PREPARATION BIASES IN SILICIFIED FAUNAS REVEALED BY PAIRED MICRO-CT AND ACID MACERATION

Silicified fossils in limestone are common subjects of taxonomic and paleoecological study. Fine-grained replacement preserves excellent detail, and the resistance of silica to acid allows fossils to be easily extracted through acid maceration. This preparation method has drawbacks, however, insofar as calcareous faunas and trace fossils, as well as depositional features such as directional sorting and sedimentary structures, will be lost during acid digestion. While paired petrography and acid maceration has shown that preferential silicification biases recovery based on taxon or body size, this approach lacks three-dimensional context within the sample. In order to document additional missing features and quantify biases, we analyze silicified limestones from the Upper Ordovician Edinburg Formation using non-destructive X-ray tomographic microscopy (μCT) paired with acid maceration.

Samples in this study comprise 15 cylindrical cores ~2 cm in width and height representing 5 distinct fossiliferous horizons with varying lithologies and fossil content. Common silicified and rarer pyritized fossils from these horizons include trilobites, ostracods, bryozoans, gastropods, and bivalves. Low-density siliceous fossils and high-density pyritized fossils were identified in μCT imagery through segmentation by voxel x-ray opacity. Morphometrics were assessed from both μCT imagery and photographed residues. Our results illustrate the complementary advantages and limitations of both methods. We were able to observe undisturbed fossils in μCT imagery that were damaged or missing in the corresponding maceration residues, along with signs of directional sorting and likely trace fossils. Residues, meanwhile, tended to better recover small (<3 mm) fossils not resolved in μCT imagery due to limitations of sample size, density separation, and noise. We also document taxon-specific pyritization and signs of early diagenetic mineralization in the samples; these are resolved well in μCT but poorly in residues. Overall, the performance of the two methods depends heavily on the abundance and size distribution of fossils within sampled material and the extent of mineralization during diagenesis. By assessing what is recovered versus lost in acid maceration, we can better inform analyses of silicified fauna.