In order that a discussion of sedimentary strata be meaningful it is necessary that a system of classification be used so that various units may be designated and named for recognition. Also, these units must be correlated from place to place or the classification and description has significance only for an individual locality. These facts have been recognized since the beginning of stratigraphical geology and general principles applicable to the classification and correlation of marine strata have been developed and used by many workers.
In the case of glacial deposits the recognition of a significant stratigraphic sequence developed relatively late; and as these deposits were studied principally by observation of their physiographic expression, they were considered by many geologists to be unique in stratigraphy and not adapted for inclusion in the general framework of classification and correlation evolved for marine strata. Although such features as degree of dissection, intersecting end moraines, and other physiographic features have been used extensively, sound stratigraphic criteria also have been applied in the study of glacial Pleistocene sediments for more than 50 years. The number of localities exposing superposed tills, or tills and other sediments, is relatively small but a general understanding of the sequence of glacial events in the upper Mississippi River basin region has become well established.
A completely different approach to these young sediments existed in the region beyond the limits of glaciation or marine transgressions. While a Pleistocene chronology was being developed by the study of glacial deposits in north-central United States, fossil vertebrates were being collected at many other localities. The Great Plains of Kansas and adjacent states contributed importantly to these collections. A chronology, unrelated to the glacial sequence, developed from the study of these fossil vertebrates. It was based in part on evolutionary changes in land mammals, and later on attempted correlation of these vertebrate remains with described forms occurring in other continents. In this independent chronology, even though localities showing superposition of distinctive assemblages of fossil vertebrates are exceedingly rare, the definition, classification, and correlation of stratigraphic units rested almost entirely on the one criterion of fossil vertebrates. The lack of adequate assemblages of fossil vertebrates from known stratigraphic positions in the glacial sequence retarded a correlation of the two chronologies.
The zone of transition from the glacial province to the non-glacial province crosses several of the states in the central Missouri River basin. The specialized techniques of neither the vertebrate paleontologist nor the glacial geologist are usable in the other's province. These facts, coupled with expanded programs of ground-water and engineering geology, have prompted a relatively large number of workers in the Missouri River basin region to attempt to place the study of Pleistocene deposits on a sound, well-rounded stratigraphic basis; to remove these deposits from the "special" category; and to use bases and principles of classification and correlation generally accepted for use in other parts of the rock column. Techniques of correlation now in general use in this region include, in addition to physiographic expression, fossil vertebrates, and stratigraphic succession, three general types of evidence that cross the glacial boundary and occur abundantly in both the glacial and nonglacial provinces. These are fossil molluscan faunas, morphology and continuity of buried soils, and petrographically distinctive volcanic ash.
In any classification system the units must be defined for consistent use. Unfortunately, as independent chronologies developed for the glacial and nonglacial Pleistocene deposits, conflicting philosophies have developed concerning the nature of stratigraphic units. Although these divergent viewpoints are rarely expressed in published papers, it becomes evident on critical examination of the literature that in dealing with glacial deposits many geologists accept lithologic continuity of till sheets as indicating contemporaneity throughout the traceable extent of the deposit, whereas many vertebrate palentologists are prone to use a formation name to include all deposits yielding similar fossils, regardless of lithologic differences. That is, while the glacial geologist may assign, quite properly, a formation name to a till sheet he often assumes improperly that the formation represents an identical unit of time throughout its extent. In contrast, some vertebrate paleontologists having collected similar fossils from several localities and presuming that the fossils establish a faunal zone that may be of the same age at the several localities, have assigned a formation name to the faunal zone and treated it in a manner appropriate to a lithologic stratigraphic unit. Obviously where these two practices are followed by different workers in the same area the resulting nomenclature becomes confused beyond comprehension.
The problem of classification is complicated in the nonmarine Pleistocene by the degree of subdivision and predominance of deposits reflecting several strongly contrasting nonmarine environments but the basic problem of definition of kinds of units is the same for these youngest sediments as for all strata of the rock column. This general problem has attracted attention in recent years and Moore (1947) has proposed in a recommendation to the American Commission on Stratigraphic Nomenclature that three kinds of units be recognized--time units, time-rock units, and rock units. He has further proposed that appropriate spellings be used to differentiate clearly rock units (written as nouns) from the units designating a time interval (written as adjectives). A consistent differentiation of these kinds of units serves to clarify the confusion that has resulted from conflicting bases of stratigraphic classification.
In the marine rocks, segmentation of the sequence for the establishment of time-rock and time units has commonly been based on unconformities, hiatuses in sedimentation, or marked change in adjacent faunas where physical evidence is lacking. As Pleistocene time is relatively short and includes the present, it is desirable to divide it into shorter and more sharply restricted units than are used for older time intervals of Epoch rank. Therefore mere recognition of unconformities in an individual section becomes inadequate for a full definition of the several Ages. Continental glaciation was the most distinctive event of the Pleistocene and glacial till the most distinctive lithology. The interval of time characterized by the active deposition of glacial till during an episode of continental glaciation is taken as defining the span of a glacial Age, and the interval of time separating two such episodes is taken as defining an interglacial Age. Deposits made during an Age constitute a Stage.
For example, a continuously traceable till sheet, such as the Kansas till, is a rock unit of formational rank and is called by the same name throughout a region extending hundreds of miles north and east from Kansas although it is known to have been made by a migrating ice margin and to be of somewhat different age at different places. The Kansan Age, on the other hand, is a unit of time that meets, without overlap or hiatus, comparable units of time called Aftonian Age and Yarmouthian Age below and above. The time, like the till sheet, takes its identity from the particular advance and retreat of a continental glacier that characterized or produced it, but unlike the till sheet it embraces the same interval of years everywhere regardless of the particular events of deposition or erosion at any one spot. Therefore the Kansas till at most places where it is observed represents only a fraction of Kansan time. Furthermore, stratigraphic units in addition to till were deposited during Kansan time.
In this report, as is standard practice in publications of the Kansas Geological Survey, lithologic units are named as nouns, while adjectival endings on names denote time and time-rock units.
All units of first rank subdivision of the Pleistocene may be considered, as proposed by Flint and Moore (1948), both time units (Ages) and time-rock units (Stages). In application to the interglacial intervals the term Age is quite appropriate; however, the application of Stage is not so clear because the physical record of much of this time is preserved in soil profiles that represent the alteration of earlier rocks. Nevertheless, as these profiles of weathering furnish a physical record of interglacial time, it is deemed appropriate to class these terms also as both Ages and Stages.
The term "Pleistocene" is generally assigned to the last major unit of Series (Epoch) rank in the standard scheme of stratigraphical classification. Although the term is used on all continents, it has been unsatisfactory because of the striking inconsistencies in the defined span of time and rocks included within it in the various regions and depositional provinces. "Pleistocene" was introduced by Lyell in 1839 to apply to marine strata in the Mediterranean region that contain molluscan faunas, the species of which are more than 70 percent living. In the light of present knowledge such a definition would include much of the time now universally assigned to the late Tertiary. In 1846 Forbes used the word Pleistocene to apply to the "glacial epoch," thus giving a climatic implication--a redefinition to which Lyell agreed in 1873. A definition based on climatic change as evidenced by continental glaciation is almost universally used for Pleistocene in central and northern North America as indicated by the official usage of the U. S. Geological Survey (Wilmarth, 1925, p. 49); ". . . Pleistocene epoch includes the deposits of the Great Ice Age, as it is popularly known, and contemporaneous marine, fiuviatile, lacustrine, and volcanic rocks. In some areas it also probably includes some preglacial deposits and some postglacial deposits older than those of the Recent epoch."
Nevertheless, in the typical region of the northern and western Mediterranean "Pleistocene" has been used in diverse ways. In recent years the disagreements among workers in this region have been reduced largely to two possible stratigraphic positions that might properly be regarded as the base of the Series, namely the top of the marine Calabrian (Gignoux, 1943) and its presumed nonmarine equivalent, Villafranchian, or at the base of these same units (Movius, 1949; Migliorini, 1950). As meaningful classifications of rocks must be based on a standard of reference (albeit our correlation with a type may be faulty), a resolution of this problem is vital.
At the 18th International Geological Congress in Great Britain (1948) a commission was appointed to make recommendations on the Pliocene-Pleistocene boundary line. The commission report, unanimously accepted by the Council of the Congress, was as follows (Oakley, 1950, p. 6):
(1) The Commission considers that it is necessary to select a type-area where the Pliocene-Pleistocene (Tertiary-Quaternary) boundary can be drawn in accordance with stratigraphical principles.
(2) The Commission considers that the Pliocene-Pleistocene boundary should be based on changes in marine faunas, since this is the classic method of grouping fossiliferous strata. The classic area of marine sedimentation in Italy is regarded as the area where this principle can be implemented best. It is here too that terrestrial (continental) equivalents of the marine faunas under consideration can be determined.
(3) The Commission recommends that, in order to eliminate existing ambiguities, the Lower Pleistocene should include as its basal member in the type area the Calabrian formation (marine) together with its terrestrial (continental) equivalent the Villafranchian.
It is hoped, for sake of clarity and understanding among workers, that geologists generally will accept the decision of this international body. It is a particularly happy decision for workers in the United States (Moore, 1949) because both in Europe (Migliorini, 1950; Venzo, 1952) and, as will be shown later, in central United States this boundary coincides rather closely with the advent of the first major glaciation and thus matches the definition currently in use by glacial geologists.
The Kansas Geological Survey recognizes Tertiary and Quaternary as the Period-Systems within the Cenozoic Era. This rather outmoded usage is retained primarily because these terms are deeply rooted in the literature and they are in current use by many other official Geological Surveys. This has little bearing on the present discussion as in Kansas usage Quaternary Period covers the same time span as its only contained Epoch, the Pleistocene.
Subdivisions of Pleistocene time now in general use have evolved largely as the result of work in the glaciated region of the upper Mississippi Valley. Although arbitrary, the general framework developed in that area constitutes a usable Pleistocene time scale and, with minor modification of the Wisconsinan Sub-Ages, is accepted for use in Kansas. The time scale accepted as standard in Kansas is shown in Figures 1 and 2.
The earliest Age (Stage) is called the Nebraskan. Seventy years ago the concept of multiple glaciation was established in Iowa by the work of McGee (1878, 1879, 1881) and others and in 1894 Chamberlin (pp. 753-764; 1895) (also Calvin, 1896) applied the name Kansan to the lower of two tills in the Afton Junction-Thayer area of Union County, Iowa, on the basis of the supposed extent of this till into Kansas. In 1897 Bain (p. 464) stated: "A preliminary examination as far south as Kansas City seemed to show that the older drift did not come to the surface, and accordingly the upper drift at Afton Junction is presumably the surface drift of eastern Kansas, though the matter has not been fully studied." As a result of Bain's work the term Kansan was transferred to the upper drift in the Union County, Iowa, exposures and the name Albertan or sub-Aftonian assigned to the lower one (Calvin, 1897). The name Nebraskan was later proposed for this lower till by Shimek (1909) on the basis of its supposed westward extent in Nebraska, and this term has enjoyed general acceptance in the Missouri basin region.
The Aftonian Age (Stage), the second time unit in the Pleistocene, is an interglacial interval. This name first was proposed for sediments exposed in the Afton Junction, Iowa, region (Chamberlin, 1894a; Calvin, 1897) and has been generally used during the present century to designate the post-Nebraskan pre-Kansan interglacial interval.
The Kansan Age (Stage), the third segment of Pleistocene time and the second glacial interval, takes its name from the State of Kansas. Although the name Kansan was proposed originally for deposits in Union County, Iowa, the name was transposed from the lower to the upper of the two tills exposed there after southward tracing revealed that the upper till is equivalent to the surface till of northeastern Kansas. Therefore, it may be assumed that the Kansas region is the type locality for this Stage.
The Yarmouthian Age (Stage) is the interglacial interval next following the Kansan. The name Yarmouth was proposed by Leverett (1898b) to apply to the soil that served to separate the tills of Kansan and Illinoian ages in east-central Iowa. It is the oldest of the interglacial Ages to be based on a buried soil. The name Yarmouth is now used in Kansas for the buried soil in this stratigraphic position while the name Yarmouthian is assigned to the Age (Stage). In his description Leverett (1898b, p. 239) stated: "The presence of this soil horizon was first brought to the writer's notice by a well section at Yarmouth in Des Moines County, Iowa. For this reason, and because the name of this village is less likely to be confusing than names which are more common, it seems appropriate to apply the name Yarmouth to this weathered zone."
The Illinoian Age (Stage) is the glacial interval next following Yarmouthian. Leverett in 1899 proposed the name Illinois glacial lobe, from the State of Illinois, to include the drift sheet so extensively developed in that state. Subsequent usage has established the name Illinoian for this interval of Pleistocene time.
The Sangamonian Age (Stage) is the youngest of first-rank interglacial intervals within the Pleistocene. Like the Yarmouthian, the name was originally based on a buried soil, and similarly the name Sangamon is now used in Kansas to apply to the buried soil formed during this time. The name Sangamon was proposed by Worthen in 1873 for a soil developed in Illinois drift and overlain by Iowa loess in Sangamon County, Illinois. Leverett in 1898 used the term to denote the interval between the Illinoian and Iowan glacial episodes.
The Wisconsinan Age (Stage) is the fourth and last of the first-rank subdivisions of Pleistocene time that is characterized by continental glaciation. Unlike the preceding Ages, however, the Wisconsinan contains repeated glacial advances and retreats and has been subdivided into Sub-Ages (Substages). Kay and Graham (1943, p. 89) have summarized the early development of this classification in Iowa as follows:
The name East Wisconsin was first used by Chamberlin in 1894 for the most recent of the glacial ages of the Pleistocene period. The following year at the suggestion of Upham the name was shortened to Wisconsin (Chamberlin, 1895). Soon two substages, early Wisconsin and late Wisconsin, were recognized. More recently Leverett (1929) described three substages, early, middle and late Wisconsin. The early Wisconsin drift was interpreted to have come from the Labradorean center, the Middle Wisconsin from the Patrician center, and the Late Wisconsin from the Keewatin center. Then in 1931, as a result of detailed field studies in Illinois, Leighton (1931) proposed a modification of the use of the name Wisconsin in Pleistocene classification. The change involved the elimination of the Peorian as an interglacial stage and the inclusion of the Iowan stage as the oldest substage of the Wisconsin. The evidence in Illinois had convinced Leighton and his associates that there was continuous deposition of loess, previously interpreted to belong to the Peorian interglacial age, from Iowan time until after early Wisconsin time; and furthermore, that the interval heretofore called Peorian was so short as to necessitate its elimination as an interglacial age from the classification of the Pleistocene. Moreover, Leighton proposed that since the Iowan glacial age cannot longer be recognized in Pleistocene classification as being independent in age from Wisconsin age, the usage of the name Wisconsin of our present classification be modified to include the Iowan as the earliest of its substages. He proposed also that the Wisconsin substages be named, from oldest to youngest, Manitoban (Iowan), Quebecan (Early and Middle Wisconsin), and Hudsonian (Late Wisconsin). Later he realized that the terms herein used were not appropriate. . . . Leighton (1933) therefore proposed other names. . . They are, from oldest to youngest, the Iowan, the Tazewell (Early Wisconsin), the Cary (Middle Wisconsin), and the Mankato (Late Wisconsin).
The names proposed by Leighton in 1933 have been in general use since that date as the Substage names for the Wisconsinan. However, recent work in the Missouri River basin region has produced strong evidence indicating that the first two of these substages were closely linked, as were the last two, but that these two pairs of glacial episodes were separated by a significant time interval during which continental glaciers, if they were in existence, had small influence on the region of north-central United States. This grouping of events suggests that the recognition of three Sub-Ages--an early glacial Sub-Age, an intermediate interglacial Sub-Age, and a late glacial Sub-Age--might be appropriate. Such a reorganization would involve the introduction of three new Sub-Age names and reduce the refinement of the Pleistocene time scale, neither of which is judged to be appropriate at this time. Therefore an additional (interglacial) Sub-Age, Bradyan, has been proposed for use in the Kansas region (A. B. Leonard, 1951; Frye and Leonard, 1951; Frye, 1951). The Bradyan Sub-Age is between the Tazewellian and Caryan glacial Sub-Ages and is named for the Brady soil, extensively developed in the top of Tazewellian sediments in western Nebraska (Schultz and Stout, 1945) and in Kansas, and overlain at many places by the Bignell loess of presumed Caryan-Mankatoan age. Thus the Sub-Ages of the Wisconsinan as used in Kansas are, from oldest to youngest, Iowan, Tazewellian, Bradyan, Caryan, and Mankatoan.
The Recent Age (Stage), predominantly comparable in character to an interglacial interval, is the last major subdivision of the Pleistocene and includes the time since the Mankatoan glaciers ceased to exert a recognizable influence in the central interior region of North America. The term Recent was first used in a geologic time sense by Lyell (1833, pp. 52-53) to designate the interval during which man has inhabited the earth. Since Pleistocene has come to be a virtual synonym for "glacial period" Recent has been generally used to designate the post-glacial interval. This should not be construed to mean that Recent time has been lacking in climatic changes or geologic events. It may be that a minor episode of continental glaciation (Cochrane) occurred during this Age and sediments in Nebraska have been correlated with this event (Schultz, Lueninghoener, and Frankforter, 1951). However, the Recent was predominantly a time free of important effects of continental glaciation and for usage in Kansas has not been subdivided into named Sub-Ages (Sub-stages).
The foregoing discussion of the subdivisions of Pleistocene time shows at once a cyclic repetition of events. Recognition of this cyclic arrangement is not new and has been emphasized by classification schemes used in Iowa and Illinois. In the continental interior this cyclic arrangement of sediments, even far beyond the glacial margins, is clearly due to the alternate advance and retreat of continental glaciers. An indirect effect of glaciation--sea level fluctuation--may have exerted a major influence on sedimentation and erosion in coastal areas but is judged to have had slight effect on the Kansas region.
A Pleistocene cycle in the glaciated region and in an extensive marginal belt consists of a glacial and an interglacial interval. It is characterized by (1) valley cutting in the early part of the glacial interval plus some local sedimentation caused by the disruption of former drainage by advancing glaciers; (2) deposition of till and outwash in the glacial region at the glacial maximum; (3) deposition of coarse-textured outwash beyond the glacial limit, and in shallow valleys cut across the till plains as the ice margins retreated; (4) deposition of progressively finer alluvial materials as the glaciers shrank and finally disappeared; and (5) development of mature soil profiles over much of the region that presented surface conditions of essential equilibrium. As the major source of loess was the flood plains of outwash-carrying streams, these deposits occur in positions 2, 3, and 4 of the cycle. It can be seen then that the greatest volume of sediments of a Pleistocene cycle is not evenly distributed through the time of either a glacial or interglacial age but is concentrated primarily in the time span extending from about the midpoint to the end of a glacial Age, and perhaps into the very early part of the succeeding interglacial Age. The soils that typify the interglacials formed during a time span extending from early in an interglacial Age to approximately the beginning of the next succeeding glacial Age (Frye, 1951).
In the Kansas region evidence is lacking concerning the deposition of sediments during the early part of a glacial Age beyond the limits of the Kansan ice sheet. Therefore beyond the ice margin the cyclical unit of sediments represents an age extending from mid-glacial to very early interglacial. However, the formation of normal soil profiles was probably terminated early during the following glacial Age by the sharp acceleration of stream erosion.
In Iowa and Illinois the Pleistocene is classed as a Period (System) and four Epochs (Series) are used to include a glacial-interglacial pair each. These are Grandian (Nebraskan and Aftonian), Ottumwan (Kansan and Yarmouthian), Centralian (Illinoian and Sangamonian), and Eldoran (Wisconsinan and Recent). Of these units each of the first three essentially coincides with a glacial cycle; however, present data indicate that the youngest (Eldoran) includes two distinct cycles, each of which is complex within itself. These terms have not been adopted for official use in Kansas partly because of this inconsistency and partly because the retention of Quaternary as the System-Period with Pleistocene as its contained Series-Epoch would force the erection of a new category of names to include these terms and thus produce a further complication of the classification system.
The practical field geologist, the person concerned with securing water supplies, ceramic clays, volcanic ash, or sand and gravel deposits, or the soil scientist investigating parent materials of soils, is primarily concerned with the classification of rocks and only secondarily with the classification of time. Although on the map and cross sections in Plates 1 and 2 the Pleistocene deposits of Kansas are shown by Stages, for detailed geologic mapping and the several fields of applied geology the investigator deals with physical entities or rock units, commonly classed as formations, members, and beds or lentils. In principle the recognition of units of rocks is based on physical characters observable in the field and the assignment of a specific exposure to such a rock unit does not necessarily imply that it is contemporaneous with other exposures so classed. It has been recommended (Ashley and others, 1933) that a formation be a unit of rocks mappable by ordinary field methods on maps of conventional scale. Members may also be mappable units but beds rarely are.
It should not be construed, however, that age is of no value in field mapping. As the sediments of the Pleistocene consist of repeated cycles of sediments that have much in common, their physical appearance may have striking similarities. Therefore placement within a Stage by paleontological or other means may be necessary before final specific formational or member assignment is possible.
As has been described, the most striking arrangement of Pleistocene deposits in Kansas is in cyclic bundles and these bundles are in general comparable to Stages. The primary physical basis for classifying these sediments is therefore the buried soils that cap each cyclic bundle or the period of erosion which was the opening episode of the next succeeding cycle and which in many places destroyed the soil profile that had developed during the preceding interglacial Age.
Rock units of formational rank ideally have lithologic unity and contrast with contiguous formations. Thus the Kansas till is predominantly ice-deposited till, but it contains zones or beds of water-laid sand and gravel which were deposited as outwash during minor fluctuations of the ice margin, or by englacial streams, or by incorporation into the till from the deposits overriden by the glacier. It has stratigraphic continuity over a wide region and contrasts strongly with the underlying Atchison formation (silt, sand, and gravel) and the overlying Meade formation. At such places where it rests directly on an older unit of similar lithology (Nebraska till) it is separated from it by a buried soil, or an erosion surface. While a till sheet, because of its relatively sharp and easily recognizable top and bottom, is readily mappable, comparably sharp limits are not present in some sequences of alluvial deposits. For example, the Nebraska Geological Survey classes both the Sappa and Grand Island as formations. It is true that the Sappa commonly overlies the Grand Island but it is also true that the two units are gradational nearly everywhere they occur and any boundary drawn between them must be purely arbitrary. A boundary line drawn arbitrarily within a gradational series from gravels to silts can hardly be used as an objective field basis for mapping as several observers cannot be expected to draw such a line at the same place. For that reason the two units are classed as members in Kansas terminology and the Meade formation is used to include the entire gradational sequence which does have clearly recognizable limits at top and bottom. Such a practice retains lithologic similarity for the formational unit from place to place while establishing recognizable limits.
Stratigraphic continuity cannot be construed to have the same meaning in alluvial deposits as it does in marine strata. Glacial till and loess each transgress divides and have regional continuity. Fluviatile deposits, on the other hand, are restricted to the valleys that contain them and do not normally possess such regional continuity. Deposits within several drainage basins, if they possess the same lithologic character, such as a gradational sequence of gravel, sand, and silt, are nevertheless placed in the same formation when it is possible to demonstrate that they resulted from the same regional cycle of alluviation. In many places physical continuity of alluvial deposits can be demonstrated for hundreds of miles along an individual valley system. This practice seems required for Pleistocene deposits in order to avoid a maze of comparable names and because the complexities of drainage adjustments would otherwise extend names of older units across present divide areas while requiring several names for comparable age sediments in some existing valleys (note section on drainage history).
Names are not commonly assigned to beds or lentils (subdivision of a member) in Kansas Pleistocene deposits. A striking exception to this general practice is the Pearlette volcanic ash bed which occurs within the Sappa member of the Meade formation. In this case the unique character of this bed requires a formal name for proper reference.
The assignment of a name to a buried soil is a problem rarely encountered in sediments older than Pleistocene and does not fall clearly within the accepted practice of stratigraphic nomenclature. These soils are developed in the top of a stratigraphic unit and so are included entirely within sediments that are assigned a proper name. For example the soil in the top of the Kansas till is entirely within the Kansas till formation even though it may be called the Yarmouth soil. Therefore named buried soils do not become formations or members but their upper surface may be taken as the boundary between units of any rank. It has been general practice by some workers to assign to a buried soil the name of the formation or member in which it is typically developed (Loveland soil, or Kansan gumbotil). This practice becomes awkward when such a buried soil is found developed on another formational unit, such as "Loveland soil" on the Crete member, or even on Greenhorn limestone of Cretaceous age.
It has become standard practice in Kansas to use, as a noun, the name of the Age during which a buried soil developed as the proper name for the buried soil, and to consider these soils not as stratigraphic units but as bearing the same relation to rock units as a combined faunal zone and unconformity. This is particularly appropriate as most interglacial Ages have been named for buried soils and much of our knowledge of these intervals of time is derived from the corresponding soil. In this usage the Sangamon buried soil is known by the same name where it is developed in Loveland loess, Crete sand and gravel, or any one of the several older units on which it has been observed.
Stratigraphic correlation is essentially the matching of rock or time-rock units from one place to another. As these two categories of units are different there are philosophically two kinds of correlation. However, in the Pleistocene the unit of regional correlation is generally the Stage and therefore several basically different types of evidence may with propriety be brought to bear on a correlation problem. These types are: (1) lines of evidence establishing physical continuity from place to place, such as the tracing of a till sheet, a loess sheet, or a terrace surface; (2) evidence of synchronous deposition, usable only by the recognition of separated deposits made by the same volcanic ash fall; (3) evidence from the matching of similar assemblages of fossil organisms; (4) the matching or tracing of weathering effects on a former surface; and (5) the matching of similar stratigraphic or physiographic sequences. Although all may constitute an integrated team each of these methods actually correlates different things. Tracing of a lithologic unit correlates physical continuity but these physical features may be of slightly different ages from place to place; provable simultaneous deposition establishes contemporaneity of date from place to place but says nothing about the continuity of the containing sediments; fossil faunas (if the assemblages are adequate) establish similarity of animal populations which may imply a correlation of time or environment or perhaps both; the use of buried soils, if performed by the matching of morphological types, establishes a similarity between environments, but if performed by tracing (physical continuity) falls in the same category as the first group; and the fifth method matches similarity in sequence of events, but unless the matched sequence is sufficiently long and intricate these matched events may be similar but are not necessarily contemparaneous.
Lithologic continuity--Continuous tracing of a bed, or soil, is the most direct means of correlating stratigraphic units from place to place. The continuous tracing of units, however, is rarely possible in the field. In some areas in north-central Kansas individual loess sheets and their associated soils can be traced continuously for tens of miles. What workers usually mean by the "tracing" of beds or soils is the observation of such units at closely spaced intervals, either in surface exposures or by penetration in test holes. This method is used extensively in Kansas and is particularly applicable to correlation of loess sheets and tills in the hundreds of test holes that have been drilled by the cooperating Geological Surveys.
Petrographically distinctive volcanic ash--A method of establishing synchroneity of date from place to place has resulted from the discovery that fresh volcanic ash deposits of the rhyolitic type can be distinguished petrographically from other ash falls of different ages (Swineford and Frye, 1946). Bentonite, or altered volcanic ash has been used in many parts of the rock column as means of correlation but petrographically distinct fresh volcanic ash deposits have been used for interregional correlation only in the Pleistocene deposits west of Mississippi River and east of the Rocky Mountain Front Range. Only one such bed of ash (the Pearlette bed, Sappa member, Meade formation) occurrs extensively in the Kansas Pleistocene and therefore this method is effectively limited to the establishment of one time line. This bed has been studied from more than 100 localities in the State. The validity of the method has been summarized as follows (Frye, Swineford, and Leonard, 1948, p. 514):
Such properties as the refractive index, specific gravity, and shape of shards depend not only upon the chemical composition of the magma but also upon such highly variable factors as the temperature of quenching, pressure, and gas content. Glass from acid magmas in particular may show much variation because of a wide range in temperature and other conditions at eruption (E. F. Osborn, oral communication). Therefore, it is unlikely that the glass shards from several different ash falls will have the same characteristics.
The particular petrographic features that are usable in distinguishing the Pearlette volcanic ash bed from other volcanic ash deposits of the region have been summarized as follows (Frye, Swineford, and Leonard, 1948, p. 513):
- The color includes certain light shades of orange-yellow, which are not characteristic of fresh ash described from the Pliocene.
- The refractive index is consistently 1.499-1.501.
- The shape of the shards is characteristically sharply curved, with thickened glass at the bubble junctures, which are commonly curved and branching at wide angles. Fibrous shards are present in all samples.
- Many shards have groups or clusters of elongate vesicles which are seldom found in ash of Pliocene age.
- The percentage of iron oxide is less than 2 . . . whereas in . . . Pliocene ash the Fe2O3 content ranges from 1.66 to 3.09 . . . .
- The specific gravity ranges from 2.21 to 2.32, whereas in the Pliocene ash it ranges from 2.33 to 2.37.
Molluscan faunal assemblages--Assemblages of fossils are useful in determining the relative age and stratigraphic sequence of deposits only under rather rigidly controlled conditions. First of all, the vertical sequence and range of distinctive assemblages must be ascertained, and as far as practicable, the areal extent of each assemblage must be learned. Finally, of course, a correlation of established faunal zones with the standard stratigraphic sequence must be made, without which any sequence of faunal assemblages lacks meaning to the geologist.
In Kansas, a distinctive molluscan faunal assemblage is associated with each of the major cycles of Pleistocene deposition, and in the case of the complex Wisconsinan cycles, a stratigraphic sequence of faunal zones occurs. Fortunately, in the midcontinent region, the association of a distinctive molluscan assemblage with the petrographically unique Pearlette volcanic ash provided a key to the correlation of the succession of molluscan faunal assemblages with the accepted stratigraphic sequence. Once such a correlation has been made, molluscan faunal assemblages become a useful tool with which to identify sediments in situations where the stratigraphic sequence is not clear. Occasionally, assemblages of fossils may be useful in the correlation of units of deposits which are not subdivisible by lithologic or other standard criteria of the stratigrapher. The correlation of the massive Peoria silt member of the Sanbom formation with the Farmdale, Iowa, and Tazewell loesses of the glaciated region is an example of this use of fossils.
Morphology and continuity of buried soils--Buried soil profiles have been used in two distinctly different ways as tools of stratigraphic correlation. These two techniques are (1) matching or contrasting degree or type of profile morphology and color and (2) the tracing of a soil by many closely spaced observations in much the same way as a rock unit is traced. The end product of these two methods is quite different and for either method to have validity a worker must understand the genetic implications of the many soil types. Of the three orders of soil recognized in soil science (Baldwin, Kellog, and Thorp, 1938) only zonal (or normal) and intrazonal soils are of stratigraphic value as the azonal soils do not exhibit distinct vertical zonation and therefore cannot be interpreted in a sequence of clastic sediments.
The matching of physical characteristics of soils over wide regions is extremely precarious unless it is known that all factors affecting the development of the soil were the same. Most prominent among these factors are parent material, climate, drainage conditions, floral cover, and animal population. Of these factors only the nature of the parent material can be determined with certainty by the geologist by examination of a sequence of sediments containing a buried profile. Therefore, this technique should be used, in stratigraphic correlation, with great caution and over short distances. Even more precarious is the use of profile morphology of surface soils to date the immediately underlying sediments. Although of value in establishing relative ages of adjacent surfaces with similar topographies, this method has not been used regionally in Kansas. The present surface soils that extend east-west across Kansas, if examined at localities where they are developed in materials of similar type and age, illustrate this point. They range from Brown soils and Chestnut soils in the northwest through Chernozems and dark Planosols to Prairie soils and Prairie-Forest soils in the northeast.
The tracing of a buried soil is a valuable tool to stratigraphic correlation and this method is not beset with the principal dangers involved in the matching of profile morphology. The Sangamon soil, in spite of its progressive change in morphology, has been traced virtually the entire 400 miles east-west across Kansas.
Stratigraphic succession--The measurement of a sequence of rocks, noting physical appearances and distinctive characters, constitutes one of the earliest techniques of stratigraphic correlation. It is essentially the tracing of rock units by moderately to widely spaced observations, and as a group of associated contrasting rock units is being traced simultaneously stratigraphic placement can be maintained by the particular sequence. In other words, this method relies on the consistence of a sequence of past events that resulted in distinctive rock layers. When dealing with widespread tabular bodies of rock (loess sheets, till sheets) the method has a high degree of reliability. However, where it is applied to alluvial sediments it should be used with caution as deposits of restricted geographic distribution may be easily confused with other cyclic units which display similar physical appearance.
Physiographic expression--The physiographic character of tills has been used for many years as a means of correlation, particularly of the younger tills. In the Kansas region this technique is not applicable in the glaciated region as Kansas till everywhere overlaps Nebraska till and younger tills do not occur in the State.
Topographic form is quite usable as an aid to correlation, however, along the valleys flanked with alluvial terraces, particularly in the northern part of the State. The tracing of an alluvial terrace is much the same as the tracing of a rock unit, and in effect the deposits under the surface of an alluvial terrace constitute a lithologic unit. Many special problems arise in using terraces for correlation purposes. Perhaps the most persistent problem is the determination of the erosional history of the particular valley along which the terraces are being traced. The presence of a sequence of terraces is evidence that erosion has been predominant over deposition in the valley system. In Kansas valleys, erosional histories have not been consistent and in some valleys a physiographic or inverted sequence displays the several Pleistocene Stages in descending steps. In other valleys most of Pleistocene time was typified by predominant alluviation and only Wisconsinan terraces occur. In still others, Nebraskan terraces occur at high levels, Kansan sediments are overlain by Illinoian sediments under an intermediate terrace surface, and a complex of low terraces and flood plain are of Wisconsinan and Recent age. Obviously a terrace sequence cannot be projected with safety from one valley to another.
Other problems of terrace correlation are inherent in the technique. Along some valleys terrace levels converge or diverge and in some areas in Kansas former stream gradients cross. Influence on former stream gradients by resistant units in the bedrock may cause alternate divergence and convergence of the same pair of terraces when traced along a valley. Even the recognition of a terrace surface may be rendered difficult by erosion of the outer margin, and local deposition on the inner part of the terrace surface. In general, physiographic expression is a valuable tool in Pleistocene correlation when used in conjunction with other techniques in relatively local areas.
As preface to a discussion of the stratigraphy of the Pleistocene deposits in Kansas it is desirable to review briefly the stratigraphy of these deposits in the midcontinent region of which this State is a part. It is especially desirable to do so as the Pleistocene deposits of Kansas occur in three distinct provinces which are in some cases more closely related to the stratigraphy of adjacent states than to that of the other Kansas regions. Glacial deposits are extensive in northeastern Kansas and are similar to deposits in Iowa and eastern Nebraska. In east-central and southeastern Kansas locally derived stream-deposited gravels are the predominant Pleistocene sediment and show little direct effect of glaciation. In central and western Kansas stream-laid and eolian sediments are closely related to similar deposits in the states to the north and south and indirectly to glacial events in the area to the northeast and to the west.
Northeastward from Kansas, in the upper Mississippi valley region of Iowa, Illinois, Minnesota, and Wisconsin, there are extensive deposits made by continental glaciers. It is within this area that the generally accepted glacial sequence was worked, out. The history of this work has been summarized by Kay and Apfel (1929, pp. 71-73) as follows:
In the year 1837, Louis Agassiz, then living in Switzerland, put forward the theory that there had been continental glaciation in Europe. This somewhat startling interpretation stimulated investigation of mantle rocks both in Europe and in America. Soon in both countries abundant evidence had been found to place Agassiz's views of continental glaciation upon a firm basis. From that time glacial phenomena have been studied by many geologists, and year by year as investigations have continued, more and more of the complex phases of the history of the deposits which were made by glaciers during the Pleistocene or Glacial Period have been unraveled. Nor has finer work been done anywhere than by students of the glacial deposits of the Mississippi Valley. At first it was believed that all the phenomena could be explained in relation to the advance and retreat of a single ice sheet. But soon evidence was found which indicated to some geologists that there had been two ice sheets separated by a long interglacial epoch. This evidence consisted in many places of a forest bed or buried soil separating two tills. For instance, buried soils between tills were described in Illinois as early as 1868 by Worthen. In a report of the Ohio Geological Survey for 1869 but published in 1871 Orton called attention to a buried peat near Germantown, Ohio. Moreover, in the report of the Geological Survey of Ohio, Volume I, written in 1872, Orton stated that the interglacial stage was coming to be clearly recognized both in Europe and in America. Another interesting early reference to the significance of vegetable material in relation to till was made by N. H. Winchell in 1873. He stated that he found leaves and wood in clay in the midst of tills in southeastern Minnesota, and expressed the view that the clay "may consist of the remains of a previous glacial sheet, upon which rested vegetable growths of the surface, accumulating between the periods of glacial epochs." In the 3rd and 4th Annual Reports of the Geological and Natural History Survey of Minnesota, published in 1875 and 1876, respectively, Winchell described occurrences of two tills separated by peat, and in Geology of Minnesota, Volume I, 1884, he stated that in southeast Minnesota the peat separates an "old" drift or upper drift from an "older" drift or lower drift. At this time he recognized also in other parts of Minnesota a "younger" drift which is younger than his "old" drift.
Between the years 1875 and 1886 Chamberlin published several important papers in which differences in topographic form and degree of alteration were emphasized as bases for differentiating tills. In fact he put greater emphasis upon the significance of these features than upon the forest beds. He recognized two tills in Wisconsin separated by a long interglacial interval. He called the older till the First Glacial and the younger the Second Glacial. In 1886 he stated that the subdivisions within each of these glacial epochs remained to be worked out but that some evidence of the older drift area pointed to a dual division of the first epoch.
From these splendid beginnings, investigations have continued year after year to the present time. As a result of these studies many chapters of the whole story of the Pleistocene of the Mississippi Valley have now been clearly outlined.
Concerning the early work in Iowa, Kay and Apfel, (1929) state that McGee made the first important contributions during the period 1878-1891 by recognizing two distinct tills separated by a forest bed. During the last decade of the last century work by McGee, Chamberlin, Bain, and Calvin led to the recognition of essentially all the major glacial units now known in Iowa. In 1897 Calvin presented a classification of Iowa Pleistocene deposits that is strikingly similar, in spite of altered names, to the accepted classification over 50 years later. A summary of Calvin's (1897) classification is as follows:
10. The recent stage, since the retreat of the Wisconsin ice.
9. Fifth glacial stage, Wisconsin.
8. Fourth interglacial stage, Toronto (?).
7. Fourth glacial stage, Iowan.
6. Third interglacial stage (unnamed).
5. Third stage of glaciation, Illinois.
4. Second interglacial stage, Buchanan.
3. Second glacial stage, Kansan.
2. First interglacial stage, Aftonian.
1. First stage of glaciation, Albertan.
Although it is the work in central and eastern Iowa, Illinois, and Wisconsin that served to define the standard glacial sequence on which the subdivision of Pleistocene time is based, it is with deposits of the Missouri Valley region in western Iowa and eastern Nebraska that the glacial deposits in Kansas have been correlated. The existence of glacial deposits of Nebraskan and Kansan age have for more than 50 years been known to occur along the Missouri River valley of Iowa and Nebraska, and recently it has been demonstrated (Smith and Riecken, 1947) that the Iowan glacier crossed this valley into northeastern Nebraska. Younger Wisconsin tills occur farther north in South Dakota and to the east in north-central Iowa. Deposits of all the major episodes of continental glaciation except the Illinoian are known to occur along Missouri River Valley northward from Kansas.
The late Pleistocene loesses are also well developed in this area. The Loveland loess (Illinoian) was described by Shimek (1909) from exposures in the Missouri Valley bluffs at Loveland, Iowa. Peoria loess has been studied extensively along this segment of valley, and the Bignell loess has been described from exposures at Sioux City, Iowa, and several localities in eastern Nebraska.
The exact placement of deposits of sand, gravel, silt, and volcanic ash that occur at several places along the valley is important to regional correlations. This is particularly so as the volcanic ash has been determined petrographically to be the Pearlette bed and is associated with an important molluscan fauna (Frye, Swineford, and Leonard, 1948; A. B. Leonard, 1950) which establish these deposits as age equivalents of the Grand Island and Sappa members of the Meade formation of Kansas terminology. The evidence bearing on the placement of these beds in the glacial section has been discussed by Frye, Swineford, and Leonard (1948) who concluded that they occur above Kansas till and below Loveland loess. The principal formational units described from the midcontinent region and their correlation with the Kansas section are shown in Figure 2.
The east-central region of Kansas Pleistocene is genetically and physiographically related to southern Missouri and northeastern Oklahoma. No detailed studies of Pleistocene deposits have been made in either of these adjacent areas and although the term "Lafayette gravels" has been applied to some deposits by a few workers, a formal stratigraphic nomenclature is not in existence. As the deposits of Pleistocene age in these areas are quite thin and discontinuous the problems are primarily physiographic rather than stratigraphic. Furthermore, the available evidence in Kansas indicates that some of the gravel veneers on relatively high levels are late Tertiary in age.
The Pleistocene geology of western Kansas is closely related to the stratigraphy of adjacent areas in the High Plains--western Oklahoma, northwestern Texas, and western Nebraska. The High Plains of northwestern Texas are far removed from continental glaciation or the direct effects of these glaciers. However, episodes of erosion and deposition are judged to have taken place in this region essentially synchronously with erosion and deposition farther north in the High Plains. The Pleistocene geology of this area has been reviewed by Evans and Meade (1945). They describe the Blanco formation (and equivalent Rita Blanca deposits) and conclude that this formation is Nebraskan in age is unconformable on Pliocene sediments, and that it accumulated in basins. The Tule formation (and probably equivalent Spring Creek deposits) unconformably overlies the Blanco formation. Although Evans and Meade do not give a definite age assignment to the Tule formation, work on volcanic ash petrography and fossil molluscan faunas by Frye, Swineford, and Leonard (1948) shows clearly that this formation is equivalent to the Meade formation (Sappa and Grand Island) of Kansas classification, and is late Kansan and early Yarmouthian in age. Deposits of Illinoian age have so far not been recognized in northwestern Texas, but Peoria loess with a distinctive pre-Bradyan Wisconsinan snail fauna has been described in the northern part of the panhandle area (Frye and A. B. Leonard, 1951). Evans and Meade (1945, p. 495) proposed the name Tahoka clay for young basin deposits that occur at several localities in this part of Texas and adjacent New Mexico. They conclude on the basis of stratigraphic relations, fossil vertebrates, and artifacts, that the Tahoka formation is Wisconsinan in age.
In the Nebraska region work in the western part of the State has in general been coordinated with studies of the Pleistocene deposits in the eastern (glaciated) area. Attention was drawn to the early Pleistocene deposits of western Nebraska by the naming and description of the Broadwater formation (Schultz and Stout, 1945, p. 232) and its fossil vertebrate fauna. This formation, recognized only in western Nebraska, occurs unconformably on Ogallala formation in a high terrace position along the North Platte Valley, has been correlated in part with the Holdrege and Fullerton formations which occur farther east (Condra and Reed, 1950), and is considered to be Nebraskan and Aftonian in age.
Several general summaries of the Pleistocene stratigraphy of Nebraska have been published (Lugn, 1935; Condra, Reed, and Gordon, 1947; Condra and Reed, 1950; Schultz, Lueninghoener, and Frankforter, 1951) and the formations now recognized by the Nebraska Geological Survey are listed on Figure 2. For the most part the formational names in current use in Nebraska appear also in the Kansas nomenclature as members or formations. As they will be described in some detail in the following section of this report, they will not be discussed here. Two important Pleistocene regions in Nebraska should be mentioned, however. These are the vast sand hills tract which lies north of the Platte River Valley in north-central Nebraska and the extensive region of thick loess in southwestern Nebraska. These two areas and the southward extension of the loess cover into Kansas present an expanse of thick and virtually continuous Wisconsinan deposits that rivals even the Wisconsinan till plains province of Ohio, Indiana, Illinois, Wisconsin, and Minnesota. In this northern High Plains area these late Pleistocene sediments approach 200 feet in thickness and dominate the surface of a region extending more than 300 miles both north-south and east-west.
It is apparent that Kansas lies across the boundaries separating three distinctly different Pleistocene provinces. This geographic circumstance is at once the frustration and the challenge to the Pleistocene geologist working in the State. It renders inoperative some commonly used techniques of Pleistocene correlation but those correlations that have been possible serve to link dissimilar chronologies and nomenclatures.
Kansas Geological Survey, Pleistocene Geology
Comments to firstname.lastname@example.org
Web version August 2005. Original publication date Nov. 1952.