Does cement-induced sulfate limitation account for Burgess Shale-type preservation?

In a recent PNAS paper, Gaines et al. (1) proposed a mechanism of early diagenetic sealing and oxidant limitation to account for the preservation of Burgess Shale-type (BST) fossils (i.e., the “exceptional” preservation of carbonaceous compression fossils in Cambrian-aged marine mudstones) (1). It is an interesting hypothesis, but it raises a number of issues.

If early carbonate cementation is the critical factor in BST preservation, then BST biotas should be consistently associated with bed-capping authigenic carbonate. Such a relationship simply does not exist. In terms of anatomical detail, by far the finest BST preservation is found in shales and mudstones with no detectable carbonate, most notably in the Mount Cap and Deadwood biotas of western Canada. Carbonate is certainly associated with some BST assemblages (1), but even here, it is not demonstrably occurring as early cements. Most of the cemented horizons in the Burgess Shale, for example, are associated with coarser-grained, more permeable horizons rich in carbonate bioclasts (2). As such, the cements were most likely emplaced during later-stage diagenesis (and subsequent greenschist facies metamorphism), with marine δ¹³C_carb signatures derived from the original bioclasts. Notably, there is no carbonate cap associated with the exceptionally fossiliferous Great Marrella Layer and no regular association of BST fossils with cemented horizons (figure 2 in ref. 2). More generally, the collapse of 3D carcasses to 2D fossils is difficult to reconcile with pervasive early diagenetic cementation (2), and there is little evidence for elevated occurrences of marine hardgrounds or seafloor precipitates in Cambrian seas. The preservation of unoriented clays in lightly lithified, 1-km-thick Pliocene turbidite sequences (O'Brien, et al. 1980, cited in ref. 1) precludes the use of such fabrics as a proxy for early cementation.

Even if many/most BST biotas are not associated with bed-capping cements, it is still possible that they were starved of sulfate. At face value, their relatively heavy δ³⁴S_pyrite signatures might reasonably be linked to sulfate depletion and the interruption of bacterial sulfate reduction (BSR) (1). Such data, however, are only relevant in the context of corresponding marine sulfate signatures—and the early–middle Cambrian interval stands out for its extraordinarily heavy δ³⁴S_sulfate (>+50‰) (3). As such, the observed δ³⁴S_pyrite signatures most likely derive from sulfate-replete BSR fractionation, and the absence of down-bed isotopic depletion observed in some units (1) might be viewed as a consequence (rather than a cause) of prematurely terminated decay.

Even if sulfate was entirely depleted, it is still worth asking whether such conditions are sufficient for inducing BST preservation. Significantly, it is not BSR that is responsible for the primary breakdown of structural biopolymers in the absence of oxygen, but autolysis, hydrolysis, and fermentation (4). In other words, morphological degradation proceeds rapidly regardless of external oxidant supply. The ineffectiveness of simple sealing as a preservational mechanism has also been shown by taphonomic experiments on wax-embedded flies (5), as well as its limited applications in the food-processing and embalming industries. More generally, it is worth recognizing the dearth of organic-walled fossils from oxidant-depleted (and preferentially cemented) Archean and early Proterozoic sediments—and from stratified freshwater deposits of any age.