The Upper Cretaceous strata of the Gulf Coast region of México, Texas, and southwestern Arkansas contain rich, abundant, well-preserved, and, hitherto, poorly studied assemblages of planktonic foraminifera. The rapid evolution and cosmopolitan nature of these planktonic microfossils make them an ideal biostratigraphic tool for the development of detailed, long-distance systems of zonation. In the present report, the planktonic foraminifera have been utilized to subdivide the Upper Cretaceous of the western Gulf Coast region into the following biostratigraphic units: (1) The Rotalipora s.s. Assemblage Zone: Rotalipora evoluta Subzone to Rotalipora cushmani-greenhornensis Subzone. Late Washitian to early Eagle-fordian (early to late Cenomanian). (2) The Marginotruncana helvetica Assemblage Zone: Marginotruncana sigali Subzone to Whiteinella archaeocretacea Subzone. Middle to late Eagle-fordian (early to late Turonian). (3) The Marginotruncana renzi Assemblage Zone: Early Austinian (Coniacian). (4) The Globotruncana bulloides Assemblage Zone: Marginotruncana concavata Subzone to Globotruncana fornicata Subzone. Middle to late Austinian (early to late Santonian). (5) The Globotruncana fornicata—stuartiformis Assemblage Zone: Archaeoglobigerina blowi Subzone (Dictyomitra multicostata Zonule to Planoglobulina glabrata Zonule); Globotruncana elevata Subzone (Pseudotextularia elegans Zonule to Globotruncana calcarata Zonule); and Rugotruncana subcircumnodifer Subzone (Globotruncana lapparenti s.s. Zonule to Rugotruncana subpennyi Zonule). Archaeoglobigerina blowi Subzone = early Taylorian (early Campanian); Globotruncana elevata Subzone = late Taylorian (late Campanian); and Rugotruncana subcircumnodifer Subzone = early Navarroan (early Maestrichtian). (6) The Globotruncana contusa—stuartiformis Assemblage Zone: Globotruncana gansseri Subzone to Abathomphalus mayaroensis Subzone. Middle to late Navarroan (middle to late Maestrichtian). This system of zonation is based (1) on the association of diagnostic taxa at given stratigraphic horizons; (2) the range zones and concurrent range zones of the various taxa; (3) the relative abundance of important taxa at various stratigraphic horizons; and (4) the phylogeny and evolution of Upper Cretaceous planktonic foraminifera. It has become established through the analysis of more than 1000 fossiliferous samples from the surface and sub-surface in both the western part of the Gulf Coast region and in the Caribbean region. Where possible, samples were collected within the framework of measured sections of the lithic units under study. Samples for planktonic foraminifera were collected as far south in the western Gulf Coast region as the approximate latitude of Tampico, México (22° north latitude) and as far north as Brownstown, Sevier County, Arkansas (34° north latitude). At a given stratigraphic horizon, such as the Globotruncana elevata Subzone, there are few species of planktonic foraminifera that do not occur in both the Tethyan faunal province and southern part of the Boreal faunal province. In most cases planktonic species that do not occur in both areas are new species, whose stratigraphic and geographic distribution are yet unknown. The present study indicates that the above-mentioned system of zonation is applicable at the zonule level at least as far north as the latitude of Brownstown, Arkansas. Furthermore, Olsson’s (1964) work in New Jersey and investigations in progress in California by the writer, Douglas, and others suggest that the system of zonation introduced here can be applied at the subzone level as far as 40° north latitude in eastern North America and as far as 34° north latitude in western North America. One of the chief by-products of this study is the creation of a regional correlation chart for the Upper Cretaceous of the western Gulf Coast region that is based on planktonic-foraminiferal zonation. Accurate biostratigraphic dating of lithic units in eastern México utilizing planktonic foraminifera indicates that a number of formational units are time transgressive from north to south. Units like the San Felipe Formation and Agua Nueva Formation are considerably older in northern México near Monterrey than they are in southern México near Tampico. The present study has yielded few radical changes in the dating of lithic units in Texas and Arkansas previously established on the basis of megafossils (Stephenson and others, 1942). Notable among these changes are Navarroan ages for the Upson Clay and San Miguel Formation of the Rio Grande area of Texas (approximate latitude of Eagle Pass) and all, but perhaps, the lowermost part of the Marlbrook Marl of Arkansas. Correlation with the type-European Upper Cretaceous stages is rendered difficult (1) by the imprecise and often obsolete definition of the stages in their type areas and (2) by a lack of accurate data concerning the stratigraphic distribution of planktonic foraminifera in the type sections of the stages. The first of these problems affects all Upper Cretaceous biostratigraphy irregardless of the group of organisms used for correlation. It can perhaps only be solved by an international stratigraphic commission chosen to modernize the definition of the European stages. The second problem can be solved by detailed sampling of strata included in the type sections of each European stage for planktonic foraminifera. Until a detailed reanalysis of the majority of the type-European stages is made, the writer prefers to use North American stage names — particularly those of the standard Gulf Coast section. European stage names are used in this report only in a tentative way. The Eaglefordian Stage of the standard Gulf Coast Upper Cretaceous section has been subdivided into 3 new substages: (1) the Lozierian; (2) the Bocian; and (3) the Sycamorian. The Boquillas Formation in Val Verde and Terrell Counties, Texas, has been subdivided into a lower unit termed the Rock Pens Member and an upper unit termed the Langtry Member. The terms Ateo Chalk and Bruceville Chalk Marl, first used informally by Durham (1957), have been formally introduced here.

A knowledgeable choice for a stage boundary stratotype is dependent upon obtaining high-resolution stratigraphic data. Detailed analyses conducted for the two potential Turonian–Coniacian stage boundary stratotypes that were considered at the Second Cretaceous Stage Boundary Symposium provide both positive and negative insights for consideration. The Salzgitter-Salder Quarry in central Germany (which was recommended by the symposium) contains abundant bivalve fossils, including the recommended boundary datum: the lowest occurrence of the inoceramid bivalve Cremnoceramus deformis erectus. Foraminifera are also abundant, but extensive diagenetic recrystallization seriously degrades nannofossil and palynomorph recovery and limits the potential of the section for stable isotope stratigraphy and radiometric dating. Furthermore, palaeoenvironmental analysis indicates that much of the Salzgitter stage boundary interval has resulted from allochthonous sedimentation, indicating that the well-developed lithological cyclicity between limestone and marlstone that occurs in the section is largely autocyclic. The orbital cyclostratigraphic potential of the section is therefore also in question.The Wagon Mound outcrop in northeastern New Mexico, USA, has good recovery and biostratigraphic control for all three microfossil groups, but the base of C. deformis erectus occurs above the section, by definition placing the section entirely in the Upper Turonian Well-preserved ammonites and inoceramid bivalves are also recoverable in over half of this section. Facies have not been recrystallized and represent continuous autochthonous sedimentation with sharply defined lithological cyclicity between limestone/marlstone couplets on a fine stratigraphic scale. In addition, a number of bentonites with proven datability occur in the section. Thus the bio- and chemostratigraphic dating potential, as well as the radiometric dating potential, of the section are good. However, much of the section is composed of carbonaceous, dysoxic facies with abnormal marine micro- and macrofossil assemblages, limiting study of the stratigraphic or palaeoecological trends leading up to the boundary.The absence of the datum at Wagon Mound is puzzling, because microfossil biostratigraphy suggests that the section is substantially coeval with the Salzgitter-Salder section. The C. deformis erectus datum may thus be diachronous. Until the suitability of the recommended boundary datum is addressed, a reasoned choice of a section for the boundary stratotype is not possible. In any case, the absence of C. deformis erectus and the abnormal facies in the lower part of the Wagon Mound section, and the extensive diagenesis and partly allochthonous nature of the Salzgitter-Salder section, are serious enough problems to warrant rejection of both sections as stratotypes. AbstractA knowledgeable choice for a stage boundary stratotype is dependent upon obtaining high-resolution stratigraphic data. Detailed analyses conducted for the two potential Turonian–Coniacian stage boundary stratotypes that were considered at the Second Cretaceous Stage Boundary Symposium provide both positive and negative insights for consideration. The Salzgitter-Salder Quarry in central Germany (which was recommended by the symposium) contains abundant bivalve fossils, including the recommended boundary datum: the lowest occurrence of the inoceramid bivalve Cremnoceramus deformis erectus. Foraminifera are also abundant, but extensive diagenetic recrystallization seriously degrades nannofossil and palynomorph recovery and limits the potential of the section for stable isotope stratigraphy and radiometric dating. Furthermore, palaeoenvironmental analysis indicates that much of the Salzgitter stage boundary interval has resulted from allochthonous sedimentation, indicating that the well-developed lithological cyclicity between limestone and marlstone that occurs in the section is largely autocyclic. The orbital cyclostratigraphic potential of the section is therefore also in question.The Wagon Mound outcrop in northeastern New Mexico, USA, has good recovery and biostratigraphic control for all three microfossil groups, but the base of C. deformis erectus occurs above the section, by definition placing the section entirely in the Upper Turonian Well-preserved ammonites and inoceramid bivalves are also recoverable in over half of this section. Facies have not been recrystallized and represent continuous autochthonous sedimentation with sharply defined lithological cyclicity between limestone/marlstone couplets on a fine stratigraphic scale. In addition, a number of bentonites with proven datability occur in the section. Thus the bio- and chemostratigraphic dating potential, as well as the radiometric dating potential, of the section are good. However, much of the section is composed of carbonaceous, dysoxic facies with abnormal marine micro- and macrofossil assemblages, limiting study of the stratigraphic or palaeoecological trends leading up to the boundary.The absence of the datum at Wagon Mound is puzzling, because microfossil biostratigraphy suggests that the section is substantially coeval with the Salzgitter-Salder section. The C. deformis erectus datum may thus be diachronous. Until the suitability of the recommended boundary datum is addressed, a reasoned choice of a section for the boundary stratotype is not possible. In any case, the absence of C. deformis erectus and the abnormal facies in the lower part of the Wagon Mound section, and the extensive diagenesis and partly allochthonous nature of the Salzgitter-Salder section, are serious enough problems to warrant rejection of both sections as stratotypes.