Extinct plant or animal communities that can be correlated with similar living communities, and thereby be used as paleoclimatic indicators, require a lot of interpretation and guesswork to infer climatic conditions for past ages. Also, such indicator communities become scarcer and harder to identify the farther into the past one tries to look. Strictly geological (i.e. non-biotic) paleoclimate indicators are even more scarce and problematical. Yet, despite the paucity of evidence, one thing is abundantly clear; climates of the past have not always been the same, nor enjoyed so great a diversity, as what we find today.
Evidence of reduced climatic diversity abounds. One particularly telling indicator is the distribution with latitude of individual species. Today, very few species can endure the great variations in climate that occur at widely differing latitudes. Tropical species tend to be restricted to the tropical zones; temperate species to more temperate climates; and polar species to the frigid nether regions of the globe. A more cosmopolitan distribution of species has been noted for the Late Devonian and Early Carboniferous (= Mississippian), and the Late Permian (Dickens, 1993). The same cosmopolitan distribution seems to have occurred during the Late Jurassic and Cretaceous periods.
Hexacorals, the dominant reef-builders of tropical waters today, flourished in the Triassic and into the Jurassic, but languished throughout the Cretaceous, only to reappear and flourish again in the Tertiary (Stanley, 1987). Bands of arid deserts were prominent in the Triassic and Early Jurassic, just as they are today, but were apparently unknown in the Carboniferous, Permian, and Cretaceous. The discovery of Cretaceous crocodillians and sub-tropical vegetation at paleolatitudes above 70° attest to much milder conditions in polar and near-polar regions then, while frigid climates and arctic tundra are found at that latitude today. (Paleolatitude is the latitude at the time the fossils were deposited, adjusted for later motions of Plate Tectonics.)
Existing paleoclimate models are at a loss to explain the climatic conditions inferred from some of these and other biotic indicators (Dickens, 1993, Chandler, 1992, Crowley, 1994). In this article is proposed a new hypothesis which presents arguments for a long-term cyclicity (ca. 190 Million years) in paleoclimates, and attempts to correlate this cyclicity with major extinction events.