Paleolatitude Distribution of Phanerozoic Marine Ooids and Cements

Home , Ooid

Palaeogeography, Palaeoclimatology, Palaeoecology, 78 (1990): 135 148 135 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands

Paleolatitude distribution of Phanerozoic marine ooids and cements

BRADLEY N. OPDYKE and BRUCE H. WILKINSON

Department of Geological Sciences, The University of Michigan, Ann Arbor, MI 48109-1063 (U.S.A.)

(Received June 5, 1989; revised and accepted October 31, 1989)

Abstract

Opdyke, B. N. and Wilkinson, B. H., 1990. Paleolatitude distribution of Phanerozoic marine ooids and cements. Palaeogeogr., Palaeoclimatol., Palaeoecol., 78: 135-148.

Data on 493 Phanerozoic marine ooid and cement occurrences indicate that the dominance of calcite versus aragonite in tropical marine settings has changed in response to variation in atmospheric CO2 and/or oceanic temperature gradient. Holocene ooid and cement precipitation occurs over similar latitudes, with means centered around 24° and 28°, respectively. Aragonite and calcite also display roughly comparable distributions, with average occurrences between 25° and 28°. Surface seawater saturation values requisite for ooid-cement carbonate precipitation are at least 3.8 (~,rs) for aragonite and 3.4 (~a,s) for calcite. Ancient ooid-cement occurrences vary in space and time, with depositional zones generally closer to the equator during continental emergence; greatest extent correlates with periods of maximum transgression. Aragonite formation is favored in more equatorial localities than calcite when cement-ooid distributions are narrow and continents are emergent. Similarity of latitude distribution of marine ooids, cement, and biogenic carbonate suggests that physicochemical processes that control levels of carbonate saturation were more important in predicating sites of limestone accumulation in Phanerozoic seas than biological processes related to net productivity of various carbonate platform communities. Continental position and tropical shelf area available for carbonate accumulation dictates the relative abundance of shallow water inorganic carbonate precipitates in space and time.

Introduction 1964). Occurrences of Paleozoic tillite are still widely used to calibrate apparent polar wander Geologists have long used the distribution of paths (e.g. Van der Voo, 1988; Scotese and different rock types or floral/faunal assem- Barrett, 1989), and temperature indicators blages to interpret ancient climates. Even such as fossil fauna and flora remain the before general acceptance of plate tectonics, principal evidence that Cretaceous and Eocene continental mobility was inferred from anom- climates were generally warmer than at pre- alous lithologic occurrences such as evaporite sent (e.g. Bailey and Sinnott, 1915; Smiley, and/or carbonate sequences at high latitudes 1967; Douglas and Williams, 1982; Wolfe and (e.g. Wegener, 1915; Koeppen and Wegener, Upchurch, 1987). 1924; DuToit, 1939). Early paleomagnetic Limestone-dolostone sequences are particu- studies also relied heavily on the coincidence larly notable in this regard in that carbonate of specific facies within projected paleolati- accumulation is generally thought to be tudes as support of continental reconstruc- favored in warm shallow seas that are far tions (e.g. Opdyke, 1959, 1962; Blackett, 1961; removed from sources of terrigenous clastic Irving and Briden, 1962; Briden and Irving, sediment (e.g. Milliman, 1974; Wilson, 1975;

Leeder, 1982; Tucker, 1985). Modern platforms Late Cretaceous, a time period characterized occur preferentially on tropical to subtropical by continental submergence, elevated CO2, shelves (Emery, 1968; Lees, 1975) where warmer climate, and possibly greater poleward seawater reaches its highest saturation (Morse extent of carbonate-saturated seawater. Briden et al., 1980), and available data suggest that and Irving (1964) and Ziegler et al. (1984), on paleoplatforms exhibit a similar distribution the other hand, concluded that areas of carbon- (e.g. Briden and Irving, 1964; Parrish, 1982; ate deposition do not expand poleward during Ziegler, 1984); Mesozoic-Cenozoic limestones warmer time intervals, and suggested that occupy a range concentrated about the equa- limits of carbonate generation are primarily tor, but exhibit a skewed distribution that controlled by biologic factors such as the suggests primary accumulation at about 20 ° interdependence between latitude, degree of (Fig.l). light penetration, and the direct or indirect Given this relatively narrow range, the fixation of calcium carbonate by algae in paleolatitude distribution of cratonic lime- shallow platform settings. stone sequences should serve as a sensitive Clearly, a number of interrelated processes record of changes in those parameters that act to determine areas of carbonate generation control seawater saturation, including temper- in shallow marine environments. At the coars- ature, atmospheric CO2 (e.g. Mackenzie and est scale, rates of limestone accumulation must Pigott, 1981; Berner et al., 1983), and climate reflect rates of delivery of Ca and CO~- ions (e.g. Fisher, 1981, 1984; Sandberg, 1983, 1985; to global oceans which, in turn, depend on Worsley et al., 1986). Previous studies of areas of exposed crust, rates of weathering, carbonate accumulation, however, have ar- and amounts of hydrothermal alteration at rived at different conclusions concerning oceanic ridges. In addition, finite ion source specific details of the reliability of this record. fluxes are partially controlled by dissolution Opdyke and Wilkinson (1989), for example, processes in the ocean and are ultimately suggested that maximum expansion of shallow partitioned between deep marine and shallow carbonate environments occurred during the platform depositional settings, depending on those factors that control net carbonate productivity of planktic and benthic marine com- 80 munities. Given that planktic calcifiers only arose and diversified since the Triassic, and 6o that a variety of taxa have dominated benthic cnnit communities over the past 590 m.y., secular change in the distribution of platform carbon- 4o O ate might indeed reflect a complex of biologic factors largely independent of variables such 20 as temperature and salinity that determine levels of carbonate saturation in shallow marine settings. -70 -50 -30 -10 10 30 50 70 In short, patterns of cratonic carbonate DEGREES LATITUDE accumulation probably record diverse rela- Fig.1. Paleolatitude distribution of Mesozoic-Cenozoic tions between biological and physicochemical shallow water carbonate sequence localities, modified from Zielger et al. (1984). Bars represent actual data whereas the processes that preclude simple correlation shaded curve represents the likely sum of northern and between limits of biogenic carbonate produc- southern hemisphere occurrences. To eliminate bias intro- tion, ambient seawater saturation, and cli- duced by the dominance of Mesozoic-Cenozoic shallow mate. In order to examine these relations and marine basins in the northern hemisphere, the southern hemisphere is inferred to have a similar potential for the to evaluate the appropriateness of carbonate distribution of carbonates. sediment distribution as a record of paleo- PALEOLATITUDE DISTRIBUTION OF PHANEROZOIC MARINE OOIDS AND CEMENTS 137 climate, while striving to avoid many of the al., 1980; Mucci, 1983; Cooke and Kepkay, 1984; factors that influence the distribution of differ- Feely et al., 1984). but, given that calcium/ ent biotic communities, we have evaluated salinity ratios of modern seawater vary by Phanerozoic limits of inorganic calcium no more than about 1.5% (Culkin and Cox, carbonate precipitation through tabulation of 1966), seawater saturation variation is largely occurrences of marine ooids and cement. Such controlled by differences in carbonate ion an approach may yield a better measure of concentration. Because alkalinity values are physicochemical variation in the marine realm relatively constant in surface oceans, change because these are largely abiotic precipitates, in temperature is therefore the most important the distribution and mineralogy of which may factor in controlling carbonate saturation respond to changes in physical and chemical levels (e.g. Takahashi et al., 1982) and, hence, parameters of surface seawater throughout where ooids and cement originate, (Fig.2). As a geologic time. result, greater carbonate saturation is found in equatorial regions as CO 2 is released to the Abiotic carbonate precipitates atmosphere, while in polar regions atmospheric CO 2 is absorbed by the colder waters. The inorganic precipitation of calcium Because of the strong dependence of carbonate carbonate from marine fluids results either in saturation on temperature and hence latitude, the formation of layers of pore-lining cement or massive carbonate precipitation from marine their centripetal counterpart as cortical lami- fluids is generally restricted to tropical seas. nae on ooids and other coated grains. While Regional deviation from this pattern, how- these forms of calcium carbonate generally ever, may give rise to variation in ooid-cement record the accumulation of limestone in distribution. Western boundary currents, for shallow tropical settings, variation in their example, carry warmer water to temperature abundance as sediment components, and the latitudes, somewhat expanding depositional dominance of different calcium carbonate poly- range along western basin edges, while cooler morphs among cortical-intergranular phases, may reflect regional variation in composition 30 related to seawater carbonate saturation. o

A ~mm • Controls on formation c.) • ~u mlmm LU 2O • .o Q:: BD [] Obviously, abiotic ooids and cements form in I-- < response to ambient physicochemical condi- rr LU ." • ATLANTIC tions in the surface ocean. Physically, ooid Q- 10 [] PACIFIC # • • • INDIAN LD cortex growth requires elevated levels of I-- ~'o. .a calcium carbonate supersaturation and kinetic % g. m',,%~ o energy via waves or strong currents for grain 0 ,~ ,m , , i , , i , , i r , i i agitation (e.g. Ball, 1967). Marine cement also 14 18 22 2.6 3.0 34 38 4.2 46 OMEGA ARAGONITE only originates in supersaturated solutions but is most common in reefal voids where precipi- Fig.2. Relation between carbonate saturation values (~ar~) tation is also facilitated by strong fluid flow and temperature for surface water of the Atlantic, Indian, and Pacific Oceans. ~,,~ is defined as o=[Ca2÷][CO~-]] (e.g. Lighty, 1985). Kar8 where COl- concentrations for surface water are Chemically, CO2, pH, alkalinity, and ionic taken directly from data tables in Takahashi et al. (1980, strength all play important roles in determin- 1982). Ca 2÷ concentrations are assumed to vary as a ing saturation state (MacIntyre, 1965; Edmond function of salinity (Culkin and Cox, 1966). K,,~ varies as a function of temperature and salinity (Mucci, 1983). Note a and Gieskes, 1970; Berner, 1976; Takahashi et linear increase in saturation values (fl, rg) from polar (0°C) al. 1976; Plath and Pytkowicz, 1980; Morse et to equatorial (30°C) regions. 138 S.N. OPDYKE AND B. H. WILKINSON currents flowing toward the equator may (Land et al., 1979), and the Mediterranean at narrow tropical ranges along eastern margins. Gabes Bay, Tunisia (Fabricius et al., 1970) and Similarly, excess CO2 generally occurs in areas Neapolis, Greece (Richter, 1976). Ooids from of upwelling along narrow equatorial belts and Tunisia and Greece are not actively forming along eastern basin margins which are regions today, and are largely relict deposits of the of low temperature and low saturation state Holocene transgression. Modern Indian and (Keeling, 1968; Takahashi and Azevedo, 1982). Pacific ocean carbonate deposits are largely Where tropical currents mix with cooler high void of ooids (Lees, 1975), and small deposits of latitude water, such as along the east and west ooids within areas such as the Great Barrier coasts of Australia, surface water CO2 is Reef are somewhat enigmatic (Marshall and depressed relative to the atmosphere. These Davies, 1975; Davies and Martin, 1976). warm currents become undersaturated relative to atmospheric CO2 as they move to higher Controls on mineralogy latitudes and slowly cool. The Antilles current, which has its origin off the northwest African Although ooids and cement are generally coast, for example, is mature and cooling by restricted to warm, wave-agitated environ- the time it reaches the northeastern coast of ments, either consist of magnesium calcite or the Bahamas (Takahashi et al., 1982). Here, aragonite, a mineralogical diversity that may lower CO2 and higher CO~- concentration also serve as a record of ocean saturation state. between 20 ° and 30°N on the western Atlantic With increasing temperature, calcite precipi- margin correlate with rapid inorganic carbon- tated from seawater contains greater amounts ate precipitation. of incorporated Mg in the crystal lattice, Additional variation in the distribution of ranging from only a few mole % in deep cold ooids and cement may relate to the develop- waters below the oceanic thermocline (e.g. ment of enclosed and/or evaporative basins Schlager and James, 1978) to nearly 20 mole ~/o such as the Mediterranean Sea, where saturation state is higher than that of open oceans at in warm shallow settings (e.g. Alexandersson, similar latitudes. This type of deviation should 1972). Moreover, at tropical temperatures of be recorded as ooid/cement "outliers" from any 25-28°C, the solubility of precipitated calcite paleolatitude distribution, much in the same with about 12 mole % Mg approaches that of manner as Mediterranean Sea occurrences are aragonite (Walter and Morse, 1984). Aragonite at the northern limit of a normal distribution and magnesium calcite may form in intimate for Holocene precipitates. association as intergranular and intragranular Because of such regional variations, ooids cement, and can co-occur as alternating lami- and cement are not ubiquitous in low-latitude nea within single ooids (Land et al., 1979; carbonate platform-reefal environments. Major et al., 1988). Among the modern areas of carbonate accumu- Both mineralogies also formed ooid-cement lation, only a few are significant examples of carbonate during much of the Phanerozoic, but ooid generation. The Bahama Platform (e.g. exhibit considerable secular variation in their Ball, 1967), Yucatan (Ward and Brady, 1973), relative abundance. Whereas aragonite is pres- southern Cuba (Daetwyler and Kidwell, 1959), ently the dominant phase as cortical carbonate Pedro Bank (Zans, 1958), Serrana Bank (Milli- within Holocene ooids, and is volumetrically man, 1969), the Persian Gulf (e.g. Loreau and nearly equal to micritic magnesian calcite as a Purser, 1973), the Gulf of Suez (Sass et al., marine cement, abiotic aragonite (or calcitized 1971), Lizard Island, Australia (Davies and equivalent) is virtually absent from large Martin, 1976), and Shark Bay (Davies, 1970) portions of the Phanerozoic sedimentary rec- are notable in this regard. Other Holocene ooid ord that correspond to periods of continental deposits include those from Baffin Bay, Texas flooding (e.g. Sandberg, 1985). PALEOLATITUDE DISTRIBUTION OF PHANEROZOIC MARINE OOIDS AND CEMENTS 139

Changes in seawater Mg/Ca ratio have been logues can be evaluated in terms of surface suggested as a controlling factor of calcite water saturation. versus aragonite dominance (Mackenzie and In order to reduce the effect of redundant Pigott, 1981). Berner (1975, 1978) has shown references to a single stratigraphic unit con- that higher Mg/Ca ratio inhibits calcite taining marine ooids and/or cement; multiple growth rate, while laboratory studies by Bur- reports of abiotic carbonate of the same age ton (1988), Burton and Walter (1988) and Mucci within a continuous geographical area were et al. (1989) suggest that aragonite precipita- combined into a single "occurrence". With tion rate exceeds that of calcite over a wide respect to Quaternary units, this lumping of range of Mg/Ca and SO~- concentrations in reports results in occurrences at a scale of the seawater or seawater-like solutions. When Bahamas, the Persian Gulf, or Yucatan. Older saturated with respect to both polymorphs, Phanerozoic occurrences are mostly at the calcite precipitation rate increases relative to stratigraphic level of formations. that of aragonite at low temperatures, at low Mg/Ca ratios, and at low sulfate activities Quaternary precipitates (Burton, 1988). From a perspective of paleoclimate, aragon- Latitudes of 78 Pleistocene-Holocene ooid- ite versus calcite dominance in shallow car- cement occurrences are more-or-less normally bonate sequences has also been linked to distributed within the northern hemisphere atmospheric CO 2 (e.g. Sandberg, 1983) and (Table 1), and exhibit a mean latitude of 27 ° carbonate saturation (Given and Wilkinson, (Fig.3A). Reported modern cements are much 1987), with greater abundance of abiotic cal- more numerous (n=56) than coeval ooids cite generally recording higher sealevel, ele- (n=22), and exhibit a northern hemisphere vated CO2, and lower seawater saturation mean of 28 ° (Fig.3B). Aragonite (n=22) and state. If correct, absence or latitude restriction calcite (n=34) cement exhibit statistically of aragonite might then be considered an indistinguishable distributions with means at indication of warmer climate, as would a wider 27 ° and 28 ° , respectively. It is notable, how- distribution of ooids and/or cement. ever, that several examples of calcite cement In order to evaluate ooid/cement distribu- range to 60 ° (Fig.3B). These are somewhat tions as records of temperature-saturation anomalous among modern carbonate precipi- variation in Phanerozoic oceans, data on the tates in that they largely comprise occurrences age, geographic location, stratigraphic posi- within terrigenous clastic sediment, and are tion, and fabric of 316 ooid and 177 cement probably only recognized as early cement occurrences were tabulated from the phases because of their presence in otherwise literature. Methanogenic ('3C-enriched) cements were not included in this compilation, TABLE I nor were occurrences of ooids or cement replaced by dolomite or quartz. All of these Latitudinal distribution of Phanerozoic ooids and cement ooid-cement occurrences are associated with Quaternary Phanerozoic relatively shallow water units, and most occur in carbonate sequences. n X SD n X SD In the following, these data are first con- Ooids and Cement 78 27 10 415 16 10 sidered in terms of the distribution of ooids and All ooids 22 24 8 294 16 10 cement, then in terms of ooid versus cement Aragonite ooids 19 25 7 36 12 9 occurrences, and finally in terms of the distri- Calcite ooids 4 21 9 138 17 109 bution of aragonist versus calcite. Quaternary All cement 56 28 12 121 16 10 occurrences are considered first in order to Aragonite cement 22 27 7 25 9 9 establish a framework in which ancient ana- Calcite cement 34 28 14 96 18 10 140 B. N. OPDYKE AND B. H. WILKINSON

8~ A MARINE PRECIPITATES ' I cement renders comparisons between aragon- 6-4 ite and calcite statistically meaningless. Distribution of modern (recent) marine ooids t , . .0 .l!llpIH, .1! and cements suggests that carbonate precipita- B MAR,.EOEME.T /! tion closely reflects sea-surface saturation °° ,. [] ARAGONiTE ...llrl • I gradient, in that no occurrences within lime- o • O,LO,, Irlllll I stone sediment lie poleward of 35 °. Quaternary ; FI FI FII I IIIIIIIIIIIii1 I I l ooid-cement distribution is strikingly similar =~41C ...... MARI.E oo,Ds [1 I1 to that of primarily biogenic carbonate (Fig.l); " [] ...Go.,T~ II II FIN virtually all occurrences are in the northern =: • IIII IIII hemisphere, the Caribbean region in particu- .... N!rl, ,I,rli!lii,lllN , lar. D TOTAL Theoretically, such a skewed distribution --~ 2 SHELF AREA 1 could in part reflect warmer northern hemisphere sea surface temperatures. The inter- tropical convergence zone, the equatorial area -60 -40 -20 0 20 40 60 where warm air rises and moves poleward, is LATITUDE skewed about 5 ° to the north due to the cooling Fig.3. Latitudinal distribution of Quaternary carbonate influence of Antarctica. However, climate precipitates. zones are generally symmetrical on either side A. Distribution of ooids and cement. Note similar mean of the equator and average temperature is occurrences at about 28°, but that cement is reported from a somewhat broader range. likely to be similar at both 20°N and S B. Reported calcite-aragonite cement distributions which (Neiburger et al., 1982). indicates aragonite has a narrower range than calcite Almost certainly, the disproportionately relative to distance from the equator. C. Quaternary ooids. Virtually all Holocene deposits are large number of ooid and cement occurrences aragonite, but calcite grains are forming in Baflin Bay, north of the equator merely reflects the pre- Texas, and occur as several relict deposits in deeper shelf dominance of land area presently in the settings. Differences between total ooid occurrences in A northern hemisphere and, hence, the greater and C reflects the fact some deposits include both aragonite and calcite as cortical phases. D. Late Cenozoic areal extent of shallow shelf environments in shelf areas from Parrish (1985). Note predominance of this hemisphere as well. The distribution of shallow shelf settings at about 30°N, and a general Quaternary ooid/cement occurrences, Meso- correspondence between this pattern, that of ooid-cement zoic-Cenozoic platform carbonate sequences, distributions (Fig.3A), and that of shelf carbonate sequences (Fig.l). and estimates of late Cenozoic shelf areas by latitude from Parrish (1985), all exhibit similar patterns (Fig.3D). This suggests that, while unlithified units. Ancient analogues are global temperature gradients and poleward largely unidentified in terrigenous sequences. extents of saturated surface waters indeed All Quaternary ooids occur within 40 ° of the control limits of carbonate precipitation, the equator (Fig.3C) but are presently forming in abundance of ooid/cement occurrences more only 6 locations to any significant degree. closely reflects the areal size of agitated Grains in all but one of these (Baffin Bay; shallow water settings in which precipitation Land et al., 1979) are exclusively aragonite. can occur. Although the presence of calcite among three relict occurrences may reflect variation in Phanerozoic precipitates seawater chemistry and dominant cortical mineralogy on very short time scales, the In order to determine an appropriate latitude scarcity of Quaternary ooids relative to marine for each reported occurrence of Phanerozoic PALEOLATITUDE DISTRIBUTION OF PHANEROZOIC MARINE OOIDS AND CEMENTS 141 ooids (n=294) and cement (n= 121) (Table 1), aragonite. Original cement mineralogy can be all were plotted on appropriate paleogeogra- inferred employing similar textural criteria phic maps by Denham and Scotese (1988). (e.g. James and Klappa, 1983), with botryoidal Errors in paleomagnetic data on which these masses common to former aragonite and iso- reconstructions were based are about 5 ° , with pachous crusts typical or calcite. uncertainties for the lower Paleozoic perhaps Taken together, the distribution of ancient ranging to 10 ° (Van der Voo, 1988). ooids and cement are not statistically different, Data on the distribution of these 415 Phaner- with both exhibiting a mean of 16 ° from the ozoic ooids and cements demonstrate that equator with standard deviations of about 10 ° Paleozoic occurrences are dominantly in the (Fig.5A). Former aragonite cement however, southern hemisphere, while the majority of shows a narrower distribution about the equa- Mesozoic and Cenozoic localities occur north tor than does calcite (Fig.5B). Differences of the equator. These define a mean distribu- between aragonite and calcite ooids are more tion extending from about 30°S in the Cam- subtle (Table 1), but radial calcite ooids ex- brian to about 30°N in the Tertiary (Fig.4). tended to higher latitudes than did coeval Virtually all occur within 35 ° degrees of the grains of aragonite (Fig.5C). equator, a band also defined by most data on Quaternary ooids and cement. Discussion Although ancient carbonate components are now calcite and/or dolomite, original mineral- On the basis of reported aragonite-calcite ogy can be inferred from preserved fabrics due marine ooids and cement, several conclusions to the fact that diagenetic calcitization of can be drawn concerning the distribution of aragonite typically results in the obliteration inorganically precipitated carbonate over the of textural detail (Sandberg, 1983, 1985; past 590 m.y. These pertain to their distribu- Tucker, 1984; Chow and James, 1987). Ancient ooids that exhibit radial arrangement of con- stituent acicular crystallites are generally 30 " A - I I MARINE PRECIPITATES interpreted as having been precipitated as ~w 20- [] OOiDS ! I . ~° n- 10 " CEMENT i calcite, while sparry cortices of interlocking o " JFtilil,n,/l!l.q anhedral calcite crystallites were formerly 15"B ~') [] ARAGONITE_ I MARIHE CEMENT I

6O ,,o ,, • c. c,..lll o I r . LU MARINE OOIDS AND CEMENT "3 40 "'i,i" A.,NOO,OS •:i " A.A.ON,.E I , ~< 20 CLC,i Llihl]ll, I LL o -60 -40 -20 0 20 40 610 PALEOLATITUDE -20 : !... : !-...-. Fig.5. Paleolatitude distribution of Phanerozoic marine O -40 ooids and cement exclusive of Quaternary examples. L.,Uc~ A. Distribution of ooid-cement occurrences. Although -60 reports of marine oolite are much more common than those o 2;o 4~0 6OO of cement, distribution of the two populations is statisti. AGE (my) cally indistinguishable. Fig.4. Paleolatitudinal distribution of Phanerozoic marine B. Latitudes of aragonite and calcite marine cement ooids and cement exclusive of Quaternary occurrences. suggesting that occurrences of aragonite are generally Latitudes determined from pa]aeogeographic maps of more equatorial than those of calcite. (Denham and Scotese, 1986). Note a striking progressive C. Aragonite and calcite oolite distribution. While less shift in sites of carbonate precipitation from the southern pronounced than cement, aragonite occurrences define a to the northern hemisphere over the past 590 m.y. narrower latitudinal distribution than do calcitic grains. 142 B.N. OPDYKE AND B. H. WILKINSON tion: (1) relative to saturation values in part reflecting higher rainfall and lower evapo- modern oceans, (2) relative to dominant allo- ration in these areas. As a result, surface chem mineralogy, (3) relative to different waters of greatest saturation occur between component types, (4) relative to paleolatitude, 10 ° and 30 ° and include equatorial regions on (5) relative to biogenic debris that makes up the western side of basins where upwelling is most shallow marine carbonate sediment, and weak. (6) relative to their utility as records of ocean Holocene cement is most abundant in these temperature-saturation variation. areas, but exhibits a distribution extending somewhat beyond that for ooids, both to the Present distributions north and to the south. Aragonite ooids and cement most commonly form in shallow seas In order to understand those processes near open ocean water where ~arg exceeds which control carbonate precipitation and about 3.8 and summer temperatures exceed polymorphism during cement crystal growth about 25°C (Fig.6). In such areas, seawater CO2 and/or ooid cortex accretion, preliminary data is commonly 10-15 ppm below that in equilib- on saturation of modern surface water with rium with the atmosphere (Takahashi and respect to aragonite were compared to areas of Azevedo, 1982). Calcite cement can form in ooid and cement formation. Surface water more temperate regions, and the minimum ~ars chemistry data from Takahashi et al. (1980, for these precipitates is about 3.4. Because 1982) exhibit increasing saturation values calcium carbonate saturation data represent (~,rg) from poles to latitudes of about 20 °, and open ocean surface water, these saturation constant or slightly decreasing ~,,g from there values probably represent minima from which to the equator. As a result, equatorial zones shallow shelf waters evolve. exhibit a slight saturation minima (Fig.6). Decreased saturation at equatorial latitudes Aragonite-calcite distributions primarily reflects reduced CO~- concentration, largely due to the higher CO2 of upwell- With respect to original mineralogy, aragon- ing water (Broecker and Peng, 1984), but in ite ooid-cement occurrences generally have narrower ranges than coeval calcite, both in modern (Fig.3) and ancient (Fig.5) sequences. The reason for this difference probably reflects relations between phase solubility and miner- 4 U.I alogy or Mg content. As noted above, at typical z [] tropical marine temperatures, calcite with 0 (9 about 12 mole% Mg has a solubility roughly < o • m IMI rv3 equivalent to that of aragonite (Walter and [] o o • : < o Morse, 1984). Moreover, available composi- U..I [] • ATLANTIC tional data on shallow Holocene cements o [] PACIFIC • 02 ,.~m m~ [] INDIAN ~ • indicated that Mg content decreases from about 18 to about 9 mole°//o with increasing

-70 -50 -30 -10 10 ,30 50 70 latitude. Although it is not (as yet) possible to DEGREES LATITUDE determine original Mg contents for ancient Fig.6. Relation between latitude and carbonate saturation cortical or cement carbonate, there is no values for surface waters of the Atlantic, Indian and reason to believe that Phanerozoic calcites did Pacific Oceans. From data in Takahashi et al. (1980, 1982). not exhibit a similar (or greater) range of Note an increase in ~arg values from polar to equatorial compositions. Hence, there is no reason why regions. Diagonally hachured box delimits the range of saturation values that correspond to the limits of aragoni- calcite ooids and/or cement, including those tic Quaternary ooids and cement. with lower Mg contents, should not have PALEOLATITUDE DISTRIBUTION OF PHANEROZOIC MARINE OOIDS AND CEMENTS 143 extended over a wider latitude range than high and/or sediment accumulation rates low, coeval aragonite. and most ooids form in shoals along non-reefal With respect to secular variation in original but wave-agitated shelf margins as well. The mineralogy, data on ooids and cement indicate similarity of these settings is probably that both have changed with time (Fig.7), with reflected in the equivalence of ooid-cement aragonite dominance corresponding to periods distributions. of continental emergence (e.g. Sandberg, 1983, 1985). If change in mineralogy reflects varia- Biochemical-physicochemical tion in seawater carbonate saturation, then an distributions absence of aragonite probably records periods of lower carbonate saturation in surface Perhaps more importantly, data on ooid- seawater, and perhaps lower Mg/Ca ratios as cement distributions suggest that latitudes well. represented by these components are little different from those reported for platform Ooid-cement distributions carbonate sequences composed primarily of metazoan hardpart debris. Such a similarity Available data indicate no significant differ- strongly indicates that, at least from a stand- ences in the distribution of marine ooids versus point of carbonate generation, biochemical cement in modern or ancient sequences, a and physicochemical processes operate simul- similarity of occurrences that probably reflects taneously in largely coincident shallow marine the fact that both require supersaturated fluids settings. Most of these are within 35 ° of the and elevated rates of fluid flow for their equator; although a few cratonic limestones formation. Although Quaternary cement ex- formed poleward of these limits (e.g. early hibits a broader distribution than ooids (Fig.3), Tertiary bryozoan limestones of Australia) and this is not true for older Phanerozoic se- range to 60 °, a few Holocene marine cements quences with a much larger number of occur- occur at similar latitudes as well (e.g. Fig.3). rences (Fig.5). Modern cement predominates in Conventional wisdom has long held that reefal and grainstone voids where fluid flow is carbonate deposition is enhanced in low latitudes, in shallow water, and in areas far- removed from areas of significant terrigenous A 4o DE] MARINE CEMENT clastic influx (e.g. Wilson, 1975; Bathurst, LLI 2O r-, 1975). Although the distribution of ooid-cement Z3 0 ...... },.- and limestone occurrences are largely in t-'- -20 < accord with these generalizations, fundamen- -.J -40 [] CALCITE O U.J tally different mechanisms are responsible for B n I n r n .< the formation of each type of sediment compo- B... 40 I~l MARINE OOIDS (,9 nent. Most limestone units consist of hardpart UJ 20 IJ..J n" 0 ...... debris, and many individuals have therefore WL~ -20 suggested that biological mechanisms act as -40 the primary factors in predicating carbonate -60 | .... [] sediment generation (e.g. Ziegler et al., 1984). 0 200 400 600 AGE (my) Correspondence between limestone accumu- Fig.7. Paleolatitudes of ooids and cement plotted by lation and saturated seawater, on the other inferred original mineralogy. Note similar aragonite hand, has led others to suggest that physico- dominance relative to calcite (dotted line portions) for chemical mechanisms are primarily respon- both components during the early Paleozoic, late Paleo- zoic-early Mesozoic, and late Cenozoic. These correspond sible for determining rates of biogenic carbon- to periods of continental emergence. This plot does not ate generation. include Quaternary data. Collectively, we know little concerning the 144 B.N. OPDYKE AND B. H. WILKINSON

relative importance of biological versus physi- -- NORTHERN]NDIA SAUDIAARABIA cochemical processes in the generation of ...... PARIS BASIN ...... MOSCOW ...... WESTERN AUSTRALIA ...... AMAZON DELTA 60 carbonate sequences, even in modern shelf- .~, CARLSBAD, NEW MEXICO reefal settings, nor is it well understood what 40 carbonate distributions and accumulation 2O rates might be like in the absence of either UJ C~ now-dominant metazoan calcifiers or now- ~---:.~-,.:.~-_-..,-- :~- ...... ""K,., ...... prevalent carbonate-supersaturated ambient '~ -20 ', \\ ,' -. marine fluids. If biological systems were prim- ',,. ',...../._._."~,~ \t,--,,". , ,t./ ary causative agents of carbonate accumu- -40 ,, ...... - . ,,/ , \ ~ '.\ ~/ "\ ~/',\, lation, some difference in the distribution of -60 biotic and abiotic carbonate components might 200 400 600 be anticipated. However, there appears to be AGE MY no discernable dissimilarity in the distribution Fig.8. Cambrian to recent drift tracks for seven sedimen- of carbonate components generated by direct tary basins with a global distribution (India, Europe, precipitation versus those formed during bio- Australia, Saudi Arabia, Asia, South America, and North mineralization. This correspondence seem- America) showing general movement of the most conti- ingly records an intimate link between nents to more northern latitudes over the past 510 m.y. Note similarity between ooid-cementlatitudes (Fig.4) and seawater composition and the fundamental the drift tract for the Permian Basin, New Mexico (heavy controls on carbonate formation, and suggests line) (Denham and Scotese, 1988). that limestone deposition will largely correspond to areas of supersaturated seawater, regardless of the dominance of various taxo- distribution of shallow tropical seas than any nomic groups in shallow marine settings. relation with seawater saturation gradient; mean latitudes of inorganic carbonate precipi- Ooid-cement distribution and climate tates are strongly dependant on paleopositions of shallow water shelves. Regardless of likenesses and/or differences Whereas virtually all reported Paleozoic in distributions based on mineralogy, allochem ooids and cements formed south of the equator, type, or mechanism of precipitation, virtually northern hemisphere dominance began when all Phanerozoic ooid-cement occurrences North America, which includes 64% of tabu- define a rather striking trend of increasingly lated data, crossed the equator during the northern sites of deposition over the past 510 Carboniferous (Fig.4 and 8). Such strong corre- m.y. (Fig.4). Calculation of drift tracks for lation between patterns of continental drift seven depositional basins in Africa, Asia, and patterns of ooid-cement deposition signifi- Australia, Europe, Indian, North America, and cantly limits the usefulness of average precipi- South America, demonstrates that most conti- tate latitude as a record of change in ocean nents have moved northward during the Phan- temperature and/or carbonate saturation. erozoic (Fig.8). To the degree that reported What then can be said concerning ooid- marine ooid and cement occurrences reflect cement occurrences and secular change in sea- actual abundances in carbonate sequences, water saturation? Whereas distribution aver- such a relation indicates an intimate relation ages relative to the equator are significantly between tropical areas of flooded continent biased by continental area, one other possible and areas/volumes of carbonate accumulation measure of saturation gradients is variation in in shallow epicratonic settings. the actual extent of oolite-cement occurrences. One immediate result of such dependence is Theoretically, and all other parameters re- that mean or average latitudes of accumu- maining constant, increasing CO 2 should relation should more closely reflect the average sult in lower seawater carbonate saturation PALEOLATITUDE DISTRIBUTION OF PHANEROZOIC MARINE OOIDS AND CEMENTS 145 which, at any specified time, decreases pole- restricted range during the Permo-carbonifer- ward with surface temperature. Hence, aragon- ous, a time period that corresponds to general ite dominance, cooler climate, lower sealevel, continental emergence (Vail, 1977), and per- and depressed pCO2 (Berner et al., 1983; haps to the largest glacial epoch in Phanero- Sandberg, 1983, 1985), may be recorded as zoic history (Frakes and Francis, 1988; Fig.9). compressed limits of ooid-cement occurrences. Relationships such as these suggest a qualita- Such a relation, of course, is dependent on the tive link between eustasy, seawater saturation, occurrence of shallow shelf depositional areas ooid-cement mineralogy, pole-to-equator tem- across poleward limits of carbonate precipita- perature gradient, and the latitude range over tion in at least one hemisphere, on the which carbonate precipitation is likely to occur. preservation of marine sediment from that setting, and on the evaluation of that sequence Conclusions in the ooid-cement tabulation. Calculation of maximum Phanerozoic ooid- In an effort to evaluate relations between cement latitudes yields a trend with a more climate and shallow water carbonate distribu-

}... o_. O < z ~ w

u~ u~ w O 0 ,~ z

8O w C3 70 I-- ~_ 60

o'~ 50 w <~ 40 ~ a0 I I

..3 :.~.-.:.::~.::~...... V ...... :::::::::::::::::::::: iiii~.-"~'~::::~.-"::::~i~:~::::::~:.

09 3 7~~ :'~'~!:~:~:" ":::'~" "::~..... !:~:~:~:~:~..:~.-:~:~:~:~:{~:::~-.-,;...:*;z~:-.'.:#~:::i~%i~;!{iC~ili~.;.(:M::iiii~i!~ili!i~ii::iS!i~iii~iiiiiti!~i!i!~:::,: ...... I I W 30- ...~• ....~:~:~:~:~: 2O 10 -;,:~:~:~:;~:~.::~ ~:~' ...:..:~i~i~!~;i~...... ~-- ...~:~:::~:::~.~~ ~~:~;" ...... ~!:~*~~:~::~:::~- :~::~"t ~.~:~:::~:~:~:~!:~:~:~:~;;: ...... !~ .~:!~:~.'~.~:~:~~ ...... ~! ~:.;:!:~:: :~:~:~:~ i ,=No c er, I I I I oev [si, I 100 200 300 400 500 AGE (my) Fig.9. Extent of Phanerozoic ice-rafted debris in Phanerozoic sedimentary sequences (Frakes and Francis, 1988), estimates of global sealevel (Vail, 1977), and maximums of ooid-cement occurrences. Note general correspondence between equatorial expansion of ice-transported material, greater continental emergence, and generally narrower distribution of abiotic ooid- cement precipitates from about 350 to 200 m.y. 146 B.N. OPDYKE AND B. H. WILKINSON tion, latitude and original mineralogy of Phan- Van der Voo for advice on the lower Paleozoic erozoic ooid and cement occurrences were paleomagnetics, and David Dettman for re- tabulated from 493 reports in the literature viewing early versions of the manuscript. (Table 1). Mineralogical dominance of aragon- Steve Boss and Nancy Opdyke helped with ite over calcite in both ooids and cement is, as statistical manipulations of ooid-cement data. pointed out by Sandberg (1983, 1985) coinci- Angela DiMartino was a great help during the dent with periods of continental emergence. literature search. An anonymous reviewer Mean latitudes of modern and ancient ooid and helped to improve the final draft. Research on cement occurrences suggest a wider latitude patterns of Phanerozoic carbonate accumu- range for calcite relative to aragonite, and an lation at The University of Michigan is sup- identical range for ooids relative to cement and ported by the National Science Foundation, oolite-cement relative to biogenic debris. NSF grants EAR-86-07970 and EAR-88-03910. Ooid-cement formation is an inorganic processes that only occurs when critical References thresholds of physical agitation and chemical saturation have been surpassed. Global ocean Alexandersson, T., 1972. Intragranular growth of marine carbonate saturation data indicate that satura- aragonite and Mg-calcite -- evidence of precipitation from supersaturated seawater. J. Sediment. Petrol., 42: tion levels (~arg) must be at least 3.4 for calcite 441-460. cement growth and 3.8 for aragonite ooid- Ball, M. M., 1967. Carbonate sand bodies of Florida and the cement precipitation. Additional data on ooid Bahamas. J. Sediment. Petrol., 37: 556-591. shoal water compositions are necessary to Bailey, I. W. and Sinnott, E. W., 1915. A botanical index of Cretaceous and Tertiary climates. Science, 41: 831-834. more tightly constrain this value. Bathurst, R. G. C., 1975. Carbonate Sediments and Their Finally, ooid-cement occurrence data suggest Diagenesis (Developments in Sedimentology, 12). Elsev- that marine carbonate systems have responded ier, New York, N.Y., 658 pp. to changes in temperature and CO 2 concentra- Berner, R. A., 1975. The role of magnesium in the crystal growth of calcite and aragonite from seawater. Geochim. tions in global oceans. Carbonate precipitates Cosmochim. Acta, 29: 489-504. have not been uniformly distributed over invari- Berner, R. A., 1976. The solubility of calcite and aragonite ant areas of deposition. Rather, Phanerozoic in seawater at atmospheric pressure and 34.5% salinity. ranges have evidently narrowed while calcite Am. J. Sci., 276: 713-730. Berner, R. A., 1978. Equilibrium kinetics, and the precipi- precipitation gave way to aragonite dominance tation of magnesian calcite from seawater: Am. J. Sci., during periods continental emergence and in- 278: 1475-1477. ferred cooler climate. This behavior implies a Berner, R. A., Lasaga, A. C. and Garrels, R. M., 1983. The higher low-latitude saturation state relative to carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million calcite seas, while at the same time surface years. Am. J. Sci., 283: 641-683. waters between about 30 ° and 40 ° from the Blackett, P. M. S., 1961. Comparisons of ancient climate equator have lower saturation states than with the ancient latitude deduced from rock magnetic calcite seas. In effect, aragonite seas correspond measurements. Proc. R. Soc. Lond., A263: 1-30. Briden, J. C. and Irving, E., 1964. Paleolatitude spectra of to more highly saturated waters in lower sedimentary paleoclimatic indicators. In: A. E. M. Nairn latitudes. Calcite oceans, generally correspond (Editor), Problems in Paleoclimatology. Wiley, New to periods of continental flooding, elevated CO2 York, N.Y., pp. 199-224. warmer temperature, lower equatorial calcium Broecker, W. S. and Peng, T. H., 1982. Tracers in the Sea. Eldigio Press, Palisades, New York, N.Y., 690 pp. carbonate saturation state, and an expanded Burton, E. A., 1988. Laboratory investigation of the effects latitude range of ooid-cement deposition. of seawater chemistry on carbonate mineralogy. Thesis. Washington Univ., St. Louis, 306 pp. Acknowledgments Burton, E. A. and Walter, L. M., 1988. Experimental investigation of Mg]Ca and sulfate as controls on temporal variations in mineralogy of marine carbonates. We thank Taro Takahashi for the kindly Geol. Soc. Am. Abstr. with Programs, 20, p. 74. sending us the TTO North Atlantic data, Rob Chow, N. and James, N. P., 1987. Facies-specific, calcitic PALEOLATITUDEDISTRIBUTION OF PHANEROZOICMARINE OOIDS AND CEMENTS 147

and bimineralic ooids from Middle and Upper Cambrian Irving, E. and Briden, J. C., 1962. Paleolatitude of platform carbonates, western Newfoundland, Canada. J. evaporite deposits. Nature, 196:425 428. Sediment. Petrol., 57: 907-921. James, N. P. and Klappa, C. F., 1983. Petrogenesis of early Cooke, R. C. and Kepkay, P. E., 1984. Apparent calcite Cambrian reef limestones, Labrador, Canada. J. Sedi- supersaturation at the ocean surface: Why the present ment. Petrol., 53: 1051-1096. solubility product of pure calcite in seawater does not Keeling, C. D., 1968. Carbon dioxide in surface ocean predict the correct solubility of the salt in nature. Mar. waters -- 4. Global distribution. J. Geophys. Res., 73: Chem., 15: 59-69. 4543-4553. Culkin, F. and Cox, R. A., 1966. Sodium, potassium, Koeppen W. and Wegener, A., 1924. Die Klimate der magnesium, calcium and strontium in seawater. Deep- geologischen Vorzeit. Borntraeger, Berlin, pp. 255. Sea Res., 13: 789-804. Land, L. S., Behrens, E. W. and Frishman, S. A., 1979. The Daetwyler, C. C. and Kidwell, A. L., 1959. The Gulf of ooids of Baffin Bay, Texas. J. Sediment. Petrol., 49: Batabano, a modern carbonate basin. In: Fifth World 1269-1278. Pet. Congr., 1., Pap., 1, pp. 1-20. Leeder, M. R., 1982. Sedimentology: Process and Product. Davies, G. R., 1970. Carbonate bank sedimentation, eastern Allen and Unwin, Winchester, Mass, 344 pp. Shark bay, Western Australia. In: Carbonate Sedimenta- Lees, A., 1975. Possible influence of salinity and temper- tion and Environments, Shark Bay, Western Australia. ature on modern shelf carbonate sedimentation. Mar. Am. Assoc. Pet. Geol. Mem., 13:85 168. Geol., 19: 159-198. Davies, P. J. and Martin, K., 1976. Radial aragonite ooids, Lightly, R. G., 1985. Preservation of internal reef porosity Lizard Island, Great Barrier Reef, Queensland, Austra- and diagenetic sealing of submerged early Holocene lia. Geology, 4:120 122. barrier reef, southeast Florida shelf. In: N. Schneider- Denham, C. R. and Scotese, C. R., 1988. Terra Mobilis -- a mann and P. M. Harris (Editors), Carbonate Cements. plate tectonics program for the macintosh: v. 2.1. Earth Soc. Econ. Paleontol. Mineral. Spec. Publ., 36:123 153. in Motion Technol., Houston, Tex. Loreau, J. P. and Purser, B. H., 1973. Distribution and Douglas, J. G. and Williams, G. E., 1982. Southern Polar ultrastructure of Holocene ooids in the Persian Gulf. In: Forests: The early Cretaceous Floras of Victoria and B. H. Purser (Editor), The Persian Gulf. Springer, New their palaeoclimatic significance. Palaeogeogr., Palaeo- York, N.Y., 471 pp. climatol., Palaeoecol., 39: 171-185. MacIntyre, W. G., 1965. The temperature variation of the DuToit, A. L., 1937. Our Wandering Continents. Oliver and solubility product of calcium carbonate in sea water. Boyd, London, 366 pp. Fish. Res. Board Can., Ottawa, Rep., 200. 153 pp. Edmond, J. M. and Giesky, J. M. T. M., 1970. On the Mackenzie, F. T. and Pigott, J. D., 1981. Tectonic controls calculation of the degree of saturation of sea water with of Phanerozoic sedimentary rock cycling. J. Geol. Soc. respect to calcium carbonate under in situ conditions. Lond., 138: 183-196. Geochim. Cosmochim. Acta, 34: 1261-1291. Major, R. P., Halley, R. B. and Lukas, K. J., 1988. Emery, K.O., 1968. Relict sediments on continental shelves Cathodoluminescent bimineralic ooids from the Pleisto- of the world. Am. Assoc. Pet. Geol. Bull., 52: 445-464. cene of the Florida continental shelf. Sedimentology, 35: Fabricius, F. H., Berdau, D. and Munnich, K. O., 1970. 843-855. Early Holocene ooids in modern littoral sands reworked Marshall, J. F. and Davies, P. J., 1975. High-magnesium from a coastal terrace, southern Tunisia. Science, 169: calcite ooids from the Great Barrier Reef. J. Sediment. 757-760. Petrol., 45:285 291. Feelly, R. A., Byrne, R. H., Betzer, P. R., Gendron, J. F. and Milliman, J. D., 1969. Carbonate sedimentation on four Acker, J. G., 1984. Factors influencing the degree of southwestern Caribbean atolls and its relation to the saturation of the surface and intermediate waters of the ~'oolite problem". Trans. Gulf Coast Assoc. Geol. Soc., North pacific Ocean with respect to aragonite. J. 19: 195-206. Geophys. Res., 89: 10,631-10,640. Milliman, J. D., 1974. Marine carbonates. In: J. D. Fischer, A. G., 1981. Climate oscillations in the biosphere, Milliman, G. Mueller, and U. Foerstner (Editors), in biotic crises. In: M. Nitecki (Editor), Ecological and Recent Sedimentary Carbonates. Springer, New York, Evolutionary Time. Academic Press, New York, N.Y., N.Y., 1, 375 pp. pp. 103-131. Milliman, J. D. and Barretto, H. T., 1975. Relict magnesian Fischer, A. G., 1984. The two Phanerozoic supercycles. In: calcite oolite and subsidence of the Amazon shelf. W. A. Berggren and J. A. Van Couvering (Editors), Sedimentology, 22: 137-145. Catastrophes and Earth History -- The New Uniformi- Morse, J. W., Mucci, A. and Millero, F. J., 1980. The tarianism. Princeton Univ. Press, Princton, N.J., solubility of aragonite in seawater of 35% salinity at pp. 129-150. 25°C and atmospheric pressure. Geochim. Cosmochim. Frakes, L. A and Francis, J. E., 1988. A guide to Acta, 44: 85-94. Phanerozoic cold polar climates from high-latitude ice- Mucci, A., 1983. The solubility of calcite and aragonite in rafting in the Cretaceous. Nature, 333: 547-549. seawater at various salinities, temperatures and 1 Given, R. K. and Wilkinson, B. H., 1985. Kinetic control of atmosphere total pressure. Am. J. Sci., 283: 780-799. morphology, composition, and mineralogy of abiotic sedi- Mucci, A., Cancuel, R. and Zhong, S., 1989. the solubility of mentary carbonates. J. Sediment. Petrol., 55: 109-119. calcite and aragonite in sulfate-free seawater and the 148 B.N. OPDYKE AND B. H. WILKINSON

seeded growth kinetics and composition of the precipi- chemical Models, Ozone and Temperature Measure- tates at 25°C. Chem. Geol., 74: 309-320. ments. Am. Inst. Phys., New York, N.Y., pp. 83-109. Neiburger, M., Edinger, J. G. and Bonner, W. D., 1982. Takahashi, T., Broecker, W. S., Bainbridge, A. E. and Understanding Our Atmospheric Environment. Free- Weiss, R. F., 1980. Carbonate chemistry of the Atlantic, man, San Francisco, 2nd ed., pp. 453. Pacific, and Indian Oceans -- the results of the Opdyke, B. N. and Wilkinson, B. H., 1989. Surface area GEOSECS expeditions 1972-1978. Lamont-Doherty control of shallow cratonic to deep marine carbonate Geol. Obs. Tech. Rep., 1, CU-1-80, Palisades, N.Y., 211 pp. accumulation. Palaeoceanography, 3: 685-705. Takahashi, T., Chipman, D. and Volk, T., 1982. Geographi- Opdyke, N. D., 1959. The impact of paleomagnetism on cal, seasonal, and secular variations of the partial palaeoclimatic studies. Int. J. Bioclimatol. Biometerol., 3 pressure of CO2 in surface waters of the North Atlantic (4A): 1-11. Ocean -- the results of the North Atlantic TTO program. Proc. Carbon Dioxide Res. Conf., Carbon Dioxide, Opdyke, N. D., 1962. Paleoclimatology and continental Science and Consensus. Dep. Energy,. Washington, D.C., drift. In: S. K. Runcorn (Editor), Continental Drift. pp. 123-145. Academic Press, New York, N.Y., pp. 41-65. Tucker, M. E., 1981. Sedimentary Petrology: An Introduc- Plath, D. C. and Pytkowicz, R. M., 1980. The solubility of tion, In: A. Hallam (Editor), Geoscience Texts. Wiley, aragonite in seawater at 25.0°C and 32.62%o salinity, New York, N.Y., 3, p. 252. Mar. Chem., 10: 3-7. Tucker, M. E., 1984. Calcitic, aragonitic and mixed Parrish, J. T., 1982. Upwelling and petroleum source beds, aragonitic ooids from the mid-Proterozoic Belt Super- with reference to Paleozoic. Bull. Am. Assoc. Pet. Geol., group, Montana. Sedimentology, 31: 627-644. 66: 750-774. Vail, P. R., Mitchum, R. M., Todd, R. G., Widmier, J. M., Parrish, J. T., 1985. Latitudinal distribution of land and Thomson, S., Sangree, J. B., Bubb, J. N. and Hatleild, W. shelf and absorbed solar radiation during the Phanero- G., 1977. Seismic stratigraphy and global changes of sea zoic, U.S. Geol. Surv. Open File Rep., 803-827, p. 1-21. level, In: C. E. Payton (Edior), Seismic Stratigraphy -- Richter, D. K., 1976. Gravitational meniscus cement in a Applications to Hydrocarbon Exploration. Mem. Am. Holocene oolite near Neapolis (South-Peloponnes, Assoc. Pet. Geol., 26: 49-212. Greece). Neues Jahrb. Pal~iontol. Abh., 151: 192-223. Van der Voo, R., 1988. Paleozoic paleogeography of North Sandberg, P. A., 1983. An oscillating trend in Phanerozoic America, Gondwana, and intervening displaced terrains

non-skeletal carbonate mineralogy. Nature, 305: 19-22. -- comparisons of paleomagnetism with paleoclimatol- Sass, E., Weiler, Y. and Katz, A., 1971. Recent lagoonal ogy and biogeographical patterns. Geol. Soc. Am. Bull., oolites in the Gulf of Suez. In: Abstr. 8th Int. Sedimentol. 100: 311-324. Congr., Heidelberg, p. 87. Walter, L. M. and Morse, J. W., 1984. Magnesium calcite Sandberg, P. A., 1985. Nonskeletal aragonite and CO2 in stabilities -- a reevaluation. Geochim. Cosmochim. the Phanerozoic and Proterozoic, In: E. T. Sundquist and Acta, 48: 1059-1069. W. S. Broecker (Editors), The Carbon Cycle and Ward, W. C. and Brady, M. J., 1973. High-energy carbonates on the inner shelf, northeastern Yucatan Peninsula, Atmospheric CO2: Natural Variations Archean to Pre- Mexico. Trans. Gulf Coast Assoc. Geol. Soc., 23: 226-238. sent (Am. Geophys. Union Monogr., 32). Am. Geophys. Wegener, A., 1915. Die Entstehung der Kontinente und Union, Washington, D.C. pp. 585-594. Ozeane, Vieweg, Braunschweig, pp. 94. Schlager, W. and James, N. P., 1978. Low-magnesium Wilson, J. L., 1975. Carbonate Facies in Geologic History. calcite limestones forming at the deep-sea floor, Tongue Springer, New York, N.Y., 471 pp. of the Ocean, Bahamas. Sedimentology, 25: 675-702. Wolfe, J. A. and Upchurch Jr., G. R., 1987. North American Scotese, C. R. and Barrett, S. F., 1989. Gondwana's nonmarine climates and vegetation during the Late movement over the South Pole during the Paleozoic: Cretaceous. Palaeogeogr., Palaeoclimatol., Palaeoecol., Evidence from lithologic indicators of climate. In: W. S. 61: 33-77. McKerrow and C. R. Scotese (Editors), Palaeozoic Worsley, T. R., Nance, R. D. and Moody, J. B., 1986. Palaeogeography and Biogeography. Geol. Soc. Lond. Tectonic cycles and the history of the Earth's biogeo- Mem., 12: 75-86. chemical and palaeoceanographic record. Palaeocean- Smiley, C. J., 1967. Paleoclimatic interpretations of some ography, 1: 233-263. Mesozoic floral sequences. Am. Assoc. Pet. Geol. Bull., Zans, V. A., 1958. The Pedro Cays and Pedro Bank. Report 51: 849-863. on the survey of the cays, 1955-1957. Jamaican Geol. Takahashi, T., Kaiteris, P., Broecker, W. S. and Bain- Surv. Dep. Bull., 3: 1-47. bridge, A. E., 1976. An evaluation of the apparent Ziegler, A. M., Hulver, M. L., Lottes, A. L. and Schmachten- dissociation constants of carbonic acid in seawater. berg, W. F., 1984. Uniformitarianism and paleoclimates Earth Planet. Sci. Lett., 32: 458-467. -- inferences from the distribution of carbonate rocks. Takahashi, T. and Azevedo, A. E. G., 1982. The oceans as a In: P. Brenchley (Editor), Fossils and Climate. Wiley, CO2 reservoir. In: Interpretation of Climate and Photo- New York, N.Y., pp. 3-27.