). The Midlle Caballos sequence is composed of intercalations of shale and sand, being
predominantly muddy. The different sediments and structures found in this sequence are
interpreted as fluvial channels, floodplains, coastal floodplain, low energy bay, and distal bay
deposits (Ramon and Fajardo 2004). The Upper Caballos sequence is predominantly sandy with
thin muddy intercalations being interpreted as estuarine deposits. The Mangos field marine
deposits are found just before the inundation that gave rise to the calcareous and mudstone rocks
of the Villeta Formation (Ramon and Fajardo 2004). Some authors have named each of these
three sequences as individual Formations: Alpujarra, El Ocal and Caballos, respectively (Florez
and Carrillo 1994, Vergara et al. 1995). However, because this nomenclature has been highly
contested (Vergara et al. 1995), I do not use it in this paper.
F_____ BOGOTA OuK-2 0e
ECUAODOS e VALLEY
0 S Km
Fig. 1-1. Upper Magdalena Valley (UVM) in Colombia showing the geographic location of Los
MATERIALS AND METHODS
The total extension of the rock core that contains the Caballos Formation in the Los
Mangos 31 well (-75 32; 27.89" W and 2 37' 14.56"N) is approximately 145.7 meters, but it
presents some missing intervals. A total of 33 samples were taken from this core using a 3 to 4.5
meter interval when possible. The stratigraphic column of the core analyzed with the sample
locations is shown in Fig. 2-1. Some samples did not fit exactly in this interval because of
missing rock intervals. In those cases, a sample was taken from the closest depth available and
from this point the interval was recalculated. Additionally, some samples were taken between
intervals when layers with potentially rich organic content were located. Each sample contained
approximately 30 to 50 grams of rock. The samples were taken with a geological hammer and
stored in a plastic bag that was labeled with the name of the core and the respective depth. Five
additional samples from Los Mangos 7 (75 32' 47.93" W and 2 37' 11. 39"N) and one from
Los Mangos 4 (750 32' 33" W and 2 37' 29"N) were used to complete the column were the
gaps were very long or samples were barren (location and lithology of those samples are shown
in Fig. 2-1). Since these two cores were very close to Los Mangos 31 core (< 2 km, see Fig. 1-
1.) I simply calculated their equivalent depth in Los Mangos 31 core when extracting their
samples. The Los Mangos 31 core is stored in the Yaguara Field (Neiva, Colombia) and is
property of Petrobras. The Los Mangos 4 and 7 cores are stored at the Colombian Institute of
Petroleum (Bucaramanga, Colombia), and are property of Ecopetrol. The total number of
samples taken from the three cores was 39.
The samples were prepared in the Geological Samples Preparation Laboratory at the
Colombian Institute of Petroleum (Bucaramanga, Colombia). The preparation of the samples
followed standard palynological preparation techniques described in (Traverse 1988). First, 10
grams of each sample were macerated with a mortar and pestle with the resulting powder put in
250ml beakers and the remaining portion of rock stored. Hydrochloric acid (HC1) at 10% was
mixed with each sample for a period of 90 minutes to eliminate carbonates. Samples were then
washed and left in water for 10 minutes to remove the acid. After discarding the water, samples
were transferred to hydrofluoric acid (HF) at 52% for 12 hours to eliminate silicates. Samples
were then concentrated using a centrifuge for 10 minutes and washed with distilled water. The
centrifugation and washing of the samples was repeated 3 times. Next, the samples were
thoroughly mixed with a saturated solution of Zinc chloride (ZnCl2) and centrifuged for 60
minutes to separate the organic matter through density gradient. After a new centrifugation, the
organic suspended portion (upper dark layer) was transferred to a test tube and washed with
water four times. The samples were then centrifuged again at 3500rpm, washed several times
through a 10um sieve to eliminate debris and centrifuged for another 10 minutes. After sieving
the resulting material, a first mount for each sample was made on glass slides. The remaining
material was centrifuged again and Nitric acid (HNO3) added to oxidize the material. Potassium
hydroxide (KOH) at 5% was added to remove the humic acids and samples were centrifuged
again at 3500rpm for 6 minutes. After a final wash with distilled water, the samples were sieved
and an oxidized mount placed on the same slide of the previous mount for a given sample. The
permanent mounts of the samples were made using PVC and Canadian balsam. All resulting
slides were labeled with well name, depth at which the sample was taken and a unique sequential
Three hundred palynomorphs were counted per slide when possible. This number of
palynomorphs was selected because it allows for a good statistical approximation of the real
proportion of species in a population (Hayek and Buzas 1997). The oxidized mount of each
sample was always the first to be scanned, and in the case that it did not contain 300
palynomorphs, the non-oxidized mount was scanned. After reaching a count of 300
palynomorphs, counting stopped and the remaining of both mounts were fully scanned to register
all species present in the sample but not recovered in the 300 count. The location of morphotypes
and other palynomorphs (e.g. well-preserved, rare, or first encounters of a form in each sample)
were registered using an England Finder for 2/3 of the slides and with X/Y coordinates for the
remaining 1/3. The analysis of the samples was made with a Nikon E200 and a Nikon Eclipse
600 light microscopes. All the counts were included in an excel datasheet that is presented in
Appendix A. Each morphotype was photographed, described and drafted. Photographs were
taken using an oil 63X magnification objective for most morphotypes and dry 40X magnification
objective for a few large palynomorphs with a digital Axiocam Zeiss camera integrated to an
Axiophot Zeiss microscope. The identification of palynomorphs was made by comparison with
descriptions and photographs from published studies of similar ages (Brenner 1963, 1974,
Brenner and Bickoff 1992, de Lima 1979, 1980, Doyle et al. 1982, Herngreen 1973, 1974, 1975,
Jansonius et al. 2002, Jardine and Magloire 1965, Kemp 1970, Pocock 1962, Schrank 1987,
2002, Schrank and Ibrahim 1995, Srivastava 1975). When possible palynomorphs were
identified to species level. Species not identified in the literature were named using the genus
followed by consecutive species numbers (e.g. Verrutriletes spl, Verrutriletes sp2, etc). For
palynomorphs whose identification was not possible beyond the genus level, the particle ssp.
(unknown species) was added following the genus name (e.g., Verrutriletes spp.). All slides were
deposited in the Paleobotany Collection of the Florida Museum of Natural History, in
Gainesville Florida, USA.
All samples with counts of 200 individuals or more were used for the analyses made in this
study. Samples were rarefied to 200 individuals to make them comparable. The rarefaction was
performed using the software Past (Hammer et al. 2001). The result of the rarefaction is the total
number of species that would be present in a 200-count sample. Rarefaction per group was made
possible by assuming that the relative abundance of the three palynomorph groups observed in a
sample (angiosperm pollen, gymnosperm pollen and spores) is kept constant as counts per
sample change. This is a reasonable assumption given that the chance of a grain belonging to any
of the three palynomorph groups studied is independent of the palynomorph found before or after
any given grain. More specifically, observations show that grains belonging to the same
palynomorph group are not spatially clustered or autocorrelated. Given the assumption above, I
rarefied the richness of each group based in the expected abundance of that palynomorph group
at 200 individuals.
Distribution and Variance Test
A Shapiro-Wilk test was performed to determine whether or not the distribution of the
data deviated significantly form normality. Additionally the variances were checked to
determine their homogeneity. Both tests were made using JMP software (SAS Institute 1998).
When distribution of data was non-normal, median was used instead of media, because using
averages may not give an accurate idea of the central tendency of the data. There are several
comparisons made and described in the following steps. For those comparisons using normally
distributed data and homogeneous variances, a standard student-t test was used and for non-
normal data and/or with not-homogeneous variances a Kruskal-Wallis test was used. This test is
a non-parametric equivalent to the student-t test. All comparisons were made using an alpha
value of 0.05 to determine statistical significance, and performed in JMP software (SAS Institute
The absolute abundance of palynomorphs was calculated as the total number of
palynomorphs in each group (angiosperm pollen, gymnosperm pollen and spores) present in a
sample. After determining if data was or not normally distributed and if their variances were
homogeneous, comparisons of the abundance of palynomorphs between both portions of the core
were made for each group studied (angiosperm pollen, gymnosperm pollen and spores) using the
proper statistical test as mentioned above.
Absolute species richness was calculated for these samples. The absolute species richness
is simply the total number of species of each palynomorph group found in the samples. After
determining if the distribution of data was normal or not, and if their variances were
homogeneous, comparisons of the numbers of species between both portions of the core were
made for each group of palynomorphs using the proper statistical test as mentioned above.
Cluster Analysis: Grouping Samples with Similar Composition
A cluster analysis was made for grouping the samples depending on similarities in
species abundance and distribution. A hierarchical agglomerative clustering with Wards linkage
method and Euclidean distance measure was used, as recommended by McCune and Grace
(2002). Wards's clustering is recognized as a very effective method that gives distinct clusters
and has been used and recommended to interpret biostratigraphical data (Kovach and Batten
1994). The results are presented as dendrograms scaled by the distance between merged groups,
in this case, the distance between sample compositions. Before making this analysis samples
(depths) with less than 5 individuals were excluded. Then an outlier analyses based on the depth
was performed to remove data more than 2 standard deviations from the mean of the distribution.
Finally, to improve distance calculations, a Beal smoothing transformation of the data was made
to reduce the skewness in the distribution of the data that is common in count data (McCune and
Grace 2002). The interpretation of the obtained dendrogram showing the relationship of the
samples was made by the incorporation of labels in front of each sample. First, it was evaluated
comparing the lithology of each cluster, and then the depositional environments of each sample.
The cluster analysis was made using the software PCORD (MCCune and Mefford 1999).
Relationship between Species Distribution and Lithology
Multi-response Permutation Procedure (MRPP) analyses were made to determine the
effect of lithology on species distribution. MRPR is a nonparametric method to distinguish
differences among two or more groups (Mielke and Berry 2001). This method was used instead
of others similar multivariate methods like MANOVA, because the later is not appropriate to use
with data exhibiting nonlinear relationships and extremely skewed frequencies, both
characteristics of community data (McCune and Grace 2002). For this analysis the samples were
grouped based in 6 lithological groups: Sandstone (14 samples), Siltstone (8 samples), Sandy
siltstone (3 samples), shale (6 samples), Wackstone (5 samples) and Packstone (3 samples). Prior
data screening included removal of samples with less than 5 individuals and species with
distribution more than 2 standard deviations from the mean. The procedure involved the
calculation of a distance matrix among all samples (using Euclidean distance) for the calculation
of average within-lithology distance. The average distance indicates the degree of compositional
similarity of the samples within a lithology. The greater the average distance values the greater
the differences in composition of the samples within a lithology and vice versa (McCune and
Grace 2002). Lastly, the probability of a smaller weighted mean within-group distance was
calculated using a statistical approach equivalent to multiple random reassignments of samples to
groups respecting sample size differences. MRPP output includes a p-value and a change-
correlated within-group agreement (A). The p-value is the probability of observing equal or
greater similarities within lithological groups merely due to chance. The A value is a measure of
within-group homogeneity compared with random expectation. In this case, for identical species
distribution for samples within a lithology A = 1. When heterogeneity within lithologies equals
the expectation by chance, then A = 0. Lastly, if there is less agreement within lithologies than
expected by chance, then A < 0. This analysis was performed using the software PCORD
(MCCune and Mefford 1999).
Comparison between Paleotropical and North American Samples
After determining the normal or non- normal distribution of all the data sets used in this
comparison (description above), a comparative test was made to evaluate differences between
the floral composition of the paleotropical site analyzed with middle and high paleolatitudes. For
those comparisons using normally distributed data and with homogeneous variances, a standard
ANOVA test was used. For ANOVA comparisons that resulted in significant differences among
groups a Tukey test was performed to determine which groups were statistically different from
each other. For non-normal data and/or populations with non-homogeneous variances, a Kruskal-
Wallis non-parametric test was performed. For the Kruskal-Wallis comparisons that resulted in
significant differences among groups, a Tukey test was performed on the ranks of the data to
determine which groups were statistically different from each other (Zar 1999). The quantitative
palynological data for middle and higher paleolatitudes was derived from a North American
dataset compiled by (Lupia et al. 1999). To make comparisons valid, only samples which age
ranges from middle Albian to middle Aptian were selected from this data set. This selection
resulted in a total of 67 samples for the abundance comparisons (all with a minimum of 100
individuals counted) and 297 samples for the species richness comparisons (all with a minimum
of 10 species recorded). The abundance and species richness data sets were divided into middle
(below 420N) and high (above 420N) paleolatitudes. Due to differences in sampling effort
between the datasets, comparison tests were performed for relative abundances and relative
richness only, using an alpha value of 0.05 to determine statistical significance. I performed all
the tests mentioned in this section using JMP statistical software (SAS Institute 1998).
c f LM4
Fig. 2-1. Lithological column of Los Mangos 31 core and sample locations. Red lines indicate
samples taken from the wells Los Mangos 4 and Los Mangos 7 used to complete the
section. Their respective lithology is shown in front of each red line. Ammonite
zonation and age were taken from (Etayo 1993) and lithology from (Ecopetrol ICP
Of the thirty-nine samples analyzed eighteen were barren or presented counts lower than
200 individuals those were excluded from the following analyses. All samples were rarefied to
200 individuals. A table with all species and counts per sample is annexed on Appendix A. All
parameters presented in this chapter were defined in the materials and methods section.
Spores were the dominant palynomorph in nearly all samples analyzed with 122 of the
palynomorphs in average per sample, followed by gymnosperm pollen with an average of 65
pollen grains per sample and angiosperm pollen with a median of 5 pollen grains (Fig. 3-1).
When pollen and spore abundance are examined, there are two distinguishable portions of the
core. The lower portion goes from the base to the sample LM 2625' and the upper portion goes
from the sample LM 2885' to the top.
Since the distribution of angiosperm pollen abundance was non-normal, a non-
parametrical statistical analysis was used for this variable. The amount of angiosperm pollen was
relatively constant with a median of 4 angiosperm pollen grains for the entire lower part of the
sequence. An increase in the number of angiosperm pollen grains is clearly observable in the
upper portion of the core, presenting a median of 19 angiosperm pollen grains for these samples
(Fig. 3-1). The Kruskal-Wallis test showed that there are significant differences in the number of
angiosperm pollen between both portions of the core (p < 0.05).
Since the distributions of abundance data for gymnosperm pollen and spores were
normal, I used parametric statistical analyses for them. In the case of the gymnosperm pollen, I
observed an average of 65 pollen grains per sample for the complete core. There was not a
significant difference between the average number of gymnosperm pollen grains between the
lower and the upper portion of the core (68 vs 53 grains in average per slide, respectively; t-test:
p < 0.37). Spores presented an average of 122 grains per sample through the core. Spores did not
show a significant difference in abundance between upper and lower portions of the core (120 vs.
129 grains in average per slide, respectively; t-test: p > 0.66) (Fig. 3-1).
In total 113 species of palynomorphs were found, with an average of 23 species per
sample (Fig. 3-2). The number of species per systematic group was: 36 angiosperm species, 26
gymnosperm species and 51 spore species. From this total, 20 species where singletons and their
distribution into the three palynomorphs groups was 7 singletons for angiosperm, 5 singletons for
gymnosperms and 8 singletons for spore species. Considering all three palynomorph groups,
there was an average of 21 species found per sample. However, there is a significant increase in
the number of species in the upper portion of the core, having a median of 18 species in the
lower portion and 28 sp in the upper portion of the core (Krukal-Wallis test: p > 0.011) (Fig. 3-
Non-parametric statistical analyses were used for the species richness of angiosperm
pollen since data exhibited a non-normal distribution. The median number of angiosperm species
found for the whole core was 3. However, there is a significant increase in the number of species
by the upper portion of the sequence (Fig. 3-2). The median number of angiosperm species for
the initial portion of the sequence is 2 while the median number of angiosperm species for the
upper section of the sequence is 9 (Kruskal-Wallis test: p < 0.0021). This comparison was made
including two samples that presented an abnormally high number of angiosperms in the lower
portion of the core. These samples were LM 2887' and LM 2749'7". Both samples are different
from all the other samples in the core, because they have an elevated number of angiosperm
pollen. The first sample (LM 2887') had a total of 74 angiosperm pollen grains out of a total of
300 pollen and spores counted (this value is extracted from the original data, before the
rarefaction to 200). Two species contributed to most of that abundance: Retipollenites sp.2 with
54 grains and Retimonocolpites cf. mawhoubensis Schrank with 17 individuals. In the case of
the second sample (LM 2749'7"), the abundance of angiosperms is even higher, with a total of
133 angiosperm pollen grains out of 300 (before rarefaction). In this slide most of those grains
belonged to the same species: Pennipollis cf reticulatus (Brenner) Friis, Pedersen & Crane with
108 individuals. The other angiosperm species present in that sample were: Scabramonocolpites
sp2 with 16 individuals and Asteropollis spl and Retimonocolpites sp. 4 with 4 and 1 individuals,
respectively (the remaining amount are non identified angiosperm pollen grains). The most
abundant angiosperm pollen species in the core was Pennipollisperoreticulatus with a total of
124 individuals, but most of them (108) are present in a single sample (LM 2749'7").
The data of number of species of gymnosperm pollen showed a normal distribution, and
then parametric statistical analyses were used. The percentage of gymnosperm species was 7 in
average per slide through the total length of the core. The number of gymnosperm species
remained constant in the lower and upper portions of the core (7 and 7, respectively; t-test: p <
0.67). (Fig.3-2). Some samples presented peaks in the number of gymnosperm pollen. For
example, the depth LM 2932'8" has a majority of gymnosperm pollen, with 215 pollen grains
out of 290 individuals present in the sample. On the other hand, some samples had an abnormal
low number of gymnosperm pollen, such as sample LM 2868'7" with just 4 gymnosperm pollen
grains out of 272 individuals and sample LM 3496'5" with 4 gymnosperm pollen grains out of
119 individuals present in the sample. The only sample with a single species of gymnosperm
pollen was LM 2868' 7", containing only Inaperturopollenites sp.2. The most abundant
gymnospermous species was Callialasporites dampieri (Balme) Dev, being present in almost
every sample along the core (Appendix A).
Spores presented a non-normal distribution in the lower portion of the core, then a non-
parametrical test was performed for the comparison. There core had a median of 20 spore
species in average per sample. There is a significant increase in the number of species when
comparing the lower and upper portions of the core. The lower portion has a median of 18
species and the upper a median of 28 species (Kruskal- Wallis test: p < 0.042) (Fig. 3-2).
Quantitatively, the most important spore species were C),Qthi/,ire, minor Couper and C),Qthi/,ire,
australis Couper. It was found in most of the samples and was the most abundant species in the
study (in average 14% and 31% respectively of the total number of palynomorphs found in the
Hierarchical Cluster Analyses: Sample Associations
The resulting dendrogram shows groups of samples that have similarities on species
distribution and abundance (Fig 3-3). The dendrogram shows 6 clusters: A, B, C, D and E.
However, not all clusters are meaningful when interpreting them in lithological terms, because
most of the clusters grouped different lithologies in the same cluster. Cluster A presents only
samples with sandstone and siltstone lithology, being the best-defined cluster with medium grain
composition. Cluster B and cluster D show dominance of sandy lithology, but there are also
shales and lime rocks in both. Cluster C is dominated by lime rocks (Wackstone and Packstone).
The remaining cluster (E) is composed mainly by fine particle sediment shaless), but there are
also sandy and lime components.
Analyzing the dendrogram in terms of depositional environments seems to be more
convenient because the association patterns are clearer (Fig.3-4). There are two main groups of
the dendrogram showing continental and marine composition. The clusters A and B compose
the continental group and the clusters C, D, E and F are all coming from marine environments.
The cluster A is dominated by samples coming from floodplains. Cluster B is dominated by
samples coming from channels. Middle shore-face environments, with one continental sample
present, dominate the cluster C. Cluster D is clearly dominated by samples coming from
offshore environments. Cluster F has samples from offshore and shore-face environments.
Multi-response Permutation Procedure (MRPP): Species Distribution and Lithology
This analysis shows that species composition in the samples is not strongly dependent on
lithology. The analysis result ofp = 0.15 indicate that there is a 15% chance I could find a better
grouping of the samples (greater within group similarity) by chance compared to the one made
based on lithology. The A value resulted equal to 0.0339, which means that the samples do not
have an identical species distribution within a lithology (that would be A = 1). This also means
that the heterogeneity within lithologies does not equals the expectation by chance (that would be
A = 0). Additional graphs were made as an attempt to illustrate the results of this analysis (Figs.
3-5 and 3-6), given that the MRPP does not give a graphic result. What is shown on the graphs
is that the relative amounts and number of species of each palynomorph have a similar pattern
independent of the lithology analyzed. In terms of the relative number of individuals and relative
number of species there was a constant trend in all the different lithologies: spores are dominant
with the highest values, followed by gymnosperms and finally by angiosperms with the lowest
values. The only exception to this trend was that relative number of species of angiosperm pollen
was higher than the gymnosperms species in shales.
Comparison of Abundance and Number of Species between the Paleotropical Site Studied
with Middle and High Paleolatitude Sites of North America
After checking if data were or not distributed normally, and if their variances were
homogenous, the comparisons of relative abundance and species richness between the three
different paleolatitudes (low, middle and high) were made using an ANOVA for the spore
richness, which showed normal distribution and homogeneous variances. For the abundance and
species richness in angiosperms and gymnosperms, and for the abundance of spores, which
showed a non-normal distribution, a Kruskal-Wallis test was performed.
The comparison of the relative abundances of angiosperms pollen between low, mid and
high paleolatitudes revealed that there are significant differences within the three paleolatitudes
(Kruskal-Wallis: p< 0.0004). The median relative abundances of angiosperm pollen for low, mid
high paleolatitudes were 4, 2 and 0, respectively. The Tukey test performed on the ranks of the
data showed that there are not significant differences between the relative abundance of
angiosperm pollen between low and mid paleolatitudes (overlapping circles), but that there are
significant differences between these two latitudes (low and mid) and high paleolatitudes (no
overlapping circles). There is a latitudinal gradient of decrease in the relative abundance of
angiosperms as latitude increases (Fig. 3-7)
For the relative abundance of gymnosperm pollen, the comparison made showed that
there are significant differences between the three paleolatitudes (Kruskal-Wallis: p< 0.0001).
The median relative abundances of gymnosperm pollen for low, mid high paleolatitudes were 26,
55, and 97, respectively. The Tukey test showed that there are significant differences between
the relative abundance of angiosperm pollen between the three paleolatitudes compared (no
overlapping circles). In this case, there is a clear latitudinal gradient of increase in the relative
abundance of gymnosperms as latitude increase (Fig. 3-8).
The comparison performed for the relative abundance of spores showed significant
differences between the three different paleolatitudes compared (Kruskal-Wallis: p< 0.0001).
The median relative abundances of spores for low, mid high paleolatitudes were 62, 34 and 3,
respectively. The Tukey test performed on the ranks revealed significant differences in the
relative abundance of spores between the three paleolatitudes. For spores the latitudinal gradient
is opposite to the one found for gymnosperms, in this case the relative abundance of spores
decreases as the latitude increases (Fig. 3-9).
The comparison of the relative richness of angiosperm pollen showed that there are
significant differences between low, mid and high paleolatitudes (Kruskal-Wallis: p < 0.0001).
The median relative richness of angiosperm pollen for low, mid high paleolatitudes were 14, 10
and 0, respectively. The Tukey test performed revealed that there are not significant differences
in the relative richness of angiosperm pollen between low and mid paleolatitudes (overlapping
circles), but that there are significant differences between these two (low and mid, with high
paleolatitudes (no overlapping circles). The latitudinal gradient is similar to the one exhibited by
the relative abundance, there is a decrease in the relative richness of angiosperm pollen as the
latitude increases (Fig. 3-10).
The comparison of the relative richness of gymnosperm pollen between low, mid and
high paleolatitudes showed that there are significant differences between the three paleolatitudes
(Kruskal-Wallis: p< 0.0019). The median relative richness of gymnosperm pollen for low, mid
high paleolatitudes were 31, 33, and 39, respectively. The Tukey test showed that there are not
significant differences between the relative abundance of gymnosperm pollen between low and
mid paleolatitudes (overlapping circles), but that there are significant differences between these
two paleolatitudes (low and mid) and high paleolatitudes (no overlapping circles). There is a
latitudinal gradient of increase in the relative richness of gymnosperms as latitude increase
The comparison performed for the relative richness of spores showed significant differences
between the three paleolatitudes (Kruskal-Wallis: p< 0.0001). The median relative richness of
spores for low, mid high paleolatitudes were 51, 53, and 60, respectively. The Tukey test showed
that there are not significant differences between the relative abundance of spores between low
and mid paleolatitudes (overlapping circles), but that there are significant differences between
these two paleolatitudes (low and mid) and high paleolatitudes (no overlapping circles). There is
a latitudinal gradient of increase in the relative richness of gymnosperms as latitude increase
LM 2541'9" --
LM 2562' _
M 25710'i -- -
LM 2625 I .-
LM 26 6' ........... .
M 2 6 1 .......... ........... .... ........ .....
LM 295' 11 -
LM 2720 I
LM 2867' 5" I
LM 2932'8'" I ____._
0 60 120 2 60 1a 0 200
IN) 12 2
Number of Individuals
Fig. 3-1. Absolute abundance of angiosperm pollen, gymnosperm pollen and spores represented
as the total number of individuals found in each one of the samples (all samples rarefied
to 200 counts).
LM 2,571' 10i
LM 2 85'
M 2625' |
LM 2696' *
L 2749'7" -
LM 28679'5 -
LM2g6g'7 5 -
0 10 20 0 10 20 0 10 20
Number of species
Fig. 3-2. Absolute richness of angiosperm pollen, gymnosperm pollen and spores represented as
the total number of species found in each one of the samples (all samples rarefied to
IE-02 I.7E+00 3.4E+00 5. 1 E+00 6.9E+00
Information Remaining (%)
100 75 50 25 0
: LM 2956 2 _
LM 293261 __
A LM 27420
B~L 26 2868.
LC LM 2686
LM 2677 -
DLM 26746 1 "-
E LM 2695 I I
D LM 269585 1
LM 2571 10
LM 254S I"
LM 2677 '
Fig. 3-3. Dendrogram showing the different lithological associations based upon their species
composition for the analyzed core. There are 5 clusters of samples that showed a
similar species composition. The lithological convention is the same used for the
methods figure (Fig.2-1).
methods figure (Fig.2-1).
I E-2 1.7E+00 3.4E+00 5. I E -' 6.9E+00
Information Remaining (%)
100 75 50 25 0
CCb LM 2956 '2-
CFp LM 29 15"8
CFp LM 2SS7'
CCb LM 2868'7" _
B CCh LM 2867'5
nCCb LM 27498'7"
CCb LM 2769'
C msf LM 2696'1..5-"--....
MIlf LM 2695"1. 1 -
MTn LM 274S'
D Mn LM 2720"
Ms"f LM 26907' 11
MmSf LM 2625
MOf LM 265 1.0"
MOf LM 2599"
MOf LM 2677 -
MOf LM 2674 "1 "
MOf LM 2642'1.
Mm.Sf LM 2607 -
MOf LM 255'
MItS LM 2571"'0 -
MI]f LM 2487"
MmF M LM 2541'9"-
MOf LM 2529
MISf LM 24965"' 5
MOf LM 2562' 2
MOf LM 25010'10G
Fig. 3-4. Dendrogram showing the different associations of depositional environments based
upon their species associations for the analyzed core. There are 6 clusters of samples
that showed a similar species composition. The particle preceding the sample ID
represents the depositional environment. The first letter of that particle represents
whether is C- continental or M-marine, the second part the specific environment: Ch-
channels, Fp-floodplains, Sf-Shore face, 1Sf-lower shoreface, mSf- middle shoreface,
In-inter-tidal and Of-offshore.
50 0 Gymnosperms
300 T I
Packstone (3) Sandstone (14) Sandy siltstone Shale (6) Siltstone (8) Wackstone (5)
Fig.3-5. Relative abundances of palynomorphs found in each kind of lithology. The number in
parenthesis indicates the number of samples.
10 I Angiosperms
8 E T M Spores
Packstone (3) Sandstone (14) Sandy siltstone Shale (6) Siltstone (8) Wackstone (5)
Fig.3-6. Relative species richness of palynomorphs found in each lithology. The number in
parenthesis indicates the number of samples.
Fig. 3-7. Comparison of the relative abundances of angiosperm pollen for the Aptian-Albian
interval between low (site studied), mid and high paleolatitudes.
Fig. 3-8. Comparison of the relative abundances of gymnosperm pollen for the Aptian-Albian
interval between low, mid and high paleolatitudes.
Fig. 3-9. Comparison of the relative abundance of spores for the
low, mid and high paleolatitudes.
Aptian-Albian interval between
Fig. 3-10. Comparison of the relative species richness of angiosperms for the Aptian-Albian
interval between low, mid and high paleolatitudes.
Fig. 3-11. Comparison of the relative species richness of gymnosperms for the Aptian-Albian
interval between low, mid and high paleolatitudes.
Fig. 3-12. Comparison of the relative species richness of spores for the Aptian-Albian interval
between low, mid and high paleolatitudes
Floral Composition of the Tropical Site Analyzed
The main purpose of this study was to determine the floral composition of a paleotropical
ecosystem through the analyses of quantitative palynological data. I had hypothesized that the
paleotropical site studied contained a similar floristic composition to other low paleolatitude
flora of similar age (Hypothesis 1).
Because floral composition is derived from species abundance and richness, the
hypothesis above was analyzed in both terms. The results of this study demonstrate that the
floristic abundance and richness of the Colombian low paleolatitude ecosystem analyzed are
different from the abundance and richness observed at a low latitude Brazilian site of similar age.
In general, the abundance pattern from the Brazilian site during the Albian was: gymnosperms
dominant with more than half of the individuals in average per sample, followed by angiosperms
and finally spores had the lowest abundance (Herngreen 1975). In contrast with the abundance
pattern above, the results of this study show spores as the most abundant palynomorph in nearly
all the samples, with an abundance of 61% in average per slide, while gymnosperms were a
secondary component of the flora with an average of 32% of the individuals per sample (Fig. 3-
1). Angiosperms were the minor component in the site studied with abundances less than 7% in
average per slide.
The relative richness patterns were also distinct between both low paleolatitude
ecosystems. The Brazilian site presented a higher relative number of gymnosperm species (more
than 50% in average per slide) and a similar relative number of species of angiosperms and
spores. In this study I found a higher number of species of spores (11% in average per sample),
followed by gymnosperms (7 species in average) and angiosperms had the lowest number of
species (4 in average).
In conclusion, these comparisons do not support the hypothesis that the floristic
composition of both low paleolatitude floras (Colombian and Brazilian) was similar. The
differences in the composition could be derived from essential differences in the flora of east and
west South America. It has been hypothesized that the African South American province in
the Lower Cretaceous had very dry conditions (Herngreen et al. 1996); which may have favored
gymnosperm dominance exhibited at the Brazilian site. Supporting this reasoning, the high
amounts of Classopollis classoides and gnetalean grains -pollen grains previously related to dry
environments- are present in high abundances at the Brazilian site. The hypothetical dry
conditions of east South America may also explain the low amount of spores in the samples,
because ferns are more abundant in humid environments. On the other hand, the high amounts
of spores found in the western South American site studied suggests that during the Lower
Cretaceous this area presented a humid environment. However, the lack of quantitative studies
in west South America limit the support this hypothesis.
Interesting patterns were found when analyzing the distribution of palynomorphs in Los
Mangos core. There were significant compositional differences between lower and the upper
portions of the core. Firstly, angiosperm pollen was significantly more abundant and with higher
number of species in the upper portion of the core than in the lower portion. The abundance
comparisons between the upper and lower portions of the core were made including two unique
samples that presented an abnormally high numbers of angiosperms in the lower portion of the
core. Yet, despite the inclusion of these outliers, comparisons showed statistically significant
higher abundance of angiosperm pollen in the upper portion of the core (Fig.3-1). Additionally,
the comparison of relative angiosperm richness between lower and upper portions of the core
showed the same pattern present for the abundance data with significantly more angiosperm
species in the upper portion of the core than in the lower portion (Fig.3-2). An increasing
diversity of forms and ornamentation can be observed in the upper portion of the sequence (mid
Albian), which could be the result of the increasing speciation of angiosperms during the Albian.
These results agree with the initial increase in angiosperm diversity that was taking place during
the Barremian to Albian interval (Heimhofer et al. 2005). This first increment was followed by
an even more dramatic increase of angiosperm species that took place during the Albian -
Cenomanian interval (Crane and Lidgard 1989, 1990, Lidgard and Crane 1990, Lidgard and
Crane 1988) or Albian Turonian interval in middle paleolatitudes (Lupia et al. 1999). The
samples analyzed in this study ranges from the mid Aptian to the mid Albian, which means that
the increase in angiosperm species demonstrates that the pattern of diversification found in other
quantitative studies worldwide also took place in the neopaleotropical flora analyzed here.
The first pollen grains that appear in the core are monosulcate grains identified as:
Pennipollis perireticulatus, Brenneripollis sp4 and Clavatipollenites hughesii. Some
inaperturate grains were also found at the bottom of the sequence, identified as Afropollis
jardinus andRetipollenites sp3. The first tricolpate species was found pretty early in the core
(approximately mid Aptian), which is congruent with the age of the occurrence of tricolpate
grains at low paleolatitudes in Israel (Brenner 1996). This tricolpate grain was psilate with
simple colpi and was identified as Psilatricolpites spl. The lower part of the core is
approximately middle to upper Aptian in age. More ornamented forms of tricolpate grains
appeared in the upper portion of the core, approximately in the lower to middle Albian. Two
tricolpate species, Rousea cf micupullis andR. cf georgensis, are relegated to the upper part of
the core (Appendix A).
Gymnosperms did not show differences in their abundance or species richness when
comparing the lower and upper portions of the core (Figs. 3-1 and 3-2). Spores did not show any
difference when comparing their abundance. However, when comparing species richness spores
showed a similar pattern to angiosperms, with significantly higher number of species in the
upper portion of the core than in the lower one (Fig. 3-2).
Summarizing all these comparisons, angiosperm pollen increased in number of
individuals and number of species through time, while spores increased in number of species and
gymnosperms did not show any significant change. There are at least three possible explanations
for these observed differences between the upper and lower portions of the core: lithological
differences between both portions of the core, differences in the depositional environments or
changes in the floristic composition of the ecosystem.
Lithology could potentially explain differences in composition between upper and lower
portions of the core given that, at first glance, the core presents more shales in the upper portion
than in the lower portion. Because sedimentary rocks derive their composition and texture from
source rock material and environment under which where it is deposited (Boggs 2006) and these
characteristics also define the kind of fossil preserved in these rocks, it was expected to find
significant differences in the composition of the different lithologies. To determine if lithology
was a factor related to observed differences between upper and lower portions of the core two
analyses were made: hierarchical cluster analysis (Fig. 3-3) and Multi Response Permutation
Procedure (MRPP). However, cluster analysis results show that resulting sample clusters based
on species composition are not related to lithology. Furthermore, the MRPP analysis showed a
weak relation between lithology and the composition of samples for this study (p = 0.14). The
related graphs (Figs. 3-5 and 3-6) exhibit the general relationships between floral composition
and lithologies and show that there was a constant trend in species richness and abundance for all
the different lithologies: spores were first with the highest values, followed by gymnosperms and
with angiosperms last (with the exception of the relative number of species of angiosperm pollen
being higher than that of gymnosperms in shales). However, these graphs (Figs.- 3-5 and 3-6)
show that although the trends in composition are similar between lithologies, the relative
numbers of individuals and species are variable for each lithology. These small differences
could explain the poor relationship between sample composition and lithology captured by the
MRPP analysis. In conclusion, both analyses (MRPP and hierarchical cluster analysis)
demonstrated that the lithology was not strongly influencing the samples composition.
The second possible reason for the observed differences between the upper and the lower
portions of the core is that the depositional environment of both portions was different. The kind
of palynomorphs found in a sample depends on the flora of the place that is finally represented in
the sediments by pollen grains and spores. In the case of this study samples were either from
marine or continental environments. Marine samples represent the flora of a wider area because
the sediments and palynomorphs are carried out by channels through a wide variety of floras to
be finally deposited in the sea (Srivastava 1994). Continental sediments represent a more local
flora, because the sediments are not transported from other environments. To determine if facies
determined the differences observed between lower and upper portions of the core a samples
were clustered based on species composition similarities and results interpreted using the
depositional environments of the samples. The results show the samples were clearly divided
into two clear groups: samples from marine environments and samples from continental
environments. At first look this seems to explain the observed differences, but is important to
note that all continental samples are located at the bottom of the core and marine samples are
distributed between the lower and upper portions of the core. The change in facies occurred in
the lower portion of the core, while the increase in the abundance and species richness of
palynomorphs occurred only in the upper portion of the core. This means that the increase on
angiosperm abundance and richness, and spore richness occurred in a similar depositional
environment (marine). Even removing the continental samples from the core the differences
between the upper and lower portions of the core are significant. Thus, based on this cluster
analysis it is concluded that the depositional environment did not determine the increase of
angiosperm richness and abundance and spore richness observed in the upper portion of the core.
The third possible explanation is that the increase in angiosperm abundance and richness,
and spore richness is reflecting the diversification that these groups experienced starting in the
Lower Cretaceous. Reported results suggest that in the paleotropical latitudes, angiosperms were
becoming a more prominent taxonomic and ecological component of the ecosystems by the mid
Albian with the increase in the number of species and abundance. Before the diversification of
angiosperms during the Cretaceous (Coiffard et al. 2006, Crane et al. 1995, Crane and Lidgard
1989, Lupia et al. 1999), gymnosperms, ferns and Bennettitales dominated terrestrial ecosystems
worldwide (Lupia et al. 1999). Angiosperms started their diversification during the Barremian -
Albian interval and by the Upper Cretaceous were the dominant component of paleotropical
vegetation and an important component of the middle and upper paleolatitude floras (Crane and
Lidgard 1989, Lidgard and Crane 1990, Lidgard and Crane 1988, Lupia et al. 1999).
Quantitative palynological studies in middle and high paleolatitudes have shown the same
pattern of continued increase in angiosperm diversity and abundance between the Aptian and
beginning of the Campanian (Lidgard and Crane 1990, Lupia et al. 1999). However, those
studies show a decline of spore species for that period of time. Spore richness in the low latitude
ecosystem studied increased concurrently with angiosperm richness and abundance, suggesting
that ferns were diversifying with angiosperms as found in other studies (Schneider et al. 2004).
In conclusion, the results show initial patterns of angiosperm and ferns diversification in
the low latitude site analyzed during the Albian, with the significant increase on their number of
species, which means they were becoming a more prominent taxonomic component of tropical
ecosystems in the Lower Cretaceous.
Differences in Floristic Composition between the Paleotropical Site Analyzed and
Based on the widespread hypothesis of angiosperm origin and radiation from lower
paleolatitudes during the Lower Cretaceous, I hypothesized that the abundance and number of
angiosperm species should be higher in the low latitude paleotropical site compared to medium
and high paleolatitudes during the Aptian-Albian interval (Hypothesis 2). The results partially
support the formulated hypothesis by showing that angiosperms had significantly higher relative
abundance and species richness in the paleotropical site analyzed compared to data from high
latitudes of North America. However, the results show that there is no significant difference
between the angiosperm abundance and richness between low and mid paleolatitudes (Figs. 3-7
and 3-10). The results show a latitudinal gradient in both abundance and richness data, with
higher values in low and mid latitudes and with significantly lower values in high latitudes.
Gymnosperm abundance was significantly lower in the low latitude paleotropical site
analyzed than in high and mid latitudes (Fig. 3-8). The species richness of gymnosperms was
significantly higher in high paleolatitudes compared with low and mid latitudes, which did not
present significant differences. These results were expected because gymnosperms were a very
important component of high latitude floras during the Lower Cretaceous (Crane and Lidgard
1989, 1990, Lupia et al. 1999) as they are in modem floras, and only a minor component in
Abundance of spores showed a similar pattern to angiosperm abundance with an
increasing latitudinal gradient from high to low latitudes. This pattern was expected because
spores were a minor component of high latitude ecosystems (Lupia et al. 1999, Crane and
Lidgard 1989) while being the dominant component in low latitudes floras as shown in this
The tropical origin of angiosperms with subsequent radiation to higher latitudes was first
proposed by (Axelrod 1959) and since then numerous studies have attempted to determine the
patterns followed by angiosperms in their radiation during the Cretaceous (e.g. Crane and
Lidgard 1989, 1990, Hickey and Doyle 1977, Retallack and Dilcher 1981). However,
quantitative studies are more numerous in middle and high latitudes than they are in the lower
paleolatitudes. Based on the results of this study, angiosperms were a more conspicuous
component of low and mid paleolatitude ecosystems during the Aptian- Albian interval than they
were in high latitudes. These findings partially support the widespread hypothesis that
angiosperms appeared in low paleolatitudes and later on time radiated to higher latitudes (Crane
and Lidgard 1989, 1990, Hickey and Doyle 1977, Retallack and Dilcher 1981). However,
although there is a clear gradient in the data from high to low latitudes, there was no significant
difference in angiosperm abundance and richness between mid and low paleolatitudes. Based on
the results of this study is not possible to determine if angiosperms were present in low latitude
prior to mid latitudes, or vice versa. These small and non- statistically significant differences
may be related to the hypothesized expansion of the tropics during global warming periods
(Jaramillo et al. 2006) and the need for greater amount of data from tropical paleolatitudes. The
world's temperature was higher during the Lower Cretaceous (Bice et al. 2006), making mid
paleolatitudes warmer and making possible that tropical plants could live in those warmer
paleolatitudes. As consequence of those more uniform temperatures between them, low and mid
paleolatitudes presented very similar floristic composition. This scenario could explain why I did
not find significant differences in the floristic composition of low and mid palelatitudes. High
latitudes in other hand, do not present temperatures as warm as tropical areas even with higher
worldwide temperatures, then tropical taxa cannot live there and vice versa, making the floral
composition very different, as it is in modern floras and as it was expected.
A possible source of error in the analyses made is that the comparison made was between
one low latitude paleotropical site versus numerous sites in mid and high paleolatitudes. These
differences in sampling effort could be introducing bias in the comparison, making necessary the
use of more low latitude paleotropical sites to obtain a more reliable comparison of the
paleofloras in different latitudes and thus a better determination of the patterns followed by
angiosperms and other taxa during the radiation of flowering plants.
Difference in the preservation of palynomorphs is another possible factor that could have
affected the comparisons made between the floras of low, mid and high paleolatitudes. For this
study we used samples from a single site, which after deposition were exposed through time to
similar environmental conditions. Those conditions could have been responsible in many cases
for the damage or alteration of the characteristics of sensitive species, which will imply their
elimination from the record or their damage to the point of not being identified as different
species (e.g. loss of perine in spores). On the other hand, the study from North America has a set
of samples from a wide range of geographical sites, each one with different preservation
conditions. Due to this, there is a higher probability that the sensitive species that could have
been damaged or altered from some of the sites, are still present in the record of other sites.
Finally, underestimation of the number of species has been always a concern when
working with light microscopy (Lidgard and Crane 1990). The use of a scanning electron
microscope (SEM) leads to the identification of a greater number of characters that often lead to
the assignment of more species. All the samples in this study and most of the samples in the
dataset from higher paleolatitudes were analyzed only with light microscopy, which decrease the
bias of having different methods determining the morphological characters of the palynomorphs
and therefore the species richness.
To summarize, I found that the floristic composition of the Aptian- Albian flora of a
paleoequatorial site is significantly different from high paleolatitudes, but similar to mid
paleolatitudes' composition. I found that angiosperms were more abundant and diverse in the
paleotropical ecosystem analyzed, gymnosperms were more abundant and diverse in higher
paleolatitudes and spores were more abundant in the paleotropical ecosystem analyzed but more
diverse in the paleotemperate region. It is recommended to include more paleotropical samples
to enlarge the dataset and support or reject the patterns found in this study.
SPECIES COUNTS PER SAMPLE
LM 2487' 2 1___
LM2496'5" 2 3 2 27 2 3 4
LM2510'10" 7 4 1 75 1 2 11 4
LM 2529' 1 4 13 3 4
LM2541'9" 2 3 40 2 3 14 5 2
LM 2562' 3 18 4 4 59 1 8 2
LM2571'10" 1 1 8 2 14 1
LM 2585' 2 3 15 1 1 8
LM2599'4" 4 1 5
LM 2607' 6 1 7
LM 2625' 2 1 8 1
LM 2642' 11" 2 27 10 1
LM 2658'10" 2 117 30
LM2674' 1" 4 11 35 11
LM 2677' 11 3 9 1
LM 2686' 15 4 63 6
LM 2690'11" 4 1 80 4
LM 2695' 11" 2 1 45 10
LM 2696' 1.5" 1 1 10
LM 2720' 2 13 3
LM 2735' 1 2 30 1
LM 2748' 1 4 33
LM 2749'7" 4 1
LM 2789' 2 1
LM2836'3" Barren interval
LM 2867'5" 2
LM 2868'7" 1 1
LM 2887' 16
LM 2915'8" 1 1 7 1
LM 2932'8" 1 4 1 4 1
LM 2956'2" 2 1 5 ____18 1___
LM 2965' _____________
LM 2496' 5" 1 2
LM 2510' 10" 1 3
LM 2541'9" 2 1 1
LM 2562' 2
LM 2571' 10" 1
LM 2585' 1 1 1 1 1
LM 2599' 4"
LM 2607' 2
LM 2625' 58 1
LM 2642' 11" 1 52 1
LM 2658' 10" 2
LM 2674' 1" 27 2
LM 2677' 12 1
LM 2686' 2 10
LM 2690' 11" 15 10
LM 2695' 11" 70 23
LM 2696' 1.5" 1 12 12
LM 2720' 20
LM 2735' 1 12 1 1
LM 2748' 1 8
LM 2749'7" 3
LM 2769' 2
LM 2789' 1
LM2836'3" Barren interval
LM 2867'5" 1 3
LM 2868'7" 11
LM 2887' 1 81 21
LM 2915'8" 5 5 1 17 11
LM 2932'8" 1 3 1 5 11 136
LM 2956'2" 2 1 2___ 2_ 9
I I I I I I II I I I II
LM 2496' 5" 1
LM 2510' 10" 1 1
LM 2529' 2 4
LM 2541' 9" 1 1
LM 2562' 1 1 4
LM 2571' 10" 1 1
LM 2585' 3_ 2 1 2
LM 2599' 4" 1
LM 2607' 1
LM 2625' 1
LM 2642' 11" 1 1
LM 2658' 10" 1
LM 2674' 1" 1 4 1
LM 2677' 4 1 3 1
LM 2686' 1 1 1 1
LM 2690' 11" 1 1
LM 2695' 11" 1 3
LM 2696' 1.5" 1
LM 2720' 1 5 1 1 3
LM 2735' 1
LM 2749'7" 1 5 3 1
LM 2789' 3 1
LM2836'3" Barren interval
LM 2867'5" 1 2 1 3
LM 2887' 1
LM 2915'8" 1 2 1 2 4 2
LM 2932'8" 1 1 1 2 5 3
LM 2956'2" 1 1 5 4
I II II II I IIII I I
LM 2487' 1
LM 2496' 5" 1 1 2
LM 2510'10" 1 1 1 47
LM 2529' 1 1 3 14
LM2541'9" 2 1 1 34
LM 2562' 4 17
LM 2571' 10" 1 1 28
LM 2585' 1 1 1 1 9
LM 2599' 4" 1
LM 2607' 4
LM 2625' 43
LM 2642' 11" 2 15
LM 2658' 10" 1 34
LM 2674' 1" 1 11
LM 2677' 2 5 1 24
LM 2686' 7 28
LM 2690' 11" 14
LM 2695' 11" 2 1 4 17 21
LM 2696' 1.5" 1 2 20
LM 2720' 1 9 16
LM 2735' 1 1 5 4
LM 2748' 2 14 2
LM 2749'7" 1 2 57 44
LM2836'3" Barren interval
LM 2867'5" 3 4 2 4
LM 2868'7" 4
LM 2887' 4 1 1 4
LM 2915'8" 1 1 1 5 3 16
LM 2932'8" 3 3 4 6 1 43
LM 2956'2" 1 1 5
LM 2965' 1
I I III I II IIIIIIII
'3 e "
LM 2496' 5"
LM 2510'10" 2 1 1
LM 2529' 3
LM2541'9" 5 2 1
LM 2562' 1 1 1
LM 2571' 10" 7 1 1
LM 2585' 1 1 1 1 3
LM 2599'4" 6
LM 2607' 1 1
LM 2625' 2 4
LM 2642' 11" 1 5
LM 2658' 10" 28
LM 2674' 1"
LM 2677' 6
LM 2686' 9 1 3
LM 2690' 11" 29
LM 2695' 11" 20 1 1
LM 2696' 1.5" 3 1
LM 2720' 1 2 12
LM 2735' 5 4 12
LM 2748' 8 5
LM 2749'7" 109 1 1
LM2836'3" Barren interval
LM 2867'5" 2 1 6 12
LM 2868'7" 1 45
LM 2887' 5 1 12 3
LM 2915'8" 8 3 3 2 2
LM 2932'8" 3 4 6
LM 2956'2" 1 1 _5 1
I I I I I I I I I I I I
LM 2487' 8 3
LM 2496'5" 37 17
LM2510'10" 44 48 1 1 1 1 8
LM 2529' 33 30 1 6
LM2541'9" 86 96 1 2 11 1
LM2562' 54 84 2 1 5 1 1 8 5 4
LM 2571' 10" 15 149 1 6
LM 2585' 16 220 1 2 6
LM 2599'4" 32
LM 2607' 9 31
LM 2625' 3 129
LM 2642' 11" 145
LM2658' 10" 6 115
LM2674' 1" 1 15 70
LM 2677' 27 190
LM 2686' 43 113 1
LM2690' 11" 20 20___
LM2695' 11" 78 62
LM 2696' 1.5" 39 10
LM 2703' 1
LM 2720' 78 115
LM2735' 2 6 80 107 1
LM 2748' 3 52 163 1
LM 2749'7" 28 11 3 1
LM 2769' 10
LM 2789' 2
LM 2806' 1
LM2836'3" Barren interval
LM 2867'5" 198 40
LM 2868'7" 52 145
LM 2887' 30 107 23 1 74
LM2915'8" 55 36 1 1 3
LM 2932'8" 8 6 1
LM 2951'9" 1 1
LM 2956'2" 45 2 1____
I I I. I I
LM 2496'5" 1 2 3
LM 2510'10" 3 2 1 1 1 1
LM 2529' 2 2 1
LM2541'9" 1 1 1 6 8 1
LM 2562' 1 2 1 1
LM 2571' 10" 1 1
LM 2585' 1 1 2 2 2 1 1
LM 2599' 4" 2
LM 2607' 1
LM 2625' 1 5
LM 2642' 11" 2 1 4
LM 2658' 10" 12
LM 2674' 1" 1 1
LM 2677' 1 1 2
LM 2686' 3
LM 2690' 11" 2 1
LM 2695' 11" 1
LM 2696' 1.5"
LM 2720' 2
LM 2735' 18
LM 2748' 1
LM2836'3" Barren interval
LM 2867'5" 1
LM 2887' 1 1 8
LM 2915'8" 1 1 5
LM 2932'8" 1 1 2 1
LM 2956'2" 1 1
IIIIIIII I I I III II
LM 2487' 1
LM 2496'5" 2 1 1
LM 2510'10" 6 8
LM 2529' 7 2 3 1 6
LM2541'9" 14 18
LM 2562' 8 2 1 3
LM 2571' 10" 14 2 13
LM 2585' 1 1 1 11
LM 2599' 4" 1 1
LM 2607' 1 1 3
LM 2625' 7
LM 2642' 11" 1 18
LM 2658' 10" 1 1 12
LM 2674' 1" 2 8
LM 2677' 2 1 12
LM 2686' 27
LM 2690' 11" 2 16
LM 2695' 11" 32
LM 2696' 1.5" 1 1 1
LM 2720' 1 2 4
LM 2735' 1 8
LM 2748' 1
LM 2749'7" 16 1
LM 2789' 5
LM2836'3" Barren interval
LM 2867'5" 5 2
LM 2868'7" 1 1 10
LM 2887' 1 10
LM 2915'8" 2 5
LM 2932'8" 1 1
LM 2951'9" 3
LM 2956'2" 2 1 3 9_
LM 2965' 1______
LM 2487' 1 3 6 1
LM 2496'5" 4 2 3 8 1
LM 2510'10" 3 3 1
LM 2529' 4 6 1 1 5 4
LM2541'9" 5 11 5 37 4 1
LM2562' 7 13_ 5 1 2 48 13
LM 2571'10" 7 24 2 9
LM2585' 1 5 18 1 1 20 6
LM 2599'4" 1 2 17 5
LM 2607' 1 10 1
LM 2625' 4 2 1 18 2
LM 2642' 11" 5 3 7 4
LM2658'10" 10 15 3 115 28
LM 2674' 1" 7 5 29 1
LM2677' 2 11 23 2
LM2686' 9 13 1 1 161 7
LM 2690' 11" 3 5 245 4
LM 2695' 11" 3 9 _142 10
LM 2696' 1.5" 1 3 1 63 9 1
LM 2703' 3
LM 2720' 3 5 1 1
LM 2735' 9 10 3 11
LM 2748' 1 4 1 1
LM 2749'7" 3 3 1 1
LM 2769' 1
LM 2789' 3 1- 2
LM 2806' _
LM2836'3" Barren interval
LM 2867'5" 1 4 2
LM 2868'7" 1 2
LM 2887' 13
LM 2915'8" 37 4 6 7 1 3 1 7 6 12
LM 2932'8" 2 1 1 5 2 1 2 2 8
LM 2956'2" 5 1 1 1 1 2 1 9
LM 2965' 2___ 2
1. Afropollis cf jardinus (Brenner) Doyle and Doerenkamp; LM31-2562'; England Finder
2. Afropollis sp. 1; LM31 2562'; E.F: X54
3. Pennipollis cf. reticulatus (Brenner 1963) Friss, Pedersen & Crane; LM31-2956'2"; E.F:
W22/2 (Label on left)
4. Liliacidites sp. 1; LM7-2 766'; E.F: Y21/3
5. Psilatricolpites sp. 1; A131-2585 '; E.F: P60
6. Psilatricolporites sp. ; LM31-2585'; E.F: N52/4
7. Retimonocolpites sp.2; LM31-2585 '; E.F: P60
8. Retimonocolpites sp. 3; LM31-2562'; E.F: N69/2
9. Clavatipollenites cf hughesii Couper; LM31-2585'; E.F: L59
* England finder coordinates (E.F) were found having the label of the slide on the right side,
otherwise it is indicated as: label on left. Scale = 10tm.
10. Retimonocolpites spp; LM7 2766'; E.F: L62/3
11. Retimonocolpites cf. mawhoubensis Schrank; LM31 2887'; E.F: U70/4
12. Brenneripollis sp. 1; LM7 2750'; E.F: W28
13. Brenneripollis sp. 2; LM 2585' E.F: V61/4
14. Brenneripollis sp.3; LM31 2562'; E.F: G64/2
15. Retipollenites sp. 3; LM7 2750'; E.F: D23/2
16. Retipollenites sp. 4; LM31 2541 '9"; E.F: M33
17. Retimonocolpites sp. 4; LM31 2585'; E.F: H39/1
18. Retimonoporites sp. 1; LM31 2585 '; E.F: E58
19. Tricolpites sp. 1; LM31 2677'4"; E.F: D57
20. Phimopollenites cf pannosus (Dettmann & Palyford) Dettmann; LM31 2585'; E.F: F49/2
21. Rousea cf georgensis (Brenner) Dettmann; LM31 2510'10"; E.F: S56/1
22. Phimopollenites cf pseudocheros Srivastava; LM31 2642'1"; E.F: T66/3
23. Rousea cf miculipollis Srivastava; 3LM31 2529'; E.F: L38
24. Retitricolporites sp.1; LM31 2571'10"; E.F: J50/3
25. Schrankipollis microreticulatus (Brenner) Doyle et al.; LM4 3160'; E.F: Y57/2
26. Stellatopollis doylei Ibrahim; LM7 2766'; E.F: H56
27. Araucariacidites australis Cookson; LM7 2766'; E.F: X58
28. Callialasporites dampieri (Balme) Dev; LM31 2496'5"; E.F: J54/4
29. Callialasporites trilobatus (Balme) Dev; LM31 2658'10"; E.F: K58/1
30. Classopollis cf classoides (Pflug) Pocock y Jansonius; LM31 2696' 1.5"; E.F: R21/2
31. Classopollis cf intrareticulatus Volkheimer; LM31 2696' 1.5"; E.F: L58/1
32. Cycadopites sp. ; LM31 2562'; M62/1
33. Ephedripites cf barghoornii Pocock; LM 31 2487'5"; E.F: M54/4
34. Ephedripites irregularis Hemgreen; LM31 2496'5"; E.F: M57/1
35. Equisetosporites cf leptomatus de Lima; LM7 2766'; E.F: R65
36. Equisetoporites cf dudarensis (Deak) de Lima; LM7 2766'; E.F: V57/4
37. Ephedripites cf multicostatus Brenner; LM31 2956'2" (Label on left); Q18/3
38. Eucommiidites sp2; LM7 2766' E.F: G57/3
39. Equisetosporites cf ambiguus Hedlund; LM7 2750'; E.F: 056/2
40. Equisetosporites sp.2; LM31 2956'2"(Label on left); E.F: L12/1
41. Equisetosporites cf fragilis de Lima; LM31 2529"'; E.F: E60/2
42. Equisetosporites sp.1; LM31 2965'; E.F: V24/3
43. Equisetosporites cf albertensis (Singh 1964) de Lima, LM31 2887'; E.F: R21
44. Eucommiidites spl; LM31 2585'; E.F: L55/2
45. Inaperturopollenites sp2; LM7 2766'; E.F: Y65
46. Inaperturopollenites spl; LM31 2686; E.F: V60
47. Gnetaceapollenites spl; LM31 2677'; E.F: 064
48. Steevesipollenites spl; LM31 2887'; E.F: U64/3
49. Baculotriletes sp.1; LM31 2562'; E.F: S62/1
50. Gabonisporites sp.2; LM7 2750'; E.F: W61/4
51. Baculotriletes sp.2; LM7 2766'; E.F: P52/2
52. Gabonisporites sp.3; LM31 2571'10"; E.F: L54/4
53. Chomotriletes cf almegrensis Pocock; LM31 2541'9"; E.F: W50/4
54. Echimonoletes "sphericus" Informal ICP; LM7 2766'; E.F: W54
55. cf. Cicatricosisporites subrotundus Brenner; LM31 2956'2" (Label on left); E.F: U14/4
56. cf. Appendicisporites dentimarginatus Brenner; LM31 2956'2" (Label on left); E.F: U25
57. Cicatricosisporites dorogensis Potonie and Gelletich; LM31 2585'; E.F: F51
58. Appendicisporites cf erdtmanii Pocock; LM31 2585'; L40/4
59. Appendicidisporites cf jansonii Pocock; LM7 2750'; E.F: N51
60. Cicatricosisporites cf hughesii Dettmann; LM7 2750; E.F: R30/4
61. cf. Exesipollenites tumulus Balme; LM7 2766'; E.F: Y69/4
62. Concavissimisporites variverrucatus (Couper) Brenner; LM31 2887'; W60
63. Baculatisporites comaumensis (Cookson) Potonie; LM7 2766'; E.F: K30/2
64. Echitriletes sp.1; LM 2750'; E.F: D23
65. cf. Muerrigerisporites coronispinalis Srivastava; LM7 2750'; E.F: U66
66. Echinatisporis varispinosus (Pocock) S.K. Srivastava; LM7 2750'; E.F: U29/1
67. Echitriletes sp.3; LM31 2690'11"; E.F: H60
68. Microfoveolatosporis skottsbergii (Selling) Srivastava; LM31 2642' 11"; E.F: J40
69. C)yuthidliit minor Couper; LM31 2496'5"; E.F: T63/1
70. Gleicheniidites cf senonicus Ross; LM31 2562'; E.F: M70/3
71. Impardecispora trioreticulosa (Cookson and Dettmann) Venkatachala, Kar & Raza; LM7
2750'; E.F: U57/4
72. Zlivisporis sp.1; LM7 2766'; E.F: S66
73. Matonisporites sp. 1; LM31 2956'2"; E.F: N23
74. Psilamonoletes sp.1; LM31 2541'9"; E.F: P63/3
75. Klukisporitesfoveolatus Pocock; LM7 2766'; E.F: L62/4
76. Microreticulatisporites spl; LM31 2562'; E.F: 052/3
77. Foveotriletes sp.1; LM7 2766'; E.F: J64/3
78. Camarazonosporites cf insignis Norris; LM31 2529'; E.F: Y63/3
79. Rugutriletes sp.2; LM31 2585'; E.F: X20/2
80. Cicatricosisporites hallei/venustus; LM4 3160'; E.F: W64
81. Cicatricosisporites sinuosus Hunt; LM7 2766'; E.F: 026
82. Verrumonoletes sp.1; LM7 2766'; E.F: F52/1
83. cf Verrucosisporites rotundus Singh; LM7 2766'; E.F: U69/1
84. Converrucosisporites spl; LM31 2585'; E.F: U57/2
85. Converrucosisporites sp2; LM31 2956'2"; E.F: T17
86. Leptolepidites sp.1; LM 31 C21/4
87. Verrutriletes sp.1; LM7 2750'; E.F: S67/1
88. Verrutriletes sp.4; LM7 2766'; E.F: W21/2
89. Perotriletes spl; LM7 2766'; E.F: X64/4
LIST OF REFERENCES
Axelrod, D. I. 1959. Poleward Migration of Early Angiosperm Flora. Science 130(3369):203-
Barrio, C., and D. Coffield. 1992. Late Cretaceous stratigraphy of the Upper Magdalena Basin in
Payande-Chaparral segment (western Girardot Sub-Basin), Colombia. Journal of South
American Earth Sciences 5(2):123-139.
Beltran, N., and J. Gallo. 1968. The geology of the Neiva Sub-Basin Upper Magdalena Basin,
southern portion. Pp. 253-275. Ninth Ann. Field Conf. Col. Soc. Petrol Geol. & Geoph.
Bice, K. L., D. Birgel, P. A. Meyers, K. A. Dahl, K. U. Hinrichs, and R. D. Norris. 2006. A
multiple proxy and model study of Cretaceous upper ocean temperatures and atmospheric
CO2 concentrations. Paleoceanography 21(2).
Blau, J., L. Vergara, and H. Stock. 1992. First planktonic foraminifera from the Early Cretaceous
(Albian) of the Upper Magdalena Valley, Colombia. Journal of South American Earth
Boggs, S. 2006. Principles of sedimentology and stratigraphy. Pearson, Prentice Hall.
Brenner, G. 1963. The spores and pollen of the Potomac Group of Maryland. Department of
Geology, ed., Baltimore, Maryland.
Brenner, G. 1968. Middle Cretaceous spores and pollen from Northeastern Peru. Pollen et Spores
Brenner, G. 1974. Palynostratigraphy of the Lower Cretaceous Gevar'am and Talme Yafe
Formations in the Gever'am 2 well (Souther coastal plain Israel). Geological Survey of
Israel Bull. 59:1-27.
Brenner, G. 1976. Middle Cretaceous floral provinces and early migrations of angiosperms. In C.
Beck, ed. Origin and early evolution of angiosperms. Columbia University Press, New
Brenner, G. 1996. Evidence for the earliest stage of angiosperm pollen evolution: A
paleoequatorial section from Israel. Chapman & Hall.
Brenner, G., and I. Bickoff. 1992. Palynology and age of the Lower Cretaceous basal Kumub
Group from the coastal plain to the Northern Negev of Israel. Palynology 16:137-185.
Burnham, R., and K. Johnson. 2004. South American palaeobotany and the origins of neotropical
rainforests. Philosophical Transactions of the Royal Society of London Series B-
Biological Sciences 359(1450):1595-1610.
Coiffard, C., B. Gomez, J. Kvacek, and F. Thevenard. 2006. Early angiosperm ecology: evidence
from the Albian Cenomanian of Europe. Annals of Botany 98:495-502.
Corrigan, H. 1967. The geology of the Upper Magdalena Basin (Northern portion). Pp. 221-251.
Eigth Field Conf. Col. Soc. Petrol. Geol. & Geoph., Bogota, Colombia.
Crane, P. 1987. Vegetational consequences of the angiosperm diversification. Pp. 107-144. In E.
M. Friis, W. G. Chaloner, P. R. Crane., ed. The origin of angiosperms and their biological
consequences. Cambridge University.
Crane, P., E. Friis, and K. Pedersen. 1995. The origin and early diversification of angiosperms.
Crane, P., and S. Lidgard. 1989. Angiosperm diversification and paleolatitudinal gradients in
Cretaceous floristic diversity. Science 246(4930):675-678.
Crane, P., and S. Lidgard. 1990. Angiosperm radiation and patterns of Cretaceous palynological
diversity. Pp. 377-407. In P. D. Taylor, and G. P. Larwood, eds. Major evolutionary
radiations. Oxford University.
de Lima, R. 1978. Palinologia da Formacao Santana (Cretaceo do Nordeste do Brasil).
Introducao geologica e descricao sistematica dos esporos da subturma azonotriletes.
Ameghiniana Revista de la Asociacion Paleontologica Argentina XV(3-4):333-365.
de Lima, R. 1979. Palinologia da Formacao Santana (Cretaceo do Nordeste do Brasil). II.
Descricao sistematica dos esporos da subturma zonotriletes e turma monoletes, e dos
polen das turmas saccites e aletes. Ameghiniana Revista de la Asociacion Paleontologica
de Lima, R. 1980. Palinologia da Formacao Santana (Cretaceo do Nordeste do Brasil). III.
Descripcao sistematica dos polens da turma plicates (subturma costates). Ameghiniana
Revista de la Asociacion Paleontologica Argentina XVII(1): 15-47.
de Lima, R. 1987. Estudo palinologico da sondagem estratigrafica da Lagoa do Forno, Baic do
Rio Peixe, Cretaceo do Nordeste do Brasil. Bol. IG-USP, Ser. Cient. 18:67-83.
de Lima, R. 1989. Palinologia da Formacao Santana (Cretaceo do Nordeste do Brasil). IV.
Descricao sistematica dos polens das turmas plicates e poroses, esporos, incertae sedis e
microplancton marinho. Ameghiniana Revista de la Asociacion Paleontologica Argentina
Dino, R., d. S. O., and D. Abrahao. 1999. Palynological and stratigraphic characterization of the
Cretaceous strata from the Alter do Chao Formation, Amazonas Basin. Boletim do 5
simposio sobre o Cretaceo do Brasil:557-585.
Doyle, J., S. Jardine, and A. Doerenkamp. 1982. Afropollis, a new genus of early angiosper
pollen, with notes on the Cretaceous palynostratigraphy and paleoenvironments of
Northern Gondwana. BCREDP 6:39-117.
Ecopetrol ICP. 2000. Registro de description sedimentologica y estratigrafica: Los Mangos 4,
Los Mangos 7 y Los Mangos 31. Stratigraphic and lithologic column. Ecopetrol,
Etayo, F. 1993. A modo de historic geologica del Cretacico en el Valle Superior del Magdalena.
In F. Etayo, ed. Estudios geologicos del Valle Superior del Magdalena. Universidad
Florez, J. M., and G. A. Carrillo. 1994. Estratigrafia de la sucesi6n litol6gica basal del Cretacico
del Valle Superior del Magdalena. Pp. 1-25. In F. Etayo, ed. Estudios geol6gicos del
Valle Superior del Magdalena. Universidad Nacional, Bogota.
Friis, E., W. Chaloner, and C. P. 1987. Introduction to angiosperms. Pp. 1-15. In E. M. Friis,
W. G. Chaloner, P. R. Crane., ed. The origin of angiosperms and their biological
consequences. Cambridge University.
Friis, E., K. Pedersen, and P. Crane. 2006. Cretaceous anglosperin flowers: Innovation and
evolution in plant reproduction. Palaeogeography Palaeoclimatology Palaeoecology
Hammer, O., D. Harper, and P. Ryan. 2001. PAST: Paleontological Statistics Software Package
for Education and Data Analyses. Palaeontologia Electronica 4(1):9pp.
Hayek, L., and M. Buzas. 1997. Surveying Natural Populations. Columbia Univeristy Press, New
Heimhofer, U., P. Hochuli, S. Burla, J. Dinis, and H. Weissert. 2005. Timing of Early Cretaceous
angiosperm diversification and possible links to major paleoenvironmental change.
Herngreen, G. 1973. Palynology of the Albian-Cenomanian strata of Borehole 1-QS-1-MA, State
of Maranhao, Brasil. Pollen et Spores XV(3-4):515-555.
Herngreen, G. 1974. Middle Cretaceous palynomorphs from the Northeastern Brazil. Sci. Geol.
Herngreen, G. 1975. Palynology of the Middle and Upper Cretaceous strata in Brazil.
Medelingen Rijks Geologische Dienst, Nieuwe Serie 26(3):39-91.
Herngreen, G., M. Kedves, L. Rovnina, and S. Smirnova. 1996. Cretaceous palynofloral
provinces: a review. Pp. 1157-1188. In J. Jansonius, and D. McGregor, eds. Palynology:
Principles and applications. American Association of Stratigraphic Palynologists
Hickey, L. J., and J. A. Doyle. 1977. Early Cretaceous Fossil Evidence for Angiosperm
Evolution. Botanical Review 43(1):3-104.
Ibrahim, M. 1996. Aptian-Turonian palynology of the Ghazalat-1 Well (GTX-1), Qattara
Depression, Egypt. Review of Palaeobotany and Palynology 94:137-168.
Jansonius, J., L. Hills, and Hartkopf-Froder. 2002. Genera file of fossil spores. Dept. of Geology
and Geophysics, University of Calgary, Calgary.
Jaramillo, C., M. Rueda, and G. Mora. 2006. Cenozoic plant diversity in the neotropics. Science
Jardine, S., and H. Magloire. 1965. Palynologuie et stratigraphie du Cretace des Bassind du
Senegal et Cote d'Ivoire. Memoires du Bureau de Recherches Geologiques et Minieres
Kemp, E. 1970. Aptian and Albian miospores from Souther England. Palaeontographica 131(1-
Kovach, W. 1993. Multivariate techniques for biostratrigraphical correlation. Journal of the
Geological Society 150:397-705.
Kovach, W., and D. Batten. 1994. Association of palynomorphs and palynodebris with
depositional environments: quantitative approaches. In A. Traverse, ed. Sedimentation of
organic matter. Cambridge University Press.
Lidgard, S., and P. Crane. 1990. Angiosperm Diversification and Cretaceous Floristic Trends a
Comparison of Palynofloras and Leaf Macrofloras. Paleobiology 16(1):77-93.
Lidgard, S., and P. R. Crane. 1988. Quantitative-Analyses of the Early Angiosperm Radiation.
Lupia, R. 1999. Discordant morphological disparity and taxonomic diversity during the
Cretaceous angiosperm radiation: Notrh American pollen record. Paleobiology 25(1):1-
Lupia, R., P. Crane, and S. Lidgard. 2000. Angiosperm diversification and Cretaceous
environmental change. In S. Culver, and P. Rawson, eds. Biotic response to global
change The last 145 million years. University Press, Cambridge.
Lupia, R., S. Lidgard, and P. Crane. 1999. Comparing palynologycal abundance and diversity:
implications for biotic replacement during the Cretaceous angiosperm radiation.
Magurran, A. 2003. Measuring biological diversity. Blacwell Publising.
McCune, B., and J. Grace. 2002. Analysis of ecological communities. MjM Sofware Design,
MCCune, B., and M. Mefford. 1999. Multivariate Analisys of Ecological Data, Version 4.01.
Mielke, P. W., and K. J. Berry. 2001. Permutation methods: a distance approach. Springer.
Muller, H. 1966. Palynological investigations of the Cretaceous sediments in Northeastern
Brazil. Pp. 123-136. 2nd W. African Micropal. Coll., Ibadan.
Pocock, S. 1962. Microfloral analysis and age determination of strata at the Jurassic-Cretaceous
boundary in the Western Canada plains. Palaeontographica 111(1-3):1-95.
Prossl, K. 1992. Preliminary results of palynological investigations on the Cretaceous of
Colombia, South America. Review of Palaeobotany and Palynology 71:275-268.
Prossl, K., and L. Vergara. 1993. The Yavi Formation (Lower Cretaceous), Upper Magdalena
Valley, Colombia: an integrated sedimentological and palynological study. N. Jb. Geol.
Palaont. Abh. 188(2):213-240.
Ramon, J., and A. Fajardo. 2004. Sedimentologia y estratigrafia secuencial de la Formacion
Caballos, Subcuenca de Neiva, Valle Superior del Magdalena. III Convencion Tecnica
ACGGP. Bogota, Colombia.
Regali, M., and C. Viana. 1989. Late Jurassic-Early Cretaceous in Brazilian sedimentary Basins:
Correlation with the International Standard Scale. Petroleo Brasileiro S.A. Service de
desenvolvimiento de recursos humans, Rio de Janeiro.
Retallack, G., and D. Dilcher. 1981. Arguments for a Glossopterid Ancestry of Angiosperms.
SAS Institute. 1998. JMP Statistics and graphic guide, Version 5.1. SAS Institute.
Schneider, H., E. Schuettpeltz, K. Pryer, R. Cranfill, S. Magallon, and R. Lupia. 2004. Ferns
diversified in the shadow of angiosperms. Nature 428:553-557.
Schrank, E. 1987. Palaeozoic and Mesozoic palynomorphs from Northeast Africa (Egypt and
Sudan) with special reference to Lat Cretaceous pollen and dinoflagellates. Berliner
geowiss. Abh. 75(1):249-310.
Schrank, E. 1994. Nonmarine Cretaceous palynology of northern Kordofan, Sudan, with notes
on fossil Salviniales (water ferns). Geol Rundsch 83:773-786.
Schrank, E. 2002. Barremian angiosperm pollen and associated palynomorphs from the Dakhla
Oasis area, Egypt. Paleontology 45(1):33-56.
Schrank, E., and M. Ibrahim. 1995. Cretacceous (Aptian-Maastrichtian) palynology of
foraminifera-dated wells (KRM-1, AG-18) in northwestern Egypt.
Srivastava, S. 1975. Microspores from the Frederiscksburg Group (Albian) of the Southern
United States. Paleobiologie Continentale VI(2):1-119.
Srivastava, S. 1994. Evolution of Cretaceous phytoprovinces, continents and climates. Review of
Palaeobotany and Palynology 82:197-224.
Sun, G., and D. Dilcher. 2002. Early angiosperms from the Lower Cretaceous of Jixi, eastern
Heilongjiang, China. Review of Palaeobotany and Palynology 121(2):91-112.
Taylor, W. D., and L. J. Hickey. 1996. Evidence for and implications of an herbaceous origin for
angiosperms. Pp. 232-266. In W. D. a. H. Taylor, L. J., ed. Flowering plant origin,
evolution and phylogeny. Chapman & Hall.
Traverse, A. 1988. Paleopalynology. Unwin Hyman, London.
Vergara, L. 1992. Lower Cretaceous stratigraphic sequences in the Quebrada Bambuca (Aipe),
Upper Magdalena Valley, Colombia. Giessener Geologische Schriften 48:183-200.
Vergara, L., J. Guerrero, P. Patarroyo, and G. Sarmiento. 1995. Comentarios acerca de la
nomenclatura estratigrafica del Cretacico Inferior del Valle Superior del Magdalena.
Geologia Colombiana 19:21-31.
Wing, S., and L. Boucher. 1998. Ecological aspects of the Cretaceous flowering plant radiation.
Annual Review of Earth and Planetary Sciences 26:379-421.
Zar, J. H. 1999. Biostatistical analysis. Prentice Hall.
Paula Mejia was born on Dec 23, 1979 in Medellin, Colombia. The oldest of three
children, she grew up mostly in her city of origin, graduating from the INEM "Jose Felix de
Restrepo" High School in 1996. She earned her B.S. in Biology with emphasis in botany and
palynology from the Universidad de Antioquia in 2004. While in college she worked as research
assistant in the Chagas disease Lab research in 2000, in the HUA herbarium 2000 2003, and
then in 2003 she moved to Bucaramanga, Colombia to work in her undergrad thesis in
palynology. Upon graduating in May 2004 with her B.S. in biology, Paula worked as research
assistant in the Colombian Institute of Petroleum. In Aug 2004 she moved to Gainesville, Fl in
order to attend to the University of Florida to get her master's degree in science.
Upon completion of her master's program, Paula will continue in graduate school in the
University of Florida as a PhD student in the Botany Department. She recently got married to
Lucas Fortini, a PhD student in the Forestry Department.