We welcome the Comment by Wolkenstein (2014). Extraction of organic molecules from Paleozoic fossils is an emerging area of research, so establishing protocols for these analyses are critical. Wolkenstein raises important points regarding our research that our response will hopefully address. As acknowledged by Wolkenstein, our primary conclusions (O’Malley et al., 2013) are that biomarkers can be extracted from Mississippian crinoids and that the extracts from different taxa are a mixture containing different organic molecules. With this established, the next phase of our research will be to delineate individual organic molecules, as suggested by Wolkenstein, and to test the resolution of taxon specificity.
The page-length limit of a Geology article necessitates making choices about which details to include in a manuscript. Locality information for samples was provided by listing a reference (Ausich, 1983) that includes maps and coordinates.
The samples analyzed for this study were from the Edwardsville Formation of Indiana, as clearly stated in the manuscript. Illustrations of the crinoids from the Maynes Creek Formation from Iowa were presented (again clearly labeled) and used because these Iowa crinoids are one of the most striking examples of fossil crinoid species preserved with vivid colors. Analysis of material other than that from the Maynes Creek Formation was a purposeful choice to demonstrate the widespread nature of biomarker preservation in fossil crinoids. Only subtle color differentiation occurs among species of Edwardsville Formation crinoids, yet biomarkers are preserved.
It was an oversight on our part not to mention that more than one sample of a taxon was analyzed and consistent results were obtained. Most samples studied were isolated radial plates free from matrix, because these plates are readily identifiable on many Mississippian crinoids. However, we also analyzed other portions of the body and demonstrated that, for a given taxon, the same biomarker signature was present throughout the calyx and column.
The concerns that Wolkenstein has regarding the fluorescence excitation emission matrix (EEM) spectroscopy are well-founded had we relied only on that tool to reach our conclusions. Because living crinoids produce quinone compounds, we were exploring the possibility of using a faster, less-costly, and non-destructive method to determine the organic biomarkers in fossil crinoids. The application of EEM spectroscopy has been used extensively to tease out the “quinone-like” components in humic materials. Elegant and robust tools to deconstruct EEMs, such as parallel factor analysis (PARAFAC) are now used to determine “quinone-like” fluorescent moieties. Indeed, the field has moved well beyond the reference cited by Wolkenstein (i.e., Coble, 1996), and much of our work is based on research done by Cory and McKnight (2005) and more recent papers (Miller et al., 2009; SanClements et al., 2012; Gabor et al., 2014, plus the references cited in our paper). Unfortunately, we lacked the samples needed to create a PARAFAC model (typically n > 70), but we have plans to develop one if we can obtain the number of requisite samples. Nonetheless, our data demonstrated how much the fossil extracts have fluorescent moieties that resemble quinones on a qualitative basis, and we explicitly state in our paper that the identity of the molecule(s) is “tentative” and “quinone-like.” This last point is important, as none of the papers that utilize EEMs and PARAFAC (including ours) ever imply that the elucidated components are actual quinones, but rather are substances that fluoresce in the same region as quinones.
Wolkenstein’s concern regarding siloxane artifacts is legitimate; however, our group and the lab facility (the Campus Chemical Instrument Center, or CCIC, at Ohio State University) that processed and ran our samples are also cognizant of these contamination issues. Had there been a systemic blank issue we would have seen the peaks with characteristic siloxane mass-to-charge ratios (m/z) in all the samples. As a rule, we do not use silicone grease (an organic siloxane and possible contaminant) for any applications involving the extraction of samples in our lab. The CCIC is equally careful with respect to siloxane contamination, but the masses cited by Wolkenstein were not in the blanks, and as stated above, would have appeared in all of our samples. Further, the most common m/z present in the CCIC blanks in the past (not for this study) are 371 and 413, neither of which were in any of our samples. Thus, we are confident that the m/z we observed are real and not artifacts. Finally, we disagree with Wolkenstein regarding the inability to ionize neutral hydrophobic species. The CCIC routinely runs neutral hydrophobic compounds in their mass spectrometry facility using methods that facilitate ionization for these substances in the source (Chen et al., 2013), and they used the same approach for our samples.