Author Posting. © The Author(s), 2018. This is the author's version of the work. It is posted here under a nonexclusive, irrevocable, paid-up, worldwide license granted to WHOI. It is made available for personal use, not for redistribution. The definitive version was published in Geochimica et Cosmochimica Acta 233 (2018): 115-134, doi:10.1016/j.gca.2018.04.035.
The dependence of thorium scavenging by particles on particle composition is examined at
selected stations of the U.S. GEOTRACES North Atlantic Section (GA03). Scavenging is here
described by the apparent, first-order rate constant of Th adsorption onto particles (k1), as estimated from an inversion of Th radioisotope and radioactive parent data. Our k1 estimates are
regressed against particle phase data using two different models. Model I considers biogenic
particles (POC+PIC+bSi), lithogenic particles, Mn (oxyhydr)oxides, and Fe (oxyhydr)oxides
as regressors, and k1 as the regressand. Model II considers ln(POC+PIC+bSi), ln(lithogenic
particles), ln(Mn (oxyhydr)oxides), and ln(Fe (oxyhydr)oxides) as regressors, and ln(k1) as
the regressand, where ln() denotes the natural logarithm. Thus, models I and II posit that the
effects of particle phases on k1 are, respectively, additive and multiplicative. These models are
applied to three groups of stations: (i) all selected stations, (ii) stations west of theMauritanian
upwelling region (“western stations”), and (iii) stations within that region (“eastern stations”).
We find that model II appears to better describe the effect of particle composition on k1 than
model I. Particle composition explains a larger fraction of the variance of k1 for the eastern
stations (R2 = 0.60 for model I and 0.67 for model II) than for the western stations (R2 = 0.26
for model I and 0.39 for model II). When considering all stations, the variance of k1 explained
by particle composition is intermediate (R2 = 0.50 for model I and 0.51 for model II). According to model II, the variance of k1 explained by particle composition is predominantly due
to biogenic particles at the eastern stations and to Mn (oxyhydr)oxides at the western stations.
Additionally, we find that particle composition does not explain a significantly different proportion of variance of k1 than particle concentration. It is thus concluded that, at our selected
stations, (i) biogenic particles andMn (oxyhydr)oxides more strongly influence Th scavenging
than any other phases considered, and (ii) particle composition and particle concentration have
comparable effects on this process.
We acknowledge the U.S. National Science Foundation for supporting this study (grant OCE-1232578) and the U.S. GEOTRACES North Atlantic section ship time, sampling, and data analysis.
Woods Hole Open Access Server