A biostratigraphic revision of the Eocene and Oligocene foraminiferal type localities of Trinidad described by Cushman & Stainforth (1945) and Cushman & Renz (1948)

This study constitutes a revision and reappraisal of the samples used in the two classic studies of the calcareous benthic foraminifera of the Navet, San Fernando and Cipero Formations of Trinidad: Cushman & Renz’s 1948 study of the Eocene, and Cushman & Stainforth’s 1945 study of the Oligocene. Biostratigraphical calibration is achieved by reference to the lithological schemes of Bolli (1957) and other studies, and by directly assigning the samples to the standard planktonic foraminiferal and nannofossils zonal schemes. This direct calibration enables a more precise biostratigraphic framework for the type localities of 55 benthic foraminiferal taxa described by Cushman & Renz and Cushman & Stainforth.

greater accuracy for determining the age of these sediments.
Advances in both technology and biostratigraphy since Cushman's day makes re-examination of the faunas from these type localities worthwhile.
During a recent visit to the Smithsonian Institution, Washington D.C., one of us (MAK) traced the whereabouts of C u s h m a n ' s original samples from the Eocene a n d Oligocene of Trinidad. In total, Cushman & Renz and C u s h m a n & Stainforth described 55 n e w taxa of foraminifera from the Navet, San Fernando, and lower Cipero Formations, many of which to our knowledge have escaped the attention of subsequent workers (Appendix 1). The p r i m a r y goal of this s t u d y is to revise t h e biostratigraphy of Cushman's type samples used in these classic s t u d i e s w i t h detailed observations, thereby providing more accurate age assignments for the type localities of foraminiferal taxa described by Cushman & Renz a n d C u s h m a n & Stainforth.
Because the lithostratigraphic units in Trinidad can be precisely dated by the use of nannofossils and planktonic foraminifera, this provides an excellent opportunity to calibrate bathyal benthic foraminifera to the s t a n d a r d low-latitude biostratigraphic framework. This study is part of a revision of the Paleogene benthic foraminifera described from Trinidad.

PREVIOUS STUDIES
The rich fossiliferous deposits of Trinidad have been s t u d i e d since t h e m i d d l e of the nineteenth century. Knowledge of their micropalaeontology was already advanced by 1892, when Guppy published a general study of the "Microzoic Deposits of Trinidad" which spoke of a "series of grey marls" containing "Glohigevina and other foraminifera". Detailed foraminiferal research of the Navet and Hospital Hill marls and the Cipero Formation was undertaken subsequently by Cushman & Rena (1948) and Cushman & Stainforth (1945).

Navet Formation
In 1948, Cushman & Renz described the extremely rich foraminiferal fauna that had been deposited in "typical marine, open sea a n d w a r m water" environmental conditions similar to those of the underlying Lizard Springs Formation (Cushman & Renz, 1946) and the overlying Cipero Formation. The Navet formation was characterised by "light gray and greenish-gray, khaki-weathering marls a n d marly clays". It displayed n o "continuous o r undisturbed outcrop sections" but in southern Trinidad was seen to grade conformably down into the Lizard Springs Formation and u p into the overlying Hospital Hill marl. A stratigraphic sequence based upon faunal evidence was identified from top to bottom (see Cushman and Renz, 1948 for exact localities): Penitence Hill marl Fitt Trace -Navet River -Nariva River marl Friendship Quarry -Dunmore Hill marl Ramdat marl Cushman & Renz dated the Navet Formation as occurring between the "Upper Cretaceous" Lizard Springs Formation and below the upper Eocene Hospital Hill marl. They believed it thus to be "?Palaeocene to lower-middle Eocene". The age w a s based u p o n t h e occurrence of the genus Hantkeizina, present throughout the Navet Formation, excluding the Ramdat marl. This indicated that the Ramdat marl at the base of the Formation had an age of "older than early Eocene" (Cushman & Renz, 1948). The Paleogene Formations of Trinidad are described below.

San Fernando Formation.
The San Fernando Formation contained a rich foraminiferal fauna between the Navet and Cipero Formations but there were a number of species restricted to it. These were typified by Hail tkeiziiia alabamerzsis Cushman var. prinzitiva Cushman & Jarvis and, more rarely, Bulimina jacksoizensis Cushman, which led Cushman & Renz to regard this fauna as "typically upper Eocene" in age. Bolli (195%) believed Cushman & Renz's stratigraphic sequence to be "tentative", as it was based on "isolated, small outcrops and subsurface occurrences". In a t t e m p t i n g to p r o v i d e a planktonic zonation scheme, h e s t r o v e to counterbalance these unfavourable conditions by studying a large number of samples. Age assignments for the Navet and San Fernando Formations can be compiled by reference to Bolli (195%) w h o calibrated his s a m p l e s by m e a n s of planktonic foraminifera ( Figure 1).

I Planktonic Foraminifera1 Zone I Type Locality
Navet and San Fernando Formations and regarded them as upper lower Eocene to lower upper Eocene, and uppermost Eocene respectively. Bolli's chronostratigraphy can be updated by comparing his original scheme to the tabulated planktonic foraminiferal zonal schemes of Toumarkine & Luterbacher and Bolli & Saunders in Plankton Stratigraphy (1985), and also to the revised P-zones of Berggren & Miller (1988).
Revision of the chronostratigraphy by these authoritative studies reveals that the samples can be dated accordingly: t h e Navet Formation samples extend from Zone P10 (Friendship Quarry marl) through P11 (Dunmore Hill marl), Zone P12 (including both the Navet River and Near Fitt Trace marls) and Zone P13 (Penitence Hill marl). The age of the Hospital Hill marl proved more difficult to establish - Bolli (1957) tentatively assigned it to the T. rokri and G. semiinvoluta Zones, which correlate to Zones P14 and P15.

Cipero Formation.
Research on the Cipero Formation suggested a variety of ages. Cushman & Stainforth (1945) stated that Nuttall's (1928) dating of this formation as Oligocene to Miocene had been revised by Cushman, with other collaborators, as Eocene. C u s h m a n & Stainforth again revised the age correlating the Cipero Formation with the Oligocene. These authors subdivided the Cipero Formation into three zones: Zone I (the lower Globigerina concinna Zone) rests conformably on top of the Hospital Hill marl but does not contain its characteristic fauna of Bulimina jacksonerisis and Hantkenina alabaniensis. This zone was dated as lower Oligocene to "middle" Oligocene.
The "Bamboo Clay" unit divides Zone I from Zone I1 yielding an age of "middle" Oligocene by the occurrence of larger foraminifera of the genus Lepidocycliin~ (Vaughan & Cole, 1941).
Zones I1 and 111 [middle (Globigeriiiatella insueta) and upper (Globorotalia foksi) Zones] are positioned above the "Bamboo Clay" and were regarded as " not older than middle Oligocene." In 1957, Bolli added another zone to that of Cushman & Stainforth's scheme, and in his 1959 study of planktonic foraminifera from the Cipero Formation, Bolli regarded it as lower Oligocene to lower Miocene. The Cipero Formation is now regarded as extending from the lower Oligocene to the lower middle Miocene (Bolli & Saunders 1985) (figure 2).

Hospital ill formation Fitt Trace marl-Navet River mart
This study is based upon samples taken both from the Dunmore Hill mad-Nariva River marl original studies of Cushman

Methods
All the samples used in this study had previously been washed, and no further preparation was needed. The samples were picked, divided into benthic and planktonic foraminifera and mounted onto faunal reference slides. Preservation of the specimens was good, except in the Friendship Quarry and Dunmore Hill marl samples in which the planktonic foraminifera displayed calcitic overgrowths. Planktonic foraminifera were collected in order to facilitate a direct calibration to the standard planktonic foraminiferal zonations. For the Eocene Navet and San Fernando Formations, the zonation scheme of Berggren & Miller (1988) was used throughout this study. The same procedure was adopted for updating the biostratigraphy of the Cipero Formation samples; using zonal schemes of Bolli (1957), Bolli & Premoli Silva (1973, Bolli & Saunders (1985), against which the P-zones of Berggren & Miller (1988) were compared.
Nannofossil slides were made from each sample either by picking small fragments of loose matrix or by selecting several planktonic foraminifera and crushing them to free the matrix enclosed within. The smear preparation technique was used whereby a drop of distilled water was added to the crushed sediment on the slide, smeared and then dried on a hot plate. Ultraviolet light curing mounting medium was used to fix the coverslip. All slides are housed in the Micropalaeontological Collections of University College London. The recorded nannofossil assemblages have been interpreted in terms of Martini's (1971) worldwide zonal scheme. In some instances the nominate taxa for the Martini zones have not been recognised during this study. However, the assemblages are diverse and contain many of the species that are characteristic of the particular Martini zone cited in this text. The stratigraphic range of some taxa have been derived from Perch-Nielsen (1985).

Biostratigraphical Calibration
Calibrating the biostratigraphical ranges of the benthic for a mini f e r a to the stand a r d p 1 a n k t o ni c framework necessitated examining characteristic planktonic foraminifera1 species in each sample and comparing the zonal assignments yielded by the planktonic specimens with the benthic foraminiferal data. Using the zonal scheme of Berggren & Miller (1988), planktonic zonal markers were identified in all of the samples and the results are listed in Table 1. The sample also contained Acariiiiira, Trrrncorotaloides, and Hairtkeniria riiexicaiia but did not contain Morozovdla aragoiiensis, Globigeriiiatheka iiiexicaiia, kiigleri or Harzlkeiiim alabameiisis, which would indicate a younger age. The nannofossil assemblage in this sample is poor. Specimens are poorly preserved and display much syntaxial overgrowth. Taxa recorded include Discoaster  388 Navet River marl This sample was originally thought to be from Zone P12 (Morozovella lehneri), but is here reassigned to the uppermost part of Zone P11 based upon the occurrence of

Specimens of Globigerinatheka mexicana mexicana, Morozovella lehneri, Subbotina f r o n t o s a and a transitional form between
Turborotalia cerroazulensis pomeroli and T . cerroazulensis posagnoensis were also present. This sample did not yield nannofossils.
m e x i ca n a bar r i .

Near Fitt Trace marl
This sample correlates with the middle part of Zone P12. Globigerinatheka subconglobata curryi was present, as well as two specimens that are transitional to Globigerinatheka subconglobata euganea. Turborotalia cerroazulensis pomeroli, Morozovella lehneri and H a n t k e n i n a dumblei were also present. However Morozovella aragonensis and Orbulinoides beckmanni were not found. The nannofossils in this sample were indeterminate because of their low abundance and poor preservation. 387 Dunmore Hill marl Bolli (1957b)

GI o biger i n a t he ka s u bco ngl o ba ta euganea.
Orbulinoides beckmanni was not found in this sample. A diverse, but only moderately abundant assemblage of nannofossils was recorded from this sample. Principal taxa are Sphenolithus furcatolithoides, S. predistentus, C h i a s m o l i t h u s g r a n d i s , Reticulofenestra umbilica, Discoaster barbadiensis, and D . saipanensis. This assemblage is characteristic of Zone NP16, which corresponds to Zones P12-Pl3. The presence of a few specimens of D. saipanensis indicates this sample is from the upper part of Zone NP16.
390 Penitence Hill marl The presence of Orbulinoides beckmanni indicates that this sample is from Zone P13.

Truncorotaloides rohri, Hantkenina alabamensis, Clavigerinella eocanica eocanica
and Morozovella spinulosa were also present. The nannofossils in this sample were indeterminate because of their low abundance and poor preservation. 391 & Stop 9: Hospital Hill marl Previous studies were unable to establish a precise age for this sample. However we determined that Sample 391 correlates with the lower part of Zone P15. This age is based upon the presence of G l o b i p r i n a t h e k a semiinvoluta, and-the co-occurrence of Clavigerinella eocaenica eocaenica and C. eocaenica jarvisi, which become extinct in the middle of Zone P15. Because muricate forms should have been extinct by the late Eocene, the discovery of a couple of specimens of Acarinina leads us to believe that this sample experienced slight reworking or contamination. However, the sample contains high relative abundances of Turborotalia cerroazulensis, T. cocoaensis, and Catapsydrax dissimilis, which are typical for the late Eocene in the Atlantic. This sample contains a poorly-preserved, lowabundance nannofossil assemblage consisting of late Eocene and reworked early to middle Eocene taxa. The in-situ late Eocene taxa include Cyclicargolithus floridanus, Ericsonia formosa, Discoaster barbadiensis, D.

saipanensis, Reticulofenestra u m b i l i c u s , S p h e n o l i t h u s p r e d i s t e n t u s , S . radians, and
Helicosphaera euphratis. This total assemblage is characteristic of Zone NP18. Early to middle Eocene taxa present in the assemblage include Discoaster lodoensis (common) and D. keupperi.
The "Stop 9 "sample was dated as Zone P15 because it also contains Globigerinatheka semiinvoluta in addition to the above species, but without Clavigerinella and the muricate forms. However, this sample yielded a mixed assemblage of early late Eocene and early to middle Eocene nannofossils. The in-situ late Eocene taxa include Discoaster saipanensis, D. barbadiensis, Dictyococcites bisecta, Sphenolithus predistentus, Cribrocentrum reticulatum, Reticulofenestra umbilica, Ericsonia formosa, and Helicosphaera euphratis. This assemblage is considered to be typical of Zone NP18, which corresponds to Zone P15. Early to middle Eocene taxa also present in this assemblage include Discoaster lodoensis, D. keupperi, and Sphenolithus spiniger.

Globigerina concinna and Stop 15: Globorotalia opima opima Zone
Sample 380 is known also as sample "Rz. 90', and in 1957 Bolli designated it as the type sample of the Globorotalia opima opima Zone (P20 / P21 late Oligocene). The presence of Globorotalia opima opima in both of these samples confirms this assignment. Sample 380 contains a reasonably diverse but sparse nannofossil assemblage with Sphenolithus d i s t e n t u s , Helicosphaera recta, Dictyococcites bisecta, Cyclicargolithus floridanus, H. euphratis, and S. moriformis. Although this is a poor assemblage in terms of numbers of individuals, the co-occurrence of S. distentus and H. recta is typical of Zone NP24, corresponding to foraminiferal Zone P21.

Stop 16: Globigerina ciperoensis ciperoensis Zone
Globigerina ciperoensis ciperoensis is observed in this sample, indicating Zone P22. It had not been found in the sample from the Globorotalia opima opima Zone. The sample yielded an assemblage of long-ranging nannofossil taxa, including Coccolithus pelagicus, Helicosphera euphratis, Cyclicargolithus jloridanus, and Discoaster deflaizdrei.

Stop 8: Globorotalia kugleri Zone
This sample was characterised by the presence of Globorotalia kugleri, thus giving it an age of latest Oligocene. However, the nannofossil assemblage from this sample was poor and lacked age-diagnostic taxa. The assemblage included Triquetrorhabdulus carinatus, Cyclicargolithus jloridanus, C. abisectus, Discoaster(?) druggii, D. deflandrei, and Helicosphaera euphratis. This poor assemblage is probably indicative of the early Miocene Zone NNl based upon the presence of a poorly preserved single specimen of D . druggii; otherwise the assemblage ranges from late Oligocene to early Miocene (NP25 to "1).

CONCLUSIONS
The results of our planktonic foraminiferal and nannofossil analyses provide a chronostratigraphic framework for revising the classic taxonomic studies of Cushman and Renz (1948) and Cushman and Stainforth (1945). Our age calibrations of Cushman's samples and those of previous studies although substantially similar, show clear points of differentiation. Sample 388 (Navet River marl), thought previously be Zone P12 in age, is now reassigned to the upper part of Zone P11. A distinct change in the stratigraphic sequence can be seen in Sample 387 (Dunmore Hill marl). The Dunmore Hill marl was previously thought to be in Zone P11, but Cushman's sample is here reassigned to the upper part of Zone P12. It is possible that this sample was not taken exactly from the type locality of the Dunmore Hill marl, or that more than one "P zone" was exposed at the Dunmore Hill locality. This is possible, because the Navet Formation's stratigraphic sequence is tentative, with samples taken from "isolated, small outcrops" (Bolli 1957