Desert rivers - Climatic Change in Deserts - page 177

have not pursued this idea, but it is not without merit (see Wasson, 1982 ). In the
semi-arid Flinders Ranges of South Australia, the fine-grained late Pleistocene valley
fills represent reworked eolian dust or loess mantles (Williams et al., 2001 ; Chor et al.,
2003 ; Williams and Nitschke, 2005 ; Williams et al., 2006a ; Williams and Adamson,
2008 ; Haberlah et al., 2010a ; Haberlah et al., 2010b ). Similar valley fills derived from
reworked loess are a feature of the Matmata limestone uplands in Tunisia (Coude-
Gaussen et al., 1987 ), the Namib piedmont valleys (Eitel et al., 2001 ; Heine and Heine,
2002 ;Eiteletal., 2005 ) and the wadis within the Sinai Desert (Rogner et al., 2004 ).
Aword of caution is needed here in relation to changes from braided to meandering
stream patterns. It is widely assumed in much of the geomorphic literature relating to
meandering and braided river channels that there are distinct hydrologic thresholds
leading from one state to another. This is a view challenged by Tooth and Nanson
( 2004 ) as a result of their work on the Plenty and Marshall rivers in arid central
Australia. Both rivers flow close to one another on roughly parallel courses in one
sector of 70 km and have similar gradients and local climates. Despite this, and
contrary to accepted geomorphic theory, they display very different channel patterns
and cross-sectional forms. The Plenty River in the reach studied is a single-thread, low-
sinuosity channel up to 2 m deep and 1,200 m wide. In strong contrast, the Marshall
River has numerous narrow anabranches, usually less than 60 m wide, separated by
vegetated ridges and broader islands. Arising from their field investigations, Tooth and
Nanson ( 2004 ) concluded that there were two main reasons for this difference, both
of which are quite subtle. The Marshall River carries a slightly coarser bed load and
receives occasional tributaries, enabling tree growth and accumulation of sediment
immediately upstream of these trees, leading to anabranch formation on either side of
the obstructions. Hence, minor differences in sediment and water inflow to the trunk
stream channel can lead to major changes in stream channel morphology.
10.7 Quaternary Blue Nile paleochannels on the Gezira plain,
semi-arid Sudan
In this section, we focus in some detail on the Nile, not only because it has been intens-
ively studied over many decades (Lombardini, 1865a ; Lombardini, 1865b ; Willcocks,
1904 ; Lyons, 1906 ; Lawson, 1927 ; Hurst and Philips, 1931 ; Hurst and Phillips, 1938 ;
Hurst, 1952 ; Williams and Faure, 1980 ; Williams and Adamson, 1982 ; Hassan, 1981 ;
Said, 1993 ;Said, 1997 ; Williams et al., 2000 ; Woodward et al., 2001 ; Woodward
et al., 2007 ; Williams, 2009b ; Williams et al., 2010b ; Williams, 2012a ), but espe-
cially because it illustrates very nicely how different lines of evidence can be used to
construct a coherent history of river response to environmental change in arid areas.
The Nile is the longest river in the world and carries a sediment load of about
100 million tonnes/year, most of which comes from its Ethiopian headwaters, as
Herodotus correctly surmised some 2,500 years ago. The two major Ethiopian