IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.344.209.2018

Introduction

The objective of the Costa Rica Seismogenesis Project (CRISP) undertaken by Expeditions 334 and 344 was to understand the processes controlling fault zone behavior during earthquake nucleation and rupture propagation at erosional subduction zones. Offshore Costa Rica, the erosional subduction zone is formed by the incoming oceanic Cocos plate subducting under the Caribbean plate. The CRISP study area is located offshore the Osa Peninsula, an area characterized by low sediment supply, fast convergence rates, abundant plate interface seismicity, and variable relief of the subducting plate along strike (see the Expedition 344 summary chapter [Harris et al., 2013a]). An important component to evaluating processes occurring at erosional subduction zones is the geochemical characterization of the interstitial pore fluids, which may yield insights to the changes in fluid-rock interactions occurring in the depths of seismogenesis. Additionally, important feedback mechanisms exist between chemical reactions in the sediments and the hydrologic characteristics of the sediments including porosity, permeability, and fluid pressure, which in turn influence the mechanical state of the plate interface at depth.

Rare earth elements (REEs) provide one tool to characterize the geochemistry of the interstitial pore fluids. REEs are categorized as the 14 naturally occurring members of the lanthanide series that share the same valence electron shell and thus display similar group chemical properties and characteristics. Importantly, however, REEs differentiate subtly in chemical reactivity as the ionic radius contracts with increasing atomic number. This slight differentiation leads to consistent and predictable fractions between the informally designated light (La to Nd), middle (Sm to Dy), and heavy (Ho to Lu) REEs during physical, chemical, and biological processes (Byrne and Li, 1995; Byrne and Sholkovitz, 1996; Elderfield and Greaves, 1982). To evaluate these relative fractionations among REEs, the standard practice is to normalize the concentrations to a reference material. For oceanic studies, the majority of REEs are sourced from continental weathering and delivery by river systems; therefore, Post-Archean Australian Shale (PAAS) is often used as the reference material (Elderfield et al., 1990; Pourmand et al., 2012; Sholkovitz et al., 1999; Taylor et al., 1981). Normalized to PAAS, dissolved oceanic REEs generally display enrichment of heavy REEs relative to middle and light REEs (e.g., De Baar et al., 1985; Elderfield and Greaves, 1982; Sholkovitz et al., 1994). This signature develops through the greater affinity of light REEs to adsorb onto marine particles and the stronger complexation of heavy REEs to dissolved carbonate, silicate, or dissolved organic compounds (e.g., Akagi, 2013; Bau and Koschinsky, 2009; Byrne and Kim, 1990; Byrne and Li, 1995). Additionally, whereas dissolved REEs exist dominantly in a +3 oxidation state, cerium (Ce) can be oxidized through microbial activity to a +4 state that is more particle reactive, thus removing Ce from solution at a greater rate and creating a negative Ce anomaly relative to its neighboring REEs (e.g., Bau and Koschinsky, 2009; Moffett, 1990; Sholkovitz and Schneider, 1991). Europium (Eu) may also exist in a reduced +2 oxidation state, though this state is typically only observed in high-temperature reducing conditions such as those found at hydrothermal vents (e.g., Elderfield, 1988; Klinkhammer et al., 1983).

In marine pore fluids, the signatures of dissolved REEs have been documented and associated with various diagenetic reactions in the sediments, although understanding the diagenetic cycling of REEs is still incomplete (e.g., Abbott et al., 2015; Haley et al., 2004; Kim et al., 2012; Sholkovitz et al., 1989). In the oxic zone of the sediments, pore fluid REE measurements attributable to particulate organic carbon (POC) show linear enrichment of heavy REEs relative to light REEs, reflecting remineralization from POC and complexation of heavy REEs to dissolved organic carbon (DOC) (Haley et al., 2004). Conversely, measurements of fresh, labile organic matter in the sediments that had been isolated through an H2O2 extraction have shown middle-REE enrichment (Freslon et al., 2014). In the ferruginous zone of the sediments, where the respiration of organic matter continues through the reductive dissolution of Fe oxides, a characteristic middle-REE enriched signal in the pore fluids has been observed by several studies (Abbott et al., 2015; Du et al., 2016; Haley et al., 2004). Although there has been some debate as to the mechanisms behind the middle-REE enrichment, Fe oxide phases isolated through hydroxylamine and dithionite extractions have shown similar middle-REE enrichments, suggesting that the pore fluid signatures in this zone are a direct consequence of Fe oxide dissolution (Abbott et al., 2016; Du et al., 2016). In the methanogenic zone of the sediment column, dissolved REEs in the pore fluids take on a heavy-REE enriched pattern (Kim et al., 2012; Soyol-Erdene and Huh, 2013). A recent study of methanotrophs in the Deepwater Horizon oil spill has demonstrated that light REEs (La, Ce, Pr, and Nd) are important cofactors in methanol dehydrogenase, providing a mechanism for a pore fluid signal depleted in light REEs relative to heavy REEs in methanogenic sediments (Shiller et al., 2017).

For the CRISP sites, we measured the pore fluid signals at three different locations and from four different cores: Sites U1378/U1380 from the middle slope on the continental margin, Site U1381 from the incoming plate, and Site U1414, also from the incoming plate (Fig. F1). We used a relatively novel sample preconcentration system (seaFAST2 from Elemental Scientific, Inc., Omaha, NE) with direct introduction to an inductively coupled plasma–mass spectrometer (ICP-MS) to measure REEs in these pore fluids.