to them, at Quaternary time scales of hundreds to tens of thousands of years. As examples, Figure 3.1 shows typical analyses of Quaternary climate change, as measured by stable isotopes and CO2 data depicting temperature changes over the past 50,000 yr; the values for 1000-yr intervals (the typical duration of 1-2 cm of shale or limestone in Phanerozoic basinal facies) are highlighted by triangles and connected by a bold line. The thousand-year Phanerozoic level of analysis effectively replicates the significant patterns of global change in the more finely analyzed Quaternary data. Further, interregional correlation of Quaternary global change data relies heavily on the isotopic record of glaciation and sea-level fluctuations, as regulated through Milankovitch climate cyclicity. Milankovitch cyclostratigraphy is also widely recognized in ancient sedimentary sequences representing nonglacial greenhouse intervals (e.g., for the Cretaceous; Barron et al., 1985; Fischer et al., 1985; Kauffman, 1988a; Glancy et al., 1993), and extensively used for regional correlation as a part of HIRES chronostratigraphy (Hattin, 1971, 1985; Elder, 1985; Kauffman, 1988a; Kauffman et al., 1991). Quaternary and older rocks can thus be correlated at the same scales. Further, event chronostratigraphic resolution in the older stratigraphic record commonly exceeds that possible from Quaternary climate cyclostratigraphy. Whereas Quaternary scientists have the advantage of finer time scales of observation, students of older geological intervals have the advantage of longer continuous records of global change. Both groups have much to contribute to the documentation, interpretation, and prediction of global change phenomena and their short- and long-term effects on Earth ecosystems. An example of high-resolution stratigraphic analysis of an interval of significant past global change and biodiversity crisis is drawn from the Cretaceous record of North America, the Cenomanian-Turonian mass extinction interval (92 to 94 million years ago (Ma)).
An abrupt change in the global marine biota at the Cenomanian-Turonian boundary was first noted by d'Orbigny (1842-1851), who named these stages (Hancock, 1977). Raup and Sepkoski (1984, 1986) and Sepkoski (1990, 1993) statistically determined that the Cenomanian-Turonian boundary interval was a secondary mass extinction on the basis of both familial and generic-level data. They recorded a 5% loss of families and a 15% loss of genera (per-genus extinction rate 0.026), utilizing stage-level data with an average duration of 6 million years (m.y.); this broad resolution did not, however, allow rates and patterns of Cenomanian extinction to be determined. For example, a review of global Cenomanian ammonite extinction data (Collom, 1990) resolved to substage level showed several discrete extinction events among these molluscs within the Cenomanian.
A number of highly detailed regional studies of Cenomanian-Turonian extinction patterns, and associated environmental perturbations, have been conducted during the last three decades which shed light on the dynamic nature of this interval. Of particular interest are the new data from the Western Interior of North America (Figure 3.2), from western and central Europe, from North Africa, and from northern South America (Colombia, Venezuela, and Brazil). These studies do not support the idea of a single "catastrophic" boundary extinction among ecologically and genetically diverse taxa, but rather a 1.46-m.y. interval of ecologically graded extinction characterized by several discrete, short-term events with global expression ("steps" of Kauffman, 1988b), most of them in the last 520,000 yr of the Cenomanian. Each of these discrete extinction events is associated with geochemical evidence for environmental chaos in the ocean-atmosphere system—short-term, large-scale perturbations in temperature, ocean chemistry, and carbon and nutrient cycling in the sea, upon which are superimposed a somewhat exaggerated expression of Milankovitch climate cyclicity at 41,000-and 100,000-yr intervals (subsequent discussion). Collectively, these integrated, high-resolution (centimeter to meter scale) paleobiological, geochemical, and sedimentological studies provide a global perspective of the rates, patterns, causes, and consequences of the C-T mass extinction interval.
In the Western Interior Basin of North America (Figure 3.2), bed-by-bed stratigraphic and paleontologic studies by Cobban and Scott (1972) and later Cobban (1985, 1993) documented an abrupt change in ammonite assemblages across the C-T boundary; they noted the loss of typical late Cenomanian ammonites at five to six discrete levels within 4 m of the C-T boundary, spanning the biozones of Sciponoceras gracile (Vascoceras diartanum and Euomphaloceras septemseriatum subzones) and, above it, Neocardioceras spp. (Eoumphaloceras navahopiensis, Neocardioceras juddi, and Nigericeras scotti subzones, ascending). Cobban (1985, 1993) and Kennedy and Cobban (1991) defined the C-T boundary between the latest Cenomanian Neocardioceras biozone, Nigericeras scotti subzone, and the overlying Watinoceras biozone, W. devonense subzone. Kauffman (1975) and Kauffman et al. (1976, 1993) also noted a major change in Bivalvia at the ammonite-based C-T boundary of North America and Europe, with the disappearance of latest Cenomanian Inoceramus pictus and I. tenuistriatus lineages, overlain by the first common occurrence of early Turonian Mytiloides