Comprehensive zircon thermometry that takes into account zircon saturation temperatures, Ti-in-zircon measurements, and zircon morphologies and microstructures can provide key information on the thermal evolution of a granite batholith. The Variscan South Bohemian batholith (Germany, Austria, and Czech Republic) comprises a series of granitoid units that intruded between ca. 330 and ca. 300 Ma. We categorize the granitic rocks according to their emplacement temperature into very low temperature (T) (VLT; <750 °C), low T (LT; 750–800 °C), medium T (MT; 800–850 °C), high T (HT; 850–900 °C), and ultrahigh T (UHT; >900 °C). The first stage of batholith formation (ca. 330–325 Ma) is characterized by LT to MT melting of mainly metasedimentary sources driven by their isothermal exhumation. In turn, ca. 322 Ma HT and UHT granites in the southern half of the batholith reveal an ephemeral thermal anomaly in the subbatholithic crust, which is presumably linked to a hidden mafic intrusion. The HT and UHT granites are weakly peraluminous, high-K, I-type rocks. Although sharing some features with A-type granites such as high Zr and rare earth element contents, they differ from classical A-type granites in being magnesian, not enriched in Ga over Al, and having high Ba and Sr contents. A ring structure of ca. 317 Ma MT and/or LT plutons is observed around the HT and/or UHT granite complex and interpreted as an aftermath of the hotspot event. This study is an example of how deep-crustal hotspots, presumably caused by mantle magmatism, can significantly enhance the effects of decompressional crustal melting in a post-collisional setting.

Experimental studies show that magmas of granitoid composition may form (and evolve) over a wide temperature (T) range between 700 and 1100 °C (see the data compilation of Gao et al. [2016]). This opposition of “cold” and “hot” granites has been highlighted in the work of Chappell et al. (1998), Miller et al. (2003), Collins et al. (2016, 2021), and Volante et al. (2020). Current granite research is strongly focused on identifying magma sources (e.g., I-type, S-type, mantle versus crust). We suggest here that a categorization of granitic rocks into very low T (VLT; <750 °C), low T (LT; 750–800 °C), medium T (MT; 800–850 °C), high T (HT; 850–900 °C), and ultrahigh T (UHT; >900 °C) is complementary to the classical S-, I-, A-, and M-type granite classification. This approach can be put into practice by using comprehensive zircon thermometry, which takes into account zircon saturation temperatures, Ti-in-zircon measurements, zircon morphologies, and zircon microstructures.

The South Bohemian batholith (also termed the Moldanubian batholith) is one of the largest plutonic bodies of the Variscan orogenic belt in Europe, extending over ~10,000 km2 from northern Austria into the Czech Republic and Germany (Žák et al., 2014). It formed in the middle to late Carboniferous at the end of the Variscan subduction-collision orogeny (Schulmann et al., 2014). The batholith (Fig. 1) intruded into high-grade, mainly metasedimentary country rocks that were diachronously exhumed (Žák et al., 2014).

S-type two-mica granites of the Eisgarn suite are prominent along the northern flank of the batholith and were emplaced between ca. 328 Ma and ca. 325 Ma (Gerdes et al., 2003; Žák et al., 2011). The coarse-grained K-feldspar-phyric Weinsberg granite is dominant in the southern half of the batholith (Fig. 1). It includes I- and S-type subunits but has entirely crustal isotopic signatures (Liew et al., 1989; Vellmer and Wedepohl, 1994; Gerdes, 2001). Finger and Clemens (1995) proposed that the Weinsberg magma formed in the lower crust from greywacke or tonalitic protoliths in connection with fluid-absent biotite + quartz melting. Gerdes et al. (2003) dated different subunits of the Weinsberg granite, referenced as Weinsberg I, II, and III granites in this paper, by means of high-precision zircon and monazite geochronometry to ca. 330 Ma (I-type biotite granite), ca. 326 Ma (S-type biotite-muscovite granite), and ca. 322 Ma (I-type biotite granite), respectively.

Eisgarn- and Weinsberg-type granites are traditionally summarized as the “Older Granites” (Frasl and Finger, 1991). Smaller plutons of mostly fine-grained granites and granodiorites, also with crustal isotope signatures (Liew et al., 1989; Vellmer and Wedepohl, 1994), intrude the Weinsberg granite in the southern half of the batholith. The post-Weinsberg plutons include

  1. An older suite of biotite-rich I-type granites (Migmagranite, Engerwitzdorf granite, Karlstift granite); these were grouped by Frasl and Finger (1991) to the Older Granites because of weak deformation fabrics, and we term them here the “Late Older Granites”. The Late Older Granites were dated by Gerdes et al. (2003) to ca. 322 Ma. They form only small plutons that are spatially and temporally connected to the Weinsberg III unit.

  2. A group of ca. 318–316 Ma plutons (Gerdes et al., 2003) that seemingly form a ring structure around the Weinsberg III unit (Fig. 1). These plutons include cordierite-bearing and two-mica S-type granites (Schärding-Peuerbach granite, Altenberg granite) as well as I-type biotite granites (Mauthausen granite).

  3. Biotite granodiorite bodies with an age of ca. 302 Ma (Freistadt granodiorite).

Frasl and Finger (1991) have summarized (2) and (3) as the “Younger Granites”.

In order to categorize the granite suites of the South Bohemian batholith in terms of emplacement temperatures, we applied comprehensive zircon thermometry that takes into account zircon saturation temperatures, Ti-in-zircon measurements, and zircon morphologies and microstructures. The Ti-in-zircon analyses are based on conventional laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) methods (see details in the Supplemental Material1 and Data Set S1 therein).

The initial temperature (emplacement temperature) of a granitic magma can be estimated based on the whole-rock Zr content and the zircon saturation temperature (TZr) (Watson and Harrison, 1983). However, this method sometimes gives incorrect results due to the presence of restite, zircon accumulation, Zr undersaturation, and other complications (Clemens and Stevens, 2012; Siégel et al., 2018; Barnes et al, 2020).

Titanium-in-zircon thermometry (TIZT) is another tool for constraining the temperature evolution of a granite (Ferry and Watson, 2007). Zircons that precipitate from a granite melt typically display a wide spectrum of TIZT temperatures. In the case of Zr saturation, the initial magma temperature (i.e., the emplacement temperature) is indicated by the highest measured TIZT temperatures (TIZTmax). However, unrecognized zircon antecrysts and xenocrysts may lead to misinterpretations (Siégel et al., 2018; Collins et al., 2021).

A cross-evaluation of TZr and TIZT data (Schiller and Finger, 2019; Volante et al., 2020) reduces the risk of misinterpretation. In addition, morphological and microstructural (cathodoluminescence-based) zircon investigations can help to identify problematic inherited, xenocrystic, and antecrystic zircons in a sample (Data Set S1).

Zirconium undersaturation is a potential problem for zircon thermometry. In Zr-undersaturated granites, TIZTmax and TZr values are both lower than the initial magma temperature. Zr undersaturation is difficult to recognize and has been reported even for granites that carry inherited zircon (Villaros et al., 2009). Warning signals for Zr undersaturation are a suspiciously low whole-rock Zr content when compared to TiO2 or SiO2, and positive or chaotic Zr-SiO2 trends within the suite (Chappell et al., 1998). If inherited zircons dominate the Zr budget of a granite, this can be an indication for a Zr-undersaturated melt phase as well (disequilibrium melting). Clemens and Stevens (2012) pointed out that zircon entrainment may remain unrecognized if the grains were subsequently dissolved in the ascending (decompressed) magma. TZr and TIZTmax values may still give a reasonable temperature estimate in that case, depending on the available Zr and the amount of entrained zircon.

Figure 2 shows a compilation of TZr data for granite samples from the South Bohemian batholith, calculated from published geochemical analyses (for data sources, see the Supplemental Material). In addition, TIZT data are presented for 15 selected (mainly Zr-rich) samples in combination with a morphological and cathodoluminescence-based zircon investigation (Table1). An important result is that TIZT and TZr gave sympathetic T estimates in all 15 cases (Fig. 2). Deviations between TIZTmax and TZr were 40 °C at most, which is less than the calibration errors of the methods. The T estimates are also consistent with other petrological constraints; e.g., with fusion experiments on Eisgarn granite samples (René et al., 2008) and with the presence of orthopyroxene in some Weinsberg III rocks (Finger and Clemens, 1995). We thus conclude that zircon thermometry is, in general, a proper means to categorize granites according to their emplacement temperature.

Appreciable problems with xenocrystic or inherited zircons were encountered in only one sample (sample JZ 262; Table1): in this relatively Zr-poor sample, more than half of the zircons showed pre-granitic U-Pb ages, which implies that the melt phase was Zr undersaturated. The TZr value of 807 °C (Table1) thus must be considered meaningless. The TIZTmax value of 831 °C calculated from the autocrystic zircons can be interpreted as a minimum emplacement temperature.

Collins et al. (2021) have recently proposed that most crustal granites are Zr saturated and that TZr values commonly reflect initial magma temperatures. Our data from the South Bohemian batholith seem to confirm this view, but it must be emphasized that our work was strongly focused on Zr-rich samples (see Fig. S3).

Assuming that most temperatures estimates from Table 1 and Figure 2 are approximate emplacement temperatures, the following scenario emerges: the two small occurrences of ca. 330 Ma Weinsberg I granite (Fig. 1) represent early HT intrusions, while the voluminous ca. 328–325 Ma S-type granite domains of the batholith (Eisgarn and Weinsberg II granites) are mainly LT to MT magmas. The ca. 322 Ma Weinsberg III granite complex in the southwestern sector of the batholith yields significantly higher temperatures, between 850 and 920 °C, with some samples classifying as UHT. Samples of Late Older Granites (with high Zr contents >550 ppm) gave temperatures of 900–950 °C and represent UHT magmas. With one exception (Freistadt granodiorite; see Table 1), TIZT data are presently unavailable for the Younger Granites (SPG, ABG, MHG in Fig. 2). However, their whole-rock Zr contents of <200 ppm imply a LT to MT origin.

Early Batholith Evolution (ca. 330–325 Ma)

Based on pressure-temperature-time data from country-rock xenoliths and field observations, Žák et al. (2011) argued that the granite-forming processes of ca. 330–325 Ma were mainly exhumation driven. The MT nature of many Eisgarn and Weinsberg II samples (Fig. 2) seems consistent with a fluid-absent decompression melting model as proposed by Žák et al. (2011). However, there are also low-Zr samples in this group, which (if not just products of differentiation or Zr undersaturated) would leave room for water-fluxed crustal melting models (Bea et al., 2021; Collins et al., 2021) as well. Small thermal anomalies may have existed at the eastern batholith flank, producing the ca. 330 Ma HT Weinsberg I bodies (Fig. 1).

Hotspot Event at ca. 322 Ma

The novel and most intriguing result of this study is that the investigated ca. 322 Ma granite units (Weinsberg III granite and Late Older Granites) involve much hotter, partly UHT magma with formation temperatures of ~900–950 °C. These UHT granites are encountered only in a relatively small area, located off center of the Weinsberg III unit (dashed red line in Fig.1). This delineates a thermal anomaly of limited size, most likely caused by a deep mafic intrusion (Petford and Gallagher, 2001; Annen et al., 2008). Considering the spatial and temporal connections, the entire assembly of the Weinsberg III granite and Late Older Granites is most likely a product of this “hotspot” event. Whether the Weinsberg III granite and Late Older Granites varieties with lower Zr contents (classified as MT or LT based on TZr values) represent differentiated melts from an UHT parental magma or pristine LT-MT magmas that feature a thermal gradient around the hotspot remains to be clarified in a future study. A more detailed zircon investigation (including TIZT) must be carried out for these rocks to control for the significance of the TZr data.

Later Batholith Evolution

The ca. 322 Ma granites (Weinsberg III granite and Late Older Granites) are surrounded by a conspicuous ring structure of ca. 318–316 Ma plutons (Fig. 1). Obviously, these have to be considered as an aftermath of the hotspot event. Sharp intrusive contacts with the Older Granites suggest that these plutons postdate a phase of crustal uplift (Frasl and Finger, 1991). The magmas may thus have formed through decompressional crustal melting. The TZr data (if reliable; confirmation through TIZT is necessary; see Fig. 2) suggest that crustal anatexis at ca. 316–318 Ma reached MT conditions. This is confirmed by a thermobarometric study on ca. 316 Ma migmatites from the southern part of the batholith (Sorger et al., 2018). Thus, biotite dehydration melting is a possible granite-forming scenario. However, models of water-fluxed crustal melting (Bea et al., 2021; Collins et al., 2021) are also within the realm of possibility and would be especially attractive for the low-Zr Altenberg granite (Fig. 2), which has been described as undifferentiated, pristine magma based on petrographic criteria (Frasl and Finger, 1991).

The Freistadt granodiorite suite (ca. 302 Ma) represents an independent plutonic event (Gerdes et al., 2003). These relatively K-poor rocks (Fig. S3 and Data Set S1) have a Zr content of 100–200 ppm and classify as LT according to both TZr and TIZT data. Thus, the water content of these magmas was probably relatively high (Collins et al., 2021).

Most granites worldwide are LT, MT, or HT, while UHT granites (T > 900 °C) are relatively rare. The South Bohemian UHT granites share some geochemical features with A-type granites (Collins et al., 1982; Eby, 1990), such as high Zr and rare earth element contents (Fig. 3; Fig. S3). However, they differ significantly from classical A-type granites in three aspects: they have lower Fe/Mg ratios (Fig. 3), higher Ba and Sr contents, and a source that was likely not melt depleted (no high Ga/Al ratios, no depletion of melt-affine elements like Rb, Cs, or U).

The South Bohemian UHT granites are moderately felsic, fine- to coarse-grained biotite granites with 60–73 wt% SiO2, high K2O (4–6 wt%), and FeOt/(FeOt + MgO) between 0.65 and 0.85. Their alumina saturation index, A/CNK, values are mostly between 1.0 and 1.1 (see the data in Data Set S1). Thus, they are most likely derived from a biotite-rich, intermediate (I-type) crustal source.

The recognition of UHT granites in the South Bohemian batholith comes amid a general debate about the role of mantle events and transcrustal processes in the Variscan orogeny. For instance, the widespread low pressure–high temperature metamorphism in the core zone of the orogen is readily explained in terms of delamination of mantle lithosphere (Henk et al., 2000; Sorger et al., 2018) or put into the context of a mantle plume (Franke, 2014).

The simplest explanation for the newly discovered “thermal eye” in the South Bohemian batholith is certainly mantle magmatism and an associated heat anomaly (Clemens and Stevens, 2021). Our work thus corroborates models that invoke a significant late-stage mantle heat addition into the Variscan crust, though only rare signs of mantle magmatism (i.e., mafic intrusions) are observed at the present-day exposure level. However, our study implies that larger bodies of mafic magma ponded at depth (Vielzeuf et al., 1990) and may also have supplied extra water for crustal melting (Collins et al., 2021). Small diorite stocks and lamprophyre dike swarms were mapped in many places throughout the Variscan orogen (Žák et al., 2014) and could be signals for deep-seated mafic bodies. Tracing the spatial and temporal distribution of deep hot zones may become an important new clue for geodynamic interpretations of the Variscan orogen.

This study was supported by the Austrian Science Fund (project I1993 N29), through the Czech-Austrian mobility project 8J20AT004, and by Charles University (Prague, Czech Republic) through projects PROGRES Q45 and the Center for Geosphere Dynamics (project UNCE/SCI/006). We gratefully acknowledge the constructive contributions of Gary Stevens, Fawna Korhonen, and an anonymous reviewer during peer review. Marc Norman is thanked for editorial handling.

1Supplemental Material. Zircon thermometry protocol and geochemical data. Please visit https://doi.org/10.1130/GEOL.S.18737441 to access the supplemental material, and contact editing@geosociety.org with any questions.
Gold Open Access: This paper is published under the terms of the CC-BY license.