Imagine the floor of one of today's tropical rainforests.  It is covered with an uncountable number of leaves, branches and seeds.  Plant material is falling every day and covering up previous material.  This continues until the rainforest floor is a soft, spongy deposit of leaves.  Most of the plant matter decomposes, never to be seen again.  However, under certain circumstances, like occasional heavy rains and flooding, water moves large quantities of sand and mud around the rainforest.  If this does not occur, it unlikely that fossil leaves would be preserved from a rainforest environment.

With this in mind, what possible sources of mud deposits might have contributed to Castle Rock's fossil preservation?  Let's look at three possibilities: rivers; lakes; and mudslides.  But, first, an understanding of the specific geologic data found in our rock and leaf layers is required.  At Castle Rock, we find well-defined leaf layers (or leaf mat layers) with alternating layers of mudstone (fine grained silt/mud/sand).  There are three clear sequences of leaf layer and mudstone layer.  The mudstone layers found in between leaf layers contain few fossil leaves.  The occasional leaves found in the mud layers are usually in a non-horizontal position.  They are found at angles to the horizontal plane and are often curved so as not to fit even in an angled plane.  We've even found several leaves completely folded over at the primary vein resulting in a well-preserved leaf on both side, with a layers of mudstone in between.

The mudstone layers are nearly pure sediments of mud and sand, with occasional small tunnels in a vertical position.  All of these clues are important information in understanding the ancient environment and determining the processes that took place to fossilize Castle Rock's leaves.  Now, with all of these clues, the 'detective' work begins.

The Lake Theory

Did these leaves fall into a lake?  Were they covered by silt and mud accumulating at the bottom of the lake?  This is unlikely for several reasons.  If these leaves were falling into a lake, the result would be that the leaves would be somewhat evenly dispersed in the mud layers.  Leaves and sediment would be mixed and we would find fossil leaves throughout.  However, the distinct layers of leaf mat and nearly fossil-free mudstone indicate that that the Castle Rock leaves were not preserved in the bottom of a lake.  The mud appears to have been deposited in relatively quick events.  And, leaves at the bottom of a lake would generally all have the same horizontal orientation, or be folded over with little or no mud in between the folds.  The occasional leaves we find in the mudstone layers are at angles other than the horizontal leaf mat layers.  They are three-dimensional in that they are wavy and when they are found folded over, they have sedimentary deposition in between each half of the leaf.

The Mudslide Theory

The Castle Rock rainforest was near the uplifting Rocky Mountains.  Could it be that a mudslide from the surrounding hills and mountains provided the preserving mud.  This, too, is unlikely.  A mudslide typically contains a wide range of sediments from fine silts and sands to pebbles and larger rocks.  We don't find this in our layers of mudstone.  Also, a mudslide is a violent event that would have ground up our leaves so as not to be unrecognizable.  The leaves in our mat and the mudstone layers are very well preserved.  And, a mudslide event would leave an uneven horizontal leaf mat layer.  Areas of the leaf mat would be depressed by the weight and force of the mudslide.

Tree trunk in vertical growing position in mudstone layers

Block diagram of a meandering river (Utrecht University [1])

The River Theory

Was Castle Rock's rainforest near a river?  Were the mud deposits the result of occasional flooding?  This appears to be the most probably answer.  During a flood, the river breaches its banks and deposits sediment in the surrounding area.  The coarseness (fine, coarse, mixed) of the sediments can help us determine what type of flood helped to preserve the fossil rainforest.  Let's look at the various flood scenarios that occur around a river.  To do this, we will refer to the concepts, structure and dynamics of a 'meandering river'.  If you are unfamiliar with these terms, please read the "Meandering River" page first.

As you can see in the discussion and stratigraphic sections of a meandering river, there are numerous depositional environments surrounding the river.  The following discusses the most likely sources of deposition and the likelihood that they correspond to our Castle Rock environment.

• Crevasse-Splay: These are caused by a break in the river's bank (or levee).  Water gradually overflows the levee and erodes it.  Soon the water is rushing out of a breach in the levee.  Large amounts of sediment are deposited in a lobe shape extending from the river.  The resulting lobe of sediments often contain a chaotic mix of sediments ranging from fine silts to pebbles and rocks.  The leaves would likely not have survived a crevasse-splay in the well-preserved state we find in our quarries (if they survived at all).  The more violent nature of this event would likely have ground them up so as not to be recognizable.  However, the leaves in our mat layers and the mudstone layers are very well preserved and don't contain the larger pebbles and rocks.  We can't completely rule out a crevasse-splay though.  There are many types of meandering rivers with a wide range of sediments in their channels and banks.  If the surrounding area was composed of fine silts and sands, it's conceivable that a crevasse-splay event wouldn't be violent enough to disturb the rainforest in the manner described above.  But, wouldn't it disturb the leaf mat more and churn more leaves up into the mud layers?

 
• Lateral Accretion Units: These are deposits of sediment from the river channel resulting from the corkscrew motion of the water as it hits the cut bank.  The layers found in these units are not horizontal, but slope downward from the point bar towards the base of the cut bank.  They contain a large percentage of sand sediments.  These 'lateral accretion units' on the slipoff-slope could certainly be a source of 'some' fossilized leaves.  However, one wouldn't expect to find large layers of leaf mat in this very dynamic section of the river.  For these reasons, it's unlikely that this river environment represents what we see at Castle Rock.

 
• Overbank Deposits:  In a gradual rise in flood waters, the river may not break the levee as in a Crevasse-Splay.  It may simply overflow the bank.  Such waters usually contain much sediment in suspension, resulting in fine layers of silt/sand deposition.  This is precisely what we find in Castle Rock's mudstone layers.  This type of flood deposit would explain the undisturbed leaf mat layers.  These layers of large numbers of horizontal leaves appear to be the rainforest floor.  A gradual deposition of mud from overbank deposition would likely not disturb the leaf mat much.  As the waters recede and the suspended sediments settle, some floating leaves would be trapped in the mud/sand layers.  It's also likely that these leaves would not come to rest in horizontal positions.  Again, this is what we find in our mudstone layers.


The picture above of the tree trunk growing in a horizontal position in the mudstone layers adds more weight to the overbank deposit argument.  This tree is actually a medium sized trunk.  We have found much larger trees (6 foot diameter) and much smaller trees (2-3 inches).  Would the smallest trees have survived a crevasse-splay?  Not likely.  And, looking back at the Lake Theory, we wouldn't expect to find trees growing in a lake bottom.  The same is true for the lateral accretion units because this is an ever-changing environment not well-suited for trees.

But, what about those vertical tunnels in the mudstone layers?  It turns out that they are often filled with sand.  And, they are always found starting at the top of the mudstone layer.  Why is this?  Picture an overbank flooding event with large amounts of fine silts and sands.  It's likely that there was a considerable amount of standing water after the flood.  These vertical tunnels (or burrows) are fossil traces of previous worms and gastropods that swam around in the shallow ponds left by the flood.  The ponds may have dried up or been filled in by subsequent floods.  The fact that the burrows are filled with sand is revealing.  In subsequent floods, in a fine sand and silt depositional environment, the sands would be the heaviest and sink to the bottom first.  Therefore, these burrows would be primarily filled with sand.  Again, this is exactly what we find at Castle Rock.

Looking back at the crevasse-splay theory, it's less likely that this event would leave large shallow ponds in which burrowing creatures could survive.  The same is true for a mudslide event.  Both of these events typically leave large deposits with surfaces that dry and harden rather soon.

Another interesting observation is that we very rarely find smaller plants growing from seeds in our leaf mat layers.  In one or two cases, we see a seed begin to branch out horizontally in the leaf mat layer.  However, the mudstone layers do not have small roots growing, only remnants of past trees.  We find many seeds in the leaf mat layers, but they are intact and most have never developed.  What does this mean?  It's likely that the flood events that generated our mudstone layers occurred in rapid succession.  There was time for many leaves to fall to the forest floor, but we find no evidence that seeds took hold.

It's interesting to note that we don't know the size of the ancient Castle Rock rainforest.  As with modern rainforests, it was probably widespread.  However, it may have been preserved only in certain areas that benefited from the fossil preserving mud deposits.  In the future, other parts of this rainforest may be discovered at other bends and flood plains near the ancient meandering river.
 

A Rainforest Ending Flood

One thing is fairly certain in the Castle Rock rainforest.  A massive flood occurred sometime after our leaf layers were deposited.  It's possible that it even wiped out our portion of the rainforest.  Directly above our leaf layers, there is something we call the massive layer.  It's very thick and void of leaves.  It contains much coarser sediments and appears to be the result of a more violent episode.  We don't find tree trunks surviving this event.  This could have been a huge crevasse-splay or even a mudslide.  Much further above the massive layer, we find something we've termed the 'wrong layer'.  There are leaves in this layer, but it's not a thick leaf mat layer like our others which appear to represent the tropical rainforest floor.  Could it be that these were deposited from nearby rainforest areas that survived the massive flood event?  The further we get in the stratigraphic layers above our leaf mats, the more we get into the unknown.  Let's continue on with the story of the fossilized rainforest as we observe in our in our leaf mat layers...

Continuing the Fossilization Story

After the river's flooding provided the necessary mud/sand deposits, the fossilization process may begin.  The sediments entomb the leaves in a protective layer which can slow or stop their decomposition, providing the right environment to begin the fossilization process.  In subsequent flood events, the depositional process continues to build layer upon layer of alternating plant material, and sand and mud layers.  The lower layers are gradually compressed by the weight of the upper layers which begins to push out water and oxygen.  This too is required in the fossilization process because water and oxygen will rapidly decompose leaves.  Under the weight of many layers of matter, the lowest layers begin hardening and turning to mudstone or sandstone.  Eventually, environmental factors such as a massive flood event, severe climate changes, fires or falling volcanic ash cause the rainforest to die.  The area dries up and the rock layers continue to harden which fossilizes the leaves over a period of millions of years.

The 'Return' of the Fossil Rainforest

Most fossils are never found.  They lay in rock layers thousands of feet below the earth's surface.  But, occasionally, geologic forces bring the fossils back to the surface.  This is often accomplished by geologic uplift from faults and mountain building, combined with erosion which slowly reveals the lower layers containing the fossils.  Many times, the removal of the final layers to expose the fossils is done by development efforts such as excavation of land for new buildings or roads.  The Castle Rock fossil rainforest was found in such a place.  Many years ago, a train track was built upon an embankment.  Excavation and development was going on all around the area. Earlier developers may have had the fossil rainforest right at their feet.  They could have either missed it or didn't know what it was. (Or, they could have kept it a secret for fear that their construction would be halted.  But, without knowing, well give them the benefit of the doubt.)  Alternatively, erosion of the embankment may have slowly exposed the fossils.  In 1994, Steve Wallace, a CDOT paleontologist, was scouting the area in preparation for future road widening projects when he discovered many large leaves.  He contacted the Denver Museum of Nature and Science which brings us to today's fossil excavation efforts and the end of the very long story for the Castle Rock tropical rainforest - a story spanning 64.1 million years.  However, the ongoing efforts of fossil recovery and identification have started a new chapter of scientific discovery in learning about the ancient past of the planet on which we live.  To continue the story, please see "How Fossils Leaves Are Recovered".

1. Borehole archives, Rhine-Meuse delta studies, Department of Physical Geography (Utrecht University, Netherlands)
2. Principles of Sedimentology, Gerald Friedman & John Sanders, 1978, John Wiley & Sons.
3. "Meandering River Channels", Colorado Water Resources, Colorado State University