Research paper
Control of seasonal and inter-annual rainfall distribution on the Strontium-Neodymium isotopic compositions of suspended particulate matter and implications for tracing ENSO events in the Pacific coast (Tumbes basin, Peru)

https://doi.org/10.1016/j.gloplacha.2019.103080Get rights and content

Highlights

  • Suspended sediments were sampled monthly in the Tumbes River along 2 hydrological years.

  • εNd signatures indicates source provenance in relation with rainfall distribution.

  • 87Sr/86Sr signatures is particularly sensitive to anomalous wet conditions.

  • Nd and Sr isotopes are powerful tracers of paleo-ENSO and sediments dynamics.

Abstract

The geochemistry of riverine sediments exported to the oceans is important for paleo-hydro-climatic reconstruction. However, climate reconstruction requires a good understanding of the relationship between geochemistry and hydrological variability and sediment sources. In this study, we analyzed the major elements, the strontium‑neodymium radiogenic isotopes signatures (87Sr/86Sr and εNd) and the mineralogy of the suspended particulate matter (SPM) sampled monthly during two hydrologic years (2007–2008, a wet year, and 2010–2011, a normal hydrological year) upstream the Tumbes River outlet. The hydroclimate of this Ecuador-Peru binational basin is particularly sensitive to ENSO (El Niño Southern Oscillation) events.

While mineralogy (dominated by illite) and the chemical alteration index (from 75 to 82) remain almost constant along the two hydrological years, 87Sr/86Sr (0.7115 to 0.7176) and εNd (−7.8 to −1.9) signatures are particularly sensitive to discharge and SPM concentration variations. Along the hydrological year, two sources control the εNd variability: (1) volcanic rocks, which dominate during the dry season, and (2) plutonic/metamorphic sources, whose contribution increases during the wet season. This behavior is confirmed by the correlation between εNd signature and the monthly rainfall contribution from volcanic area (R = 0.58; p-value < 0.01), and also with the daily discharge at the outlet (R = -0.73; p-value < 0.01). For most of the samples, 87Sr/86Sr is less variable along the hydrological year. However, two exceptional high discharge and SPM concentration conditions sampled exhibit more radiogenic (higher) 87Sr/86Sr signatures when plutonic/metamorphic rocks derived sediments are released in sufficient quantities to notably change the SPM isotopic Sr value of the Tumbes River.

Hence, this study demonstrates that 87Sr/86Sr and εNd signatures can be used as powerful proxies for paleoclimate reconstructions based on sediment core's analysis in relation with spatial rainfall distribution and intensity in Pacific sedimentary basins submitted to the diversity of ENSO events.

Introduction

The hydrological and hydro-sedimentological regime of the Andes are particularly sensitive to extreme hydrological events like those related to the El Nino Southern Oscillation (ENSO) system. The El Niño and its counterpart, La Niña (the two expressions of the ENSO) are drivers of the strongest year-to-year climate fluctuations on the planet. They control the hydrology and sediment production in Andean basins residing both along the Pacific coast (Sulca et al., 2018; Rau et al., 2017; Lavado and Espinoza, 2014, Armijos et al., 2013) and in Amazonian slopes (Espinoza et al., 2012). These events have been responsible for extreme flooding in Pacific coastal areas (particularly in Northern Peru and in Ecuador) and droughts in the Andes and in the Amazon region (Lavado and Espinoza, 2014; Sulca et al., 2018; Espinoza et al., 2012). Importantly, the two main El Nino events of the last 40 years account for around 45% of the sedimentary flux exported to northern Peru Pacific Ocean (Morera et al., 2017). However, ENSO events present a large diversity associated with their tropical Pacific Sea Surface Temperature (SST) anomaly patterns (e.g. Capotondi et al., 2015) and the variability of these patterns can affect even the sign (positive or negative) of the precipitation anomalies in northern Peru (Lavado and Espinoza, 2014; Sulca et al., 2018). Thus, predicting the impact of extreme hydrological events associated with ENSO remains difficult because of the relatively short time-scale of hydrological and riverine suspended matter export monitoring.

The reconstruction of paleo-ENSO events is necessary to understand the main forcing on these events from the Pliocene (e.g. Wara et al., 2005) to Quaternary timescale, including the Holocene (e.g. Carré et al., 2012). Paleo ENSO events have been identified in onshore/continental geological record based on oxygen stable isotope compositions of speleothems (e.g. Bustamante Rosell et al., 2016), authigenic calcite lake sediment cores (Bird et al., 2011) and ice cores from the Ecuadorian and Peruvian Andes (e.g. Thompson et al., 2013). In all of these studies, most of the geochemical tracers used to reconstruct paleo ENSO events in the geological record were aimed at the identification of temperature (oceanic cores) or precipitation (lake cores, speleothems, ice cores) anomalies based on stable isotope geochemistry. To date, there has been little attempts to reconstruct paleo ENSO events based on the identification of peak sediment fluxes linked to rainfall increase in the Andean coast, which can be reflected in the sedimentary record by changes in the provenance of the associated sediments. For instance, change in provenance based on Nd isotopic composition of the detrital sediment fraction during the past 45,000 years have been used to reconstruct climate-driven changes in the provenance of clays deposited along the Mozambique Margin (van der Lubbe et al., 2016). Similarly, changes in the provenance of sediments deposited along the tropical South American continental margin between Andes and shield regions, identified based on Ndsingle bondSr isotopic composition variations, were also used for reconstructing both erosional and associated rainfall patterns on continental source regions during the Quaternary (e.g. Höppner et al., 2018). Even more recently, variation of Nd isotopic composition of Amazon River suspended particulate material (hereafter designated as “SPM”) during a one year hydrological cycle has been related to seasonal changes in the rainfall distribution patterns across the Amazon basin that are associated with latitudinal migrations of the Intertropical Convergence Zone (Rousseau et al., 2019). The relationship between long-term (Holocene) climate change and changing hydrology and sediment sources using Sr and Nd has been studied in the Nile basin (Woodward et al., 2015). This work illustrates the benefits of using Sr and Nd isotopes in tandem to tease out changes in catchment runoff and sediment delivery. However we have only identified 3 studies exploring Sr and Nd isotopes modern SPM variability along a hydrological year (Viers et al., 2008; Mao et al., 2011; Rousseau et al., 2019). These studies reported significant variability of the Sr and Nd signatures along the hydrological year. While εNd appears to mainly trace the lithological composition of the SPM source, 87Sr/86Sr can track both the SPM source (Mao et al., 2011) and grain size sorting effects due to either erosional processes (Viers et al., 2008) or hydrodynamical sorting behavior (Rousseau et al., 2019).

In this scenario, the combined use of Sr and Nd isotopes in sedimentary rock as potential tracers of paleo ENSO event may be particularly useful providing that isotopically contrasted rocks are differentially eroded during ENSO and normal years. To explore this possibility, we present the geochemistry and Nd and Sr isotopes composition of the SPM exported by the Tumbes River along contrasted hydrological periods at both the seasonal (dry vs wet season) and inter-annual (wet vs normal years) time scale. For this purpose, we have analyzed monthly sampled SPM at the lower reach of the basin along two hydrological cycles and interpreted the corresponding data as function of discharge, SPM concentration, SPM fluxes, seasonal and inter-annual rainfall distribution and geochemical characteristic of the SPM sources.

The Tumbes River basin is located in Southern Ecuador and Northern Peru (latitude −79.35 and −80.70 decimal degrees). It drains the western slope of the Andes (between longitude −3.47 and −4.25 decimal degree) over an area of 4.8 103 km2 including ~70% in the Andean mountains above 500 m.a.s.l. (meter above sea level; Moquet et al., 2018). It originates in the Andes (~3800 m.a.s.l.) and flows through a narrow coastal plain until its outlet to the Pacific Ocean. The river drains three main lithologically-contrasted domains: a volcanic, a plutonic/metamorphic and a sedimentary domain representing respectively 17, 25 and 58% of the area (Fig. 1; table S1). The Upper Tumbes basin drains the volcanic domain which consists in Cenozoic and Mesozoic volcanic rocks (elevation = 2014 ± 667 m.a.s.l.; average ± 1sd). They are composed of andesites, basalts, and locally derived pyroclastic rocks. The mid-altitude Tumbes basin drains the Paleozoic plutonic and metamorphic domain (elevation = 1036 ± 371 m.a.s.l.), which is mainly composed of schists, gabbro, granite and intermediate intrusive rocks (Fig. 1). The remaining part of the basin (elevation = 735 ± 333 m.a.s.l.) is formed by Cenozoic-Mesozoic mudstones, shales, and sandstones as well as locally derived modern alluvial fan deposits and limestones (Fig. 1).

The basin receives a rain amount of around 1000 mm·yr−1 which leads to a runoff of around 750 mm·yr−1 (Lavado et al., 2016; Lavado Casimiro et al., 2012). The cumulative annual rainfall tends to increase with elevation. The rainfall regime, and consequently the discharge regime, shows a strong seasonality both in term of quantity and geographical distribution. The rainfall period occurs during austral summer/fall between December and May (Segura et al., 2019), peaking between February and April in relation to the southernmost position of the Intertropical Convergence Zone (ITCZ) (Huaman and Takahashi, 2016; Fig. 2). The rainfall period contributes to around 85% of the annual discharge at El Tigre station (1985–2015 period; using data from SENAMHI - Servicio Nacional de Meteorología e Hidrología -, PEBPT - Proyecto Especial Binacional Puyango Tumbes - and HYBAM - Contrôle géodynamique, hydrologique et biogéochimique de l'érosion/altération et des transferts de matières dans les bassins de l'Amazone, du Congo et de l'Orénoque). The relative contribution of the rainfall amount varies also along the year. While the plutonic/metamorphic domain contribution is almost constant along the year (~30%), the volcanic domain contributes more than ~35% along the September–October-November period and decreases to ~20% during the rest of the year. Therefore, the sedimentary area contributes between ~35% and ~50% of the total amount of rainfall received by the basin during the dry and wet seasons, respectively (Fig. 2).

The main anthropogenic activity is urbanization throughout the city of Tumbes located close to the outlet of the basin, downstream the El Tigre hydrological station (Fig. 1). Small-scale gold mining activity has also been reported upstream in the Puyango and Portovelo-Zaruma sub-basins (Marshall et al., 2018). But, overall, the anthropogenic influence is rather small on sediment production at the Tumbes basin scale.

Section snippets

Studied sites

We analyzed the geochemistry and the mineralogy of the SPM sampled monthly during two contrasted hydrologic years (2007–2008, a wet year, and 2010–2011, a normal hydrological year) at the El Tigre hydrological station located near the Tumbes River outlet (Fig. 1). We compared these data with the daily discharge and the SPM concentration monitored at this station and with the monthly rainfall data recorded over the basin (see next sections for details). To identify potential source effects which

Hydro-sediment budgets

During the sampled years, the daily water discharge corresponding to sampling ranged from 12 m3·s−1 (November 2010) to 736 m3·s−1 (March 2008) and the SPM concentration varies from 2 to 7350 mg·l−1. The latter follows broadly a power law relationship with daily discharge for discharges conditions up from around 40 m3·s−1 (Fig. 4). The SPM concentration remains almost constant for discharge lower than this value (Morera et al., 2017). During the 2007–2008 and the 2010–2011 hydrological years,

Homogenous mineralogical composition of the Tumbes River

The SPM mineralogy values recorded in the Tumbes river and its monolithological sub-basin tributaries are relatively homogenous (Fig. S1, Table S2). This parameter is therefore not discriminant for tracing the source variability of the riverine suspended sediments. Interestingly, the mineralogical composition of the analyzed samples is enriched in illite and depleted in kaolinite and smectite by comparison with those of the Peruvian and Ecuadorian Andino-Amazonian basins. There, kaolinite,

Conclusion

We investigated the geochemistry of the riverine SPM (Suspended Particulate Matter) transported by the Tumbes River at a monthly frequency along two hydrological years, including a wet (2007–2008) and a normal (2010−2011) year. We also analyzed the SPM of 2 monolithological tributaries representative of the lithological diversity of the basin (i.e. volcanic and plutonic/metamorphic) and a sedimentary sub-basins located below a volcanic domain. This constructed geochemical database, never

Acknowledgments

This study was supported by the Peruvian Ministerio de Economía y FinanzasPPR-068 program “Reducción de Vulnerabilidad y Atención de Emergencias por Desastres”, INNÓVATE PERÚ (www.innovateperu.gob.pe) and FONDECYT through the projects “Monitoreo, caracterización identifcación de las principales fuentes de erosión y sedimentos durante fuertes crecidas o eventos extremos El Niño en las cuencas binacionales Puyango-Tumbes y Zarumilla” and “Monitoreo de Sedimentos Ante Riesgos y Desastres

References (54)

  • T. Li et al.

    Sr–Nd isotopic constraints on detrital sediment provenance and paleoenvironmental change in the northern Okinawa Trough during the late Quaternary

    Palaeogeogr. Palaeoclimatol. Palaeoecol.

    (2015)
  • Z. Liu et al.

    Source-to-sink transport processes of fluvial sediments in the South China Sea

    Earth Sci. Rev.

    (2016)
  • J.-S. Moquet et al.

    Chemical weathering and atmospheric/soil CO2 uptake in the Andean and Foreland Amazon basins

    Chem. Geol.

    (2011)
  • J.S. Moquet et al.

    Comparison between silicate weathering and physical erosion rates in Andean basins of Amazon river

    Procedia Earth Plan. Sci.

    (2014)
  • J.-S. Moquet et al.

    Temporal variability and annual budget of inorganic dissolved matter in Andean Pacific Rivers located along a climate gradient from northern Ecuador to southern Peru

    Compt. Rendus Geosci.

    (2018)
  • C. Pin et al.

    Concomitant separation of strontium and samarium-neodymium for isotopic analysis in silicate samples, based on specific extraction chromatography

    Anal. Chim. Acta

    (1994)
  • M. Roddaz et al.

    Sediment provenances and drainage evolution of the Neogene Amazonian foreland basin

    Earth Planet. Sci. Lett.

    (2005)
  • M. Roddaz et al.

    Evidence for the control of the geochemistry of Amazonian floodplain sediments by stratification of suspended sediments in the Amazon

    Chem. Geol.

    (2014)
  • T.C.C. Rousseau et al.

    Controls on the geochemistry of suspended sediments from large tropical South American rivers (Amazon, Orinoco and Maroni)

    Chem. Geol.

    (2019)
  • J. Woodward et al.

    Shifting sediment sources in the world’s longest river: a strontium isotope record for the Holocene Nile

    Quat. Sci. Rev.

    (2015)
  • M.-A. Ancellin et al.

    Across-arc versus along-arc Sr-Nd-Pb isotope variations in the Ecuadorian volcanic arc

    Geochem. Geophys. Geosyst.

    (2017)
  • E. Armijos et al.

    Yields of suspended sediment and dissolved solids from the Andean basins of Ecuador

    Hydrol. Sci. J.

    (2013)
  • J. Bendix et al.

    El Niño meets La Niña—Anomalous rainfall patterns in the “traditional” El Niño region of southern Ecuador

    Erdkunde

    (2011)
  • B.W. Bird et al.

    A 2,300-year-long annually resolved record of the south American summer monsoon from the Peruvian Andes

    Proc. Natl. Acad. Sci.

    (2011)
  • P.E. Biscaye

    Mineralogy and Sedimentation of recent Deep-Sea Clay in the Atlantic Ocean and Adjacent Seas and Oceans

    GSA Bull.

    (1965)
  • J. Bouchez et al.

    Grain size control of river suspended sediment geochemistry: clues from Amazon River depth profiles

    Geochem. Geophys. Geosyst.

    (2011)
  • M.G. Bustamante Rosell et al.

    Holocene changes in monsoon precipitation in the Andes of NE Peru based on δ18O speleothem records

    Quat. Sci. Rev.

    (2016)
  • Cited by (5)

    • Spatial variation in specific sediment yield along the Peruvian western Andes

      2023, Catena
      Citation Excerpt :

      When focusing on regions with few observational data, direct application of geomorphic transport laws is not straightforward. Following the Tropical Andes from north to south, the specific sediment yield ranges from 5 to 1150 t km−2 yr−1 (Moquet et al., 2020; Restrepo et al., 2009; Tenorio et al., 2018; Vanacker et al., 2007). These authors attributed the observed variation in specific sediment yield to multiple environmental factors amongst which mean annual runoff, relief ratio, lithology and catchment area (Restrepo et al., 2006; 2009; Vanacker et al., 2015; Litty et al., 2017) and land use and management (Molina et al., 2008; Restrepo et al., 2015).

    • Tracing impact of El Niño Southern Oscillation on coastal hydrology using coral <sup>87</sup>Sr/<sup>86</sup>Sr record from Lakshadweep, South-Eastern Arabian Sea

      2022, Science of the Total Environment
      Citation Excerpt :

      Proxy derived SST records of Lakshadweep coral reefs reveal the dominant influence of ENSO on SST variability (Sagar et al., 2015). However, only few attempts have been made so far to reconstruct past precipitation changes induced by ENSO events and its impact on regional climatic and hydrological conditions (Aalto et al., 2003; Braganza et al., 2009; Fowler et al., 2012; Moquet et al., 2020). Further, these proxy records do not provide detailed information about the past hydrological changes such as precipitations, continental runoff, salinity, and coastal upwelling.

    • Using neodymium isotope ratio in Ruditapes philippinarum shells for tracking the geographical origin

      2022, Food Chemistry
      Citation Excerpt :

      The neodymium isotopic ratio (143Nd/144Nd or conventionally expressed as ƐNd) is heterogeneously distributed in the global ocean because of the short residence time of Nd in the ocean (500–700 years) (Rempfer et al., 2011; Tachikawa et al., 2003). In particular, the distribution of 143Nd/144Nd in coastal waters and sediments displays a distinct and systematic variability, largely owing to riverine Nd supply from the catchment areas (Laukert et al., 2017; Moquet et al., 2020; Rousseau et al., 2015). Notably, no measurable fractionation of Nd isotopes occurs during biological uptake and subsequent incorporation into marine biogenic carbonates.

    • Provenance constraints on the Cretaceous-Paleocene erosional history of the Guiana Shield as determined from the geochemistry of clay-size fraction of sediments from the Arapaima-1 well (Guyana-Suriname basin)

      2021, Marine Geology
      Citation Excerpt :

      Many tools exist to determine the provenance of sedimentary rocks including sandstone petrography, clay mineralogy, heavy mineral analyses, geochronological dating on accessory minerals, major and trace element geochemistry and Sr-Sm-Nd isotopic compositions of bulk sedimentary rocks (see for instance Basu et al., 1990; Caracciolo, 2020; Cullers, 2000; Dickinson, 1985; Dickinson and Suczek, 1979; Garzanti, 2016, 2019; Garzanti and Andò, 2019; McLennan et al., 1993; Weltje, 2012 and references therein). All of these tools are subjected to biases related to source-rock mineral fertility (see for instance Chew et al., 2020; Dickinson, 2008; Malusà et al., 2016, Malusà et al., 2013), pre and post depositional processes including weathering, grain-size sorting and diagenesis (see for instance (Bouchez et al., 2011; Fedo et al., 1995; Garzanti et al., 2008, 2009, 2014; Lawrence et al., 2011; McLennan et al., 1993; Moquet et al., 2020; Nesbitt et al., 1996; Roddaz et al., 2006, 2014; Rousseau et al., 2019; Viers et al., 2008) and analytical procedures (Andersen et al., 2019; Chew et al., 2020; Malusà et al., 2013; Vermeesch, 2018; Vermeesch, 2004). In addition, the low amount of material available for well or core sample makes geochronological dating methods (especially U-Pb dating on zircon) difficult to use.

    View full text