This is a classic geological detective story, set in remote eastern Mexico, full of unique characters, mystery and intrigue. Like all geological stories, it is incomplete, but may turn our understanding of the Gulf of Mexico literally upside down, and could possibly answer the question of what caused one of the greatest climate changes in geological history, the Paleocene/Eocene Thermal Maximum.
The story starts in 2004 when I visited the turbidites of the Chicontepec Formation in the Tampico-Misantla Basin in the foothills of the Sierra Madre Oriental in eastern Mexico – a wild and timeless area and setting of the famous movie The Treasure of the Sierra Madre starring Humphrey Bogart.
I was accompanied by Mark Bitter, who had done his MSc fieldwork here in the 1980s. He showed me the photos and descriptions of his original outcrops, many now overgrown, including a small outcrop of what he and others described as a ‘coal’ within a sequence of turbidites in the Paleocene/Eocene Chicontepec Formation. His original photos revealed a steeply dipping section of thin-bedded sandstones with a variable thickness black bed in the middle. Mark did not remember its exact location, but knew it was close to the village of Chicontepec.
Initially, I did not think this outcrop significant in the overall story of the Tampico-Misantla Basin, but finding it soon became an obsession, because any coal in a turbidite sequence is anomalous. I drove all the roads near Chicontepec comparing the outcrop dips to those in the photograph and finally discovered it, set back about 100m from the road and very overgrown (Figure 1). The black ‘coal’ bed was clearly visible, recessed somewhat by erosion.
Over the next eight years I regularly visited the outcrop, including it on field trips that I led to the basin, always explaining that it was a ‘coal’ overlain and underlain by turbidites, but I said little about its environment of deposition, suggesting it was probably a raft of waterlogged trees carried out into the deep basin, sunk, buried and converted to ‘coal’. I was not happy with this interpretation, but it seemed the only logical explanation at the time and no one ever offered an alternative.
Collecting the Puzzle Pieces
In 2006 colleague Lynne Goodoff alerted me to the presence of several unconformities in the Chicontepec Formation, clearly visible on seismic. I had not found evidence of these in outcrop, but was intrigued by their number and basinwide nature. From cores I realized that high gamma ray features interpreted as shales overlying the unconformities were actually pebbly mudstones, probably deposited as debrites. Whenever high-resolution biostratigraphy was present in the wells, they confirmed the large hiatuses across the unconformities.
In February 2010, I collected samples from either side of the ‘coal’ bed, which were dated by foram specialist Peter Thompson as Foraminiferal Zone P5, spanning the Paleocene/Eocene boundary (Gradstein et al., 2004; Pearson et al., 2006).
Fast forward to 2012 when I was involved in a study for Pemex, building a basinwide stratigraphic framework for the Chicontepec Formation and thus identifying the sequence boundaries in about 99 wells. One of the team members, Dr. Don Van Nieuwenhuise, was using graphic correlation to identify sequence boundaries from well data. At the end of the study, one of the major boundaries was recognized at about 54–55 Ma, close to the Paleocene/Eocene boundary – the same unconformity which had been designated Discordancia (Unconformity) “A” by Pemex geologists for many years. We now had several pieces of the puzzle needed to put together the complete story of the basin.
The mystery of the ‘coal’ bed still baffled me. I knew it was approximately at the Paleocene/Eocene boundary and I had recently read the 2003 paper by Rosenfeld and Pindell on the possible isolation and drawdown of the Gulf of Mexico, in which they propose that the Gulf could have been isolated from the world’s oceans in the Paleogene, evaporated and then catastrophically refilled, rather like the Messinian crisis in the Mediterranean in the Miocene (Vai, 2016). However, they did not know exactly when it happened or how long it lasted, proposing only that it was close to the Paleocene/Eocene boundary.
The mysterious ‘coal’ bed was close to the Paleocene/Eocene boundary and overlain and underlain by turbidites – but where were the shallow marine facies we would expect in a normal regression and transgression? Could this so-called coal at approximately the P/E boundary be the ‘smoking gun’ that would prove Rosenfeld and Pindell’s theory?
The Search is On!
Determined to solve this, I returned to Mexico in March 2015. I obtained more detailed samples from above and below the ‘coal’ bed for age dating, to see if the results coincided with the sequence boundary we had identified in 2012. I also took samples of the bed itself, and asked specialists to determine their origin and environment of deposition. To my surprise, the samples were not coal; they contained no macerals or minerals and the ‘coal’ was, in fact, some form of fossil bitumen. It had no stratification, showed conchoidal fracture and immediately released oil and gas when exposed to fluorescent light.
The 14 samples from above and below the bitumen bed were again confirmed to be Foraminiferal Zone P5, spanning the Paleocene/Eocene boundary (Gradstein et al., 2004; Pearson et al., 2006). Peter Thompson also noted that there were variable quantities of fragmentary limonite tubes in the samples below the bitumen bed, but none above it, which from his experience suggested evidence of rooting. He also noted that there were no shelf-restricted foraminifera in the samples immediately surrounding the bed. Meanwhile, Dr. Van Nieuwenhuise had graphically correlated the same samples and confidently concluded that the bitumen bed represented a hiatus of 850,000 years (54.95–55.8 Ma).
So, is this bitumen layer a fossil oil seep, a depositional bed or injected into a void? And what was the depositional environment? I was convinced it was a bed and not injected because of thin clay layers observed above and below it. Small clay clasts found within the layer could also be seen in Mark Bitter’s 1984 photograph, when the outcrop was about 5m further forward than it is today. Mark’s work provided other clues. His measured section of the outcrop in 1984 recorded the bed as 60 cm thick, but today it is only about 12 cm. Why such big thickness changes over such a short distance?
There was only one way to answer these questions and get some fresh samples of the bed: we must core the outcrop and determine its lateral extent and thickness changes. A coring program was organized for September 2015 and a cast of actors assembled for the adventure, including lecturers and students from UNAM in Queretaro. Mark Bitter also decided to return to the area with us.
Coring and Excavating
We arrived in Chicontepec on a wet September evening and the next day we chose the core locations and started drilling. In case the core recovery was unsuccessful due to weathering, we also started digging out the bitumen bed by hand, following it along strike away from the bank, and soon began to see some unusual features.
Initially we thought the bed split along strike but then we discovered a large deformed clay clast within it, along with smaller clasts. The bed got deeper the further we moved away from the natural outcrop; by the end of the first day we had dug about 5m along strike, but were now about 1m below the surface (Figure 2, above). At the end of our excavation pit something strange was happening; the bitumen bed seemed laterally discontinuous because we could see a thick sandstone on strike with the part of the bitumen bed that we had exposed – as we’d seen on the 1984 photo. This mystery would have to wait until tomorrow.
We returned next day feeling more like archeologists than geologists! Extending the trench, we discovered that the bitumen bed was not discontinuous, but turned and headed north; we were excavating along an erosional unconformity, confirming the hiatus that Dr. Van Nieuwenhuise had predicted from the earlier samples. By the time we could no longer hand-dig we had discovered that the bed thickened along strike, took a 40° turn along an erosional unconformity (Figure 3, opposite) and continued in a more northerly direction. The two boreholes, 16m apart, showed that the bitumen bed was present 10m down dip, below the present outcrop face, but was absent in the subsurface along strike and 16m behind the outcrop face – evidence that it was pinching out eastwards.
Proving the Drawdown Theory
With this new evidence and more samples we returned home to analyze everything and discuss this unique outcrop and its possible relevance to the Gulf of Mexico drawdown theory. In our first paper (Interpretation, Cossey et al., 2016), we described the bitumen bed in detail, proposing that it was direct evidence for the Gulf of Mexico drawdown, but also hinting that the drawdown of the Gulf was somehow connected to the Paleocene/Eocene Thermal Maximum (PETM). After its publication I was contacted by Dr. Jerry Dickens at Rice University in Houston, who was intrigued by the hypothesis connecting the bitumen bed outcrop to the Gulf of Mexico drawdown and to the PETM, which he had been studying for 20 years. He was keen to follow it up.
We needed to go back to Mexico to continue the excavation northwards and trace the bitumen bed to its source to prove that it was a seep, not an injection feature. A small team, including Jerry Dickens, returned in May, 2016. This time we would excavate mechanically, since the fresh outcrop was going to be too deep for hand-digging!
We marked out the area to be excavated based on the previous year’s results and by the next morning the hole was almost completely excavated. Unfortunately, our excavation up the side of the unconformity showed that the bitumen bed only continues for about 5m, thickening in places to 30 cm, before it disappears, presumably at the source of the paleoseep (Figure 3, above). Jerry collected about 50 samples from the entire outcrop for nannos, forams and palynology analysis.
A modern-day analog to the Chicontepec bitumen bed might be the oil seeps from an unconformity seen along the cliffs near Carpenteria, California (Figure 4). The scale is about the same, and features such as small bitumen ‘fans’ are preserved at the base of the cliff.
An Ongoing Story
Our focus switched to a long, clean road cut near the village of Acatepec, about 10 km west of Chicontepec. Its importance is that we interpret it as a submarine canyon-fill sequence overlying Unconformity “A” – the same unconformity we had been excavating at the bitumen bed outcrop. I believe that this canyon fill represents the post-PETM sequence overlying Unconformity “A”, which formed when the Gulf of Mexico drawdown occurred. Jerry took more samples through this section and about 100 samples from the two outcrops are now being analyzed for forams, nannos and palynology and we await the results eagerly.
What we know for sure is that this is not the end of this adventure or the story, but just the beginning. The pieces of the puzzle will come together and the Sierra Madre will once again reveal some of its secrets – but only when it is ready.