Evolutionary Patterns in the Reef Coral Siderastrea During the Mio-Pliocene of the Dominican Republic." MS (Master of Science) Thesis, University of Iowa, 2005

University of Iowa Iowa Research Online

Theses and Dissertations

2005

Evolutionary patterns in the reef coral Siderastrea during the Mio- Pliocene of the Dominican Republic

Brian Robert Beck University of Iowa

Follow this and additional works at: https://ir.uiowa.edu/etd

Part of the Geology Commons

This thesis is available at Iowa Research Online: https://ir.uiowa.edu/etd/94

Recommended Citation Beck, Brian Robert. "Evolutionary patterns in the reef coral Siderastrea during the Mio-Pliocene of the Dominican Republic." MS (Master of Science) thesis, University of Iowa, 2005. https://doi.org/10.17077/etd.psw64cdj

Follow this and additional works at: https://ir.uiowa.edu/etd Part of the Geology Commons

EVOLUTIONARY PATTERNS IN THE REEF CORAL SIDERASTREA DURING THE MIO-PLIOCENE OF THE DOMINICAN REPUBLIC

by Brian Robert Beck

A thesis submitted in partial fulfillment of the requirements for the Master of Science degree in Geoscience in the Graduate College of The University of Iowa

December 2005

Thesis Supervisor: Professor Ann F. Budd

Graduate College The University of Iowa Iowa City, Iowa

CERTIFICATE OF APPROVAL ______

MASTER'S THESIS ______

This is to certify that the Master's thesis of

Brian Robert Beck has been approved by the Examining Committee for the thesis requirement for the Master of Science degree in Geoscience at the December 2005 graduation.

Thesis Committee: ______Ann F. Budd, Thesis Supervisor

______Jonathan M. Adrain

______Christopher A. Brochu

ACKNOWLEDGMENTS

I am grateful to the following people for their help during this research: Dr. Ann Budd for assistance in fieldwork, data analyses, and comments and suggestions on several drafts; Dr. Jonathan Adrain and Dr. Christopher Brochu for helpful comments and suggestions during the editorial process; Kay Saville for help with thin-sections. I would also like to then the University of Iowa Geoscience department and the Geoscience Littlefield Fund.

TABLE OF CONTENTS

LIST OF TABLES...... iv LIST OF FIGURES ...... v

INTRODUCTION ...... 1 Systematics...... 2 Localities...... 4

METHODS AND MATERIALS ...... 7 Sampling...... 7 Morphometrics...... 8

RESULTS ...... 11 Species recognition...... 11 Stratigraphic ranges...... 11 Morphologic change through time...... 12 Comparison of fossil and modern species...... 12 Comparison of morphometrics ...... 13

DISCUSSION ...... 14

FUTURE WORK...... 17

CONCLUSIONS ...... 19

REFERENCES ...... 86

APPENDIX A. LANDMARK DESCRIPTION ...... 81

APPENDIX B. SAMPLE INFORMATION ...... 83

iii

LIST OF TABLES

Table 1. Morphologic characters distinguishing eight Caribbean species of Siderastrea...... 73 2. A list of the localities sampled, the number of samples collected, the Formation name, the river the samples were collected from, and the age of samples collected from those localities...... 74 3. Eigenvalues displaying the percent variation covered by each of the 3 discriminant functions...... 75 4. Wilks' Lambda chart displaying the significance of each of the three discriminant functions...... 76

5. Mahalanobis distance matrix reporting the significance between groups...... 77

6. Structure matrix for the discriminant analysis displaying the significant variables for each discriminant function...... 78

7. Results of the Kruskal-Wallis test of the three known DR Morphospecies throughout the Dominican Republic formations...... 80

A1. Landmarks taken on transverse thin-sections of corallites of Siderastrea...... 82

B1. List of specimens analyzed in morphometric analyses...... 84

LIST OF FIGURES

Figure 1. An example of transverse section through a corallite of a species of Siderastrea...... 21 2. Map of the study area in the Dominican Republic with the sampled rivers color coded...... 23 3. Comparison of age-depth plots of the Río Gurabo and Río Cana sections based on existing biostratigraphic and paleomagnetic (Gurabo section only) data...... 25 4. A stratigraphic column of Rio Cana with NMB localities marked at their position in the stratigraphic column...... 27

5. A stratigraphic column of Rio Gurabo with NMB localities marked at their position in the stratigraphic column...... 29

6. Locality map for Arroyo Bellaco...... 31

7. A stratigraphic column of Rio Yaque del Norte with NMB localities marked at their position in the stratigraphic column ...... 33

8. A composite stratigraphic column for each of the sampled rivers of the Northern Dominican Republic with the rivers indicated along the top...... 35

9. Transverse sections of a colony of Siderastrea showing a close up view of the landmark scheme used in the 2-D geometric morphometric analysis...... 37

10. Transverse sections of a colony of Siderastrea showing 1/12 of the corallite used for landmark measurement in the 2-D geometric morphometric analysis...... 39

11. A photograph of a transverse section through a corallite of DR Morphospecies 1...... 41

12. A photograph of a transverse section through a corallite of DR Morphospecies 3...... 43

13. A photograph of a transverse section through a corallite of DR Morphospecies 4...... 45

14. A photograph of a transverse section through a corallite of DR Morphospecies 2...... 47

15. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR Morphospecies 1-4. Groups are labeled by DR Morphospecies...... 49

16. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR Morphospecies 1-4. Groups are labeled by formation...... 51 17. Boxplots illustrating the variation of discriminant function 1 throughout the sampled formations for DR Morphospecies 1, 3, and 4...... 53

18. Boxplots illustrating the variation of discriminant function 2 throughout the sampled formations for DR Morphospecies 1, 3, and 4...... 55

19. Boxplots illustrating the variation of variable x20 throughout the sampled formations for DR Morphospecies 1, 3, and 4...... 57

20. Boxplots illustrating the variation of variable y4 throughout the sampled formations for DR Morphospecies 1, 3, and 4...... 59

21. Boxplot illustrating the variation of number of septa throughout the 4 DR Morphospecies...... 61

22. Boxplot illustrating the variation of coral radius throughout the 4 DR Morphospecies...... 63

23. A photograph of a transverse section through a corallite of a modern Siderastrea siderea from Jamaica...... 65

24. A photograph of a transverse section through a corallite of a modern Siderastrea radians from Jamaica...... 67

25. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR Morphospecies 1–4 and modern S. siderea and modern S. radians. Groups are labeled by species groupings...... 69

26. Estimated depositional water depths for Cercado and Gurabo Formation from Nehm and Geary (1994)...... 71

vi 1

INTRODUCTION

In 1972 Eldredge and Gould argued that the current evolutionary paradigm of the time (gradualism punctuated by stratigraphic gaps) did not explain patterns observed in the fossil record. They proposed an alternative hypothesis, punctuated equilibrium, which suggests that organisms go through long periods of stasis and new species are formed during short periods of rapid evolutionary change (Eldredge and Gould, 1972; Gould, 1977). Since Eldredge and Gould’s original studies, several additional studies have shown examples of both phyletic gradualism and punctuated equilibrium (e.g., Bell et al., 1985; Cheetham, 1986; Lazarus, 1986; Stanley and Yang, 1987; Nehm and Geary, 1994). This study looks at the evolutionary patterns in Siderastrea (de Blainville, 1830) to see if the observed patterns follow phyletic gradualism or punctuated equilibrium. Studies have been done in the Neogene of the Caribbean testing these alternative modes of evolution in various organisms (bryozoans, mollusks) (e.g., Cheetham, 1986, Nehm and Geary, 1994). To date, only cursory work has been done studying the evolutionary patterns in corals (e.g., Potts et al., 1993; Budd et al., 1994). Cheetham (1986) and Nehm and Geary (1994) found punctuated equilibrium, but Nehm and Geary

found gradual change associated with speciation. The corals in my study are from the same localities in the Dominican Republic as the bryozoans and mollusks that were used in Cheetham (1986) and Nehm and Geary (1994), and because of the similar localities, I will test the hypothesis that the corals in my study followed punctuated equilibrium and display either rapid change or morphologic stasis. Cheetham (1986) and Nehm and Geary (1994) used morphometrics to distinguish species and track evolutionary change over time. I also use morphometrics in my study of Siderastrea, but instead of basic linear measurements, I use two dimensional (2-D) 2

geometric morphometrics to better capture the changes within each species. With 2-D geometric morphometrics, 2-D landmarks are digitized on the photo instead of the traditional linear landmarks. Two dimensional landmarks not only capture the traditional measurements of length, width, etc., but they also capture many other measurements not obtained by traditional morphometrics (e.g. angles, ratios, etc.). These methods are used to determine if species exhibited stasis or gradual evolutionary change over approximately 9 million years of geologic time during the Mio-Pliocene. This is the period of time just before the Isthmus of Panama closed, blocking the flow from the Atlantic to the Pacific. This closure created a turnover in the Pleistocene, and several species of corals went extinct and several new species originated. This study will help to see how the corals were evolving just before this important event in Caribbean coral history. I also examine the morphometric data to determine whether the morphologic changes were evenly scattered through time or concentrated during certain intervals. If corals changed constantly through time that would indicate phyletic gradualism, but if a period of rapid change or periods of no change at all are detected, that would support the hypothesis of punctuated equilibria.

Systematics

Order: Scleractinia (Bourne, 1900) Family: Siderastreidae (Vaughan and Wells, 1943)

Genus: Siderastrea (de Blainville, 1830) Siderastrea (Figure 1) is a zooxanthellate scleractinian coral of the family Siderastreidae, and although its modern geographic distribution is restricted primarily to the Atlantic (Veron, 1995), some species of Siderastrea are found in the Pacific and 3

Indian Oceans (Budd and Guzman, 1994). Colonies of Siderastrea are usually massive, have on average 48 septa per corallite, range widely in diameter (2.5-8 mm), vary in thickness of corallite walls, and are abundant in back reef environments. Siderastrea is cerioid (corallites are juxtaposed) and has extramural budding (new corallites bud outside the wall of the parent corallite). Along with many septa, Siderastrea also has many synapticulae and a synapticulothecal wall (Veron, 1995; Budd and Guzman, 1994) (Table 1). There are five extant species recognized in the world, two of which occur in the Caribbean (S. radians and S. siderea) (Veron, 1995), and there are four extinct species found in the Caribbean (S. conferta, S. mendenhalli, S. pliocenica, and S. silecensis) (Budd et al., 1994). Siderastrea is a common coral used in many modern studies (Vermeij, 2005; Garcia et al., 2005; etc.), but the fossil history of this coral is unknown by comparison. This study helps to further the paleontological knowledge of Siderastrea by looking at the morphological history. Understanding how this reef coral has changed through geologic time will help both paleontologic and modern studies of Siderastrea and other reef corals.

Species: Siderastrea siderea (Ellis and Solander, 1786) This species of Siderastrea forms colonies larger than most other living

species of Siderastrea. The corallites, which are also larger than most other species of Siderastrea, average 4-5 mm in diameter, but can be as small as 3-3.5 mm in diameter. There are typically four complete septal cycles. The tertiary septa are fused to the secondary septa, and the quaternary septa are seen to be fused to the tertiary septa closer to the wall than S. silecensis. The columella is not wide in comparison to the other species of Siderastrea and is not as well defined. This species has a thin synapticulothecal wall and it has high calice relief.

Species: Siderastrea silecensis (Vaughan, 1919) Siderastrea silecensis has a massive colony form with a domed upper surface. The corallites are all polygonal and the corallite wall is sometimes raised. An adult corallite is on average 5 mm in diameter, but oblong corallites can reach lengths of 7 mm and widths of 5 mm. The calice depth of most corallites of S. silecensis is 1.5 mm. This species undergoes four complete septal cycles, but in larger corallites, a few quinaries are often seen. The synapticulae are well developed and rather coarse. The columella width is average for species of Siderastrea. Its synapticulothecal wall is thin and it has medium calice relief.

Species: Siderastrea mendenhalli (Vaughan, 1917). The colonies are massive in form and have a somewhat flat upper surface. Colonies are larger in size with diameters on average of 440 mm in diameter. Corallites are found to be polygonal and sometimes deformed. The diameter of corallites range from 4 mm to 7 mm with the more deformed and elongated corallites reaching the diameters around 7 mm. Four complete septa cycles are seen in average sized corallites. The septa are thin and crowded within the corallite. The synapticulae are well developed and extend more than halfway from the wall of the corallite to the columella. The columella is small in comparison to the other species of Siderastrea. It has a thick

synapticulothecal wall and has low calice relief.

Localities The Dominican Republic is located in the Caribbean Sea and shares the island of Hispaniola with Haiti. The sampling localities for my study are well-preserved river cut sections in the Northern Dominican Republic. The river cut sections from which I collected were Rio Gurabo, Rio Cana, Arroyo Bellaco, Cañada de Zamba, and Yaque del Norte (Figure 2). The sections consist of Middle Miocene to Middle Pliocene siliciclastic 5

sediments and span a continuum of more then 9 million years. The original dating of the stratigraphic sections in the Dominican Republic was done using foraminifera and other microfossils (Saunders et al., 1986). While the current stratigraphic dates are sufficient for this study, better chronostratigraphy is needed to expand this study further. McNeil (Klaus et al., 2004) has recently revised the age dates of these formations by integrating microfossil and paleomagnetic data (Figure 3), and is currently in the process of analyzing magnetostratigraphy, biostrarigraphy, and strontium isotopes to gain an accurate grasp on the timing of paleontological and sedimentological events of the Northern Dominican Republic stratigraphy. This is important not only for accurate timing of evolutionary events, but it is also needed for reconstruction of the timing of paleoenvironmental changes. The stratigraphy of the Northern Dominican Republic was originally studied by Saunders et al. (1986). On the basis of their fossil descriptions coupled with their stratigraphic columns and sedimentological interpretations, I assembled a list of potential collecting localities. These localities occurred in four stratigraphic formations: the Baitoa Formation, Cercado Formation, Gurabo Formation, and Mao Formation (Figure 4). These four formations are comprised mainly of silts and sands with conglomeratic beds between the formations (Saunders et al., 1986).

The Baitoa Formation is composed of calcareous silts which are occasionally separated by thin, sandy conglomerates. The sections of the Baitoa Formation that are dense with corals are a sandy limestone in which the coral heads are often in living position, or dense coral patches where many corals are in living position. The Cercado Formation, which lies above the Baitoa Formation, is similar in composition. The lower part of the formation is sandy with interbedded conglomerates. The sorting of these sands are poor and cross-bedding is often seen. The upper part of the Cercado Formation is silty and thick with coral and algal debris. The Gurabo Formation is marked with a 6

conglomerate at the base, and above the conglomerate are beds of calcareous and fossiliferous silts. The upper beds of these silts are extremely rich in coral colonies, and at the uppermost part of the coral rich beds of the Gurabo Formation are interbedded biostromal corals and coralliferous silts. The Mao Formation is composed of sandy silts with some sands and conglomerates. These beds are followed by calcareous silts that have a significant amount of clay in them. Above the silts is a massive series of limestones. Corals are common in these upper limestones (Saunders et al., 1986). The corals used in my study were collected from areas in the silty layers that were packed with coral debris. The overall depositional environment of these beds is shallow marine, but the depth and exact water conditions vary. These changes in shallow marine conditions were interpreted on the basis of overall faunal changes (Saunders et al., 1986). The energy levels of the depositional environments range from high energy levels represented by the conglomeratic formation boundaries to low energy levels represented by the calcareous clays (Saunders et al., 1986). 7

METHODS AND MATERIALS

Sampling Siderastrea is a relatively uncommon coral in the Dominican Republic sequences. I searched for colonies throughout the exposed sequence and made collections at all localities where Siderastrea was found by Saunders et al. (1986). At each locality, all colonies of Siderastrea were collected unless there were more than 20. Collecting colonies of Siderastrea within the rock sequence involved removing them from the face of the outcrop and packaging them for shipping back to the University of Iowa. For each sample, I recorded data on geographic location, stratigraphic position, and the interpreted geologic age (Table 2 and reference stratigraphic columns Figure 5-8). After the samples were shipped back to Iowa I cleaned off the excess dirt and prepared one thin section per colony to a thickness of 30 µm, and mounted it 1 ½” x 3” slides. The Dominican Republic sections are continuous, which is extremely helpful in this study. With a continuous section, there are no missing time gaps in which changes could take place unseen. Table 2 shows a list of the localities that were sampled and how

many samples were taken, and it also shows the river those localities are from, the formation the sample was collected from, and the age of each locality. While the sampling was not entirely even throughout the section with a few large samples in the Gurabo Formation and the Cercado Formation, samples were collected throughout the four formations. The stratigraphic placement of each locality can be seen on figures 5-8. In addition, I also obtained supplementary collections from the Natural History museum in Basel, Switzerland which were also made through the same sections in the late 1970’s. These collections were used to increase sample size and to extend the 8

temporal range of my analysis. The samples from Basel increased the temporal range of my sample due to the fact that they were the only samples from the Baitoa Formation, and since the time they were collected, the Baitoa section has become covered in water due to the construction of a dam near Santiago.

Morphometrics I used 2-D geometric morphometric analysis to study the morphologic changes between colonies. Twenty-eight landmarks (Figure 9) were digitized using Image J (ver. 1.31, 2004) on six corallites per colony to measure change between corallites and colonies. Due to the symmetry of corallites, landmarks were only taken from a primary septa to a secondary septa (Figure 10). Landmarks were selected to capture all of the measurements that were traditionally gained by using linear measurements (e.g., corallite diameter, length of primary septa, etc.) along with other measurements not easily obtained with linear measurements (e.g. septa length rations, septa width ratios, etc.). The landmarks selected were also chosen to help capture details about distance between septa, aspects of the synapticulae, and septal angles. Traditionally species of Siderastrea have been distinguished mainly on number of septa, corallite diameter, and thickness of corallite wall (Table 1). After species were established with 2-D geometric morphometrics, these traditional measures were then compared to the species groupings to see if they were in fact an accurate delimitation of species. Bookstein coordinates (Bookstein, 1991) were calculated using a baseline consisting of landmarks 1 and 8 (Figure 9), and I then used SPSS (Statistical Package for Social Science) (ver. 11.0.0, 2001) to analyze the measurements statistically. Besides landmarks 1 and 8 which were used as the baseline and left out of the analysis because they were normalized to 0,0 and 1,0, all of the other x and y coordinates of the other 26 landmarks. In combination with c-size (centroid size) and number of septa, a total of 54 variables were entered into the analysis. 9

I performed canonical discriminant analysis on the colonies from each formation in order to make initial groupings. These initial groupings were determined using a Mahalanobis distance matrix for each formation. After I had the groupings for each formation, the groups were then combined into one data set. Once I had all of the colonies into one analysis, I examined pairwise comparisons among colonies. If two colonies were not statistically significantly different (having a p-value of 0.05 or smaller) I would group them together and rerun the analysis. This process was continued until all groupings were found to be statistically different from one another and my final Mahalanobis distance matrix had p-values all below 0.05 (Table 5). In order to see if there was change over time within each of these species, I compared the morphospace occupation from one formation to the next within each species. If morphologic stasis occurred, the morphospace covered by each formation would completely overlap. If some sort of evolutionary change was taking place, the morphospace covered by each formation would not completely overlap and a directional change would be seen. I created boxplots and performed Kruskal-Wallis tests on the most significant variables and significant discriminant functions to compare one formation to the next to see if there was a statistically significant difference change with a species over time. Significant variables were determined by examining the structure matrix that was

produced from the canonical discriminant analysis (Table 6). To aid in proper identification of the species of Siderastrea I added 3 colonies of modern Siderastrea siderea and 3 colonies of Siderastrea radians collected in Jamaica into the analysis. These modern species were added to aid in species identification, and adding the modern species also was necessary to help estimate the possible morphologic change from the Pliocene until the recent.

A comparison of traditional morphometrics and 2-D geometric morphometrics was performed using the species groupings that were identified using 2-D geometric 10 morphometrics. This test was performed to examine if 2-D geometric morphometrics is a better identification tool for species, or if traditional measures are as equally affective as 2-D geometric morphometrics. Traditional measurements were calculated using TMorphGen (ver. 6c, 2004). Corallite radius was calculated by using the distance from landmark 1 to landmark 10 (Figure 9). After calculating corallite radius and making boxplots with number of septa and corallite radius versus the species groupings found using 2-D geometric morphometrics (Figures 11 and 12), Mann-Whitney U tests were performed to examine differences between DR morphospecies groups. 11

RESULTS

Species recognition After analyzing 64 coral colonies (384 corallites) using canonical discriminant analysis with groupings based on iterative group comparison, 4 groupings were found when plotting discriminant function 1 versus discriminant function 2 (Figure 13). Of these four species groupings 3 were previously known fossil species (DR morphospeciess 1, 3, and 4) (Figures 14, 15, and 16), and one a possible new species (DR morphospecies 2) (Figure 17). There was no overlap between any of the four species groups. Also, the three known species were established by the species of Siderastrea found in the Dominican Republic section by Budd et al. (1994). Figure 13 shows the distribution of the colony means and the hypothesized species groupings on a plot of discriminant function 1 versus discriminant function 2. The three landmarks that were most influential in discriminant function 1 are x20, x26, and x19 (Table 6). These three variables represent the outermost point of the secondary septa (x20), the right junction of the secondary septa with the corallite wall (x19), and the right junction of the tertiary septa with the corallite wall (x26). Discriminant function 1 explained 78.9% of the variance, discriminant function 2 explained 14.5% of the variance, and discriminant function 3 explained 6.7% of the variance (Table 3). Of these three function, the first two were found to be significant with p-values below 0.05, and the third function was found to not be significant with a p-value of 0.052 (Table 4).

Stratigraphic ranges Figure 18 shows the same plot as Figure 13 but with the markers set to display the formation in which the colony was sampled. The sampling in the 4 formations covers 12

approximately 9 million years of geologic time. Of the three known species (S. siderea, S. mendenhalli, S. silencensis) the 9 million year period that was sampled covers the majority of the temporal range of these species. The entire range of S. mendenhalli was sampled in this analysis, and for S. siderea and S. silencensis, over 50% of the total range was sampled. During this time period these three species of Siderastrea were known from Florida to the Dominican Republic to Costa Rica (Budd et al., 1994).

Morphologic change through time Figures 19-22 are boxplots of DR morphospecies 1, 3, and 4 (species 2 was left out because it is only found in a single horizon) examining the change of discriminant function 1, discriminant function 2, variable x20, and variable y4 throughout the Dominican Republic formations. Kruskal-Wallis tests showed the p-values of discriminant function 1, discriminant function 2, x20, x19, x26, and y4 throughout the formations for DR morphospecies 1, 3, and 4 to be over 0.05 (Table 7), indicating that there was no significant change in discriminant function 1 between any of the formations. Combined with Figure 18 which shows that the formations almost completely overlap within each species group, the data indicates that the different coral species occupied the same morphospace during the 9 million year time period represented in this study.

Comparisons of fossil and modern species When the modern samples of S. siderea (Figure 23) and S. radians (Figure 24) were added to the fossil data set, the results further supported the original groupings based solely on the fossil data. As seen in Figure 25, S. siderea grouped with DR morphospecies 3 (Figure 15), allowing us to interpret DR morphospecies 3 as being the same species as S. siderea. S. radians did not group with any fossil group which is expected due to the fact that S. radians has no fossil record in the Dominican Republic and its morphology is considerably reduced (Table 1) in comparison to the other species 13

of Siderastrea in the analysis. These two species were obviously distinct when using 2-D geometric morphometrics, but they were also easy to separate using traditional methods. The corallite diameter and number of septa were obviously different between the two modern species. The remaining two abundant DR morphospecies (DR morphospecies 1 and 4) can be classified as S. silecensis and S. mendenhalli. These identifications are based upon the species found in the study done by Budd et al. (1994). Traditional methods of identification identify S. silencensis as having an intermediate columella width and a thin corallite wall, and S. mendenhalli as having a thick corallite wall. Based solely on these traditional identifications, it is hard to correctly identify DR morphospecies 1 and 4 due to the fact that DR morphospecies 1 (Figure 14) has a thick columella and a thin wall, and DR morphospecies 4 (Figure 16) has a thick wall and a thin to intermediate columella.

Comparison of morphometrics The two traditional measures used for identifying species of Siderastrea are corallite diameter and number of septa. After making the boxplots of number of septa and corallite radius versus the species groupings determined with 2-D geometric morphometrics (Figures 11 and 12), no significant differences were found when examining the number of septa between the four DR morphospecies, but when looking at coral radius, a significant difference (p-value <0.05) was found between DR morphospecies 1 and DR morphospecies 4. The difference in sizes found between DR morphospecies 1 and 4 helps differentiate the 2 species. Siderastrea mendenhalli has a smaller corallite size in comparison to Siderastrea silecensis (Budd et al., 1994). 14

DISCUSSION

The scatter plot showing the fossil species of corals only (Figure 13) clearly outlines the three fossil species that were expected to be found in the Dominican Republic sections. DR morphospecies 1, 3, and 4 have absolutely no overlap and have significant gaps between them. This was important to establish before looking at morphometric change over time because I started this analysis with no initial species groupings. So, if overlap was seen it would make it more difficult to correctly parse out the species that were expected to be found and even more difficult to correctly observe how the species changed morphometrically through time. Along with the three fossil species that were expected to be seen in the Dominican Republic formations, a fourth fossil species was also found (DR morphospecies 2). Only three colonies of this fourth fossil species were found and analyzed. These three colonies only occurred in the Baitoa Formation which was sampled by Saunders et al. (1986). The Baitoa localities that contain colonies of Siderastrea are now under water due to dam construction near Santiago. The current three colonies of DR morphospecies 2 are only available in thin section which makes identification of DR morphospecies 2 difficult.

The identification of this species will be explored in future studies. With species groupings being established and with a well defined separation, it is possible to look at the morphological evolutionary history of each species. Due to the fact that DR morphospecies 2 is only found in the Baitoa Formation, it is not possible to get any data on how it changed over time so it is excluded from any statements on how Siderastrea changed morphometrically through time in the Dominican Republic. The first test of morphometric evolution was to mark cases by formation on the same scatter plot that was initially marked with fossil species. Figure 18 shows that the 15 formations completely cover the morphospace range of each fossil species (with the exception of DR morphospecies 2). If these species of Siderastrea where changing morphometrically through time, you would expect to see a directional shift in the area that each formation covered. Instead we see each formation completely cover the range of each species. Combined with the fact that there was significant environmental change with the depositional environments of the Gurabo and Cercado Formation, with it deepening up section (Nehm and Geary, 1994), it supports the hypothesis that these species of Siderastrea display morphologic stasis during the Mio-Pliocene of the Dominican Republic. To further test the hypothesis of stasis a Kruskal-Wallis was performed to see if discriminant function 1, which explained 78.9% of the total variance, showed any significant change between any of the formation in which species of Siderastrea were sampled. Along with discriminant function 1, Krustal-Wallis tests were also performed on discriminant function 2, variable x20, variable x19, variable x26, and variable y4. For all of the Kruskal-Wallis tests the p-values were greater than 0.05 (Table 7) which signifies that there is no significant difference in discriminant function 1 throughout the formations sampled. Boxplots showing discriminant function 1, discriminant function 2, variable x20, and variable y4 versus the formations sampled were also made to help illustrate the variance examined by the Kruskal-Wallis tests (Figures 19-22).

By adding the modern species of Siderastrea, it not only aided in species identification, but it gave some clues as to what the evolution history of S. siderea was from the Pliocene until the Recent. In Figure 25, the modern S. siderea plots to the upper right edge of the species boundary of the fossil S. siderea. This could indicate that from the Pliocene to the Recent, S. siderea has undergone some change and was no longer in stasis. This potential morphologic change could be attributed to the closing of the Isthmus of Panama and the Pleistocene coral turnover. This could also be an artifact of the geographic difference between the Dominican DR morphospecies and the Jamaican 16

Recent coral. In future studies colonies of S. siderea from the Pleistocene and Holocene can be added to this study to see if there is any morphologic change within this species during that time period. Another line of evidence that helps to support the hypothesis of evolutionary stasis for the species of Siderastrea is the apparent environmental change through part of the section. Nehm and Geary (1994) show that the water depth of the depositional environment of the Cercado and Gurabo Formation was deepening rapidly from the Cercado Formation through the Gurabo Formation (Figure 26). The morphologic stasis of Siderastrea during an obvious environmental change through this period of time further supports the hypothesis that Siderastrea was in morphologic stasis. While it was found that traditional morphometric techniques alone were not sufficient enough to successfully parse out all of the species, with no differences between number of septa and only 1 difference between corallite radius, they are still useful as a supplement to the 2-D geometric morphometric data. With both the 2-D geometric morphometric data and the tests done on the traditional measurements, I was able to identify DR morphospecies 1 as being Siderastrea silencensis, DR morphospecies 3 as Siderastrea siderea, and DR morphospecies 4 as Siderastrea mendenhalli. 17

FUTURE WORK

In order to complete this study and to be able to correctly identify if Siderastrea is undergoing punctuated equilibrium during the Mio-Pliocene, more samples of Siderastrea will need to be obtained. This study only looks at the Dominican Republic, and in order to accurately identify punctuated equilibrium, I will need to analyze colonies of Siderastrea from all across the Caribbean. I will contact repositories from other universities that have samples of Siderastrea from other areas in the Caribbean, and I will add them to this analysis. Not only will I need to increase my study geographically, but I will also need to attempt to expand my study temporally. While this study covers almost all of the range of S. mendenhalli, it can be expanded to cover the entire range of S. siderea, S. silecensis, and S. radians. Once the study has been expanded to more accurately represent the entire geographic and temporal range of Siderastrea, it will be possible to tell if the genus Siderastrea shows patterns of punctuated equilibrium or phyletic gradualism. Besides accurately analyzing Siderastrea, it would be beneficial to analyze other genera of corals. It would be interesting to see if punctuated equilibrium is isolated to certain clades of corals. In order to do this I will need to look at the most recent phylogenetic analysis of corals and select corals that have a variety of relationships with Siderastrea. By analyzing those closely related and those more distantly related to Siderastrea, it will be possible to see if evolutionary patterns follow phylogenetic relationships or if they are uncorrelated. The two prominent theories on evolutionary patterns have often been studied, but little has been done to see if evolutionary patterns are correlated to phylogenetic patterns. It would seem logical that if there are changes in evolutionary patterns, these patterns should be able to be marked on a cladogram. If 18 there was a shift in evolutionary patterns in corals, it would be nice to see that change separate out clades of corals. 19

CONCLUSIONS

Traditional measures have been used to identify colonies of Siderastrea in the past. While some separations were seen with corallite radius, all of the species were not parsed out. Also, there were no differences found between species when number of septa was analyzed. This data suggests that these traditional measures are not the most accurate methods to identify species. The similarity of these measures might have caused the misidentification of many colonies of Siderastrea in other analyses. 2-D Geometric morphometrics has been proven to be a much better way to accurately parse out species of Siderastrea, and with traditional measurements as a supplement all of the known fossil species in this study were identified. Using stepwise group comparisons and canonical discriminant analysis, three known DR morphospecies were found in the Mio-Pliocene of the Dominican Republic. These three known species were identified as S. siderea, S. mendenhalli, and S. silecensis. A fourth unknown species was also found in the Baitoa Formation. The successful parsing and identification of these species of Siderastrea supports the further use of these morphometric methods for similar studies.

Several lines of evidence of supported the theory that these species of Siderastrea have undergone evolutionary stasis during the Mio-Pliocene. Based on the scatterplot with markers set by formation (Figure 18), the formations cover the entire morphospace of the three known species. If there were any sort of morphologic change, one would expect to see some sort of linear change in the area covered by each formation. The boxplots of the discriminant functions and major variables (Figures 19–22) show that there is no significant difference between any of the variables through geologic time for each species. ANOVA tests also support the data illustrated by the boxplots (Table 7). 20

These data all support the original hypothesis of evolutionary stasis. Nehm and Geary (1994) showed the water depth of depositional environment of the Cercado and Gurabo Formation was increasing through time (Figure 26). During a period of non- evolutionary stasis, this environmental change would be expect to be reflected in coral morphology. This is not the case in this study and so it is likely that Siderastrea was in evolutionary stasis during the 9 million year time period of this study. 21

Figure 1 – An example of transverse section through a corallite of a species of Siderastrea.

Figure 2 - Map of the study area in the Dominican Republic with the sampled rivers color coded.

Figure 3. Comparison of age-depth plots of the Río Gurabo and Río Cana sections based on existing biostratigraphic and paleomagnetic (Gurabo section only) data. The sections are tied to established ostracode datum Radimella confragosa. The green and blue arrows are intervals of significant age.

Figure 4. A composite stratigraphic column for each of the sampled rivers of the Northern Dominican Republic with the rivers indicated along the top. 28

Figure 5. A stratigraphic column of Rio Cana with NMB localities marked at their position in the stratigraphic column.

Figure 6. A stratigraphic column of Rio Gurabo with NMB localities marked at their position in the stratigraphic column.

Figure 7. Locality map for Arroyo Bellaco. Sampling locality Bel-1, Bel-7, and Bel-8 correspond with CCE-12, Bel-4 corresponds with CCE-4, and Bel-3 corresponds with CCE-5. (NMITA)

Figure 8. A stratigraphic column of Rio Yaque del Norte with NMB localities marked at their position in the stratigraphic column.

Figure 9. Transverse sections of a colony of Siderastrea showing a close up view of the landmark scheme used in the 2-D geometric morphometric analysis.

Figure 10. Transverse sections of a colony of Siderastrea showing 1/12 of the corallite used for landmark measurement in the 2-D geometric morphometric analysis.

Figure 11. Boxplot illustrating the variation of number of septa throughout the 4 DR morphospecies. Kruskal-Wallis tests of the data for each graph gives a p-value above 0.05 indicating no difference between the DR morphospecies groups.

Figure 12. Boxplot illustrating the variation of coral radius throughout the 4 DR morphospecies. Kruskal-Wallis tests of the data for each graph gives a p- value above 0.05 indicating no difference between the DR morphospecies groups, with the exception of a difference between DR morphospecies 1 and DR morphospecies 4.

Figure 13. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR morphospecies 1-4. Groups are labeled by fossil species.

Figure 14. A photograph of a transverse section through a corallite of DR morphospecies 1.

Figure 15. A photograph of a transverse section through a corallite of DR morphospecies 3.

Figure 16. A photograph of a transverse section through a corallite of DR morphospecies 4.

Figure 17. A photograph of a transverse section through a corallite of DR morphospecies 2.

Figure 18. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR morphospecies 1-4. Groups are labeled by formation.

Figure 19. Boxplots illustrating the variation of discriminant function 1 throughout the sampled formations for DR morphospecies 1, 3, and 4 (labeled at top of each graph). Kruskal-Wallis tests of the data for each graph gives a p-value above 0.05 indicating no difference between the formations.

Figure 20. Boxplots illustrating the variation of discriminant function 2 throughout the sampled formations for DR morphospecies 1, 3, and 4 (labeled at top of each graph). Kruskal-Wallis tests of the data for each graph gives a p-value above 0.05 indicating no difference between the formations.

Figure 21. Boxplots illustrating the variation of variable x20 throughout the sampled formations for DR morphospecies 1, 3, and 4 (labeled at top of each graph). Kruskal-Wallis tests of the data for each graph gives a p-value above 0.05 indicating no difference between the formations.

Figure 22. Boxplots illustrating the variation of variable y4 throughout the sampled formations for DR morphospecies 1, 3, and 4 (labeled at top of each graph). Kruskal-Wallis tests of the data for each graph gives a p-value above 0.05 indicating no difference between the formations.

Figure 23. A photograph of a transverse section through a corallite of a modern Siderastrea siderea from Jamaica.

Figure 24. A photograph of a transverse section through a corallite of a modern Siderastrea radians from Jamaica.

Figure 25. Scatterplot of discriminant function 1 vs. discriminant function 2 of DR morphospecies 1–4 and modern S. siderea and modern S. radians. Groups are labeled by species groupings.

Figure 26. Estimated depositional water depths for Cercado and Gurabo Formation from Nehm and Geary (1994).

Table 1 – Morphologic characters distinguishing eight Caribbean species of Siderastrea. Number of Corallite Species Distribution septa per diameter Columella Corallite wall corallite (mm) S. radians Middle Pliocene to thick, solid; thick, 2-3 synap. Recent; Caribbean, 30-40 2.5-3.5 intermediate rings; septa usually Bermuda, Brazil, W. fossa depth continuous between Africa calices

S. siderea early Miocene to think, thin, 3-5 synap. Recent; Caribbean 44-50 3-5 papillose; rings; septa deep fossa alternate between calices

S. stellata Recent; Brazil thin thin, 3-4 synap. ~48 ~3 papillose; rings; septa usually very deep continuous between fossa calices

S. mendenhalli early Miocene to early thick; thick, 3-4 synap. Pliocene; Dominican 48-54 3-5 shallow fossa rings; septa Republic, California continuous between calices

S. silecensis early Miocene to early intermediate thin, 3-5 synap. Pleistocene; Florida, 48-60 >4.5 thickness; rings; septa Dominican Republic deep fossa continuous between calices

S. pliocenica middle Pliocene to early thick, solid; thick, 4-5 synap. Pleistocene; Florida 40-48 4.5-5 shallow fossa rings; septa usually continuous between calices

S. glynni Recent; eastern Pacific Intermediate intermediate 40-48 2.5-3.5 thickness, thickness; 3-4 papillose; synap. rings; septa shallow fossa usually continuous between calices S. savignyana Recent; Red Sea, Indian thick, solid; very thick, 2-3 Ocean 28-35 2.5-4 intermediate synap. rings; septa fossa depth continuous between calices

Source: Budd and Guzman, 1994

Table 2 - A list of the localities sampled, the number of samples collected, the Formation name, the river the samples were collected from, and the age of samples collected from those localities. Number of Formation River collected Location age samples name from collected 15830 1 Mao Gurabo Pliocene 15845 2 Gurabo Gurabo Miocene 15846 1 Gurabo Gurabo Miocene 15848 2 Gurabo Gurabo Miocene 15859 5 Gurabo Gurabo Miocene

16815 2 Gurabo Cana Pliocene 16817 2 Gurabo Cana Pliocene 16818 24 Gurabo Cana Pliocene 16819 1 Gurabo Cana Pliocene 16883 1 Gurabo Gurabo Miocene

16884 10 Mao Cana Pliocene 16934 1 Gurabo Gurabo Miocene 16937 4 Baitoa Yaque del Norte Miocene 16939 3 Baitoa Yaque del Norte Miocene 16943 8 Baitoa Yaque del Norte Miocene

16944 5 Baitoa Yaque del Norte Miocene 17273 1 Cercado Yaque del Norte Miocene 17282 1 Baitoa Yaque del Norte Miocene 17289 1 Baitoa Yaque del Norte Miocene CCE 4 1 Cercado Arroyo Bellaco Pliocene

CCE 5 1 Cercado Arroyo Bellaco Pliocene CCE 12 23 Cercado Arroyo Bellaco Pliocene 75

Table 3. Eigenvalues displaying the percent variation covered by each of the 3 discriminant functions.

Function Eigenvalue % of Variance Cumulative % Canonical Correlation 1 59.657 78.9 78.9 .992

2 10.933 14.5 93.3 .957

3 5.034 6.7 100.0 .913

Table 4. Wilks' Lambda chart displaying the significance of each of the three discriminant functions.

Test of Function(s) Wilks' Lambda Chi-square df p-value

1 through 3 .000 297.560 147 .000

2 through 3 .014 151.824 96 .000

3 .166 63.809 47 .052

*The first 2 functions are found to be significant with a p-value below 0.05, but the third function has a p-value of 0.052 and is therefore not significant.

Table 5. Mahalanobis distance matrix reporting the significance between groups. DR morphosp Step ecies 1.00 2.00 3.00 4.00 49 1.00 F 4.179 3.627 7.387 p-value .007 .012 .001 2.00 F 4.179 6.852 10.034 p-value .007 .001 .000 3.00 F 3.627 6.852 3.554 p-value .012 .001 .013 4.00 F 7.387 10.034 3.554 p-value .001 .000 .013 * All p-values are below 0.05 indicating that each DR morphospecies group is statistically significantly different from the other DR morphospecies groups. 78

Table 6. Structure matrix for the canonical discriminant analysis displaying the significant variables for each discriminant function.

Function 1 2 3 x20_1 -.070(*) -.047 -.028 x26_1 -.065(*) -.029 -.051 x19_1 -.060(*) -.042 -.055 x21_1 -.059(*) -.038 -.058 x15_1 -.054(*) -.025 .000 x27_1 -.054(*) -.020 -.040 y2_1 .052(*) .008 -.039 x13_1 -.052(*) -.007 .010 x12_1 -.038(*) -.021 .024 y4_1 .047 .093(*) -.029 y15_1 .015 .080(*) -.037 y24_1 -.002 .062(*) .043 y18_1 -.012 .062(*) -.029 y19_1 -.007 .061(*) -.038 y21_1 .003 .060(*) -.040 y17_1 -.009 .057(*) -.013 y22_1 -.002 .057(*) -.033 y23_1 -.007 .053(*) -.017 x9_1 -.023 -.050(*) .027 septa_1 .015 -.046(*) -.021 y5_1 .004 .041(*) -.027 y11_1 .015 .039(*) -.039 y7_1 -.015 .036(*) -.030 x4_1 .063 .089 -.267(*) x5_1 .001 .008 -.208(*) x3_1 .072 .102 -.208(*) csize_1 .057 -.083 .202(*)

Table 6-continued x2_1 .048 .111 -.180(*) x6_1 .007 -.006 -.150(*) x7_1 -.021 -.071 -.128(*) x14_1 -.063 -.004 .123(*) x28_1 -.083 -.012 .121(*) y3_1 .077 .063 -.115(*) y28_1 .024 .078 -.110(*) y14_1 .028 .034 -.103(*) y16_1 .011 .077 -.085(*) x16_1 -.049 -.025 .084(*) y9_1 .011 .075 -.078(*) y12_1 -.056 .031 .076(*) x25_1 -.062 -.022 -.069(*) y20_1 .022 .057 -.068(*) x23_1 -.049 -.005 -.065(*) y25_1 -.002 .036 -.064(*) x17_1 -.049 -.030 -.063(*) y27_1 -.005 .060 -.062(*) y26_1 -.016 .048 -.062(*) x24_1 -.061 -.031 -.061(*) y13_1 .027 .061 -.061(*) x18_1 -.051 -.031 -.055(*) x22_1 -.053 -.014 -.054(*) x10_1 .016 .019 -.053(*) x11_1 -.017 .018 -.034(*) y10_1 .008 -.006 -.022(*) y6_1 -.011 .003 -.012(*) * The largest absolute correlation between each variable and any discriminant function is marked with a *. 80

Table 7. Results of the Kruskal-Wallis test of the three known DR morphospecies throughout the Dominican Republic formations.

Discriminant Discriminant group2 y4_1 x19_1 x20_1 x26_1 Function 1 Function 2 1.00 Chi-Square 7.767 2.420 2.381 2.120 6.528 3.974 df 4 4 4 4 4 4 p-value .100 .659 .666 .714 .163 .410 3.00 Chi-Square 5.836 4.709 5.142 4.315 2.137 5.363 df 4 4 4 4 4 4 p-value .212 .318 .273 .365 .711 .252 4.00 Chi-Square 2.690 4.967 4.765 4.230 3.123 1.331 df 3 3 3 3 3 3 p-value .442 .174 .190 .238 .373 .722 * All p-values are above 0.05 indicating that there is no significant change in any of the variables analyzed over the period of time sampled. 81

APPENDIX A LANDMARK DESCRIPTION

Table A1. Landmarks taken on transverse thin-sections of corallites of Siderastrea. Number Type Description 1 3 Outermost point of primary septa 2 1 Left junction of primary septa with corallite wall 3 1 Outer left junction of first synapticulae with primary septa 4 1 Inner left junction of first synapticulae with primary septa 5 1 Inner right junction of first synapticulae with primary septa 6 1 Outer right junction of first synapticulae with primary septa 7 1 Right junction of primary septa with corallite wall 8 1 Left junction of primary septa with columella 9 1 Right junction of primary septa with columella 10 3 Center of corallite 11 1 Right junction of secondary corallite with columella 12 1 Left junction of secondary corallite with columella 13 1 Inner right junction of secondary septa with tertiary septa 14 1 Inner left junction of secondary septa with tertiary septa 15 1 Outer right junction of secondary septa with tertiary septa 16 1 Outer left junction of secondary septa with tertiary septa 17 1 Inner right junction of first synapticulae with secondary septa 18 1 Outer right junction of first synapticulae with secondary septa 19 1 Right junction of secondary septa with corallite wall 20 3 Outermost point of secondary septa 21 1 Left junction of secondary septa with corallite wall 22 1 Outer left junction of first synapticulae with secondary septa 23 1 Inner left junction of first synapticulae with secondary septa 24 1 Right junction of quaternary septa with corallite wall 25 1 Left junction of quaternary septa with corallite wall 26 1 Right junction of tertiary septa with corallite wall 27 1 Left junction of tertiary septa with corallite wall 28 3 Inntermost point of quaternary septa * Types are: 1 = juxtoposition of structures; 3 = extremal points.

APPENDIX B SAMPLE INFORMATION

Table B1. List of specimens analyzed in morphometric analyses.

Colony # Formation locality # of septa Catalog # Photo # 1 Low mid Gurab 16818 48 102554 6684 2 Cercado CCE-12 44 102555 6718 3 Cercado CCE-12 46 102556 6700 4 Cercado CCE-12 48 102557 6690 5 Cercado CCE-12 50 102558 6706 6 Cercado CCE-12 50 102559 6712 7 Mid Gur 15845 48 102560 6977 8 Mid Gur 15845 50 102561 6971 9 Low mid Gurab 16818 44 102562 6965 10 Low mid Gurab 16819 52 102563 6995 11 Cercado CCE-12 48 102564 6983 12 Cercado CCE-12 34 102565 6989 13 Low mid Gurab 16817 44 102566 7001 14 Low Gurabo 16837 48 NA 7067 15 Mid Mao 16884 46 102567 7061 16 Baitoa 16937 50 D5781 7025 17 Baitoa 16937 56 D5781 7073 18 Baitoa 16943 60 D5782 7031 19 Baitoa 16943 50 D5782 7037 20 Baitoa 16943 48 D5783 7043 21 Baitoa 16943 56 D5784 7085 22 Baitoa 16944 58 D5785 7049 23 Baitoa 16944 62 D5786 7055 24 Baitoa 17273 52 SH Norte7 7079 25 Mid Gur 15846 46 102568 1752 26 Mid Gur 15859 38 102569 1728 27 Mid Gur 15859 68 102570 1905 28 Mid Gur 15859 48 102571 1912 29 Low Gur 15885 58 102572 1780 30 Low mid Gurab 16817 50 102573 1812 31 Low mid Gurab 16818 46 102574 1938 32 Low mid Gurab 16818 46 102575 1800 33 Low mid Gurab 16818 48 102576 1806 34 Low mid Gurab 16818 48 102577 1845 35 Low mid Gurab 16818 40 102578 1679 36 Low mid Gurab 16818 44 102579 1773 37 Low mid Gurab 16818 50 102580 1953 38 Low mid Gurab 16818 48 102581 1708 39 Low mid Gurab 16818 48 102582 1864 40 Low mid Gurab 16818 48 102583 1884 85

Table B1-continued

Colony # Formation locality # of septa Catalog # photo # 41 Low mid Gurab 16818 98 102584 1824 42 Low mid Gurab 16818 34 102585 1831 43 Low mid Gurab 16818 56 102586 1699 44 Low mid Gurab 16818 50 102587 1870 45 Low mid Gurab 16818 46 102588 1715 46 Low mid Gurab 16818 46 102589 1858 47 Mid Mao 16884 48 102590 1877 48 Mid Mao 16884 44 102591 1966 49 Cercado CCE4 48 102592 1786 50 Cercado CCE5 52 102593 1973 51 Cercado CCE12 46 102594 1685 52 Cercado CCE12 44 102595 1691 53 Cercado CCE12 32 102596 1759 54 Cercado CCE12 48 102597 1838 55 Cercado CCE12 50 102598 1959 56 Cercado CCE12 54 102599 1745 57 Cercado CCE12 62 102600 1739 58 Cercado CCE12 44 102601 1818 59 Cercado CCE12 48 102602 1721 60 Cercado CCE12 52 102603 1945 61 Cercado CCE12 45 102604 1851 62 Cercado CCE12 46 102605 1931 63 Cercado CCE12 50 102606 1767 86

REFERENCES

Bell MA, Baumgartner JV, and Olson EC (1985) Patterns of temporal change in single morphological characters of a Miocene stickleback fish. Paleobiology 11: 258-271. Bookstein FL (1991) Morphometric tools for landmark data. Cambridge: Cambridge University Press. Budd AF, Foster Jr. CT, Dawson JP, Johnson KG (2001) The Neogene Marine Biota of Tropical America (“NMITA”) Database: Accounting for Biodiversity in Paleontology. Journal of Paleontology 75(3) 743-751. Budd AF, Guzman HM (1994) Siderastrea glynni, a New Scleractinian Coral (Cnidaria: Anthozoa) From the Eastern Pacific. Proceedings of the Biological Society of Washington 107(4) 591-599. Budd AF, Stemann TA, Johnson KG (1994) Stratigraphic Distributions of Genera and Species of Neogene to Recent Caribbean Reef Corals. Journal of Paleontology 68(5) 951-977. Cheetham AH (1986) Tempo of evolution in a Neogene bryozoan: Rates of morphologic change within and across species boundaries. Paleobiology 12: 190-202. Eldredge N, and Gould SJ (1972) Punctuated equilibria: An alternative to phyletic gradualism. In T. J. M. Schopf, ed., Models in Paleobiology, pp. 82-115, San Francisco: Freeman, Cooper. Erwin DH, Anstey RL (1995) New Approaches to Speciation in the Fossil Record. Columbia University Press, New York. Garcia E, Ramos R, and Bastidas C (2005) Presence of cytochrome P450 in the Caribbean corals Siderastrea siderea and Montastraea faveolata. Ciencias Marinas 31 (1A): 23-30. Gould SJ (1977) Eternal Metaphors of Paleontology. Patterns of Evolution. Ed. A. Hallam. Elsevier Scientific Publishing Company, Amsterdam. Image J, Rel. 1.31. 2004. U.S. National Institutes of Health. Klaus JS, Budd AF, McNeill DF (2004) Reef coral assemblages of the Miocene to Pliocene of the northern Dominican Republic. Geological Society of America, north- central section, 38th annual meeting. Abstracts with Programs - Geological Society of America 36, (3). Lazarus D (1986) Tempo and mode of morphologic evolution near the origin of the radiolarian lineage Pterocanium prismatium. Paleobiology 12: 175-189. Nehm RH (2005) Patterns and processes of evolutionary stasis and change in Eratoidea (Gastropoda: Marginellidae) from the Dominican Republic Neogene. Caribbean Journal of Science 41(2): 189-214. 87

Nehm RH, Geary DH (1994) A Gradual Morphologic Transition During a Rapid Speciation Event in Marginellid Gastropods (Neogene: Dominican Republic). Journal of Paleontology 68(4): 787-795. Potts DC, Budd AF, and Garthwaite RL (1993) Soft tissue vs. skeletal approaches to species recognition and phylogeny reconstruction in corals. Courier Forschungs- Institut Senckenberg 164: 221-231. Saunders JB, Jung P, Biju-Duval B (1986) Neogene Paleontology in the Northern Dominican Republic: 1. Field Surveys, Lithology, Environment, and Age. Bulletins of American Paleontology 89(323). SPSS for Windows, Rel. 11.0.0. 2001. Chicago: SPSS Inc. Stanley SM, and Yang X (1987) Approximate evolutionary stasis for bivalve morphology over millions of years: A multivariate, multilineage study. Paleobiology 13: 113-139. Vermeij MJA (2005) Substrate composition and adult distribution determme recruitment patterns in a Caribbean brooding coral. Marine Ecology-Progress Series 295: 123- 133 Veron JEN (1995) Corals in Space and Time. UNSW Press, Sydney.