Nova Terra 52

metaigneous complexes rocks (HREE = 128.58 and 116.14, respec- tively). There is no evidence of presence of Ti-rich phases in these rocks to explain this extraordinary depletion in the HFSE. However, presence of these phases in the residue causing fractionation of these elements cannot be ruled out. Normalized-REE ratios have also been used traditionally in igneous rocks generated in the active margin setting as proxies to constrain the magma source. According to the Yb N vs. (La/Yb) N and Sr/Y vs. Y ratios (Martin, 1986; MacPherson et al., 2006; Supplementary Table S2 and Fig. 6a and b) the rocks belonging to Don Álvaro and Valle Real metaigneous complexes plot in the field of ‘‘normal” volcanic arc granitoids (avg. La N /Yb N of 6.13 and 6.11, and avg. Sr/Y of 14.73 and 14.67, respectively), significantly apart from the samples belonging to San Andrés and Valverde metaigneous complexes (avg. La N /Yb N = 30.55 and 14.94, and avg. Sr/Y of 38.99 and 39.43, respectively), which plot within the field of rocks with ada- kitic signature. The average (La/Yb) N and Sr/Y are positively corre- lated with dm (dm = crustal thickness quantified after Profeta et al., 2015; Chiaradia, 2015) values referred to these ratios. Its use requires an age precision of < 5 m.y., which prevents it from being applied to the Montemolín metabasites. San Andrés and Val- verde samples show the highest values of dm (La/Yb) N and Sr/Y (45.47 and 40.62; respectively and 52.31 and 52.16, respectively; Supplementary Table S2). Don Álvaro and Valle Real samples share an average value of dm Sr/Y = 24 with an average of dm (La/ Yb) N = 19.62 and 23.52, respectively (Supplementary Table S2). According to Th/Yb - Nb/Yb diagram (Fig. 6c; Pearce, 1982, 2008), the metabasite samples analysed have high Th/Yb contents and plot displaced from the MORB-OIB array, showing a clear calc- alkaline arc affinity. High Th contents from MORB composition requires sediment melting (Johnson and Plank, 2000) or subduction-derived fluids, which agrees with their location over the MORB-OIB array. Others conservative and non-conservative trace element ratios (Supplementary Fig. S3a) also track the role of sediment melts and slab-derived fluids in these rocks. The calc-alkaline character of the metabasites of Montemolín Formation is also reflected in the Ti vs. V diagram (Fig. 7a, Shervais, 1982) showing an average Ti/V ratio of 26.62, close to the values associated with mafic magmas generated in active sub- duction zones (Ti/V = 20). The distribution of these rocks coincides with the distribution shown by the mafic calc-alkaline rocks of the Guevgueli complex (Saccani et al., 2008), which have been also used by Pearce (2014) as an example of mafic rocks linked with subduction-initiation settings. The increase in the crustal contribution or sediment melt is also seen in the metagranitic complexes, mostly in the rocks belonging to the San Andrés Metaigneous Complex (avg. Th = 7.22 ppm). This agrees with the average U/Th ratio (Sun and Sterns, 2001) close to the mantle range values (0.26–0.39 ppm, Sun and McDonough, 1989) shown by all metaigneous groups (Supplementary Table S2), with the exception of the most depleted values in San Andrés Metaigneous Complex. On the other hand, Pearce et al. (1984) and Pearce (1996b) use Rb vs. (Y + Nb), as well as Rb vs. (Yb + Ta) to discriminate the four main tectonic environments of granitoid formation. In both diagrams (Fig. 7b and 7c), all samples plot within the volcanic arc granite (VAG) field. Areas representing modern Alaskan and Chilean volcanic arc types (Pearce, 1996b) have been plotted into Rb vs. (Y + Nb) diagram. Don Álvaro and Valle Real samples overlap the Alaskan arc type, whereas the Val- verde and a part of San Andrés samples plot near the Chilean field. Although the Rb contents are similar in all groups, the values in the Y + Nb and Ta + Yb ratios continue to reflect a clear separation between the samples belonging to the felsic metaigneous com- plexes and the intermediate-mafic ones. In order to unify previous constraints in a single common setting, the tectono-magmatic Th- Hf-Nb (Fig. 7d; Wood, 1980) diagram has been used. According to this classification, all studied samples are uniformly represented within the field of calc-alkaline rocks pointing to a common sub- duction setting for the magma source. 5.2. U-Pb zircon ages 5.2.1. Metatonalite of the Don Álvaro metaigneous complex (DAMIC metatonalite) CL-images (Fig. 8) show that most of the zircon grains are frag- ments, and some are subidiomorphic crystals. Zircons are charac- terised by oscillatory or sector zoning with no evidence of metamorphic overgrowth. Some zircon grains contain inclusions. U-content ranges from 252 to 1578 ppm and the Th/U ratio is fairly homogeneous, ranging from 0.32 to 0.51. A total of 61 zircons were analyzed (one analysis per grain). The results are shown in Supple- mentary Table S3.1. 36 analyses with discordance > 5% were dis- carded (see above). The remaining 25 concordant analyses (Fig. 10a) have ages ranging from 580 to 517 Ma (Ediacaran to Cambrian). The majority of concordant ages (24/25) are Ediacaran, and a few are Early Cambrian, with clusters at 570 ± 4 Ma (n = 8; MSWD = 6.2) and 541 ± 3 Ma (n = 16; MSWD = 5.2). The latter is considered to represent the best estimate for the crystallization age (Fig. 10b) of this rock. A single younger analysis (spot located close to an inclusion) has a concordant age at 517 ± 11 Ma (DAMIC 222, Supplementary Table S3.1). 5.2.2. Felsic gneiss of the San Andrés metaigneous complex (SAMIC felsic gneiss) The zircon population is mainly made up of small zircon grains which appear as subidiomorphic crystals or broken grains. Many grains show cores with sector zoning bordered by irregular, thin rims (Fig. 8). A total of 150 analyses were performed on 126 zircon grains. 16 of those analyses have been directly discarded due to analytical errors. The remaining 134 analyses have U-content rang- ing from 80 to 1667 ppm (Supplementary Table S3.2), and the Th/U ratio ranges from 0.28 to 0.5, except for the oldest two analyses, whose Th/U is close to the unit. 43 of these analyses are concordant (Fig. 10c) and 91 have discordance > 5% or an anomalous common lead composition. This sample shows the oldest ages, ranging from Paleoproterozoic to Ediacaran (Supplementary Table S3.2). The two oldest grains analysed have Paleoproterozoic 1632 ± 30 Ma (SAMIC 168) and Tonian-Cryogenian 727 ± 44 Ma ages, respectively. The remaining 41 concordant analyses have Cryogenian to Ediacaran ages, with clusters at 645 ± 8 Ma (n = 4; MSWD = 0.73), 625 ± 5 Ma (n = 9; MSWD = 0.068) and 602 ± 3 Ma (n = 25; MSWD = 6.1). The Concordia age (Fig. 10d) obtained from the youngest age cluster is considered to be the best estimate for the crystallisation of the granitic protolith. The youngest three analy- ses (582 ± 13 Ma (SAMIC 139), 578 ± 12 Ma (SAMIC 84), 549 ± 11 Ma (SAMIC 122), falling outside the uncertainty bound- aries of the youngest age cluster cannot be considered as a popula- tion in themselves and their significance is unclear. 5.2.3. Metagranodiorite of the Valle Real metaigneous complex (VRMIC metagranodiorite) Zircon grains appear as idiomorphic to subidiomorphic thin and elongated prisms as well as other more rounded and stubby crys- tals. Broken grains are scarce. Most of the grains display complex internal structures, combined with oscillatory and sector-zoned cores variably resorbed, surrounded in some cases by a discontin- uous non-luminescent to luminescent rims. Some zircon crystals contain inclusions and inherited cores (Fig. 8). U-content ranges from71 to 543 ppm and the Th/U ratios range from 0.3 to 0.5, only three analyses have a lower (Anal. #64; 0.17) and higher (#164; 0.83 and #183; 0.69) ratio than the average. A total of 200 analyses were performed on 163 zircon grains and the results are shown in E. Rojo-Pérez, U. Linnemann, M. Hofmann et al. Gondwana Research 109 (2022) 89–112 The Ediacaran arc section