Geochemical Fingerprinting - Application of Geochemical Tracers to Fluvial Sediment - page 19

gullies (Fox and Papanicolaou 2008b ). Thus, spatially defined sources fail to directly
identify the geomorphic processes responsible for sediment generation. Second, soil
types and land-use/land-cover categories are often transitional to one another, con-
founding their spatial delineation within the catchment as well as the geochemical
differences in their sediments (Rowan et al. 2012 ). Differences in the underlying
geological units may also complicate the issue. Third, recent changes from one land-
use/land cover type to another may limit the ability of geochemical parameters to
distinguish between sediment source areas (Miller et al. 2013 ). In other words, the
geochemistry of the sediment sources may reflect both its current and past land
cover history, making it difficult to distinguish between sediments associated with
the various land-use/land-cover categories.
In light of the above, an alternative method of defining sediment sources is by
the erosional process through which the sediments are generated and delivered to
the river. Referred to as the 'source type', a distinction is most often made between
sediments generated near or at the ground surface in upland (hillslope) areas by sheet
or rill erosion and sediment derived from the 'subsurface' by means of gully or bank
erosion (Fig. 2.3 ) (Walling and Peart 1979 ; Gellis et al. 2009 ; Gellis andWalling 2011 ;
Massoudieh et al. 2013 ). Differentiation between surface and subsurface sediments
requires the use of geochemical parameters that differ as a function of depth below
the ground surface, such as organic matter or short-lived radionuclides (described in
the next chapter).
As neither the spatial or process approach to defining sediment sources is ideal
on its own, it is not uncommon for investigators to combined the two methods,
thereby defining sediment sources on the basis of both spatial and type categories
(Russell et al. 2001 ; Juracek and Ziegler 2009 ; Wilkinson et al. 2009 ), particularly
for catchments less than about 200km 2 (Mukundan et al. 2012 ). Within larger catch-
ments, the heterogeneity of sediment source properties defined by land-use, soil type,
or geomorphic process is likely to increase, making it more difficult to distinguish
between the sources and hindering source ascription (Collins et al. 1998 ). In addition,
sediment contributions from relatively minor sources, which may still cover large
areas, could be underestimated (Mukundan et al. 2012 ). As a result, the application
of geochemical fingerprinting methods to large basins (

500km 2 ) is more difficult,
although a number of studies have shown that sediment sources may be effectively
defined according to the underlying geological units within the catchment (Walling
et al. 1999 ; Bottrill et al. 2000 ; Douglas et al. 2003 ) or by tributary catchment areas
(Collins et al. 1996 ; Walling et al. 1999 ), both of which tend to exhibit less property
heterogeneity than sediments defined according to land-use, soil type, or erosion
process (Collins et al. 2012 ; Mukundan et al. 2012 ; Wilkinson et al. 2013 ).

>

2.3.2 Collection and Characterization of River Sediment

A wide range of river sediments have been targeted for geochemical fingerprin-
ting (Fig. 2.4 ). The sediments which are selected dictate the timeframe under