Observations from mines and boreholes through the Galatia channel and related features, as well as similar channels associated with older and younger coal beds, indicate that the old model of a Mississippi-style delta having natural levees and crevasse splays is due for revision. Combined with a better understanding and climate influences not available to earlier workers, these observations lead to a new model of channel development that has connotations for larger concepts of cyclic sedimentation.
Model of Channel Development
Linkage of Climate and Eustasy
Peat Developed at Lowstand
Our model calls for the Springfield Coal (and, by implication, other Pennsylvanian coal seams) to be formed during eustatic lowstand under ever-wet conditions. Some previous authors, such as Flint et al. (1995), Bohacs and Suter (1997), and Heckel (2008), maintained that peat developed during transgression. According to their view, preserving a thick peat deposit requires a rising water table because otherwise peat is oxidized and lost. Flint et al. further held that most economic coal deposits developed in domed or raised mires, which exclude clastic sediment derived from nearby rivers.
However, evidence from coal-body geometry, coal petrography, geochemistry of coal and enclosing strata, and fossil-plant patterns strongly suggests that Desmoinesian coal in the Illinois Basin developed as planar, not domed, peat deposits (Cecil et al. 1985; Eble et al. 2001; Greb et al. 2002, 2003; Neuzil et al. 2005). Thus, peat accumulated at grade with the Galatia channel, with plants filtering clastics through the flanking belts now preserved as shaly coal. The Galatia was a river without banks or natural levees. Perennial flooding from the channel, coupled with an ever-wet climate, ongoing basin subsidence, and ongoing compaction of underlying sediment, maintained a consistently high water table throughout the duration of Springfield peat accumulation.
Rapid Transgression, Gradual Regression
Evidence from the Pleistocene, particularly the most recent deglaciation, indicates that melting can be surprisingly rapid, even “catastrophic.” Blanchon and Shaw (1995) inferred, based on drowned reefs in the Caribbean and Gulf of Mexico, three “catastrophic, metre-scale sea-level rise events” during the last Pleistocene deglaciation. Gregoire et al. (2012) calculated that a sea-level rise of 45.9 to 59 ft (14 to 18 m) took place within a span of about 350 years close to 14,000 years ago, and a rise of 29.5 ft (9 m) took place within 500 years about 8,200 years ago. Data from Greenland ice cores indicate two remarkably sudden Late Pleistocene warming events. One at 11,700 years B.P. lasted 60 years, and an earlier event at 14,700 years ago spanned a mere 3 years (Steffensen et al. 2008). Other evidence and examples may be found in Webster et al. (2004), Wright et al. (2009), Bird et al. (2010), and Törnqvist and Hijma (2012). In contrast, the growth of a continental ice cap requires a much longer interval of time, measured in thousands to tens of thousands of years (Muhs et al. 2011; Dutton and Lambeck 2012).
The authors cited above suggest that any triggering event that raises the sea level may set off a chain of events leading to the rapid destruction of ice caps. The trigger could be the onset of a warmer climate or the failure of ice dams that hold back large bodies of fresh water, such as glacial Lake Agassiz and Lake Missoula. The rising sea breaks up floating ice sheets and releases fleets of icebergs, which melt in warmer waters and raise the sea level further. Moreover, the loss of sea ice changes the ocean and atmospheric circulation, leading to more melting. In contrast, the growth of continental glaciers is “glacially slow” because snow and ice can accumulate only so fast. In fact, the global cooling required to bring on an “ice age” reduces the capacity of the atmosphere to hold water vapor and yield snow.
The rapid drowning and burial of the Springfield peat swamp has counterparts in other late Paleozoic deposits. Ravinement surfaces are reported from other coals, such as the Herrin Coal of Illinois (DeMaris et al. 1983), which also has a buried autochthonous flora adjacent to the Walshville paleochannel (DiMichele and DeMaris 1987). Lower Permian carbonate cycles of Kansas commonly begin with a thin transgressive lag of fish bones, ostracods, and intraclasts, implying abrupt, erosive transgression (Miller et al. 1996). The rapid marine transgression that terminated peat accumulation repeats throughout the Pennsylvanian in the Illinois Basin among coal beds that lack channels and gray shale roof. Most commonly, the coal is sharply overlain either by offshore black, phosphatic black shale or by marine limestone; at the contact, coal laminations are generally truncated, indicating some erosion of the top of the peat body. At the base of black shale, a thin pyritic shell breccia may be present (Zangerl and Richardson 1963). Where nonmarine gray shale overlies low-sulfur coal, in situ tree stumps and tidal rhythmites are commonly in evidence. We have logged examples including the Lower Block, Murphysboro, Colchester, Herrin, and Danville Coals in addition to the Springfield.
Such conditions of rapid, pulse-like sea-level rises likely also occurred during the Pennsylvanian ice age (Archer et al. 2015, 2016). Such considerations further militate against the notion of “keep up” time-transgressive peat-swamps, created by rising base level and driven across the landscape by sea-level rise (e.g. Heckel 1995). Peat formed much too slowly to keep up with the abrupt sea-level rise of a typical deglaciation.
Relationship of Effingham and Galatia Channels
The Effingham channel clearly was cut, filled, and abandoned prior to development of the Galatia channel. In fact, the Galatia channel crosses the Effingham at a right angle. Thus, the two channels represent separate cycles of sedimentation. Previously, all strata between the Houchin Creek and Springfield Coals were assumed to belong to a single cycle (Summum cyclothem), reflecting a single episode of marine transgression and regression.
In the Midcontinent Basin, the correlative interval contains two cyclothems. The minor Upper Blackjack Creek cycle (Figure 61) falls between the major Lower Fort Scott (Excello Shale) and Upper Fort Scott (Little Osage Shale, correlative with Turner Mine Shale) cycles (Heckel 1994, 2002, 2013). As Heckel (2002, p. 110) stated, “The upper part of the Blackjack Creek Limestone extends as a bed into the upper part of the Morgan School Shale in Iowa, where it contains moderately abundant conodonts [and] thus apparently represents a minor transgression.” Heckel further noted that “the Lower Fort Scott cyclothem loses both its lower and upper bounding paleosols a short distance south of the Kansas border in northern Oklahoma [see Figure 11 on p. 20 in Heckel 2013]. This supports the idea that the lower part of the Midcontinent shelf in southern Kansas and northern Oklahoma was at a lower Pennsylvanian elevation that the entire Illinois Basin” (P.H. Heckel, written communication to W.J. Nelson, June 6, 2014).
We propose the following scenario: During late highstand and early regression of the major Lower Fort Scott cycle, deltaic sediments of the Delafield Member essentially filled the Illinois Basin. As the sea level continued to fall, the Effingham channel became incised and established a meander belt. Then came the minor Upper Blackjack Creek transgression, drowning the Effingham channel and backfilling it with sediment. When sea level again declined, a new fluvial system—the Galatia channel—became established on the exposed shelf. No limestone or marine fossiliferous shale marks the Upper Blackjack Creek event in this basin because the sea-level rise was relatively brief and low in amplitude. Tidal rhythmites in the upper Effingham channel fill in the Elysium core and local, thin coal in other boreholes points to estuarine conditions, not fully marine.
As an alternate hypothesis, an autocyclic process or tectonic movement in the basin might have led to the abandonment of one channel (Effingham) and the establishment of another (Galatia) during a single eustatic cycle. Earth movements might have changed the regional gradient from southeast to southwest, inducing a change in channel orientation. This idea is not far fetched because both the Springfield and Herrin Coals thin across the La Salle Anticlinorium and other basin structures, indicating syndepositional tectonism. The channel-forming process clearly was complex and required a substantial amount of time.
Nelson, W.J., S.D. Elrick, W.A. DiMichele, and P.R. Ames, 2020, Evolution of a peat-contemporaneous channel: The Galatia channel, Middle Pennsylvanian, of the Illinois Basin: Illinois State Geological Survey, Circular 605, 85 p., 6 pls.