Welcome to Part 2 in our discussion of the position of the chemocline in ancient basins. In Part 1 we introduced the issue at hand: Can we determine where in the water column the chemocline was? That is, what thickness of bottom water was anoxic, and how much of the overlying water column was oxygenated? If you haven’t read that piece yet and you would like to, you can find it HERE. In Part 2 I would like to review the framboid model, what these types of data sets tell us, and the pitfalls of doing this analysis in ancient rocks.
Framboids are spherical aggregates of microcrystallites. They get their name from the word “framboise” which is French for raspberry. The occurrence of framboids in sedimentary environments is well documented in Paleozoic basins. Moreover, they have been recovered from both sediments and the euxinic water columns in modern environments.
A robust body of literature exists on the occurrence and formation of pyrite in sedimentary environments. Indeed, much of our understanding is owed to the fundamental work of R.A. Berner in the 1960s. Some of the earliest documentation and discussion of framboids also occurred at this time. Early work by Vallentyne, 1963 and L.G. Love in 1965 suggested framboids were the result of biotic activity. However, numerous successes with synthesizing framboidal pyrite in a laboratory later disproved the biotic origin hypothesis. At this point, it is worth noting that there is no better synopsis and discussion of the morphology, formation, and redox implications of framboids than R. T. Wilkin et al.’s 1996 paper: “The size distribution of framboidal pyrite in modern sediments: An indicator of redox conditions” which appeared in volume 60 of Geochimica et Cosmochimica Acta (you can access the article HERE). Indeed, one would be hard pressed to find a recent study of framboids that does not reference this seminal paper. Much of the following discussion on framboid formation relies heavily their work and refences therein.
Pyrite framboids are thought to form by a series of steps to more sulfur-rich iron sulfides. The earliest formed microcrystallites occur as disordered mackinawite, followed by ordered mackinawite (FeS8), which further reacts to greigite (Fe3S4) before finally becoming pyrite (FeS2). While precursor phases are rarely observed in natural environments, greigite, and greigite-cored framboids have been found on occasion in modern environments. Further, it is assumed that the ferrimagnetic properties of greigite draw the microcrystallites together which forms the framboid morphology. Greigite is of particular interest, owing to the process by which it forms. The formation of greigite requires three necessary reactants: ferrous iron and dissolved sulfide in the presence of an oxidant, namely oxygen (or dissolved polysulfides or elemental sulfur). This step in the framboid formation process is significant because it allows us to use framboids as a paleo-redox indicator. Indeed, the co-occurrence of the reactants necessary to produce greigite is limited to the area immediately below the sulfide chemocline, the boundary separating oxygenated, from anoxic, hydrogen sulfide-bearing water.
In fact, multiple studies on the distribution of greigite have found it present in the greatest abundance at this redox interface (either in the water column or sediment). To summarize, pyrite framboids form through a series of steps involving progressive sulfurization of iron sulfide. The key step in the production of pyrite framboids is the formation of greigite, as 1) its magnetic properties lead to the formation of the spherical aggregates, the framboid, and 2) its formation is tied to the position of the sulfide chemocline.
Under normal conditions, the chemocline resides at or below the sediment-water interface (SWI). However, under certain circumstances a basins connectivity to global oceans may be impaired or cut off altogether. Under such conditions the supply of oxygen to bottom water may be less than the rate at which organic matter is oxidized. Such a scenario may cause the oxic-anoxic interface to rise above the SWI and into the water column. Moreover, if the rate of hydrogen sulfide (H2S) generation from bacterial mediated sulfate reduction exceeds the rate at which H2S is consumed by the formation of sulfide minerals, the anoxic bottom water will become euxinic (anoxic with free H2S in the water).
When the sulfide chemocline resides at or below the SWI, framboids form in anoxic sediment porewater. The size of the pores and the availability of reactants exert the greatest control on framboid growth. As a result, framboids formed in sediment are often large (10s to 100s of microns), and diverse in size. Conversely, when the chemocline resides in the water column, framboids are generally smaller (5-6 µm or less) and more uniform in size. Here, the weight of the framboids exceeds the ability of the water column to support them, causing them to sink away from the chemocline, and arresting their growth. Therefore, in ancient sediments, a population of small framboids with a narrow standard deviation may indicate the accumulation of sediment under an anoxic water column of some thickness. On the other hand, a population of large framboids with a high standard deviation may point to the growth of framboids in anoxic pore waters under an oxygenated water column.
Framboid diameter studies in ancient sediments are not without their pitfalls. For example, framboids may be small in diameter owing to extreme iron limitation. Recent sediments from the Santa Barbara Basin provide an excellent illustration of this phenomenon. Here, framboids with a uniform diameter of about 4 µm consistently accumulate under an oxygen-bearing water column. Likewise, in my recent paper on the Ordovician Point Pleasant Limestone, mudstone facies-host framboids that are uniformly small in diameter, while all other sedimentological and geochemical indicators suggest an oxic-dysoxic water column. Moreover, there is a correlation between the number of framboids and proxies for clastic influx to the basin. The most likely explanation for the small framboids in both the Santa Barbara Basin and Point Pleasant Limestone is iron limitation. In the case of the former, felsic source terranes provide iron-poor sediments to the basin. Similarly, a link between framboid occurrence and the flux of reactive detrital iron to the seafloor is also suggested by the Point Pleasant data.
We must also accept that in ancient sediments, framboid data provides a “time-averaged window”. Mud dominated strata often undergo 80% or more compaction. To get statistically valid sample sizes we must measure at least 100 framboids per sample. To accomplish this, we often analyze an appreciable thickness of rock that may represent 10s, 100s or even 1000s of years of sediment accumulation. This pitfall is quite evident in the Marcelles Shale where syngenetic framboids (those which grew in the water column) occur along side diagenetic framboids (those which grew in the sediment).
With consideration of the framboid formation process, and its uses and pitfalls as a paleo redox indicator, framboid diameters in the Marcellus Shale suggest accumulation of sediment under a bottom water that was dominantly anoxic-euxinic. However, the co-occurrence with diagenetic, and often quite large framboids, indicates that the chemocline fluctuated at a scale beyond the resolution of sample sizes needed to do the analysis, perhaps decadal or more in scale. However, we can assume that framboid diameters provide insight into the dominate redox conditions of Marcellus Shale bottom waters.
Returning to the topic of this series (Where is the Chemocline?), in ancient sediments, pyrite framboids give us useful information about the position of the chemocline. We can determine with some certainty where the chemocline resided, and the permanence of water stratification. However, we cannot say what thickness of bottom water was anoxic. It is worth noting though that pyrite framboids are often recovered from anoxic water columns tens to hundreds of meters above the SWI in modern day anoxic basins, such as the Black Sea, Lake Kivu (East Africa rift), Kau Bay (Indonesia), and Green Lake (New York, USA). It is worth considering that, although water depths are currently debated, most authors agree that the Marcellus Shale likely accumulated in 150 meters of water or less. Moreover, in some of these modern basins, the thickness of the anoxic water column exceeds the thickness of the oxygenated water column, in which case, “anoxic basin” is a most applicable term.
Thanks for reading along. Next week I’d like to discuss faunal assemblages in grey shale deposits overlying the Marcellus Shale and what insight they may give us into the thickness of “bottom water” redox conditions.