Evaluation of the paleoecology and paleoclimatology was accomplished on the basis of floral and faunal evidences. Fossil forms were compared directly with modern forms that were considered most similar in morphological respects. It was assumed that the fossil plants or animals being studied lived under ecological and climatological conditions similar to those under which their closest modern counterparts now live. It necessarily also was assumed that the most reliable evidence must come from plants and/or animals whose climate requirements were restricted and whose migratory abilities were low. Obviously when reconstructions of past conditions are made using this technique, caution must be practiced (i.e., "seeds" may be ingested by birds and deposited at great distances from their origin). Nevertheless, much may be inferred from this type of analysis.
Ideally one might hope that, because of their abundance, the fossil plants would provide substantial evidence concerning the paleoecology and paleoclimatology of the deposits. Unfortunately this is only partially true. While the fossil plants leave little doubt of the presence of grasslands environment (most probably, savanna parkland) with adjacent riparian communities, they do not suggest a specific climate. All of the fossil plants recovered to date have modern counterparts inhabiting a wide variety of climatological regimes (i.e., modern Stipa comata Trin. & Rupr. is distributed from Indiana to the Yukon Territory and south to Texas and California). At the present time certain fossil vertebrates (i.e., tortoises) appear to be more accurate paleoclimatological indicators.
While a detailed analysis of all the vertebrate fossils recovered in Ellis County is not complete, the presence of the variety of taxa so far identified (Zehr, 1974; Tallan, 1978; Zakrzewski, Holman, personal communications) would seem to indicate that the major community represented is a savanna with widely and evenly spaced trees or perhaps a savanna parkland in which a mosaic of grassland and small and large stands of trees and shrubs occur. The occurrence of grazers and semigrazers such as horses (Pliohippus), camels, and cursorial rhinoceras (Aphelops) is representative of a grassland community in which extensive stretches of grassland favoring straight-line escape would have been advantageous. The presence of peccary (Prostennhops) probably indicates a savanna parkland (Schultz and Martin, 1975). Modern tortoises, according to Brattstrom (1961), "occur in tropical jungles, forests, or savannas, most often in open areas in these habitats." There is no evidence in the Ellis County deposits to suggest jungles and forests, so savannas are the most probable community in which the tortoises occurred. Widespread treeless and shrubless grasslands are not suggested since tortoises require at least some shade to prevent absorption of too much solar radiation and consequent death upon reaching a lethal temperature (Brattstrom, 1961). Brattstrom suggested a need for forests, but this is not necessary since a savanna parkland environment would provide abundant situations to meet the shade requirements of tortoises. The presence of succulent vegetation is also suggested by the tortoises, which are primarily herbivores consuming this type of plant growth. Hibbard (1960) viewed deposits with similar tortoises as representing a "moist subhumid, subtropical savanna with forests and tall grasses along the river valleys, with shrubs and tall grasses on valley walls and on the low divides. Some short grasses may have occurred on the higher and well-drained divides." Although his description may encompass more detail than is justified by the evidence, certainly the pervasive idea of grasses, and hence a grassland community of some sort, is defensible. Fossil ground squirrels (Spermophilus) found at some sites also are indicative of a grassland community.
In recent years several authors have published articles on the paleoclimatological significance of giant land tortoises. Hibbard (1960) examined the available evidence with regard to such modern and fossil tortoises and presented convincing evidence that, "the large Pliocene and Pleistocene land tortoises could not stand freezing, but they may have existed in an area where very few light frosts occurred at night but the temperature during the day warmed to 60° or more Fahrenheit." He indicated that during the period when the thin-shelled tortoises (c.f., Geochelone orthopygia Cope) existed (all Ellis County tortoises are thin-shelled) the climate was subtropical and subhumid. Brattstrom (1961) reviewed the climatological requirements of modern tortoises and their implications with regard to the occurrence of similar fossil tortoises. Since the large tortoises are cold-blooded and presumably did not burrow, moderate temperatures were considered a necessity. Too low or too high temperatures would have led to the rapid death of the tortoises. He concluded that "the presence of gigantic tortoises in more northern latitudes than today (e.g., Kansas in the Pliocene, Montana in the Miocene, Wyoming in the Oligocene, etc.) suggests that the climates were much warmer and more constant then." Holman (1971) has indicated that the presence of thin-shelled giant land tortoises in upper Miocene deposits he studied suggests a "subtropical climate, with winter temperatures not lower than 32° F. . . ."
Summarizing the vertebrate evidence, there seems little doubt that during the deposition of the Tertiary, Ellis County strata in which the giant land tortoises are found, the climate could be reasonably characterized as subhumid and subtropical, with temperatures rarely, if ever, reaching O° C or exceeding 38° C. At least some seasonal variation might have occurred. I have found fossil wood in these deposits with distinct annual rings, probably indicating the occurrence of wet and dry seasons.
Conflicting opinions concerning the original genesis of the present-day treeless grasslands of central North America are abundant in the literature. Some authors consider the grasslands to have been in existence as early as Late Miocene (Elias, 1932, 1942; Beetle, 1961), while others believe the widespread treeless grasslands are a result of Pleistocene climates and thus of recent derivation (MacGinitie, 1962; Wells, 1970). At least one author (Ross, 1970) has proposed the possible expansion and contraction with each interglacial and glacial period of a grassland initially developed during the late Tertiary, thus combining both the pre-Pleistocene and Pleistocene genesis theories. Dix (1964) has recently reviewed the history of biotic and climatic changes that may have had an effect on the origin of grasslands in the Great Plains area. He concluded that the original grasslands evolved from elements of forests during the Oligocene into temperate and tropical savannas. However, he suggested that it was not until late Pliocene and Pleistocene that the present-day treeless grasslands evolved.
Of all the evidence indicating the development of treeless grasslands, that provided by the evolution of particular groups of vertebrates is the strongest. The evolution of hares and horses seems especially important in this respect.
Presently in North America, the lagomorphs exist as two families: Leporidae, the rabbits and the bares, and Ochotonidae, the pikas, short-legged, rabbit-like mammals found mainly in the western United States. The lagomorphs as a group are vegetarians whose diet consists mostly of leaves and young sprouts, although in severe winters they may turn extensively to woody stems and bark.
Lepus, the genus of hares in Leporidae including the jackrabbits of the High Plains, would seem to be closely linked to the development of treeless grasslands. Adapted for rapid movement in one direction over open areas, they seem ideal as an indication of treeless or nearly treeless grasslands. In areas such as Argentina where grasslands are present, but there are no jackrabbits, the "jackrabbit niche" is filled by the Patagonian cavy, a relative of the guinea pig which has long hind legs.
Wood (1940) summarized the trends in evolution of the fossil lagomorphs of the plains:
It is unquestionable that the increasing aridity of the plains during the later Tertiary brought about a reduction in the numbers of places in which rabbits could take refuge, and that as a result those individuals, races, species or genera that were the best adapted for sustained flight had an increased survival rate. There would be a positive selection of those mutations that tend to increase the rapidity of the animal's running and its endurance.
Later Dawson (1958, 1967) developed a phylogeny of the Leporidae, in which there is an increased hypsodonty and cement deposition on teeth during evolution from primitive to more advanced forms from the Oligocene through the Pleistocene. She also notes that Hypologus, a genus of fossil rabbits extant during the later Miocene through the early Pleistocene, was generally not developed for the rapid straight-line type of locomotion characteristic of an open-area genus like Lepus. It possibly became extinct because Hypologus, "lagged behind Lepus in adaptations of the dentition for gnawing and a skeleton for running." This extinction did not take place until Late Pliocene and early Pleistocene and would indicate the appearance of an open treeless grassland.
Shotwell (1961) reviewed the evidence on the evolution of the fossil horses of the late Tertiary. He concluded that the three-toed Hipparion horses occupied savanna and that the two lateral toes provided traction in dodging maneuvers in areas where obstructions (such as trees and brush) were present, whereas Pliohippus, a presumed ancestor of modern horses, with its one big toe showed an adaption to high-speed escape in straight lines where agility was a hindrance, and thus occupied open grassland.
The Hipparion horses were present throughout the central High Plains in the Miocene and Pliocene but had become restricted to an area below southern Kansas by early Pleistocene. Equus (Plesippus), on the other band, became more widespread. The habitat of the Hipparion horses, the savanna, had become increasingly restricted southward, but this did not happen until Early Pleistocene. The fact the Equus (Plesippus) remained, and the Hipparion horses left, indicates the presence of widespread, treeless grasslands. Thus, we conclude that the first treeless grasslands did not become the dominant vegetation in the High Plains until at least very late in the Tertiary.
Summarizing, although grasslands were undoubtedly present in some form from the Oligocene on, treeless grasslands are only a development of the latest Tertiary and earliest Pleistocene. Certainly, there is no evidence to suggest the presence of widespread treeless grasslands during the late Miocene-Early Pliocene deposition of sediments in Ellis County.
Kansas Geological Survey, Geology
Placed on web May 1, 2009; originally published September 1979.
Comments to firstname.lastname@example.org
The URL for this page is http://www.kgs.ku.edu/Publications/Bulletins/218/05_paleo.html