089E0377.TXT

PL.LOUVAIN 89E SERVICE GEOLOGIQUE DE BELGIQUE

377 Bulletin de la Société belge de Géologie T. 91
fasc. 1 pp. 61-69 Bruxelles 1982

Bulletin van de Belgische Vereniging voor Geologie V. 91
deel 1 blz. 61-69 Brussel 1982

A PLIOCENE RIVER DEPOSIT IN MID-BELGIUM

by Dirk Goossens (*)

(*) Aspirant N.F.W.O., Instituut voor Aardewetenschappen, K.U.L.,
16 bis, Redingenstraat, B-3000 Leuven.

ABSTRACT - In the summer of 1979, a Pliocene continental
formation was discovered on the Gasthuisberg, a Diestien hill of
Louvain. An examination of the profiles and a sedimentological
analysis of the material (chiefly coarse sand and gravel)
leads to the conclusion that the formation was deposited by
an important river. Because of the deep-red and very strong
weathering of the material, the structure of the deposit
(clearly pointing to a braided river), and the topographical
position (at the summit of a hill), we consider the formation
to be deposited during the Pliocene.
This is believed to be the first paper where unmistakable
Pliocene continental formations in Belgium are described.
Therefore, we propose the name formation of Gasthuisberg for
the continental facies of Pliocene Belgian deposits.

INTRODUCTION.

The pliocene deposits thusfar in Belgium described, are all
sea deposits. Pliocene continental deposits were until
recently in Belgium unknown. There is no certainty about the
exact age of the ONX-gravel (either Late Pliocene or Young
Pleistocene), but there seems to be a tendency to consider
them to the Pleistocene rather than to the Pliocene.
In the Ardennes there have been found remnants of a very old
weathering zone (the so-called rubified horizonts),
but these rubified horizonts are developed in top of the
Paleozoic formations, and do not form a proper deposit.

As a result of the construction of the A2 highway,
important ground-works were carried out S.W. of Louvain at the
end of 1979. In one of the excavations we discovered an important
gully, which was developed on the top of the Diestian, and which
was completely covered with Late Pleistocene Weichselian loess.
The sediments of this gully were most probably deposited
during Pliocene times, for reasons which shall be discussed
below. They show remarkable and very striking structures.

Several samples of the deposit were taken by the Centre de
Physique du Globe (Dourbes) for paleomagnetical examination.
The results of this examination are not yet known;
consequently, in this paper the interpretation of the river
deposit will be limited to a simple lithostratigraphical
interpretation. The magnetostratigraphical interpretation
will be published later.

It has to be noticed that as a result of ground-works in
November 1979, nearly the whole deposit has been dug away.
Only at a few places is it still perceptible.


SITUATION.

The excavation where the gully (we called it "B.H. Gasthuisberg
gully") was discovered, is situated very close to the summit of
the Gasthuisberg, a S.W. - N.E. oriented hill S.W. of Louvain.
Fig. 1 gives the exact location.

The Gasthuisberg forms part of the southernmost Diestian hill
of the Hageland. As can be show in fig.1, this Diestian hill
is bordered in the south by the depressions of the Voer
(west of the Dijle) and the Molenbeek (east of the Dijle).
The Dijle itself has eroded a wide gap into the hill, thus
dividing it into two portions. The gully was found at an absolute
height of about 73 m; this is 50 m above the actual alluvial
plain of the Dijle.

DESCRIPTION OF THE PROFILES.

In the excavation, a pit in which a bridge was constructed,
the river gully could be studied by two walls standing
perpendicular to each : a western wall and a northern wall
(see fig. 2) In consequence of this it was possible to
determine the exact orientation of the river-bed.
This orientation seemed to be due N-S. The zones of the
western wall are in general also found in the northern wall ;
some zones however are restrictes either to the western, or
the northern wall. The numbering of the zones, which probably
can come over somewhat confused, has been influenced by this.

1. THE WESTERN WALL.

The western wall gives a nice profile trough the main channel
of the river deposit. See fig. 3. From the bottom upward,
we distinguish the following zones :

Zone 1 : coarse green (2,5 Y 4/6) sands, very rich in glauconite.
Close by the gully, the colour becomes browner (2,5 Y 4/4).
The sands of this zone are the wellknown sands of Diest, which
occur at the top of the Diestian hills of the Hageland.

Zone 8 : a small zone in the Diestian sand, which is not
coloured greenish-brown, but cleary yellowish-brown (2,5 Y 5/6).

Zone 7 : the basic gravel of the gully. We can distinguish two
sub-zones:
zone 7a : coarse gravel(diameter of the pebbles usually more than
2 cm), in a mould of coarse sand.
zone 7b : fine gravel (diameter of the pebbles 4 at 5 mm),
in a mould of coarse sand.

Zone 6 : a zone consisting of yellowish-brown to yellowish-green
(10 YR 5/8), medium sized sand, in which a lot of small pebbles
(diameter 5 mm) are found.
The zone is characterized by a strikingly cross-stratification.

Zone 2 : practically the uppermost part of zone 6.
The material is the same (medium-sized sand with a lot of small
pebbles), but in contrast with zone 6, where the material was
coloured yellowish-brown to yellowish-green (10 YR 5/8),
the colour of zone 2 is clearly brownish-red to red (5 YR 4/6)
Moreover, in zone 2 a markedly, concentrated iron precipitation
appears. This iron precipitation appaers in different forms
(see fig. 3) :
- as long-stretched, thin iron bands with a thickness of 2 cm,
- as small rings with a diameter of 1 to 1,5 cm,
- as large (often not fully closed) rings with a diameter of
more than 5 cm.

According to D.E.B. BATES & J.F. KIRKALDY (1976), we consider
both the small and the large rings as topical Liesegang rings.

Zone 3 : forms the top of the river deposit.
We can distinguish two sub-zones :
zone 3a : a gravel layer with large pebbles (sometimes 3 cm diameter).
The pores between the pebbles are filled with very coarse sand.
zone 3b : contains only the very coarse sand. Pebbles are absent.
As zone 2, both zone 3a and zone 3b have an intense red colour
(2,5 YR 4/6).

Zone 5 : this zone doesn't belong to the river deposit anymore. It's the
Quaternary loess layer, which covers both the gully material
and the Diestian sand. We distinguished between :
zone 5b : calcareous loess, which still contains a bit pebbles and
Diestian sand.
zone 5a : calcareous loess without pebbles or sand.

As can be seen in fig. 3, several frost cracks and frost fissures
occur in the western profile. At some places these frost cracks
can reach considerable proportions (for example the two large
frost cracks right in the profile).


2. THE NORTHERN WALL.

The northern wall permitted us to get an idea of the total width
of the river gully (see fig. 2). The profile is shown in fig. 4.
We distinguish the following layers :

Zone 1 : the greenish-brown (2,5 Y 4/4)
Diestian sand we already described earlier. In contrast to the
western wall, the Diestian of the northern wall contains several
hard iron crusts of limonite (thickness : about 0,5 cm).
See fig. 4.

Zone 7a : appears only in a tiny gully in the west of the profile.
As in the western wall, the zone consist of coarse gravel with
a markedly red colour (2,5 YR 5/6).

Zone 6 : appears also only in the tiny gully in the west of the profile.
Zone 6 consists of very coarse sand with small pebbles in it.
The colour is the same as that zone 7a : (2,5 YR 5/6).

Zone 2 : appears only in the west and the east of the profile.
It's a small zone of very coarse sand, the colour of which is
again intense red (5 YR 4/6).

Zone 3 : in the northern wall, only zone 3a is present.
It consists of coarse gravel, with in the pores very coarse sand.
Also here, the colour is intense red (5 YR 4/6).
As can be seen in fig 4, zone 3a occurs over the whole width of
the profile. The contact with zone 1 (Diestian sand) is very
irregular : it shows a lot of gullies, some of which are more
than 0,5 m deep and wide.

Zone 4 : a zone which occurs only in the east of the profile.
It's a 10 cm thick, brownish-red (5 YR 4/6) coloured band of
very coarse sand.

Zone 5 : the Quaternary loess layer we described earlier.
It is composed of :
zone 5b : calcareous loess with pebbles and sand,
zone 5a : calcareous loess without pebbles or sand.

Also in the northern wall of the excavation, a lot of frost
cracks and frost fissures are present. Their depth can reach
more than 1,5 m.


SEDIMENTOLOGICAL INTERPRETATION

The profiles show very clearly that we have to deal here with
a river deposit. The zones 7a and 7b (undermost gravel zones)
form the bed gravel, that during the period that the gully still
was active, settled down as the most heavy material.
The zones 6 and 2, which have the same composition
(mixture of sand and small pebbles) and the same structure
(cross-stratification) form the main mass of the main mass of
the material deposited by the river.

The zones 4 and 3 (uppermost coarse sand and uppermost gravel)
form the top top of the deposit. They were also deposited by
the river, in a period where the stream velocity and/or the
quantity of water was considerably great (this explains the
presence of the great mass of gravel; the finer material (fine
and medium-sized sand) is not present because the water
possessed enough energy to transport it).

Zone 5b, consisting of a mixture of loam, sand and gravel,
doesn't belong to the river deposit anymore, but agrees with
the rinsed zone that is found everywhere on the base of the
Quaternary loess deposits.
Zone 5a is the typical Weichselian loess, which covers almost
everywhere in Mid-Belgium the Tertiary substratum.

The total depth of the river deposit alone (i.e. the zones
7,6,4, an 2) reaches at maximum 1,5 m. However, there are some
arguments to assume that the original depth has been
considerably greater. These arguments will be discussed in part
"Stratigraphical interpretation".

The total width of the river deposit is at least 46 m.
Indeed, for the construction of the A2 highway, a notch of
46 m wide was dug out, and we could establish that the river
deposit was present over the whole width of the notch.
Therefore, the total width of the river during times of peak
discharges has been at least 46 m, and probably more than 50 m
(there is no indication that the river has moved up in the
course of time, for example in a meander). So we can conclude
that we have to deal with a reasonably important river.


LABORATORY ANALYSIS OF THE SEDIMENTS.

1. THE GRAVEL

The river deposit contains two gravel zones : zone 7a and zone 3a.
Zone 7a was examined once, zone 3a twice : a first time with
gravel above zone 3b, and a second time with gravel underneath
zone 3b.

a. The petrography.

From each of the samples, a random test of 200 pebbles was taken.

The results are given in table I.

From the 600 examined pebbles, only 2 are no silex or
ironsandstone crust (1 sandstone in zone 3a (above 3b)
1 sandstonein zone 7a). The silexes are the most represented :
91,5 % for zone 3a (above 3b); 81,5 % for zone 3a
(underneath 3b) ; 91,5 % for zone 7a. The rest of the gravel
consists exclusively of ironsandstone crusts.

The rounded silexes are re-worked from the Diestian.
Indeed, from the Tertiary formations in the vicinity of Louvain,
only the top of the Diestian contains similar rounded and
unflattened silexes.

The ironsandstone crusts are also reworked from the Diestian.

The disc-shaped black silexes are primary of Rupelian origin
(R 1a), but at many places they are also present at the base
of the Diestian, because the Diestian sea has reworked them.
The fact that these diskshaped black silexes are also found
in the B.H. - Gasthuisberg gully, proves that somewhere at a
certain place the river has eroded the base of the Diestian.


Number of pebbles in %
Petrography zone 7a zone 3a zone 3a
underneath zone 3b above zone 3b

ironsandstone crusts 8,0 18,0 8,5

( (unbroken 16,0 27,0 14,0
rounded (cachalonized (broken 7,5 12,0 16,0
(
silexes ( not (unbroken 57,0 32,0 48,5
(cachalonized (broken 9,5 8,5 11,5

disc-shaped black silexes 1,5 1,5 1,0

sandstones 0,5 0,5 -

TABLE I : petrography of the gravel


b. The morphometry.

In order to determine the mean morphometry of the gravel,
we took from each gravel zone a random test of 100 silex pebbles.
From each pebble we measured the length (L), the breadth (B),
the thickness (D), and the smallest rounding radius in the plane
of the length and the breadt (ri).

The shape.

The mean shape was determined by calculating for each pebble
the ratios B/L and D/B and by setting out the obtained values
in a diagram (according to ZINGG 1935).
Fig. 5 shows that for the three zones, the result is nearly the
same : the great majority of the pebbles belong to the classes
"disc-shaped" and "sperical".
Therefore we can conclude that the gravel of the B.H.
Gasthuisberg gully in general consists of disc-shaped pebbles,
which are not yet considerably flattened, and in which the
spherical shape is still discernible.

The rounding.

In order to quality the rounding of the pebbles, we used the
index 2ri/B x 100, as proposed by KUENEN (1956).
The more rounded the pebbles, the higher the value of the
index is 100; in this case the pebble is perfectly circular
in the length and the breadth.

Fif. 6 shows that the pebbles generally are rounded very well :
the mean rounding index for zone 7a is 81,14 (a:8,37), for zone
3a (underneath 3b) the value is 78,17 (a : 14,73), and for zone
3a (above 3b) the value is 74,96 (a : 13,02). The dispersion
around the mean values is not high, as is indicated by the
values of a between the brackets. For each of the gravel samples
the KUENEN index rarely falls under 60; clearly bad rounded
pebbles are therefore exceptional.

The obtained values are significantly higher than those of
normal river deposts (For comparison : for the gravel of the
Kempian Plateau, E. PAULISSEN (1973) found values of 48,2
(a : 18,0) at 43,5 (a : 19,4) (for quartz), and 53,8 (a : 21,7)
at 47,3 (a : 20,3) (for quartzite).

It is striking that the closer the pebbles are situated to the
top of the gully, the worse rounded they are : the KUENEN index
is 81,14 for the undermost gravel zone, 78,17 for the middle
gravel zone, and 74,96 for the uppermost gravel zone.

c. The ratio broken/unbroken pebbles.

If we compare the total number of broken pebbles broken pebbles
(broken by the frost) of each gravel zone, we see that this
number decreases as we penetrate deeper into the gully :

- uppermost gravel zone (3a(above 3b)) : 55/200 broken pebbles,

- middle gravel zone (3a(underneath 3b)) : 41/200 broken pebbles,

- undermost gravel zone (7a) : 34/200 broken pebbles.

The frost cracks we found in the northern and the western
profile, indicate that since the deposition of the river gravel
at least one ice age has been passed. During ice age times,
congelifraction (breaking of the pebbles as a result of frost
action) occurs frequently. It however is known that
congelifraction occurs more frequently at the surface than at
large depths, because of the larger number of frost-thaw cycli.
One therefore can expect that the total number of broken pebbles
will decrease from the surface into the depth. This agrees
completely with the results we found above.

d. The surface structure of the pebbles.

Nearly all the silex pebbles have a surface that is
characterized by the typical "semicolon structure"
(very numerous tiny dot-shaped or comma-shaped notches
in the surface of the pebble, caused by the numerous
powerful collisions of the pebbles in the surf zone on a beach).
Hence, a semicolon structure is typical for a sea gravel and
not for a river gravel. The semicolon structure on the surface
of the pebbles of the B.H. Gasthuisberg gully is therefore not
originated in the gully itself, but in the Diestian sea.

e. Conclusion.

The different analyses show clearly that the gravel is in origin
a sea gravel, which has been re-worked from the Diestian
deposits by the B.H.Gasthuisberg river. The numerous broken
pebbles indicate in addition, that since the deposition of the
material, at least one ice age has been passed.

2. GRAIN SIZE DISTRIBUTION OF THE SAND LAYERS.

From each of the sand layers (i.e. the zones 1, 2, 3b and 6),
20 g. was sieved in order to determine the grain size
distribution. The results are represented in fig. 7. An analysis
of the histograms shows the following :

1. The material of the gully (zones 2, 3b and 6) is claerly coarser
than the autochthonous Diestian sand (zone 1).

2. The Diestian sand is very heterogeneous : the histogram shows

nowhere a peak. This is claerly in contrast to the gully sand,
which in each of the zones 2, 3b and 6 appaer to be
significantly more homogeneous.

3. Zone 3b (i.e. the sand between the gravel layer 3a) is
significantly coarser than the other sand zones (94 % of the
sand is coarser than 250 mm). The fraction <177 µm is neglectible.

4. The great majority of the gully sand (zones 2 and 6) belongs to
the classes 177-250 µm and 250-354 µ m, and can therefore be
labeled as moderate fine sand.
The second peak (above 500 µm) in zone 6 is caused by the small
pebbles that occur between the sand of this zone.

We can conclude that the gully sand is significantly coarser
and significantly more homogeneous than the Diestian sand.
An explanation for this can be the following. Running water
does not transport all particles evenly easy. Fine particules
(clay, silt and fine sand) are transported without many problems,
whereas coarse particles (coarse sand and gravel) are
transported more difficultly. Of the initially very heterogeneous
Diestian sand that was eroded and transported by the river,
only the finest particles could be transported over a large
distance. The coarse particles initially were transported too,
but as soon as the stream velocity decreased, they sinked towards
the bottom. Along these lines, the finest particles were washed
out the Diestian sand, and so it can be explained why the gully
sands are significantly coarser and significantly more homogeneous
than the autochthonous Diestian sand.


STRATIGRAPHICAL INTERPRETATION.

We know that the gully was formed after the Diestian period,
because it is incised in top of the Diestian deposits.
Furthermore the gully has to be formed before Weichselian times,
because it is completely covered with Weichselian loess.
Therefore it can be of Pliocene, Early-Pleistocene or
Mid-Pleistocene origin. There are however three elements,
which plead for Pliocene origin :

1. THE COLOUR OF THE DEPOSIT.

When we described the profiles of the western and the northern
wall of the pit, we more than once emphasized that the upper
part of the gully sediment (both the sand and the gravel) is
coloured markedly brownish red to red (mostly 2,5 YR 4/6;
sometimes 5 YR 4/6).
This brownish-red colour has found to be due to the weathering
of the ferruginous glauconite towards hematite. The deep
brownish-red colour indicates that the weathering has been
very strong. A suchlike intense weathering is only possible in
a warm and above all very dry climate (necessary for the
deshydration). Hence, since the disposition of the gully
sediments, at least one such-like warm and dry period has to be
passed.

In the Pleistocene, the warm periods are restricted to the
interglacial times. During these interglacial times a weathering
horizont (mostly including a soil) developed into the
outcropping formations. From the brownish-red we found in the
B.H.-Gasthuisberg gully, we cannot say that it represents a
typical soil : it's only a deep weathered zone, and soil
horizonts are completely absent. It's much more probable that we
have to do here with the undermost part of the weathering layer,
i.e. the part that is developed underneath the proper soil
(perhaps also the undermost part of the soil is still present).
The uppermost part of the soil was developed higher up in the
river gully, but because of later erosion it has disappeared
together with the uppermost part of the river deposit. If we
accept that the uppermost part of the river deposit has been
eroded away, we also can explain why the (actual) depth of the
deposit (maximum 1,5 m) is so small compared with the width
(more than 50 m).

From some of the interglacials, the fossil soils have
been found. These soils however manifest in no way the
same intense weathering colour as in the case of the
B.H.Gasthuisberg gully :


interglacial soil (in Belgium) colour

Holocene Holocene soil 10 YR 5/4 (see ref. 3)
Eemian Rocourt soil 7.5 YR 4/4 (see ref. 1)
Holsteinian As soil 5-7.5 YR 5/6 (see ref. 2)
? B.H.-Gasthuisberg 2.5 YR 4/6

Table II : colour of the interglacial soils in Belgium.


We see that the older the weathering, the intenser it becomes.
The claerly very deep weathering colour of the material of the
B.H.-Gasthuisberg gully suggests that the river deposit has to
be reasonably old : probably older than Holsteinian times.
Because of the fact during the Pliocene the climate in our
country was warm and dry (mediterraneous or sub-tropical),
it even is possible that the gully could be of Pliocene origin.
Of course we have to take into account that the strong red colour
also is due in great measure to the large quantity of available
iron minerals (glauconite). It certainly is not dependent on the
intensity of the weathering alone.

2. THE STRUCTURE OF THE DEPOSIT.

There are two elements that indicate that the B.H.-Gasthuisberg
river was a braided river :

1. The river gully is very wide (at least 46 m and probably more
than 50 m), but very shallow (original depth probably nowhere
much more than 1,5 m, except in the main channel).
Furthermore, the river bed does not consist of one single
channel, but of a very great quantity of small secondary
channels (the most important of them are shown in fig. 2)
and one main channel. These characteristics (very wide and
shallow river bed with numerous small channels) are typical
for a braided river, and are not characteristic of any
other river type.

2. The material deposited by the river is very heterogeneous
(both fine sand and large pebbles). It was deposited in
clearly distinctive layers, which either consist exclusively
of coarse material (gravel), or exclusively of fine material
(sand). Such a deposition is very typical for a braided river.

Braided rivers only appear in two climates :
- in a cold climate (glacial or periglacial),
- in a arm and dry climate (mediterranean or sub-tropical).

The gully therefore could be formed either in a cold climate
(Pleistocene) or in a warm and dry one (Pliocene);
the structure of the deposit does not produce new data with
regard to this.

3. THE TOPOGRAPHICAL POSITION OF THE DEPOSIT.

A very strong argument to make from the B.H.-Gasthuisberg
river a Pliocene (Tertiary) river and not a Pleistocene
(Quaternary) one, is given by the topographical position of
the deposit. The gully lies at an absolute height of about
73 m. The summit of the Gasthuisberg lies at a height of 85 m
(see fig. 1), but the thickness of the loess layer at that
place is about 7 m (D.GOOSSENS, 1981) as a consequence of
which the Tertiary summit of the Gasthuisberg lies at an
absolute height of 78 m. This means that, if we think away
the Quaternary loess cover, the river deposit appears nearly
exactly on the summit of the hill.

A river however does not flow over hilltops, but in
depressions. We thus have to accept that at the time
that the river still was active, the actual summit of
the Gasthuisberg was not a hilltop but a depression.
That means that at that time, the relief in the
vicinity of Louvain had to be totally different from
the actual one. The whole Dijle depression did not
exist, otherwise the river would have flowed there.
Actually the Dijle flows (near Louvain) at an absolute
height of about 20 m; the Dijle depression consequently
is incised more than 50 m with regard to the gully.
The Dijle depression is also very wide (several km).
The total amount of material that has been eroded away
since the fossilisation of the gully is therefore
tremendously great.

Such an erosion cannot possibly be carried out in a few
ten thousands or hundreds of thousands years by a river
of the calibre of the Dijle. We have to calculate here
with millions of years. That means that the B.H.-Gasthuisberg
river does not date from the Quaternary, but from the
Tertiary. Because we know that it in every respect must be
younger than the Diestian, we consider it as a Pliocene river.


CONCLUSION.

The deposits of the B.H.-Gasthuisberg river are the
first unmistakable continental Pliocene deposits found
in Belgium. The red zone at the top of the deposits
forms the undermost part of the weathering zone, which
was developed during the Pliocene. Perhaps also the
undermost part of the Pliocene soil is taken up in it.
As a result of later (Pleistocene) erosion, the
uppermost part of the soil) has disappeared. Only the
undermost part has been kept.

We propose the name formation of Gasthuisberg for the
continental facies of the Pliocene deposits in Belgium
For the river deposit we described in this paper,
we propose the name sands and gravels of Gasthuisberg.


REFERENCES.

QUATERNAIRE DU NORD DE LA FRANCE ET DU SUD DE LA BELGIQUE.

Guide-book of the A.F.E.Q.-excursion 19-21/5/1972,
excursion point Tongrinne, p.2.

EXCURSION GUIDE-BOOK co-ordinating committee for
periglacial research (Working group of the
International Geographical Union), symposium 17-28
Sept. 1978, pp. II 3-5.

GOOSSENS, D (1980) - Een onderzoek naar de
sedimentologie en het verband loessdikte-relief van
Weichseliaan-loess ten S.W. van leuven.
Dissertation Katholieke Universiteit Leuven, pp.33-34

GOOSSENS, D (1981) - De dikte van het loessdek ten S.W.
van Leuven.
Werkstukken Nationaal Centrum voor Geomorfologisch Onderzoek,
vol. XX,pp. 122-140.

KUENEN, Ph. (1956) - Experimental Abrasion of Pebbles II :
With Rolling by Current. Journ. of Geology, vol. 64,
nr. 4, pp. 336-368.

PAULISSEN, E. (1970) - Morfologie en Kwartair-stratigrafie
van de Maasvallei in Belgisch Limburg.
Doctoral dissertation Katholieke Universiteit Leuven, p. 78.

ZINGG, Th. (1935) - Gestalt und Abrollung. Schweizer
Mineral. Petrog. Mitteilungen, vol. 15, pp. 31-140.



Manuscript received on December 1981.