|Original published in D.F. Merriam, ed., 1964, Symposium on cyclic sedimentation: Kansas Geological Survey, Bulletin 169, pp. 31-42|
McMaster University, Hamilton, Ontario
Allocyclic sequences are those generated outside the sedimentary system by changes in discharge, load, and slope. They differ from autocyclic alternations in their lateral extension across the alluvial plain and even into other depositional basins; they may differ from each other in the proportions of biological and chemical sediments and in development of erosion surfaces and weathered zones.
Unequal subsidence, differential compaction, depositional topography, and substrate modify the effects of the various cyclic mechanisms and may obscure them altogether, exaggerate some at the expense of others, or produce megacyclothems or cyclothem bundles. The cyclic mechanisms may also dampen by interference or exaggerate by interaction the sedimentary alternations. With allowance for such variations and exceptions, cyclothemic sequences should be typical of alluvial deposition, and refinement of interpretive criteria should permit the recognition of the various types significant to correlation and environmental interpretations.
Clearly, theories of cyclothem development that envisage a single mechanism of origin operating during a limited time span are inadequate. Theories of local origin (e. g. Robertson, 1948; Van de Heide, 1950; Moore, 1959; Goodlet, 1960; Read, 1961; Allen, 1962) cannot adequately explain the classical Illinois cyclothem. Theories of worldwide eustatic fluctuation (Wanless, 1963; Wells, 1960) are inapplicable to enclosed lacustrine basins, do not provide for cyclic deposits of purely local extent, and, since they are related to temporally circumscribed events, i.e. glaciation, do not explain the total temporal distribution of cyclothemic deposits. Climatic theories (e.g. Brough, 1928, Beerbower, 1961) must be strained toward absurdity to provide for localized and highly variable cyclothems. Diastrophic theories (e. g. Weller, 1956) require the operation of a cyclic diastrophic mechanism on a wide (at a minimum, regional) scale with remarkable regularity; such diastrophic cycles have not been demonstrated independently of the cyclothems and face numerous a priori objections.
In these circumstances, I believe it desirable to review the potential mechanisms of cyclothemic deposition as deduced from models of sedimentation in terrestrial and transitional environments and to suggest certain criteria for the recognition of each of these mechanisms and for testing various theories of cyclothem origin. This paper deals with cyclic deposition involving only alluvial environments. I hope to prepare a similar analysis of delta and of bar-lagoon environments at a later date.
Rhythms, Cycles, and Cyclothems--Duff and Walton have recently pointed out (1962, p. 235-240) ambiguities in the definition of cyclothems, in the use of the terms "rhythm", "cycle", and "cyclothem", an in various abstract cyclothem concepts such as "typical", "normal", "complete", "standard", "characteristic", "full", "idealized", "theoretical", and "composite." They suggest that "cycle", "rhythm", and "cyclothem" be considered as synonomous except as it is necessary to restrict "cycle" to time and "cyclothem" to a rock sequence. They define (p. 239) a cycle as ". . . a group of rock units which tend to occur in a certain order and which contains one unit which is repeated frequently through the succession." They further recognize (p. 239) a "modal cycle" as that sequence which occurs most frequently and a "composite sequence" as that which comprises all lithologic types arranged in the order in which they tend to occur within cycles. Finally, they suggest that "ideal cycle" be reserved for a sequence erected on the basis of a theoretic model of sedimentation.
Although their suggestions emphasize the need for clarification of terminology, not all the modifications seem desirable in their present form. They prefer not to distinguish in their cyclic concept between the asymmetric pattern of elements [a, b, c, a, b, c] and the symmetric one [a, b, c, b, a], although the original concept of cyclothem as defined by Weller (1930) emphasizes the asymmetric pattern of their sequence. If one follows Weller, as I believe one must in such a case, cyclothems represent only one of the two possible subclasses of cyclic deposition, the asymmetric one. Further restriction or codification of the term may be desirable, but the reduction of "cycle" to synonomy with "cyclothem", contrary to original definition and subsequent usage, is unlikely to make the problem of interpretation any simpler. Their other terms appear useful and could be employed in modified form: modal cyclothem, composite cyclothem, ideal cyclothem. Since, in this paper, I will deal primarily with theoretical models of sedimentation, I find it helpful to distinguish the most probable sequence in a particular model as an "ideal modal cyclothem."
Finally, in discussing mechanisms of cyclothemic deposition I shall distinguish between those that require no change in the total energy and material input into a sedimentary system but involve simply the redistribution of these elements within the system, and those that result from changes in the supply of energy or material. The first, hereafter called autocyclic, are generated in the depositional prism and include such items as channel migration, channel diversion, and bar migration. The second or allocyclic type result from changes external to the sedimentary unit such as uplift, subsidence, climatic variation or eustatic change.
Acknowledgments--The ideas reported here are an outgrowth of work on cyclothems of the Dunkard Group reported earlier (Beerbower, 1961), supported by NSF Grant G 2156 and by grants from the Society of Sigma Xi, American Academy of Arts and Sciences, and Geological Society of America. Since publication of the Dunkard paper in 1961, I have benefited from the comments, criticisms, and suggestions of a large number of individuals whom I cannot begin to name individually; I also profited recently by the opportunity to discuss my ideas with seminar groups at Lehigh University and Illinois Geological Survey and by the chance to review with Dr. Harold Wanless his current work on Pennsylvanian regional stratigraphy. Finally, I should like to acknowledge the assistance of those who have furnished reprints of pertinent studies.
Since few alluvial plains have been studied in careful detail, I have relied largely on two sources, Fisk (1944) on the Mississippi alluvial valley and Sykes (1937) on the Colorado delta. Recent works by Leopold, Wolman, Miller, and Schumm on river dynamics and morphology have also been helpful.
Within any particular alluvial segment, width, depth, velocity, and channel pattern are entirely dependent on the other factors. Discharge reflects the input of water into the segment as well as losses of water within it; consequently, it is partly independent. Similarily, slope results from eustatic and diastrophic changes external to the segment as well as aggradation, degradation, and compaction within it; load results from sediment input as well as losses or gains within the segment; roughness from the substrate through which the channels operate as well as from the deposits on their floor. (Note: Leopold and Wolman (1957, p. 73) consider slope and roughness as dependent factors, but their conclusions clearly apply only to stream segments in quasi-equilibrium whose internal morphology ia completely determined by discharge and load.)
Geographic extent is partly a consequence of geomorphic and diastrophic events external to the erosive and depositional history of the stream; the distribution of the geomorphic elements over the segment is partly determined by events in the history of the plain prior to the current configuration of processes.
From the sedimentological viewpoint, the simplest model of an alluvial plain might well consist of an energy distribution map displaying three components, current velocity, organic activity, and nonbiological chemical reactions. If input, i.e. discharge, load, slope, organisms, and solar energy, are held constant, if the extent of plain does not change, and if the system is allowed to reach quasi-equilibrium, this energy map will change only if a different substrate is encountered or if the geomorphic elements are repositioned on the plain. If such changes are for any reason cyclic, they will be, by the definition above, autocyclic. Any natural stream will, in fact, show short period cycles in discharge and load, and therefore the energy input will vary from instant to instant. The morphology of the alluvial plain, however, adjusts primarily to two stages of discharge, the channel to bank-full discharge and the floodplain to the most frequent flood discharge (Wolman and Miller, 1960, p. 65). The relation to load fluctuations is less well known, but, since load adjusts rapidly by scour or deposition, and to discharge changes which typically are coupled with them, it seems reasonable to disregard the effect of short term load cycles on the characteristics of the alluvial plain.
With these limiting assumptions: (1) no changes in input or extent of plain, (2) system in equilibrium (no permanent gain or loss of energy within the system), (3) significant fluctuations in discharge comprising only two stages, bank-full and frequent flood, and (4) the load fluctuations disregarded, the energy distribution system of an alluvial plain can be expressed in a series of distinct environmental elements as shown in Table 1. Since energy distribution controls sedimentation, translation of these elements into sedimentary parameters yields the last column of the table.
Table 1--Characteristics of alluvial plain environments (based principally on Fisk, 1944, and Sykes, 1937).
|Effective energy states||Predominate sedimentary
|high||insignificant||insignificant||coarse clastics, gravel to sand, and/or erosion|
b. Point bar
|insignificant||insignificant||coarse clastics, gravel to sand, and/or erosion|
c. Channel bar,
|high to moderate||insignificant||insignificant||coarse clastics, gravel to sand, and/or erosion|
|slight||slight||coarse clastics, gravel to sand|
|3. Levee||moderate||moderate||slight||moderate clastics, fine sand to silt; root and burrow structures|
|moderate||moderate||slight||moderate clastics, fine sand to silt; root and burrow structures|
|moderate to low||moderate||moderate||moderate clastics, fine sand to clayey silt; root and burrow structures;
some organic material; minor chemical accumulations
|fine clastics, silt and clay; plus organic debris and minor
|7. Flood plain
|moderate||moderate||very fine clastics, silt and clay; plus minor chemical
accumulations and root, burrow, and dessication structures
|7. Flood plain
|slight to insignificant||intense||moderate||very fine clastics, clay; plus minor chemical accumulations
and large amounts of organic material
|7. Flood plain
|strong||very fine clastics, clay; plus major chemical accumulations
and some organic material
|7. Flood plain
d. Collection channel
|slight||insignificant||nondeposition or even erosion if velocity is high or substrate
It is commonplace to note that energy and the consequent sedimentary and topographic features are very unevenly distributed over an alluvial plain at any particular time. It is equally obvious that the energy system has local gradients with definite vectoral properties: erosion on the outside of meanders and deposition on the inside; crevassing of the natural levees and alluvial cone formation at outer ends of crevasses; trapping of sediment behind floodplain barriers and scour by streams that collect the sediment-free water below the barrier. In consequence of these vectoral properties, the various energy components shift progressively. This migration tends to equalize the cumulative energy distribution across the plain, and energy levels tend to succeed each other in a definite order.
If the floor of the plain is not subsiding nor its upper surface lowered by compaction or elevated by aggradation, the energy distribution in a cross section of plain will likewise approach uniformity. The sediments representing this uniform energy field should approach uniformity and consist principally of deposits that accumulated at the higher energy levels. The only heterogeneity in energy or sediment will be on the "skin" of the alluvial plain which reflects the current and recent distributions of energy values. In such a plain, most of the alluvium will consist of point-bar or channel-bar deposits. Wolman and Leopold (1957) describe examples of plains of this type.
This class of equilibrium floodplains contrasts with those whose floors are subsiding but on which aggradation exactly balances subsidence and compaction. The lateral shifts of energy values are probably more rapid and frequent because subsidence and aggradation tend to exaggerate the local energy gradients, e.g., between the channels where aggradation equals or exceeds subsidence and the floodplains where subsidence exceeds aggradation. On the other hand, part of the alluvial cross section is continuously lowered below base level of the channel and removed from reach of the alluvial energy system. Although the cumulative energy distribution across the surface still tends to uniformity, it will not be uniform in cross section, and the sedimentary section may preserve some degree of heterogeneity. If the internal energy gradients are high and, therefore, the rate of energy redistribution high compared to the rate of subsidence, the cumulative cross section energy distribution will again approach uniformity. The sediments retained in the alluvial section will consist principally of those deposited at high energy levels--except where such levels are insufficient to erode cohesive, fine-grained sediment.
This model suggests that the sedimentary records of high velocity streams with relatively heavy loads, including a considerable portion of large caliber material (and, therefore, noncohesive), will consist principally of channel-bar gravel and sand, unless subsidence is very rapid. The only evidence retained of overbank deposits will be mud chips and, in spite of periodic changes in energy distribution across the plain, no significant sedimentary alternations will be preserved. The basal sand strata of the Mississippi alluvium (Fisk, 1944, p. 49) apparently represent such a situation. On the other hand, streams with a lesser amount of coarse, noncohesive material, such as the lower Mississippi (Fisk, 1944, p. 20), will shift less rapidly, and the alluvial cross-section will include a greater proportion of overbank deposits and will preserve an alternation of coarse and fine layers.
Will these alternations be cyclic or, in particular, cyclothemic? If one assumes, for simplicity, three environmental elements: (a), high velocity and insignificant biological or thermal activity; (b), intermediate velocity, moderate biological activity, and insignificant chemical energy; and (c), low velocity, moderate biological activity, and moderate chemical energy; three different lithotopes would develop, a, well-bedded sand, b, silt with disrupted bedding, and c, clay without bedding and with limy nodules. If one further assumes that energy changes either occur suddenly without a transition in sediment character (symbolized by "/") or gradually with a sedimentary transition (symbol, "-"), a very large variety of sequences is possible--actually 48 for a five element sequence. If the redistribution is purely random, the probability of any two successive, five-element sequences being identical is relatively small, and no cyclicity would be apparent. In fact, the redistribution sequence (Table 2) is, apparently, most frequently in the pattern [a-b-c/a-b-c] for channel migration (Fisk, 1944, cross sections,) and [a-b-c/b/a-b-c/b] for channel diversion by crevassing (Sykes, 1937, p. 171-173). Since the high energy channel base level is below the depositional level of the other environmental elements, channel scour will tend to reduce these sequences locally to [a-b/a-b] and [a-b-c/a-b-c] or even to [a/a] and [a-b/a-b]. If a channel does not develop at a particular site, sequences will be limited to sediments band c and the pattern to [b-c/b-c] or possibly [b-c-b-c]. Departures from ideal conditions may complicate the section and eliminate much of the apparent cyclicity. Thus, Duff and Walton (1962, p. 243) report 320 different types of sequences in 1,200 measured alternations, although one type occurred nearly 180 times, a second about 160 times, a third about 90 times, and a fourth nearly 60 times. In the most common types of sequences the pattern is consistently asymmetric, and therefore cyclothemic, if the contacts (transitional or abrupt) between sedimentary units are considered.
Table 2--Modes of energy redistribution on alluvial plain.
|Sequence at site
initially on flood plain
|Energy system||Morphologic and
|A. Meander migration||A.1. Slow overbank flow.||Biological and chemical elements predominate.||Sheet deposit, structureless clay. May include organic and chemical deposits.|
|A.2. Rapid overbank flow.||Current strong; biological
|Sheet formed by overlapping, thin wedges of silt and sandy silt. Laminations may be disturbed by root action.|
|A.3. Channel cutting.||Current predominates.||Channel through part or all of A.2. and perhaps into A.1.|
|A.4. Point-bar formation.||Current predominates.||Lensatic, channel-fill deposit. Silt, sand, and possibly gravel.|
|A.5. Bar topstratum.||Current moderate; biological
|Sheet consisting of small lenses and local sheets of silt and sandy silt. May include organic deposits. Laminations may be disturbed.|
|A.6. Slow overbank flow.||As in A.1.||As in A.1.|
|B. Chute type cutoff||B.1. Slow overbank flow.||As in A.1.||As in A.1.|
|B.2. Channel cutting.||As in A.3.||As in A.3.|
|B.3. Rapid channel abandonment.||Currents slight; biological
and thermal energy high.
|Isolated, swampy oxbow lake or slough.|
|B.4. Slow overbank flow.||As in A.1.||Clay plug fill in channel succeeded by sheet clay.|
|C. Neck type cutoff||C.1. Slow overbank flow.||As in A.1.||As in A.1.|
|C.2. Rapid overbank flow.||As in A.2.||As in A.2.|
|C.3. Channel cutting.||As in A.3.||As in A.3.|
|D. Crevasse||D.1. Slow overbank flow.||As in A.1.||As in A.1.|
|D.2. Crevasse cutting and alluvial cone formed.||Currents strong to moderate.||Local channeling with terminal fan of lensatic and sheet sand and silt bodies.|
|D.3. Silting of crevasse.||Currents moderate to slight.||As in C.4.|
|D.4. Slow overbank deposition.||As in A.1.||As in A.1.|
|E. Diversion||E.1. Slow overbank flow.||As in A.1.||As in A.1.|
|E.2. Alluvial cone formed.||Currents moderate.||As in A.3.|
|E.3. Channel cutting.||Currents strong.||Thin fanlike sheet of lensatic and sheet sand and silt.|
|E.4. Point-bar deposition.||Currents strong to moderate.||As in A.4.|
|E.5. Gradual channel abandonment.||Currents moderate to slight.||As in C.4.|
|E.6. Slow overbank flow.||As in A.1.||As in A.1.|
From this consideration of the characteristics of alluvial plains, I suggest the following significant conclusions in the analysis of cyclic sedimentation on alluvial plains:
(1) Sedimentary cycles may result from the internal redistribution of depositional environments induced by internal energy gradients with vectoral properties.
(2) All alluvial plains undergo such internal cyclic sedimentation, but cycles are recorded only when the rate of energy redistribution is low in relation to summed rate of subsidence and compaction.
(3) The depositional cycles are most frequently cyclothemic, i.e. asymmetric.
The most significant mechanisms of energy redistribution on an alluvial plain appear to be (1) channel migration and cutoff, (2) crevassing, and (3) channel diversion. A detailed, albeit idealized, sequence for these mechanisms is shown in Table 2. The differences between types of cyclothems are less in lithologic sequence than in geographic extent and in three-dimensional shape. Channel migration and cutoff are mainly confined to a narrow strip, the meander belt, on the total width of the plain. Clay plug, floodplain, levee, and the finer bar sediments would be preserved as truncate wedges between channel fills consisting of coarse, point- or channel-bar deposits. These deposits should not differ substantially from anastomosing, fine and coarse, bar deposits in a braided stream. The belt sandstones (Potter, 1962a, p. 1891-1892) of the Pennsylvanian in Illinois with their intercalcations (or "islands") of silty or clayey sediments within anastomosing, coalescing, dendritic, sand bodies provide an example of the channel migration and cutoff.
Crevasse sequences, on the other hand, extend laterally from the meander or braided belt and consist of a sheet of distributary sand and silt with a lobate form extending from a "stem" of coarser sand in the crevasse. The lithologic sequence would most frequently consist of [b-c/b-c] but would probably include such variants as [a-b-c/a-b-c] and [b-c-b-c]. Moore (1960, p. 222-223) has described crevasse sequences in the Carboniferous of Great Britain that resemble those postulated, and some of the sheet sands associated with the channels of the Pennsylvanian in Illinois are probably of similar origin.
Channel diversion resembles crevassing, in fact is identical in early stages, but results finally in a through channel with its own meander or braided belt and with its own fringe of crevasse channels and sheet sands. The most frequent sequences would be [a-b-c/a-b-c], [b-c/b-c], and [b-c-b-c] so that discrimination from crevasse alternations would depend on extensive mapping of the rock bodies with delineation of belt sandstone representing the channel belt and of sheet sandstones representing crevasses.
Allen (1962, p. 692-694) argues that cyclic sequences in the Old Red Sandstone of Shropshire are consequences of channel diversion. The occurrence of several separate belt sandstones within a single unit of the Mississippian of Illinois (Potter 1962b, p. 16-19), suggests extensive channel diversion across the alluvial plain. Unfortunately, neither subsurface nor outcrop evidence demonstrates clearly that the latter channels postdate the sheets and thus represent the sequence suggested in Table 2: deposition of sand and silt in shallow channels and sheets on an alluvial cone (E.2) followed by the concentration of discharge into a single deep channel (E.3).
Probably no alluvial plain maintains a quasi-equilibrium state for long whether this equilibrium is a consequence of a stable floor, and thus a balance of deposition and erosion, or of an equality of aggradation with subsidence. Eustatic variation of sea level, climatic shifts, irregular elevation of the source area, and spasmodic depression of the basin are surely much more common than eustatic stability, unvarying climate, and uniform diastrophic movements. Any or all of these factors may operate to produce variation in slope, load, discharge, biological activity, and chemical deposition and, thus, will change the energy input into the system.
If the changes are sufficiently great, a canyon, a delta, or a submerged coastal plain may replace the alluvial plain and produce an entirely new pattern of energy distribution as well as a change in the total energy. To limit the scope of discussion, however, I restrict the analysis here to changes that retain the basic energy distribution characteristic of the alluvial plain. In this circumstance the only potential pattern change would be in channel morphology from braided to meandering or vice versa.
With a change in energy input, morphology of the alluvial plain is no longer in equilibrium with the energy system and erosion and sedimentation will proceed until it again reaches quasi-equilibrium. Changes in total energy input may change: (1) values in the energy distribution pattern; (2) gradients between the principal elements of that pattern; and (3) areal extent of the principal elements. The input energy comprises, of course, rate processes; input changes then involve acceleration (and deceleration) of rates. Morphology of the alluvial plain will be out of equilibrium during the accelerative phase, but when and if the rate returns to a steady state, the plain will adjust rapidly.
Since sediments deposited during a disequilibrium period must differ from those laid down during the subsequent as well as the prior steady state periods, an alternation of sedimentary types is inevitable. To the extent that sediments formed during the two steady states may be more similar than the intervening ones deposited during transition, the sequence will display at least a crude cyclicity. Of course, if the rate change is reversed, cyclicity is inevitable. If the rates show progressive change, whether intermittently or continuously, the alternations will no longer be cyclic. Obviously in any real alluvial plain sequence, a1, b1, c1, a2, b2, c2, a1, will not equal a2 exactly, and identification of cyclicity becomes a matter of judgment. Thus to some observers, the sequence clay-fine sand-siltstoneclay/coarse sand-siltstone-clay-siltstone, may be clearly cyclic; to others this may represent a crude alternation without cyclicity.
If, as I have suggested, fluctuations in climate, sea level, and diastrophism are as frequent as indicated by the record of the last million years, then any considerable section of alluvial plain sediments should retain some alternation of deposits. As in the case of allocyclic sequences, sedimentary evidence of auto cyclicity will be clearly preserved only if subsidence and compaction carry some parts of low-energy environments below the level of high energy environments before the shifting fronts of those environments occupy a particular cross section of the plain. Probability of preserVation in this case depends on the duration of the period between changes, on the duration of the high-energy phase, on the intensity of the change as well as on the rate of energy redistribution during the high-energy phase of the cycle, and finally on subsidence rates. Only changes induced by periodic subsidence are certain of preservation.
The quasi-equilibrium of a river is ultimately determined by load (quantity and caliber) and by discharge (Leopold and Wolman, 1957, p. 73). Changes in these factors will produce a new quasi-equilibrium characterized by a new distribution of energy values, a new sedimentary pattern, and a new morphology. In contrast, changes in slope, whether diastrophic or eustatic, do not change the quasi-equilibrium defined by discharge and load; they do change temporarily the characteristics of the alluvial plain so that they are out of adjustment with the quasi-equilibrium form, and by this means induce erosion or deposition until the plain again approaches that meta-stable form.
Table 3 summarizes allocyclic mechanisms and indicates the sedimentary sequences which would result from their operation. Basically, such sequences differ from the autocyclic types in their extent since they affect the total sedimentation on the alluvial plain and not merely its spatial distribution; they may coincide as well with cycles in other alluvial plains or in deltas and marine deposits unconnected to the original stream and its deposits. Since allocyclic variations involve changes in the source due to modification of erosion rates, they should also show a sequence of change in inherited mineralogy.
Table 3--Allocyclic changes in an alluvial plain.
|Mechanism||Energy system changes||Morphological and sedimentary consequences|
|A. Discharge||1. Increase||Current energy increase; frequency, extent, and velocity of overbank flow may increase. Effects of biological and chemical systems diluted.||Channels deepened with erosion of substrate and deposition of coarse sand and gravel on bars; floodplain receives coarser sediment or may be eroded if velocity is relatively high or cohesiveness of substrate is relatively low. Large channel fills associated with sheets of interlaminated sand, silt, and clay on floodplains. Little diversion until sand sheet is accumulated.|
|2. Decrease||Current energy decreases; frequency, extent, and velocity of overbank flow may decrease, but channel diversion may increase temporarily. Effects of biological and chemical systems concentrated.||Channels silted with deposition of fine sand on bars; floodplain receives sheet of clay with evidence of weathering, groundwater deposition, root action, etc. Shallow channel fills with frequent diversions over sandy substrate until accumulation of clay restricts such changes because of cohesiveness.|
|B. Load||1. Increase||Current energy decreases; diversions increase; effects of biological and chemical systems diluted.||Channels silted; floodplain receives sheet of clay with interventions of crevasse sands and silts. Shallow channel fills with frequent diversions.|
|2. Decrease||Current energy increases; diversion decreases. Effects of biological and chemical systems concentrated.||Channels deepened; floodplain receives sheet of clay with few interventions of sand or silt--may be eroded if reduction of load is great or cohesiveness low; strong admixture of biological and chemical sediments.|
|C. Slope||1. Increase||Current energy increases; diversions and overbank flow decrease; effects of biological and chemical systems concentrated.||Channels deepened; floodplains receive little or no sediment and will show progressive erosion as tributary channels develop; increasing evidence of root action, dessication, weathering, and groundwater deposition.|
|2. Decrease||Current energy decreases; diversions and overbank flow increase; effects of biological and chemical systems diluted.||As in B.1.|
Alluvial plain cyclothems of allocyclic type are typified by those of the Southern Appalachian Coal Basin (Wanless, 1946) which can be traced laterally into other basins. Localized "splits" and intercalcations of local cycles in that basin (Wanless, 1946, p. 128-129) as well as in others may be due to imposition of autocyclic phenomena or to the operation of other, more localized allocyclic mechanisms. Criteria suggested in the table are clearly inadequate for discrimination if the stratigraphic information is limited to isolated sections and the shapes and distributions of the various lithotopes are inadequately known.
Similarly, simultaneous operation of two or more allocyclic mechanisms will complicate the simple cyclic sequence. Some of the mechanisms, e.g., discharge variation induced by climatic change and slope variation resulting from glacial change in sea level, are at least partly interdependent. If these operate in phase because of their interdependence, they will invariably reinforce or cancel each other. On the other hand, mechanisms operating independently will vary in their interaction as they go in and out of phase. If their periods and intensities are irregular, the probability of exact repetition in a finite interval declines until the sequence of events and sediments approaches randomness.
Distribution of subsidence values varies not only through time but also from point to point over the plain. Variation in these values tends to cancel or reinforce vectors in the energy system and, thus, to confine major elements of the system to restricted parts of the plain. It also affects differentially the preservation of cyclic deposits on various parts of the plain. Sub-basins with rapid subsidence will show, obviously, the greatest cyclic variation since they draw into themselves stream channels with coarse deposits and, when the channels occupy other sub-basins, accumulate in contrast fine lacustrine clays, chemical deposits, and organic sediments. Platforms would show little variation since they receive sediments only during highest floods or at times of maximum aggradation.
Differential compaction would seem also to restrict the redistribution of energy values. When a particular sub-basin is receiving the bulk of the coarse sediment, initial rapid compaction of the underlying fine material would hold the surface at a low level and retain the channel in this position for some time. Fine sediments and peat of adjoining sub-basins with a slower rate of sedimentation would compact less rapidly and would remain relatively high for a time in comparison with the channel-occupied sub-basin. After initial compaction of underlying fines, however, accumulating channel and channel margin deposits would compact very little and would gradually increase the potential energy gradient into other sub-basins until diversion occurred. In the sedimentary record therefore, a channel diversion from sub-basin to sub-basin may be more striking rather than a large allocyclic change which occurs during the same period. If channel positions persist through several allocyclic changes, they may induce a megacyclic pattern as the sub-basin evolves; they will certainly impose local limits on the variation of environments during the allocyclic fluctuations. Environment maps of the Pennsylvanian of Illinois (Wanless and others, 1963) clearly show such limitations in delta development.
Depositional topography, as well as compaction and subsidence, restricts the rate and direction of energy redistribution on the alluvial plain. Unless subsidence and floodplain alluviation are extremely rapid, levees of earlier channels will tend to rise above the level of the floodplain and thus confine and direct overbank flow. Again, differential compaction may tend to exaggerate this local relief. Since these topographic ridges create vectors contrary to those of normal energy distribution, they impede energy redistribution, and, at the same time, redirect the channels when they do shift to sites not occupied for some time previous. To the extent that local deposition exceeds local subsidence and compaction, it will tend to overcome the restrictions they impose, but it will produce the same gross effects on cyclic sedimentation, exaggeration or damping of the evidence of cyclicity and, if longer period than the allocyclic mechanism, cyclothemic bundling or a megacyclothemic succession.
Finally, the character of underlying sediments affects the morphology of the alluvial plain. Channels cut in sand will have higher roughness, and in consequence a wider, shallower cross section, than those in clay and clayey silt (Schumm, 1960, p. 29). Logically, one might expect such channels to be more susceptible to migration and diversion. I have suggested in preceeding paragraphs that various factors may tend to confine a channel and near-channel deposits to a single belt for some time. In such cases, successive channels in the belt would be wider and shallower and diverted within the belt more often as the progressive superposition of overlapping sand deposits changed the substrate. The sedimentary section would display successively more numerous, but shallower, channel fills so that the sandy members of successive cyclothems would change from deep channel fills to sheet sands. The sequence would display a "megacyclic" character.
Observation indicates that the external factors affecting an alluvial plain change at intervals. If these changes involve an alternation of values, they will tend to produce cyclothemic sedimentation even if their periodicity is quite irregular.
Unequal subsidence, differential compaction, depositional topography, and the substrate differences tend to modify the effects of cyclic mechanisms and may produce in themselves a crude cyclicity by progressive changes in their effectiveness.
In all, cyclic phenomena and cyclothemic sedimentation are natural to any alluvial plain, and perhaps the absence of cyclothems or of certain types of cyclothems requires explanation. The apparent complete absence of cyclothems from an alluvial plain cross section may be the result, on one hand, of slow subsidence and aggradation with the consequent preservation of only the high-energy deposits or, on the other, of the random interaction of several cyclic mechanioms to produce a sequence without apparent pattern.
Absence of cyclothem sequences in a local section of a plain may result from localized subsidence rates which modify local energy values, and may thus tend to perpetuate a particular local environment in spite of overall cyclicity, or it may result from lack of sufficient local subsidence to preserve the evidence of cycles.
Preservation of recognizable cyclothems thus represents one of three possible cases in alluvial plain deposition. It occurs when subsidence rates are favorable, and when either (1) several mechanisms are in phase, additive, and therefore predominate in effect, or (2) a single mechanism predominates, or (3) the periodicity of the various predominant mechanisims is sufficiently long that each will typically occur in an otherwise steady state system. Because of internal restrictions--uneven subsidence, differential compaction, etc.--the dominant cyclothems on one portion of an alluvial plain may have quite a different origin than those on another part.
From the model, one must conclude that cyclothemic patterns in alluvial plain sediments should be very widespread temporally and geographically.
If the model is correct in general form, stratigraphers might well devote more attention to the search for cyclothems beyond the Carboniferous. Detailed information on the morphological characteristics of the sedimentary bodies in cyclothems could permit refinement of historical interpetation and the valid use of cyclicity in correlation.
If the model has this general validity, then its refinement is essential--by additional study of recent alluvial plains as well as tests against the stratigraphic record. Deductions and criteria offered here must be considered at best first approximations.
Finally, an argument over the origin of cyclothems is absurd; the argument must always deal with a particular cyclothem or a carefully defined class of cyclothems.
Alen, J. R. L., and Tarlo, L. B., 1963, The Downtonian and Dittonian facies of the Welsh Borderland: Geol. Mag., v. 100, p. 129-155.
Beerbower, J. R., 1961, Origin of cyclothems of the Dunkard group (upper Pennsylvanian-lower Permian) in Pennsylvania, West Virginia, and Ohio: Geol. Soc. America Bull., v. 72, p. 1029-1050.
Bersier, A., 1950, Les sedimentations rythmique synorogeniques dan l'avantfosse molassique alpine: 18th Internat. Geol. Cong., London, Rept., pt. 4, p. 83-92.
Brough, James, 1928, On rhythmic deposition in the Yoredale series: Univ. Durham Philos. Soc. Proc., v. 8, p. 118-126.
Duff, P. McL. D., and Walton, E. K., 1962, Statistical basis for cyclothems: a quantitative study of the sedimentary succession in the East Pennine Coalfield: Sedimentology, v. 1, p. 235-255.
Ferm, J. C., and Williams, E. G., 1960, Stratigraphic variation in some Allegheny rocks of western Pennsylvania. Am. Assoc. Petroleum Geologists Bull., v. 44, p. 495-497.
Fisk, H. N., 1944, Geological investigation of the alluvial valley of the Lower Mississippi River: Miss. River Comm., Vicksburg, Miss., p. 1-78.
Goodlet, G. A., 1960, Mid-Carboniferous sedimentation in the Midland Valley of Scotland: Edinburgh Geol. Soc. Trans., v. 17, p. 217-240.
Leopold, L. B., and Wolman, M. G., 1957, River channel patterns: braided, meandering, and straight: U. S. Geol. Survey Prof. Paper 282-B, p. 39-85.
Moore, Derek, 1959, Role of deltas in the formation of some British lower Carboniferous cyclothems: Jour. Geology, v. 67, p. 522-539.
Moore, Derek, 1960, Sedimentation units in sandstones of the Yoredale series (lower Carboniferous) of Yorkshire, England: Jour. Sed. Pet., v. 30, p. 218-227.
Potter, P. E., 1962a, Regional distribution patterns of Pennsylvanian sandstones in Illinois Basin: Am. Assoc. Petroleum Geologists Bull., v. 46, p. 1890-1911.
Potter, P. E., 1962b, Late Mississippian sandstones of Illinois: Illinois Geol. Survey Circ. 340, p. 1-36.
Read, W. A., 1961, Aberrant cyclic sedimentation in the Limestone Coal Group of the Stirling Coalfield: Edinburgh Geol. Soc. Trans., v. 18, p. 271-292.
Robertson, Thomas, 1948, Rhythm in sedimentation and its interpretation; with particular reference to the Carboniferous sequence: Edinburgh Geol. Soc. Trans., v. 14, p. 141-175.
Schumm, S. A., 1960, The shape of alluvial channels in relation to sediment type: U. S. Geol. Survey Prof. Paper 352-B, p. 17-30.
Sykes, Godfrey, 1937, The Colorado delta: Carnegie Inst. Washington Publ. 460, p. 1-193.
Van der Heide, S., 1950, Compaction as a possible factor in upper Carboniferous rhythmic sedimentation: 18th Internat. Geol. Cong., London, pt. 4, p. 38-45.
Wanless, H. E., 1946, Pennsylvanian geology of a part of the southern Appalachian coal field: Geol. Soc. America Mem. 13, p. 1-162.
Wanless, H. E., 1963, Origin of late Paleozoic cyclothems (abs.): Am. Assoc. Petroleum Geologists Bull., v. 47, p. 375.
Wanless, H. E., and Shepard, F. P., 1936, Sea level and climatic changes related to late Paleozoic cycles: Geol. Soc. America Bull., v. 47, p. 1177-1206.
Wanless, H. E., Tubb, J. B., Jr., Gednetz, D. E., and Weiner, J. L., 1963, Mapping sedimentary environments of Pennsylvanian cycles: Geol. Soc. America Bull., v. 74, p. 437-486.
Wanless, H. E., Ziebell, W. G., Ziemba, E. A., and Carozzi, A., 1957, Limestone texture as a key to interpreting depth of deposition: Cong. Geol. Internacional, 20th Sesion, Mexico City, sec. 5, v. 1, p. 65-82.
Weller, J. M., 1930, Cyclical sedimentation of the Pennsylvanian period and its significance: Jour. Geology, v. 38, p. 97-135.
Weller, J. M., 1956, Argument for diastrophic control of late Paleozoic cyclothems: Am. Assoc. Petroleum Geologists Bull., v. 40, p. 17-50.
Wells, A. J., 1960, Cyclic sedimentation: a review: Geol. Mag., v. 97, p. 389-403.
Wolman, M. G., and Leopold, L. B., 1957, River floodplains: some observations on their formation: U. S. Geol. Survey Prof. Paper 282-C, p. 87-109.
Wolman, M. G., and Miller, J. P., 1960, Magnitude and frequency of forces in geomorphic processes: Jour. Geology, v. 68, p. 54-74.