'The Oligocene-Miocene Transition on Coral Reefs in the Falcón Basin (NW Venezuela)'

Home , Culebra Formation

Johnson, K G; Sánchez-Villagra, M R; Aguilera, O A (2009). The Oligocene-Miocene transition on coral reefs in the Falcón basin (NW Venezuela). Palaios, 24(1):59-69. Postprint available at: http://www.zora.uzh.ch University of Zurich Posted at the Zurich Open Repository and Archive, University of Zurich. Zurich Open Repository and Archive http://www.zora.uzh.ch Originally published at: Palaios 2009, 24(1):59-69.

Winterthurerstr. 190 CH-8057 Zurich http://www.zora.uzh.ch

Year: 2009

The Oligocene-Miocene transition on coral reefs in the Falcón basin (NW Venezuela)

Johnson, K G; Sánchez-Villagra, M R; Aguilera, O A

Posted at the Zurich Open Repository and Archive, University of Zurich. http://www.zora.uzh.ch

Originally published at: Palaios 2009, 24(1):59-69. The Oligocene-Miocene transition on coral reefs in the Falcón basin (NW Venezuela)

Abstract

The Oligocene-Miocene Transition (OMT) was an interval of regional environmental and biotic change on Caribbean reefs. During the late Oligocene, a diverse Tethyan biota contributed to extensive reef building across the region, but by the early Miocene, reef building had declined, and a regional extinction had removed up to 50% of the late Oligocene diversity. The general decline in reef building in the Caribbean has been attributed to changes in regional water quality. New collections of scleractinian reef corals from four different units in the northwestern Falcón Basin of Venezuela include distinct late Oligocene and early Miocene assemblages. The distribution of fossil coral species and reef limestones suggests that the thick carbonates of the San Luis Formation were deposited during the late Oligocene and that changing water quality during the OMT resulted in the demise of San Luis coral-reef ecosystems. Previous studies have noted an increase in surface-water productivity after the OMT in the region and have suggested oceanographic reorganization as the primary cause; this interval, however, also coincides with a switch in the outlet position of an ancestral Orinoco River. A change in the terrestrial geography of South America might have caused the regional decline in reef building by altering surface-water characteristics, just as modern Orinoco and Amazon outflows exert strong control on shallow-water habitats off the coast of northeastern South America. PALAIOS, 2009, v. 24, p. 59–69 Research Article DOI: 10.2110/palo.2008.p08-004r

THE OLIGOCENE–MIOCENE TRANSITION ON CORAL REEFS IN THE FALCO´ N BASIN (NW VENEZUELA)

KENNETH G. JOHNSON,1* MARCELO R. SA´ NCHEZ-VILLAGRA,2 and ORANGEL A. AGUILERA3 1Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 5BD, UK; 2Universita¨t Zu¨rich, Pala¨ontologisches Institut und Museum, Karl Schmid-Strasse 4, CH-8006 Zu¨rich, Switzerland; 3Universidad Nacional Experimental Francisco de Miranda, Centro de Investigaciones en Ciencias Ba´sicas, Complejo Docente Los Perozos, Carretera Variante Sur, Coro, 4101, Estado Falco´n, Venezuela e-mail: [email protected]

ABSTRACT al., 2001). Decreased water quality in these regions simultaneously inhibits carbonate production and increases rates of bioerosion resulting in The Oligocene–Miocene Transition (OMT) was an interval of region- declining carbonate accumulation (Jokiel et al., 1993). al environmental and biotic change on Caribbean reefs. During the Understanding the effects of low water quality on a regional scale has late Oligocene, a diverse Tethyan biota contributed to extensive reef recently become a more important priority for reef biologists because building across the region, but by the early Miocene, reef building modern coral reefs are in decline globally, in part due to deterioration of had declined, and a regional extinction had removed up to 50% of water quality in coastal environments associated with changing patterns the late Oligocene diversity. The general decline in reef building in of land use during the past century (Hughes et al., 2003). Similar eco- the Caribbean has been attributed to changes in regional water qual- logically rapid changes in regional water quality, however, are not un- ity. New collections of scleractinian reef corals from four different precedented in the geological past (Hallock and Schlager, 1986), and the units in the northwestern Falcoń Basin of Venezuela include distinct fossil record contains examples of the long-term consequences of rapid late Oligocene and early Miocene assemblages. The distribution of regional-scale change in water quality on ancient coastal marine ecosys- fossil coral species and reef limestones suggests that the thick car- tems (e.g., Wignall, 2004). In this paper we describe a Cenozoic example bonates of the San Luis Formation were deposited during the late of the potential impact of changing water quality on Caribbean coral-reef Oligocene and that changing water quality during the OMT resulted ecosystems during the late Oligocene and early Miocene. in the demise of San Luis coral-reef ecosystems. Previous studies have The Oligocene–Miocene transition (OMT) was an important time of noted an increase in surface-water productivity after the OMT in the regional biotic change in the Caribbean. During this time, there was ex- region and have suggested oceanographic reorganization as the pri- tinction of a widespread Oligocene Tethyan reef biota and the origin of mary cause; this interval, however, also coincides with a switch in many extant lineages of reef corals that are characteristic of the western the outlet position of an ancestral Orinoco River. A change in the Atlantic today (Vaughan, 1919; Frost, 1977; Budd, 2000). For example, terrestrial geography of South America might have caused the re- the late Oligocene biotas of Antigua (Johnson, 2007) and Puerto Rico gional decline in reef building by altering surface-water character- (Frost et al., 1983) are distinct from early Miocene units in Panama´ (John- istics, just as modern Orinoco and Amazon outflows exert strong son and Kirby, 2006), Florida (Weisbord, 1973), and Trinidad (Johnson, control on shallow-water habitats off the coast of northeastern South 2001). Typical Oligocene taxa include globally extinct genera, such as America. Antiguastrea and Astrocoenia, as well as taxa such as Diploastrea and Favites, which are now extinct in the Western Atlantic but have living representatives in the Indo-West Pacific (Budd, 2000). In contrast, Mio- INTRODUCTION cene reef-coral assemblages are dominated by lineages that comprise the Most geologists would probably agree that a principal defining char- extant western Atlantic fauna, including Porites, Montastraea, and plate- acteristic of reefs is the in situ production and accumulation of biogenic forming, agariciid genera. carbonate (Fagerstrom, 1987). In modern tropical oceans, most of the The OMT also coincides with a regional decline in reef building that biogenic carbonate in a reef framework is produced by scleractinian cor- persisted until reinvigoration of carbonate production following Plio– als, calcareous algae, and other phototrophic or mixotrophic taxa that Pleistocene environmental change in the region (Johnson et al., 2008). have the potential to secrete carbonate at high rates in suitable environ- Studies of fossil corals and other reef biota from Puerto Rico (Edinger mental conditions (Hallock, 1997). The modern distribution of coral reefs and Risk, 1994, 1995) suggest that the OMT was associated with an is in part controlled by the environmental tolerances of these organisms, increase in surface productivity that inhibited carbonate accumulation. and regions characterized by high surface productivity are less likely to Large buildups are common in the Oligocene, including the Antigua For- support high rates of carbonate accumulation than regions characterized mation of Antigua (Weiss, 1994), the White Limestone of Jamaica (Ste- by oligotrophic seas (Birkeland, 1987). Therefore, although corals and mann, 2004), and the Lares and Juana Diaz Formations of Puerto Rico coral communities can apparently thrive in habitats characterized by high (Frost et al., 1983), but similar deposits are less common in the Miocene. turbidity (e.g., Riegl and Piller, 1997; Perry, 2005), extensive coral-reef Miocene coral communities are more common in small patch reefs, fringing ecosystems and thick accumulations of reefal carbonates usually do not reefs, or siliciclastic-dominated settings in units such as the Anguilla For- develop along the eastern margins of oceans characterized by upwelling mation (Budd et al., 1995), the Tamana Formation of Trinidad (Johnson, systems and on coasts adjacent to large rivers (Wells, 1957). For example, 2001), or the La Boca and Valiente Formations of Panama´ (Coates et al., 2003; Johnson and Kirby, 2006). Reef building was suppressed in the the combined outflow of the modern Amazon and Orinoco rivers sup- Caribbean for Ͼ20 myr, but there is no evidence for declining coral presses reef growth over several thousands of kilometers off the coast of growth during this interval (Johnson and Pe´rez, 2006), suggesting that South America from central Venezuela to Fortaleza, Brazil (Spalding et lack of carbonate accumulation was not a function of reduced coral growth rates. This result is in agreement with studies on extant reef sys- * Corresponding author. tems where reef accumulation is not necessarily related to coral growth

from Oligocene and Miocene rocks from the Falcoń Basin in northern Venezuela. These corals were collected from newly discovered sections containing rich biotas that provide a detailed record of the OMT in this area. Our objectives are to (1) document the coral assemblages and compare them with a regional database of coral occurrences, (2) explore the implications of the results for the age of thick carbonates preserved in the Falcoń Basin, and (3) better understand the timing and ecological effects of the configuration of the Orinoco in order to (4) determine the relationship between within-basin environmental and biotic change and regional changes.

MATERIALS AND METHODS Stratigraphy, Age, and Environments The Falcoń Basin extends across Falcoń State and parts of Lara, Yar- acuy, and Zulia states in northwestern Venezuela (Fig. 1; see Wheeler, 1960, 1963). In the Oligocene and Miocene, the Falcoń Basin formed the western end of the larger Bonaire Basin that still extends offshore to the north of the Venezuelan coast range. During the middle–late Miocene, NW–SE compression resulted in folding, tectonic inversion, and exposure of a nearly continuous record of sedimentation from the Eocene to the Pliocene (Audemard, 2001). Subsequent right lateral movement along the Oca Fault zone has resulted in the development of a set of NW–SE- trending normal faults (Fig. 1B; see Boesi and Goddard, 1991) that define a series of horst and graben structures named (from west to east) the Dabajuro Platform, Urumaco Trough, Coro High, and La Vela Bay (Fig. 1B). The sedimentation, tectonics, and geological history of the Falcoń Basin and other basins in northern South American are the subject of reviews in Gonza´lez de Juana et al. (1980), Boesi and Goddard (1991), and Macellari (1995). In this paper, we discuss material collected during May–June 2006 from Oligocene and Miocene sections in the western margin (W in Fig. 1B) and central (C in Fig. 1B) part of the basin; global positioning coordinates for specific sites are included in Table 1. Corals were collected from a total of four stratigraphic units: the Castillo, Agua Clara, San Luis, and Pedregoso Formations. Western Falcoń Basin.—The western margin of the basin contains a thick section of upper Oligocene–lower Miocene lithologically complex siliciclastics and carbonates deposited in a range of terrestrial, fresh-water, and brackish to fully marine environments (Wheeler 1960, 1963). For this study, fossils were collected from the Castillo and Agua Clara For- mations. Within the Castillo Formation, facies relations vary laterally and FIGURE 1—Map showing localities studied. A) Caribbean region with the position of the Falcoń Basin in northern Venezuela. Inset box ϭ regional map in Figure 1B. vertically and include shales, sandstones, and conglomerates up to 1500 B) General outline of the Falcoń Basin (dashed line) showing position of major faults m thick in the western part and southwestern part of the basin. The for- (thick gray lines) and tectonic units (after Boesi and Goddard, 1991). The general mation thins eastward and northward where it intergrades with the more position of localities in the central part (C) and the northwestern (W) margin of the open-marine facies of the Churuguara Formation to the southeast (Wheel- basin are shown by dots. Precise global positioning coordinates for collections are er, 1960) and the deeper-water facies of the central part of the basin, included in Table 1. Stars ϭ major towns. The position of the Miocene sections at including the Pecaya and Pedregoso Formations (Dıáz de Gamero, 1977). Cerro La Cruz (Castillo Formation) is also marked. Along the southern and western flanks of the basin, the Castillo lies disconformably on Eocene sediments. The Castillo is interpreted to have rates (Edinger et al., 2000). Other possible mechanisms for the Miocene been deposited as an extensive coastal and marginal marine complex decline in reef development include decreasing production of carbonate within a full range of paleoenvironments including terrestrial, fluvial, tid- by organisms other than corals, increased physical removal, or dissolution al, and brackish to open marine subtidal settings. Wheeler interpreted the of carbonate. age of the Castillo Formation as extending from the late Oligocene to the In this paper we report on new occurrences of fossil scleractinian corals early Miocene (Fig. 2), which was confirmed by palynology (Lorente,

TABLE 1— List of collecting localities considered in this study. Map positions of general localities in northwestern (W) and central part (C) of the Falco´n Basin are included in Figure 1B.

Section Locality Faunule Latitude Longitude

W Cerro Guariro E Cauderalito 11Њ 0.398Ј N70Њ 20.403Ј W W Cerro Guariro W Castillo 11Њ 0.513Ј N70Њ 30.686Ј W C Cerros del Agua E San Luis 11Њ 5.912Ј N69Њ 55.928Ј W C Cerro los Ahumados San Luis 11Њ 4.39Ј N69Њ 53.595Ј W C Cerros del Agua W Pedregoso 11Њ 5.968ЈN69Њ 56.049Ј W PALAIOS FALCO´ N BASIN CORALS AND THE OMT 61

and organic-rich sediments. To the north, the formation intergrades with sandstones and conglomerates of the Patecitos and Guarabal Formations, which have been interpreted as back-reef and near-shore facies compris- ing debris shed from the emergent Paraguana high (Wheeler, 1960; Dıáz de Gamero, 1977). The Pedregoso Formation includes dark-gray calcareous shales with rhythmically interbedded limestones and rare sandstones (Wheeler, 1963). The limestone and sandstone layers are Ͻ1 m thick, and the section contains sedimentological evidence that these interbeds were deposited as debris flows or turbidites; they include clasts transported from shallow water into the basinal setting (Dıáz de Gamero, 1997). The unit is highly fossiliferous and commonly includes corals, mollusks, and larger and smaller foraminifera. Wheeler (1963) correlated the Pedregoso with the Castillo to the west, the San Luis to the northeast, and the Churuguara Formation to the south. In contrast, based on field relationships, sedimentology, and fossil content, Dıáz de Gamero (1977) interpreted the Pedregoso as a wedge-shaped sheet of debris deposited as a submarine fan in the deep-water (Ͻ1000 m) facies of the Pecaya Formation at the base of the reef preserved in the San Luis Formation. FIGURE 2—Summary correlation charts for the upper Oligocene–lower Miocene Various ages have been reported for the San Luis Formation, and a units near collecting sites in the northwestern Falcoń Basin and along the boundary detailed discussion is worthwhile to better assess the significance of the between the Urumaco Trough and Coro High (see Fig. 1). Our new interpretation of the age of the San Luis Formation as late Oligocene is indicated. Dashed lines new coral assemblages reported here from the San Luis and Pedregoso indicate boundaries between subepochs. Formations. Senn (1935) found the larger benthic foraminiferan taxa Spi- roclypeus spp. and Miogypsina hawkinsi in the upper part of the San Luis, indicating an age of early late Oligocene (W. Renema, personal 1997). The upper contact of the formation is diachronous so that the unit communication, July 2006). Wheeler (1963) reported a list of larger ben- includes younger sediments in the southern flanks of the Falcoń Basin. thic foraminifera from the unit, including Lepidocyclina (Lepidocyclina) Assemblages of larger benthic foraminifera at the type locality in the forresti Vaughan, and interpreted the assemblage to be characteristic of northern part of the unit are consistent with a late Oligocene age, while the middle–upper Oligocene. Wheeler (1963) also reported an assemblage smaller benthic foraminifera in the southern part suggest an age of lower of smaller benthic foraminifera and corals that he interpreted as Miocene, Miocene. Vertebrate and invertebrate fossils have been studied in detail and he assigned an age of Oligocene–lower Miocene to the San Luis. from the Miocene in the southern part of the basin near Cerro la Cruz The base of the San Luis Formation has not been defined, but Dıáz de (Fig. 1B) and include diverse assemblages of mollusks, crustaceans, fish, Gamero (1977) correlated the onset of significant carbonate production turtles, crocodiles, and terrestrial, aquatic, and marine mammals (Sańchez- with transported bioclastic carbonates in the Pecaya Formation. These Villagra and Clack, 2004). carbonates were deposited in the Globigerinoides primordius zone or The Agua Clara Formation is generally characterized by massive- slightly earlier, which corresponds to the lower Aquitanian stage (lower bedded, fossiliferous dark-gray shales interbedded with rare thin sand- Miocene; see Berggren et al., 1995). Wheeler (1963) correlated a set of stones and limestones (Wheeler, 1960; Dıáz de Gamero 1989). In the siliciclastics overlying the San Luis with the Agua Clara Formation; Woz- northwestern Falcoń Basin, the Agua Clara has been divided into a lower niak and Wozniak (1987) concurred, and placed the end of reef devel- Cauderalito Member and an upper Santiago Member. The Cauderalito opment in Blow (1969) zone N6 (ϭ middle Burdigalian; see Berggren Member consists of highly fossiliferous reefal limestones interbedded et al., 1995). Dıáz de Gamero (1989) disagreed with this correlation, with shales and fine sandstones (Wheeler, 1960), and the carbonate li- however, and correlated beds overlying the San Luis Formation with the thologies range from blue-gray massive crystalline units to friable yellow Querales and Socorro Formations rather than the Agua Clara, which she impure limestones. The upper Santiago member includes thick-bedded restricted to older units deposited in the Urumaco trough. Dıáz de Gam- shales typical of exposures of the Agua Clara Formation elsewhere in the ero’s analysis of planktonic foraminifera in the shales, however, is con- basin. The Cauderalito Member lies conformably on the Castillo For- cordant with those of Wozniak and Wozniak (1987) and indicates a max- mation in the northwestern part of the basin, with the contact placed to differentiate the limestones, shales, and fine-grained sandstones of the imum age of Blow zone N6 for the shales above the San Luis. This Cauderalito Member from the coarse sandstones of the Castillo Formation restricts the apparent age of the San Luis from Aquitanian–middle Bur- (Wheeler, 1963). Based on analysis of foraminiferal assemblages, the type digalian. section of the Agua Clara falls within the upper Aquitanian-Burdigalian The Pedregoso Formation is considered lower Miocene based on plank- (Dıáz de Gamero 1977, 1989), and Wheeler places the Castillo-Cauderalito tonic foraminifera and calcareous nannoplankton, ranging from the Glo- contact in the northwestern part of the basin at the Oligocene-Miocene bigerinoides primordius zone to the lower part of the Catapsydrax dis- boundary (Fig. 2). similis zone (Dıáz de Gamero, 1977). If this unit was deposited as a Coro High and Urumaco Trough.—Fossils were also collected from forereef fan adjacent to the San Luis reef, then the age of at least part of the San Luis and Pedregoso Formations, from the structurally complex the San Luis Formation is within the Aquitanian–early Burdigalian (Dıáz western margin of the Coro High in the central region of the Falcoń Basin de Gamero, 1989). Both Dıáz de Gamero (1977) and Wozniak and Woz- (C in Fig. 1B). The San Luis comprises a thick accumulation of dense niak (1987), however, supposed that reef development could have been gray, indurated limestones that form prominent ridges in the higher parts initiated in the late Oligocene, and they indicate an age of late Oligocene– of the Serrania de San Luis. Wheeler (1963) estimated the maximum early Miocene for the San Luis reef complex. thickness of the unit as Ն1000 m. A variety of carbonate facies are preserved in this formation, including reef bioherms. According to Beck Localities (1951; cited in Dıáz de Gamero, 1977), these bioherms are more common in the eastern slopes of the Serrania de San Luis, while on the western Western Falcoń Basin.—Collections from the Castillo and Agua Clara sides, they are more likely to be intercalated with siliciclastic units, marls, Formations were made near Cerro Guariro, 18 km southeast of Dabajuro 62 JOHNSON ET AL. PALAIOS

TABLE 2— The number of scleractinian coral specimens collected from four faunules in the Agua Clara, Castillo, Pedregoso, and San Luis Formations. The column labeled OM summarizes the stratigraphic distribution of each species in the Caribbean; O ϭ species restricted to Oligocene units; M ϭ taxon restricted to Miocene units; OM ϭ taxa found in both Oligocene–Miocene units. Taxa only found in the faunules described in this paper are indicated by an E. The numbers and proportions of each group of species within each faunule are included in the lowermost rows of the table.

Species San Luis Castillo Pedregoso Cauderalito OM

Astrocoenia portoricensis 0200O Astrocoenia guantanamensis 1000O Stephanocoenia duncani 0203M Stylophora afﬁnis 0214OM Stylophora granulata 0100M Stylophora undata 1010M Pocillopora arnoldi 0404M Acropora saludensis 0109OM Acropora sp.A1000E Astreopora goethalsi 0020M Siderastrea conferta 1 4 0 13 OM Porites portoricensis 0201OM Porites trinitatis 0100M Porites waylandi 2502M Porites baracoaensis 0 9 0 10 OM Goniopora imperatoris 0010OM Alveopora tampae 2700O Agathiphyllia tenuis 11300O Agathiphyllia antiguensis 1100O Goniastrea canalis 0400O Colpophyllia willoughbiensis 0100O Antiguastrea cellulosa 52700OM Montastraea imperatoris 0142OM Montastraea limbata 5013M Montastraea sp.L0300O Montastraea canalis 0 7 10 4 OM Montastraea cavernosa 0200O Solenastrea sp.B0002E Diploastrea crassolamellata 1900O Diploastrea magniﬁca 1000O Diploastrea sp.T0100O Trachyphyllia sp.1010OM Total Occurrences 33 99 21 57 Total Species Richness 13 23 8 12 Number O taxa 6 10 0 0 Number M taxa 3534 Number OM taxa 3857 Number E taxa 1001 Proportion O taxa 0.462 0.435 0.000 0.000 Proportion M taxa 0.231 0.217 0.375 0.333 Proportion O ϩ OM taxa 0.692 0.783 0.625 0.583 Proportion M ϩ OM taxa 0.462 0.565 1.000 0.917

(Fig. 1B, Table 1). Cerro Guariro is a small hill capped by a several- were not in life position, along with other skeletal debris. Although the meter-thick carbonate unit, and two assemblages of corals were collected coral heads were in contact with one another, there was no evidence of from the vicinity. The first was a diverse suite of massive, columnar, and framework development, and individual colonies were not growing on thick plate-shaped colonies collected 200 m southeast of the top of the top of one another. Instead, the lower part of the carbonate beds com- hill in the Castillo Formation. The corals in this assemblage included prised a single layer of rolled coral heads. A small sample from the shale distinctive Oligocene genera such as Antiguastrea cellulosa and Diploas- lithology was collected for micropaleontological analysis. trea crassolamellata (Table 2). The second assemblage was collected The San Luis faunule was collected from two sites (Table 1). The first from near the top of the southeastern slope of Cerro Guariro and was was a small ridge to the northeast of the Pedregoso exposure. A larger dominated by thin-branching taxa, including Porites and Acropora, along collection was obtained from the eastern and southeastern flanks of Cerro with common, small massive heads of Siderastrea (Table 2). The two los Ahumados northwest of Pecaya on the southern limb of the San Luis localities at Cerro Guariro were 500 m distant from one another and 250 Anticline. The limestones included packstones and floatstones with abun- m distant in the direction of dip. We refer to the two assemblages as the dant calcareous algae, larger benthic foraminifera, and mollusks, as well Castillo and the Cauderalito Faunules. as coral framestones and micritic limestone, all indicative of a range of Central Falcoń Basin and Coro High.—Three collections were studied shallow carbonate platform and reef framework facies. from the San Luis and Pedregoso Formations along the road from Agua Clara to Pecaya (Table 1). The Pedregoso faunule was collected from a Taxonomic Methods, Data Sets, and Analysis 10-m-thick section exposed where the road cuts through the linear ridges The collections were assigned preliminary identifications in the field, called Cerros del Agua (Fig. 3). The section included mudstones with and a subset was transported to the Natural History Museum, London for thin carbonate units and sandstones. The carbonate beds were Ͻ1 m thick preparation, imaging, and comparison with reference collections. Stan- and largely composed of corals with rounded, massive colony shapes that dard colony-level and calical characters were used within the taxonomic PALAIOS FALCO´ N BASIN CORALS AND THE OMT 63

FIGURE 4—Collector’s curves illustrating sample completeness in the four faunules. Each curve shows how species richness increases as collections are pooled one by one. The shape of the curve differs depending on the order in which collections are considered, so the plotted curves indicate median values of all permutations of input order (e.g., for the 5 collections within the Castillo faunule, there are 5 ϭ 120 possible permutations).

the OMT. Although scleractinian corals are not the preferred biostrati- graphic tools for the Cenozoic, they do have considerable stratigraphic utility because most species have restricted stratigraphic ranges in the Caribbean, especially because of the regional extinctions associated with the OMT and the Plio–Pleistocene extinctions (Vaughan, 1919; Budd et al., 1994; Budd, 2000, Johnson et al., 2008). The data set is a subset of FIGURE 3—Exposure of the Pedregoso Formation at Cerros del Agua West. A) the Caribbean Coral Database (K.G. Johnson, A.F. Budd, and J.S. Klaus, Seven-meter-thick section of massive bedded siltstones with thinner beds of sand- personal communication, 2007) and includes taxonomic lists of corals stone and limestone. B) Close-up of a 50-cm-thick bed that includes rounded colo- from the upper Oligocene Antigua and Lares Formations from Antigua nies of Montastraea imperatoris and other skeletal debris. Rock hammer ϭ 30 cm long. and Puerto Rico, the lower Miocene Culebra and Anguilla Formations from Panama´ and Anguilla (Fig. 1A), and the lower–middle Miocene Baitoa and Tamana Formations from the Dominican Republic and Trin- framework of Budd et al. (1994), supplemented with information from idad. A brief summary of these units is available in Johnson and Kirby Stemann (2004), Johnson (2001, 2007), and Budd et al. (2007). A few (2006), and more detailed information can be found in the references taxa were left in open nomenclature pending detailed description and listed in Table 3. comparison with type material. A reference collection has been deposited Cumulative sampling curves were produced to assess sampling com- at the Natural History Museum. pleteness from each of the four faunules (Fig. 4). Examination of these We compared the four Falco´n faunules with assemblages from Oligo- curves suggests that sampling is uneven and far from complete for several cene and Miocene units elsewhere in the Caribbean Basin in order to of the faunules. In addition, sampling is in general more complete for the determine whether the Falco´n material was deposited prior to or following stratigraphic units included in the Caribbean Coral Database than for the

TABLE 3— Sampling statistics for Miocene and Oligocene stratigraphic units from the Caribbean included in this study. An occurrence is a record of a taxon from a sample; each occurrence may result from one or more specimens.

Stratigraphic Unit Country Age Species Samples Occurrences Specimens References

Baitoa Formation Dominican Republic Middle Miocene 19 19 69 69 Budd et al., 1994 Tamana Formation Trinidad Lower–Middle Miocene 44 11 116 509 Johnson, 2001 Anguilla Formation Anguilla Lower Miocene 21 3 51 88 Budd et al., 1995 Emperador Limestone Panama Lower Miocene 21 18 86 188 Johnson and Kirby, 2006 Cauderalito Member Venezuela ? 12 4 21 57 this paper Pedregoso Formation Venezuela ? 8 1 8 21 this paper Antigua Formation Antigua Upper Oligocene 44 36 248 548 Johnson, 2007 Lares Formation Puerto Rico Upper Oligocene 45 33 254 774 Johnson, unpublished data, 2007 Castillo Formation Venezuela ? 23 5 49 99 this paper San Luis Formation Venezuela ? 13 3 17 33 this paper Total 115 103 734 1808 64 JOHNSON ET AL. PALAIOS

TABLE 4— Probabilistic dissimilarity matrix calculated using the approach of Raup and Crick (1979) to compare the Castillo, Cauderalito, Pedregoso, and San Luis faunules with assemblages of corals from other Oligocene and Miocene units.

Cauderalito Formations Baitoa Tamana Anguilla Culebra Member Pedregoso Antigua Lares Castillo San Luis

Baitoa — 0.299 0.567 0.768 0.269 0.078 1 0.996 0.827 0.878 Tamana 0.299 — 0.935 0.918 0.633 0.642 1 1 0.965 0.994 Anguilla 0.567 0.935 — 0 0 0.121 0.664 0.346 0.02 0.439 Culebra 0.768 0.918 0 — 0.015 0.029 0.926 0.704 0.157 0.703 Cauderalito Member 0.269 0.633 0 0.015 — 0 0.613 0.421 0 0.096 Pedregoso 0.078 0.642 0.121 0.029 0 — 0.876 0.388 0.151 0.022 Antigua 1 1 0.664 0.926 0.613 0.876 — 0 0.007 0.102 Lares 0.996 1 0.346 0.704 0.421 0.388 0 — 0.007 0.14 Castillo 0.827 0.965 0.02 0.157 0 0.151 0.007 0.007 — 0.002 San Luis 0.878 0.994 0.439 0.703 0.096 0.022 0.102 0.14 0.002 —

new collections from the Falco´n Basin, so we used a probabilistic dis- Luis and Castillo faunules fall within the upper Oligocene cluster, and similarity index (Raup and Crick, 1979) to compare assemblages (Table both the Cauderalito and Pedregoso faunules fall within the lower Mio- 4). This index differs from indices more commonly used in community cene cluster. Assuming that the OMT coincided with the Oligocene– paleoecology in that it explicitly considers sample size and the relative Miocene boundary, the presence of pre-OMT assemblages in the Castillo completeness of sampling among assemblages. The Raup and Crick dis- and San Luis faunules suggests that these units were deposited during an similarity index between two assemblages is computed by comparing the earlier interval than the Cauderalito and Pedregoso faunules. The Castillo number of taxa found in both assemblages with the distribution of ex- and San Luis faunules include abundant, massive faviid species charac- pected numbers of shared taxa if the taxa had been distributed randomly teristic of regional Oligocene assemblages, including Agathiphyllia ten- among the entire set of assemblages. The randomization distribution was uis, Antiguastrea cellulosa, and Diploastrea crassolamellata. Overall, created by 1000 permutations of taxon occurrences under the null hy- 46% and 43% of the taxa in the San Luis, and Castillo and San Luis, pothesis that the change of any particular species occurring in an assem- respectively, are known exclusively from Oligocene rocks, and 69% and blage is proportional to the number of assemblages in which the taxon 78% of the taxa in the San Luis and Castillo faunules have been reported is observed to occur, that is, taxa are not all present in equal numbers. from Oligocene rocks in the Caribbean region (Table 2). In contrast, all The dissimilarity indices are computed as the proportion of the permu- of the taxa recovered from the Pedregoso and Cauderalito faunules have tation distribution that is less than the observed number of shared taxa. been reported from Miocene rocks elsewhere in the Caribbean, and at The resulting matrix of indices of dissimilarity was put into an average least one third of the taxa found in both faunules have only been reported linkage cluster analysis to visualize the relationships among biota. All from post-Oligocene units. Based on these results, we conclude that the analyses were performed using the R statistical programming environ- San Luis and Castillo faunules are pre-OMT and the Pedregoso and Cau- ment (ver. 2.5; see R Development Core Team, 2007). deralito faunules are post-OMT.

RESULTS DISCUSSION A total of 34 taxa were recovered from the four sections in the Falcoń The presence of distinct pre- and post-OMT assemblages within sec- Basin (Fig. 5, Table 2), including 32 zooxanthellate and 2 azooxanthellate tions in both the northwestern Falcoń Basin and the margins of the Coro taxa (as defined following the terminology of Schuhmacher and Zibrow- ius, 1985). Four of the taxa have been left in open nomenclature, includ- High has significant implications for the local stratigraphy and paleoge- ing species of Acropora, Montastraea, Solenastrea, and Diploastrea. ography, as well as the timing and regional effects of the OMT. Species-level identification was not attempted for poorly preserved molds Age of Pre-OMT and Post-OMT Faunules.—Comparison of the reef- of Trachyphyllia. coral assemblages from the Falcoń basin with assemblages from the Species richness was higher in assemblages from the San Luis and the Caribbean region indicates a late Oligocene age for the pre-OMT assem- Castillo faunules, and these units included abundant massive taxa and blages from the Falcoń Basin and an early Miocene age for the post- less-abundant thin branching units, suggesting deposition in relatively OMT assemblages in the basin. The earliest record of post-OMT assem- high-energy habitats. In contrast, the Cauderalito fauna includes abun- blages with ages constrained by multiple taxonomic groups is from the dant, thin branching Acropora, Stylophora, and Porites species. Sedi- Emperador limestone in the Culebra Formation of Panama´ (Johnson and mentology of the unit and the local abundance of thin branching colonies Kirby, 2006), and this unit has been placed in the Catapsydrax stainforthi and heads of Siderastrea conferta suggested that corals were living in a zone of Blow (1969; also see van den Bold, 1973). Based on these results, small patch reef in a turbid, protected setting characterized by generally and studies of the distribution of corals from other late Oligocene and unfavorable conditions. Extant Siderastrea species and thin branching early Miocene sites (Budd et al. 1995; Budd, 2000; Johnson, 2001; John- Porites furcata are characteristic of such turbid habitats in the south- son and Kirby 2006; Johnson, 2007), it is extremely unlikely that the western Caribbean (Guzmań and Guevara, 1998). observed distinct compositional variation between the pre-OMT and post- Examination of the dissimilarity indices calculated among the Falcoń OMT faunules resulted from facies control rather than taxonomic turn- faunules and other units from the Caribbean region suggests significant over in the regional species pool. Coexistence of distinct facies-controlled temporal variation in the distribution of reef corals during the Oligocene assemblages has not been documented in other Oligocene–early Miocene and Miocene (Table 4). Both Oligocene assemblages from the Antigua units, in part because these two facies types rarely coexist in the Oligo- and Lares Formations and early Miocene assemblages from the Culebra cene and Miocene record in the Caribbean. To date, few extensive reef and Anguilla Formations are indistinguishable from one another at a limestones have been discovered in early Miocene sections from any- probability level of 0.001. Strong stratigraphic grouping is shown on the where in the Caribbean. The overwhelming majority of early Miocene cluster-analysis results (Fig. 6), with the upper Oligocene, lower Miocene, coral assemblages from the Caribbean have been recovered from thin and lower–middle Miocene units each forming coherent clusters. The San limestones deposited in siliciclastic settings, while most late Oligocene PALAIOS FALCO´ N BASIN CORALS AND THE OMT 65

FIGURE 5—Selected coral species recovered from the Agua Clara, Castillo, Pedregoso, and San Luis Formations and deposited at the Natural History Museum, London. A–G, I–J, L ϭ colony surface; H, K, M ϭ polished transverse sections. A) Astrocoenia portoricensis (NHM AZ 3118). B) Stephanocoenia duncani (NHM AZ 3120). C) Stylophora afﬁnis (NHM AZ 3130). D) Acropora saludensis (NHM AZ 3105). E) Acropora species A (NHM AZ 3137). F) Siderastrea conferta (NHM AZ 3104). G) Porites baracoaensis (NHM AZ 3108). H) Agathiphyllia tenuis (NHM AZ 3134). I) Agathiphyllia antiguensis (NHM AZ 3136). J) Goniastrea canalis (NHM AZ 3122). K) Antiguastrea cellulosa (NHM AZ 3115). L) Montastraea canalis (NHM AZ 3114). M) Diploastrea crassolamellata (NHM AZ 3111). Scale bars in A–B, F ϭ 5 mm; all others ϭ 10 mm. 66 JOHNSON ET AL. PALAIOS

two units are distinct, and the corals from the San Luis Formation are uniquely found in late Oligocene rocks in all other known sites in the Caribbean region. Therefore, we assign a restricted late Oligocene age to the San Luis Formation. The Pedregoso Formation has been previously correlated with the first half of the lower Miocene (Dıáz de Gamero, 1989) and includes a post-OMT coral assemblage similar to early Mio- cene assemblages from Anguilla, the Dominican Republic, and Panama´. This age is consistent with the correlation of a sample from the Cerros del Agua locality with calcareous nannoplankton zones NN1–NN2 (J. Young, personal communication, July 2006). History of River Drainage in the Area, the Demise of the San Luis Reefs, and the OMT.—The modern drainage of the region is dominated by the Orinoco River, which flows from west to east and drains a large section of the northern Andes and Guyana Shield. This river system has had a complex history, largely because the tectonic history of northern South America has been complicated. The precise details of this history remain controversial, but the rise of the Andes in western and northern South America must have resulted in substantial changes in the regional drainage, and the sedimentary basins of northern Venezuela contain a record of these changes that extends into the Paleogene (Macsotay et al., FIGURE 6—Dendrogram illustrating the results of an average linkage cluster anal- 1998). ysis of Oligocene–Miocene units and the coral faunules from the Falcoń Basin. Some authors suggest that during the Eocene and Oligocene a proto– Orinoco River drained into the Maracaibo or Llanos Basins, to the west assemblages originate from thick bioclastic limestones deposited in hab- and south of the Falcoń Basin, but that the outlet shifted eastward to the itats characterized by oligotrophic sea water (Johnson et al., 2008). This Falcoń Basin by the early Miocene (Dıáz de Gamero, 1996). A record of difference has been interpreted as a result of regional decline in water this shift is contained in the Miocene sections of Castillo Formation, in quality caused by paleoceanographic changes during the OMT (Edinger particular the deposits exposed at Cerro la Cruz (Sańchez-Villagra and and Risk, 1994, von der Heydt and Dijkstra, 2005). Clack, 2004), which include a thick section of sediments deposited in Our interpretation of a late Oligocene age for the San Luis reef does marginal marine, freshwater, and terrestrial environments. Overall, there not conflict with the water-quality hypothesis. In contrast, the early Mio- is a large body of geological and paleontological evidence from the Mio- cene age suggested by previous authors for the reefal deposits of the San cene supporting the presence of a large river system that flowed from the Luis Formation would contradict the water-quality hypothesis because far south in western Amazonia into the Andean foreland basin, which continuous reef building throughout the OMT in the Falcoń Basin would would have drained into the Caribbean in northern Venezuela (Hoorn, suggest that water quality did not decline in that part of the region. Based 1993, 1994; Hoorn et al., 1995; Lundberg et al., 1998). Several authors on the distribution of coral species in the studied sections, however, we (e.g., Rod, 1981; Lundberg et al., 1988; Aguilera, 1994; Dıáz de Gamero, consider the Miocene age to be incorrect. Our proposed late Oligocene 1996) discuss evidence for a major outlet of a paleo-Orinoco system into age for the San Luis Formation is based on the observation that contem- the Caribbean from northwestern Venezuela. poraneous pre- and post-OMT coral assemblages have not been reported This river system would have contained a common freshwater fauna, from any site in the Caribbean Region, and pre-OMT assemblages always including fish species represented as fossils in Falcoń Basin Miocene occur in upper Oligocene units. rocks and from the Acre region in southwestern Amazonia (Aguilera, Upper Oligocene and Lower Miocene Stratigraphy.—The presence of 1994; Aguilera and Rodrigues de Aguilera, 2003; Lundberg and Aguilera, distinct late Oligocene and early Miocene coral assemblages in the Falcoń 2003; Sańchez-Villagra et al., 2003; Sańchez-Villagra and Clack, 2004; Basin has implications for the stratigraphy of units from which the corals Sańchez-Villagra and Aguilera, 2006; Sabaj Pe´rez et al., 2007). These were collected. In the northwestern margin of the basin, Wheeler placed fossils include large specimens of the distinctive pimelodid catfish genus the contact between the Castillo Formation and the Cauderalito Member Phractocephalus, such as P. nassi from the late Miocene Urumaco For- of the Agua Clara Formation at the Oligocene-Miocene boundary (Fig. mation in the Falcoń Basin (Lundberg and Aguilera, 2003). Other fossils 2). The present study confirms this interpretation, at least at Cerro Guar- from this exclusively freshwater genus have been described from the iro. middle Miocene La Venta fauna of central Colombia (Lundberg, 1997) Previous studies (Wheeler, 1963; Wozniak and Wozniak, 1987; Dıáz and late Miocene fluvial sediments of Acre, Brazil (Bocquentin et al., de Gamero, 1989) have correlated the San Luis Formation with upper 1997). These occurrences demonstrate a previous trans-Andean distribu- Oligocene–lower Miocene units elsewhere in the Falcoń Basin (Fig. 2); tion for this genus, even though the single extant species is currently however, we find no evidence for lower Miocene corals in this unit. The restricted to the lowlands of the Amazon, Orinoco, and Essequibo river new coral data do support the Oligocene part of Wheeler’s (1963) age basins (Aguilera et al., 2008). These large fossil catfishes provide im- for the San Luis Formation, which was based on larger benthic forami- portant corroborating evidence for ancient large rivers draining the An- nifera in the formation. In the same paper, Wheeler extended the range dean foreland basin northward to the proto-Caribbean Sea that needs to of the formation into the Miocene based on correlation with the Guarabal be considered when evaluating alternate models for the evolution of Formation and a small coral fauna that included Montastraea canalis, drainage patterns in South America during the late Cenozoic. Porites toulai, Stylophora imperatoris, and Stylophora panamensis. Each In contrast to the paleontological data, paleogeographic reconstructions of these taxa or related species with very similar morphologies have all based on structural, stratigraphic, and sedimentological data from exposed been recorded from Oligocene rocks in the region since Wheeler’s study, and subsurface units from the Maracaibo Basin and other basins of north- so the occurrence of these corals does not provide strong support for a western Venezuela and northeastern Colombia (Stephan et al., 1990; Vil- Miocene age for the upper part of the San Luis Formation. The other lamil, 1999; Guzmań and Fisher, 2006) find no evidence for deposition evidence for a lower Miocene age for part of the formation was the of thick and extensive deltaic deposits that would be expected with the correlation of the San Luis with the Pedregoso Formation (Dıáz de Gam- outflow of a major river system. The authors of these reconstructions ero, 1989). We show here, however, that coral assemblages from these instead conclude that during the Oligocene and Miocene, the drainage of PALAIOS FALCO´ N BASIN CORALS AND THE OMT 67 northern South America was isolated from main continental drainage by remains to be determined after continued detailed study of the stratigra- highlands underlain by the Lara Nappes, so that the ancestral Orinoco phy, sedimentology, and paleontology of the late Cenozoic sedimentary drainage flowed eastward but well to the south of the Falcoń Basin as basins of northern South America. early as the early Oligocene, and that subsequently there was never a connection between the southwestern Caribbean and a large, persistent ACKNOWLEDGMENTS ancestral Orinoco River. This hypothesis remains to be tested by detailed documentation of the uplift history of the Lara Nappes. We gratefully acknowledge the field assistance of Roberto Lozsań and This controversy cannot be resolved here; it would require a regional Rodolfo Sańchez, laboratory work of Dione Rodrigues de Aguilera, cur- synthesis of the structural geology, stratigraphy, and paleoenvironmental ation of samples by J. Darrell, and photographic work of the Imaging data for northern Venezuela within a refined stratigraphic framework. The Department, Natural History Museum, London. Previous versions of the alternate hypotheses, however, each have implications for mechanisms manuscript were improved by detailed reviews by F. Audemard, E. Edin- controlling reef development in the Falcoń Basin and potentially over ger, and O. Macsotay, although we accept full responsibility for our in- larger spatial scales during the regional decline of reef building associated terpretations. The Department of Palaeontology at the Natural History with the OMT. During the late Oligocene, conditions in the Falcoń Basin Museum, London, the Percy Sladen Memorial Fund of the Linnean So- were amenable to the production and accumulation of thick and geo- ciety, and the National Geographic Society provided funding for this pro- graphically extensive reef-framework deposits. This indicates that no sig- ject. nificant river outflow could have existed in proximity to the Falcoń Basin during the late Oligocene, an interpretation that is consistent with both REFERENCES alternate reconstructions. During this time, either the basin was isolated from the Maracaibo and Barinas-Apure Basins by topographic highs AGUILERA, O.A., 1994, Ictiofauna neo´gena del noroeste de Venezuela y su relacioń (Guzmań and Fisher, 2006), or the outlet of the river was far to the south con el paleo-Orinoco y el paleo-Caribe: Unpublished Ph.D. dissertation, Univer- of the Falcoń Basin. In the second alternative, the thick deposits of Car- sidad Central de Venezuela, Caracas, 139 p. AGUILERA, O.A., BOCQUENTIN, J., LUNDBERG, J.G., and MACIENTE, A., 2008, A new bonera Formation in the Los Llanos Basin of Colombia 500 km to the cajaro catfish (Siluriformes: Pimelodidae: Phractocephalus) from the late Miocene south may represent the former outlet (Dıáz de Gamero, 1996). of Southwestern Amazonia and its relationship to Phractocephalus nassi of the Extensive reef building had declined in the Falcoń Basin by the early Urumaco Formation, in Sańchez-Villagra, M.R., and Aguilera, O.A., eds., Verte- Miocene, suggesting a change in local environmental conditions. Did this brate Palaeontology in Venezuela: Palaöntologische Zeitschrift, v. 82, p. 231–254. decline result from changes in the position of the outflow of the proto- AGUILERA, O.A., and RODRIGUES DE AGUILERA, D., 2003, Two new otolith-based Orinoco, or was it part of a Caribbean-wide environmental change re- sciaenid species of the genus Plagioscion from South American Neogene marine sediments: Journal of Paleontology, v. 77, p. 937–942. sulting in the regional decline of reef building during the OMT? A variety AUDEMARD, F.A., 2001, Quaternary tectonics and present stress tensor of the inverted of global and regional-scale environmental events are associated with the northern Falcoń Basin, northwestern Venezuela: Journal of Structural Geology, v. OMT, and many have been implicated as primary or secondary causes of 23, p. 431–453. the observed biotic change. For example, a rapid decline in atmospheric BECK, R.H. 1951. Geological survey of the San Luis Range (Central Falcoń). Com- carbon dioxide concentrations (Pagani et al., 2005) ended a prolonged pan˜ia Shell de Venezuela, unpublished report. period of globally warm climates during the late Oligocene warm period, BERGGREN, W.A., KENT, D.V., SWISHER, C.C., and AUBRY, M.-P., 1995, A revised Cenozoic geochronology and chronostratigraphy, in Berggren, W.A., Kent, D.V., resulting in a brief but intense episode of Antarctic glaciation near the Aubry, M.-P., and Hardenbol, J., eds., Geochronology, Time Scales, and Global Oligocene-Miocene boundary (Zachos et al., 2001). Analyses of fully Stratigraphic Correlation: SEPM (Society for Sedimentary Geology) Special Pub- coupled climate models hypothesize that reconfiguration of ocean gate- lications, v. 54, p. 129–212. ways in both high and low latitudes would result in a flow reversal across BIRKELAND, C., 1987, Nutrient availability as a major determinant of differences the Panama Seaway in the early Miocene that would have brought cooler among coastal hard-substratum communities in different regions of the tropics, in Birkeland, C., ed., Differences between Atlantic and Pacific Tropical Marine water into the Caribbean (Maier-Reimer et al., 1990; von der Heydt and Coastal Ecosystems: Community Structure, Ecological Processes, and Productiv- Dijkstra, 2005). Edinger and Risk (1995) showed that coral taxa tolerant ity: UNESCO, Paris, p. 45–90. of low temperature and high turbidity were more likely to survive the BLOW, W.H., 1969, Late Middle Eocene to Recent planktonic foraminiferal biostra- OMT than taxa tolerant of turbidity alone. This oceanic source of de- tigraphy, in Bro¨nnimann, P., and Renz, H.R., eds., Proceedings of the First Inter- clining water quality must be implicated for the demise of the San Luis national Conference on Planktonic Microfossils (Geneva, 1967): vol. 1, E.J. Brill, reefs if we accept the hypothesis that the ancestral Orinoco River never Leiden, p. 199–421. BOCQUENTIN, J., JE´ GU, M., and BRITO, P., 1997, An extinct Phractocephalus species drained into the southern Caribbean through the Falcoń Basin. (Siluriformes, Pimelodidae) from the Mio–Pliocene Solimo˜es Formation of Acre If we accept the alternate model that the outflow of a significant river State, Brazil, in International Symposium on Phylogeny and Classification of Neo- was discharging into the southern Caribbean through the Falcoń Basin, tropical Fishes, Abstracts: Museo de Ciencias e Tecnologıá: PUCRS (Pontificia however, as suggested by the Miocene sedimentary and fossil record of Universidade Cato´lica do Rio Grande do Sul), Porto Alegre, Brazil, p. 56. the basin, then it is likely that this land-based source of declining water BOESI, T., and GODDARD, D., 1991, A new geologic model related to the distribution quality was responsible for the demise of the San Luis reef system. Fur- of hydrocarbon source rocks in the Falcoń Basin, Northwestern Venezuela, in Biddle, K.T., ed., Active Margin Basins: AAPG (American Association of Petro- thermore, the effects of this outflow are likely to have had regional-scale leum Geologists) Memoir no. 52, p. 303–319. impacts on carbonate systems throughout the Caribbean. In terms of dis- BUDD, A.F., 2000, Diversity and extinction in the Cenozoic history of Caribbean charge, the modern Orinoco is the third largest river in the world (36,000 reefs: Coral Reefs, v. 19, p. 25–35. m3·sϪ1; see Lewis and Saunders, 1989), and an extensive delta complex BUDD, A.F., FOSTER, C.T., JR., ADRAIN, T., and DAWSON, J.P., 2007, Neogene Marine has formed where the river drains into the western Atlantic to the south Biota of Tropical America: http://nmita.geology.uiowa.edu. Checked September 2008. of Trinidad (Fig. 1). The zone of increased productivity associated with BUDD, A.F., JOHNSON, K.G., and EDWARDS, J.C., 1995, Caribbean reef coral diversity discharge of the Orinoco and Amazon covers an area of ocean equivalent during the early to middle Miocene: An example from the Anguilla Formation: in size to the modern Caribbean Basin (Hu et al., 2004). Poor water Coral Reefs, v. 14, p. 109–117. quality (low salinity, high turbidity) inhibits the development of signifi- BUDD, A.F., STEMANN, T.A., and JOHNSON, K.G., 1994, Stratigraphic distributions of cant carbonate buildups, including coral reefs, within a broad zone ad- genera and species of Neogene to Recent Caribbean reef corals: Journal of Pale- jacent to the delta, and fluctuations in discharge of the Orinoco have ontology, v. 68, p. 951–977. COATES, A.G., AUBRY, M.-P., BERGGREN, W.A., COLLINS, L.S., and KUNK, M., 2003, measurable effects on water quality throughout the eastern Caribbean Early Neogene history of the Central American arc from Bocas del Toro, western (Hellweger and Gordon, 2002; Watanabe et al., 2002). Whether or not a Panama: Geological Society of America Bulletin, v. 115, p. 271–287. proto-Orinoco exerted such a large influence on early Miocene habitats DıÁZ DE GAMERO, M.L., 1977, Estratigrafıá y micropaleontologıá del Oligoceno y 68 JOHNSON ET AL. PALAIOS

Mioceno inferior del centro de la Cuenca de Falcoń, Venezuela: GEOS, v. 22, p. JOHNSON, K.G., and PE´ REZ, M.E., 2006, Skeletal extension rates of Cenozoic Carib- 2–50. bean reef corals: PALAIOS, v. 21, p. 262–271. DıÁZ DE GAMERO, M.L., 1989, El Mioceno temprano y medio de Falcoń septentrional: JOKIEL, P.L., HUNTER, C.L., TAGUCHI, S., and WATARAI, L., 1993, Ecological impact GEOS, v. 29, p. 25–35. of a fresh-water ‘‘reef kill’’ in Kaneohe Bay, Oahu, Hawaii: Coral Reefs, v. 12, DıÁZ DE GAMERO, M.L., 1996, The changing course of the Orinoco River during the p. 177–184. Neogene: A review: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 123, LEWIS, W.M., JR., and SAUNDERS, J.F., III, 1989, Concentration and transport of dis- p. 385–402. solved and suspended substances in the Orinoco River: Biogeochemistry, v. 7, p. DıÁZ DE GAMERO, M.L., 1997, Pedregoso Formation, in Committee for Stratigraphy 203–240. and Nomenclature (CIEN), ed., Stratigraphical Lexicon of Venezuela: http:// LORENTE, M.A., 1997, Castillo Formation, in Committee for Stratigraphy and No- www.pdvsa.com/lexico/lexicohi.htm. Checked September 2008. menclature (CIEN), ed., Stratigraphical Lexicon of Venezuela, http://www. EDINGER, E.N., LIMMON, G.V., JOMPA, J., WIDJATMOKO,W.,HEIKOOP, J.M., and RISK, pdvsa.com/lexico/lexicohi.htm. Checked September 2008. M.J., 2000, Normal coral growth rates on dying reefs: Are coral growth rates good LUNDBERG, J.G., 1997, Freshwater fishes and their paleobiotic implications, in Kay, indicators of reef health?: Marine Pollution Bulletin, v. 40, p. 404–425. R.F., Madden, R.H., Cifelli, R.L., and Flynn, J.J., eds., Vertebrate Paleontology EDINGER, E.N., and RISK, M.J., 1994, Oligocene–Miocene extinction and geographic in the Neotropics: The Miocene Fauna of La Venta, Colombia: Smithsonian In- restriction of Caribbean corals: Roles of turbidity, temperature, and nutrients: PA- stitution Press, Washington, DC, p. 67–91. LAIOS, v. 9, p. 576–598. LUNDBERG, J.G., and AGUILERA, O.A., 2003, The late Miocene Phractocephalus catfish EDINGER, E.N., and RISK, M.J., 1995, Preferential survivorship of brooding corals in (Siluriformes: Pimelodidae) from Urumaco, Venezuela: Additional specimens and a regional extinction: Paleobiology, v. 21, p. 200–219. reinterpretation as a distinct species: Neotropical Ichthyology, v. 1, p. 97–109. FAGERSTROM, J.A., 1987, The Evolution of Reef Communities: Wiley, New York, LUNDBERG, J.G., LINARES, O.J., ANTONIO, M.E., and NASS, P., 1988, Phractocephalus 600 p. hemiliopterus (Pimelodidae, Siluriformes) from the upper Miocene Urumaco For- FROST, S.H., 1977, Cenozoic reef systems of Caribbean: Prospects for paleoecologic mation, Venezuela: A further case of evolutionary stasis and local extinction synthesis, in Frost, S.H., Weiss, M.P., and Saunders, J.B., eds., Reefs and Related among South American fishes: Journal of Vertebrate Paleontology, v. 8, p. 131– Carbonates: Ecology and Sedimentology: AAPG (American Association of Petro- 138. leum Geologists) Studies in Geology, v. 4, p. 93–110. LUNDBERG, J.G., MARSHALL, L.G., GUERRERO, B.H., MALABARBA, M.C., and WESSE- FROST, S.H., HARBOUR, J.L., BEACH, D.K., REALINI, M.J., and HARRIS, P.M., 1983, LINGH, F., 1998, The stage for neotropical fish diversification: A history of tropical Oligocene reef tract development, southwestern Puerto Rico: Sedimenta, v. 9, p. South American rivers., in Malabarba, L.R., Reis, R.E., Vari, R.P., Lucena, Z.M., 1–144. and Lucena, C.A.S., eds., Phylogeny and Classification of Neotropical Fishes: GONZA´ LEZ DE JUANA, C., ITURRALDE DE AROZENA, J.M., and PICARD, X., 1980, Geologıá EDIPUCRS (Editora Pontificia Universidade Catolica do Rio Grande do Sul), Por- de Venezuela y de sus cuencas petroliferas: Foninves, Caracas, Venezuela, to Alegre, Brazil, p. 13–48. 1031 p. MACELLARI, C.E., 1995, Cenozoic sedimentation and tectonics of the southwestern GUZMA´ N, H.M., and GUEVARA, C.A., 1998, Arrecifes coralinos de Bocas del Toro, Caribbean pull-apart basin, Venezuela and Colombia, in Tankard, A.J., Sua´rez, R., Panama´: pt. I, Distribucioń, estructura y estado de conservacioń de los arrecifes and Welsink, H.J., eds., Petroleum Basins of South America, AAPG (American continentales de la Laguna de Chiriquı´ y la Bahıá Almirante: Revista de Biologıá Association of Petroleum Geologists) Memoir no. 62, p. 757–780. Tropical, v. 46, p. 601–622. MACSOTAY, O., VIVAS,V.,WEHRMANN, M., HARTENBERGER, J.L., and CHACHATI, B., GUZMA´ N, J., and FISHER, W.L., 2006, Early and middle Miocene depositional history 1998, Tectono-sedimentary molasse cycles along northern Venezuela and Trinidad, of the Maracaibo Basin, western Venezuela: AAPG (American Association of in Ali, W., Paul, A., and Young, V., eds., Transactions of the 3rd Geological Petroleum Geologists) Bulletin, v. 90, p. 625–655. Conference of the Geological Society of Trinidad and Tobago and the 14th Ca- HALLOCK, P., 1997, Reefs and reef limestones in earth history, in Birkelund, C., ed., ribbean Geological Conference: Geological Society of Trinidad and Tobago, Port Life and Death of Coral Reefs: Chapman & Hall, New York, p. 13–42. of Spain, v. 2, p. 584–593. HALLOCK, P., and SCHLAGER, W., 1986, Nutrient excess and the demise of coral reefs MAIER-REIMER, E., MIKOLAJEWICZ, E., and CROWLEY, T., 1990, Ocean general circu- and carbonate platforms: PALAIOS, v. 1, p. 389–398. lation model sensitivity experiment with an open Central American isthmus: Pa- HELLWEGER, F.L., and GORDON, A.L., 2002, Tracing Amazon River water into the leoceanography, v. 5, p. 349–366. Caribbean Sea: Journal of Marine Research, v. 60, p. 537–549. PAGANI, M., ZACHOS, J.C., FREEMAN, K.H., TIPPLE, B., and BOHATY, S., 2005, Marked HOORN, C., 1993, Marine incursions and the influence of Andean tectonics on the decline in atmospheric carbon dioxide concentrations during the Paleogene: Sci- Miocene depositional history of northwestern Amazoˆnia: Results of a palynostra- ence, v. 309, p. 600–603. tigraphic study: Palaeogeography, Palaeoclimatology and Palaeoecology, v. 105, PERRY, C.T., 2005, Structure and development of detrital reef deposits in turbid near- p. 267–309. shore environments, Inhaca Island, Mozambique: Marine Geology, v. 214, p. 143– HOORN, C., 1994, Fluvial palaeoenvironments in the intracratonic Amazonas Basin 161. (early Miocene–early middle Miocene, Colombia): Palaeogeography, Palaeocli- RAUP, D.M., and CRICK, R.E., 1979, Measurement of faunal similarity in paleontol- matology, Palaeoecology, v. 109, p. 1–54. ogy: Journal of Paleontology, v. 53, p. 1213–1227. HOORN, C., GUERRERO, J., SARMIENTO, G.A., and LORENTE, M.A., 1995, Andean tec- RDEVELOPMENT CORE TEAM, 2007, R: A language and environment for statistical tonics as a cause for changing drainage patterns in Miocene northern South Amer- computing, version 2.5: R Foundation for Statistical Computing, http://www. ica: Geology, v. 23, p. 237–240. r-project.org. Checked September 2008. HU, C., MONTGOMERY, E.T., SCHMITT, R.W., and MULLER-KARGER, F.E., 2004, The RIEGL, B., and PILLER, W.E., 1997, Distribution and environmental control of coral dispersal of the Amazon and Orinoco River water in the tropical Atlantic and assemblages in northern Safaga Bay (Red Sea, Egypt): Facies, v. 36, p. 141–162. Caribbean Sea: Observation from space and S-PALACE floats: Deep Sea Re- ROD, E., 1981, Notes on the shifting course of the ancient Rio Orinoco from late search, pt. 2: Topical Studies in Oceanography, v. 51, p. 1151. Cretaceous to Oligocene time: GEOS, v. 26, p. 54–56. HUGHES, T.P., BAIRD, A.H., BELLWOOD, D.R., CARD, M., CONNOLLY, S.R., FOLKE, C., SABAJ PE´ REZ, M.S., AGUILERA, O.A., and LUNDBERG, J.G., 2007, Fossil catfishes of GROSBERG, R., HOEGH-GULDBERG, O., JACKSON, J.B.C., KLEYPAS, J.A., LOUGH, J.M., the family Doradidae and Pimelodidae (Teleostei: Siluriformes) from the Miocene MARSHALL,P.,NYSTRO¨ M, M., PALUMBI, S.R., PANDOLFI, J.M., ROSEN, B.R., and Urumaca Formation of Venezuela: Proceedings of the Academy of Natural Sci- ROUGHGARDEN, J., 2003, Climate change, human impacts, and the resilience of ences of Philadelphia, v. 156, p. 157–194. coral reefs: Science, v. 301, p. 929–933. SA´ NCHEZ-VILLAGRA, M.R., and AGUILERA, O.A., 2006, Neogene vertebrates from Uru- JOHNSON, K.G., 2001, Middle Miocene recovery of Caribbean reef corals; new data maco, Falcoń State, Venezuela: Diversity and significance: Journal of Systematic from the Tamana Formation, Trinidad: Journal of Paleontology, v. 75, p. 513–526. Palaeontology, v. 4, p. 213–220. JOHNSON, K.G., 2007, Reef-coral diversity from the late Oligocene Antigua Formation SA´ NCHEZ-VILLAGRA, M.R., AGUILERA, O.A., and HOROVITZ, I., 2003, The anatomy of and temporal variation of local diversity on Caribbean Cenozoic Reefs, in Hub- the world’s largest extinct rodent: Science, v. 301, p. 1708–1710. mann, B.P., and Werner, E., ed., Fossil Corals and Sponges: Proceedings of the SA´ NCHEZ-VILLAGRA, M.R., and CLACK, J.A., 2004, Fossils of the Miocene Castillo Ninth International Symposium on Fossil Cnidaria and Porifera (Graz, Austria): Formation, Venezuela: Contributions on Neotropical Palaeontology, Special Pa- O¨ sterreichische Akademie der Wissenschaften Schriftenreihe der Erdwissenschaf- pers in Palaeontology, v. 71, p. 1–117. tlichen Kommissionen, v. 17, p. 471–491. SCHUHMACHER, H., and ZIBROWIUS, H., 1985, What is hermatypic? A redefinition of JOHNSON, K.G., BUDD, A.F., and JACKSON, J.B.C., 2008, Coral reef development was ecological groups in corals and other organisms: Coral Reefs, v. 4, p. 1–9. independent of coral diversity in the Caribbean over 28 million years: Science, v. SENN, A., 1935, Die stratigraphische Verbreitung der tertiaren Orbitoiden, mit spezieller 319, p. 1521–1522. Beru¨cksichtigung ihres Vorkommens in Nord-Venezuela und Nord-Marokko: JOHNSON, K.G., and KIRBY, M.X., 2006, The Emperador Limestone rediscovered: Eclogae Geologicae Helvetiae, v. 28, p. 51–113. Early Miocene corals from the Culebra Formation, Panama: Journal of Paleon- SPALDING, M.D., RAVILIOUS, C., and GREEN, E.P., 2001, World Atlas of Coral Reefs: tology, v. 80, p. 283–293. University of California Press, Berkeley, 424 p. PALAIOS FALCO´ N BASIN CORALS AND THE OMT 69

STEMANN, T.A., 2004, Reef corals of the White Limestone Group of Jamaica: Cai- WEISBORD, N.E., 1973, New and little-known corals from the Tampa Formation of nozoic Research, v. 3, p. 83–107. Florida: Florida Bureau of Geology, Geological Bulletin, v. 56, p. 1–147. STEPHAN, J.-F., MERCIER DE LEPINAY, B., CALAIS, E., TARDY, M., BECK, C., CARFANTAN, WEISS, M.P., 1994, Oligocene limestones of Antigua, West Indies: Neptune succeeds J.-C., OLIVET, J.-M., VILA, J.-M., BOUYSSE,P.,MAUFFRET, A., BOURGOIS, J., THERY, Vulcan: Caribbean Journal of Science, v. 30, p. 1–29. J.-M., TOURNON, J., BLANCHET, R., and DERCOURT, J., 1990, Paleogeodynamics maps WELLS, J.W., 1957, Coral reefs, in Hedgpeth, J.W., ed., Treatise on Marine Ecology of the Caribbean: 14 steps from Lias to present: Bulletin de la Socie´te´Geólogique and Paleoecology: Geological Society of America Memoirs, v. 67, p. 609–631. de France, ser. 8, v. 6, p. 915–919. WHEELER, C.B., 1960, Estratigrafıá del Oligoceno y Mioceno inferior de Falcoń occidental y nororiental: Memorias III Congreso Geolo´gico Venezolano, v. 1, p. VAN DEN BOLD, W.A., 1973, Ostracoda of the La Boca Formation, Panama Canal 407–465. Zone: Micropaleontology, v. 18, p. 410–422. WHEELER, C.B., 1963, Oligocene and lower Miocene stratigraphy of western and VAUGHAN, T.W., 1919, Fossil corals from Central America, Cuba, and Porto Rico, northeastern Falcoń Basin, Venezuela: American Association of Petroleum Geol- with an account of the American Tertiary, Pleistocene, and Recent coral reefs: ogists Bulletin, v. 47, p. 35–68. U.S. National Museum Bulletin, v. 103, p. 189–524. WIGNALL, P.B., 2004, Causes of mass extinctions, in Taylor, P.D., ed., Extinctions in VILLAMIL, T., 1999, Campanian–Miocene tectonostratigraphy, depocenter evolution the History of Life: Cambridge University Press, Cambridge, UK, p. 119–150. and basin development of Colombia and western Venezuela: Palaeogeography, WOZNIAK, J., and WOZNIAK, M.H., 1987, Bioestratigrafıá de la region nor-central de Palaeoclimatology, Palaeoecology, v. 153, p. 239–275. la Serrania de Falcoń, Venezuela nor-occidental: Boletıń de Geologıá (Caracas), VON DER HEYDT, A., and DIJKSTRA, H.A., 2005, Flow reorganizations in the Panama v. 16, p. 101–139. Seaway; a cause for the demise of Miocene corals? Geophysical Research Letters, ZACHOS, J.C., SHACKLETON, N.J., REVENAUGH, J.S., PA¨ LIKE, H., and FLOWER, B.P., 2001, v. 32, DOI: 10.1029/2004GL020990. Climate response to orbital forcing across the Oligocene–Miocene boundary: Sci- WATANABE,T.,WINTER, A., OBA,T.,ANZAI, R., and ISHIOROSHI, H., 2002, Evaluation ence, v. 292, p. 274–278. of the fidelity of isotope records as an environmental proxy in the coral Montas- trea: Coral Reefs, v. 21, p. 169–178. ACCEPTED AUGUST 5, 2008