sign in sign up
<<
Home , Bone Valley Formation, Citronelle Formation, Cormohipparion, Equinae, Goliad Formation, Gomphothere

TULANE STUDIES TN GEOLOGY AND

Volu me 22, Number 2 Sepl<'mber 20. l!J8~)

PLIOCENE THREE-TOED FROM . WITH COMMENTS ON THE

EAHL M. MANNING MUSP.UM OF'GEOSCIF:NCE. LOUISJJ\NA STATE UNIVF:RSlTY. JJATO.\I ROI.JG/<. LOL'/S//\;\':1 and llRUCE J. MACFADDlrn DEJ>ARTM/<:NTOF NATUH/\LSCIENCES. F'LORJD/\ MUSf:UM Of<'NJ\TUIV\/, lllSTOUY UNIVERSITY OF FLOH!IJJ\. GJ\/NESVlU.E. Fl.OH/DA

CONTENTS Page T. ABSTRACT 3.5 II INTRODUCTION :l5 Ill. ACKNOWLEDGMENTS :rn TV . ABBREVIATIONS :l7 V. SYSTEMATIC PALEONTOLOGY ;37 VI. AGE OF THE TUNICA HILLS HIPPARIONINES 38 VIL STRATIGRAPHIC PROVENIENCE 38 Vlll. TERRESTRIAL OF THE GULF AND ATLANTIC COASTAL PLAIN .JO IX. COMMENTS ON THE CITRONELLE FORMATION .JI X. AGE OF THE CITRONELLE 42 XL TH E CITRONELLE FORMATION IN nm TUNICA HILLS .t:1 XII. LITERATURE CITED l.J

January of 1985, the senior author was L ABSTRACT shown a large collection of late Teeth and metacarpals of early Pliocene (Rancholabrean land- agel ver­ (latest land-mammal age) tebrate from the Tunica Hills of three-toed (hipparionine) horses are de­ Louisiana (Fig. I) by Dr. A. Bradley scribed from the Tunica Hills of West McPherson of Centenary College, Feliciana Parish in east-central Louisiana. Shreveport. McPherson and Mr. Bill Lee An upper perta ins to of Balon Rouge had collected fossils from minor, known from the Hcmphillian of that area since about 1981. Among the Central and , and two teeth standard assemblage of Rancholabrean and two distal metacarpals pertain to a re­ taxa (e.g. , , ground sloth, cenlly described species of Cormohippar­ tapir and ) were an upper molar and ion, C. emsliei, known only from Pliocene distal metacarpal of very small horses col (latest Hemphil\ian and land \ected by them in 1983, which were quite mammal ages) sites in peninsular . different from the common large of T he material represents the first record of the fauna. In 1986 Mc Pherson and Lel:- col Pliocen e non-marine fossils from Louisi­ lectcd two more teeth and anolhPr distal ana as well as the first record of' a metacarpal of a small. primitive' horse. hor~e from the sta te oth er than Pleistocene These fi\'l' spt•cimens arL', thus for th<.• Equus. Found in sur face stream gravels, only record of a pre-lat<.• Pk•istoeenc fauna the horses proba bly were derived from the among the hundreds of verU_•bratc fossils otherwise a lmost nonvertcbrate-bcaring that havt• been found m tfw Tunica Hills Citronelle Forma tion. area. Thev indicate the presence of ter restnal or. near-shore early Pliocene de posits in Louisiana. The purpose of' the II. INTRODUCTION present study is to describ<.• this matt.•nal Fossil land mammals of pre-Pleistocene and to discuss its irnplications for Louisiana age a rc exceedingly rare in Louisiana. In biochron(Jlogy and stratigrapl1y 35 36 Tulane Studies in Geology cmd Paleontology Vol. 22

e1°3o'w 90°W {( • P1nckneyv1lle MISS.

LA.

Bluff LOCALITIES

30°45'N

0-- - 6 KM e1 °3o'w

Figure 1. Location of the Pliocene horse sites a nd the Tunica Hills, West F eliciana Parish, Louisiana and Wilkinson County, . Modified from Givens and Givens (1987, fig. ! ) and Alford et al. (1983, fig. l ).

III. ACKNOWLEDGMENTS Tunica Hills, the mate ria l would never have come to light. Dr. R ichard C. Hul­ The authors a re e specially grateful to bert, Jr. (U niversity of Florida , Gaines­ Dr. A. Bradley McPherson of Shreveport ville) k indly allowed us the use of a the n and Mr. Bill L ee of Baton Rouge for per· unpublished manuscript on Coastal P lain mission to describe these specimens. With­ horses that was critical to the present study out their persistanl collecting efforts in the a nd also provided useful discussion of o ur No. 2 Louisiana Pliocene Horses 37 identifications. Dr. Frank Whitm ore (U.S. Feliciana Parish. east-central Louisiana (Fig. lJ. Geological Survey, Washington, D. C.) gra­ From surface stream gravel, probably derived ciously permitted one of u s (EMM) to from the Citronelle Formation (see discussion examine the fascinating Mauvilla fauna of below). . Dr. Richard Kese l (L SU Depart­ Age. Early Pliocene (latest Hemphillianl. ment of Geography and Anthropology, Discussion: The dental pattern in the Baton Rouge) examined the sites where Louisiana specimens exhibits the diagnos­ the specimens were found, and provided tic characters listed in Hulbert (19871. The useful information on the stratigraphy of parastyle and mesostyles are grooved on the area. Dr. Whitney J. Autin (Louisiana the anterior side. The fossette borders are Geological Survey, Baton Rouge) re­ richly plicated with a count of 4-5 + 4-1. viewed the manuscript, and provided The pli caballin consists of a large fold with many helpful comments . T he manuscript five secondary folds. The hypoconal was also reviewed by Dr . Judith A. groove is moderately deep. The protocone Schiebout (LSU Museum of Geoscience, is elongate with a concave lingual side. Baton Rouge). This is University of Florida The is in middle wear and seems Contribution to Paleobiology number 321. moderately hypsodont (Table 1 ). The lower tooth, a right P 2 , exhibits a IV. ABBREVIATIONS diagnostic anterior extension of the Fossil collections refe rred to in the text metaconid that connects to the posterior are: CCVC - Centenary College portion of the paraconid. The protoconid Collection, Biology De partment, Cente­ and hypoconid are flattened and there is a nary College, Shreve port LA; UF - Ver­ prominent pli caballinid. Although the size tebrate Paleontology Collection, Florida of the Louisiana sample is smaller than the State Museum, University of Florida, type from the Blancan, it falls within the Gainesville, Florida. size range published by Hulbert (19871 for the large sample from the late Hemphillian upper Formation. V. SYSTEMATIC P ALEONTOLOGY There are two distal metacarpals, CCVP 86 (Fig. 3) and 535, from the Tunica Hills Class MAMMALIA L innaeus, 1758 that seem to be referable to Cormohippar­ Order PERISSODACTYLA Owen, 1848 ion emsliei based on the primitive (hip­ Family Gray, 1821 parionine) aspect of the distal tuberosities Subfamily Gray, 1821 (26.9 and 28.9 mm) being greater than the Genus width of the distal articular keel (26.4 and Skinner & Ma c F adden, 1977 27.3 mm; see Hussain, 1975) and similar CoRMOIIIPPARION F:MSLH:I Hulbert, 1987 proportions to that of probable C. emsliei Figures 2 c-f, 3, Table 1 metapodials from the Bone Valley. Distribution: Early P li ocene {latest Hemphil­ Because North American horse species lian land-mammal age) of the upper Bone Valley are generally widespread in geographic Formation, Polk Co. , central peninsular distribution, the occurrence of C. emsliei Florida; late Pliocene (late Blancan land mam­ in only Florida and Louisiana suggests that mal age), of the Pine crest beds, Macasphalt it is an unusual, regional endemic species Shel! Pit, 3 km N of F ruitvill e (type localiiy), of the late Cenozoic Gulf Coastal savanna. Sarasota Co., west-centra l peninsular Florida; and the Tunica Hills area, West Feliciana Genus NANNIPPUS Matthew, 1926 Parish, east-central Louisiana. NANNIPPUS >!!NOH (Sellards, 1916) Referred Louisiana specimens: CCVC 1628, Figures 2 a-b, Table 1 right P 3 or P 4 ; CCVC 1627 , right P:,:; CCVC 86 and CCVC 535, distal ends of MC III.

TABLE 1 Measurements (in mm) of the teeth of Pliocene three-toed horses from the Tunica Hills. Greatest antero- Greatest transverse Height posterior length width (excluding (upper-mesostyle; Specimen No. Element (excluding cement) cement) lower-paraconid)

Cormohipparion emsliei ccvc 1628 RP' or P 4 18.8 17.0 37 .5 ccvc 1627 RP2 20.6 9.0 30.4 Nannippus minor ccvc 585 LM 2 16.5 15.0 35.7 emsliei. Although broken on the occlusal reviews in Domning, 1969 and Lowery, surface, the parastyle and metastyle seem 1974). Recent work has shown that nearly unconstricted. The fossettes are moder­ all of the vertebrates have derived from ately plicated (3-3-4-2). There is a well de­ low level (Tl) terrace deposits of late Pleis­ veloped pli caballin. The protoconc is elon­ tocene age (Givens and Givens, 1987). No gate with a concave lingual border. This older vertebrates have previously been re­ tooth, which is in middle wear stage, ported from the area. seems too low-crowned to be referable to The fossil horse material described here Nannippus peninsulatus (senior synonym was found loose in the surface stream of N. phlegon). It falls within the large end gravel along Tunica Bayou (Fig. 1), as are of the observed range and has a dental pat­ most vertebrate fossils from the Tunica tern consistent with that of Nannippus Hills. In situ recovery of vertebrates is un­ minor (MacFadden, 1984; Hulbert, 1987). common and the discovery of even par­ tially articulated material (such as the par­ VI. AGE OF THE tial mastodon skeleton figured in Lowery TUNICA HILLS HIPPARIONINES 1974, p. 519) is extremely rare. Even when found in place, most vertebrate material The presence of the two hipparionines consists of isolated, often badly stream Cormohipparion emsliei and Nannippus worn, elements in channel gravel or sand. minor suggests a latest Hemphillian, or The outcrops exposed in the area of early Pliocene, age for the sub-Pleistocene Tunica Bayou where the horse material stratum of the Tunica Hills. This is based was found consist of reworked loess, sand on: 1) the age of the i.;pper Bone Valley and thin gravel stringers typical of much of Formation of Florida, the only other place the Tunica Hills Pleistocene. One of the where these two taxa are known to co­ Pliocene horses, CCVC 1628, was even occur; and 2) the primitive stage of evolu­ found in the same area as Rancholabrean tion of C. emsliei from Louisiana, which is (late Pleistocene) reported by of similar size to the Bone Valley occur­ Damning (Arata and Domning MS, in ences of this species (and smaller than the Domning, 1969, p. 393). The Ranchola­ Blancan specimens). The biochronology of brean fauna found in the stream beds of the upper Bone Valley fauna is well estab­ the Tunica Hills probably derives from the lished (e.g .. MacFadden, 1986) and proba­ slumping of this material into the bayous, bly represents a time interval between 5.4- with the gravel-sized fraction concentrated 4.5 myr ago, or early Pliocene, using the later at the stream surface. The composi­ time-scale of Berggren et al. (1985) . tion of the loess in the area, containing stringers of sand and fine gravel, suggests VII. STRATIGRAPHIC PROVENIENCE that it is reworked colluvium, and is not in The Tunica Hills have long been known situ (Fisk, 1938; Delcourt and Delcourt, for their rich fossil flora (see review in Gi­ 1977; Alford, Kolb and Holmes, 1983). If vens and Givens, 1987), freshwater mol­ this is the case, then the Pleistocene ver­ lucks (Richards, 1938) and vertebrates (see tebrate fauna may have been reworked No. 2 Louisiana Pliocene Horses

E

1 ms

2 3 4 5cm

Figure 2. Cormohipparion emsliei (C-F) and No.nnippus minor (A-B) from the early Pliocene of the T unica Hills, West Feliciana Par., La. A-B: CCVC 585, left M'. Iatl'ral and occlusal views. C-D: CCVC 1628, right P" or F', lateral and occlusal views. E-F· CCVC 1627, right P 2, in occlusal and lateral views. twice. Specime n wear (including nearly Formation is known to be of late rounded cobbles of mastodon bone) tends age (see sect. X) and, therefore, it is most to corroborate this hypothesis. likely that the Pliocene horses were de­ rived from the Citronelle Formation. Because the hipparionine horses de­ which is the bed usually underlying the scribed here are of Pliocene age (based on Pleistocene beds in the Tunica Hills. distribution of the species elsewhere, par­ It is worth noting that while Rancholab­ ticularly Florida), they must have origi­ rean fossils have been found in nearly nally derived from a bed below that in every major drainage of West Feliciana which the much larger Rancholabrean Parish (Tumca Bayou, Kimball Creek, Lit~ fauna was buried. Because of the relative­ tie Bayou Sara, Bayou Sara, Thompson's ly unworn state of the fossils described Creek, etc.), Pliocene vertebrates are only here (despite evident reworking), it is un­ known from a small area on Tunica Bayou likely that they were derived from a very This possibly suggests a limited, local distant source. The underlying Pascagoula source bed. 40 Tulcme Studies in Geology and Paleontology Vol. 22

ta! Pliocene faunas are both earliest Pliocene (latest Hemphillian land mammal age), whereas later Pliocene (Blancan land mammal age) Coastal Plain faunas are un­ known outside Florida. The following brief review of Pliocene terrestrial vertebrates from the Coastal Plain discusses the princi­ pal localities of this age. - Although late Hemphillian and Blancan faunas are well known from the High Plains Ogallala Group of the north Texas panhandle (see review in Schultz, 1977), and Blancan fossils are recorded from the Basin and Range bolsons of west­ ern Texas (Akersten, 1972\ Pliocene ter­ restrial vertebrates have not been re­ corded from the Texas coastal plain. This is surprising in light of the extensive record of Miocene terrestrial vertebrates of the Texas coastal plain (Quinn, 1955; Wilson, 1956; Patton, 1969; Patton and Taylor, 1971; Forsten, 1975; Prothero and Man­ ning, 1987, etc.). The has, at times, been identified as Pliocene, but vertebrate faunas of the Goliad are of middle to late Miocene (late Barstovian/ early ) age (Prothero and Manning, 1987). Louisiana and Mississippi - Until the I ffi B present report, no Pliocene terrestrial fos­ si ls had been reported from either state (see review in Domning, 1969). The Miocene terrestrial vertebrate record from 3 4 Scm these states is little better, limited as yet to a single pair of tusks from the part of the Fleming Figure 3. Cormohipparion emsliei from Formation of western Louisiana (Arata, the early Pliocene of the Tunica Hills. 1966). CCVC 86, dorsal view of distal metacarpal Alabama. - "Middle Pliocene" terrestrial III. vertebrates have been reported from southwest Alabama (lsphording and Lamb, 1971; Thurmond and Jones, 1981); VIII. PLIOCENE TERRESTRIAL but this age assignment was based on an VERTEBRATES OF THE GULF outdated correlation of the Hemphillian. AND ATLANTIC COASTAL PLAIN Except for the latest faunas, the Hemphil­ Unlike Miocene terrestrial vertebrates, lian is now generally considered to be of which have an extensive, if' spotty 1 record late Miocene age (Tedford, et al., 1987). from the southeastern North American This Mau villa, or Mobile, Local Fauna (see Coastal Plain (see review in Tedford and Whitmore in Isphording and Lamb, 1971 , Hunter, 1984), non-marine Pliocene ver­ p. 776 and lsphording and Flowers, 1983 tebrates are well represented only in pp. 71-72; Tedford and Hunter, 1984, p. Florida and North Carolina. Considering 144) can be dated as early (or middle) the extensive surface exposures of Hemphillian, rather than late, on the basis Pliocene sediments in the area, it is sur­ of the presence of cf. prising that Pliocene terrestrial faunas arc tricoronatus in assocation with advanced not better known. The better known coas- rhinocerotid taxa. A middle Hemphillian No. 2 Louisiana Pliocene Horse.'> 41 age has been indicated by Tedford and die. Matson named the beds for outcrops Hunter (1984 p. 144). With the Mauvilla near the city of Citronelle, in southwest Local Fauna reassigned to the late Alabama. The outcrop area has now ex­ Miocene, Alabama also has no Pliocene panded across the Gulf and Atlantic Coas­ terrestrial vertebrates. tal Plain. Doehring (1958, p 773) included Florida - Unlike Gulf Coastal slates to "a belt of up lo 100 miles wide which ex­ the west, Florida has an extensive Plio~ tends for 1,500 miles from the Nueces cene terrestrial vertebrate record. An ex­ River in south Texas to the vicinity of cellent late Hemphillian fauna has long Washington, D.C." been known from the fluvio-deltaic upper Sediment - Citronelle elastics vary, but part of the , exten­ they are dominated in outcrop by quartz sively exposed in the phosphate mines of sand and gravel (Matson, 1916, p. central peninsular Florida (see review in 174). The sand and gravel is gray-white MacFadden and Webb, 1982). Several when fresh (Rosen, 1968), but weathers to Blancan faunas, e.g., the Haile XV A site pink, red, or orange as its iron cement (Robertson, EJ76) and Santa Fe River oxidizes to ferric hydroxide {Parsons, localities have been reported from penin­ 1967). Generally, these elastics are only sular Florida (see MacFaddcn and Wal­ slightly consolidated. Sedimentary struc drop, 1980). Unlike the Gulf Coastal states lures such as channels, cut and fill, scour to the west, Florida also has an extensive surfaces, and cross-bedding are locally early Pliocene marine fossil record, in the common in the Citronelle (Autin, et al , upper Choctawhatchee Group. The differ­ 1986, p. 420). Interbedded with the coarse ence relative to Louisiana, for example, is elastics are minor beds of silts and clays due to the distance of Florida away from (Parsons, 1967, pp. 41 and 44). the elastic-dominated delta facies of the Mississippi River, Paleoecology - Early work on these hPcl: Despite extensive surface exposures of suggested that they might be glacial "drift' Pliocene sediments on the Atlantic Coastal (Hopkins, 1872: Lerch, 1892l or marinl' (McGee, 1891: Harris and Veatch, 18991. Plain (Cronin, et aL, 1984), Pliocene ter­ restrial vertebrates in this area are also but most workers in this century havP generally poorly represented, agreed that the Citronelle is predominai<'ly Georgia - Voorhies (1974 ) reported the of fluviatile origin. It has been ck•scribvd as " ... an alluvial plain built up by braiding, occurrence of a possible early Pliocene coalescing streams crossing a gently slop­ (latest Hemphillian) Nannippus minor from ing coastal plain." (Parsons. 1967, p 49) upland gravel beds in central . This interpretation has bcPn supported by North Carolina - A relatively large early more recent work {Smith and Mey ar Pliocene (latest H emphillian) terrestrial 1983, p. 430l. The presence of two Pliocene vertebrate fauna is known from the horses with very high-crowned (grazm; Yorktown Formation at Lee Creek Mine, type) teeth in the Tunica Hills suggt";ts th lt near Aurora. The fauna has been listed (R. at least the upland part of the Citront•lk· Eschelman in Tedford and Hunter, 1984, must have had nearby grasslands dunr, pp. 138-139) , but has not yet been fully de­ deposition. There an• considL'rable chan· scribed. ges in the Citronelle from north to south In summary,. the horses described here Clastics are generally coarser to the north are the only Pliocene terrestrial verte­ and finer to the south, and the southern brates presently known from the Gulf deposits art• party estuarine (Matson, 191(), Coastal Plain west of peninsular Florida. p. 187). Matson notPd that the varying sedi ment types suggested that the sourc( IX. COMMENTS ON THE streams varied considerably in both din.'c CITRONELLE FORMATION tion and velocity {IHW. p. 1871, Some bt.'ds Extent - As originally studied by Matson suggest flood deposits while others ..;uggest (1916), the Citronelle included reddish and quiel pond deposition. orange sands and gravels in a belt along Source Derivation from northern glacial the northcentral Gulf Coast from western out wash (e.g., Fisk, 19:~9), or an area west Louisiana to the western Florida panhan- of the Mississippi iClendcnin. 1896, p.1961, 42 Tulane Studies in Geology and Paleontology Vol. 22 or an ancestral Mississippi drainage (Rus­ Citronelle was not contemporaneous with sell, 1987), have all been suggested but the Williana has been indicated by Russ mineralogical analysis points strongly to a (1976, p. 1081), who noted that the Williana southern Appalachian source area, at least surfaces cut into the Citronelle in the Red for Citronelle deposits in Mississippi and River area of Louisiana. Alabama (Rosen, 1968 and 1969; Isphord­ Fossils - Despite its enormous surface ing, 1976: lsphording, 1983: Isphording and exposure, the Citronelle has produced Flowers, 1983; Smith and Meylan, 1983). very few fossils. There are probably sev­ The later work tends to support an earlier eral reasons for the scarcity of fossils. High hypothesis of a southern Appalachian energy gravels often destroy any con­ source area (Brown, 1967). The Citronelle tained fossils. Shallow braided streams heavy mineral suite differs from both that generally preserve fewer vertebrates than of late Pleistocene glacial terrace sediment do deeper channels. Leaching during the and that of modern Mississippi River sedi­ weathering process may have dissolved ment (Rosen, 1968 and 1969). Middle some fossils, though Matson noted that Paleozoic invertebrates found in chert even some unweathered beds he pebbles from central Gulf Coast Citronelle examined were non-fossiliferous (1916, p. gravels (and reworked into Pleistocene 186). The major exception is the plant ma­ gravels) may also derive from the southern terial described by Berry (1916) from un­ Appalachians (Rosen, 1968, p. 1O; Smith weathered clays in southwest Alabama. and Meylan, 1983, p. 424). He described well-preserved leaves, Relationship to the Coastal Plain Se­ seeds, and nuts of at least 18 species, reve­ quence - As emphasized by Doehring aling a coastal flora similar to that of coas­ (1958), unlike the underlying Coastal Plain tal Alabama today, with abundant bald beds, the basal Citronelle is often strikingly cypress and live oak. He suggested that unconformable. While most Coastal Plain the plants lived near a quiet, coastal la­ sediments rest only on the next older bed, goon, which was protected by a barrier the Citronelle often overlies much older beach. Although it has been argued that units. On the eastern flank of the Sabine the plants are from a bed below the Cit­ Uplift in northwest Louisiana, the Cit­ ronelle (Roy, 1939; Carlston, 1950), more ronelle overlies Eocene beds. This distinc­ recent work (Stringfield and LaMoreaux, tive unconformity, in addition to it non­ 1957, p. 746; Isphording and Flowers, 1983, marine origin, suggests that the Citronelle p. 68) has corroborated the view that the is not part of the typical Coastal Plain se­ plants occur in the Citronelle. No in situ quence. Contrary to Doehring (1958), this vertebrate material is known from the Cit­ does not mean that the Citronelle is part of ronelle Formation. a Pleistocene sequence. Relationship to Pleistocene Beds - There X. AGE OF THE CITRONELLE has been tremendous confusion about the relationship of the Citronelle to overlying Flora - Matson (1916) based his Pliocene Pleistocene beds. This is mostly due to the age assignment on the floral determina­ fact that many of the Pleistocene beds in­ tions made by Berry (1916, published at corporated large amounts of Citronelle the same time as Matson's paper) as well sand and gravel during deposition (Is­ as the fact that the Citronelle is overlain by phording and Lamb, 1971, p. 775). As a re­ Pleistocene sediment. Berry suggested a sult, they often look very similar and con­ late Pliocene age (1916, p. 195) because tain similar lithologies. This makes con­ some Citronelle plants are not present in tacts difficult to discern and resulting maps the Pleistocene. It must be added that, at and thickness estimates of the Citronelle the time, no Pliocene floras from North suspect. Doehring (1958, p. 778) states that America were available for comparison. some Citronelle outcrops have been map­ Doehring (1958) suggested that the ped as Williana Terrace, and Rosen (1968) plants might be of early Pleistocene, rather noted that parts of Matson's mapped Cit­ than Pliocene, age. This argument ap­ ronelle outcrop was included by Fisk pears to have been part of an effort to (1939) as Williana, Bentley, and even place the Citronelle in the same deposi­ Montgomery terrace deposits. That the tional system as the overlying Pleistocene No. 2 Louisiana Pliocene Horses 43

deposits. Doehring considered the Cit­ Valley Formation of the central peninsula, ronelle a pre-glacial Pleistocene deposit. and the Tamiami Formation of the south. Doehring's age assignment of Berry's flora The similarity of horse taxa described here has been contested by Stringfield and with those of the Bone Valley corroborate LaMoreaux, who included a confirmation these correlations. of Berry's Pliocene date by R. W. Brown of the USGS (1957 , p. 746). On the other XI. THE CITRONELLE FORMATION hand, a study of pollen from the middle IN THE TUNICA HILLS and upper Citronelle Formation from As elsewhere in the Gulf Coast, the Cit­ western-most Florida by Estella Leopold ronelle Formation has often been confused indicated a Quaternary age (Marsh, 1964, p. 83). with both older and younger beds in Louisiana. As a result, much basic geologic Underlying Beds - A maximum age for data on the unit (outcrop distribution, the Citronelle can be obtained from the thickness, and even stratigraphic terminal age of the youngest underlying beds. As ogy) in the state is equivocal. noted by Matson (19 16 , p. 172), the Cit­ Hilgard ( 1860) included Citronelle beds ronelle lies unconformably upon the Pas­ in the Tunica Hills of Louisiana and Missis cagoula Formation in Mississippi. A small sippi as part of a Quaternary unit that he invertebrate fauna from the Pascagoula in identified as "Orange Sand." In a later southeastern Mi ssissippi was reported by work (1869), he grouped Citronelle mate· Mincher (1941 ), who compared its age to rial with Eocene sediment under the same that of the Arca zone in Florida, now name. Hopkins (1872) included Citronelle thought to be of late Miocene (planktonic material as part of a "Northern Drift " foraminiferal zone NI 7) age (Akers, 1972, McGee (1891) included Citronelle beds in p. 14). In southwest Alabama, the Citonelle his Lafayette Formation, a unit he consid· overlies a sand and clay bed (possibly the ered to be a pre-Pleistocene marine bed. Ecor Rouge Member of the Catahoula For­ Lerch ( 1892) included Citronelle sediment mation), which contains the previously in beds he identified as "Drift." Harris and mentioned Mau villa L ocal Fauna (Isphord­ Veatch (1899) included Citronelle beds ing and Flowers. 1983, pp. 69-71) of late with Eocene sediment in their usage of Miocene age. Thus, in at least two places, McGee's "Lafayette." They considered the the Citronelle overlies late Miocene (late Lafayette to be a late Pliocene coastal de­ H emphillian and probably Messinian posit. Matson (1916), in the original de­ European stage) sediments. scription of the Citronelle, described the Overlying Beds - A minimum age for the bed in the Tunica Hills area as "Yellow Citronelle can be obtained from the age of and red sands and clays, locally gray the overlying beds . Several authors have where unweathered. Much gravel near noted Pleistocene beds overlying the Cit­ the landward margin and in valleys of prin ronelle (Matson, 1916; Mayer, 1932, p. 34; ciple streams." Parsons, 1967; Rosen, 1968; Isphording Parsons (1967) noted that the Citronelle and Lamb, 1971 , p. 775; Russ, 1976; Is­ overlies silts and clays in Louisiana, but phording and Flowers, 1983; etc.). did not identify these beds. Brown and Stringfield and LaMoreaux (1957, p. 746) Guyton (1943) identified these underlymg note that the Citronelle is overlain by the beds as Pascagoula Formation. Parsons oldest Pleistocene marine terrace in noted (1967, pp. 37-38) that the basal Cit­ Florida. ronelle contact in the Tunica Hills area Correlation ~ The above data argue could be either a sharp unconformity or a strongly for a Pliocene age for at least part transitional zone. of the Citronelle, as originally proposed by Rosen (1968, 1969) studied the mineral· Matson (1916). It has been suggested (Ted­ ogy of the Tunica Hills Citronelle in detail. ford and Hunter, 1984, Fig. 4.) that the Cit­ and noted that it is the principal aquifer of ronelle may be a late ral equivalent of sev­ the Baton Rouge area. eral early Pliocene transgressive beds in Because many Pleistocene beds (par­ Florida, i.e., the Jackson Bluff Formation ticularly the gravels) in the Tunica Hills are of the panhandle, the uppe rmost Bone derived from reworked Citronelle, it is dif 44 Tulane Swdies in Geology and Paleontology Vol. 22 ficull lo draw a boundary between the two. 98-L: 193-208, pis. 45-47. Many people have included nearly all of BROWN, B. W., 1967, A Pliocene Tennessee West Feliciana Parish in the Citronelle out­ River hypothesis for Mississippi. Southeast­ crop area (Malson, 1916; Parsons, 1967; ern Geo!. 8:81-84. Rosen, 1968). The presence of la le P leis­ BROWN, G. F., and W. F. GUYTON, 1943, Geology und ground water supply al Camp tocene (Rancholabrean) vertebrates in Van Dorn. Bull. Mississippi Geo!. Surv. 56: bayous across the parish (as well as the 68 p. presence of extensive loess outcrop) shows CARLSTON. C. W., 1950, Pleistocene history of this lo be incorrect. While Citronelle may coastal Alabama. Bull. Geo!. Soc. Amer. 61 : underlie much of the parish, most of the 1119-1130. surface exposure is clearly of late P leis­ CLENDENIN. W. W., 1896, A preliminary re­ tocene age or younger. Recent geologic por·t upon the Florida parishes of east work, such as the geologic map of the stale Louisiana. p. 187-192. I n Geology and Ag­ by the Louisiana Geological Survey riculture of Louisiana pt. 3, Rept. Louisiana St. Expt. Sta., p. 159-256. (Snead and McCulloh, 1984), unites the Cit­ CRONIN, T. M. , L. M. BYBELL, R. Z. ronelle with the Williana a nd highest POORE. B. W. BLACKWELDER, J.C. LID­ Bentley terraces simply as "High Ter­ DICOAT. and J. E. HAZEL, 1984, Age and races," avoiding the boundary problem en­ correlation of emerged Pliocene a nd Pleis­ tirely. It is worth noting that, as yet, no tocene deposits, U.S. Atlantic Coastal Plain, early Pleistocene (Irvingtonian land mam­ Palaeogeog., Palacoclimatol, Palaeoecol. mal age) vertebrates have been recovered 47(1/2): 21 -51. from the Tunica Hills. An early Pliocene/ DELCOURT, P. A., and H. R. DELCOURT, late Pleistocene unconformity may exist in 1977, The Tunica Hills, Louisiana - Missis­ much of the area. The present description sippi: late glacial locality for spruce and de­ ciduous forest species. Quat. Res. 7: 218-237. provides strong evidence for a nearly DOEHRING, J.A. , 1958, Citronelle age prob­ Pliocene age for the Citronelle Formation lem. Bull. Amer. Assoc. Petrol. Geol. 42(4): in east-central Louisiana, based on the 764-786. presence of biochronologically diagnostic DOMNING. D. P., 1969, A list, bibliography, late Hemphillian horses. and index of the fossil vertebrates of Louisiana and Mississippi. Trans. Gulf Coast XII. LITERATURE CITED Assoc. Ge o!. Socs. 19: 385-422. FISK, H. N ., 1938, Pleistocene exposures in AKERS. W. H., 1972. Planktonic foraminifera western Florida parishes of Louisiana. Bull. and of some forma­ Louisiana Geo!. Surv. 12: 3-25. tions, northern Florida and Atlantic Coastal FISK, H. N., 1939, Depositional terrace slopes Plain. Tulane Stud. Geo\. Paleont. 9(1-4): in Louisiana. Jour. Geomorphology 2(2): 181- 139 p. 200. AKERSTEN, W. A., 1972, Red Light Local !'ORSTEN A. , 1975, The fossil horses from the Fauna (Blancan) of the Love Formation. Texas Gulf Coastal P lain. Jour. Paleontology southeastern Huds peth County, Texas. 49(2): 395-399. Texas Mem. Mus., Bui!. 20: 52 p. GIVENS, C. R., and F. M. GIVENS, 1987, Age R. ALFORD, J. J., C. KOLB, and J . C. and significance of fossil white sprnce (Picea HOLMES, 1983, Terruce stratigraphy in the glauca), Tunica Hills. Louisiana - Mississippi. Tunica Hills of Louisiana. Quat. Res. 19:55- Quot. Res. 27, 283-296. 63. HARRIS, G. D. , and A. C. VEATCH, 1899, A ARATA, A. A., 1966. A Tertiary proboscidean preliminar·y report on the geology of from Louisiana. Tulane Stud. Geol. 4(2): 73- 74. Louisiana. Louisiana St. Expt. Sta., Rept. AUTIN, W. J., J. J. ALFORD, B. J. MILLER, Geol. Agr., pt. 5, 139 p. and R. P. SELF, 1986, The Florida parishes HILGARD, E. W., 1860, Report on the geology of southeastern Louisiana. Geol. Soc. Amer. and agriculture of the state of" Mississippi. E. Barksdale. Jackson, xxiv and 391 p. Cent. Field Guide, Southeastern Sect.. 1986'419-424. IIILGARD, E. W., 1869, P relimina ry report of a BERGGREN. W. A .. D. V. KENT, J . J. geological reconnaissance of Louisiana. FLYNN, and J. A. VAN COUVER!NG, DeBow'sNewOrleansMon. Rev., p. 11. 1985. Cenozoic . Bult. Geo!. HOPKINS, F. V., 1872, Third annual report of Soc. AmeL 9G:l407-1418. the Geological Survey of Louisiana. BERRY, E.W., 1916, The flora of the Citronelle Louisiana St. Univ., Ann. Rept. of 1871, p. Formation. U.S. Geol. Surv. Prof. Paper 163-206. No. I Louisiana Pliocene Horses 4;)

HULBERT, R. C., Jr., 1987, A new Cornwhip­ MATSON, G. E., 1916, The Pliocene Citronelle parion (Mammalia, Equidae) from the Formation of the Gulf Coastal Plain. U. S. Pliocene (latest Hemphillian and Blancan) of Geol. Surv .. Prof. Paper 98-L: 167-192. pis. Florida. Jour. Vert. P aleontology 7(41: 451- 33-42. 468. MA YER, M. J ., 1932, A contribution to the gl'o\. H USSAIN, S. T., 1975, Evolutionar·y and func­ ogy of Bienville Parish, Louisiana. Unpubl. tional anatomy of the pelvic limb in fossil and MS thesis, Louisiana State Uni\". Baton recent Equidae (Perissodactyla, Mammalia). Rouge, 62 p. Anat., Histo!., Embryo!. 4: 179-222. McGEE, W. J., 1981. The Lafayette Formation. ISPHORDING, W. C., 1976, Multivariate min­ 12th Ann. Rept.. U.S. Geo!. Sun: .. pt. 1. p. eral analysis of Miocene - P liocene Coastal 347-521. Plain sediments. Trans. Gulf Coast Assoc. MINCHER. A. R., 1941, The fauna of the Pas· Geol. Socs. 26> 326-331. cagoula Formation. Jour. Paleontology 151.t1: ISPHORDING, W. C., 1983, l nlecprelive 337-348. pis. ·16-47. mineralogy: examples from Miocene Coastal P ARSONS, B. E., 1967. Geologic factors in· Plain sediments. Trans. Gulf Coast Assoc. f1uencing recharge to the Baton Rougt._• Geol. Socs. 33, 295-305. ground-water system, with emphasis on the !SPHORDING, W. C. , a nd G. C. F LOWERS, Citronelle Formation. Unpubl. MS thesi::., 1983, Differentiation of unfossiliferous elastic Louisiana State Univ., Baton Rouge, vii and sed im ents: solutions from the southern por­ 74 p. lion of the Alaba ma - Mississippi Coastal PATTON, T. H .. 1969, Miocene and Plioc(.•ne Plain. T ulane Stud. Geol. Paleont. 17(3): 59- ar tiodactvls, Texas Gulf Coastal Plain. Bull 83. Florida St. Mus., Biol. Sci. 14. 115-226. ISPHORDING, W. C., and G. M. LAMB, 1971 , Age and origin of the Citronelle F ormation in PATTON, T. H., and B. E. TAYLOR. 1971 The Alabama. Bull. Geo!. Soc. Amer. 82(3): 775- Synthetoceratinae (Mammalia, Tylopoda. 779. Protoceratidael. Bull. Arner. Mus. Nat. Hist 145>I 19-218. LER CH , 0., 1892, A prelimina ry report upon the hills of north Louisiana no rth of the U.S. PROTHERO, D. R., und E. M. MANNING and P . Railroad. Bui!. Louisiana St. Expt. 1987, Miocene from tlw T!.'xas Stal., Repl. Geol. Agr. pl. I, 51 p. Gulf Coastal Plain. Joui-. Paleontology 61{2 : 388-123. LOWERY, G. H., 1974, T he mammals of Louisiana and its a djacent waters. LSU QUINN, J. H., 1955, Miocene Equidae of the Texas Gulf Coastal Plain. Pub!. Uni\.· Texas press, Baton Rouge, 565 p. 155161> 102 p., l4 pis. MacFADDEN, B. J., 1984, Systematics and RICHARDS. H. G .. 1938, Some Plcistm·ene phylogeny of , Neohi pparion. Nan­ freshwater mollusks from Louisiana and :\1b­ nippus and Cormohipparion (Mammalia, sissippi. Bull. Louisiana Geol. Sun· 12:27 -16, Equidae) from the Miocene and P liocene of 105-113. the New World. Bull. Amer . Mus. Nat. Hist. ROBERTSON, J. S., 1976, Latest Pliocene 179, act. I> 195p. MacFADDEN, B. J., 1986, Late Hemphillian mammals from Hailt~ XV A. Alachua Co., Florida. Bull. Florida. SL Mus. 20C1l: 111 monodacty! horse s (Mamma lia, Equidae) 186. from the Bone Va lley Formation of Central ROSEN, N. C., 1968. Heavy minerals of tht._ Cit Florida. J our. P a leontology 60: 466-475. ronelle Formation of th~ Gulf Coastal Plain. MacFADDEN, B. J., a nd J . S. WALDROP, Unpubl. Ph.D. dissert., Louisiana State .1980, Nannippu s phlegon (Mammalia, Univ.. Baton Rouge, vii and 187 p. Equidae) from the Pliocene (B\ancan) of ROSEN, N. C., 1969, Heavy minerals and sizl.' F lorida . Bull. Florida State Mus., Biol. Sci. analysis of the Citronelle Formation tlw 25, 1-37. or Gulf Coastal Plain. Jour Sed Petrol. :~!}: MacFADDEN, B. J., and S. D. WEBB, 1982, 1552-1565. The succession of Miocene (Arikareean ROY, C. J., 1939, Type locality oftfw Citrorwlh.• through IJe mphillian) terrestrial mammalian Formalin. Citronelk>, Alabama. Bull. Amer localities and faunas in F lorida. In T. M . Assoc. Petrol. Gcol. 23: 1553-155!). SCOTT and S. B. UPCHURCH. Miocene of the southeastern . F lorida BuL RUSS, D. P., 1976, Rcinterµretation of !lw \\'ii Geol., Spec. Publ. 25>1 86-199. liana-Citronelle relationship I abstr. J. Geo] MARSH, 0. T., 1964, Geology or Escambia and Soc. Amer Abstr Prng. 8161: 1080-1081 Santa Rosa counties, western Florida RUSSELL, E. E .. 1987, Gravel aggregate Ill panhandle. Florida Geo!. S urv., Bull. 46: Mississippi its origin and distribution. :\l1s 140 p. sissippi Geo!. 713l: ·1 p. 46 Tula11e Studies in Geology and Paleontology Vol. 22

SCHULTZ. G. E. (eel). 1977. Guidebook. Field Mammals of North America: Geochronology Conference on !ate Cenozoic biostratigraphy and Biostratigari:)hy. Univ. of Califon1ia of the Texas panhandle and adjacent Ok­ Press, Berkeley, xv ! ~1:{6 p. lahoma. Special Publ. Killgore Res. Center, TEDFORD, R H. and M. E. HUNTER, 1984, Dept. of Geo!. and Anthro. (I): vii and 160 p. Miocene marine-nonmarine correlations, At­ SMITH, M. L., and M. MEYLAN, 1983, Red lantic and Gulf Coastal Plains, North Ameri­ Bluff, Marion County, Mississippi: a Cit­ ca. Palaeogeog., Palaeoclirnatol., Palaeoe­ ronelle braided stream deposit. Trans. Gulf col. 47(1-2): 129-151. Coast Assoc. Geo!. Socs. 33: 419-432. THURMOND, J. T., and D. E. JONES, 1981. SNEAD, J. !., and RP. McCULLOH. comps., Fossil vertebrates of Alabama. Univ. of 1984, Geologic map of Louisiana. 1984. Alabama Press, University. ix t- 244 p. Louisiana Geo I. Surv. Baton Rouge, I: VOORHIES, M. R., 1974, The Pliocene horse 500,000. Nannippus minor in Georgia: geologic impli­ STRINGFIELD, V. T., and P. E. cations. Tulane Stud. Geol. Paleont. 11(2): LaMOREAUX, 1957, Age of Citronelle Fo•·• 109-113. mation in Gulf Coastal Plain. Bull. Amer. WEBB, S. D., 1974, Chronology of Florida Pleis­ Assoc. Petrol. Geo!. 41(4): 742-746. tocene mammals. p. 5-31. In S. D. WEBB TEDFORD, RH., T. GALUSHA, M. F . SKIN­ (ed), P leistocene mammals of Florida, Uni­ NER. B. E. TAYLOR, R W. F IELDS. J. R versity Presses of Florida, Gainesville, x + MACDONALD, T. H. PATTON, J . M. RE­ 270 p. NSBERGER, and D. P. WHISTLER, 1987, WILSON, J. A., 1956, Miocene formations and Faunal succession and biochronology of the vertebrate biostratigraphic units, Texas Ar·ikcireean through Hemphillian interval Coastal P lain. Bull. Amer. Assoc. Petrol. (late Oligocene th1·ough e