The relative proportions of the cement and clasts were measured by classical point counting on thin sections (30 × 45 mm²; 4000 points/section; Table 1). The uncertainty on modes can reach up to 2 vol.%, as estimated from standard deviations from repeated modal analyses on granitic rocks at Géosciences Rennes.

Chemical compositions were measured by the Service d’Analyse des Roches et des Minéraux (SARM, CRPG-CNRS, Nancy, France). Analyses were performed by ICP-AES for major elements and ICP-MS for trace elements. Uncertainties and detection limits are provided in Table 2.

Stable isotope analyses were performed at the stable isotope laboratory at Géosciences Rennes. About 10 mg of powder was reacted with anhydrous H₃PO₄ at 50.0 °C overnight. Isotopic compositions were measured on a VG-SIRA 10 triple collector mass spectrometer. The results are quoted in Table 1 using the δ notation with respect to SMOW for δ¹⁸O and PDB for δ¹³C. Isotopic measurements were normalized using repeated analyses of an “in-house” standard (Prolabo Rennes) and of the NBS 19 international limestone standard. Taking into account (1) the analysis of five duplicates and (2) the uncertainty on measurements of the standards, we estimate the total uncertainty on isotopic compositions at ±0.1‰ and ±0.15‰, for δ¹³C and δ¹⁸O, respectively.

Nd isotope analyses were performed on 100 mg of rock powders using a 7-collectors Finningan MAT-262 mass spectrometer at Géosciences Rennes (Table 2). Powders were dissolved twice with a mixture of 6 N HCl acid in a sealed Savilex beaker on a hot plate during 3 to 4 days. They were then dried and taken up with concentrated 2.5 N HCl acid. During the analytical session, measurements of the AMES Nd standard gave a mean ¹⁴³Nd/¹⁴⁴Nd ratio of 0.511945. Blanks values for Nd were lower than 300 pg and no correction was made to the measured isotopic ratios.

5.1.1 Description

5.1.2 Modal composition

The Comblanchien limestone and the Granité exhibit rather homogeneous δ¹⁸O (25.4–27.7‰) and δ¹³C (1.24–2.25‰) values, which compare well with unaltered marine Jurassic limestones (Veizer et al., 1999) (Table 1, Fig. 4). Some samples from the SETP quarry have lower δ¹⁸O values down to 24.6‰ and δ¹³C values down to –1.24‰, which are likely signs of some secondary alteration.

The VATC sample has a δ¹⁸O value comparable to most of the Comblanchien samples (26.5‰) but shows a negative δ¹³C value (–0.40‰) that thus compares well with some SETP samples. Samples A, B and F have a comparable δ¹⁸O value (near 26.5‰) and a somewhat variable δ¹³C value between –1.55 and 0.41‰.

5.3.1 Major and trace elements

The chemical composition of the Comblanchien limestone reflects the nearly total absence of terrigenous mineral phases: excluding the major oxides classically entering in carbonate structural formulas (CaO, MgO and LOI assumed as mainly CO₂), the sum of oxides is less than 1 wt.% (Table 2). Consistently, characteristics of trace elements are typical of marine limestones: Sr content is about 200 ppm, REE patterns are not fractionated relative to the upper continental crust and exhibit a large Ce negative anomaly (Ce/Ce* about 0.49) (Fig. 5).

5.3.2 Nd isotopes

The Comblanchien limestone samples define a rather homogeneous population. Nevertheless, small heterogeneities occur in the data. Petrographically, the most prominent feature is the presence or the absence of coated grains in the samples. Actually, coated grains were only observed in the intermediate layers of the exploited strata and can thus be considered as a good marker of these layers. Geochemically, the stable isotope compositions of some samples from the SETP quarry are distinguished from the others by lower δ¹³C and δ¹⁸O values. In particular, the negative δ¹³C value, which corresponds to the lowest δ¹⁸O value (sample SETP6d), has to be related to a secondary effect because no negative value is to date documented in sedimentary elements from Bathonian sequences (see fig. 9 E in Brigaud et al., 2009). Interestingly, there is no correlation between the petrographic peculiarity (presence of coated grains) and the isotopic anomaly (low δ¹³C and δ¹⁸O values). These two features should thus relate to two independent processes that did not significantly alter the whole rock chemical compositions or the Nd isotope compositions. Both are indeed invariant throughout the Comblanchien population. The Sr content of sample SETP6d lower than that of the other SETP samples may reflect some alteration. In carbonate rocks, as the Comblanchien limestone, the main causes of heterogeneities are lateral and vertical variations of facies, variations in deposition environment, variations of the terrigenous input and of the diagenetic overprint. The latter may be related to heterogeneities in fluid circulations during compaction and re-crystallization controlled by variations in porosity and permeability. For the Comblanchien limestone, it is reasonable to exclude effects of recent weathering to explain the differences of some SETP samples because the rocks of lesser quality are excluded from the commercial channels. Likewise, local isotopic perturbations due to climatic effects during deposition are likely negligible because of the relatively short time span recorded by the studied sequence.

The specific isotopic signature of some SETP samples is not restricted to given layers in the sequence, but is rather characteristic of the SETP quarry as a whole. As a consequence, the alteration lowering both δ¹³C and δ¹⁸O values of some SETP samples should relate to the location of the SETP quarry on a map-scale fault (Fig. 1B). A possible explanation for the isotopic alteration of the SETP samples is therefore preferential fluid flow along the fault. Such fluid-rock interactions might have occurred either as early as the diagenetic history of the limestones or later during the Cenozoic tectonic activity of the area (Rocher et al., 2003), depending on the age of the fault. Both contexts are indeed capable of lowering the isotopic signatures through carbonate-water interactions.

There are some implications of the above observations in the context of building stones characterization. Because coated grains are restricted to intermediate layers, they represent good markers of these layers. By extension, as they locate within intermediate layers, it becomes possible to assign a precise stratigraphic origin for the coated grain-bearing Comblanchien limestone. In addition, because stable isotope anomalies are restricted to the SETP samples, it is possible to guarantee that a given sample with such anomaly (taken for example from a recent building) comes from that specific quarry. Finally, the so-called Granité, which forms the top of the sequences worked in Rocamat, Pierre Bourguignonnes and SETP quarries, appears as a strict equivalent of the Comblanchien limestone. So, even if the Granité was sold as a building stone, which is actually not the case as it is used to make aggregates, one would have had difficulties in discriminating between the Granité and the Comblanchien limestone with the tools used here. This illustrates the difficulty to precisely define commercial building stones.

Our study is restricted to the four quarries near the Comblanchien village. It is therefore not possible to ascertain that no other quarry, in the Paris Basin or the Jura, mining the Comblanchien Formation, could or not provide limestones with the same petrographical, geochemical and paleontological characteristics as the present Comblanchien limestone.

The VATC sample displays similarities with some Comblanchien limestones, both petrographically and geochemically. Most of the qualitative and quantitative characteristics of the VATC sample can be recognized in some Comblanchien limestones. Actually, as foraminifera Meyendorffina (Kilianina) bathonica and Orbitammina elliptica (Delance, 1964) are absent from VATC, the faunal content could constitute an efficient tool of discrimination. However, the determination of foraminifer species can only be made by specialists and dating using biozones requires strong expertise. So the present strategy constitutes an alternative and quite easily accessible method of discrimination.

Only the few Comblanchien samples from the intermediate coated grain-bearing layers with a sparitic cement compare to VATC. From petrographic observations alone, it is thus impossible to discriminate between the coated grain-bearing Comblanchien and VATC limestones. Conversely, coated grain-free building stones from Comblanchien cannot be confused with VATC type ones, and petrography alone seems there discriminative.

VATC sample has O and C isotope compositions in the range defined by the Comblanchien limestone. In details, a single Comblanchien sample (SETP6d) has a δ¹³C lower than that of the VATC sample. Considering the large number of analyses of the Comblanchien limestone, one could say that the δ¹³ C tool is a rather good marker, but not absolute since a negative δ¹³C value does not guarantee the VATC origin. To circumvent this uncertainty, one could follow an isotopic reasoning in which the negative δ¹³C value of sample SETP6d is accompanied by a lower δ¹⁸O relative to the other Comblanchien samples, this peculiar δ¹⁸O being absent from the VATC sample. One could then argue that the low δ¹³C values of SETP6d and VATC samples reflect distinct processes, a fluid-related one (as discussed above) and a primary signature one, respectively.

The efficient discrimination between the Comblanchien limestone and the VATC sample comes from a combination of analytical methods. VATC is coated grain-bearing, as do some Comblanchien limestones (11 samples, which coated grain contents are highlighted in bold in Table 1). The latter have δ¹³C values between +1.75 and +2.25‰ (bold values in Table 1), well above the VATC value (−0.4‰). In addition, the chemical compositions of eight Comblanchien samples (Table 2) differ from that of the VATC sample, especially for MgO and REE (see also Fig. 5). Thus, we infer that the combination of the coated grains and δ¹³C tools is efficient and sufficient to discriminate between Comblanchien and VATC.

Samples A, B and F were collected on a building site near Paris in early 2010. On this site, building stones were assumed to come from the Comblanchien area. Our previous results provide some clues confirming that these samples did not origin from the Comblanchien area. Indeed, coated grain-bearing samples A, B and F have δ¹³C values between +0.41 and −1.55‰, which are different from the Comblanchien coated grain-bearing limestone isotopic characteristics.

Finally, from our results, the question to know if samples A, B and F come from Portugal as does sample VATC can also be addressed. At this stage, it is somewhat hazardous to conclude since the intrinsic petrographical and geochemical variability of VATC cannot be defined from a single sample. Nevertheless, sample VATC and samples A, B and F share many petrographical (sparitic cement, coated grains as clasts) and geochemical characteristics (low to negative δ¹³C value, identical δ¹⁸O value, low MgO, Ba and Sr content, high REE and Y content), so that all these samples likely share the same Portuguese provenance.