CC-BY 4.0
1. Introduction
Copper (Cu) is an essential element for the physiological functioning of living beings, but it can also be toxic in high concentrations (Flemming and Trevors, 1989; Brewer, 2010). It is also an indispensable metal for our societies, thanks to its many uses (transport, industry, communication, construction, agriculture). Demand for Cu is growing exponentially to meet the needs of both the world’s growing population and the energy transition. Demand is such that even in Europe, where mining activity has been excessively low for the past thirty years, former mining areas such as Andalusia (Huelva region, Spain) are experiencing a significant mining revival (Buu-Sao, 2021).
Because of its economic importance, numerous works have been carried out to understand the development of Cu-rich mineralization (e.g., Mathur, Titley, et al., 2009) and thus facilitate mineral exploration. Given its massive use in all sectors of our societies, many studies have been carried out on the dispersion of copper in the continental and oceanic environments (e.g., Little, Vance, Walker-Brown, et al., 2014; Blotevogel, Schreck, et al., 2019). At the end of the 90s, the first precise measurements of the copper isotope ratio (i.e., 65Cu/63Cu) were carried out by MC-ICP-MS (Maréchal et al., 1999), Cu having two stable isotopes (63Cu and 65Cu). The scientific community then began systematic measurements of this isotopic ratio in various environments (rocks, soils, rivers, sediments etc.) (Johnson et al., 2004; Teng et al., 2017; Moynier et al., 2017). The purpose of these studies was initially to make measurements in these reservoirs, and see if we could detect different isotope signatures. In a second step, the objective was to see if these signatures could be used as a source tracer or to understand the processes that induce them. In the latter case, understanding these processes opens a window of understanding on mechanisms that counteract the mobility of elements, in this case copper, in the natural environment (e.g., Borrok et al., 2008; Navarrete et al., 2011; Coutaud et al., 2018; Komárek et al., 2021).
The objective of this work is to make an inventory of copper isotope signatures (what amplitudes? what signs?) in the main reservoirs (rock, soil, river, oceans). After almost 25 years of work carried out with precise measurements of these isotopic ratios, it is important to consider whether they can provide us useful constraints in the understanding of the cycle of large-scale elements. The goal is not to discuss all the hypotheses of each paper, of each site, but to present the compilation of data in a neutral way and identify the significant features. This work should provide a simple, hard-hitting overview of the potential of this isotopic tool as a source tracer, but also of the main mechanisms at play within and between reservoirs. All the isotopic data presented are expressed in the form of δ65Cu (in units of ‰) whose reference material is SRM-NIST 976:
| \begin {equation}\label {eq1} \updelta {}^{65}\mathrm {Cu} (\textperthousand) = \left ( \left (\frac { \left (\frac {{\,}^{65}\mathrm {Cu}}{{\,}^{63}\mathrm {Cu}}\right )_{\mathrm {sample}}}{ \left (\frac {{\,}^{65}\mathrm {Cu}}{{\,}^{63}\mathrm {Cu}}\right )_{\mathrm {NIST976}}} \right )-1 \right )\times 1000. \end {equation} | (1) |
2. Rocks, mineralizations and alteration products (Figure 1)
Rocks (upper continental crust, mantle) show relatively low isotopic variability with the following mean values for δ65Cu (‰): 0.02 ± 0.06 (granite), 0.41 ± 0.16 (granodiorite), −0.25 ± 0.44 (peridotite), 0.11 ± 0.08 (basalt), and 0.11 ± 0.16 for sedimentary rocks. These data are consistent with the value proposed by Savage et al. (2015) for the Bulk Silicate Earth (0.07 ± 0.1). The most striking result is the high isotopic variability of Cu-rich primary or secondary mineralization and associated reservoirs (ore, mining waste, precipitates, river sediments impacted by acid mine drainage). While primary sulfides (e.g., chalcopyrite) show an average value of 0.33 ± 1.66, variability is extreme for secondary products in natural or mining contexts (−16.49 < δ65Cu(‰) < 9.98) (Mathur, Titley, et al., 2009) with an average value in the present data set of 0.07 ± 2.86. These secondary products may be either zones of secondary Cu enrichment (e.g. secondary chalcocite) (1.32 ± 1.26) or zones displaying the typical mineralogical assemblages of environments impacted by acid mine drainage with, in particular, ferrous minerals (e.g., goethite, hematite) or secondary sulfate minerals (e.g., jarosite, schwertmannite) (−1.59 ± 3.60). However, it seems that zones with secondary enrichment (chalcocite) are generally clearly positive, while leaching zones with a mixture of goethite, hematite and sulfates are negative (Mirnejad et al., 2010; Mathur, Ruiz, Casselman, et al., 2012; Mathur, Jin, et al., 2012). Experimental works (e.g., Ehrlich et al., 2004) have clearly shown that the dissolution of primary sulfide minerals results in the preferential departure of the heavy isotope (in the fluids), leaving an in-situ light-enriched solid residue. This result may seem contradictory to the synthesis of the data presented in Figure 1, where secondary products exhibit both positive or negative values. However, this high variability can be explained by the complexity of the water–rock interaction processes occurring in mineralized and/or mining areas. Indeed, in these environments, the dissolution of sulfides results in very charged percolating waters which can re-precipitate minerals, the latter may be dissolved under favorable conditions and so on (Lottermoser et al., 1999; Olias et al., 2006; Nieto et al., 2007). This succession of precipitation/dissolution may result in a high degree of fractionation (isotopic variability) in secondary products, solids and liquids, which can be both negative and positive (Perez Rodriguez et al., 2013). The strong isotopic amplitude in mining areas make copper isotopes a potentially relevant tracer for the study of metal transfer processes within these environments but also to assess the impact of these mining districts on continental and marine ecosystems.
Type of sample and associated bibliographic reference for rocks, mineralization (ore) and alteration products; all data are presented in the form of δ65Cu (in units of ‰) whose reference material is SRM-NIST976 (see Equation (1)). This is the case for all figures. For each reference, is reported a mean value associated with a standard deviation and the number of samples. Sometimes there is only one data. The points surrounded by a dotted line are from the same study and the reference (ex: S-6) allows to find the author and the associated values in the legend. R1: Granite from Lachlan Fold Belt (SE Australia) (0.08 ± 0.34 (n = 32); W. Li et al., 2009); R2: Granite from porphyry-type ore bodies (Lhasa, Tibet) (0.04 ± 0.23 (n = 4); Y.-C. Zheng et al., 2019); R3: Diorite from Central Asian Orogenic Belt (West China) (−0.06 (n = 1); Zhao et al., 2022); R4: Granodiorite (Colorado, USA) (= GSP-2, Certified Reference Material) (0.3 (n = 1); R.-R. Wang et al., 2022); R5: Granodiorite from porphyry-type ore bodies (Lhasa, Tibet) (0.34 ± 0.24 (n = 36); Y.-C. Zheng et al., 2019); R6: Granodiorite from Qulong porphyry copper deposit (Tibet) (0.59 ± 0.13 (n = 14); Nie et al., 2012); R7: Mafic enclave (Lhasa, Tibet) (0.42 ± 0.3 (n = 12); Y.-C. Zheng et al., 2019); R8: Lherzolite from Central Asian Orogenic Belt (West China) (−0.12 ± 0.27 (n = 11); Zhao et al., 2022); R9: Hornblende Lherzolite from Central Asian Orogenic Belt (West China) (−0.63 ± 0.42 (n = 6); Zhao et al., 2022); R10: Olivine Gabbro from Central Asian Orogenic Belt (West China) (−0.3 ± 0.2 (n = 8); Zhao et al., 2022); R11: Cratonic peridotite (China) (−0.64 ± 0.68 (n = 30); S.-A. Liu, Huang et al., 2015); R12: Basalt (Iceland) (= BIR, Certified Reference Material) (−0.02 (n = 1); W. Li et al., 2009); R13: Columbia River Basalt (USA) (= BCR-1/2, Certified Reference Material) (0.19 (n = 1); K. V. Sullivan et al., 2022); R14: Diabase (Virginia, USA) (= W-2a Certified Reference Material) (0.04 (n = 1); K. Sullivan et al., 2020); R15: Hawaiian basalt (USA), (= BHVO-2, Certified Reference Material) (0.08 (n = 1); R.-R. Wang et al., 2022); R16: Basalt from Country Rock (Huangshannan, Huanshandong, Hulu and Tulaergen deposits, Eastern Tianshan, China) (0.22 ± 0.04 (n = 2); Zhao et al., 2022); R17: Basalt (Southwest Pacific Ocean, Niuatahi volcan, Tonga) (0.13 ± 0.10 (n = 3); Z. Wang et al., 2019); R18: Dacite (Southwest Pacific Ocean, Niuatahi volcan, Tonga) (0.21 ± 0.1 (n = 9); Z. Wang et al., 2019); R19: Basalt (Hainan Island, China) (0.07 (n = 1); J.-H. Liu et al., 2022); R20: Andesite (Oregon, USA), (= AGV-2, Certified Reference Material) (0.06 (n = 1); K. V. Sullivan et al., 2022); R21: Andesitic breccia tuff from Country Rock (Huangshannan, Huanshandong, Hulu and Tulaergen deposits, Eastern Tianshan, China) (0.10 ± 0.03 (n = 3); Zhao et al., 2022); R22: Cambrian micashist (southwest France) (0.07 (n = 1); El Azzi et al., 2013); R23: Biotite quartz schist from Country Rock (Huangshannan, Huanshandong, Hulu and Tulaergen deposits, Eastern Tianshan, China) (0.46 ± 0.04 (n = 2); Zhao et al., 2022); R24: Shale (Marcellus formation, USA) (0.04 ± 0.16 (n = 8); Mathur, Jin, et al., 2012); R25: Black shales (South Africa) (−0.05 ± 0.42 (n = 11); Chi-Fru et al., 2016); R26: Black shales (Gabon) (0.23 ± 0.09 (n = 8); Chi-Fru et al., 2016); R27: Black shales (Maokou formation, China) (0.14 ± 0.09 (n = 5); Lv et al., 2016); R28: Cody Shale (USA) (= SCo-1, Certified Reference Material) (−0.16 (n = 1); Sossi et al., 2014); R29: Loess (China) (0.01 ± 0.03 (n = 2); W. Li et al., 2009); R30: Greywacke (Goa state, India) (0.03 (n = 1); Little, Munson, et al., 2019); R31: Sandstone (0.26 ± 0.12 (n = 2); Zhao et al., 2022); R32: Bedrock (Fluvio-lacustrine sediments) (China) (0.12 ± 0.01 (n = 2); Ren et al., 2022); R33: Tuffite (Slovakia) (0.14 (n = 1); Bigalke, Weyer, Kobza, et al., 2010); R34: Bulk Silicate Earth (MORB + OIB + Komatite + Peridotite) (0.07 ± 0.1; Savage et al., 2015); R35: Mix of chalcopyrite, chalcocite, djurleite, and covellite from Canariaco district (Peru) (0.03 ± 0.85 (n = 37); Mathur, Ruiz, Casselman, et al., 2012); R36: Mix of goethite and hematite from Canariaco district (Peru) (−5.35 ± 2.28 (n = 7); Mathur, Ruiz, Casselman, et al., 2012); R37: Mix of chalcopyrite and pyrite from Urumieh-Dokhtar magmatic belt (Iran) (0.43 ± 0.4 (n = 10); Mirnejad et al., 2010); R38: Mix of hematite and goethite from Urumieh-Dokhtar magmatic belt (Iran) (−3.84 ± 4.49 (n = 10); Mirnejad et al., 2010); R39: Supergene chalcocite from Urumieh-Dokhtar magmatic belt (Iran) (3.16 ± 1.29 (n = 11); Mirnejad et al., 2010); R40: Cu–Co ore (= SU-1, Certified Reference Material) (−0.02; Chapman et al., 2005); R41: Mix of Pyrite, chalcopyrite, galene, and sericite from Prospect Gulch abandoned mine (Colorado, USA) (−0.13 ± 0.1 (n = 2); Fernandez and Borrok, 2009); R42: Chalcopyrite ore (skarn type) (China) (−1.61 ± 0.55 (n = 3); Mathur, Munk, et al., 2014); R43: Chalcopyrite and pyrite (massive sulfide, Spain) (−1.91 ± 2.6 (n = 15); Mathur, Munk, et al., 2014); R44: Chalcopyrite and pyrite (porphyry copper deposit, Chile) (3.98 ± 2.38 (n = 4); Mathur, Munk, et al., 2014); R45: Chalcopyrite and pyrite from 9 porphyry copper deposits (USA, Chile, Turkey) (1.21 ± 2.74 (n = 7); Mathur, Titley, et al., 2009); R46: Secondary chalcocite (0.53 ± 4.89 (n = 15); Mathur, Titley, et al., 2009); R47: Leach cap (mix of hematite, jarosite and goethite) (−2.49 ± 4.01 (n = 5); Mathur, Titley, et al., 2009); R48: Oxide (1.98 ± 4.81 (n = 10); Mathur, Titley, et al., 2009); R49: Chalcopyrite (USA, Peru) (0.04 ± 0.52 (n = 38); Larson et al., 2003); R50: Pyrite from Mining district of Taxco (southern Mexico) (0.18 ± 0.28 (n = 4); Dótor-Almazán et al., 2017); R51: Tailings and precipitates from mining district of Taxco (southern Mexico) (−0.18 ± 0.70 (n = 14); Dótor-Almazán et al., 2017); R52: Chalcopyrite (Colorado, USA) (−0.01 (n = 1); Kimball et al., 2009); R53: Enargite (Colorado, USA) (0.29 (n = 1); Kimball et al., 2009); R54: Chalcopyrite (0.37 ± 0.23 (n = 29); Y.-C. Zheng et al., 2019); R55: Chalcopyrite (0.36 ± 0.2; Z. Q. Li et al., 2009); R56: Chalcopyrite (Schwarzwald area, South Germany) (0.014 ± 0.5 (n = 25); Markl et al., 2006); R57: Oxidized chalcopyrite (Schwarzwald area, south Germany) (−1.14 ± 0.89 (n = 3); Markl et al., 2006); R58: Chalcopyrite (Dexing mine, China) (0.33 ± 0.57 (n = 9); Song et al., 2016); R59: Pyrite from Dexing mine (China) (2.81 ± 4.47 (n = 15); Song et al., 2016); R60: Tailings from Dexing mine (China) (5.1 ± 0.7 (n = 4); Song et al., 2016); R61: Chalcopyrite (Lau Basin, Pacific) (0.32 (n = 1); Maréchal et al., 1999); R62: Chalcopyrite (Chuquicamata, Péru) (0.27 (n = 1); Maréchal et al., 1999); R63: Chalcopyrite (Morenci, USA) (5.74 (n = 1); Maréchal et al., 1999); R64: Malachite (Zaire) (−0.24 (n = 1); Maréchal et al., 1999); R65: Chrysocolla (Bagdad, USA) (−0.12 (n = 1); Maréchal et al., 1999); R66: Azurite (Chessy, France) (2.05 (n = 1); Maréchal et al., 1999); R67: Hypogene zone from Kerman Porphyry Copper Belt (Iran) (1.49–7.31 (range); Sarjoughian et al., 2024); R68: Leached zone from Kerman Porphyry Copper Belt (Iran) (−3.41 to 5.82 (range); Sarjoughian et al., 2024); R69: Supergene enriched zone with secondary minerals (Kerman Porphyry Copper Belt, Iran) (5.18–8.71 (range); Sarjoughian et al., 2024); R70: Chalcopyrite (Cu–Pb–Zn deposit from Mount Isa, Australia) (−0 ± 0.26 (n = 12); Mahan et al., 2023); R71: Cu sulphide minerals within porphyries (−2.31 ± 0.68 (n = 9); Asael et al., 2007); R72: Cu sulphide minerals within Cambrian dolomite (−2.03 ± 0.6 (n = 13); Asael et al., 2007); R73: Chalcopyrite (0.13 ± 0.47 (n = 3); Kim et al., 2023); R74: Chalcopyrite (Chile) (0.09 ± 0.38 (n = 24); Mathur, Ruiz, Titley, et al., 2005); R75: Chalcopyrite from the skarn ore bodies (Hongshan-Hongniu deposit, China) (0.34 ± 0.22 (n = 28); Wang et al., 2017); R76: Chalcopyrite from Dabu porphyry Cu–Mo deposit (Tibet) (−0.10 ± 0.60 (n = 22); S. Wu et al., 2017); R77: Secondary minerals from bed rivers and open pits impacted by Tharsis Mine (Huelva, Spain) (−1.82 ± 3.18 (n = 9); Viers, Freydier, et al., 2023).
3. Soils (Figure 2)
Type of sample and associated bibliographic reference for soils; The points surrounded by a dotted line are from the same study. S1: Fluvisol (Glen Feshie chronosequence, Scotland, United Kingdom) (−0.25 ± 0.11 (n = 20); Vance, Matthews, et al., 2016); S2: Andisol from Hawaii (USA) (−0.08 ± 0.38 (n = 54); Vance, Matthews, et al., 2016); S3: Lateritic soils from Amazonia (Brazil) (−0.39 ± 0.12 (n = 8); Guinoiseau et al., 2017); S4: Lateritic soils from Goa state (India) (−0.37 ± 0.33 (n = 14); Little, Munson, et al., 2019); S5: Lateritic soils from Penglai weathering profile (Hainan Island, China) (−0.05 ± 0.15 (n = 19); J.-H. Liu et al., 2022); S6: Saprolitic soils on basalt (South Carolina, USA) (0.4 ± 1 (n = 17); S.-A. Liu, Teng, et al., 2014); S7: Soils on basalt from Hainan Island (China) (0.01 ± 0.16 (n = 19); S.-A. Liu, Teng, et al., 2014); S8: Arenosol from the French Soil Quality Monitoring (France) (−0.15 (n = 1); Fekiacova et al., 2015); S9: Cambisol (on basalt) from the French Soil Quality Monitoring (France) (0.09 (n = 1); Fekiacova et al., 2015); S10: Cambisol (on granit) from the French Soil Quality Monitoring (France) (0.01 (n = 1); Fekiacova et al., 2015); S11: Podzol from the French Soil Quality Monitoring (France) (−0.14 (n = 1); Fekiacova et al., 2015); S12: Soils on Marcellus formation (Pennsylvania, USA) (−0.5 ± 0.19 (n = 33); Mathur, Jin, et al., 2012); S13: Paddy soil, gley horizon (Suzhou, Eastern China) (−0.09 ± 0.08 (n = 33); R.-R. Wang et al., 2022); S14: Liaohe plain Soil (= GBW07425 (GSS-11, Certified Reference Material) (−0.07 (n = 1); K. V. Sullivan et al., 2022); S15: Paddy soil (GBW07443) (GSF-3, Certified Reference Material) (−0.04 (n = 1); K. V. Sullivan et al., 2022); S16: Soil from North China Plain (GBW07389) (GSS-33, Certified Reference Material) (−0.06 (n = 1); K. V. Sullivan et al., 2022); S17: Deeper horizon from retisol (Yonne Plateau, France) (−0.38 ± 0.15 (n = 19); Kusonwiriyawong, Bigalke, Cornu, et al., 2017); S18: Carbonatic alluvial soil (floodplain, Switzerland) (−0.08; Kusonwiriyawong, Bigalke, Abgottspon, et al., 2016); S19: Skeletic cambisol developed on slate (Germany) (−0.15 ± 0.13 (n = 4); Bigalke, Weyer and Wilcke, 2011); S20: Dystric cambisol on sandstone (Slovakia) (−0.06 ± 0.12 (n = 7); Bigalke, Weyer and Wilcke, 2011); S21: Haplic podzol (Germany) (−0.20 ± 0.31 (n = 10); Bigalke, Weyer and Wilcke, 2011); S22: Haplic podzol (Slovakia) (−0.15 ± 0.08 (n = 6); Bigalke, Weyer and Wilcke, 2011); S23: Cultivated paddy soils (Mun River Basin, Thailand) (0.04; X. Zheng et al., 2023); S24: Abandoned paddy soils (Mun River Basin, Thailand) (0.06 (n = 1); X. Zheng et al., 2023); S25: Vertic cambisol (Vineyard catchment, Italia) (0.26 ± 0.08 (n = 8); Blotevogel, Oliva, et al., 2018); S26: Calcaric cambisol (Vineyard catchment, Italia) (0.18 ± 0.03 (n = 7); Blotevogel, Oliva, et al., 2018); S27: Fungicides (Cu sulfates) (0.85 ± 0.05 (n = 2); Blotevogel, Oliva, et al., 2018); S28: Fungicides (Cu hydroxydes) (0.09 ± 0.27 (n = 8); Blotevogel, Oliva, et al., 2018); S29: Fine clays fraction from Cambisol (Vineyard catchment) (−0.23 ± 0.16 (n = 8); Babcsanyi, Chabaux, et al., 2016); S30: Silt fraction from Cambisol (Vineyard catchment) (0.23 ± 0.04 (n = 4); Babcsanyi, Chabaux, et al., 2016); S31: Fungicides (Cu sulfates) (−0.07 ± 0.01 (n = 2); Babcsanyi, Chabaux, et al., 2016); S32: Fungicides (Cu hydroxydes) (0.36 ± 0.36 (n = 4); Babcsanyi, Chabaux, et al., 2016); S33: Soils from Bordeaux (Vineyard catchment, France) (−0.28 ± 0.03 (n = 3); Petit, Schäfer, et al., 2013); S34: Soils from Banyuls (Vineyard catchment, France) (−0.03 ± 0.03 (n = 2); El Azzi et al., 2013); S35: Cu fungicides (−0.34 (n = 1); El Azzi et al., 2013); S36: Dystric Cambisol (1 km–3.8 km from smelter, Slovakia) (0.15 ± 0.13 (n = 12); Bigalke, Weyer, Kobza, et al., 2010); S37: Stagni eutric cambisol (0.14 ± 0.13 (n = 6); Bigalke, Weyer, Kobza, et al., 2010); S38: Waste products (slag, ash, solid waste) (Slovakia) (0.59 ± 0.75 (n = 5); Bigalke, Weyer, Kobza, et al., 2010); S39: Top soils from Panhihua city area (Sichuan province, China) (−0.01 ± 0.11 (n = 19); Xia et al., 2023); S40: Ore from Panhihua city area (Sichuan province, China) (1.34 (n = 1); Xia et al., 2023); S41: Fly ash from Panhihua city area (Sichuan province, China) (0.16 ± 0.16 (n = 3); Xia et al., 2023); S42: Coal from Panhihua city area (Sichuan province, China) (2.44 (n = 1); Xia et al., 2023); S43: Smelting slag from Panhihua city area (Sichuan province, China) (1.29 (n = 1); Xia et al., 2023); S44: Contaminated soils (Tsumeb district, Namibia) (0.4 ± 0.22 (n = 8); Kribek et al., 2018); S45: Slag flotation tailings, smelter dust … (Tsumeb district, Namibia) (0.43 ± 0.3 (n = 18); Kribek et al., 2018); S46: Polluted soils from Copperbelt (Zambia) (−0.23 ± 0.18 (n = 13); Mihaljevic et al., 2018); S47: Concentrates processed in the smelter (Zambia) (−0.55 ± 0.14 (n = 4); Mihaljevic et al., 2018); S48: Contaminated soils from coal mining area (Suixi County, soutwest Huaibei city, China) (0.72 ± 0.16 (n = 17); Ren et al., 2022); S49: Flying ash coal mining area (Suixi County, soutwest Huaibei city, China) (0.92 (n = 1); Ren et al., 2022); S50: Soils from mining region of Taxco area (southern Mexico) (0.94 ± 0.57 (n = 5); Dótor-Almazán et al., 2017); S51: Tailing and precipitates from mining region of Taxco area (southern Mexico) (−0.13 ± 0.7 (n = 15); Dótor-Almazán et al., 2017); S52: Soil close to abandoned Dalseong Cu–W mine (Republic of Korea) (−0.46 to 0.32 (range); Kim et al., 2023); S53: Soil downstream the adit seepage (abandoned Dalseong Cu–W mine, Republic of Korea) (0.41–1.12 (range); Kim et al., 2023); S54: Adit seepages from Dalseong Cu–W mine (Republic of Korea) (1.94± 1.15 (n = 9); Kim et al., 2023); S55: Soils (colluviosol or leptosol) on limestone Marseille (−0.41 ± 0.2 (n = 9); Gelly et al., 2019); S56: Soil from Dexing mine (China) (3.94 ± 2 (n = 2); Song et al., 2016); S57: Tailings from Dexing mine (China) (5.1 ± 0.7 (n = 4); Song et al., 2016).
The first observation we can make is that soils exhibit a very low isotopic variability (−0.11 ± 0.19) and are generally enriched in the light isotope compared to the mean value of the rocks in the upper continental crust (see Section 2). However, we find that some soils may have much higher variability (S.-A. Liu, Teng, et al., 2014; Vance, Matthews, et al., 2016). The soil profile studied by S.-A. Liu, Teng, et al. (2014) in North Carolina (S6, Figure 2) reveals a high isotopic amplitude with values that can be high (0.4 ± 1) compared to the average soil. If we look at this study more closely, it appears that it is the surface soils which show strong joint enrichments of iron and copper that show these high isotopic values (up to 3.63). These data agree with the work of Balistrieri et al. (2008) and Pokrovsky et al. (2008) showing that the sorption of Cu onto amorphous Fe(III) metallic oxyhydroxides favors the heavy isotope. Note also that some soils (Andisol) in Hawaii may have high signatures which are explained by an input of Asian dust that contributes significantly to the Cu budget (Vance, Matthews, et al., 2016). However, in general, the global isotopic value we obtained for soils suggests that the pedogenetic processes induce the preferential leaching of the heavy isotope (65Cu). This signifies that the products (dissolved and particulate matter) of these weathering processes exported out of the soil should therefore exhibit an isotopic signature higher than that of soils. In other words, this suggests that the material transported by rivers must have a heavier overall signature than “average” soils.
Soils impacted by strong anthropogenic activities also show differences in isotopic composition compared to the average soil. Among soils impacted by winegrowing, it appears that they can present a negative signature (e.g., Bordeaux region, France, −0.37 to −0.28; S33, Figure 2; Petit, Schäfer, et al., 2013) or positive (e.g., Soave region, Italy; 0.18–0.26; S25–S26, Figure 2; Blotevogel, Oliva, et al., 2018). This contrast can be explained by the high isotopic variability of copper brought into vineyards as a fungicide (−0.34 < δ65Cu(‰) < 0.85; 0.3 ± 0.4). We obtained an average value of 0.02 ± 0.24 for soils impacted by winegrowing. Another particularity concerns soils impacted by mining activities and/or processing (in particular pyrometallurgy) which generally show a significantly heavier signature than average soils (0.96 ± 1.5). However, if this general trend is emerging, it should not obscure the fact that differences can be found locally, at certain sites. This can be illustrated by the work of Mihaljevic et al. (2018). Indeed, on this site of Zambia impacted by mining activity, the soils are isotopically negative (−0.43 ± 0.06‰) as are the mine tailings (−0.75 to −0.45‰; (Mihaljevic et al., 2018)). These soils impacted by viticulture or mining activities do not seem to follow a general rule regarding the sign and amplitude of fractionation. It is therefore important to properly characterize their reservoirs/sources and their isotopic signatures at a local scale. However, the magnitude of these fractionations makes the use of copper isotopes relevant in such contexts. Their use in natural contexts for the study of pedogenetic processes seems more limited.
4. World rivers (Figure 3)
Type of sample and associated bibliographic reference for river waters. R1: Amazon River (Brazil) (0.69 ± 0.09 (n = 3); Vance, Archer, et al., 2008); R2: Brahmaputra River (India) (0.64 (n = 1); Vance, Archer, et al., 2008); R3: Nile River (Sudan, Egypt) (0.58 ± 0.22 (n = 18); Vance, Archer, et al., 2008); R4: Chang-Jiang (Yangtze) River (China) (1.32 ± 0.1 (n = 13); Vance, Archer, et al., 2008); R5: Missouri River (USA) (0.15 ± 0.16 (n = 4); Vance, Archer, et al., 2008); R6: Kalix River (Sweden) (0.53 ± 0.07 (n = 7); Vance, Archer, et al., 2008); R7: River from Kamtchanka (Russia) (0.6 (n = 1); Vance, Archer, et al., 2008); R8: Ottawa River (Canada) (0.45 ± 0.06 (n = 2); Vance, Archer, et al., 2008); R9: Tocantins River (Brazil) (0.54 ± 0.09 (n = 3); Vance, Archer, et al., 2008); R10: Volga River (Russia) (0.64 ± 0.01 (n = 2); Vance, Archer, et al., 2008); R11: Itchen River (England) (0.64 ± 0.09 (n = 19); Vance, Archer, et al., 2008); R12: Beaulieu River (England) (0.79 ± 0.17 (n = 7); Vance, Archer, et al., 2008); R13: Suspended matter from Itchen River (England) (−0.55 ± 0.22 (n = 16); Vance, Archer, et al., 2008); R14: Global River Water (0.68; Vance, Archer, et al., 2008); R15: Suspended matter from Lot River (France) (−0.19 (n = 1); Petit, Schäfer, et al., 2013); R16: Suspended matter from Gironde (France) (−0.22 (n = 1); Petit, Schäfer, et al., 2013); R17: Suspended matter from Garonne River (France) (−0.24 ± 0.08 (n = 13); Petit, Schäfer, et al., 2013); R18: Garonne River (France) (0.12 (n = 1); Petit, Schäfer, et al., 2013); R19: Scheldt Estuary (Belgium) (0.01 (n = 1); Petit, Jong, et al., 2008); R20: Bed sediment from Scheldt Estuary (Belgium) (−0.07 ± 0.21 (n = 7); Petit, Jong, et al., 2008); R21: Palojoki River (Karely, Russia) (0.42 ± 0.05 (n = 3); Ilina et al., 2013); R22: Yangtze River (China) [1.68 (n = 1); Viers, unpublished data]; R23: River from China [0.20 (n = 1); Viers, unpublished data]; R24: River from Canada [0.27 (n = 1); Viers, unpublished data]; R25: river from Siberia (Russia) [0.08 (n = 1); Viers, unpublished data]; R26: River from Siberia (Russia) [0.05 (n = 1); Viers, unpublished data]; R27: Rivers from Khamtchanka Peninsula (Russia) [−0.25 ± 0.06 (n = 5); Viers, unpublished data]; R28: Suspended matter from Negro River (Brazil) (0.14 ± 0.14 (n = 16); Guinoiseau et al., 2017); R29: St Lawrence River (Canada) (0.38 (n = 1); K. V. Sullivan et al., 2022); R30: Floodplain sediment (China) (−0.02 (n = 1); K. V. Sullivan et al., 2022); R31: Tawara River (Japan) (1.24 ± 0.05 (n = 6); Takano, Tsuchiya, et al., 2021); R32: Uji River (Japan) (0.78; Takano, Tsuchiya, et al., 2021); R33: Glen Feshie River (Scotland) (0.03 ± 0.09 (n = 5); Vance, Matthews, et al., 2016); R34: Alluvium from Glen Feshie River (Scotland) (−0.23 (n = 1); Vance, Matthews, et al., 2016); R35: Chang-Jiang (Yangtze) River (China) (1.18 ± 0.3 (n = 20); Q. Wang et al., 2020); R36: Tributaries of the Yangtze River (China) (0.77 ± 0.28 (n = 7); Q. Wang et al., 2020); R37: Suspended matter from Zhujiang River (China) (0.17 ± 0.02 (n = 22); Zeng and Han, 2020); R38: Bed sediments from Huangpu River (tributary of the Yangtze river, China) (−0.37 to 0.18 (range); Tu et al., 2023); R39: Bed sediment from Baillaury River (Banyuls, France) (−0.03 ± 0.08 (n = 3); El Azzi et al., 2013); R40: Suspended matter from Baillaury River (Banyuls, France) (0.09 (n = 1); El Azzi et al., 2013); R41: Baillaury River (Banyuls, France) (0.31 (n = 1); El Azzi et al., 2013); R42: River from vineyard catchment (Alsace, France) (0.95 ± 0.25 (n = 9); Babcsanyi, Chabaux, et al., 2016); R43: Suspended matter from vineyard catchment (Alsace, France) (−0.20 ± 0.11 (n = 9); Babcsanyi, Chabaux, et al., 2016); R44: Sediment from urbanized area (Guanabara bay, Rio de Janeiro, Brazil) (−0.42 ± 0.05 (n = 5); Barreira et al., 2024); R45: Sediment from slightly urbanized area (Guanabara bay, Rio de Janeiro, Brazil) (−0.15 ± 0.18 (n = 5); Barreira et al., 2024); R46: Smelter impacted sediment from Ballinger lake (USA) (0.94 ± 0.1 (n = 11); Thapalia et al., 2010); R47: Sediment from Ballinger lake (USA) (0.77 ± 0.06 (n = x); Thapalia et al., 2010); R48: Spanish River (Great Lakes region, Canada) (−0.24 ± 0.35 (n = 8); Junqueira et al., 2023); R49: Mine waste seepage (Spanish River, Great Lakes region, Canada) (0.88 ± 0.1 (n = 2); Junqueira et al., 2023); R50: Trent River (Great Lakes region, Canada) (−0.26 ± 0.27 (n = 8); Junqueira et al., 2023); R51: Fischer Creek (USA) (1.43 ± 0.32 (n = 13); Borrok et al., 2008); R52: Tinto River (Iberian Pyrite Belt, Spain) (−0.45 ± 0.66 (n = 5); Borrok et al., 2008); R53: Odiel River (Iberian Pyrite Belt, Spain) (−0.74 ± 0.27 (n = 5); Borrok et al., 2008); R54: West Branch Ompampanoosuc River (−0.41 (n = 1); Balistrieri et al., 2008); R55: Meca River (Odiel River tributary, Iberian Pyrite Belt, Spain) (−0.31 ± 0.17 (n = 17); Masbou et al., 2020); R56: Meca River tributary (Iberian Pyrite Belt, Spain) (−2.72 (n = 1); Masbou et al., 2020); R57: Rivers from mining region of Colorado (USA) (1.59 ± 0.09 (n = 14); Kimball et al., 2009); R58: Cobica River (Iberian Pyrite Belt, Spain) (0.73 ± 0.71 (n = 4); Viers, J. A. Grande, et al., 2018); R59: Meca River (Odiel River tributary, Iberian Pyrite Belt, Spain) (−0.21 ± 0.47 (n = 7); Viers, Freydier, et al., 2023); R60: Rivers from Dexing mine region (China) (4.96 ± 4.58 (n = 20); Song et al., 2016).
In Figure 3, are plotted river data for dissolved phase (<0.22 μm or 0.45 μm), suspended sediment (>0.22 μm or 0.45 μm) and bottom sediments. If we consider unpolluted rivers, that is to say “just” subject to global diffuse pollution, it emerges that the dissolved phase of rivers shows relatively high isotopic variability (between −0.25 <δ 65CuDissolved (‰) < 1.68). In 2008, Vance et al. proposed a global average value for the dissolved fraction of rivers of 0.68‰. Vance, Archer, et al. (2008) suggest that the high value for the dissolved phase of rivers is related to the preferential complexation of heavy copper with organic ligands, as shown by the work of Ryan et al. (2014) and Bigalke, Weyer and Wilcke (2010). Compared to this average value proposed by Vance, Archer, et al. (2008), it appears that some rivers may have much more negative values [−0.25‰ (Khamchanka, Russia); R27, Figure 3; Viers (unp. value)] or much more positive ones (1.18 ± 0.3‰ for Yangtze River, China; R35, Figure 3; Q. Wang et al., 2020). Q. Wang et al. (ibid.) recognized that the values obtained for the Yangtze River and its tributaries are the highest ever measured; moreover, unpublished measurement (Viers) confirm these high values for the Yangtze River (R22, Figure 3). Q. Wang et al. (ibid.) attribute these high values to the effect of the Three Gorges Dam; the accumulation of sediment within the dam would favor the preferential sorption of the light isotope onto mineral surface leaving a heavier copper in the dissolved phase complexed with organic matter. Even if another average value is proposed in this paper (0.53 ± 0.45) for the dissolved phase of world rivers, to my opinion, proposing a rigorous weighted average flow value requires a broader sampling of other major rivers (e.g., Orinoco, Mekong, Congo, Yenisey, Lena) which provide a significant contribution of the Cu budget to the world ocean and for which we don’t have any data. It also seems necessary to see whether the isotopic signature in copper for these large river systems varies during the seasons, that is to say the hydrological year. There is, however, another aspect to be considered here. Some studies suggest that the flow of Cu to the World Ocean via rivers could be largely underestimated and in particular the transfer of Cu through coastal mining areas (Olias et al., 2006). This is the case, for example, of the Iberian Pyritic Belt (Huelva region, Spain) where the Tinto and Odiel rivers would export to the Atlantic Ocean almost 10% of the world’s copper flow (ibid.) due to the presence of nearly 90 abandoned mines in the watersheds of these rivers. We understand that given the particular signature of these rivers (−0.74 ± 0.27 and −0.45 ± 0.66 for Odiel and Tinto rivers, respectively; R52 and R53, Figure 3; Borrok et al., 2008), this region (and similar coastal regions) may influence the global isotopic signature of the world’s ocean. It is important to note that the rivers draining areas impacted by mining activities are those with the highest isotopic amplitude (0.28 ± 1.84). Given what we observed for solids (precipitates, sediments, ore) in these mining areas, it is not surprising to find this high amplitude for watercourses.
It also emerges from this synthesis that sediments show more negative signatures than the dissolved phase of rivers (−0.55 <δ 65CuSediment (‰) < 0.23) with an average value of −0.09 ± 0.20. If we consider, based on literature data (e.g., Boyle et al., 1977; Viers, Dupré, et al., 2009; Little, Vance, Walker-Brown, et al., 2014; Gaillardet et al., 2014; Takano, Tanimizu, et al., 2014; Li et al., 2020; Machu et al., 2024) that 90% of the copper is exported in particulates, a simple mixing calculation reveals that rivers export to the world ocean a whole copper close to 0 or slightly negative depending on whether one considers the value proposed by Vance, Archer, et al. (2008) or that of this article. Considering such a distribution, this isotopic signature of this whole material transported by rivers seems contradictory because it was thought to find the missing fraction of the soils and therefore a rather positive pool was expected. This apparent bias may be linked to a still too incomplete database that does not properly represent the reality of the isotopic signatures of these reservoirs (soils, river sediments) or the isotopic signatures observed in watercourses are impacted by secondary exchange/redistribution processes fractionating copper isotopes within watercourses.
5. Oceans (Figure 4)
Type of sample and associated bibliographic reference for Oceans; the points surrounded by a dotted line proviennent d’une même zone géographique (China sea, Pacific Ocean). O1: Mediterranean sea (0.51 ± 0.2 (n = 96); Baconnais et al., 2019); O2: South Western Mediterranean sea (0.48 ± 0.09 (n = 23); Baconnais et al., 2019); O3: South Eastern Mediterranean sea (0.55 ± 0.14 (n = 19); Baconnais et al., 2019); O4: North Western Mediterranean sea (0.50 ± 0.08 (n = 21); Baconnais et al., 2019); O5: North Eastern Mediterranean sea (0.54 ± 0.09 (n = 14); Baconnais et al., 2019); O6: North Atlantic Ocean (0.50 ± 0.09 (n = 20); Baconnais et al., 2019); O7: Sapropel (−0.09 (n = 1); Maréchal et al., 1999); O8: South Atlantic (0.64 ± 0.09 (n = 63); Little, Archer, et al., 2018); O9: Particles from South Atlantic (0.18 ± 0.15 (n = 17); Little, Archer, et al., 2018); O10: Leachable pool from particles from South Atlantic (0.4 ± 0.11 (n = 16); Little, Archer, et al., 2018); O11: Near shore seawater (Halifax Harbour, North Atlantic) (= CASS-5 reference material) (0.46; K. V. Sullivan et al., 2022); O12: Seawater (Nova Scotia, North Atlantic) (NASS-6 reference material) (0.42; K. V. Sullivan et al., 2022); O13: English channel (0.7 (n = 1); Bermin et al., 2006); O14: Sediment from Atlantic Ocean (0.22 ± 0.1 (n = 8); Maréchal et al., 1999); O15: Suspended particle (lower fraction, <250 μm) from continental shelf of northern Bay of Biscay (north eastern Atlantic Ocean) (−0.42 ± 0.04 (n = 2); Araujo, Knoery, et al., 2022); O16: Suspended particle (coarse fraction, >250 μm) from inner continental shelf of northern Bay of Biscay (north eastern Atlantic Ocean) (0.09 ± 0.03 (n = 5); Araujo, Knoery, et al., 2022); O17: Plankton, BCR 414 Certified Reference Material) (−0.27 (n = 1); Araujo, Knoery, et al., 2022); O18: Surface sediments from the Loire estuary (France) (−0.08 ± 0.08 (n = 17); Araujo, Ponzevera, et al., 2019); O19: Paleovase from the Loire estuary (France) (0.03 ± 0.03 (n = 12); Araujo, Ponzevera, et al., 2019); O20: Marine sediments from Beaufort sea (MESS-3 reference material) (0.01; K. Sullivan et al., 2020); O21: Marine sediments from British Columbia (Canada) (PACS Certified Reference Material) (0.03; K. Sullivan et al., 2020); O22: Modern marine sediments (Cariaco basin) (0.14 ± 0.08 (n = 2); Little, Vance, McManus, et al., 2017); O23: Modern marine sediments (California borderland basin) (0.14 ± 0.08 (n = 4); Little, Vance, McManus, et al., 2017); O24: Modern marine sediments (Black Sea) (0.2 ± 0.01 (n = 4); Little, Vance, McManus, et al., 2017); O25: Modern marine sediments (Peru margin) (0.32 (n = 1); Little, Vance, McManus, et al., 2017); O26: Modern marine sediments (Mexican margin) (0.16 ± 0.03 (n = 3); Little, Vance, McManus, et al., 2017); O27: North Atlantic marine sediment (HISS-1 Certified Reference Material) (0.05 ± 0.02; Takano, Liao, Tian, et al., 2020); O28: South Central Indian Ocean (<100 m) (0.5 ± 0.01 (n = 4); Takano, Tanimizu, et al., 2014); O29: South Central Indian Ocean (>300 m) (0.64 ± 0.03 (n = 12); Takano, Tanimizu, et al., 2014); O30: South Central Indian Ocean (1.1 ± 0.16 (n = 12); Vance, Archer, et al., 2008); O31: Western North Pacific Ocean (<100 m) (0.47 ± 0.03 (n = 6); Takano, Tanimizu, et al., 2014); O32: Western North Pacific Ocean (>300 m) (0.65 ± 0.04 (n = 12); Takano, Tanimizu, et al., 2014); O33: Eastern North Pacific Ocean (<100 m) (0.49 ± 0.06 (n = 7); Takano, Tanimizu, et al., 2014); O34: Eastern North Pacific Ocean (>300 m) (0.73 ± 0.06 (n = 18); Takano, Tanimizu, et al., 2014); O35: North East Pacific Ocean (1.14 ± 0.12 (n = 11); Vance, Archer, et al., 2008); O36: North East Pacific (0.85–1.35 (range); Bermin et al., 2006); O37: North Tasman sea (SW Pacific Ocean) (0.61 ± 0.16 (n = 15); Thompson and Ellwood, 2014); O38: Mid Tasman sea (SW Pacific Ocean) (0.7 ± 0.2 (n = 17); Thompson and Ellwood, 2014); O39: South Tasman sea (SW Pacific Ocean) (0.78 ± 0.17 (n = 16); Thompson and Ellwood, 2014); O40: Particles from Tasman sea (SW Pacific Ocean) (0.17 ± 0.15 (n = 7); Thompson and Ellwood, 2014); O41: Surface water (<200 m) from Kuroshio current along the East China Sea (0.47 ± 0.05 (n = 37); Takano, Liao, Ho, et al., 2022); O42: Deep water (>600 m) from Kuroshio current along the East China Sea (0.68 ± 0.02 (n = 17); Takano, Liao, Ho, et al., 2022); O43: Northern South China Sea (<400 m) (0.47 (n = 1); Takano, Liao, Tian, et al., 2020); O44: Northern South China Sea (>400 m) (0.68 ± 0.03 (n = 4); Takano, Liao, Tian, et al., 2020); 045: Particles from Northern South China Sea (0.13–0.36; Takano, Liao, Tian, et al., 2020); O46: Fe–Mn nodules from Pacific Ocean (0.54 ± 0.04 (n = 8); Little, Vance, Walker-Brown, et al., 2014); O47: Fe–Mn nodules from Indian Ocean (0.28 ± 0.39 (n = 8); Little, Vance, Walker-Brown, et al., 2014); O48: Fe–Mn nodules from Atlantic Ocean (0.33 ± 0.07 (n = 8); Little, Vance, Walker-Brown, et al., 2014); O49: Natural background sediment from Sepetiba Bay (Brazil) (0.2 (n = 1); Jeong, Araujo, Garnier, et al., 2023); O50: Sediment from Sepetiba bay (Brazil), impacted by smelter (0.32 ± 0.18 (n = 13); Jeong, Araujo, Garnier, et al., 2023); O51: Sediment from Northern Sepetiba bay (Brazil) (0.23 ± 0.1 (n = 12); Jeong, Araujo, Garnier, et al., 2023); O52: Metallurgic slag from Sepetiba bay (Brazil) (0.4 ± 0.2; Jeong, Araujo, Garnier, et al., 2023); O53: Sediments from Busan Harbor (South Korea) (0.27 ± 0.04 (n =7); Jeong, Araujo, Knoery, et al., 2023); O54: Domestic Antifoulings paints (South Korea) (0.48 ± 0.06 (n = 5); Jeong, Araujo, Knoery, et al., 2023); O55: Imported antifouling sources paints (South Korea) (0.39 ± 0.16 (n = 4); Jeong, Araujo, Knoery, et al., 2023); O56: Cu-contaminated Mediterranean sediments from marina of Port Camargue (France) (0.07 ± 0.12 (n = 25); Briant et al., 2022); O57: Antifouling paints Cu-contaminated Mediterranean marina of Port Camargue (0.54 ± 0.05 (n = 3); Briant et al., 2022).
Given the limited data available, it appears that the World Ocean has an average value of 0.63 ± 0.12‰ in agreement with literature data (Little, Archer, et al., 2018). This isotopic signature of the oceans ranges from 0.50‰ (Mediterranean Sea and Atlantic Ocean) to 0.85‰ (North Pacific Ocean). The most striking result, however, is the difference in isotopic composition between the dissolved phase (0.63 ± 0.12‰) and the particulate phase (0.07 ± 0.17‰), enriched in the light isotope. We find here a similarity with the dissolved and particulate phases of rivers, for which we have also an isotopically dissolved phase heavier than the particulate phase. A number of articles (e.g., Vance, Archer, et al., 2008; Takano, Tanimizu, et al., 2014; Little, Vance, Walker-Brown, et al., 2014; Sherman and Little, 2020) have attempted to explain the copper cycle within the oceans and the associated isotopic fractionations during intra-oceanic processes (inputs from rivers and the atmosphere, intra-oceanic exchange between dissolved and particulate matter, role of planktonic matter and Fe–Mn nodules). In particular, these authors (e.g., Little, Archer, et al., 2018) point out that if the signature of the ocean is similar (0.65‰) to that provided by rivers (0.68‰) a mass balance in stationary state requires an additional source of light Cu. This source of light Cu could be hydrothermalism or the dissolution of terrigenous particles, brought by rivers or the atmosphere. However, the establishment of these balance sheets is subject to high uncertainties regarding flows and associated isotopic signatures. While the signature of the particles brought by the rivers seems homogeneous and well constrained, it is not the same for the dissolved phase. There is also considerable uncertainty about the isotopic signature of atmospheric particles. If it seems that particles in cities have isotopic signatures between 0 and 0.6 (global diffuse pollution of urban environments; Souto-Oliveira et al., 2018; Dong, Ochoa Gonzalez, et al., 2017; Ochoa Gonzalez et al., 2016), the few measurements available for non-anthropized environments (Sahel Region (Africa), Taklamakan desert (Asia), equatorial eastern North Atlantic region of West Africa, eastern tropical Atlantic; South China Sea) reveal values in a more negative range (−0.24 to 0.83) (Dong, Weiss, et al., 2013; Packman et al., 2022; Takano, Liao, Tian, et al., 2020). Aerosols values obtained for sites heavily impacted by mineral processing activities in Central Europe appear very negative (−3.63 to 0.42) (Novak et al., 2016; Mihaljevic et al., 2018). Another striking fact resulting from this synthesis is that there is no major isotopic difference between the signature of marine particles and those of harbor sediments impacted by anthropogenic activities (Briant et al., 2022; Jeong, Araujo, Garnier, et al., 2023; Jeong, Araujo, Knoery, et al., 2023). These data appear similar to the mean value for terrestrial rocks.
Beyond these uncertainties on flux or isotopic signatures, work must be continued on particle dissolution processes (e.g., Little, Archer, et al., 2018; Thompson and Ellwood, 2014) as well as on processes occurring in estuarine zones. There is no data on these environments.
6. Conclusion
The objective of this work was to make an inventory of copper isotope signatures in the main reservoirs (rock, soil, river, oceans). What amplitudes, what signs? After almost 25 years of work carried out with precise measurements of these isotopic ratios, it is important to consider whether they can provide us useful constraints in the understanding of the cycle of large-scale elements. Some results can be confirmed while others today have too many uncertainties: (1) the isotopic signature of rocks shows a low amplitude; (2) the regions with sulfides or affected by acid mine drainage are those that exhibit the most important fractionations either for solids (rocks, precipitates, sediments) or for waters; (3) soils not impacted by human activities (industry, mine, winegrowing) have low variability; (4) Cu leached from the soils is heavier than the residual soil; (5) the global isotopic signature used today for the dissolved phase of rivers should, to my opinion, be improved by considering seasonal sampling and unsampled large river systems; (6) the particulate matter of rivers is more negative than the dissolved phase; (7) coastal mining areas are potentially a major source of copper to the global Ocean and can impact the isotopic signature of the global Ocean; (8) the dissolved phase of the Oceans is much heavier than the particulate phase; (9) the work of modelling Cu isotopic signatures in the World Ocean remains hazardous due to the uncertainty of the isotopic signature of river and atmospheric flows.
Declaration of interests
The authors do not work for, advise, own shares in, or receive funds from any organization that could benefit from this article, and have declared no affiliations other than their research organizations.
Acknowledgement
I would like to deeply thank you Nolwenn Lemaitre for helpful discussion.