Page 56

6.2 - OXYGEN ISOTOPE STAGE 4 (75-64 kabp)
No member of Unit 2 was directly dated to OIS 4. If
Unit 1 (undated) belongs to OIS 5 and the base of Unit
2 to 55 ka B.P., there are no dated sediments belonging
to OIS 4. Alternatively, Unit 1 may be OIS 4, a climatic
period known to be cool and dry in the Mediterranean
and the Atlantic (Heinrich 6), (Bond et al., 1993).
6.3 - OXYGEN ISOTOPE STAGE 3 (64-32 kabp)
The main aggradational phase, Unit 2, consisting of
repeated cyclothemic sequences only slightly disturbed
by minor erosional phases, occurred during this
Stage.
This aggradational phase begins at ca. 55 ka with
travertine deposits on gravels (layer 3), and the main
silty sand layers commence after ca. 49 ka B.P. (layer
no6), with an uniform suspended load of tractive flow.
Minor erosional phases, characterised by shallow small
channel or gully features, occur in this lower part of
Unit 2. They include colluvium and laterally derived
slope debris (layer no8) suggesting that glacis processes
were active on the lower valley side slopes,
transferring sediments to the valley floor. Such sheetflood
processes may have been associated with a drier
environment, but the duration of this episode remains
still unknown.
Above layer no8 occurs the most important
aggradational sequence, 12,50 m thick, with dates from
about 45 to 30 ka ( layers 9 to 28). This phase saw an increase
of fine sediment supply, mostly between ca.37
and 33 ka. Sedimentological records and faunistic remnants
(mostly of Bovini) indicate a local open biostasic
environment, like a steppic prairie or a dry-forest
(Wengler et al., 2002). This is probably one of the most
important palaeoclimatic inferences from this site, indicating
a more frequent occurrence of subtropical
rains in the desert margin (Rognon, 1996). An important
alluviation phase with this age is widely reported
from the Mediterranean basin (Besançon et al., 1997;
Macklin et al., 2002), but with an opposite, rhexistasic,
process significance.
The upper part of Unit 2 shows several small meandering
channel fills containing small pebbles, calcreted
root remnants and oncoliths, starting with layer no29.
Channel fill layer no31 gives dates of ca. 30,9 - 27,8 ka
B.P., dates comparable with those for channel gravels
in the Soltanian silts in the Anti-Atlas valleys further
east (Thorp et al., 2002). It may coïncide with Heinrich
event 3. The upper part of the sequence, layer numbers
29 - 32, therefore shows more variability in sedimentation
but not of such a degree as to suggest any major climate
change: tree charcoal, bones of Bovini andMiddle
Paleolithic tools are still present as they are in the layers
below. Recurrent travertine constructions and silt
deposits initiated by small shallow wandering gravel
bedded channels continue at a high aggraded valley
level during the 30 – ca. 24 ka period, suggesting little
significant change.
6.4 - OXYGEN ISOTOPE STAGE 2 (32-13 kabp)
This stage begins with a small channel (layer no33
eroded into layer no32). It is similar to layer no31 but
dates to ca. 24 - 20 ka and possibly coïncides with
Heinrich event 2. Further silt alluviation layers, numbers
34 and 35, overly these channels in the floodplain,
and pebbly bars of reworked coarse material appeared
close to the valley slopes. The top of this sedimentary
aggradation was here reached at ca. 20 ka B.P., as in
other parts of SouthMorocco (Boudad et al., 2003), but
not in the same catchment further east where it continues
till ca 11 ka B.P. (Andres, 1977; Thorp, in litteris).
The last phase of OIS 2 sedimentation, Unit 3, layer
no36, consists of both active erosion and deposition
and indicates the end of the main cyclothemic
aggradational phase. Thick sandy and silty fans,
sourced from the glacis slopes of the tributary catchments,
and also containing aeolian quartz grains and
palygorskyte, were deposited at their confluences with
the main valley after 20 ka B.P. These latest sediments
repose on the surface of layer no35 with a date of ca.
18,5 ka B.P. 2 km upstream, the base of this colluvial
fan aggradation is dated at 17 ka B.P (OSL date communicated
by N. Mercier), and at ca. 14 ka further
north in the Anti-Atlas, (Weisrock et al. , 1994). At this
time it is possible that an increase in aridity and a reduction
in vegetation cover led to increase soil erosion
by sheet-floods over the catchment slopes. On the valley
floor in Oued Noun, the gueltas became shallower,
more restricted and finally dried out, probably by 12 or
10 ka B.P.
At the end of OIS 2 and the beginning of OIS 1,
ca 10 ka B.P., deep linear incision of the whole infilling
commenced (Mathieu et al., 2004).
7 - CONCLUSIONS
7.1. Despite two radiometric ages that are not in
chronostratigraphical order, comparative dating methods
tested on the same section and even on the same layers
give generally consistent results. All but 2 of the 20 dates
lie in the range 55,9 – 18,5 ka B.P., that is OIS stages 3
and 2. Apart from the two out of sequence dates of 92,2
and 87,6 ka B.P., there are either no sediments or no dates
for OIS stages 4 and 5. The basal sedimentary Unit 1 consists
of a coarse cobble fluvial gravel but remains
undated.
7.2. The main unit (Unit 2) of the Oued Noun Soltanian
formation was deposited between ca.55 and 25 ka B.P.
The most important part of this silty aggradation (more
than 12,5 m thick) occured during OIS.3, between ca. 40
and 30 ka B.P. It comprises repeated sequences of shallow
channels, their small gravel and sands infills, their
travertine caps and overlying silts. Some of the silts contain
aeolian grains, some were deposited in former
gueltas, and most were fluvial overbank sediments. This
period was, in this area at the northern boundary of the
Sahara, a biostasic phase, with a more regular water