Page 51

turbulence. These could be interpreted as deposition
less distal in relation to the flood flow than the other
samples, and from a flow that had higher turbulence
than the others. The samples come from two within
channel sedimentary bodies (31 and 33) in the upper
part of Unit 2.
Group 2, fine sandy silts, the finest samples, possesses
an M range of 2-14 microns and a Cu range of
45-82 microns. These would be classed as type VIII,
finest “uniform” (sediment distributed through the
whole depth of flow) suspension from a non-turbulent,
perhaps non flowing body of water such as a shallow
lake. Such a lake might form in a pre existing guelta or
lie at the distal end of a waning flood, perhaps dammed
behind a dune or a downstream sill of calcrete and travertine.
Morphomicroscopy of some of the grains may
have been Aeolian derived from reworking of earlier or
adjacent fine silts and sands. All these samples came
from layers that belong to the middle, dominantly silty
fine sand part of Unit 2, except no 32b which came from
the upper part.
The remaining samples fall into Passega’s classes VI,
VII and VIII of “uniform” suspensions. Type VII samples
towards the upper right in this group, 31-2g and
35, would be classed as a low turbulence “uniform”
suspension either within a channel towards the end in
time of a flood, or laterally on the adjacent flood plain;
35 would indicate moderate turbulence. They contain
some medium sized sands that might have been part of
a far travelled or local tributary derived load. The remainder
in this group, those mid to bottom right with
Cus above 100 microns – are mainly type VIII – fine
“uniform” suspension in flows with little turbulence,
probably deposited from slowly moving water onto
floodplains, marshes and swamps or in secondary
abandoned channels.
These Group 2 samples came from the fine-grained
silty fine sandy, clayey deposits making up the bulk of
the main middle portion of Unit 2. There are numerous
Middle Palaeolithic tools and bones through these sediments
suggesting habited, moist, vegetated environments.
This group also includes layer 36 from the non-
Soltanian top layers.
The genesis of Unit 2.
1) The lower part of Unit 2 (layers 2 – 8) contains highenergy
deposits, beds of suspension load sediments
and several interbedded travertines and calcretes
(no3, 5 and 3-5 on the left bank of the wadi). A conspicuous
feature is an irregular layer of angular schist
gravels in a reddish-yellow silty matrix (layer no8)
derived from local hill slopes, that marks a specific
but short lived input during this longer period of
aggradation.
2) In its middle part (layers 9 – 28), Unit 2 begins with
an “uniform” suspended load deposit possibly
containing aeolian transported fines (no9). As aggradation
proceeded, the floodplain became wider.
During this phase, large floods carrying suspended
silt loads reached this part of the valley:
a) bed nos 10, 18 and 26 are shallow channel infills
with massive structure and sometimes small gravels
(no26) indicating sediment transport by
traction with an “uniform” suspended load under
upper regime flow. They are interpreted as
flood deposits in active channels.
b) bed nos12a, 20a, 28a are thicker ( 2,50, 1 and
2,30 m respectively) and the CM plots suggest
deposition from uniform suspended load during
periods of maximum overbank inundation. However,
in the case of 12a it is likely that the coarser
fraction may have a wind blown source. In
the narrower sections of the valley gorge, dams
from earlier sediments or temporary dunes may
have occured with subsequent extension of
marshes and gueltas.
c) bed nos 12b to 17a, 20b, 32b are thin, show lamellar
structure and often contain clay. They
were probably laid down in abandoned channels
and gueltas from the settling out of suspension
load under slack water conditions.
In this middle part of Unit 2, the successive shallow
channel-, overbank- and floodplain-deposits
represent three repeated aggradation sequences.
Flows were generally valley floor wide and had
differing discharges, velocities, turbulences and
sediment sourcing. Some may have been near the
end of the downvalley flow extent; other may
have proceeded further down gorge. The flows
were generally very turbid despite sedimentation
that must have occured in the Guelmim basin
upstream of the gorge. We may suppose that during
at least three periods, the valley floor saw a
succession of shallow palaeo-gueltas.
3) In its upper part (layers 29 – 35), Unit 2 shows an increase
of gravelly and sandy deposits (no31, 33 and
even 35) oversealed by travertines.We suggest these
indicate a new sedimentary style associated with
shifting gravel-bed channels towards the end and on
the top of themain phase of the silty aggradation and
especially near the flanking valley slopes. Groundwater
levels, however, remained high and some very
fine slack water suspension load beds were laid
down (e.g. no32). In some places the fine gravels
may have been locally reworked as discontinuous
“shorelines” around palaeo-gueltas.
Other morphological and chemical evidence for
high groundwater table levels are CaCO3 indurations
of the steep faces of beds 12 and 28, gypsum and salt
crystallisations in numerous beds and hydromorphic
features; calified bioliths of root channels indicate
vegetation in moist soils. We may suppose that Unit 2
is a record of i) flood events at the end of the flow extent,
probably widely separated in time, and ii) almost
continuous sedimentation in the gueltas, albeit with a
very low rate, between the flood events.