Finds of Late Cretaceous dinosaur remains and substantial coal deposits at near-polar latitudes in both hemispheres seemed to confirm that the end of the Mesozoic experienced hothouse conditions.  Even so, both are very odd because of the darkness of polar winters; how could plants photosynthesise and supposedly cold-blooded reptiles stay warm?  To add to these oddities, it has now been suggested that periodically there were Antarctic ice sheets substantial enough to draw down sea-level (Miller, K.G. et al. 2004. Upper Cretaceous sequences and sea-level history, New Jersey Coastal Plain.  Geological Society of America Bulletin, v. 116, p. 368-393).  The possibility comes from a detailed stratigraphic and palaeontological analysis of Late Cretaceous sequences on and off the eastern seaboard of the US.  There are 11 to 14 sequences that show shallowing-upwards changes in the near-shore environment, somewhat similar to the cyclicity of Carboniferous times.  Calibrating the section with strontium isotopes and fossil changes suggests that sea-level ups and downs greater than 25 metres occurred swiftly (much less than 1 Ma).  This is considerably faster than changes due to variations in the volume of the ocean basins that result from fluctuations in sea-floor spreading rates, but if localised in eastern North America might have resulted from local tectonics, such as episodic deepening related to extensional tectonics.  The surprise is that the changes correlate well with those in western Europe and on the stable Russian platform, pointing to global, eustatic changes in sea level.  There is some correlation with oxygen-isotope records from foraminifera, so there is a strong possibility of a glacial cause.  The degree of fluctuation matches the effect on sea level of ice volumes of the order of 10⁶ to 10⁷ km³.  This is considerably more than the volume of the present Greenland ice cap, but on Antarctica it would have occupied only a small part of the surface.  There is another alternative; that eustatic changes are not well understood and there is a bias because of the Pleistocene correspondence between them and changes in continental Arctic ice sheets.  The amplitudes of the three different records do not match well, although their timing does.

Biology and iron minerals

The principal colouring agents in rocks, especially those of sedimentary origin, are iron minerals, foremost of which are oxides and hydroxides (e.g. hematite and goethite).  It doesn’t take much of either in a sedimentary grain coating to impart the vivid colour variations seen in some sedimentary formations.  It is easy to suppose that such veneers formed while the sediments were at the surface in an unconsolidated state, but there is much evidence that at least some, if not all, formed in buried sediments saturated with groundwater.  But the problem is getting the iron into pore spaces as well as precipitating its oxides and hydroxides.  Iron in its divalent state (Fe-2) is soluble, but exists only under reducing conditions, so it does not easily enter surface waters that supply groundwater.  In its trivalent state (Fe-3) iron is highly insoluble, and that is how it occurs in oxides and hydroxides.  Yet groundwater tends to lose its oxidising potential because dissolved oxygen is consumed by aerobic bacteria, and oxidation is required to convert soluble Fe-2 to insoluble Fe-3, so that hematite and goethite skins can form around sediment grains.  A clue to the precipitation method comes from a study of slime-encrusted surfaces in old mine workings (Chan, C.S. et al. 2004.  Microbial polysaccharides template assembly of nanocrystal fibers.  Science, v. 303, p. 1656-1658).  Although oriented towards the possibility of bacteria creating materials useful in nanotechnology, this non-geological paper might ring a few bells.  It shows how filaments (of the order of a few nm) that make up bacterial slime are associated with similarly thin and long filaments of one of the precursors to goethite.  The bacteria involved use the oxidation (electron removal) of Fe-2 to Fe-3 as a source of metabolic energy.  They colonise highly reducing waters, so there is a ready source of dissolved Fe-2 for them to exploit, especially in old mine workings, but also in groundwater cut off from the air  There is a snag for the bacteria, because Fe-3 is highly insoluble and could easily snuff out processes in the cells and cause their death.  So in evolving this chemo-autotrophic metabolism they would also have to evolve a means of disposing of its by-product.  The filaments are chains of polysaccharides grown outside the cell wall that act as templates for the precipitation of Fe-3 minerals.  The techniques used to show this include very-high resolution electron microscopy.  It would be interesting to see if very high resolution images of iron-stained mineral grains reveal  relics of these intricate structures.  Less powerful methods have already shown tiny spheres of magnetite in sediments above petroleum fields that formed biogenically through another metabolic process.

How old is the Dalradian?

Half the Scottish Highlands, from the Great Glen to the Highland Boundary Fault, and their equivalent in Ireland, is occupied by a convoluted orogen that is dominated by an almost exclusively sedimentary sequence of Neoproterozoic age – the Dalradian Supergroup.  Its importance is historical, for this is where many of the fundamental tenets used in unravelling complex terrains were developed and tested.  This still goes on, building on over a century of research in an easily accessible area.  Briefly, the Dalradian orogen evolved from a series of extensional basins, in a shelf area, that imposed considerable variations in thickness of the Dalradian sequence.  Protracted deformation in the Late Cambrian to Early Ordovician developed the structural complexity of the orogen, partly controlled by the original variations in sedimentary thicknesses.  We know the youngest age of the Dalradian, because its upper parts contain Cambrian fossils, estimated to be about 509 Ma old.  The earliest age for sedimentation has so far only been guessed, and must be younger than the 800 Ma of migmatites on which its lowest members rest .  The problem is that only one series of dateable volcanic rocks occur in the pile, and they are towards the top (601 Ma old).  At most the whole sedimentary sequence spans 300 Ma, and that in itself is most peculiar.  Most geologists have assumed continuous sedimentation under a great range of environments, but only because they have never found evidence for erosion in the sequence; hardly surprising from the complexity, and not-so-good exposure.  Yet nowhere on the planet is there a sedimentary sequence spanning such a time period that does not contain several unconformities; things have never been that quiet for so long.  Probably the only feasible way to get a handle on the duration of the Dalradian sedimentation is by matching geochemistry of the numerous marine limestones in the sequence with the global record for the Neoproterozoic, that is by seeking signs of the secular variations in the composition of seawater during that Era.  Scottish geoscientists have applied that technique, using 47 samples of Dalradian limestones (Thomas, C.W. et al. 2004. ⁸⁷Sr/⁸⁶Sr chemostratigraphy of Neoproterozoic Dalradian limestones of Scotland and Ireland: constraints on depositional ages and time scales.  Journal of the Geological Society of London, v. 161, p. 229-242).  Unsurprisingly, the results do not show a smooth curve that can be matched directly with various estimates of secular change in seawater strontium isotopes; the limestones occur haphazardly through the sequence.  The effort is not helped by considerable differences between global seawater strontium isotope curves compiled by several authors, so Thomas and colleagues’ interpretation is limited.  Yes, the Dalradian is younger than 800 Ma, but by how much cannot be said with confidence.  Its base is an unconformity that represents erosion of an older 800 Ma orogen, and how long that took is anyone’s guess.  The lowest Dalradian limestone falls in a strontium-isotope span that matches that for about 700 Ma, which fits with recent evidence for continued thermal activity in the underlying complex at 730 Ma.  Around the middle of the Dalradian deposition there occurs one of the most spectacular examples of possible glaciogenic rocks in the Precambrian, the Port Askaig Formation, which has been widely regarded as a product of one of the “Snowball” Earth events of the late Precambrian.  If the Dalradian deposition did begin around 700 Ma, then this unit cannot have formed in the earliest and best documented Sturtian glacial episode at 730 Ma, but perhaps in the younger Marinoan-Varangerian one (640 to 560 Ma).  The paper concludes with the time-honoured phrase “…await the application of alternative dating techniques”.   It may be a long wait, and perhaps the most important unresolved aspects of the Dalradian are whether or not its 30 km maximum thickness represents several distinct depositional basins, and if it contains numerous breaks in deposition.