DIFFERENCES IN THE OXYGEN ISOTOPIC COMPOSITION OF UPPER ORDOVICIAN CONODONT SPECIES

In parallel d¹⁸O analyses of six conodont species from a sample of the Upper Ordovician Kimmswick Limestone, we found a difference of 1.2‰ between Panderodus gracilis (18.4‰) and Polyplacognathus ramosus (19.6‰).  Other taxa analyzed (Belodina compressa, Drepanoistodus sp., Phragmodus undatus, and Aphelognathus kimmswickensis) exhibited intermediate values. We believe this study represents the first demonstration of isotopic differences among co-occurring conodonts and argue the differences reflect primary isotopic gradients.

Overall preservation of conodonts from the Kimmswick is excellent.  The specimens lack any apparent overgrowth, have a color alteration index of 1, and exhibit weak orange luminescence when examined under a cathodoluminescent microscope.  Further, for four of the six taxa analyzed we measured 2 or 3 replicates, and for a fifth taxon (P. undatus) we measured separates composed solely of either P or S element types.  Only B. compressa was represented by a single analysis.  Variability among replicates was smaller than intraspecies variation, demonstrating that analytical precision is adequate to resolve the observed differences among taxa.  Results for the paired P and S elements argues against diagenetic artifacts as these two element types have different shapes and different proportions of mineralized tissues (i.e., variable diagenetic susceptibility), yet they yield virtually identical d¹⁸O values.

Assuming a seawater d¹⁸O value of -1‰, results yield an average paleotemperature estimate of 28°C, consistent with predictions of paleoclimate models.  The d¹⁸O variation among species could be the result of depth preferences among taxa that would translate to a ~5°C temperature or ~3ppt salinity difference between the environments inhabited by P. gracilis and P. ramosus.  Alternatively, the taxa may have lived during different times of the year or exhibited various vital effects.  However, these results are unlikely to represent sampling errors or temporal changes, because the specimens were isolated from a single sample collected from a <15cm-thick bed and because each analysis represents 50 to 100 elements.  Additional samples are being processed to test the consistency of the patterns observed.