Extant coccolithophores are widespread in all marine photic zone environments, but the most familiar typically oceanic taxa are not found at latitudes higher than 70° and populations are most diverse at low latitudes in warm, stratified, oligotrophic, open-ocean environments. Oceanic coccolithophore biogeography defines broad latitudinal belts or zones (McIntyre and Bé, 1967; Okada and Honjo, 1973; Winter et al., 1994) distinguished by variations in assemblage composition, rather than high endemicity, as many coccolithophore species are virtually cosmopolitan. These distributions reflect the temperature and nutrient characteristics of watermasses and oceanographic features such as divergence and upwelling zones, ocean gyres, and seasonal mixing. Coastal and estuarine environments usually support different taxa, many of which are small and weakly calcified and therefore without a fossil record.
Our understanding of Early Cretaceous nannoplankton biogeography is largely based on records from the European Boreal and Tethyan epicontinental basins and Atlantic Ocean DSDP/ODP sites (Roth and Bowdler, 1981; Roth and Krumbach, 1986; Bralower et al., 1989; Mutterlose, 1992a; Mutterlose and Kessels, 2000; Street and Bown, 2000). More limited data are available for the Indian (Proto Decima, 1974; Thierstein, 1974; Bralower and Siesser, 1992; Mutterlose, 1992a; Bown, 1992), Southern (Mutterlose and Wise, 1990), and Pacific (Roth, 1973, 1981; Thierstein, 1976; Erba, 1992; Erba et al., 1995; Lozar and Tremolada, 2003) oceans. The Pacific (or Panthalassa) Ocean was the largest contiguous marine habitat on the Cretaceous Earth, but much of the seafloor has since been lost through subduction.
Street and Bown (2000) concluded that Early Cretaceous nannoplankton biogeography was characterized by a wide, low- to mid-latitude zone (50°N–50°S) of relatively stable assemblage composition and diversity, flanked in both hemispheres by distinct high-latitude zones significantly lower in diversity and dominated by typically bipolar taxa (see also Mutterlose and Kessels, 2000). Further differentiation is apparent based on the distribution of rare taxa, many of which have been labeled Boreal or Tethyan based on their distributions in the European–Atlantic area.
Neritic vs. oceanic differentiation is often acknowledged but rarely unequivocally demonstrated in Cretaceous studies (Thierstein, 1976; Roth and Krumbach, 1986; Applegate et al., 1989). Most authors agree that Nannoconus and Micrantholithus were marginal or neritic taxa that were scarce or absent in oceanic settings (Table T6). Street and Bown (2000) suggested neritic assemblages were characterized by lower diversity, high unevenness, and common to dominant taxa that varied with latitude and that were essentially absent in oceanic settings, citing Nannoconus and Micrantholithus as the two most important neritic groups. These observations are particularly germane to this study, from which these two taxa are practically absent; however, their paleoecology remains contentious and will be discussed further below (see "Nannoconus"). As yet, there has been no explicit recognition of coastal nannoplankton in the Cretaceous, due, in no small part, to the paucity of data from such settings, itself a reflection of the poor preservation typical of such facies and the reticence to study such material.
This discussion will focus particularly on Nannoconus and Micrantholithus, as both represent major assemblage components with distinctive biogeography and their virtual absence at the Shatsky sites requires an explanation. In addition, they may provide important paleoceanographic information that has yet to be fully realized (but see discussion in Erba, 1994; Street and Bown, 2000; Herrle, 2003).
Site 1213 provides the most continuous Early Cretaceous nannoplankton record from the Pacific Ocean and gives particularly valuable insight into the Berriasian–Hauterivian evolutionary succession and paleobiogeography, an interval that followed a period of considerable taxonomic turnover at the Jurassic/Cretaceous boundary. Sites 1207, 1213, and 1214 also provide rich, well-preserved mid-Cretaceous nannofossil successions.
The great majority of nannofossil assemblages recorded from Shatsky Rise include taxa that are typical of this age; however, there are clear differences that set them apart from coeval Atlantic Ocean and European epicontinental assemblages. In terms of common assemblage components, L. carniolensis is consistently abundant, a feature that appears to be a characteristic of oceanic assemblages (Thierstein, 1976; Roth and Krumbach, 1986). Applegate et al. (1989) noted that Lithraphidites was associated with neritic taxa (Nannoconus and Micrantholithus) in Central Atlantic assemblages, but they appear to have been referring to the species L. bollii and Lithraphidites alatus and not to the most abundant and cosmopolitan species, L. carniolensis.
The absence of rare taxa is a subtle paleobiogeographic signal but significant nevertheless, as many of these species have been utilized in Boreal or Tethyan biostratigraphic schemes and their distribution attributed to Boreal/Tethyan biogeographic differentiation (e.g., Mutterlose, 1992a). The rarity or complete absence of many of these taxa from the Pacific (and actually much of the eastern Tethyan–Indian Ocean) suggests that these taxa may have been practically neritic in distribution and largely excluded from the expansive Cretaceous oceanic areas (Table T7) (see also Thierstein, 1976).
Almost certainly a number of taxa that have previously been labeled Boreal or Tethyan had truly oceanic distributions, controlled to some extent by temperature, and are better termed tropical or temperate/high-latitude taxa. The most restricted biogeographies are easiest to recognize in the fossil record, and these include the temperate species Repagulum parvidentatum, Crucibiscutum salebrosum, Seribiscutum primitivum, and Ceratolithina spp. and the tropical taxon H. irregularis. These interpretations are beautifully demonstrated on Shatsky Rise, where cold-water taxa were totally absent and the warm-water H. irregularis was periodically common to abundant.
Site 1213 also records unusual stratigraphically extended Early Cretaceous ranges for A. cylindratus, Hexapodorhabdus cuvillieri, and B. dorsetensis, species previously thought to have become extinct during the Late Jurassic (Tithonian) (see Bown and Cooper, 1998). Such observations suggest that these species withdrew to the Pacific Ocean prior to their final extinction. In addition, W. britannica, which, after the Late Jurassic, declined to trace levels in other areas, is recorded consistently and relatively commonly through much of the Berriasian–Valanginian. Again, this suggests that W. britannica became primarily a Pacific-restricted species during this interval before its Lazarus-taxon-like reexpansion of biogeographic range in the Aptian.
However, arguably the most striking aspect of the Shatsky assemblages, when compared with coeval sections in the Atlantic Ocean in particular, is the complete absence of Conusphaera and near absence of Nannoconus and Micrantholithus. The latter two genera are abundant and conspicuous components of many Atlantic and western Tethyan assemblages.
Nannoconus is a noncoccolith nannofossil group that appeared cryptically in the Tithonian and was a conspicuous component of Tethyan Early Cretaceous assemblages until a numerical decline in the Aptian (the "nannoconid crisis" of Erba [1994]). The genus survived until the Campanian but was rarely common after the Albian. They were abundantly present only in certain marine settings, notably the marginal basins of the circum-western Tethys, proto-Atlantic, and Caribbean, where they may be rock forming. There is also evidence from the Barremian of the North Sea for the occurrence of endemic nannoconid species adapted to temperate epicontinental waters (Nannoconus abundans and Nannoconus borealis) (e.g., Street and Bown, 2000). However, most of the more widespread geographic occurrences appear to have been temporally restricted (Mutterlose, 1989, 1992a) and limited to neritic settings or deep-sea settings adjacent to carbonate platforms. The claim that nannoconids were cosmopolitan, without qualification, is highly misleading.
A compilation of DSDP/ODP presence/absence data from the Pacific and Indian oceans, presented here in Table T8, shows that nannoconids were practically absent from these oceans, which made up ~80% of the Early Cretaceous marine ecosystem. Of the 48 DSDP and ODP sites that have recovered nannofossiliferous Lower to mid-Cretaceous sections, only two (Sites 463 and 465) recovered nannoconids in more than five samples. Both these occurrences are associated with allochthonous material sourced from shallow-water platforms or guyots, an observation previously made more broadly by Thierstein (1976). Many of the other, more sporadic, occurrences are similarly associated with transported shallow-water-sourced material (e.g., Sites 766, 800–802, 872, 878, and 879). Of the many hundreds of Indian Ocean samples that have been studied, only four have yielded nannoconids (pers. observ.). We are confident that this record is robust because nannoconids are highly distinctive nannofossils and many authors have made specific reference to the absence of the group (Bukry, 1971, 1973a, 1973b; Hekel, 1973; Thierstein, 1976; Erba and Covington, 1992; Bown, 1992). Nannoconids are far more frequently found in Atlantic Ocean sites, but this ocean was a narrow intra-Pangean basin in the Early Cretaceous, and sediment transport from surrounding shelves appears to have been common. Applegate et al. (1989) clearly document Atlantic Ocean sections of shelf-sourced turbidites with common Nannoconus and Micrantholithus interbedded with pelagic carbonates in which these taxa are rare or absent. Latitude also appears to have played an important part in the biogeography of nannoconids, and, in general, they were rarer or absent at latitudes >30° (Street and Bown, 2000).
This puzzling distribution pattern has led to a range of explanations concerning their biology and paleoecology, but most have noted the link with low-latitude (tropical), sediment-starved epicontinental basins and the association with braarudosphaerids (Roth and Krumbach, 1986; Mutterlose, 1989; Street and Bown, 2000). Busson and Noël (1991) went further and suggested nannoconids may have been meroplanktonic calcareous dinoflagellates, excluded from deep and anoxic settings by the water-depth constraints on cyst viability but which dominated in neritic environments during red-tide-like blooms. Erba (1994) postulated a deep photic zone paleoecology comparable to the extant Florisphaera profunda with abundance related to nutricline depth, although, contrastingly, F. profunda is a strictly open-ocean, "blue-water" species.
The reasonably comprehensive Early to mid-Cretaceous biogeographic data suggest there is now little doubt that the paleoecology of nannoconids was in some way related to water depth and latitude. Abundant nannoconid occurrences are generally limited to tropical to subtropical carbonate platforms and epicontinental basins around western Tethys and the Central Atlantic, a paleobiogeography highly unusual among calcareous nannoplankton but, interestingly, comparable to the incertae sedis protozoan calpionellids (Remane, 1985). Latitudinal control is most easily explained by temperature, but, notably, coeval tropical deep-sea sites do not yield nannoconids (Table T8; and see Applegate et al., 1989). Migration events may have occurred during warmer climate intervals (e.g., Mutterlose, 1989), but dispersal appears to have been prevented by large oceans such as the eastern Tethys, Indian, and Pacific. Extra-Tethyan occurrences of nannoconids were linked by shallow-water migration routes via epicontinental basins or across the narrow Atlantic Ocean (e.g., Mutterlose, 1989; Bown and Concheyro, 2004). The more exotic oceanic occurrences, such as the mid-Pacific Mountains, can be explained by island-hopping migration routes more typically associated with marine benthic invertebrates and not problematic for planktonic organisms. Furthermore, the Pacific distribution data (Table T8), which indicate a strong association with shallow carbonate platforms, suggest that nannoconids may have been living above and around certain Pacific atolls and guyots. The rare and sporadic occurrences at other truly oceanic sites were probably drifted or transported specimens. The almost total absence of nannoconids in the Indian Ocean may reflect the presence of fewer oceanic rises and guyots (Table T8).
The limiting role of water depth may be explained by a number of interrelated neritic factors including environmental stability, turbulence, transparency, salinity, and nutrients or even water depth itself if the organism had a benthic life-cycle stage (Busson and Noël, 1991), as do a number of extant coastal coccolithophores (e.g., Pleurochrysis). The fact that nannoconids flourished in detrital-free, tropical, carbonate shelf settings suggests water column transparency may well have been important and lends support to the deep photic-zone ecology suggested by Erba (1994). If water depth did exert a direct control on nannoconid distribution, then the effect of sea level change on nannoconid records should not be overlooked. Declines in nannoconid abundance in epicontinental settings may have been associated with sea level rises in much the same way as carbonate platform drowning suppressed carbonate-producing invertebrates. This interpretation is somewhat supported by records of platform drowning episodes that coincide with nannoconid "crisis" events previously reported by Weissert et al. (1998) but explained therein by climate-driven eutrophication.
The relationship between nannoconids and trophic resources has been discussed by a number of authors, and most have suggested that they were adapted to oligotrophic environments (Busson and Noël, 1991; Coccioni et al., 1992; Erba, 1994). This is based largely on their decline prior to black shale intervals, particularly OAE1a, that are interpreted by these authors as the product of high productivity but also the apparently inverse relationship between nannoconid and coccolith abundances. However, the black shales of the upper Albian OAE1b in southern France are considerably enriched in nannoconids relative to adjacent sediments (Kennedy et al., 2000; Herrle, 2002; Nagai et al., 2002), and the relationship between organic-rich sediments, productivity, and nannoconids is clearly not so straightforward (Herrle, 2002).
The Early Cretaceous Micrantholithus appears to have had a similar abundant neritic distribution to the nannoconids but over a broader latitudinal range (50°N–50°S) (Applegate et al., 1989; Street and Bown, 2000). Table T8 shows that Micrantholithus was practically absent from the Pacific and Indian oceans, and, when present, most of the records are sporadic trace occurrences of Albian Braarudosphaera, often associated with transported shallow-water material. The extant Braarudosphaera replaced Micrantholithus in the Aptian but was never as consistently abundant in the remaining Cretaceous and appears to have retained a similar eccentric ecology for much of its considerable history (Aptian–Holocene; 119 m.y.) (Roth and Bowdler, 1981; Parker et al., 1985; Siesser et al., 1992; Paleo-Alampay et al., 1999; Kelly et al., 2003). Living Braarudosphaera bigelowii is most common in neritic environments (Gran and Braarud, 1935) and fossil distributions are comparable, although anomalous occurrences are revealing, most notably abundant occurrences in post-Cretaceous/Tertiary boundary extinction assemblages and Braarudosphaera chalk occurrences in temperate to subtropical (20°–35°C) open-ocean sites, best documented for the Oligocene South Atlantic (Parker et al., 1985; Kelly et al., 2003) but also known from the Upper Cretaceous (Scarparo and Shimabukuro, 1997). These suggest Braarudosphaera is and was an opportunistic taxon whose oceanic occurrence may be limited to unusual neritic or oceanic conditions related to the upwelling of cool, nutrient-enriched, and/or low-salinity waters (Siesser et al., 1992).
Nannoconids and Micrantholithus are practically absent throughout the Cretaceous at the Shatsky Rise sites and are rare or absent in coeval Indian and Southern ocean sections (Table T8). The most obvious explanation for their absence at Shatsky is primary ecological exclusion by some aspect of water depth, as discussed above, and that the Pacific presented a barrier to their dispersal and habitation. This also best fits the observation of absence from the majority of Early Cretaceous deep ocean settings.
Alternatively, if these organisms were oligotrophic adapted (e.g., Erba, 1994) then it could be argued that they were excluded from the Shatsky area by high productivity that was associated with equatorial upwelling or the topography of the rise itself. Most preserved Early Cretaceous Pacific crust was close to, or drifting toward, equatorial regions, and the ubiquitous presence of chert in the Shatsky Cretaceous sections has been explained by the presence of a very broad (>30°) zone of equatorial divergence (Bralower, Premoli Silva, Malone, et al., 2002). The exclusion of nannoconids at Shatsky would have required a period of sustained upwelling and high productivity for at least 50 m.y. through >10° latitude. The presence of abundant chert and timing of chert deposition cessation does lend some support to this hypothesis; however, the interpretation of radiolarian-sourced chert as a proxy for high productivity is based on comparison with the modern ocean where diatoms dominate primary productivity and the silica cycle. This modern plankton system was probably not established until the late Eocene, and possibly as late as the Neogene, and thus the interpretation of chert as a high-productivity proxy is highly questionable when applied to pre-Modern ocean systems and especially so for the chert-dominated Mesozoic Pacific Ocean and margins (Baumgartner, 1987; Racki and Cordey, 2000; Shipboard Scientific Party, 2002b [p. 16]; Robinson et al., 2004). In addition, the relatively high diversity nannoplankton assemblages do not indicate long-term high productivity, and high-fertility nannofossil proxy species, such as Biscutum constans, are virtually absent through most of the interval in question.
There is also little evidence to suggest that Micrantholithus and Braarudosphaera would have been excluded by high productivity. The biogeography and paleoecology of the extant species B. bigelowii is undoubtedly enigmatic, but its occurrence and abundance is definitely not controlled by productivity alone. At present it appears to exhibit preferences for neritic environments but has occasionally flourished in open-ocean settings in the past (Kelly et al., 2003).
Finally, the absence of nannoconids and braarudosphaerids in all mature Cretaceous ocean basins is compelling evidence that productivity alone was not responsible for their exclusion. Rather, it appears likely that they were excluded from deep blue-water environments, living most abundantly in marginal ocean basins, epicontinental shelves, and shallow-water oceanic platforms and guyots.