Servicenavigation

Project Details

Back

Projekt Print View

Generation and modification of large-scale post-tectonic batholiths (Donkerhuk granite; Namibia): Incremental growth vs in-situ differentiation

Subject Area Mineralogy, Petrology and Geochemistry
Term from 2016 to 2020
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 339361994
 
Final Report Year 2021

Final Report Abstract

The present study has shown that the orogen-wide Donkerhoek batholith (Damara orogen, Namibia) with a spatial extent of ~5000 km2 is derived from various crust-derived sources. Most rock types range are true granites to leucogranites, however, isolated outcrops of more mafic granodiorites also exist. These granodiorites are limited in their spatial extend and cannot serve as the sources of the more common granites. The overall granitic composition of the Donkerhoek batholith indicate that it cannot be connected with a common subduction zone setting because of a lack of mafic and intermediate rock types. Whereas Clemens et al. (2017) suggested that almost all granites from the Donkerhoek were generated by partial melting of metasedimentary rocks and some basement rocks, the present study implies that aged meta-igneous sources are equally important. Clemens et al. (2017), obviously impressed by the existence of hundreds to thousand sheet-like intrusions, suggested that most of the chemical variability is related to heterogeneities of the inferred source rocks, whereas results of Schwark et al. (2018), Jung and Hauff (2020) and Aspiotis et al. (2021) showed that processes of crystal fractionation with and without assimilation of lower crustal basement rocks have substantially modified the isotope composition of the granites. New geochronological data have shown that the main mass of the Donkerhoek batholith intruded between ~541 Ma (Schwark et al. 2018) and ~520 Ma (Clemens et al. 2017) in which the older data were already known since the work of Jung and Mezger (2001) on a Donkerhoek granite-related injection migmatite complex. Unfortunately, these results were ignored over the years which hampered a proper interpretation of the temporal evolution of the Donkerhoek granite since then. With a spatial extent of ~5000 km2 and an intrusion interval of ~20 Ma, emplacement of the Donkerhoek magmas occurred probably two times faster than the well-investigated Tuolumne intrusive suite (1200 km2 in 10 Ma; Coleman et al., 2004) although the whole Sierra Nevada batholith with a comparable spatial extent of ~4000 km2 intruded only within 12 Ma (Coleman and Glazner, 1994). The Sierra Nevada batholith represents a typical syn-collisional subduction-related igneous complex and shows rapid crustal growth. The Donkerhoek intrusive event lasted longer implying that processes involving intracrustal reprocessing need more time to yield comparable igneous masses. These results have important implications for the interpretation of the durations of magma-producing events and whether magma production and emplacement were semicontinuous or strongly pulsed in age distribution. The new geochronological results from the Donkerhoek batholith place its origin into the syn-collisional stage that shaped the Southern Damara orogen implying that the batholith as a whole is clearly not late- to post-collisional as previously suggested (Miller, 2008). For the Donkerhoek batholith magma production and emplacement seems to be rather continuous then pulsed although this view contradicts with the suggestions of Clemens et al. (2017) who interpreted the sheet-like intrusion in places as evidence for a pulsed magmatism. Clearly, both modes of emplacement operated in tandem in which the sheet-like intrusions seem to be a local phenomenon. Clemens et al. (2017) suggested that the sheet-like intrusions were constructed from numerous discrete magma batches. They concluded that the chemical composition of these sheets is unrelated to synemplacement differentiation, but rather reflects deeper-seated processes, e.g. multiple, chemically distinct sources that were tapped at various stages of pluton assembly. However, field and petrographic evidence indicated that, in a given outcrop, sheets have rather similar 22 compositions which renders the opinions of Clemens et al. somewhat speculative. On a whole, the Donkerhoek batholith is heterogeneous but this heterogeneity may also result from AFC processes, as demonstrated by Schwark et al. (2018), Jung and Hauff (2020) and Aspiotis et al. (2021), superimposed on pre-existing source heterogeneities. Large-scale granitic batholiths are commonly heterogeneous in their geochemistry and in common syn-collisional settings, rock associations may range from gabbro to granite (Sierra Nevada batholith; Miller and Glazner, 1994). Since the continental crust is too felsic to yield primary mafic melts, such rock associations imply interaction of lithospheric mantle sources with crustal rocks, both at the site of melting and during ascent. The Donkerhoek batholith do not show this petrographic variability and thus require distinct sources. The chemical and isotope data obtained in this study in comparison with the results presented by Clemens et al. (2017) are, in part, in conflict. Clemens et al. (2017) argue that the mostly peraluminous composition testify for a metasedimentary origin of the Donkerhoek magmas, however, evolved melts from meta-igneous source may also be peraluminous (Chappell et al., 2012). Nd-Sr isotope data obtained on samples from the Donkerhoek batholith are much more heterogeneous that those from the Kuiseb schists in which the Donkerhoek granite intruded (Schwark et al. 2018) and particularly extent to more unradiogenic 87Sr/86Sr compositions. Many of them overlap with compositions of the Proterozoic basement that underlies the Southern Damara orogen. Later, Clemens et al. (2017) argue for a mixture of source rocks now including basement rocks because their two-stage Nd model ages of the Donkerhoek granites are older than those of the Kuiseb schists at a given 87Sr/86Sr ratio (1.48-2.01 Ga; Jung et al. unpublished). The reason for using two-stage Nd model ages was not justified because 147Sm/144Nd ratios seem undisturbed and cluster around the common crustal value of 0.10-0.12. Interestingly, when using single-stage Nd model ages, there is complete overlap in Nd model ages of the Donkerhoek granite and the Kuiseb schist implying derivation of both rock units (Donkerhoek granite and Kuiseb schists) from a similar source. In conclusion, the sources of the Donkerhoek magmas remain elusive and may contain distinct components of variable ages. New analyses from the alkaline rocks at Otjimbingwe and Okamutambo show that the rocks suites consist of ne-normative to qtz-normative monzosyenite, monzonite to syenite that differentiated from a mantle-derived alkaline melt. Recalculated Sr-Nd-Pb isotope data back to a hypothetical parental melt with 10 wt% MgO suggests derivation from a moderately enriched lithospheric upper mantle (87Sr/86Sr: 0.705, e Nd: -2, d18O: 6‰, 206Pb/204Pb: 19.40, 207 Pb/204Pb: 15.82). More evolved samples show increasing 87Sr/86Sr ratios but less radiogenic e Nd values and Pb isotopes with decreasing MgO, indicating assimilation of ca. 10% Archaean to Proterozoic local lower. New age constraints (Pontow et al., 2021) show that intrusion of mantle-derived melts in this segment of the belt occurred earlier that the oldest granites from the Donkerhoek granite (541±4 Ma; Schwark et al., 2018) and that the mantle melts may have provided additional heat to melt the overlying metasedimentary (Clemens et al., 2017) and meta-igneous (Schwark et al., 2018; Jung and Hauff, 2020, Aspiotis et al. 2021) crust of the southern central and southern Damara orogen. These results imply substantial mantle-crust interaction during the Pan-African orogeny.

Publications

 
 

Additional Information

© 2024 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung