MIDDLE PLEISTOCENE OCEANOGRAPHIC AND ENVIRONMENTAL CONDITIONS IN SHALLOW-MARINE WATERS OF THE SOUTHWESTERN COLOMBIAN BASIN: INFERENCES FROM THE STABLE-ISOTOPE RECORD OF A FOSSIL CORAL by STEPHEN A. SCHELLENBERG A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Geology University of South Florida August 1994 Major Professor: Terrence M Quinn, Ph.D.
Graduate School University of South Florida Tampa, Florida CERTIFICATE OF APPROVAL Master's Thesis This is to certify that the Master's Thesis of STEPHEN A. SCHELLENBERG with a major in Geology has been approved by the Examining Committee on July 15, 1994 as satisfactory for the thesis requirement for the Master of Science degree Examining Committee: Major Terrence M .t6uinn, Ph.D. MernbeP.lisa L. Ph.D / [/. Member : H. Leonard Vacher Ph.D.
DEDICATION To my family, for support and love without end or limit. To past, present, and future fellow students for always tickling my brain and kicking my arse when necessary To my teachers, from kindergarten to graduate school whose patience and dedication generally exceeded my appreciation and feedback at the time.
ACKNOWLEDGMENTS The deepest of gratitude is extended to Dr. Terry Quinn for taking me on as his student, making me go to the blackboard when necessary and advising me in intellectual, academic, and personal matters. I extend my sincere thanks to my past and present committee members for their time thoughts, and critical reviews: Dr. Bob Halley, Dr. Peter Harries Dr. Lisa Robbins, and Dr. Len Vacher. Special thanks to Dr. Lisa Robbins for joining my committee in the last moments of this venture. Funding for this project was provided by student res earch grants to the author from th e Geological Society of America, Sigma Xi, an d the Gulf Coast Association of Geological Societies. Logistical support was provided b y the Smithsonian Tropical Research Institute. Field work without the accompaniment of Pam Borne, Dr. Nancy Budd, Dr. Don McNeill, and Jamie Wineberg would have be e n significantly les s educ ationa l rewarding, and fun Coral prepara tion and sampling was done unde r the generous hospitality of Dr. Bob Halley at the USGS Center for Coastal Geology (St. Petersburg, FL). Dr. Steven Clemens (Brown University) kindly conducted pro bono 87Sr j 86Sr analyses and Dr. Harry Dowsett (USGS) kindly examined Lomas del Mar microfossils for biostratigraphy. And last but ce rtainly not l eas t thanks to my fellow PASGL studentsSuzanne Link, Rick Kayser, Leanne Roulier and J en i Wyatt-for holding m y hand and speaking slowly when n ecessa ry.
ACKNOWLEDGMENTS The deepest of gratitude is extended to Dr Terry Quinn for taking me on as his student, making me go to the blackboard when necessary, and advising me in intellectual, academic and personal matters. I extend my sincere thanks to my past and present committee members for their time, thoughts, and critical reviews: Dr. Bob Halley, Dr. Peter Harries, Dr. Lisa Robbins, and Dr. Len Vacher. Special thanks to Dr. Lisa Robbins for joining my committee in the last moments of this venture Funding for this project was provided by student research grants to the author from the Geological Society of America, Sigma Xi, and the Gulf Coast Association of Geological Societies. Logistical support was provided by the Smithsonian Tropical Research Institute. Field work without the accompaniment of Pam Borne Dr. Nancy Budd, Dr Don McNeill, and Jamie Wineberg would have been significantly less educational rewarding, and fun. Coral preparation and sampling was done under the generous hospitality of Dr. Bob Halley at the USGS Center for Coastal Geology (St. Petersburg FL). Dr. Steven Clemens (Brown University) kindly conducted pro bono 87Sr j 86Sr analyses and Dr Harry Dowsett (USGS) kindly examined Lomas del Mar microfossils for biostratigraphy. And last, but certainly not least, thanks to my fellow P ASGL studentsSuzanne Link, Rick Kayser Leanne Roulier and Jeni Wyatt-for holding my hand and speaking slowly when necessary.
TABLE OF CONTENTS LIST OF TABLES ii LIST OF FIGURES 111 ABSTRACT v I. INTRODUCTION 1 II SKELETAL CHEMISTRY AS AN ENVIRONMENTAL ARCHIV E 4 Sclerochronology 4 Oxygen Isotopes 4 Carbon Isotopes 6 Cova riance of Oxygen and Carbon Isot opes in Coral Records 6 III. E NVIRONMENTAL AND GEOLOGICAL SETTING 10 Modern Oceanography and Climate 10 Geo lo gic and Biologic Setting 12 IV. METHODS 1 3 V RESULTS 1 5 VI. DISCUSSION 19 Age of the Lomas del Mar Sect i o n 19 Inferr e d Paleoenvironment f rom Foss il Co r a l Isotopes 21 Sampling Effects 21 Kinetic Effects 22 Paleoenvironmental Inferences 24 Quantitative Estimates of SST and SSS from Coral o180 28 Annual o180 Range as a Function of Eithe r SST or SSS 29 Annual o180 Rang e as a Function of Both SST o r SSS 30 VI. CONCLUSIONS 33 VII. REFERENCES 34 VIII. APPENDIXI SO TOPIC DATA FOR MIDDLE PLEISTOCENE CORAL 40
Table 1. Table 2. Table 3. Table 4 LIST OF TABLES Statistics for High-Density and 813C Sclerochronologies Modes of Coral-Isotope Record in 813C Sclerochronology Middle Pleistocene Annual Coral o18Q Range Interpreted as Either SST or SSS, Exclusively Comparison of Middle Pleistocene Seasonality Calculated using Modern Relative Magnitudes and Phase Relationships for Temperature and Salinity from Puerto Limon 11 18 26 30 32
LIST OF FIGURES Figure 1. Map of the Study Area 3 Figure 2 Generalized Covariance Patterns of Modern Coral Isotope Records from Various Climate Regimes 7 Figure 3. (A.) Mean Monthly SST for Puerto Limon and Southwestern Columbian Basin (9-11 o N, 81-83 W) and (B.) Mean Monthly SSS and Rainfall for Puerto Limon 11 Figure 4 Positive Print of X-Radiograph (60 kV, 3 rnA, 25 sec) for Fossil Siderastrea siderea Coral 14 Figure 5. 8180 and 813C Profiles for Fossil Siderastrea siderea Coral 16 Figure 6. Comparison of (A. ) Mean Annual 8180 Values and (B.) Mean Annual 813C Values Derived from 813C and High-Density Band Sclerochronologies 17 Figure 7. Age Data on the Moin Formation and Lomas del Mar Section 20 Figure 8. (A.) Sampling Effects and (B.) Kinetic Effects on Annual 8180 and 813C Values 23 Figure 9. Covariance of Foss il Coral 8180 and 813C Values 25 Figure 10. (A.) Predicted 8180 Values and (B.) Stable Isotope Covariance Pattern for Modern Coral Growing in Puerto Limon Coastal Waters 27 Figure 11. Covariance of SST and SSS in Modern Coastal Waters Off Puerto Limon 31 iii
MIDDLE PLEISTOCENE OCEANOGRAPHIC AND ENVIRONMENTAL CONDITIONS IN SHALLOW-MARINE WATERS OF THE SOUTHWESTERN COLOMBIAN BASIN: NFERENCES FROM THE STABLE-ISOTOPE RECORD OF A FOSSIL CORAL by STEPHEN A. SCHELLENBERG An Abstract Of a thesis submitted in partial fulfillment of the requirements for the degree of Ma s ter of Science Department of Geology University of South Florida August 1994 Major Professor: Terrence M Quinn, Ph. D lV
The Plio-Plei s t o cen e Moin Forma tion (Limon, Costa Rica) contains an important r ecord of late Neogene oceanographic and environmenta l changes that o ccurred in the Columbian Basin after th e closure of the Central American Isthmus. T o bette r understand this histor y, we have inves tigat e d the age and paleoenvironment of the Lomas del Mar section of the Moin Formation and the paleoseasonality of ambient shallow-marine waters using the skeletal chemistry of a pristin e fos s il Siderastrea siderea s cleractinian coral. This study represents, to th e best of our knowledge, the first application o f the co ral-isotop e approach to r econstructin g environmental and seasonal conditions of the middle Pleistocene. Strontium-isotope analyses of coral aragonit e was u se d to constrain the age of the sec tion and indicates a n apparent age of 1.03 Ma, with upper and l o w e r age limit s of 0.38 Ma and 1.63 Ma (2cr). This age i s near the upper limit of a nannofo ssi l-b ased age of to M a and is consistent with age estimates based on planktonic foraminifera. Stable-isotope analyses of a contiguously sampling (0.5-mm interval) 60-rnrn-long section of a sing l e corallite yielded a mean o18Q value of -2.13 0.27 %o (PDB) and mean o13C value of -2 .09 0.2 3 %o (PDB). Xradiography o f th e coral r eve al 2 8 diffuse an d s ubtl e hi gh-and low-density b and couplets transve r se to the sampled co r a llit e sec t ion, altho ugh a co n s i s tent phase r e l atio n ship with eit h e r ()18Q and ()13C i s absent. Therefore, an annual scle ro chro nology based o n ()13C r e l a tive maxima was used to d e m arcate 28 years of growth, producing a n average of samples per year. Mean annua l ()18Q and ()13C r a n ges for the 28-year r ecord are 0 27.16 % o and 1.03.54 %o, respectively. Annual ranges in ()1 8Q show a n eglig ibl e corre lation to annua l sample density (r = 0 28), while annual ()13C ranges cor r ela t e more strongl y (r=0.44). In c r eases in annual g rowth rate (mm/ y r) are v
correlated with lower mean annual values in both 8180 (r=0.46) and 813C (r=0.46), and are interpreted to reflect increased kinetic fractionation with increased growth rate. Predicted modern coral 8180 and 813C values, based on long-term local sea-surface temperature (SST), sea-surface salinity (SSS), and cloudiness records, are positively correlated and consistent with numerous modern coral-isotope records from regions with a climate characterized by alternating warm-cloudy and cool-clear seasons. In the fossil coral, a strong positive correlation and phase relationship between 8180 and 813C values imply that middle Pleistocene seasonality was similar to those of modern environmental conditions off Puerto Limon today. The calculated middle Pleistocene SST range (2.4.4 C0 ) is nearly identical to observed modern SST range (2.4.1 C0), while the calculated middle Pleistocene SSS range (6.8.1 %o) i s over 2 %o larger than modern SSS range (4.6.9 %o). Although such calculations provide quantitative paleoseasonality estimates, they assume constancy of the relation between SST, SSS, and t1(8180) / t1salinity in the modern and middle Pleistocene. Abstract Approved: Major Professor: Terrence M. Quinn, Ph.D. Assistant Professor, Department of Geology Date Approved: vi
1 I. INTRODUCTION Paleoseasonality is an important, but often poorly documented, parameter in the reconstruction of ancient marine environments and climate regimes (Crowley and North, 1991) Modern seasonal fluctuations in seawater temperature and salinity reflect ocean-atmosphere processes and greatly influence the structure, diversity, and richness of marine ecosystems At lower frequencies, climate perturbations (e.g., the El Nino-Southern Oscillation (ENSO)) and net climate changes are often expressed as regional shifts in seasonal temperature and precipitation patterns (i.e., seasonality; Karl et al., 1991) which m ay induce ecological and evolutionary responses (c.f., Benn e t, 1990). Estimates of paleoseasonality are frequently derived from broader p a leoenvironmental reconstructions, fauna-based transfer functions, and climate models. Such interpretations however, are generally restricted to average summer-winter extrema over mill e nnia! or longer intervals that preclude insight into inter-and intra-annual environmental variability Records of ambient environmental conditions are often preserved in the skeletal chemistry of marine invertebrates (e.g., Wef e r and Berger, 1991). In particular, the skeletal-isotope records of scleractinian corals provide a robust archive of sea-surface conditions. A greatly increased understanding of environmental and physiological controls upon skeletal-isotopic composition, together with their fidelity with ambient seawater records in modern studies, have validated scleractinians as excellent high-resolution environmental recorders (e.g. Dunbar and Cole, 1993).
2 The success of these modern studies has prompted the use of fossil coral-isotope records to open "real-time" windows on subannual-to subcentennial-scale variability in paleoenvironmental and paleoclimatic conditions. For example, using oxygen-isotope data from fossil Acropora palmata corals of Barbados, Guilderson et al. (1993) have proposed significantly greater cooling (-5 C0 ) of tropical sea-surface temperatures (SST) at 19 ka than indicated by CLIMAP's microfossil-based SST-cooling estimate of 2 C0 Roulier and Quinn (in review), using an exquisitely preserved 49-year coral-isotope record from the Pliocene Pinecrest Beds, have reconstructed climatic conditions in southwestern Florida during the middle Pliocene warm interval to be similar to slightly cooler than today, with no evidence of ancient upwelling. Such records provide important environmental and climatic data that are unobtainable by most other methods. In this study, the age and paleoenvironm ent of the Lomas d e l Mar section of the Plio-Pleistocene Moin Formation in Limon, Costa Rica (Figure 1) and the paleoseasonality of the time are investigated using the skeletal chemistry of a pristine fossil Siderastrea siderea scleractinian coral. The Moin Formation contains an important record of oceanographic, environmental, and paleobiological changes in the southwestern Colombian Basin that occurred following the final closure of the Central American Isthmus (Coates et al. 1992). 87Sr j86Sr stratigraphy is used to refine the temporal position of the Lomas del Mar section within the Moin Formation. A 28-year record of coral oxygen-and carbon-isotope variation in the fossil coral is presented. Stable-isotope covariance is interpreted to refl e ct ambient paleoenvironmental conditions. Estimates of middle Pleistocene sea-surface temperature (SST) and s ea-surface salinity (SSS) ranges are calculated from the coral 818Q record
Figure 1. 0 1 Map of the Study Area. This map shows the location and geologic setting of the Lomas del Mar section of the Moin Formation (modified from Coates et al., 1992). Elevation of the section is -50 m. Hatched square (9-11 o N, 81-83 W) is grid of monthly SST data obtained from the Comprehensive Ocean Atmosphere Data Set (COADS) from 1940 to 1990. 3
4 II. SKELETAL CHEMISTRY AS AN ENVIRONMENTAL ARCHIVE Sderochronology The longevity and rapid growth of scleractinian corals provide an excellent medium for reconstructing past sea-surface conditions at monthly and, in some cases, weekly resolution. Annual variations in skeletal extension rate produce high-and low-density band couplets that normally represent one year of growth (Buddemeier et al., 1974). These annual density couplets, revealed by X radiography, provide a general context for determining growth rates and isotopic sampling strategies (Knutson et al. 1972; Dodge and Vaisnys, 1980) Cyclicity in skeletal 8180 and 813C values are also used to demarcate annual cycles, particularly in coral records with irregular, subtle, or absent density banding (Cole and Fairbanks 1990; Shen et al., 1992; Carriquiry et al., 1994). Oxygen Isotopes Seasonal fluctuations in the temperature and 8180 of seawater largely control seasonal fluctuations in coral 8180 records. Studies of modern corals have demonstrated skeletal 8180 to decrease by 0.22 %o per 1 oc increase in ambient seawater temperature (Weber and Woodhead, 1972 ; Druffel 1985). This relationship is the same as the classic paleotemperature equation of Epstein et al. (1953) although absolute coral-isotope temperature equations are negatively offset from the Epstein et al. paleotemperature equation by
5 various amounts at the generic level (Weber and Woodhead, 1972) This negative offset is attributed to a "kinetic" effect, whereby lighter-mass isotopes (e g., 16Q, 12C) are more preferentially reacted in C02 hydration and hydroxylation as calcification reaction rates increase (McConnaughey, 1989a). Sampling of corals along axes of maximum and most consistent growth rate minimizes this kinetics-ba se d disequilibrium variability (McConnaughey, 1989a; Cole and Fairbanks, 1990). Thus, regardless of genus, seasonal temperature cycles may be calculated from seasonal oscillations in coral 818Q values under known or constant 8180seawater conditions. Changes in the 8 1 80seawater, howeve r commonly occur at various frequencies. Daily to annual cycles in 8180seawater result from net input of 16Q-enriched freshwater (i.e., rainfall river discharge meltwater, etc.) or net r emoval of 160-enriched water vapor by seawater evaporation (Yurtsever and Gat, 1981). Because these processes also affect the concentration of dissolved solids in seawater strong co rrelations often exist between sea-surface salinity and 8180seawater (Epstein and Mayeda, 1953; Craig and Gordon, 1965; Dunbar and Wellington, 1981). Partitioning and quantification of these primary controls have allowed modern and historical coral isotope-based reconstruction of highl y accurate (e.g., up to .5 C0 ) sea-surface temperature records (Fairbanks and Dodge, 1979; Dunbar and Wellington, 1981; Patzold, 1 984; McConnaughey, 1989a), regional precip itation and temperature perturbations in response to ENSO events (Carriquiry et al., 1988; Cole and Fairbanks, 1990; Shen et al., 1992 ; Carriquiry et al., 1994), the annual latitudinal oscillation of the InterTropical Convergence Zone (ITCZ) (Linsley et al., 1994) and other regional climate histories and phenomena.
6 Carbon Isotopes Controls on the carbon isotopic composition (o13C) of coral aragonite include (1) the o13C of ambient seawater L.C02 (Weber and Woodhead, 1971; Kroopnick, 1974; Nozaki et al., 1978; Aharon, 1991), (2) bicarbonate exchange between ambient seawater and coral calcifying tissue, (3) the aforementioned kinetic effect related to phylogeny and calcification rate, and (4) the biological modulation of the o13C of the dissolved inorganic carbon (DIC) reservoir from which corals precipitate their skeletons. McConnaughey (1989a) termed the latter control the "metabolic" effect-a modulation of skeletal o13C by the preferential addition or removal of 12C02 in the calcification region via coral and zooxanthellae respiration or zooxanthellae photosynthesis, respectively. Exact partitioning of these controls upon skeletal o13C values is difficult given the variable importance of different metabolic processes and pathways in the coral-zooxanthellae symbiosis (Goreau, 1977; Swart, 1983; Muscatine et al., 1989). However, the photosynthetic activity of zooxanthellae appears a dominant control upon skeletal o13C in many coral-isotope records based on the strong positive correlation between incident solar irradiance and skeletal o13C (i.e., Goreau, 1977; Fairbanks and Dodge 1979; Swart, 1983; Muscatin e e t al., 1989 ; McConnaughey, 1989a; Coleand Fairbanks, 1990). Covariance of Oxygen and Carbon Isotope in Coral Records Different climat e r eg imes can produce coral o18Q and o13C record s with positive, absent, or negative covariance (Figure 2). A common pattern is the positive correlation of coral o18Q and o13C values corresponding to annual
Figure 2. + + + + + c. + Generalized Covariance Patterns of Modern Coral-Isotope Records from Various Climate Regimes 7
8 cycles of warm/ cloudy and cool/ clear climate conditions (Figure 2a). As SST increases during the warmer months, there is concurrent to slightly lagged increase in cloudiness that decreases incident solar irradiance. This decrease in incident solar irradiance lowers the overall photosynthesis rate of zooxanthellae endosymbionts (Barnes and Taylor, 1973), resulting in a net decrease in the o13C of the dissolved inorganic carbon (DIC) calcification reservoir. As SST cools and clearer skies return, the zooxanthellae photosynthesis increases, enriching the o13C of the DIC calcification reservoir by preferentially removing and fixing 12C. Thus, the temperature-dependent o18Q and photosynthesis-dependent ol3C are positively correlated. These warm/ cloudy and cool/ clear seasonality regimes have been linked to a positive coral-isotope covariance in numerous locations including the Philippines (Patzold 1984), the Florida Keys (Leder et al., 1991), Bermuda (Fairbanks and Dodge, 1979), and the Pacific side of southern Central America (W e llington and Dunbar, submitted). In contrast, M cConnaughey (1989a) has reported that positive o18Q and o13 C correlations may result under warm/ clear and cool/ cloudy seasonal conditions. In these cases, the coral is light-saturated year-round and photosynthesis is actually reduced during the warm/ clear season due to bleaching, photoinhibition, or expulsion of zooxanthellae (Glynn, 1983) R e latively constant incident solar irradiance may result in an absent to very w eak corre l ation between coral o18Q and ol3C values (Figure 2b). This constant incident solar irradiance produces a consistent depletion in skeletal o13C throughout th e year via photosynth es is, with annual SST cycles driving the annual o18Q cycle. Such isotopic patterns have been reported for the Galapagos Islands (Wellington and Dunbar, submitted) and the Great Barrier Reef (Aharon, 1991).
9 Negative covariance in coral oi8Q and o13C values are known to occur in two distinct climate regimes (Figure 2c). The first consists of warm/ clear and cool/ cloudy seasons that produce coral-isotope records that are 180 out of phase. This pattern has been documented in the Galapagos Islands (McConnaughey, 1989a; Shen et al., 1992), the Red Sea (Klein et al., 1992), the Pacific coast of Panama (Carriquiry et al., 1994) Jamaica (Fairbanks and Dodge, 1979), and Barbados (Fairbanks and Dodge, 1979) The second climate regime identified to produce negative coral oi8Q and o13C covariance is characterized by minimum SST variability and maximum o180seawater variability, as modulated by rainfall and evaporation. The amount and intensity of rainfall is strongly correlated with cloudiness, producing a concurrent decrease in zooxa nthellate photosynthesis and coral o13C. An excellent example of this relationship is documented by Cole and Fairbanks (1990) in their study of ENSO variability at Tarawa Atoll in the western equatoria l Pacific.
10 III. ENVIRONMENTAL AND GEOLOGIC SETTING Modem Oceanography and Climate Cyclonic ocean circulation of Caribbean Current waters dominates within the southwestern Colombian Basin (Brucks, 1971). Coastal waters off Puerto Limon Costa Rica (10 00'N, 83 02'W), l ess than three k m from our study s it e, have a mean annual SST of 27 1 1.2 C with maxima during June and September-October (1949-1970; NOAA). Annual SST range s vary from 0 8 to 4.4 c o with a mean annual range of 2.4.1 C0. Region al southwestern Colombian Basin (9-11 N; 81-83 W) SST seasonality from 1940 to 1 990 i s sim il a r to that of coastal waters, with slightly warmer temperatures a n d decreased annual SST r a n ge (COADS; Woodruff et al. 1987) (Figure 3a). Seasonal precipitation and sea-su rf ace salinit y (SSS) of the r eg ion a r e grea tly affected by the ITCZ and winte r n orther n fronts. Th e ITCZ's n arrow low-pressure zone of d ense cloud cover and high r ainfa ll r eaches its northernmost position in July (Chahine 1992)-the middle of Costa Ri ca's summe r w e t season and month of minimum summe r salinity in coastal waters (Figure 3 b) In winter months, a n eq u atorward shift and strengthening of the n ortheast trad e winds drive a ir masses ac ro ss the isthmu s (Riehl, 1979; G l ynn, 1 983), r es ultin g in a winte r r ainfa ll maxima and salinity minima. Mean annua l SSS for coas ta l wa t ers off Puerto Limon i s 33.0 9%o with a range of 3.6%o (1949-1970; NOAA). Annual SSS ranges va r y from 1.8 to 8.5%o with a mean annual r a n ge of 4.6 1.7%o.
Figure 3. 301,--------------------------------------29 SST 28 (oC) 27 26 35 34 33 sss (%o) 32 31 A. Jan B. -----Puerto Limon Mean SST( C) ---.--SW Columbian Basin Mean SST (0C) Feb ar Apr May Jun Jul Aug Sept Oct Nov Dec Montb ----o-Puerto Limon Mean SSS (%o) ----.....-Puerto Limon Mean Precipitation (mm) 50 40 3 0 20 10 Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Month PPT (mm) A Mean Monthly SST for Puerto Limon and Southwestern Colombian Basin (9-11 N, 81-83 W). Local and regional SST's are hi g hly in-phase Coastal waters 1.5 c o cooler than regional waters during any given month. 11 B. Mean Monthly SSS and Rainfall for Puerto Limon. SSS has a relative minima during July (northernmost ITCZ position) and a greater relative minima in December, r elated to strong northeast trade winds and associated weather fronts moving over the isthmus.
12 Geologic and Biologic Setting Along the Caribbean coast of Costa Rica and Panama, an assemblage of shallow-marine Neogene deposits is exposed by backarc thrust faulting and tectonic uplift that resulted from convergent zone tectonism (Dengo and Case, 1990). Coates et al. (1992), building extensively upon work by Taylor (1975), Cassel (1986), and Cassel and Sen Gupta (1989a; 1989b), combined the Neogene deposits around Limon, Costa Rica, into the Limon Group, consisting of the Uscari, Rio Banano, and Moin Formations (Figure 1). The Moin Formation, youngest of the group, is a pervasively bioturbated, fossiliferous, shallow-marine succession of alternating blue-gray silty claystone and volcaniclastic litharenite and is exposed over -50 km2 with a maximum thickness of-200m (Taylor, 1975; Cassell and Sen Gupta, 1989a; Coates et al., 1992). The age of the Moin Formation is only broadly known from biostratigraphy as Plio-Pleistocene at the uppe r and lowermost exposures (Coates et al., 1992). Reefal biostromes are occasionally present in the upper portion of the formation in the Lomas del Mar section. The section's ostracode assemblage is analogous to modern carbonate platform assemblages (Borne and Cronin pers. comm., 1994). Modern Siderastrea siderea inhabits a wide spectrum of reef environments to d epths of over 35m. The highest abundance is in shallow ( < 20 m), low-relief, marginally developed reef communities (Goreau, 1959; Goreau and W e lls, 1967; Burns, 1985 ; Wheaton and Jaap, 1988) The genus appears more tolerant of turbidity (Hubbard and Pocock 1972) and salinity fluctuations (i.e., up to %o; Muthiga and Szmant 1987) than typical zooxanthellate cor a ls.
13 V. METHODS The Siderastrea siderea coral was collected in life position from a reefal biostrome of the Lomas del Mar Section of the Moin Formation (STRI Locality AB-93-71 P4/P5). The coral was cut to a 3 mm-thick slab parallel to the growth axis. X-ray diffraction analyses verified the retention of primary aragonite and absence of secondary calcite in the sampled area Two aragonite-powder samples were analyzed for 87Sr j86Sr values by thermal ionization mass spectrometry with a Finnigan-MAT 261 mass spectrometer at Brown University (see Clemens et al., 1993 for complete procedure). The long-term average of strontium-isotope standard NB5-987, as measured at Brown University, was 0.710257 0.000014 (2 a). X-radiography of the coral slab (60 kV, 3 rnA, 25 sec) revealed subtle high-and low-density band couplets with diffuse contacts (Figure 4). Contiguously routed samples (0.5mm interval) were collected from a single corallite using a micro-drill (1.0mm bit diame ter) mounted on a computer-assisted triaxial sampler. Oxygen-and carbon-isotope analyses of the routed corallite-powder samples were performed at the University of Michigan. Samples were vacuum roasted at 200 oc for one hour and reacted with anhydrous phosphoric acid at 75 C in individual reaction vessels of a CarboKiel system mated to a Finnigan-MAT 251 mass spectrometer. Isotopic values are reported r e lative to the Chicago PDB standard. Standard deviation (1 a) was monitored by daily analyses of NBS-20 powde red calcite standard and was less than 0.08 %o for 8180 values and 0.04 %o for 813C values.
Figure 4. 14 Positive Print from X-Radiograph (60 kV 3 rnA 25 sec) of the Fossil Siderastrea siderea Coral. There are 28 high-and lowdensity band couplets in the sampled corallite section, although contrast is poor in some r eg ions and contacts are fairly diffuse.
15 V. RESULTS Strontium-isotope analyses produced coralline 87Sr j86Sr values of 0.709141.00008 and 0.709145 00008 (1 cr) (Clemens, 1994, pers. comm.). The stable-isotope record for the Siderastr e a siderea coral is presented in Figure 5 and Appendix 1. Mean 8180 and 813C values are 2.13.27 %o (PDB) and -2.09 0.23 %o (PDB), respectively. Deviation in all isotope mean values is standard deviation of population (StDevP) Total range in 8180 is 1.18 %o, while 813C total range is nearly three times as large at 3.35 %o. To partition sub-and super-annual variation and patterns, two annualscale sclerochronologies were established inde p endently using high-density bands (c.f., Fairbanks and Dodge, 1979 ; (Winter et al., 1991)) and 813C relative maxima (c.f., Cole and Fairbanks, 1990; Shen et al., 1992; Carriquiry et al., 1994). The two sclerochronologies are esse ntially identical in terms of annual scale statistica l moments (Table 1) as well as mean annual isotope time-series patterns (Figure 6). A total of 28 years of growth in the sampled corallite is indicat e d by both highdensity bands and 813C relative maxima; however, they are synchronous in only 9 of the 28 years (31 %) with high-density bands lead 813C peaks by a mean of 0.10 1.54 mm. While the overall similarity between the two sclerochronologies strengthens the inference that both represent annual cycles, the 813C-bas e d sclerochronology is used for this study because of its strong correlation with 8180 discussed later and the diffuse and subtle nature of the density banding.
-0.5 -1.0 -1.5 -2.0 8 -2.5 -3.0 -3 5 -4 0 4 5 10 Figure 5. 2 !il!il !il s 9 6 '' -.: I t l ....-----------..;;. : : : High Dens i ty Band I -- ---o13c !il Missing Valu e 20 30 40 :..;. : : ::"' : : ::. :: . . mm from Base s s s ., : " , : " I .. 50 60 70 ()18Q and ()13C Profiles for Fossil Siderastrea sider e a Coral. Numb ers above the ()13C reco r d i n dica t e end of the given year based on the ()13C re l ative maxima scleroc h ronology Note positive covariance of stable isotopes and the pronounced isotopic shift during year 14 to more negative values increased growth rates and extension of ()13C to more negative relative minima. 16
Figure 6. Mean A. --e--1>180 Annual Mean by S1 3C Yea r -1.50 ------1>180 Annual Mean by HD Year ; 2 .50 15 20 25 30 Year -1.50 -2.50 Year 17 Comparison of (A ) Mean Annual ()18Q Va l ues and (B. ) Mean Annual ()13C Values Derived from ()13C and High-Density Band Sclerochronologies. The similarity of the two sclerochronologies support an annual periodicity for each.
18 Table 1. Statistics for High-Density and ()13C Sclerochronologies. High-density Band Sclerochronology Mean StDevP Range Mean StDevP Range Growth Annual Annual Annual Annual Annual Annual Rate (%o) (%o) olSQ (%o) o 13 C (%o) o13C (%o) o13C (%o) (mm) Annual Mean -2.09.23 0.12 0 .07 0.29.16 -1. 99 0 .50 0.40.20 1.02.56 2 .05 0 4 3 (n=28 yrs) ol3C Sclerochronology Mean StDevP Range Mean StDevP Range Growth Annual Annual Annual Annual Annual Annual Rate olSQ (%o) olSQ (%o) o18Q (%o) o 1 3 C (%o) o 13 C (%o) o 13 C (%o) (mm) Annual Mean -2 .10 0.24 0.11 06 0.27.16 1 .99.52 0 41 .20 1.03 0.54 2 .07.53 (n=28 yrs) Based on the o13C sclerochronology, m ea n annual growth rate for the coral is 2 .1 .5 mm/year and range s from 1.0 to 3.5 mm/year. At a 0.5 mm sampling int e rv a l, thi s growth rate trans l ates to an average of 4.11.1 samples/year with an actual sampling of 2 t o 7 per year. For the 28 years of growth, m ea n annual range in ()18Q and ol3C are 0.27 .16 %o (PDB) and 1.030.54 %o (PDB) re s p ec tively. Annual r anges in 81 8 0 and o13C vary from 0.08 to 0 .87 %o and 0.10 to 2.10 %o, r espective ly.
19 VI. DISCUSSION Age of the Lomas del Mar Section Prior to this study, the age range of the Moin Formation was best constrained by Coates et al. (1992) in their extensive reconnaissance and synthesis of the nearshore Neogene record of Costa Rica and Panama. Coates et al. (1992) determined biostratigraphic ages of :s; 1.7 Ma and;::: 2.4 to Ma for the uppermost and lowermost exposures of the Moin Formation, respectively. Within the Lomas del Mar section of the upper Moin Formation, biostratigraphic analysis of sediment surrounding the fossil coral confirms the planktonic foraminifera-based age of :s; 1.7 Ma (Dowsett, pers. comm., 1993) while calcareous nannofossil data (Wuchang, pers. comm., 1994) reduces the age range of Lomas d e l Mar to between 0.45 to 0 9 Ma. To assess these biostratigraphic data and improve overall temporal resolution within the Moin Formation, the 87Sr j86Sr values of the fossil Sid erastrea siderea specimen were compared to a high-resolution Neogene 87Sr j86Sr seawater curve from the western Pacific Ocean (ODP Leg 130 ; Shackleton et al., in prep.) The coral 87Sr j86S r values indicate an apparent age of 1.03 Ma, with a minimum age of 0.38 Ma and maximum age of 1.63 Ma (Clemens, 1994, pers. comm.). The apparent 87Sr j86S r age is close to the nannofossil-age range, although the uncertainty is larger Hence, our isotopebased age estimate for the formation is consistent with biostratigraphy-based estimates (Figure 7).
20 Worker Data Coates !BEI!lliliJlllll!lliliJlllll!lliliJlllll!lliliJlllll!llill!lllll S 1.7 Ma (uppermost exposure) PF & NF r:: eta!. (lowermost exposure) 2: 2.4 Ma NF ... (1992) PF Figure 7 Dowsett PF Wuchang NF This Study 87 /86Sr 0 1 2 3 Ma Age Data on the Moin Formation and Lomas del Mar section Strontium-isotope age is based on comparison of coral 87Sr j86Sr values to a high-resolution Neogene seawater strontium-isotope curve (Shackleton et al., in prep.). Nannofossil dating by Wuchang (pers. comm., 1994) provides the narrowest age range for Lomas del Mar and is assumed the section's age in this study (NF = nannofossil, PF = planktonic foraminifera).
21 Inferred Paleoenvironment from Fossil Coral Isotopes The isotopic record of the Siderastrea siderea scleractinian coral contains substantial inter-and intra-annual variability. Because the sampled corallite has not been diagenetically altered, this variability represents ambient physico-chemical seawater conditions modified by coral physiology and biomineralization. Although unique partitioning and quantification of the myriad controls is difficult, constraints may be placed on ambient paleoenvironmental conditions by using modern coral-isotope and climate relationships as an interpretive guide. Variations in growth rate can control the observed coral-isotope record in two ways-attenuation of seasonal amplitude through sampling techniques and variations in isotopic disequilibrium through kinetic effects. We examined these sources of variability prior to inferring ambient environmental conditions. Sampling Effects Contiguous sampling of coral records reduces annual isotopic ranges through the averaging of annual extrema with adjacent values. Compared to predicted ranges from concurrent instrumental records, annual isotopic range reductions may vary from negligible (Fairbanks and Dodge, 1979) to over 50 % (Emiliani et al., 1978) depending upon the sampling rate and technique. Carriquiry et al. (1994) calculated a global average sampling rate for coral isotope studies to be 6.8 samples/year, although Aharon (1991) and Delaney et al. (1993) found 4 to 5 samples per year to be adequate to detect seasonality as recorded in coralline chemistry
22 The contiguous 0.5-mm sampling interval used in this study produced 2 to 7 samples/year due to variable annual growth rates. The effect of these various sampling rates on observed annual isotopic ranges were evaluated by comparing the two for each of the 28 years (Figure 8a). Annual ()180 ranges increase slightly with higher annual sampling rates and are highly variable overall (r=0 28). In contrast, annual ()13C ranges exhibit a stronger positive relationship to the number of samples per year (r=0.44). These calculations imply that sampling rate exerts a negligible minor background control on annual ()18Q ranges, while actual annual ()13C ranges may be greatly underestimated at lower sampling rates in this coral record. Kinetic Effects The aragonite of scleractinian corals exhibits significant isotopic disequilibrium in ()180 and ()13C values (Weberand Woodhead, 1972; Land et al., 1975; Swart, 1983; Gonzalez and Lohmann, 1985). This disequilibrium is attributed to kinetic isotope fractionation during C02 hydration and hydroxylation during biomineralization (McConnaughey 1989b). Thus, the amount of isotopic disequilibrium should vary with biomineralization rate, which fluctuates in response to the environment and physiological processes. The coral-isotope record was examined for kinetic effects by comparing annual isotopic mean to growth rate for each of the 28 years (Figure 8b). Growth rate exhibits a negative correlation with both 8 180 (r=0.46) and ()13C (r=0.46), with mean annual ()13C values decreasing at approximately twice the rate of mean annual ()18Q values per mrn increase in annual growth rate. These relationships indicate that kinetic effects have an influence on the absolute isotopic values of this coral record
Figure 8. 3.0,-------------------, 2.5 2.0 s Annual 1.5 Range -0.5 -1.0 -1.5 s"o Annual Mean -2.0 -2.5 -3.0 0.5 A. 2 B. I 0 0 1.0 -9-SUO Annual Range +--013C Annual Range m=0.22 r-:0.44 .. --- .. .... . m=0.04 r-:0.28 3 4 5 7 8 Samples/Year -9-SuO Anouol Meon -o13c Annual Mean I .... m=.21 r-:0.46 m=.45 r-:0.46 1.5 2.0 2.5 3.0 3.5 4 0 Annual Growth Rate (mm) 23 (A ) Sampling Effects and (B.) Kinetic Effects on Annual 8180 and 813C Values. A. Sampling effects. Annual 8180 ranges show a negligible correlation to annual sample density, while annual 813C range is positively correlated to annual sample density. These relationships imply that the sampled annual 8180 range is close to actual annual 8180 range regardless of sampling density, while annual 8130 range is underestimated at all sampling densities. B. Kinetic effects. Mean annual 8180 and 813C values decrease as annual growth rates increase. This relation is consistent with increased kinetic fractionation as growth rate increases.
24 Paleoenvironmental Inferences Fossil-coral 818Q and 813C values show a strong positive correlation (r=0.82) 'for the entire record (Figure 9). This positive correlation is also observed in the annual phase relationship between 818Q and 813C. Annual 818Q maxima coincide with 813C peaks in 15 (60%), and lead 813C maxima by one sample in 5 (20%), of the 25 years with completely sampled maxima curves (excludes years 5, 11, 26, 27 with missing data) Similarly, annual 818Q minima coincide with 813C minima in 19 (76%), and lead by one sample in 3 (12%), of the 25 years with completely sampled minima curves (excludes years 0, 6, 11, 21 with missing data) No robust phase relationships exist between 818Q and density banding. The positive correlation and phase relationships in this fossil-coral record are consistent with many modern coral-isotope records. In these records modern tropical climate regim es consist of warmcloudy and cool-clear seaso nality, where coral 818Q and 813 C values are controlled by SST and photosynthetic response of zooxanthellae to incident solar irradiance, respectively. We are unaware of any modern coral-isotope records from the southwestern colombian Basin to compar e to this fossil coral-isotope record. However, modern coral-isotope covariance for a coral living in Puerto Limon coastal waters may be pre dicted from modern monthly SST, monthly SSS, and 8180seawater-salinity relationships. Craig and Gordon (1965) report an equatorial Atlantic Ocean ,M8180seawat er)/ L1salinity of 0 11. Using the coral 8180-temperature relationship of Druffel (1985) and Craig and Gordon's (1965) 8180seawate r-salinity relationship, mean monthly 8 180 values were predicted for a modern coral using Puerto Limon mean monthly SST and SSS values by the equation:
Figure 9. -1.5 .---------------------,..------------, -2.0 -2.5 m=0.32 r=0.82 0 6>0 0 00 0 0 3 0 t-+--+-+-+-+-+-+-+-+-r->--+--<-t-+-+-+--+-t--+-+--+-++-+--+-+-+-+-+-+-+-+t-+-+-+-+--! -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0. 5 o13c 25 Covariance of Fossil Coral ()18Q and o13C Values Note the strong correlation between the isotopes and that the o13C range is approximately four times greater than that of ()18Q The low ()18Q range is attributed to the opposite effects of SST and SSS changes, interpreted to be in phase, on coral ()180.
26 8180coral = [ SST (C0 ) (-0.22 %o/ l C0)] + [ SSS (%o) (0.11 %o/ 1 %o)] Relative monthly 818Q values were calculated because of the unknown isotopic equilibrium offset for Siderastrea (Figure lOa). The cloudy season for the southwestern Colombian Basin region is from approximately June to November (COADS, 1940-1990). Thus, a modern zooxanthellate scleractinian growing at non-photoinhibition depths is predicted to show a positive correlation in 818Q and 813C val ues (Figure lOb) Two superannual isotopic modes are defined for the fossil coral record and provide important evidence on paleoenvironmental conditions (Table 2). The early portion of the record (Years 2-13) exhibits relatively slow growth rates, high 818Q and 813C values, and a small 813C range; however, in the latter portion of the record (Years 14-25), average growth rate increases by 0.46 mm/year, mean annual 818Q and 813C values decrease by 0.48 and 1.04 %o, and mean annual 813C range increases by 0.96 %o. The phase relationship of 818Q and 813C is consistent in both modes. Years 1 and 26-28 are intermediate in nature between these two modes and may represent transitional periods; however, the record's boundaries preclude verification of this hypothesis. Table 2. Modes of Coral-Isotope Record in 813C Sclerochronology Mean StDevP Range Mean StDevP Range Growth Annual Annual Annual Annual Annual Annual Rate YEAR (%o) (%o) olSQ (%o) o13C (%o) o 13 C (%o) o 13 C (%o) (mm) 1 -2.22 0.32 0.87 -2.16 0.57 1.48 2 .00 2-13 -1.85 0.10 0.24 -1.49 0.25 0.57 1.83 14-25 -2.33 0.11 0.28 -2.53 0.58 1.53 2.29 26-28 -2.11 0.09 0 .21 -1.80 0.30 0.75 2.17
0 06 A. 0.0 .,o "' > .c .. "ii 0.0 c.: "" 1l :a -0.0 0. 0 6+--:-+--::+:--:+----:+---:-:+---::+--:+:--:+---::-+-:--::-i-,---:,:+--=+----l Jan Feb ar Apr May Juo Jul Aug Sept Oct Nov Dec Month + B. + Figure 10. (A.) Predicted ()180 Values and (B.) Stable Isotope Covariance Pattern for Modern Coral Growing in Puerto Limon Coastal Waters. 27 A Predicted relative values of mean monthly coral ()180 based on modern SST and SSS for Puerto Limon Period of maximum cloudiness for the region (June to November) from COADS (1940-1990). Hatched boxes indicate mean isotopic value if 4 samples/year were collected from the coral. B. Hypothesized correlation of ()180 and ()13C of a coral growing in shallow waters off Puerto Limon today ()13C values are inferred from the correlation of decreased incident solar irradiance with decreased ()13C via photosynthetic modulation of DIC calcification pool (McConnaughey, 1989a).
28 This observed shift in the nature of isotopic variability is consistent with an increase in ambient temperature and its effect on coral physiology and skeletal isotopes. An increase in paleotemperature would produce a more negative skeletal ()18Q value and increase coral growth rate (Weil et al. 1981) further promoting depletion in both skeletal ()18Q and ()13C via kinetic effects. In addition, as Coles and Jokiel (1977) have shown, coral-algal respiration increases at a greater rate than algal photosynthesis as ambient temperature increases. Because photosynthesis increases and respiration decreases the ()13C of the DIC calcification reservoir, a lowering of the productivity / respiration ratio would decrease skeletal ()13C. Furthermore, reduced metabolic contribution from endogenous photosynthesis would be compensated by an increased intake of exogenous dissolved or particulate organic carbon, each with highly depleted ()13C signatures The effects of these processes are consistent with the superannual shift in the fossil coral-isotope record Quantitative Estimates of SST and SSS from Coral ()18Q Robust and accurate quantitative reconstruction of sea-surface seasonality from coral-isotope records is controlled largely by the ability to constrain the relative magnitudes and phase relationships of annual SST and ()180seawater cycles In historical studies, coral-isotope records are often calibrated to local instrumental records, and the interpretation of older isotopic data are guided by these earlier calibrations (Shen et al. 1992; Dunbar et al., 1994). In some regions, negligibl e seasonal changes in either SST or ()180seawater allow a direct calculation of the other parameter (i.e ., Cole and Fairbanks, 1990).
29 In this study, a unique solution for absolute sea-surface conditions for the middle Pleistocene coral isotope record is precluded because the following parameters are unknown: 1) annual relative magnitude and phase relationships between paleotemperature and c180paleoseawater, 2) the isotopic equilibrium offset for Siderastrea, and 3) the global mean c180seawater (icevolume effect) at time of coral growth as the age range contains eight middle Pleistocene glacial-interglacial cycles (Stages 12-27) (Williams et al., 1988). Despite these unknowns, relative ranges in annual paleotemperature and paleosalinity may be inferred from the fossil coral using modern SST and SSS relations by assuming that these relations were the same in the past. Annual clBQ Range as Function of Either Temperature or Salinity A direct application of the c180coraHemperature equation to the fossil annual c180coral range attributes all isotopic variability to SST changes, with the c180seawater and SSS assumed constant. Application of the relationship (e.g., -0. 22 %o per 1 C0 ; Weber and Woodhead, 1972; Druffel, 1985) to the 28-year c18Q record produced a mean annual SST range roughly half that of modern SST (Table 3). Minimum and maximum annual SST ranges for the 28-year c18Q record were within .5 co of minimum and maximum annual SST ranges in the modern. If annual SST was constant, the fossil annual c180coral range may be directly interpreted as annual SSS range through the application of modern c180seawater-salinity relationships. Application of the equatorial Atlantic relationship (e.g., 0 11; Craig and Gordon, 1965) to the annual c180coral ranges produced lower mean, minimum, and maximum annual SSS ranges than modern annual SSS ranges of Puerto Limon (Table 3).
Table 3. Middle Pleistocene Annual Coral 818Q Range Interpreted as Either SST or SSS, Exclusively. ALL I Mean Annual Minimum Annual Maximum Annual TEMPERATURE SSTRan e SSTRan e SSTRan e Mid-Pleistocene II 1.2.7 c o 0.4C0 4 .oco Modem 2 4 1.1 c o 0 8CO 4 .4C0 ALL I Mean Annual Minimum Annual Maximum Annual SALINITY SSSRan e SSS Ran e SSS Ran e Mid-Pleistocene 2.4.4 % o 0.7 %o 7.9 %o Modem 4.6.9 %o 1.8 %o 8.5 %o 30 Exclusive interpretation of the coral 818Q record as either temperature or salinity yield paleoseasonality estimates lower than modern conditions. It is unlikely, however, that either annual SST or SSS was constant in the middle Pleistocene A more parsimonious approach is to consider the simultaneous effects of seasonal SST and SSS cycles on the coral 818Q record. Annual 8180 Range as Function of Both Temperature and Salinity Annual 818Q range in the fossil-coral record may be partitioned into annual SST and SSS ranges using modern sea-surface seasonality relationships from Puerto Limon. Explicit in this approach is the assumption that relative magnitudes of SST and SSS, and their phase relationship, were the same in the middle Pleistocene as today. This assumption is supported by the positive 818Q and 813C correlation in fossil coral discussed earlier; however, the very real possibility of middle Pleistocene deviations from modern relations must be appreciated. Modern mean monthly SSS increases by 1.42 %o for each 1 c o increase in modern mean monthly SST (r=0.81; Figure 11) Assigning the annual 818Q
34 Monthly Mean 33 sss 32 25 m=1.42 r=0.81 26 27 Monthly Mean SST(o q 28 29 31 Figure 11. Relation of SST and SSS in Modern Coastal Waters Off Puerto Limon Note that SSS range is nearly three times greater than annual SST range and the r value of 0 82. This modern relationship is used in the estimation of mid-Pleistocene annual SST and SSS ranges from annual ranges in coral ol80.
32 range in the fossil coral-isotope record a coefficient of one, the annual ranges in ancient SST and SSS were back-calculated using the modern SSS:SST slope value (m=1.42) and the respective relationships of SST and SSS with coral 818Q values by the following equations : Annual SST Range = 2.00 [(Annual 8180coral Range) (1 c o I -0.22 %o)] Annual SSS Range= 2.84 *[(Annual 8180coral Range) (1 %o/ 0.11 %o)] These decoupled ancient SST and SSS equations produce a mean annual paleotemperature range nearly identical to modern mean annual SST range for Puerto Limon (Table 4). Paleosalinity range estimates are 2 2 %o larger than modern SSS range. While this approach does provide m_athernatically rigorous results caution in their interpretation is prudent given the estimate's high sensitivity to middle Pleistocene deviations from modern conditions. Thus, we view these data as a supporting element, not a cornerstone, for the interpretation that the fossil Siderastr e a siderea grew in conditions generally similar to those of modern local waters. Table 4. Comparison of Middle Pleistocene Seasonality Calculated using Modern Relative Magnitudes and Phase Relationships for Temperature and Salinity from Puerto Limon Mean Annual Minimum Annual Maximum Annual TEMPERATURE SSTRan e SSTRan e SSTRan e Mid-Pleistocene 2.4.4 c o 1.4 c o 7.2C0 Modem 2.4.1 c o 0.8 c o 4 .4C0 Mean Annual Minimum Annual SSS Maximum Annual SA UNITY SSSRan e Ran e SSS Ran e Mid-Pleistocene 6.8.1 %o 2.1 %o 22.6 %o Modem 4.6.9 %o 1.8 %o 8.5 %o
33 VI. CONCLUSIONS 1. Coral aragonite 87/86Sr values indicate a mean age of 1.03.65 Ma (2 cr) for the Lomas del Mar section. This age estimate is larger than, but consistent with, nannofossil-based age range of 0 45 to 0.9 Ma for the Lomas del Mar section. 2. The positive covariance between (5180 and (513C in the fossil Siderastrea siderea coral is consistent with modern coral-isotope records from regions with alternating seasons of warm SST-high cloudiness and cool SST -clear skies. The fossil coral stable-isotope record is also consistent with the predicted stable-isotope covariance pattern for a coral growing in modern Puerto Limon coastal waters. Hence, climate patterns during the coral's life history are interpreted as being similar to modern Southwestern Colombian shallow-marine waters. 3. Quantitative estimates of the middle Pleistocene annual SST and SSS ranges, based on modern SST-SSS relative magnitudes, SST-SSS phase relations and equatorial Atlantic Ocean 8180seawat e r-salinity relationships, are similar to modern SST and SSS ranges. Mean modern and middle Pleistocene SST ranges are identical, while mean middle Pleistocene SSS range is 2.2 %o greater than modern SSS. However, these estimates are speculative as the calculations are highly sensitive to middle Pleistocene deviations from modern SST, SSS, and (5180seawater-salinity relationships
34 VII. REFERENCES Aharon, P. (1991). Recorders of reef environment histories: stable isotopes in corals, giant clams, and calcareous algae. Coral Reefs, 10, 71-90. Barnes, b. J., & Taylor, D. L. (1973). In situ studies of calcification and photosynthetic carbon fixation in the coral Monastrea annularis Helgolander wiss. Meeresunters, 24, 284-291. Bennet, K. D. (1990) Milankovich cycles and their effects on species in ecological and evolutionary time. Paleobiology, 16 11-21. Brucks, J. T (1971). Currents of the Caribbean and adjacent regions as deduced from drift-bottle studies. Bulletin of Marine Sciences, 21, 455-465. Buddemeier, R. W Margos, J E., & Knutson, D. W. (1974). Radiographic studies of reef coral exoskeletons: rates and patterns of coral growth. Journal of Experimental Marine Biology and Ecology, 14 177-200. Burns, T. P. (1985). Hard-coral distribution and cold-water disturbances in South Florida: variation with depth and location. Coral Reefs, 4, 117124. Carriquiry, J.D., Risk, M. J., & Schwarcz, H. P. (1988). Timing and temperature record from stable isotopes of the 1982-83 El Nino warming event in eastern Pacific corals. Palaios, 3 359-364. Carriquiry, J.D., Risk, M. ]., & Schwarcz, H. P. (1994). Stabl e isotop e geochemistry of corals from Costa Rica as proxy indicator of El Nino Southern Oscillation (ENSO). Geochimica et Cosmochimica Acta, 58, 335-353. Cassell, D. T. (1986). Neogene foraminifera of the Limon Basin of Costa Rica. Unpublished Ph.D. Dissertation Louisiana State University. Cassell, D. T., & Sen Gupta, B (1989a). Pliocene foraminifera and environments, Limon Basin of Co s ta Rica. Journal of Paleontology, 63 146-157 Cassell, D. T., & Sen Gupta, B. (1989b) Foraminiferal stratigraphy and paleoenvironments of the Tertiary Uscari Formation, Limon Basin, Costa Rica. Journal of Foraminiferal Research, 19, 52-71.
35 Chahine, M. T. (1992). The hydrological cycle and its influence on climate. Nature, 359, 373-380. Clemens, S. C., Farrell, J W., & Gromet, L. P. (1993). Synchronous changes in seawater strontium isotope composition and global climate. Nature, 363, 607-610. Coates, A. G., Jackson, J. B. C., Collins, L. S., Cronin, T M., Dowsett, H. J., Bybell, L. M., Jung, P., & Obando, J. A. (1992). Closure of the Isthmus of Panama: the near-shore marine record of Costa Rica and western Panama. Geological Society of America Bulletin, 104 814-828. Cole, J E & Fairbanks, R. G. (1990) The Southern Oscillation recorded in the o18o of corals from Tarawa Atoll. Paleoceanography, 5 669-683. Coles, S. L., & Jokiel, P. L. (1977). Effects of temperature on photosynthesis and respiration in hermatypic corals. Marine Biology, 43, 209-216. Craig, H., & Gordon, L. I. (1965) Isotopic oceanography: Deuterium and oxygen-18 variations in the ocean and the marine atmosphere. In Proceedings of the Symposium on Marine Geochemistry, 3 (pp. 277374). University of Rhode Island: University of Rhode Island. Crowley, T J., & North, G R. (1991). Paleoclimatology. New York, NY: Oxford University Press. Delaney, M. L., Linn, L. J., & Druffel, E. R. M. (1993). Seasonal cycles of manganese and cadmium in coral from the Galapagos Islands. Geochimica et Cosmochimica Acta, 57, 347-354. Dengo, G., & Case, J. E. (1990). The Geology of North America, vol. H, The Caribbean Region. Boulder, CO: Geological Society of America. Dodge, R. E & Vaisnys, J. R. (1980). Skeletal growth chronologies of recent and fossil corals. In D. C. Rhoads & R. A. Lutz (Eds.), Skeletal Growth of Aquatic Organisms (pp. 493-517). Plenum Press. Druffel E R. M (1985). Detection of the El Nino and decade time scale variations of the sea surface temperature from banded coral records: implications for the carbon dioxide cycle In E. T Sundquist & W. C. Broecker (Eds ), The Carbon Cycle and Atmospheric C02 : Natural Variations Archean to Present (pp. 111-122). Washington, D.C.: American Geophysical Union. Dunbar, R. B ., & Cole, J (1993). Coral Records of Ocean-Atmosphere Variability (Special Report No. 10) Boulder, CO: NOAA Climate and Global Change Program.
36 Dunbar, R. B., & Wellington, G. M. (1981). Stable isotopes in a branching coral monitor seasonal temperature variations. Nature, 293, 453-455. Dunbar, R. B., Wellington, G. M., Colgan M W & Glynn, P. W. (1994). Eastern Pacific sea surface temperatures since 1600 A.D.: the 8180 record of climate variability in Galapagos corals. Paleoceanography, 9 291-315 Emiliani, C., Hudson, J H., Shinn, E. A & George, R. Y (1978). Oxygen and carbon isotopic growth record in reef coral from the Florida Keys and a deep sea coral from the Blake Plateau. Science, 202 627-629. Epstein, S ., Buchsbaum, R. Lowenstam, H. A., & Urey, H C. (1953). Revised carbonate-water isotopic temperature scale Geological Society of America Bulletin, 64, 1315-1326 Epstein S., & Mayeda, T (1953) Variation of 018 content of waters from natural sources. Geochimica et Cosmochimica Acta, 4, 213-224. Fairbanks, R. G. & Dodge, R. E (1979) Annual periodicity of the 0-18/0-16 and C-13 / C-12 ratios in the coral Montastrea annularis. Geochimica et Cosmochimica Acta, 43, 1009-1020. Glynn, P W. (1983). Extensive bleaching and death of reef corals on the Pacific coast of Panama. Environmental Conservation, 10, 149-154. Gonzalez, L.A. & Lohmann, K. C. (1985). Carbon and o xyge n isotopic composition of Holocene reefal carbonates Geology, 13 811813. Goreau, T. F. (1959). The ecology of Jamaican coral reefs. 1. species composition and zonation. Ecology, 40, 67-90 Goreau, T. F., & Wells, J. W. (1967) The shallow-water Scleractinia of Jamaica : revised list of species and their vertical distribution range. Bulletin of Marine Science, 17 442-453 Goreau, T J. (1977a) Carbon metabolism in calcifying and photosynthetic organisms: theoretical models based on stable isotop e data. In Third International Coral Reef Symposium, (pp. 395 -4 01). Miami, FL: University of Miami. Goreau, T J. (1977b) Coral skeletal chemi s try: physiological and environmental regulation of stable isotopes and trace metals in Montastrea annularis. Proceedings of the Royal Society of London, 196 291-3 15
Guilderson, T P., Fairbanks, R. J & Rubenstone, J. L. (1993) Tropical temperature variations since 20,000 years ago: modulating interhemispheric climate change. Science, 263, 663-665 37 Hubbard, J. A. E B & Pocock, Y. P (1972). Sediment rejection by recent scleractinian corals: a key to paleo-environmental reconstruction. Geol. Rundschau, 61,598-626 Karl, T R., Kukla, G Razuvayev, V. N Changery, M J., Quayle, R. G., Heim, R. R., Easterling, D R., & Fu, C. B. (1991). Global warming: evidence for asymmetric diurnal temperature change Geophysical Research Letter, 18, 2253-2256. Klein, R., Patzold, J., Wefer, G., & Loya, Y. (1992) Seasonal variations in the stable isotopic composition and the skeletal density pattern of the coral Porites lobata (Gulf of Eilat, Red Sea) Marine Biology, 112, 259-263. Knutson, D. W., Buddemeier, R. W & Smith, S V (1972) Coral chronometers: seasonal growth bands in reef corals. Science, 177, 270272. Kroopnick, P (1974). Correlations between 13C and .EC02 in surface waters and atmospheric C02 Earth and Planetary Science Letters, 22 397 403 Land, L. S., Lang, J C., & Barnes, D. J. (1975). Extension rate: a primary control on the isotopic composition of West Indian (Jamaican) scleractinian reef coral skeletons Marine Biology 33, 221-233. Leder, J J Szmant, A. M., & Swart, P K. (1991) The effects of prolonged ''bleaching" on skeletal banding and stable isotopic composition in Montastrea annularis. Preliminary observations. Coral Reefs, 10, 19-27. Linsley, B K., Dunbar, R. B., Wellington, G. M., & Mucciarone, D. A. (1994). A coral based reconstruction of intertropical convergence zone variability over Central America since 1707 Journal of Geophysical Research. McConnaughey, T. (1989a) 13c and 180 isotopic disequilibrium in biological carbonates : I. patterns. Geochimica et Cosmochimica Acta, 53, 151-162. McConnaughey, T. (1989b) 13c and 180 isotopic disequilibrium in biological carbonates: II. in vitro simulation of kinetic isotope effects. Geochimica et Cosmochimica Acta, 53, 163-171. Muscatine, L., Porter, J. W & Kaplan, I. R. (1989). Resource partitioning by reef corals as determined from stable isotope composition. I. 813c of zoxanthellae and animal tissue vs depth. Marine Biology, 100, 185-193.
38 Muthiga, N. A., & Szmant, A.M. (1987). The effects of salinity stress on the rates of aerobic respiration and photosynthesis in the hermatypic coral Siderastrea siderea. Biological Bulletin, 173, 539-551. Nozaki, Y., Rye, D. M., Turekian, K. K., & Dodge, R. E. (1978) A 200 year record of carbon-13 and carbon-14 variations in a Bermuda coral. Geophysical Research Letters, 5, 825-828 Patzold, J. (1984). Growth rhythms recorded in stable isotopes and density bands in the reef coral Porites lobata (Cebu, Philippines). Coral Reefs, 3, 87-90. Riehl, H. (1979) Climate and weather in the tropics. New York, NY: Academic Press Roulier, L. M., & Quinn, T. M. (in review) Seasonalto decadal-scale climatic variability in southwest Florida during the Middle Pliocene warm interval: inferences from a coralline stable isotope record Paleoceanography. Shen, G. T ., Cole, J. E., Lea, D. W., Linn, L. J., McConnaughey, T. A., & Fairbanks, R. G. (1992) Surface ocean variability at Galapagos from 1936-1982: calibration of geochemical tracers in corals Paleoceanography, 7, 563-588. Swart, P K. (1983). Carbon and oxygen isotope fractionation in scleractinian corals: a review. Earth Science Reviews 19, 51-80 Taylor, G. (1975) The geology of the Limon area of Costa Rica. Unpublished Ph.D. Dissertation, Louisiana State Univeristy. Weber, J N., & Woodhead, P M J (1971). Diurnal variations in the isotopic composition of dissolved inorganic carbon in seawater from coral reef environments. Geochimica et Cosmochimica Acta, 35, 891-902. Weber, J. N & Woodhead, P.M. J. (1972). Temperature dependence of Oxygen-18 concentration in reef coral carbonates Journal of Geophysical Research, 77, 463-473 Wefer G & Berger, W H. (1991). Isotope paleontology: growth and composition of extant calcareous species. Marine Geology, 100 207-248. Weil, S. M ., Buddemeier, R. W., Smith, S V & Kroopnick, P. M. (1981). The stable isotopic composition of coral skeletons: control by environmental variables. Geochimica et Cosmochimica Acta, 45, 11471153
Wellington, G. M., & Dunbar, R. B. (submitted) Regional variation in the stable isotopic signature of ENSO events in reef corals in the eastern tropical Pacific. Coral Reefs Wheaton, J. L., & Jaap, W. C. (1988) Corals and other prominent benthic cnidaria of Looe Key National Marine Sanctuary, Florida No. 43) Florida Department of Natural Resources Williams, D. F., Thunell, R. C. Tappa, E., Rio, D., & Raffi, I. (1988). 39 Chronology of the Pleistocene oxygen isotope record : 0-1. 88 m.y. B.P .. Palaeogeography, Palaeoclimatology, Palaeoecology 64, 221-240 Winter, A. Goenaga C., & Maul G. A. (1991). Carbon and oxygen isotope time series from an 18-year Caribbean reef coral. Journal of Geophysical Research, 16, 673-678. Woodruff, S. D., Slutz, R. J ., Jenne R. L., & Steurer, P M (1987) A comprehensive ocean-atmospher e data set. Bulletin American Meteorological Society 68, 1239-1250. Yurtsever, Y., & Gat, J. R. (1981). Atmospheric waters In J. R. Gat & R. Gonfiantini (Eds ) Stable Isotope Hydrology: Deuterium and Oxygen18 in the Water Cycle, Technical Report Series No. 210 (pp. 10 3 -142) Vienna: International Atomic Energy Agenc y.
40 VIII. APPENDIX
41 APPENDI X 1 ISOTOPIC DATA FOR MID D LE PLEISTOCENE CORAL mm H igh Sample from Density 818() error 813C error # Base Band 20 1 0.0 2.28 0 .03 -2.08 0.02 21 10.5 -2 00 0.03 -1.33 0 04 2 2 11.0 0 -2.19 0 .01 -2 .38 0 .01 23 11.5 2.58 0.01 -2 .68 0.01 24 12 0 -2 38 0 .01 2 .39 0.01 25 12.5 -1.71 0.06 -1. 20 0 .04 26 13.0 1 -1.89 0 04 -1.51 0 02 27 13.5 -1.71 0 .01 0.74 0.02 28 1 4.0 -1.74 0.03 -1.06 0 02 29 14.5 -1.68 0 .02 -1.45 0.04 30 15.0 2 -1.66 0 .06 1 .34 0 02 31 15.5 1.49 0 .05 0.98 0 02 32 16.0 -1.77 0 .03 -1. 32 0.02 33 16 5 -1.56 0.05 1.05 0 03 34 17.0 3 -1.77 0.03 1.07 0.01 35 17.5 -1.63 0 .01 -1.14 0 .04 36 18 0 -2 .01 0 .03 1.81 0 02 37 18.5 4 -2 06 0 .03 -1.93 0 03 38 19.0 -1.95 0 .03 1 92 0.02 39 19.5 -1. 90 0 .03 -1.50 0 0 3 40 20.0 -2 .01 0 .01 1.42 0.02 41 20.5 5 42 21.0 -1.99 0 .03 1.72 0.01 43 21.5 44 22.0 -1.87 0.03 -1.55 0.01 45 22.5 6 -1.79 0.02 -1.00 0 02 46 23.0 -1.94 0 .02 1.65 0.01 47 23 5 -1.86 0 05 1.40 0.01 48 24 0 7 -1.95 0.05 -1.36 0 02 49 24.5 -1.84 0 .04 1.57 0.02 50 25.0 -2 .11 0 .01 -1.84 0 .01 51 25.5 8 1.99 0 .03 -1.68 0 02 52 26 0 1 75 0 .02 -1.26 0 02 53 26.5 -1. 94 0 04 -1. 92 0.02 54 27 0 9 -2.04 0 .02 -1.96 0 .01
42 APPENDIX 1. (Co n ti nued) mm High Sample from Density ()18() error ()13C error # Base Band 55 27.5 -1.75 0.06 -0 .91 0.01 56 28 0 1 88 0 03 -1.2 1 0.00 57 28.5 10 -1.71 0.03 -1.57 0.02 58 29.0 -1.90 0.07 -1.96 0.03 59 29.5 1 52 0.09 -1.43 0.05 60 30.0 1.84 0.04 -1.94 0.02 6 1 30.5 11 62 31.0 -2.04 0.02 -1.91 0.02 63 31.5 -1.90 0.03 -1.84 0.02 32.0 12 64 32.5 2.05 0 .02 -1.71 0.01 6 5 33.0 -1.92 0 08 1.38 0.04 66 33.5 -1.97 0 .01 -2 14 0.04 67 34.0 13 -1.98 0.04 -1.59 0.02 68 34.5 1 .87 0.05 -1.48 0 02 69 35.0 -2.40 0.03 -2 88 0.03 70 35.5 2.41 0.04 -3.02 0.01 71 36.0 14 -2.41 0.03 -2.27 0.01 72 36.5 1.97 0.03 -1.51 0.02 73 37. 0 -2.25 0.01 -2.01 0 03 74 37.5 -2. 30 0.01 -3.32 0.02 75 38 0 15 2 30 0 .06 -3.15 0.01 76 38.5 2.24 0.03 -2.60 0.01 77 39 0 -2 17 0.11 -2.09 0 .07 78 39.5 -2.38 0.0 2 -3.50 0.01 79 40.0 -2 14 0.08 -2.43 0.06 80 40.5 16 2 17 0.02 -1.79 0.01 81 41.0 82 41.5 2.23 0.04 -3.15 0.02 83 42 0 -2.41 0.03 -2.86 0.02 84 42.5 17 2.29 0.05 -2.25 0.01 85 43 0 2.31 0.07 -1.68 0.03 86 43.5 -2.46 0.06 -2.91 0 02 87 44.0 2.47 0.01 -3.35 0 02 88 44.5 -2 39 0 .02 -2.81 0 02 89 45 0 18 -2.16 0.03 -2.20 0 .01 90 45.5 2.40 0.02 -2.17 0.01 91 46 0 2.52 0.03 -2.65 0.01 92 46.5 -2.31 0.06 -3 36 0 03 93 47.0 -2.52 0 .03 -3.00 0.02 94 47.5 2.23 0 04 -2.25 0 04
43 APPENDIX 1. (Continued) mm High Sampl e from Density 818() error 813(: error # Base Band 95 48.0 19 -2.29 0.03 -1.65 0.02 96 48.5 -2.33 0.06 -1.92 0 03 97 49.0 -2 60 0.02 -3.77 0.02 98 49.5 -2.58 0.07 -3.53 0 05 99 50.0 -2.5 1 0.03 -2.66 0 .01 100 50.5 20 -2 25 0.03 -1.77 0.01 101 51.0 102 51.5 -2.42 0.03 -3.43 0.01 1 03 52.0 2.32 0.03 -2.78 0 02 104 52.5 -2.42 0.03 -2.71 0.01 105 53.0 21 -2.17 0.07 -1.63 0.02 106 53.5 -2 30 0.04 -2.18 0 02 107 54.0 2 42 0.04 3.32 0.02 108 54.5 2.04 0.03 -2.13 0.02 109 55.0 22 2.32 0.05 -2.08 0.01 110 55.5 2.33 0.02 -1.87 0.03 111 56 .0 -2.59 0.01 -2.57 0.01 112 56.5 -2.67 0.05 -4 .09 0 02 113 57.0 -2.49 0 04 -2.97 0.03 114 57.5 23 -2.23 0.05 -2 14 0 .03 115 58 .0 -2 .3 2 0 02 -1.99 0.01 116 58.5 -2 17 0.06 -2.63 0.01 117 59.0 2.08 0.04 -2.35 0.01 118 59 5 2.09 0.03 -2.04 0.02 119 60.0 -2.20 0.01 -1.49 0.02 120 60.5 24 -2 44 0. 03 -2.62 0.01 121 61.0 2.45 0.04 -2.66 0.02 122 61.5 -2 48 0. 03 -2.48 0.03 123 62.0 -2 .3 0 0 .02 -1.69 0.02 124 62.5 -2.26 0. 04 -1.84 0.02 125 63.0 25 2 00 0. 02 -2.2 3 0.03 126 63.5 -2.3 5 0 .01 2.32 0.02 127 64.0 2.02 0 .02 -1.59 0.02 128 64.5 26 129 65.0 -2.02 0 .02 -1.68 0.03 130 65.5 -2.06 0.03 -1.90 0.02 131 66.0 1 98 0 .01 -1.31 0.02 132 66.5 27 133 67.0 2 28 0.03 -1.84 0.03 134 67 5 -2.09 0.03 -2.37 0 .02
44 APPENDIX 1 (Continued) m m High Sample from Density 818() error 813C error # Base Band 135 68.0 -2.08 0.02 -1.48 0.02 136 68.5 28 -2 13 0 03 -1.43 0 03 137 69 0 -2. 07 0.01 -1. 98 0 02 138 69.5 -2 .21 0.02 -2.28 0.03 139 70.0 -1.95 0.09 -1.83 0.02 140 -1.95 0 04 -1.46 0.02