This field work was conducted around the city of Sidi Bouzid in central Tunisia. The
task was to study and document different weathering forms in the area as well as differences
and uniformities between the forms created in different rock types (e.g. sand-
stone, limestone and gypsum rock). The weathering forms studied were rillenkarren,
weathering pits, alveoles, tafonis, exfoliation-, jointing, plane-, needle-, staircase- and
circle weathering. The last four weathering forms referred are not described in the literature.
The names are given by us and we have describe them and, in some cases, tried
to hypothesize about their generation. The examined rocks are either located up on
mountains (on ridges or slopes) or down in wadis were the rock is exposed in (to?) the
surface due to the occasional flow of water. We are grateful to our lectures, Lars Franzén
and Mats Olvmo, who helped us to find some of the locations that were studied. We
also thank them for describing the big context in which the area was originated as well
as the processes that have generated the landscape in this part of Tunisia.
Figure 1: Map of the investigated area with locations.
The central parts of Tunisia, as described by Campbell (1980), consists of anticlinal
mountains whose axis of fold is aligned from the south-west to the north-east. Between
mountain and basin are extensive pedimented and alluviated surfaces. Few permanent
streams exists and the sparse rainfall permits only flow occasionally. These flows gives
high erosion in the stream channels which are terraced with the surface often standing
several meters above their present channels.
The mountain ranges are relatively undissected reflecting a geologically recent origin,
resistant rocks and the irregular action of erosional processes. They consist of strongly
folded and faulted Cretaceous to early Quarternary sediments; shales, limestone, sandstone
and conglomerates. The lithological differences has influenced the landscape,
where cliffs and ridges are developed on massive limestone and valleys eroded in
weaker rock. The limestone have often a massive crystalline character with siliceous inclusions.
Between the mountain ranges lie many closed depressions, salt lakes (sebkhas) and salt
marshes (chotts) ranging from a few square kilometres up to several thousand square
kilometres (Chott el Djerid).
Small-scale solutional features, so called karren forms are primarily noted in karst regions
on limestone and dolomite, although they can be seen even on basaltic and granitic rock.
The small and geometrically regular solution flutes or rillenkarren belong to a group of
features whose form result from the very rapid dissolution. This occurs when rainwater
meets the limestone surface without CO2 diffusion.
Bögli (1960, in Dunkerley 1979) has made some specific statements about rillenkarren:
rillenkarren forms will not form on horizontal surfaces, where flow is low and CO2 diffusion
becomes involved. They occur on steeper slopes and gets larger as the surface
becomes steeper. The reaction between rain water and the rock is temperature dependent
and is fast in high temperature areas. The length of the solution flutes increases with
the intensity of the local rainfall. Their width varies little, lying usually in the range 2-3
Dunkerley also notes that the feature is best developed on south facing slopes which is
linked to the existence of a thin pre weathering layer on surfaces in that direction.
Since the morphology of rillenkarren is independent of inclination of the rock face it
follows that flute dimensions (width, cross-sectional area, etc.) are not purely
hydrodynamically determined. The discharge of water over a particular point and the intensity
of rainfall are not the determining factors in the development of rillenkarren.
Dunkerley says that the physical chemistry of the solutional process must control flute
formation and determine the rate of limestone removal. The rate-determining factor is
some step involved in the surface reaction between calcium carbonate and the moving
Because of morphologic differences between two sites it should be empathised that the
factors that influence the dissolution reaction in a certain site are very important. These
factors can be difference in texture and composition of the surface rock. Temperature is
also an important influence.
The development of a solutional flute, according to Dunkerley, is the metamorphose
from one weathering form to another. Solutional attack on the bare rock begins at scattered
points where solutional pits are developed. Downslope chains of depressions become
progressively more smoothly formed and better connected. This is because of the
universal mechanism that small irregularities in the pattern of solution tend to cause
their own removal. This process gives the rillenkarren weathering form.
Tafoni are cavernous weathering features that typically have volumes of several cubic
metres, arch-shaped entrances, concave inner walls, overhanging margins, and fairly
smooth, debris-covered floors. Cooke et al (1993). They occur in many different rock
types and their formation includes at least two phases: disintegration of the walls, and
removal of debris from the floors. The air inside a tafoni is cooler and moister then outside.
The moister air is the main factor for increased weathering inside a tafoni.
The polygonal plates range in diameter from 2 cm to some 24 cm, with the average and
mode both near the upper end of the range. Some of them are slightly curved or convex
upward in respect to the surface of the host rock. Twidele (1982). Possible explanations
for development of polygonal cracking are: Initiated beneath the land surface, disinte-
gration of sheet structures developed on the joint faces at the time of major jointing and
in consequence of it, insolation and chemical weathering.
The task was carried out as a documentation of the various weathering forms that were
found in the area. Photographs were taken and were used along with our field notes and
sketches to compare our observations with the literature. Some samples of rock were
brought back to our laboratory to judge the lime content in the various kinds of lime-
and sandstone. We used a ruler and a vernier calliper to measure the smallest karren
forms and step length for larger forms. Directions and inclination was measured with a
compass including an inclination indicator. The measurements taken of the various
weathering forms have been statistically analyzed in terms of mean values, standard
deviations and correlation coefficients between the variables.
The studied sandstone rock lies in a wadi close to Behima. The rock is a sandstone terrace,
three metres high, which forms a dried up waterfall. Parts of the terrace has got
fluvial forms with potholes and the other part is filled with Tafoni, alveoles and pitting.
Most of the exposed rock faces east but Tafoni lies in the direction from North-East
down to South. Pitting are exposed in all directions except from the north. There is no
exposed rock in the northern direction and therefore naturally no weathering forms
The weathering in the Tafonis goes backwards and a bit upwards and the bottom is covered
with a thin sedimentary layer. The depths and the heights have got similar length
but the width varies.
Pitting shows similar values concerning depth and diameter with some exeptions concerning
the diameter which in some places are greater than the depth.
In Djediane a tunnelcave has been developed through different tafoni formations. The
tafonis have eaten their way through the rock and eventually created a tunnel. The tunnel
is aligned from north-west to south-east. New tafonis have been developed inside
the walls of the tunnel.
Table 2: Tafoni on Limestone.
Figure 5: The tunnelcave.
There are alveoles in the ceiling of the tunnelcave and exfoliation on the floor. Some
desert varnish has been developed in well ventilated areas of the Tunnelcave.
The area is located on a slope exposed to the south. Weathering forms on solid sandstone
rocks and boulders were studied. The main weathering form in this area is joint-
ing, sometimes with small tafoni formations along the joint. We also noticed a peculiar
kind of weathering form that we named circle weathering. The formation was a 1 mm
deep and 15 mm wide circle on the exposed rock. Some other parts of the area showed
exfoliation surfaces with 1 to 2 mm thick flakes. Some parts were covered with desert
varnish which shows signs of slow weathering processes in the area.
Table 3: Weathering in a joint, Sandstone.
Figure 6: Photo of joint, small tafoni and circle weathering.
Weathering formations in an iron rich sandstone with large quarts particles next to a
wadi. The rock formation is covered with small tafonis, exfoliation areas, jointing and
desert varnish. We found one interesting tafoni containing two small tafonis inside. The
larger tafoni has a depth of 27 mm and diameter of 29 mm and the smaller tafonis in-
side are 10 mm high, 5 mm wide and has a depth of 3 mm. There is desert varnish and
exfoliation areas on the south-western side of the rock. The desert varnish is at least 1
mm thick and it seems to protect the rock from further exfoliation weathering.
The studied area is gypsum slope with three different kinds of weathering formations,
flat surface, staircase weathering and rillenkarren. The angle of the flat surface weath-
ering varies from 2 to 4 in the same direction as the slope, which faces towards North.
The angle of the rillenkarren is about 30 and faces towards West and East, normally on
the sides of the flat surface weathering. There are no rillenkarren facing North, here
staircase weathering divides different flat surface weathering formations. Staircase
weathering is described in figure 7
Figure 7: Staircase weathering on gypsum.
Table 4: Rillenkarren on gypsum slope.
Figure 8: Photo of rillenkarren and flat surface weathering on gypsum.
In the El Djediane area needle weathering was found on a North Eastern slope together
with small rillenkarren. The needles were up to 7 mm long and the rillenkarren next to
the needles where 3-4 mm wide. The gypsum were the needle weathering was found
were lying as pillows surrounded by more easily weathered material.
Figure 9a: Needle weathering on a slope near El Djediane.
Figure 9b: Needle weathering on a slope near El Djediane.
This is an area were a canyon cuts through a gypsum dike. There were rillenkarren and
flat surface weathering in the upper part of canyon. The area looked like a slope with
terraces. We measured rillenkarren in tree different intervals 0-15° , 25-40° and 65-70° .
Flat surface weathering are the normal formation in the 0-15° category but some wide
and shallow rillenkarren occurred. The 25-40° category had the best shaped rillenkarren
and in the 65-70° category the rillenkarren were smaller.
Table 5: Rillenkarren in a gypsum dike.
Jointing is the main weathering form further down in the canyon. The rillenkarren and
the Jointing goes in the same direction as the dike, from North to South. The whole
canyon shows a great variation in gypsum with fluvial formations, different colours and
eye gypsum (a gypsum rock type where pieces of more homogenous materials forms
something that looks like eyes).
This is a limestone area with pitting, alveoles and rillenkarren. All weathering forms are
richly represented in all directions. Small cave systems are built up by alveoles with a
size of up to 20 cm in depth and width. Depth and width of pitting and rillenkarren are
shown in table x. Some of the rillenkarren showed a depression making 10-20 cm deep
This location was the only place where Rillenkarren was found in siltstone. The slope
faced south and the weathering form was only seen in an area of one square meter.
Depth and width are shown in table X.
Table 7: Rillenkarren on siltstone.
There were also exfoliation weathering with up to three flakes of weathering on top of
each other. The weathering are located beneath an overhang on a south-western wall.
Next to Oued Seguia, hollow weathering depending on the rock texture was found. Un-
easily weathered quartzite falls out from more easily weathered rock and creates small
pits. Some of the quartzite stays and creates small boulders sticking out from the rock.
This area with exposed rock on a stone paved surface showed areas of polygonal
cracking. The width on the areas between the joints had a diameter of up to 140 mm
and the joints where 1-3 mm wide. The weathered polygons had 4, 5 or 6 sides.
The examined rocks; gypsum, sandstone and limestone shows a clear difference be-
tween gypsum and the last two mentioned. Tafoni and pitting where only observed in
sand- and limestone and rillenkarren where observed mainly in gypsum and limestone,
except for one place where we observed rillenkarren in something that looked like siltstone.
The hypothesis, stated by Dunkerley, that the? rillenkarren is made out of scattered so
lution pits forming a chain and eventually the solution flute does not fit in to our studies
of gypsum in which we observed a large amount of rillenkarren but no solution pits at
all. In gypsum the rillenkarren went in certain directions that seemed to follow the
structure of the rock. On locations with both rillenkarren, flat surface and staircase
weathering, the rillenkarren goes in one direction (tvärs sluttningen) and flat surface
and staircase weathering goes in the (längs sluttningen) other. The flow over the flat
Our hypothesis concerning the developement of flat surface weathering from a surface
covered by rillenkarren goes? as follows:
1. Flat surfaces are developed where the flow of water is in the form of a thin sheet.
Over the flat surface, the water gets to thick for the raindrops to hit the rock and
weathering acts uniformly over the surface.
2. The flat weathering surface is situated on the same small-scale base level as the sur-
3. The surrounding less resistant material is eroded, leaving the flat surface as a plateau
with rillenkarren down the margins?. Staircase weathering is developed down the
4. The rillenkarren eats its way into the old surface down to a new small-scale base
level and leaves a new flat surface beneath.
Figure 11: The weathering process of flat and staircase weathering
The staircase surface are sometimes finished with a upphöjning?. An explanation for
this could be that because of the nature of the water with a thinner layer near the edges,
the water is given less opportunity for dissolution of the rock beneath. It looked like the
water flowed down the sides of the staircase instead of over the upphöjning?
Though the amount of data taken on each location was pretty small and the fact that it
was very varied, it can up to some point show differences between the rocks.
Campbell, I. A. (1980): Structural geology and geomorphology of the Oed Maala re-
gion, central Tunisia. Zeitschrift für Geomorphologie 24:91-110.
Cooke, R., Warren, A. and Goudie, A (1993): Desert Geomorphology, UCL Press, Lon-
Dunkerley, D. L. (1979): The morphology and development of rillenkarren. Zeitschrift
für Geomorphologie 23:332-348.
Jennings, J. N. (1985): Karst geomorphology, Blackwell Ltd. Oxford.
Martinsson, M. & Swantesson, J. (1988): Some geomorphological notes from Djebel El
Kebar and surrounding mountains. Students seminar notes: Field course in the Sidi
Bouzid area Central Tunisia. Naturgeografiska inst., Göteborg University.
Figure 1: Map of the investigated area with locations. 5
Figure 2: Tafoni formations in sandstone (I). 9
Figure 3: Tafoni formations in sandstone (II). 9
Figure 4: Tafoni formations in sandstone (III). 10
Figure 5: The tunnelcave. 11
Figure 6: Photo of joint, small tafoni and circle weathering. 12
Figure 7: Staircase weathering on gypsum. 13
Figure 8: Photo of rillenkarren and flat surface weathering on gypsum. 13
Figure 9: Needle weathering on a slope near El Djediane. 14
Figure 10: Photo of polygonal cracking. 17
Figure 11: The weathering process of flat and staircase weathering 18
Front image: Our Lecteurer Lars Franzén at the Behima Wadi.
Tabell 1: Rain pits and Tafoni in sandstone. 8
Tabell 2: Tafoni on Limestone. 10
Tabell 3: Weathering in a joint, Sandstone. 11
Tabell 4: Rillenkarren on gypsum slope. 13
Tabell 5: Rillenkarren in a gypsum dike. 15
Tabell 6: Rillenkarren and rain pits on limestone. 16
Tabell 7: Rillenkarren on siltstone. 16