Effects of Past Global Change on Life (1995)
Chapter: Periodic Cycles
Often such consequences can be tested against key features of the geologic record. Thus, certain empirical data may constrain models, whereas others test model results. At present, the effective utilization of models is often limited by a paucity of pertinent geologic data. In addition, some models fail, in detail, to square with geologic data.
Any study of past changes in ecosystems demands a certain level of stratigraphic and chronological information. Such information is needed, for example, to determine whether similar events in widespread areas were contemporaneous. It is also essential for documenting global trends—for effectively collapsing data from many regions onto a single time line. Even at a single locality, one must know the approximate length of time separating two different conditions in order to calculate the rate of change that produced the second condition from the first. In general, chronological accuracy increases with decreasing geologic age. Special advantages are gained, for example, by working within the ranges of 14C dating, well-preserved glacial varves, and extant species. Farther back in the record, errors in correlation are frequently smaller than errors in actual dates. High-resolution stratigraphy based on widespread events of brief duration can yield correlations one or two orders of magnitude more accurate than conventional biostratigraphy. For example, chemical marker beds and changing isotopic ratios of carbon and oxygen, which reflect events that spanned less that 105 years, have contributed to a detailed global chronology for rapid environmental change and mass extinction at the Cenomanian-Turonian boundary, about 91 m.y. ago (see Kauffman, Chapter 3). Some events, such as volcanic eruptions and accumulations of chemical fallout from extraterrestrial impacts, have deposited widespread stratigraphic markers within less than a year (Toon et al., 1982). Quantitative statistical methods based on first and last stratigraphic occurrences of species are also yielding improved correlations.
Calibration of sedimentation rates allows for estimation not only of rates of extinction but also of rates of biotic recovery. High rates of deposition yield an expanded stratigraphic record and therefore often improve the quality of both the record and its temporal resolution. Thus, for the terminal Cretaceous event at about 65 m.y. ago, the shallow (middle neritic) deposits exposed at El Kef, Tunisia, seem to offer a more accurate picture of the sequence of events than do deep-sea cores (see Keller and Perch-Nielsen, Chapter 4). For the terminal Ordovician crises about 440 m.y ago, intervals of biotic change have been estimated by using the numbers (including fractions) of graptolite zones that they span (see Berry et al., Chapter 2). The average duration of a zone (on the order of a million years) is estimated from radiometric ages for the boundaries of longer stratigraphic intervals.
SHIFTS BETWEEN ENVIRONMENTAL STATES
Many of the changes that have altered the global ecosystem in the course of Earth history can be viewed as shifts between environmental states. The most important shifts to affect the course of biotic evolution and the nature of the biosphere have been ones that are unique and unidirectional. Others have been components of episodic or periodic cycles, some of which have been superimposed on long-term trends.
Periodic Cycles
The controversial issue as to whether mass extinctions have occurred at equally spaced intervals has stimulated interest in the periodicity of geologic events. The most striking examples of periodic oscillations between environmental states in the recent past are those between glacial maxima and glacial minima during the past 2.5 m.y. These transitions, which have affected sea-level, climates, and biotas, have been linked to periodic changes in the Earth's axial and orbital rotations—the so-called Milankovich cycles. These cycles are best documented by foraminiferal fossils from deep-sea deposits, which exhibit relative
