RESULTS AND DISCUSSION

Palaeoecology

The Neogene sediments of Falcón State in northwestern Venezuela yielded 10 echinoid species. Although we cannot make any statistically well-founded statement about echinoid diversity and palaeoecology in this region, our study demonstrates the potential of the region to provide both taxonomic and palaeocological information.

Each of these localities shows a different species assemblage (Table 1). Only one of 10 species, Mellitella falconensis, occurs in two of the three sites (in both San Gregorio sites). Encope seconensis is the only other fossil species found in the Couiza East site. The Cocuiza West site comprises five species, Melitella falconensis, Encope cf. emarginata, Agassizia excentrica, Moira atropos and Arbacia punctulata. The two San Gregorio sites are situated only 1 km apart and belong to the same stratigraphic unit; therefore, we hypothesize that these sediments and fossils come from the same habitat. The difference in the species composition may be the result of lateral variation within the habitat and the low sample size.

The site in the La Vela Formation displays a different species composition to the San Gregorio sites. At this locality, four species were found, two of the genus Clypeaster and two regular echinoids Prionocidaris sp. and Lytechinus cf. euerces.

The three localities studied here are rich in other invertebrate fossils such as bivalves and gastropods. In contrast to the San Gregorio formation sites, the La Vela site preserves coral fragments indicating different habitat-properties. This is confirmed with known palaeoenvironmental records (Ministerio de Energía y Minas 1997; Smith 2008). As previously stated, the deposits of the Cocuiza Member display characters of a moderate wave energy coastline habitat and the La Vela sediments were deposited in a middle neritic habitat, indicating a depth gradient within the Falcón Basin (Smith 2008; Sánchez-Villagra et al. 2010). After analyzing the habitat preferences (Mortensen 1928-1951; Schultz 2006) of extant relatives of the sampled fossil taxa, we compared these modes of life to the species found in the Cocuiza sites, which accordingly inhabited still and shallower waters than the living representatives of the species found in the La Vela site. These habitat differences may be one possible explanation for the heterogeneous occurrence of echinoid genera in the San Gregorio and the La Vela formations. Difference in age of deposits and geological events occurring at the time may have caused the local environmental differences described above.

Palaeogeographic Implications and the Closure of the Central American Isthmus

The most significant geological event in the Caribbean Neogene was the gradual closure of the Central American Isthmus completed in the late Pliocene, about 2.8 m.y.a. (Lessios 2008). The closure was accompanied by significant global environmental changes (Cronin and Dowsett 1996). One of them, the change in the Global Ocean Conveyor, had an impact on the global climate. On a more local scale, the formation of this land bridge resulted in contrasting environmental conditions on the Pacific and the Caribbean coasts (Lessios 2008). Hence, the Caribbean waters, in contrast to the Pacific ones, are characterized by the absence of upwelling (except in Venezuela), low seasonality, increase of carbonate deposition and decrease of the plankton productivity (Aguilera et al. 2010). The relation between the age of the deposits studied here and the degree of closure of the Central American Isthmus (Coates and Obando 1996) may have affected the differing species compositions of the collections from the three localities.

The physical changes, such as fluctuations in sea level, salinity, ocean currents and temperature, caused by the closure of the Central American Isthmus, had a great impact on benthic communities (Aguilera et al. 2010). The described material contains three taxa, Mellitella falconensis, Encope secoensis and Prionocidaris, that appear to have been affected by the accompanying effects of the closure. Mellitella falconensis and Encope secoensis are known only from the Venezuelan San Gregorio Formation; therefore, they became extinct during the last few million years. The changing environment and its conditions led to extinction of other invertebrate groups (Johnson 1995; Aguilera et al. 2010; O'Dea et. al. 2007; Todd et al. 2001). The genus Prionocidaris experienced a shift of its distribution pattern from the Mediterranean and Caribbean regions, from where the fossils of this genus are known, to the Indian and Pacific Oceans, where the extant species of this genus occur (Mortensen 1928). The route and the exact time of the shift are unknown. Thus, the genus could have spread to the Pacific before the closure of the Central American Isthmus or moved to the Indian Ocean before the separation of the Mediterranean Sea and the Indian Ocean due to a drift of the African continent. The present radiation of Prionocidaris may be a consequence of both processes (Cutress 1980).

Diversity Analyses of Venezuelan and All Caribbean Fossil Echinoids

Echinoid fossils from Venezuela have been repeatedly reported during the last decades. Most of the material is from the Cenozoic era (Jeannet 1928; Cooke 1941; Von der Osten 1957; Cooke 1961; Anisgard 1954; Weisbord 1969; Mooi and Peterson 2000), but there are nine genera known from the Early Cretaceous (Cooke 1961; Cutress 1980) and one record dates back to the Permian (Hoover 1981).

Together with our new records, the previously reported Venezuelan echinoid fossils presented in this work include 38 genera collected mostly from the northern half of the country, especially from Falcón State (Figure 9, Table 2). Since Venezuela, Trinidad and Tobago together with part of the Dutch Antilles off the coast (Aruba, Curacao and Bonaire) can be considered as a geographical unit, echinoid fossil diversity in these regions was analysed as a whole. This region holds a total of 51 genera. In the following, we will refer to this whole region when we discuss the diversity of Venezuela.

For the diversity analyses, we applied several methods: we calculated the mean standing diversity with or without singletons and counted boundary crossers. The mean standing diversity approach turned out to be a reasonable estimate of the fossil record diversity through time, since it deals well with long ranging taxa and turnovers (origination and extinction) within certain time interval (Hammer and Harper 2005). To be able to better understand the echinoid evolutionary patterns in the Caribbean, the same diversity analyses were applied to echinoid faunas of other Caribbean regions: Central America, Lesser Antilles and Great Antilles (Appendix). Echinoid fossil records of all mentioned regions were also analyzed together to obtain the general echinoid diversity patterns for the whole Caribbean (Figure 10). Since Jamaica and Cuba have a rich echinoid fossil record they were analyzed separately. We are aware of a potential bias within the Cuban fossil record and its need of revision. Furthermore, the irregular echinoids clearly outnumber the regular ones, probably due to a more rigid test that facilitates preservation (Kier 1963b; Donovan 2001). Due to the lower preservation potential of regular echinoids fossils and of the lack of thorough systematic collecting of echinoid fossils, the number of known taxa is most probably far from reflecting complete sampling.

Three diversity decreases are recognizable in the overall pattern of Venezuelan echinoids (Figure 10, Table. 3): at the end of the Early Cretaceous, around the Eocene-Oligocene boundary and after the diversity peak in the Middle Miocene. The gap in the fossil record between the Early Cretaceous and the Late Eocene is most probably an artifact of biased sampling rather than of an extinction event. All of the analyzed regions show a diversity decrease either in the beginning or at the end of the Late Cretaceous. It is not clear if this is as well an artifact of biased sampling or some other unknown event. The Eocene-Oligocene boundary extinction event, caused by severe cooling and changes in oceanographic currents, is well documented (Prothero 1989; Berggren and Prothero 1992). The extinction of echinoid taxa at the Eocene-Oligocene boundary is also documented from other Caribbean echinoid faunas (McKinney et al. 1992; Dixon and Donovan 1994), and our review of the data suggests this extinction event for the first time for Venezuela. Due to different results of our diversity analyses, it is not clear, however, whether this diversity decrease in Venezuela happened exactly on the Eocene-Oligocene Boundary or somewhat later at the end of the Early Oligocene. Other Caribbean echinoid fossil records show even earlier diversity decreases, starting already at the end of the Middle Eocene (Figure 10). The third diversity decrease occurred after the beginning of the closure of the Central American Isthmus, 14 m.y.a. (Lessios 2008). As previously mentioned, the gradual closure of the Central American Isthmus was accompanied with significant environmental changes that probably triggered the extinction of some echinoid taxa. However, extinctions of other invertebrate faunas are reported closer to the end of the closure (Johnson 1995; Budd et al. 1996; Todd 2001) rather than to the beginning, in contrast to the Venezuelan and all other analyzed Caribbean echinoid fossil records, except the Lesser Antilles. This diversity decrease continues until today in Venezuela, Cuba and Central America, but not in the Lesser Antilles and Jamaica where we observe a slight diversity increase in the Early Pleistocene. Unfortunately, the complete record of the Recent Caribbean echinoids is not available (Penchaszadeh 2003; Schultz 2006; Francisco and Pauls 2008). The knowledge of this information would bring better resolution of the evolutionary patterns in last few million years.

The highest diversity peak of Venezuelan echinoids was in the Middle Miocene in contrast to other studies of Caribbean echinoid faunas that documented the highest diversity peak in the Eocene (Kier 1977; McKinney et al. 1992; Donovan 2001). This difference may be due the poor diversity data on Venezuelan Eocene echinoids.

When comparing Venezuelan echinoid diversity patterns with other Caribbean regions, one has to be aware that northern Venezuela experiences upwelling events caused by trade wind from February to April (Gilbes and Armstrong 2004). This upwelling phenomenon in Venezuela is not restricted to today but has been shown to exist with intensity variations since the Early Miocene, i.e., upwelling events occurred both before and after the closure of the Central American Isthmus (Aguilera and Aguilera 2001). Considering that this major geological event was accompanied by changes in ocean currents and that there are two other nutrient rich intrusions in the Caribbean (intrusions from Orinoco and Amazon River), it is possible that these nutrient rich waters may have reached wider parts of the Caribbean and not only the Lesser Antilles as they do today (Aguilera et al. 2010; Gilbes and Armstrong 2004). Since it is known that a combination of upwelling events and nutrient-rich intrusions have an effect on both echinoid larvae and adults, it is fair to assume that they also influence species assemblages and diversity. It would therefore be of great importance to further explore the potential of paleontological data in revealing these complex temporal and spatial nutrient patterns in the Caribbean (McAlister 2008; Smith and Stockley 2005).