Carbon isotope values in conodont elements from the latest Devonian–Early Carboniferous carbonate platform facies (Timan-Pechora Basin)

The δCorg values and their variations in conodont elements of the latest Famennian–Tournaisian species Polygnathus parapetus Druce are considered. The conodonts studied come from the latest Famennian–Tournaisian interval of the Kamenka River section (Timan-Pechora Basin). The carbon isotope composition of conodont elements had an average value of –25.2‰, suggesting a low trophic level of this species. During the latest Famennian (praesulcata conodont Zone) δCorg shifted to more negative values (up to –30.4‰), which may be attributed to changes in the global carbon cycle and local influx of organic matter of terrestrial origin during the terminal Famennian regression.


INTRODUCTION
Conodonts were the Palaeozoic and Early Mesozoic extinct group of marine animals possessing debated affinities (e.g. Donoghue et al. 2000;Blieck et al. 2010;Turner et al. 2010). The only mineralized parts of conodonts are tooth-like elements disposed in a bilaterally symmetrical apparatus. Conodont elements are composed of complex aggregates of proteins and apatite-(CaF) (e.g. Trotter & Eggins 2006;Rosseeva et al. 2011). The mineral composition of conodont elements is known in detail; however, their organic matter is studied to a lesser degree. Previous investigations have demonstrated that organic matter, consisting of less than 4% of a conodont element, is composed of collagen-like protein (e.g. Fåhraeus & Fåhraeus-van Ree 1987;Kemp 2002;Rosseeva et al. 2011;Zhuravlev 2017a). The protein network is surrounded by aligned crystallites of apatite-(CaF) and strongly incorporated into the mineral matrix. This incorporation provides unique conservation of organic matter, demonstrating a preserved supramolecular protein structure (Zhuravlev 2017a).
The low content of carbonate ions in conodont apatite of lamellar, paralamellar and albid tissues (Trotter & Eggins 2006;Frank-Kamenetskaya et al. 2014) makes it possible to study carbon isotope values of organic matter in conodont elements without their demineralization. The first information about isotope composition of conodont organic matter was published by Over & Grossman (1992). These authors reported δ 13 C org in conodont elements of the Late Devonian genus Palmatolepis (from -24.5‰ to -24.0‰), Early Carboniferous (Mississippian) siphonodellids (-26.3‰ and -27.3‰) and Late Carboniferous (Pennsylvanian) Streptognathodus elegantulus (from -23.0‰ to -24.0‰), and noted that significant isotopic differences may be related to 'local changes in source of organic carbon, global changes in the carbon budget, or to dietary differences among conodont animals' (Over & Grossman 1992 p. A214). Some data on the trophic differentiation of the Late Viséan (Mississippian) conodonts based on δ 13 C org were published later (Nicholas et al. 2004).
This study considers δ 13 C org values and their variations in conodont elements of the latest Famennian-Tournaisian species Polygnathus parapetus Druce. The key objective is to evaluate potential implications of δ 13 C org of conodont elements for palaeoecological reconstructions, including the position of conodonts in the food web and probable causes of temporal fluctuation of δ 13 C org during the Devonian-Carboniferous transition.

MATERIAL AND METHODS
Conodont elements were studied from the Kamenka River section in the southern part of the Pechora Swell (northern Cis-Urals, N 65°04′27.4″, E 56°42′50.9″) (Fig. 1). The section was located on the northeastern margin of Laurussia during the late Famennian-Tournaisian and comprises the stratigraphical interval from the latest Famennian praesulcata Zone through the late Tournaisian crenulata Zone, being 16.5 m thick (Zhuravlev et al. 1998;Vevel′ et al. 2012;Zhuravlev 2017b). The sedimentological and palaeontological characteristics of the section reflect a palaeoenvironmental setting in shallow tropical waters (Zhuravlev et al. 1998;Vevel′ et al. 2012). About 100 limestone samples were taken from the section. The processing of samples followed the standard procedure (dissolution of limestone in 10% buffered acetic acid). The residues were washed through a sieve of 70 µm, dried and conodont elements were picked out. Rather rich and diverse conodont faunas were obtained from the Kamenka River section, which provided the biostratigraphical framework (Zhuravlev et al. 1998;Vevel′ et al. 2012;Zhuravlev 2017b) (Fig. 2). The preservation of conodont elements is fine and they have a Conodont Alteration Index (CAI) value of 1 (i.e. T < 50 °C). The low grade of thermal maturity suggests the preservation of the original carbon isotope composition of organic matter (McKirdy & Powell 1974). According to McKirdy & Powell (1974), unmetamorphosed organic matter is isotopically lighter than metamorphosed material, and postdepositional thermal alteration may lead to a positive shift in carbon isotope composition. In contrast, no significant correlation between thermal alteration and δ 13 C org was discernible from data by Strauss & Peters-Kottig (2003).
The P1 elements of the long-lived conodont species Polygnathus parapetus Druce (Fig. 3), which are abundant in the study samples, were selected for investigating organic carbon isotope values. The predominance of organic-rich lamellar and paralamellar tissues in these conodont elements makes them suitable for this study (Müller & Nogami 1971). The histological study of conodont elements demonstrates good   preservation of the primary structure of hard tissues. A number of thin and polished sections of P1 elements of P. parapetus were studied with SEM, microprobe (EDS), and optic microscopy. The hard tissues of conodont elements prove the absence of re-crystallization and post-mortem uptake of minerals such as sulphides ( Fig. 4A-C). Geochemical analysis of bioapatite of two polished sections of P1 elements, mounted in the low molecular weight epoxy resin and coated in carbon, was performed using a VEGA TESCAN microprobe with the precision of 0.1 wt.%. Microprobe (EDS) data demonstrate the absence of Fe, Mn, Al, Zn and Pb, which are supposed to be post-mortem contaminants (Trotter & Eggins 2006;Zhuravlev & Shevchuk 2017) in all hard tissue types. The Ca/P and Sr/Ca values of 2.18-2.20 and 0.005-0.011, respectively, are close to the Ca/P and Sr/Ca values of unaltered conodont elements reported earlier (Zhuravlev & Shevchuk 2017). The lamellar structure of the lamellar and paralamellar tissues is well preserved (Fig. 4A, C). These observations suggest a rather good preservation of the conodont The study of thin and polished sections of conodont elements in association with micro-CT and optic microscopy data allow elaborating the histological model of the P1 element of P. parapetus, demonstrating the distribution of albid tissue (Fig. 4D). According to the model, lamellar and paralamellar tissues compose about 94% of the P1 element of P. parapetus. Albid tissue, comprising about 6% of the element, forms cores of denticles of the fixed blade and the uppermost parts of denticles of the anterior free blade only.
Separated P1 elements of P. parapetus were washed with ethanol and distilled water and then used for analysis of carbon isotope values with the DELTA V Advantage mass spectrometer equipped with the Thermo Electron Continuous Flow Interface (ConFlo III) and Element Analyzer (Flash EA 1112). The δ 13 C org values are reported relative to the PDB standard. Isotope analyses were performed at the CKP 'Geonauka' of the Institute of Geology Komi SC UrB RAS (Syktyvkar, Russia). The international standard USGS-40 (L-Glutamic acid) was used. The precision of the δ 13 C org value is ±0.15‰.
The diagenetic degradation of the organic matter of conodont elements was estimated by their demineralization in 1N solution of HCl. The demineralization of conodont elements yields a protein 'pseudomorph' possessing the same size and shape as the original element. The presence of a protein 'pseudomorph' is likely a guarantee of the structural integrity and good preservation of organic matter (Sealy et al. 2014). The 'pseudomorphs' exist due to collagen molecules which are still bonded together to form the organic framework of conodont element tissues. In poorly preserved diagenetically altered conodont elements, demineralization does not yield a 'pseudomorph'. In this case the conodont element may yield amorphous fragments of gelatinous material composed of degraded collagen molecules, and isotopic measurements can produce unreliable results (Sealy et al. 2014).
In any case similar preservation of the studied conodont elements provides uniform degradation of organic matter and suggests good preservation of relative variations in the isotope composition.

RESULTS
Ten samples were analysed for δ 13 C org in P1 elements of Polygnathus parapetus. The δ 13 C and δ 18 O values in the host carbonate rocks were studied as well (Table 1). The δ 13 C org values in P1 elements of P. parapetus vary from -30.4‰ up to -22.5‰; the average value is -25.2‰ (Fig. 5).
During the latest Famennian (praesulcata conodont Zone) δ 13 C org shifted to more negative values (up to -30.4‰). The early Tournaisian δ 13 C org record shows weak variations around the average value (from -25.9‰ up to -22.5‰). The highest δ 13 C org value of -22.5‰ is observed in the Lower crenulata conodont Zone (Fig. 5,  bed 9). Significant intraspecific isotopic differences (about 8‰) may be related to environmental changes as well as to the influence of the host rock composition and diagenetic degradation of conodont organic matter.
Very weak negative correlation (correlation coefficient R² = 0.0965) is observed between the carbon isotope composition of conodont elements and host carbonates (Fig. 6), which allows excluding the significant influence of the composition of surrounding media on δ 13 C org values. Similar preservation of all the studied conodont specimens supports the low probability of the taphonomic control on variations in the δ 13 C org value.
The composition of the main types of hard tissues of two P1 elements from the upper part of bed 7 and bed 9 was determined by microprobe ( Fig. 4B; Table 2). All the tissue types possess a similar Sr/Ca ratio of 0.005-0.011. These values of Sr/Ca are close to those of the Frasnian polygnathid conodont elements of exceptional preservation (0.003-0.012) (Zhuravlev & Shevchuk 2017).
It was suggested that the distribution pattern and concentrations of Sr in conodont element tissues was most likely controlled by biomineralization and hardly affected by secondary processes (Trotter & Eggins 2006;Zhuravlev & Shevchuk 2017). Thus rather high Sr/Ca values in conodont bioapatite suggest a low level of biopurification of Sr in conodonts (Peek & Clementz 2012).

DISCUSSION
The average δ 13 C org values in P1 elements of Polygnathus parapetus of -25.2‰ are close to those of recent zooplankton (Bohata & Koppelmann 2013), but far from the δ 13 C org values of collagen of bones of most marine vertebrates (from -14‰ up to -10‰) (Schoeninger & DeNiro 1984). The carbon isotope ratio in consumer tissues is used as a tool in assigning the trophic status, because the δ 13 C of an organism closely reflects the δ 13 C of its food. The low δ 13 C org values in P1 elements of P. parapetus, if supposed as unaltered, suggest a low position of the species in the food chain. This δ 13 C org value is close to that of primary producers which utilize the C 3 photosynthetic plant system (δ 13 C org values range from -35‰ to -20‰) (Hare et al. 1991;Peters et al. 2005). Primary producers, which use the C 3 pathway to fix carbon from CO 2 during photosynthesis, are eukaryotic algae, autotrophic bacteria, and some marine and terrestrial plants (Peters et al. 2005). So we can suppose that marine algae may constitute a main food of P. parapetus, and conodonts of this species were low-level consumers. Besides the δ 13 C org value, the Sr/Ca ratio (0.005-0.008) in P. parapetus bioapatite is quite high in relation to high-level marine consumers (Peek & Clementz 2012). According to data on recent marine organisms, the higher Sr/Ca values, close to those of sea water, are characteristic of lower trophic levels (Peek & Clementz 2012). It is suggested that the Sr/Ca value in conodont bioapatite of P. parapetus is most likely caused by the low trophic level of this species.
There is no apparent contradiction between rather herbivorous or basal carnivorous feeding specialization and tooth-like morphology of conodont elements of this species. Tooth-like morphology, the presence of micro-wears and traces of in-vivo injuries, reported for elements of conodont apparatuses of Ozarkodinida, allow just reconstructing biomechanics of the apparatuses (Purnell 1993(Purnell , 1995Purnell & Donoghue 1997;Donoghue 2001;Zhuravlev 2007;Purnell & Jones 2012). These studies demonstrate that P1 elements of ozarkodinid conodonts, including the genus Polygnathus, are acting as occluding pairs (Nicoll 1987;Purnell & von Bitter 1992;Donoghue & Purnell 1999;Martínez-Pérez et al. 2016). A commonly supposed function of these elements is the cutting and grinding of food particles (e.g . Nicoll 1987;Purnell & von Bitter 1992;Zhuravlev 2007). Following interpretations related to the trophic specialization of conodonts are very circumstantial, and based on doubtful analogies with Vertebrata teeth (see discussion in Turner et al. 2010;Blieck et al. 2010). Thus the interpretation of the trophic position of P. parapetus based on δ 13 C org and Sr/Ca values allows us to suppose what kind of food particles were cut and ground by P1 elements. Establishing a correlation between conodont element morphology and the position of the corresponding conodont taxon in the food web may be subject to future research.
Variations in the carbon isotope ratio through time, reflected in the stratigraphical sequence (Fig. 5), may result from changes in environmentally induced conodont metabolism, the trophic level of feeding, or isotopic fractionation at the base of the food web (phytoplankton) that is transferred through the food web.
Intraspecific variations in the metabolism of P. parapetus are difficult to estimate because of lack of data on conodont physiology. Recent marine consumers demonstrate weak intraspecific fluctuations of the δ 13 C org value. DeNiro & Schoeninger (1983) reported variations of up to 1.0‰ in bone collagen resulting from inter-individual differences in metabolism. The observed fluctuations of about 8‰ are too great to be attributed to the inter-individual differences.
The variations in the isotope composition of the food source seem to be a more probable cause of the observed changes in the δ 13 C org value of P. parapetus through the stratigraphic sequence. The latest Devonian negative shift in the terrestrial organic carbon isotope ratio, followed by the early Carboniferous positive trend, was reported by Strauss & Peters-Kottig (2003). These authors attributed the variations to changes in the global carbon cycle (Strauss & Peters-Kottig 2003). The marine and terrestrial δ 13 C org fluctuations were linked via atmospheric carbon dioxide (Hayes et al. 1999;Strauss & Peters-Kottig 2003). A marine δ 13 C org shift from -25‰ to -23‰ was reported by Hayes et al. (1999) for the latest Devonian-earliest Carboniferous transition (360-355 Ma). Increase in the δ 13 C org value of P. parapetus at the late praesulcata-early sulcata conodont zones may be linked to this shift via the food web.
The very low values of δ 13 C org (from -30.4‰ to -25.4‰) in conodont elements from the latest Famennian lagoon facies (Fig. 5, beds 1a-1) may reflect the freshwater inputs of river-transported organic matter of terrestrial origin (δ 13 C org = -29‰ to -27‰). Another possible cause of these extremely low values of δ 13 C org is some temperature decrease in the shallow-water realm. However, that possibility is not proved by Ca/Mg thermometry, suggesting rather high temperatures, about 23-25 °C, in the earliest Tournaisian in this region (Vevel′ et al. 2012).
The most probable causes of intraspecific variations in δ 13 C org observed in P1 elements of P. parapetus through the latest Famennian-Tournaisian are local changes in the source of organic carbon and/or dietary changes of conodont animals.

CONCLUDING REMARKS
A case study from the Kamenka River section indicates that the carbon isotope composition of conodont elements of Polygnathus parapetus during the latest Famennian-Tournaisian had an average value of -25.2‰. Similar values of δ 13 C org are characteristic of recent zooplankton, and a low trophic level of this species is suggested. This supposition is supported by a quite high Sr/Ca value of about 0.005-0.008 in conodont bioapatite. During the latest Famennian (praesulcata conodont Zone) δ 13 C org shifted to more negative values (up to -30.4‰). This shift may be caused by an influx of organic matter of terrestrial origin during the terminal Famennian regression. The carbon isotope composition of organic matter of conodont elements provides information about the position of conodonts in the food web and peculiarities of trophic relations in the Palaeozoic pelagic ecosystems.