Geochemical Fingerprinting - Application of Geochemical Tracers to Fluvial Sediment - page 16

area is modified by the selective entrainment (erosion), transport, and deposition of
grains as they are dispersed through the system according to their size, density, and
shape (Knighton 1998 ; Miller and Orbock Miller 2007 ). Mechanical and chemical
weathering processes also lead tomodifications in the initial grain population. The net
effect of these processes is that sediment of different size, shape, and density within
the river is transported downstream at different rates, often producing a downstream
fining in particle size. Sediment also may be transported by different methods (e.g.,
by suspended and bedload process) (Weltje 2012 ), or be partitioned by the flow into
distinct depositional units at a given site (e.g., pools, riffles, point bars, floodplains,
etc.) (Miller and Orbock Miller 2007 ). The river and source area sediments, then,
may represent two very different populations of particles. Suspended sediments, for
example, may only represent the finest materials within the source areas, and their
mineralogy would be expected to differ from that of the bulk material. Moreover, the
geochemical properties of the river and source area sediments are likely to differ as
fine-grained particles characterized by large surface areas and high surface charge
tend to be more reactive and have a greater potential to collect, concentrate, and
retain ions (e.g., trace metals).
Modification of the source area sediments during dispersal by physical and chem-
ical processes is important because an assumption inherent in inverse modeling is
that the physical and geochemical composition of river sediment differs from a spe-
cific source area only because it has been mixed with sediment from another source
area(s). Thus, physical and chemical modifications of the source area sediments dur-
ing erosion, transport and/or deposition must be eliminated, or at least limited, to
effectively use inverse modeling. Provenance studies, particularly those aimed at
determining the provenance of sediments in stratified rocks, often deal with these
modifications using a concept that Weltje ( 2004 ) referred to as transport invariance.
The concept assumes that particles with similar sizes, shapes, and densities will be
entrained, transported and deposited under similar conditions. Thus, by comparing
particles from the source areas and the river that fall within a narrowly defined range of
sediment size, density or shape, the potential, transport-related modifications can be
reduced, and the composition of the river sediment will primarily reflect the relative
mixing of sediment from the various source areas. As we will see below, approaches
other than the analysis of transport invariant populations have also been proposed
to account for the physical and geochemical modification of the source sediments.
The point to be made here is that a determination of the provenance of the bulk
sediment (consisting of a wide range of particle sizes) may require the combined
analysis of multiple size ranges. In fact, it is quite possible that the predominant
source(s) of sediment found within the river may vary as a function of particle size
(Miller et al. 2013 ). Sandstone strata, or the soils developed within it, for exam-
ple, are likely to contribute more sand-sized sediment to a river than a shale and
its associated soils. Determining the provenance of multiple size fractions can be
time consuming and expensive. Thus, most studies of sediment provenance focus
on the particle size fraction that is of importance to the question at hand. For the
majority of the geochemical fingerprinting studies, the focus has been on relatively
fine-grained sediment (

< µ

63 m) as it is this size fraction that forms a significant