Role of Source, Mineralogy, and Organic Complexation on Lability and Fe Isotopic Composition of Terrestrial Fe sources to the Gulf of Alaska

Abstract

Iron (Fe) is a key trace nutrient supporting marine primary production, and its deposition in the surface ocean can impact multiple biogeochemical cycles. Understanding Fe cycling in the subarctic is key for tracking the fate of particulate-bound sources of oceans in a changing climate. Recently, Fe isotope ratios have been proposed as a potential tool to trace sources of Fe to the marine environment. Here, we investigate the Fe isotopic composition of terrestrial sources of Fe including glacial sediment, loess, volcanic ash, and wildfire aerosols, all from Alaska. Results show that the δ⁵⁶Fe values of glaciofluvial silt, glacial dissolved load, volcanic ash, and wildfire aerosols fall in a restricted range of δ⁵⁶Fe values from -0.02 to +0.12‰, in contrast to the broader range of Fe isotopic compositions observed in loess, -0.50 to +0.13‰. The Fe isotopic composition of the dissolved load of glacial meltwater was consistently lighter compared to its particulate counterpart. The 'aging' (exposure to environmental conditions) of volcanic ash did not significantly fractionate the Fe isotopic composition. The Fe isotopic composition of wildfire aerosols collected during an active fire season in Alaska in the summer of 2019 was not significantly fractionated from those of the average upper continental crust composition. We find that the δ⁵⁶Fe values of loess (<5 μm fraction) were more negative (-0.32 to +0.05‰) with respect to all samples measured here, had the highest proportion of easily reducible Fe (5.9-59.6%), and were correlated with the degree of chemical weathering and organic matter content. Transmission electron spectroscopy measurements indicate an accumulation of amorphous Fe phases in the loess. Our results indicate that Fe isotopes can be related to Fe lability when in the presence of organic matter and that higher organic matter content is associated with a distinctly more negative Fe isotope signature likely due to Fe-organic complexation.

Figure 1

Map indicating the locations of samples analyzed in this study, including from southcentral Alaska, USA, and Yukon Territory, Canada. Map generated using GMTApp. The sample sites of suspended particulate are the same as those of dissolved load.

Figure 2

Stable Fe isotopic compositions of potential Fe sources to the Gulf of Alaska (note the loess shown here is only from the river valley loess profiles). Error bars show the 95% confidence intervals.

Figure 3

Fe isotopic compositions of paired glaciofluvial suspended particulate matter and dissolved load by sample location (see Table S1 for full sample names and locations). The solid lines represent the mean values. Error bars for Fe isotopic ratios determined in this study are long-term precision. The shaded area represents the 95% confidence intervals of the mean value.

Figure 4

(A) Fe isotopic compositions of different size fractions (colored circles) and loss of ignition (LOI) values of loess (gray hexagons) from the Kanakanak bluffs, Alaska. The δ⁵⁶Fe isotopic compositions of three size fractions of loess were measured at six different depths. Error bars show the 95% confidence intervals. The lightest δ⁵⁶Fe values were observed at ∼1.5 m depth, highlighted in brown in panel (A), where oxidized Fe may have accumulated. The light gray shaded area in panel (A) represents a possible paleosol layer, characterized by relatively high organic matter. The white circles in panel (B) represent the sample depths.

Figure 5

Stable Fe isotopic compositions of glaciofluvial silts and river valley loess (all <5 μm size fraction) correlated to Al/K mass ratios (m/m). Major element data are from Koffman et al. Symbols are the same as in Figure 2. Error bars for iron isotopic ratios determined in this study are 95% confidence intervals. The black solid line is the best linear fit for Fe isotopes and Al/K ratio with a 95% confidence envelope (dark pink area) and 95% prediction envelope (light pink area) excluding three loess samples with particularly low δ⁵⁶Fe as these are hypothesized to have Fe fractionation due to the presence of organic carbon.

Figure 6

TEM images of well-crystallized Fe phases of a representative glaciofluvial silt sample AK2016-03 from Knik River in the <5 μm size fraction. TEM images (A,B) show ∼1 μm clay particle sheets. The diffraction pattern in (C) shows that this particle is crystalline. X-ray spectra (D–G) show that it contains Fe, Mg, Si, and O within the crystalline structure.

Figure 7

TEM images of amorphous Fe phases of a representative loess sample AK2016-13 from Chitina River in the <5 μm size fraction. TEM images (A, B) show ∼1 μm clay particles. The diffraction pattern in (C) shows that this particle is poorly crystalline or amorphous, and X-ray spectra (D–G) show that it contains Fe, Mg, C, and O within a poorly crystalline or amorphous structure.

Figure 8

Variation of the δ⁵⁶Fe values of glaciofluvial silt, loess, and ash correlated to the LOI data. The gray bar represents the δ⁵⁶Fe values of upper continental crust. The loess shown here is fine-grained (<5 μm) from the river valley loess profiles.