We present global and regional correlations between whole-rock values of Sr/Y and La/Yb and crustal thickness for intermediate rocks from modern subduction-related magmatic arcs formed around the Pacific. These correlations bolster earlier ideas that various geochemical parameters can be used to track changes of crustal thickness through time in ancient subduction systems. Inferred crustal thicknesses using our proposed empirical fits are consistent with independent geologic constraints for the Cenozoic evolution of the central Andes, as well as various Mesozoic magmatic arc segments currently exposed in the Coast Mountains, British Columbia and the Sierra Nevada and Mojave-Transverse Range regions of California. We propose that these geochemical parameters can be used, when averaged over the typical lifetimes and spatial footprints of composite volcanoes and their intrusive equivalents to infer crustal thickness changes over time in ancient orogens.
Magmatism at subduction zones originates in the asthenosphere above the downgoing slab and is a result of partial melting of the mantle in the presence of aqueous fluids released from the slab1. These processes produce primary basaltic melts, which then ascend into the overlying plate. Global geochemical trends of mafic rocks found in arcs are influenced in part by heterogeneities in these primary magmas2 and are complemented by subsequent fractionation of these liquids in the crust3, as well as several assimilation mechanisms of the pre-existing upper plate crust4. The end result of arc magmatism is a crustal column dominated by magmatic products5,6 and an average upper crustal composition that is variable in composition but on average is andesitic, regardless of whether the subduction is intra-oceanic or if the upper plate is continental. Island arcs tend to be more mafic than continental arcs in most places7. Certain major element oxides of mafic arc rocks (e.g. NaO, CaO) correlate with crustal thickness at global scales8,9 as do some trace elemental ratios10,11 (La/Yb or Ce/Y), likely indicating their average depth of melting in the mantle and thus indirectly placing limits on crustal thickness. Intermediate rock geochemical parameters involving major elements or their ratios, such as silica or K2O/Na2O may also correlate with crustal thickness at global scales8,12,13, however most such correlations break down where crustal thicknesses exceed 45 km. Perhaps the only major elemental oxide that correlates with crustal thickness at global scales is MnO, but the magnitude of this oxide’s variation in calk-alkaline rocks is small and the correlation is weak. In addition, there are hypothesized trace elemental parameters in intermediate rocks that could be sensitive to crustal thickness in arcs at regional scales (e.g. the slope of REEs14, or Sr/Y15).
Chapman et al.16 showed that tonalite (andesite) and granodiorite (dacite) whole-rock Sr/Y varies regionally with crustal thickness in the western North American Cordilleran interior. As an extension of that work, here we show that two trace elemental ratios, Sr/Y and La/Yb can be used as crustal thickness proxies in low magnesium intermediate calc-alkaline rocks (55–68 wt% SiO2) covering the compositional range of andesites and dacites and their intrusive equivalents (hereafter referred to as intermediate rocks), when larger global and regional datasets are averaged out. We accomplish this first by using an existing large global database for geochemistry of geologically young, Pliocene to modern arc rocks and knowledge of global crustal thickness, followed by comparison of the global to a regional dataset, the southern Andean zone, where crustal thickness varies significantly along the strike of the volcanic arc front17. We then test our global and regional empirical correlations in a few ancient subduction margins where independent geologic constraints exist on variations in crustal thickness over time.
Sr/Y and La/Yb are commonly used in petrology to qualitatively infer depths of magmatic diversification. They can distinguish deep from shallow fractionation processes due to differences in the partition coefficient between these elements in various residual phases and intermediate melt11,14,18. During partial melting of lower crustal igneous rocks such as gabbros and their metamorphic equivalents or during magmatic fractionation of mantle-derived mafic magmas, Sr is preferentially incorporated into the residue at low pressures (<~1.0 GPa) where it strongly partitions into plagioclase. However, at higher pressures (>1.2 GPa), where plagioclase is less abundant and with increasing pressure eventually becomes unstable, Sr preferentially enters the liquid phase. On the other hand, Y is incompatible at low pressures, but readily partitions into garnet at high-pressure. The presence of amphibole, a mineral that can both be consumed and generated during the formation of a residual lower part of a batholith19,20,21 will also elevate the Sr/Y ratios (Fig. 1), although separating amphibole from garnet effects using these elements alone has proven to be difficult. As a result, Sr/Y may act as a pseudo-barometer reflecting the average depth at which magmatic diversification occurred. A larger Sr/Y ratio in the volcanic and batholithic parts of an arc signifies a greater pressure or depth. Predicted values for this ratio are shown in Fig. 1, based on typical concentrations in residual sub-arc rocks22,23 and known residue-intermediate melt partition coefficients from the GERM (Geochemical Earth Reference Model) online database of partition coefficients between various minerals in equilibrium with liquids (http://earthref.org/KDD/). We use averages of partition coefficients for dacitic and andesitic liquids referenced on that online database.
A similar argument can be made for the ratios of light (L) to heavy (H) rare earth elements (REE); we use the La/Yb ratio in this analysis. Elevated La/Yb ratios are predicted in rocks that are fractionated out of garnet and/or amphibole rich residues in thick arcs, whereas La/Yb ratios are lower in thin arc residues14 (Fig. 1). Observations in exposed tilted sections of magmatic arcs have shown that the bulk of compositional diversity is achieved in the deeper parts of arc crust, where thermal efficiency is high4,17,24, regardless of crustal thickness. Therefore geochemical parameters can indirectly map out various fractionating phases, although clearly one expects complexities that need to be averaged out over time- and length-scales larger than an individual eruption.
Melting of subducted oceanic crust can also generate intermediate magmas25 and the principal reservoir that can be responsible for partial melting is the oceanic crust, which is the sum of various mafic igneous rocks, totaling about 6 km in thickness in most oceanic areas. The La/Yb of these melts will be low reflecting the light rare earth element depleted nature of mid-ocean ridge basalts (Fig. 1). Sr/Y, on the other hand should be >60 in the case of intermediate rocks generated by slab melting compared to Sr/Y of 30–50 for intermediate rocks extracted from arc roots from thick crust because of the higher Sr in altered mid-ocean ridge basalt compared to basalts/gabbros from sub arc environments23,26 (Fig. 1). The positive correlation between Sr/Y and La/Yb in virtually all intermediate arcs argues strongly for the fact that the majority of modern Pacific arc magmatic products owe their compositional diversity of the trace elements investigated here to partial melting or crystal fractionation in the crust27,28 and that slab melting plays at most a secondary role.
A large number of previously published whole rock geochemical analyses were used for this study; the methods section at the end of the manuscript details the selection of data and their manipulation.
Global and regional correlations
We observe that for Pliocene to modern subduction-related intermediate rock compositions there is a positive correlation of global arc-averaged Sr/Y and La/Yb with crustal thickness (Fig. 1). The empirical fits to the global data are:
where Sr/Y and La/Ybn are whole rock average ratios (and the subscript “n” for La/Yb implies that that ratio was normalized to chondritic values of McDonough and Sun29) and dm= crustal thickness or depth to Moho.
Our Sr/Y results are similar to the global Sr/Y correlation with crustal thickness of Chiaradia15; in that paper rocks spanning a wide range of compositions (from mafic to the most felsic) were mixed in the correlation. A limiting factor in this type of analysis is that sub-arc Moho depths obtained from seismic data may be underestimated because the crust-mantle boundary could be a complex transition involving mantle-like cumulates (e.g. gabbronorites transitioning to pyroxenites, with or without garnet5,6,22), rather than a change from typical lower crustal rocks to a peridotitic assemblage. Therefore, the sub arc crust may be thicker petrologically than seismically, an uncertainty less likely to occur in other tectonic environments.
We also show in Fig. 1 that similar trends can be seen at regional scale in the northern part of the Southern Andean volcanic zone, which is the only modern arc where significant changes in crustal thickness (from over 60 km to less than 30 km17) occur over a relatively short distance along strike. Regional data correlations show that when data from individual composite volcanoes data are averaged over the lifetime of the volcano (about 5 million years), the trace elemental ratios discussed here mimic global trends. Of the two ratios, La/Yb has less scatter in global and regional correlations and is a better indicator of thicker crust in continental arcs. How these correlations using volcanic data apply to the subvolcanic and plutonic equivalents at depth is not straightforward. Limited comparisons of volcanic versus plutonic Sr/Y and La/Yb ratios in ancient arcs7 suggest that batholiths have Sr/Y ratios that may be somewhat higher (by ~10-15%) than the volcanic cover, whereas no systematic difference between volcanoes and batholiths is noticed in La/Yb.
Applications to ancient orogens
Figures 2 and 3 show La/Yb and Sr/Y respectively as a function of time for several ancient magmatic arcs whose regional tectonic evolution is independently constrained by various geologic data. We average out rocks representing individual magmatic systems (composite volcanoes or their roots representing co-magmatic suites in a batholith). This typically corresponds to a 30 km spatial average and ~5 million years temporal average, although the numbers differ from region to region, depending on local particularities of magmatism and data availability. The two central Andes transects are consistent with the idea that the most recent crustal thickening began during an early Cenozoic orogeny (45–55 Ma), a process that continues through modern time30,31. The Coast Mountains batholith displays three significant pulses of Sr/Y and La/Yb increase, one during the Latest Jurassic, another during the Late Cretaceous and the last one during the Paleocene-Eocene, separated by periods of lower Sr/Y and La/Y. All three episodes of increase in these ratios correspond temporally to periods of crustal thickening in the arc’s foreland area32. The Mesozoic Sierra Nevada batholith of California shows a gradual increase in crustal thickness from the mid-Cretaceous on, also consistent with independent geologic constraints such significant crustal shortening in the retroarc thrust belt33. In both cases, the leading hypothesis for intra arc crustal thinning is thought to be due to episodes of lower crustal delamination occurring after significant crustal thickening, a sequence that is thought to occur cyclically in long-lived arcs34. In contrast to these arcs, the along strike segment of arc that formed south of the Sierra Nevada, represented by the Mojave-Transverse Ranges corridor, does not appear to have attained thick crust during the late Cretaceous, as determined from independent regional geologic observations35; our compiled data confirm that assertion.
Modern global and regional data and ancient regional examples lend confidence to the use of Sr/Y and La/Yb correlations in determining paleo-crustal evolution in magmatically active subduction orogens, such as the Cordilleran/Andean ones investigated here. In contrast, short lived subduction events in Mediterranean type environments and collision-related magmatism are far less understood petrologically; our data may apply to those tectonic environments but that has not been tested yet. As for sui generis subduction magmatism, such correlations averaging out several processes as well as length and timescales will inherently be subjected to large errors; they nonetheless seem to provide reasonable quantitative constraints on crustal thickness evolution in subduction-related orogens. We propose that episodes of thin versus thick crust can be distinguished in the geologic record of subduction-related arc magmatism using the geochemical ratios presented here.
Databases for modern arcs
We use whole rock analyses from 25 Plio-Quaternary Circum-Pacific volcanic arcs from the GEOROC database (http://georoc.mpch-mainz.gwdg.de/georoc/). This data set consisting of 18,200 data points was filtered to include rocks with SiO2 content (55–68 wt %) and MgO content (<4 wt %); the resulting number of individual points after this filtering was decreased to 9,800. This range of silica concentration was used to eliminate mafic rocks generated in the mantle, as well as high silica granites; in both cases, the distribution of the trace elements investigated here may reflect processes other than those presented in this manuscript and therefore may not be indicative of crustal thickness. We removed La/Yb and Sr/Y outliers from these data subsets using the modified Thompson Tau statistical method and calculated median values. Data with average Rb/Sr > 0.2 or Rb/Sr < 0.05 were discarded, providing a trace elemental filter for mantle-derived rocks or rocks formed by melting of pre-existing metasedimentary framework rocks, which could have been missed by the silica and magnesium filters (protocol as in Chapman et al.16). A total of 4,800 data points were used in the global correlations (Supplementary files 1 and 2). Crustal thickness estimates for the volcanic arcs used in this paper are also given Supplementary files 1 and 2. A regional database for the northern part of the South Andean Volcanic zone was also generated using GEOROC and filtered identically as for the global set. Averages for individual composite volcanoes spanning as much as 5 million years are used in the correlation in Fig. 1. Crustal thickness was calculated using the average elevations of the areas occupied by strato-volcanoes, assuming isostatic equilibrium at a horizontal scale of 50 km and an average crustal density of 2700 kg/m3.
Data from ancient arcs
Only data with tightly constrained ages (<5 My error), most of which are determined by isotopic techniques were used here. Geochemical data on ancient arcs are from the central Andean data base30 (http://andes.gzg.geo.uni-goettingen.de). Data for the Coast Mountains, Sierra Nevada and Transverse Range-Mojave arc segments data are from NAVDAT, the North American Volcanic (and Intrusive) Data Base, or (http://www.navdat.org) supplemented by various published and unpublished data compiled by Scott Paterson and students. Filtering applied to these data is identical to that described above for the modern global and regional databases and Chapman et al.16. Filtered data were pooled into 5 million years time intervals and 50 km diameter spatial footprints in order to mimic typical average lifetimes and spatial distribution of modern volcanoes in subduction zones7.
How to cite this article: Profeta, L. et al. Quantifying crustal thickness over time in magmatic arcs. Sci. Rep. 5, 17786; doi: 10.1038/srep17786 (2015).
Grove, T. L., Till, C. B. & Krawczynski, M. J. The role of H2O in subduction zone magmatism. Annu. Rev. Earth Pl. Sc. 40, 413–39 (2012).
Turner, S. J. & Langmuir, C. H. The global chemical systematics of arc front strato volcanoes: Evaluating the role of crustal processes. Earth Planet Sc. Lett. 422, 182–193 (2015).
Lee, C. T., Cheng, X. & Horodyskyi, U. The development and refinement of continental arcs by primary basaltic magmatism, garnet pyroxenite accumulation, basaltic recharge and delamination: insights from the Sierra Nevada, California, Contrib. Mineral. Petr. 151, 222–242 (2006).
Annen, C., Blundy, J. D. & Sparks, R. S. J. . The genesis of intermediate and silicic magmas in deep crustal hot zones. J. Petrol. 47, 505–39 (2006).
Jagoutz, O. Arc crustal differentiation mechanisms. Earth Planet Sc. Lett. 396, 267–277 (2014).
Debari, S. M. & Greene, A. R. Vertical Stratification of Composition, Density and Inferred Magmatic Processes in Exposed Arc Crustal Sections. in Brown, D. & Ryan, P. D. (Editors) Arc continent collisions, Front. Earth Sci. 121–145, doi: 10.1007/978-3-540-88558-0_17 (2011).
Ducea, M. N., Bregantz, G. & Saleeby, J. B. The arhitecture, chemistry and evolution of continental magmatic arcs. Annu. Rev. Earth Pl. Sc. 43, 299 (2015).
Leeman, W. P. The influence of crustal structure on compositions of subduction-related magmas. J. Volcanol. Geoth. Res. 18, 561–588 (1983).
Plank, T. & Langmuir, C. H. An evaluation of the global variations in the major element chemistry of arc basalts. Earth Planet Sc. Lett. 90, 349–370 (1988).
Gromet, L. P. & Silver, L. T. REE variations across the Peninsular Ranges Batholith; implications for batholithic petrogenesis and crustal growth in magmatic arcs. J. Petrol. 28, 75–125 (1987).
Mantle, G. W. & Collins, W. J. Quantifying crustal thickness variations in evolving orogens: Correlation between arc basalt composition and Moho depth. Geology 36, 87–90, doi: 10.1130/g24095a.1 (2008).
Best, M. G., Barr, D. L., Christiansen, E. H., Gromme, S., Deino, A. L. & Tingey, D. G. The Great Basin Altiplano during the middle Cenozoic ignimbrite flareup: Insights from volcanic rocks. Int. Geol. Rev. 51, 589–633 (2009).
Lee, C.-T. A. & Morton, D. M. High silica granites: terminal porosity and crystal settling, Earth Planet Sc. Lett. 409, 23–31 (2015).
Kay, S. M. & Mpodozis, C. Central Andean ore deposits linked to evolving shallow subduction systems and thickening crust. Geol. Soc. Am. Today 11, 4–9 (2001).
Chiaradia, M. Crustal thickness control on Sr/Y signatures of recent arc magmas: an Earth scale perspective. Sci. Rep. 5, doi: 10.1038/srep08115 (2015).
Chapman, J. B., Ducea, M. N., Profeta, L. & DeCelles, P. G. Tracking changes in crustal thickness during orogenic evolution with Sr/Y; an example from the Western U.S. Cordillera. Geology 43, 919–923 (2015).
Hildreth, W. & Moorbath, S. Crustal contributions to arc magmatism in the Andes of central Chile. Contrib. Mineral. Petr. 98, 455–489, doi: 10.1007/BF00372365 (1988).
Lee, C.-T. A., Morton, D. M., Kistler, R. W. & Baird, A. K. Petrology and tectonics of Phanerozoic continent formation: from island arcs to accretion and continental arc magmatism. Earth Planet Sc. Lett. 263, 370–387 (2007).
Rapp, R. P. & Watson, E. B. Dehydration melting of metabasalt at 8-32 kbar; implications for continental growth and crust-mantle recycling. J. Petrol. 36, 891–931 (1995).
Wolf, M. B. & Wyllie, P. J. Dehydration-melting of amphibolite at 10 kbar; the effects of temperature and time. Contrib. Mineral. Petr. 115, 369–383 (1994).
Davidson, J., Turner, S., Handley, H., Macpherson, C. & Dosseto, A. Amphibole “sponge” in arc crust? Geology 35, 787–790, doi: 10.1130/g23637a.1 (2007).
Ducea, M. N., Constraints on the bulk composition and root foundering rates of continental arcs; a California arc perspective, J. Geophys. Res. 107, doi: 10.1029/2001JB000643 (2002).
Otamendi, J. E. et al. Generation of tonalitic and dioritic magmas by coupled partial melting of gabbroic and metasedimentary rocks within the deep crust of the Famatinian magmatic arc, Argentina. J. Petrol. 50, 841–873 (2009).
DePaolo, D. J. A neodymium and strontium isotopic study of the Mesozoic calc-alkaline granitic batholiths of the Sierra-Nevada and Peninsular Ranges, California. J. Geophys. Res. 86, 470–488 (1981).
Defant, M. J. & Drummond, M. S. Derivation of some modern arc magmas by melting of young subducted lithosphere. Nature 347, 662–665 (1990).
Martin, H., Smithies, Rapp R., Moyen, J.-F. & Champion, D. An overview of adakite, tonalite–trondhjemite–granodiorite (TTG) and sanukitoid: relationships and some implications for crustal evolution. Lithos 79, 1–24 (2005).
Castillo, P. R. Adakite petrogenesis. Lithos 134, 304–316 (2012).
Ducea, M. N. & Barton, M. D. Igniting flare-up events in Cordilleran arcs. Geology 35, 1047–1050 (2007).
McDonough, W. F. & Sun, S-s. The composition of the Earth. Chem. Geol. 120, 223–253 (1995).
Mamani, M., Worner, G. & Sempere, T. Geochemical variations in igneous rocks of the Central Andean orocline (13 S to 18 S): Tracing crustal thickening and magma generation through time and space. Geol. Soc. Am. Bull. 122, 162–182 (2010).
Chapman, A. D. et al. Constraints on plateau architecture and assembly from deep crustal xenoliths, northern Altiplano (SE Peru). Geol. Soc. Am. Bull. 127, doi: 10.1130/B31206.1 (2015).
Girardi, J. D. et al. Elemental and Isotopic Evidence for Granitoid Genesis From Deep-Seated Sources in the Coast Mountains Batholith, British Columbia. J. Petrol. 53, 1505–1536 (2012).
DeCelles, P. G. Late Jurassic to Eocene evolution of the Cordilleran thrust belt and foreland basin system, western USA. Am. J. Sci. 304, 105–168 (2004).
DeCelles, P. G., Ducea, M. N., Kapp, P. & Zandt, G. Cyclicity in Cordilleran orogenic systems. Nat. Geosci. 2, 251–57 (2009).
Chapman, A. D., Ducea, M. N., Kidder, S. & Petrescu, L. Geochemical constraints on the petrogenesis of the Salinian arc, central California: Implications for the origin of intermediate magmas Lithos 200, 126–141 (2014).
We acknowledge funding from the National Science Foundation Petrology and Geochemistry program (EAR-1019525 and EAR-0907880 to M.N.D. and EAR-0073943 and EAR-1019636 to S.R.P) and from the Romanian Executive Agency for Higher Education, Research, Development and Innovation Funding (UEFISCDI) (project PN-II-ID-PCE-2011-3-0217).
The authors declare no competing financial interests.
Electronic supplementary material
About this article
Cite this article
Profeta, L., Ducea, M., Chapman, J. et al. Quantifying crustal thickness over time in magmatic arcs. Sci Rep 5, 17786 (2016). https://doi.org/10.1038/srep17786
Synthesis on productive, sub-productive and barren intrusions in the Urumieh-Dokhtar magmatic arc, Iran, constraints on geochronology and geochemistry
Ore Geology Reviews (2021)
From long-lived batholith construction to giant porphyry copper deposit formation: petrological and zircon chemical evolution of the Quellaveco District, Southern Peru
Contributions to Mineralogy and Petrology (2021)
Journal of South American Earth Sciences (2021)