Diagenetic History Of The Great Barrier Reef Of Australia

That giant organism is slowly giving up its deepest secrets.

From the November 2009 issue of Sedimentology : The Great Barrier Reef: a 700 000 year diagenetic history - Colin J. R. Braithwaite and Lucien F. Montaggioni

Most people rightly tend to think of the Great Barrier Reef as a living wonder. But the current living ecosystem, the coral communities and associated faunal and floral assemblages have been built on a foundation of a community of dead corals. And that ancient community when it was living grew on a foundation on an earlier community and so on.. for hundreds of thousands of years through many episodes of community growth and decline and growth again.

A borehole recovered sediments from Ribbon Reef 5 northwestern Australia. Analysis shows ten such coral reef growth episodes, ten sedimentary depositional units beginning around seven hundred thousand years ago. These episodic events correspond to major sea level changes driven by Pleistocene glacial growth (sea level fall) and melting (sea level rise). Coral reef building stopped when sea level fell and resumed when sea-level rose again.

This study delves into the mineralogical and geochemical transformations (Diagenesis) that occur in these reefs when sea-level fell and exposed the sea bed and the coral communities to fresh water.

As is the case with diagenetic research a lot of the base data is collected from micro scale analysis and this study does a detailed job. Cement mineralogy, (chemical precipitates) morphology, habits and associations are carefully documented. Textural features such as solution fabrics and pore space types are noted. The overall conclusion is that during sea-level fall meteoric (fresh water) aquifers developed in the exposed layers and there was considerable amounts of reaction of the coral skeleton with fresh water leading to destruction of minerals like aragonite and high Mg calcite and precipitation of low Mg calcite.

Based on these diagenetic textures distinct vadose zone (soil profile above the water table) and phreatic zone (below water table) demarcations are recognized in these ancient aquifers and some units show evidence of the water table being mobile i.e. migration up and down likely due to changing rainfall conditions.

That's the micro-level stuff and its a well characterized in this study.

My interest in diagenesis tends to make me look at the larger basin scale picture and I found much to think about in this paper. Diagenesis occurs during the establishment of large scale hydrologic systems in sediments and rocks. The scale and geometry of these hydrologic systems controls the extent and 3-D shape of alteration of the sediment.

Such thinking in terms of diagenetic reaction fronts, shapes and volumes is very important because these altered volumes of sediment may due to their properties of porosity and permeability act later in the sediment history as pathways and reservoirs for hydrocarbons and other economically important minerals like copper, lead and zinc.

Although not looked at from such an economic angle this study does recognize the larger controls on the development of hydrologic systems during sea-level falls.

One important diagenetic pattern revealed was that the fresh water reactions that took place cannot be tied to any particular emergent event although aquifer formation likely affected only the youngest deposition unit. This conclusion is based on the observation that there is no superimposition of diagenetic textures and cements.

This is a bit tricky. Say during a sea-level fall a particular depositional unit experiences meteoric diagenesis and then sea-level rises and drowns that unit. A new coral community will grow on that surface and form a younger depositional unit. Later sea-level falls again. Another fresh water aquifer develops. If it is thick enough it will affect not only the youngest sediment unit but the ones underlying it. So the older depositional units might then contain evidence of more than one generation of diagenetic products, a superimposition of events.

That kind of evidence of superimposition is lacking in these reefal units, leading the authors to conclude that the meteoric aquifers that formed were thin and did not penetrate deeper into the sequence.

The major reason for this is the lay of the land and the location of the reefal bodies in the context of the Australian shelf.  The Barrier Reef is located some distance away from the Australian mainland and away from any continental relief. During sea-level falls the undulating topography of the reef, made up of mounds and depressions would have meant that the reefs would have been exposed as a chain of islands.

The depth to which groundwater circulates in these exposed islands i.e. the thickness of fresh water lenses under these islands is governed by the width of the islands, the hydrologic conductivity of the sediment and the amount of recharge. These relationships observed in several Holocene islands suggest a meteoric lens thickness of just 1% of island width.

This contrasts with carbonates sequences which accumulate as bodies fringing continental land masses with coastal relief. Here during sea level falls, the hinterland relief provides a stronger hydraulic head for fresh water to flow towards and into the exposed sea bed and circulate deeper into the sediment layers, generating diagenetic reaction zones which cut across depositional units. 

One consequence of the thin meteoric lenses that episodically developed during the Barrier Reef history is that much of the sediment remained only partially altered. This is indicated not just by the sporadic presence of mineral aragonite which usually dissolves in fresh water, but by a cross plot of the oxygen and carbon isotopes of coral skeletal material in which the overall pattern of fresh water sediment interaction is brought out beautifully.

Here is the theory. Meteoric water is enriched in the lighter isotopes compared with sea water and compared with marine carbonate sediment. Imagine a marine carbonate sediment body reacting with meteoric (fresh) water. Marine carbonate is the dominant reservoir of carbon relative to meteoric water (means there is much more dissolved carbon in marine water than in fresh water) and thus the carbon composition of pore fluids quickly reaches a state of equilibrium with dissolving marine carbonate during water-rock interaction. So any product precipitated from this pore fluid will have the signature of marine carbon. In contrast, meteoric water is the main oxygen reservoir, and thus the oxygen composition of pore fluids with change towards a marine signature only after prolonged interaction with the isotopically heavier marine carbonate.

A pattern of relatively invariant oxygen and variable carbon is thus indicative of less water rock interaction during diagenesis.

That's what the bulk of data points of the key units in the above graph is pointing to. Here units 1, 9 and 2 are mostly marine and are not altered much by meteoric diagenesis. In the  rest of the units the bulk of the samples show a constrained oxygen isotope signal (compared to total range of 0 - 9.5 parts per thousand) and a very variable carbon isotope signal. The depleted but constrained oxygen isotope values suggest that the pore fluid composition were not impacted much by marine oxygen implying limited reaction with rock as does the variable carbon which would have been quickly acquired a uniform marine signature if water rock reactions had been prolonged.

All this is because of thin discontinuous meteoric lenses and thick vadose profiles where diagenesis is concentrated in pockets and along very thin reaction fronts leaving other pockets of the sediment body unaltered.

Enjoyed the paper. It's a good mix of micro-scale detailed work on cements and geochemistry but done in a way that allows you to draw a larger picture of the controls that landscape, platform geometry and hydrologic systems exert on diagenesis.

BRAITHWAITE, C., & MONTAGGIONI, L. (2009). The Great Barrier Reef: a 700 000 year diagenetic history Sedimentology, 56 (6), 1591-1622 DOI: 10.1111/j.1365-3091.2008.00982.x