Re: coal again!

Keith B Miller (
Sat, 27 Mar 1999 14:44:20 -0600

Bill Payne wrote:

>No. Can you think of a modern analog where an organic deposit is
>partially transported and partially swamp? I suppose you might say a
>river could flood and wash tree debris into a swamp, but that would not
>match the features of coal seams I observe.

Much of the marsh of the Mississippi Delta in Louisiana are floating
vegetation mats (similar in some ways to sphagnum peat bogs). These are
made primarily of grass as well as a variety of other vegetation including
some trees. The roots (including those of the trees) are confined to the
mat and separated from the organic-rich mud below by a water layer of
several inches to several feet. The vegetation decays at the base of the
mat and falls down to accumulate the peat layer below.

As several people have already stated there is a very wide range of modern
peat forming environments. Peats form in depressions within glacial
outwash deposits, in oxbow lakes and other environments within alluvial
systems, in delta environments, and some cover extensive geographic areas
as in Indonesia. Some form in tropical and subtropical climates, others in
boreal climates. Some are fed by groundwater, some by high rainfall. One
model will not be adequate to describe coal beds from the past geologic
record. This diversity of coal-forming environments is recognized in the
geological literature.

>Coal seams generally have a razor-sharp contact with the underlying clay,
>shale or sandstone. (Those underlain by sandstone generally don't have
>rootlets and would likely be classified by most geologists as
>transported). The rootlets, where they are found, are individually
>recognizable. If this had been a swamp for thousands of years, the root
>zone would have no discernable interbedding of very fine sand and shale
>because the roots would have destroyed the interbeds, and I would expect
>that the contact between organics and soil would be gradational due,
>again, to the roots that would have lived and died in the soil. I asked
>one geologist about this on a field trip, and he replied in a rather
>agitated tone, "I don't know, maybe they rotted!" I didn't press him,
>but I'm wondering, why would the roots rot and the organics above the
>roots be preserved? And, why would the tiny rootlets be preserved but
>the larger Stigmarian axial roots, from which the rootlets radiate, not
>be preserved? Incidentally, I've only seen the Stigmarian axial root
>systems in the rubble at the site in Birmingham. Directly beneath the
>coals there, I've only seen rootlets which seem to be oriented
>vertically, but no axials.

I do not work on coal-bearing cyclothems, but I have seen several
coal-bearing cyclothems as part of fieldtrips and other excursions. Not
uncommonly, the coal is genetically unrelated to the paleosol immediately
underlying it. The paleosols appear to have developed before the peat
formation began, with the saturated conditions being superimposed on the
earlier developed paleosol profile. These are often widepread coal units
that were associated with the transgressive phase of a cyclothem. As the
sealevel rose and the shoreline moved inland the watertable rose producing
conditions favorable for peat accumulation. Plants (even trees) growing on
saturated ground do not send down roots because they need oxygen to
respire. The roots would tend to be confined to the surface of the
accumulating peat.

Paleosols _are_ associated with coal seams and are ubiquitous within every
cyclothem I have ever seen. As I have stated before, the literature on
these paleosols is voluminous. Paleosols formed in saturated environments
are very poorly developed. Good soil development requires the downward
migration of water and material. What are usually called underclays are
typical of soils formed in water saturated environments. Most of the
diagnostic features, which are very subtle, can only be clearly identified
in thin section. It is not uncommon for original sedimentary features to
be preserved.

In poorly developed paleosols within the Permian that I have studied,
remanents of original lamination a may be preserved. The paleosol may even
contain fossil debris such as ostracodes. A few paleosol profiles within
the sections I have studies were only confrimed after examining thin
sections. Soils generate very unique microstructural features that are
diagnositic of particular soil forming processes.

Please before you go dismissing the existence of paleosols below coal
seams, read some of the literature on modern and ancient paleosols. Go
look at buried Holocene or Pleistocene soils. You will find very few, if
any, large root structures in them. Soils are very dynamic environments,
and roots are very quickly destroyed after the death of the plant. Only
the roots of the last generation of plants on a landscape have really any
chance of preservation and then only if the soil is rapidly buried. Fine
roots actually have a better chance of leaving a record because of the
chemical auras that surround the zone of root hairs. Within the paleosols
I study, the roots themselves are rarely preserved. What are common are
root traces produced by the chemical reduction of the soil surrounding the
root. Also coomon are rhizocretions produced by the precipitation of
calcium carbonate around roots. The Permian paleosols I study possess all
the microstructural features found in analogous modern soils.

>In summary, to falsify the floating mat model, I would like to see
>intensely and deeply rooted underclays with little or no interbedded
>structure remaining, a gradational contact between the underclay and
>coal, roots connecting to the last generation of stumps in the swamp, and
>the last generation of stumps still standing in the coal where they grew.
> Having recently seen rootlets as deep at four feet below a coal seam, I
>do not think it reasonable to suggest that the roots were shallow and

Your expectations are incorrect based on my comments above. I would make
none of the predictions you make above. Thus your falsification is cannot
be used. My prediction is for very poorly developed paleosols
characteristic of water saturated environments, or of superimposed
"gleying" of earlier paleosols formed under conditions that preceeded coal
formation. This is indeed what is observed.


Keith B. Miller
Department of Geology
Kansas State University
Manhattan, KS 66506