Lithofacies-influenced ostracod associations in the middle Ordovician Bromide Formation, Oklahoma, USA

The Bromide Formation of southern Oklahoma was deposited in a linear basin and on the adjoining platform during a marine transgressive-regressive event in the middle Ordovician. The formation displays wide lateral (platform-basin) and vertical (transgressive-regressive) sedimentary facies variation. From the prolific and diverse ostracod fauna present in the Bromide Formation two lithofaeies-related ostracod associations can be defined: a geographically and stratigraphically widespread Anisocyamus Association, occupying subtidally deposited marine sediments: and a Leperditella Association, which is restricted to marginal marine environments. The ostracods of the Bromide Formation demonstrate that the group can be utilized in the Ordovician as a tool to help establish palaeoenvironments and differentiate palaeoshoreline.


INTRODUCTION AND GEOLOGICAL SETTING
and forms part of the thick intracratonic succession of Ordovician rocks exposed in the Arbuckle Mountain:; and Criner Hills of southern Oklahoma. The formation is divided into a lowcr Mountain Lake Member and ;in upper F'oolevillc Member and, based on conodont stratigraphy (Sweet & Bcrgstriiim, 1976;Ross et ul., 1982: Sweet, 1?)84), is considered of late Whiterockian-early Mohawkian age. In terms of the British Ordovician 'Series' this equates witlh the late Llandeilo to early Caradoc.
The Bromide Formation was deposited within a major linear basin (the Southern Oklahoma Aulacogen: Shatski, 1946) and on the adjoining platform during a marine transgressive-regressive event (Fig. 2). The formation displays wide lateral (platf'orm-basin) and vertical (transgressive-regressive) facies variation, having marginal marine t o basinlal types of lithofacies, and contains one of the most prolific and diverse ostracod faunas recorded from the Ordovician of N. America. Some 48 gencra and 83 species of ostrircods are documented Crom the formation (Harris. 19.57: Williams. 1990). thus providing an appropriate sequcnce from which to investigate the patterns and possible controls of the distribution o f ostracods, particularly with respcct to prevailing lithofacies.
Detailed pa1aeoenvironment;iI studies of Ordovician ostracods are riirc. Several authors (especially Berdan, l96Y, 1976(especially Berdan, l96Y, , 1984 have noted the co-occurring relationship between IeperJitiocope-dominated ostracod faunas and marginal marine lithofacies in the Lower Palaeo/oic in general. Copelland (1982) documented bathymetrically controlled ostracod assemblages from the middle Ordovician L,ower Esbatao1.tine Formation of Canada and, on the basis of published work of other authors, attempted t o recognize similar assemblages in other N. American sequences, including the Bromide Formation of Oklahoma.
The ostracod faunas of the Bromide Formation were originally documented by Harris (1931Harris ( , 1957. Subsequent published research has addressed the taxonomy o f a few species (e.g. Martinson, 1960: Levinson, 1961 & Vannier,199.5). With the exception of the leperditiocopes the entire ostracod fauna of the Simpson (iroup has recently been studied (Williams. 1990). Pending full publication o f Williams' revised study. with some modifications we have used herein the binomen of previous authors. Where herein the generic name is placed in inverted commas (e.g. 'Ckwoholhinu' inflntir Harris, 1957: see PI. I , Fig. 3) we consider the species to belong to a new, as yet unpublished, genus. Where both the generic and specific names (as used by a previous author) are placed in inverted commas (e.g. 'Schniitltc~llu ufinix' Ulrich .sen.sii Harris. 1957: sce PI. 2, Fig. 5 ) we consider that both the genus and species are new (see Williams, 1990).

SECTIONS STUDIED AND METHODS
The ostracods detailed in the present study were recovered from six logged sections which, taken together, embrace the entire Bromide Formation (Fig. 3 Oklahoma Aulacogen. Section 1 is considered to have been situated on the palaeo-platform and section 6 is situated near the centre (ix. the presumed deepest marine parts) of the palaeo-basin. Sections 2-5 are thought to represent intervening and progressively deeper marine locations down a palaeoslope. In general, thc Mountain Lake Member of the Bromide Formation is well exposed in all 6 sections. The Pooleville Member of the Bromide Formation is well exposed in sections 1. 2 and 3; in section 4 only its top few metres is exposed; in section 5 its uppcr part is incompletely exposed and in section 6 the top of the member is periodically covered by river alluvium.
Thc ostracod distributions analyscd herein are based on faunas recovered from 117 samples collected in the field (Williams, 1990). These samples represent all of the clastic and carbonate lithologies present in the Bromide Formation and, where possible, embrace the whole of the stratigraphic range of the formation exposed in the six sections studied. Section 1 presents the richest and best preserved ostracod faunas so far encountered in the Bromide Formation.
The ostracods of sections 1 and 6 were originally studied by Harris (1957, charts 3 and 4). Harris also studied lateral equivalents of our sections 4 and 5 (see Harris, 1957, charts 2 and 1 respectively). Where appropriate, ostracod data from Harris' collections have been utilized in the present study. Account is also taken of the fact that the stratigraphy of the sections in question has been extensively revised (see Fay er al., 1982;Grahn & Miller, 1986;Williams, 1990).  In order to recover ostracods from the Bromide Formation, shale samples were broken down in baths of 10% hydrogen peroxide and then wet sieved, dried and picked for ostracods. In order to prevent possible bias due to 'facie:; samypling' the limestones were also sampled in bulk, subsequently broken into small pieces in a rock crusher and then examined for ostracod valves. Specimens recovered in this way were prepared using a vibrotool. Longmari (1976Longmari ( , 1981Longmari ( , 1982 clearly demonstrates a range of lithofacies in the Bromide Formation which, taken together, indicate ii marine transgressive-regressive event (Fig. 4) occurrence of a tidal flat environment. However, nearer to the geographic and bathymetric depo-centre of the Southern Oklahoma Aulacogen subtidal sedimentation persisted throughout the Pooleville Member. The sedimentological variation from shelf to basin is discussed below first with reference to sections 1 and 6, the end-members of the transect. Further data on the relevant lithofacies are given in Longman (1982).
Here. lithofacies of the Bromide Formation are dominantly of shallow marine aspect. Thc Mountain Lake Member begins with a thick sequence of shoreface sandstones which are succeeded by thin, interbedded, subtidally deposited shales and limestones. Longman ( 1982) estimated that water depths o f only 5 m prevailed during the acme of the marine transgression represented in this section. The succeeding Pooleville Member is dominated by shallow marine carbonate deposits: these culminate at the top of the sequence with thc occurrence of birdseye micrites which are interpreted as the product of deposition in a broad tidal flat environment (Longman. 1982).
Section 6 ( Fig. 6) Section 6 is situated near the geographic and bathymetric centre o f the Southern Oklahoma Aulacogen. Compared to the deposits of section I. the lithofacies of the Bromide Formation in section 6 have a more deep marine aspect. As in section 1 , the sequence begins with shoreface sandstones which arc succeeded by interbedded shales and limestones. Longman (1982) estimated water depths o f about 80m for the acme of the transgression for the strata represented in this section. The lithofacies of the overlying Pooleville Member. comprising richly fossiliferous limestones and shales, show no evidence of regression, and were probably deposited in water depths well below normal wave base.

Sections 2-5
Sections 2-5 ( Fig. 3) represent a gradual. intervening palaeoslope between the depositional settings of sections 1 and 6. Deposits of sections 2 and 3 were situated at the shelf-slope break (hinge line) o f the Aulacogen and show a similar development of lithofacies to those ol section 1, with regression evident in the Pooleville Member. Deposits of sections 4 and 5 were situated within the more basinal parts o f the aulacogen. though apparently in an overall shallower marine setting than those of section 6. In section 4 the strata of the uppermost Pooleville Member are characterized by birdseye micrites, thus indicating that regression also occurred here.  patterns may be expected in the Palaeozoic. For example in summarizing the ecological ranges of Silurian ostracods Siveter (1984) noted that highest diversity among palaeocope-dominated ostracod faunas occurred on the midshelf to shelf upper slope and that reduced diversity was characteristic of both deeper marine and more nearshore high energy environments. A few ostracod taxa (myodocopes) had probably adopted t o a pelagic life-style by the Silurian but most Silurian ostracods were apparently benthic (see Sivcter, 1984: Siveter el al., 1991  Ihc Bromide Formation present at the Highway 77 locality (see Harris,195'7). Copeland concluded that the ostracod t'aunas of rhe Bromide Formation (Decker's beds 8-15) of the Highway 77 locality (see Harris, 1957) correspond quite closely t o the deeper shelf marine ostracod assemblage o f the Lower Es ba t ao t t i ne Forma ti on.

OSTRACOD DISTRIBUTIONAL PATTERNS
Of the 83 ostracod species currently recorded from the Bromide Formation (Williams, 1990), 65 are present in the sections 1-6 discussed herein. The distributional patterns of ostracods in the Bromide Formation is herein presented based on presence/absencc data. N o statistical analysis has been made. partly in the knowledge that field sample size and sampling interval varied in order to maximize the number of ostracods recovered. Two lithofacies-rclated associations. of what are prcsumed to be benthic ostracods, c;in he clearly recognized within our six sampled sections: the Ani,soc.yunziis Association and the I,epc~rtlitc4ln Association. The term 'association' is here used with respect to the Bromide Formation for a recurring, discrete group of ostracods which occupy a defined range of litbofacies (palaeoenvironments) and which are taxonomically distinct from ostracod faunas (associations) which occur in other, different lithofacies. (Pls 1-3: Fig. 7) This association, which derives its name from two characteristic primitiopsacean species (Anisocyamus elegans and Ani.rocyuniic,s husslrri: see Siveter & Williams, 1988~7, h ) is present in all six sections (' Table 1). The Anisocyamus Association includes the majority of ostracod species (56) documented in the six sections studied and is dominated by palaeocope (see PIS 1-3: Fig. 7) and leiocope ostracods (see PI. 2). However, overall diversity is high, with eridostracan, binodicope and podocopid ostracod taxa also being present (see PIS 2, 3: Fig. 7). Many of the species belonging to the Anisocyamus Association (e.g. Bromidella reticulata, Cryptophyllus gibbosum and Eridoconcha simpsoni) have long stratigraphic ranges, occurring in the Mountain Lake and Pooleville members of the Bromide Formation and also in the underlying Tulip Creek Formation. Within the Anisocyczmus Association there also appears to be some cross-basinal variation in the geographical ranges of individual species, although some of this variation may be related to sampling bias in individual sections. A total of seven species (e.g. 'Schmidtella' minuia, Kuyina hybosa) appear to be present only in the shallow marine subtidal deposits of :section 1. Six species (e.g. Baltonotella parsispinosa, ',Primiiiopsis' minutu) appear to be restricted to the deeper rnarine sediments of section 6 ( Table 1). More significantly, however, most species of the Anisocyamus Association have wide geographical distribution; for example, 20 species of this association are common to sections 1 and 6 and a further seven species are common to sections 2 and 6 ( Table I).

Anisocyarnus Association
Leperditella Association (Fig. 7) This association, which derives its name from two characteristic leperditellacean species (Leperditella rex and 1,eperdiiella tumida), is present only in the upper part of the Pooleville Member and does not occur in all the sections studied. Significantly, the Lepertiitella Association is absent from section 6. By contrast, the Leperditella Association is best developed, both in terms of species diversity and number of individuals, in section I . The Leperditella Association is poorly developed in the uppermost part of the Pooleville Mernber at both sections 3 and 4. The Leperditella Association is characterized by leperditiocope ostracods which occur in limestone beds generally 10-50 cm in thickness; however, in the intervening shales (generally 2-10 cm thick), palaeocope ostracods, particularly leperditellaceans, predominate. Significantly, none of the (nine) ostracod species which comprise the Leperditella Associatioin are present in the AnisocyamuJ Association.

FACTORS INFLUENCING OSTRACOD DISTRIBUTION
The distribution of the two named ostracod associations is clearly closely related to the distribution of the lithofacies of the Bromide Formation. Deposits of sections 1 and 6, which record sedimentation on the palaeo-platform and in the palaeo-basin respectively, clearly demonstrate this pattern of occurrence (Figs 5 , 6). The most characteristic ostracods recorded from these two sections and belonging to the Anisocyamus and Leperditella associations are illustrated in Plates 1-3 and Fig. 7. In section 6 where, after the deposition of the basal shoreface sandstone sequence, sedimentation occurred in a subtidal marine setting throughout the Mountain Lake and Pooleville members, only the Anisocyamus Association is present ( Fig. 6; see also Harris, 1957, chart 4). The Anisocyumus Association is absent from lithofacies characteristic of very shallow or marginal marine deposition such as are developed in the upper part of the Pooleville Member in sections 1-4. Ostracods of the Anisocyumus Association appear to have tolerated a wide range of water depths; it is estimated that relevant marine subtidally deposited sediments of sections 1 and 6 were deposited in water depths of 5 and 8 0 m respectively (Longman, 1982).
By contrast, the Leperditella Association is restricted to lithofacies characteristic of very shallow or marginal marine environments, in that it occurs only in the upper part of the Pooleville Member in the platform sequences. Where deeper water sedimentation persisted throughout the Pooleville Member (i.e. section 6, Fig. 6) the Leperditellu Association is absent. The Leperditella Association seems to be only completely developed where both limestones and shales were deposited (i.e. section 1). Where only limestones are present (i.e. section 4) only leperditiocope ostracods are present (data of Harris, 1957, chart 1). This suggests that in addition to water depth, substrate was also a factor influencing the distribution of the Bromide Formation ostracods.
The Leperditellu Association (comprising nine species)
The ostracod associations of the Bromide Formation can hc compared with Copeland's (1982) middle Ordovician ostracod assemblages from the Lower Esbataottine Formation of Canada. The Lower Esbataottine Formation appears to have been deposited entirely within a marine subtidal setting (Copeland, 1082, p. '4). There are no marginal marine, peritidal or tidal flat depositional environments in the Lower Esbataottine Formation comparable to those of the Bromide Formation in Oklahoma. Correspondingly the Leperditalla Association, which in the Bromide Formation is characteristic of lithofacies indicative of marginal marine environments, shows no relationships to the ostracod assemblages of the Lower Esbataottine Formation. By contrast, the .Anisocyamus Association of the Bromide Formation has species in common with both the two ostracod assemblages of the Lower Esbataottine Formation. For example, Euryholhinu hispinufa (PI. 1, Fig. 12) is common to the Anisoc~yumt~s Association of the Bromide Formation and Copeland's (widespread shallow to deeper shclf marine) ostracod assemblage 1. Eohollina depressa (PI. I , Fig. 13) and Plafyrhorriboides quadratus (Fig. 7.1) are common to the Anisocyamus Association and Copeland's (deeper shelf marine) ostracod assemblage 2. Copeland