CARNIAN/NORIAN CONODONTS FROM THE PIZZO MONDELLO SECTION (MONTI SICANI, WESTERN SICILY), GSSP CANDIDATE FOR THE BASE OF THE NORIAN: INTEGRATED BIOSTRATIGRAPHY, PHYLOGENY AND STABLE ISOTOPE ANALYSES.

Pizzo Mondello section (Sicani Mountains, Western Sicily, Italy), is one of the best localities for the definition of the Carnian/Norian boundary (CNB). The section is a 430 m thick continuous succession of marine limestones, characterized by uniform facies, ranging from Late Carnian to Rhaetian (Upper Triassic), named Calcari con selce. The section is entirely exposed and it shows high sedimentation rates of 20-30 m/m.y. (Muttoni et al., 2004). Pizzo Mondello section is not affected by diagenetic overprint and it thus preserves the original paleomagnetic and geochemical record and, moreover, it yields several biostratigraphically significant fossil groups. The conodont and radiolarian records are in fact very rich and, furthermore, the Calcari con Selce of the Western Sicily, and especially of the Sicani Mountains, are well known all over the world for the exceptionally rich and well preserved Late Carnian to Early Norian ammonoid faunas studied by G.G. Gemmellaro (1904) and for the very rich halobiid record (Gemmellaro, 1882; Montanari & Renda, 1976; Cafiero & De Capoa Bonardi, 1982; De Capoa Bonardi, 1984; Levera, 2009: PhD Thesis in progress). At present conodonts are considered the most useful tool to define the CNB. Thus, in order to find the most suitable conodont bioevent as primary biomarker for the CNB and to solve some problematic issues concerning the biostratigraphy and phylogeny of some Upper Triassic conodonts, the first 143 m of the section, where the CBN interval is located, were sampled in detail to conduct integrated taxonomic, biostratigraphic, phylogenetic and ecological studies on the conodont populations. The five most widespread Upper Triassic conodont genera (Paragondolella, Carnepigondolella, Metapolygnathus, Epigondolella and Norigondolella) have been revised in detail: this revision allowed to make very accurate biostratigraphic correlations of Pizzo Mondello with the other most complete Upper Triassic sections in the world, and to recognize some discriminating morphological features for the separation of these five genera and many others for the classification of the species. The changes in the conodont platform morphology observed through the section provided the following trends in the evolution of the pectiniform elements: a) the forward-shifting of the pit; b) the shortening of the platform; c) the gradual posterior prolongation of the keel end and d) the appearance of nodes on the platform margins and the evolution of nodes into denticles. Species belonging to the five studied genera have been thus assigned to two separate monophyletic lineages: i) the Paragondolella-Metapolygnathus lineage and ii) the Carnepigondolella-Epigondolella lineage. Nevertheless, the phylogenetic relationships of some species remained still problematic. In order to solve these problems, numerical cladistic analyses have been applied to the studied species. The morphological characters used to classify conodont species have been thus codified to generate numerical matrices and then analyzed by PAUP software (Swofford, 2002) under the parsimony criteria. The cladograms evidenced morphological relationships among genera and species, showing that the evolutionary trends identified are supported by the most parsimonious character states distribution among taxa, confirming the importance of the considered characters for the conodont systematic and revealing also other new important diagnostic characters. The cladistic analyses clarified thus the phylogenetic relations between the studied taxa: Paragondolella resulted the most probable forerunner genus of all the other Upper Carnian/Norian genera and Norigondolella is strictly related to it; Metapolygnathus and Epigondolella are recognized as two monophyletic groups while Carnepigondolella as a polyphyletic group; the two lineages have been confirmed and the systematic positions of problematic (e.g. Carnepigondolella nodosa and “Metapolygnathus communisti B”) and key species (e.g. Epigondolella quadrata and Metapolygnathus communisti) for the definition of the CNB have been better delineated. Given the richness of the conodont populations, a statistical approach was also applied. The quantitative curves of the absolute abundances for each genus showed potential ecological competition between Paragondolella-Carnepigondolella and later between Metapolygnathus-Epigondolella (and partially between Epigondolella-Norigondolella). This ecological behaviour supports the phylogenetic relationships among Norigondolella, Paragondolella and Metapolygnathus, and between Epigondolella and Carnepigondolella. Furthermore, cross checks of the quantitative curves evidenced the presence of three major turnovers, named T1, T2 and T3: at event T1 Carnepigondolella is replaced by its descendant Epigondolella in an evolutionary step; at event T2 Epigondolella is substituted by the mass occurrence of Metapolygnathus and at event T3 Metapolygnathus is succeeded by more evolved Epigondolellae and by Norigondolella. Looking for environmental explanations to these biological events, the conodont assemblages were compared to coeval δ18O and δ13C isotopic curves, based on new data from Pizzo Mondello. From the comparison, a correspondence appeared between higher 13C/12C ratios and the interval between events T2-T3, but not with event T1, in accordance with the interpretation of event T1 as an evolutionary turnover. In more detail, we observe the possible influence of environmental conditions on the absolute abundances of all the studied genera: while Epigondolella proliferates when seawater δ13C ranges between 2.1‰ and 2.5‰, Carnepigondolella proliferates in the range between 1.6‰ and 2.1‰; Metapolygnathus instead appears to be limited to environmental conditions related to higher δ13C values in the seawater. Given the results provided by the phylogenetic and ecological studies, the FAD of Epigondolella quadrata results the best primary biomarker for the definition of the CNB: this species is worldwide spread, its phylogenetic lineage is, it is globally recognized and its FAD (coinciding with event T1) is not triggered by any ecological perturbation. The FAD of Metapolygnathus communisti in sample NA35 is another possible biomarker, but its FAD, coinciding with event T2, is ecologically controlled and till now not found in North America and, thus, not a good species for global correlations.