Human and climate impact on catchment development during the
Holocene -- Geul River, the Netherlands
J.J.W. de Moor a,, C. Kasse a
, R. van Balen a
, J. Vandenberghe a
, J. Wallinga b
a
Department of Paleoclimatology and Geomorphology, Faculty of Earth and Life Sciences, Vrije Universiteit, De Boelelaan 1085,
1081 HV Amsterdam, The Netherlands
b
Netherlands Centre for Luminescence dating (NCL), Delft University of Technology, Faculty of Applied Sciences, Mekelweg 15, 2629 JB Delft,
The Netherlands
Received in revised form 5 June 2006; accepted 27 December 2006
Available online 21 May 2007
Abstract
Interest in the Holocene development of small to medium-sized river catchments in Western and Central Europe in relation to
changes in land use and climate has increased over the past years. In this study we reconstruct the Holocene landscape development
and fluvial dynamics of the Geul River (The Netherlands) and the main forcing mechanisms of environmental change. Field studies
were carried out and we used OSL and 14
C dating methods to reconstruct the Holocene valley development. Our study shows that 2
periods of deforestation (during the Roman Period and the High Middle Ages) led to severe soil erosion and increased floodplain
sedimentation in the catchment of the Geul River, possibly combined with periods of increased wetness during the High Middle
Ages. Alluvial fans have been active since the Roman deforestation phase. Our results show that the Geul catchment is highly
sensitive to changes in land use.
 2007 Elsevier B.V. All rights reserved.
Keywords: Floodplain sedimentation; Alluvial fan sedimentation; Land-use change; Climate change; Geul River
1. Introduction
Small and medium-scale river catchments in Western
and Central Europe have important functions as
ecological environments, local drinking water reservoirs
and as flood retention areas during major floods of rivers
they drain into (like the Rhine or Maas rivers). However,
during the last few centuries many of these river systems
have lost their natural character and have been
straightened (Brookes, 1988; Wolfert, 2001). Interest
in the Holocene development of small to medium-sized
river catchments in Western and Central Europe has
increased in the past years. Especially the forcing
mechanisms for the development of these catchments
have been thoroughly investigated. Important questions
hereby are: (1) how are Holocene fluvial dynamics
coupled to human impact and changes in climate and;
(2) what is the sensitivity of Holocene fluvial systems to
environmental change?
Rivers, floodplains and valley environments in the
western and central European loess zone (for example
Belgium, Germany and Poland) have changed strongly
since the Last Glacial Maximum and have been affected
by human activities for about 5000 yr (e.g. Houben,
1997, 2002; Klimek, 2002; Starkel, 2002; Kukulak,
2003; Lang et al., 2003; Mäckel et al., 2003; Zolitschka
et al., 2003; Raab and Völkel, 2005; Rommens et al.,
Available online at www.sciencedirect.com
Geomorphology 98 (2008) 316­339
www.elsevier.com/locate/geomorph
 Corresponding author.
E-mail address: j.de.moor@orange.nl (J.J.W. de Moor).
0169-555X/$ - see front matter  2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.geomorph.2006.12.033
2006). Many of these catchments had a long cultivation
history and are therefore very suitable to study the effect
of land-use changes on fluvial dynamics and catchment
development. Human activities (for example ore mining
and deforestation associated with the development of
agriculture) have led to drastic changes in the discharge
regimes and floodplain dynamics, and soil erosion has
increased. Periods of increased wetness during the
Holocene have been reflected in increased floodplain
sedimentation rates during the Holocene in Poland
(Starkel et al., 2006) and Great Britain (Macklin et al.,
2006).
Since most rivers have been embanked or canalized,
nearly undisturbed fluvial systems are very hard to find.
The Geul River, a small tributary of the Maas in the
southern Netherlands (Fig. 1), although it has been
influenced by human activities over the last thousands of
years, is a nearly undisturbed system. While the Holocene
Rhine­Meuse delta in the Netherlands has been thoroughly
studied (e.g. Berendsen and Stouthamer, 2001),
knowledge of the Holocene development of smaller
catchments in the Netherlands is lacking. This study
intends to improve our knowledge about human and
climate impact on a small river catchment in the fertile and
densely populated loess area. Moreover, this study will
contribute to the regional picture of Holocene valley
development and the forcing mechanisms responsible.
This paper focuses on the Holocene landscape development
and fluvial dynamics of the Geul River catchment.
We make a reconstruction of the river dynamics and
sedimentation characteristics during different phases of
the Holocene and determine the main factors influencing
catchment development and river dynamics. A detailed
sedimentological record with numerous radiocarbon dates
and some additional OSL dates is used to investigate
historical changes in sedimentation and we compare the
fluvial record with local and regional vegetation data. We
characterize the alluvial architecture of the Holocene Geul
River and discuss the forcing mechanisms responsible for
its development.
1.1. The Geul River catchment
The Geul River catchment is situated in the
southernmost part (South-Limburg) of the Netherlands
and adjacent Belgium (Fig. 1). It originates in eastern
Belgium near the German border and flows into the
Maas River a few kilometres north of Maastricht. Its
length is 56 km and the catchment area is about
380 km2
. The altitude of the catchment varies from 50 m
above sea level at the confluence with the Maas River to
400 m above sea level in the source area. The average
discharge of the Geul River near its confluence with the
Maas River is 3.4 m3
s-1
(data from Waterboard Roer
and Overmaas). Occasional peak discharges of more
than 40 m3
s- 1
can cause local floods. The discharge
mainly depends on the amount of rainfall. Heavy rainfall
can result in overland flow on loess-covered slopes and
rapid discharge into the river. The discharge of the Geul
River can change very rapidly, for example during
heavy thunderstorms (De Laat and Agor, 2003). Smallscale,
local floods occur almost every year (mainly
during the winter), but they do not cause much damage
as only some grasslands along the river become
inundated.
An important steering factor in the long-term
landscape development of the Geul basin during the
Quaternary has been the River Maas. At the end of the
Tertiary and the beginning of the Quaternary, the Maas
had an easterly flow direction, thereby crossing the
present-day Geul catchment. Due to tectonic uplift of
the Ardennes and South-Limburg, the river started to
shift towards its current south-north position, thereby
creating a number of terraces (Van den Berg, 1996).
These terrace levels can clearly be recognized at high
altitude in the landscape today as large, flat plains (cf.
Zonneveld, 1974). Fluvial gravels are often present at
the surface of these river terraces. Following the
downcutting of the Maas, tributaries like the Geul
incised as well. In the Belgian part of the catchment
and the southernmost tip in the Netherlands, the Geul
River is incised into Palaeozoic rocks (Devonian and
Carboniferous sandstones, shales and limestones
containing Pb­Zn mineralizations). In the Dutch part
of the catchment, the river is incised mainly into
Cretaceous lime- and sandstones. Almost the complete
catchment has been covered with a blanket of loess,
deposited during the Saalian and Weichselian glacial
periods.
The present-day landscape of the Geul River
catchment is characterized by large, flat plateaus and
deeply incised, asymmetrical river valleys (Fig. 2a). The
floodplain of the Geul River is in general flat, but
several alluvial fans coming from tributary valleys cover
the floodplain (Fig. 2b­f).
1.2. Land use, vegetation change and human activity
The Late Glacial and Holocene vegetation cover
and land-use history of South-Limburg (Fig. 1) has
previously been studied by Bunnik (1999), Havinga
and Van den Berg van Saparoea (1980) and Renes
(1988). Bunnik (1999) has constructed several detailed
pollen diagrams for South-Limburg and adjacent
317J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Fig. 1. Location of the Geul River catchment and the study sites in South-Limburg.
318J.J.W.deMooretal./Geomorphology98(2008)316­339
Fig. 2. a. Slope map of the study area. b. Topographic and geomorphologic map with the location of transect Schoutenhof. c. Topographic and
geomorphologic map with the location of transect Hommerich. d. Topographic and geomorphologic map with the location of transects Burggraaf
West and East. e. Topographic and geomorphologic map with the location of transect Genhoes. f. Topographic and geomorphologic map with the
location of transect Vroenhof.
319J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Germany, while Havinga and Van den Berg van
Saparoea (1980) have constructed pollen diagrams for
two locations in the Geul River valley.
The Late Glacial vegetation pattern was characterized
by an open vegetation type with birch and pine
trees. This changed during the Preboreal when the area
Fig. 2 (continued ).
320 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
was gradually covered by a dense forest with wet
vegetation types in the river valleys. Prior to significant
human activity, the vegetation reconstructions demonstrate
that valley sides with loess soils were covered
with mixed deciduous forest (oak and lime), while the
river valley floors were covered by alder, willow and
poplar (Bunnik, 1999). During the Boreal and Atlantic
not much changed in the vegetation cover: dense forest
dominated although there were some changes in the tree
species composition.
Parts of South-Limburg have been inhabited for the
last 7000 yr. Archaeological findings just north of the
Geul River catchment indicate the presence of these first
settlers (farmers of the Bandkeramik culture, 4500 BC)
on the loess plateaus (Renes, 1988). No settlements of
the Neolithic Bandkeramik culture have been found in
the Geul River catchment (Van de Westeringh, 1980).
The Early and Middle Bronze Age (1950­1000 BC)
were characterized by an almost complete forest cover
in the area (Bunnik, 1999). During the Late Bronze Age
(1000­700 BC) the first farming took place in addition
with local deforestation. Evidence of presence of man in
the area during the Late Bronze Age is supported by the
presence of several burial mounds close to the Geul
River catchment and by the presence of heath and
agricultural indicators (cereals) in the pollen record
(Bunnik, 1999). Local deforestation continued during
the Iron Age (700­53 BC). In this period the forest in
the river valleys was replaced by grassland (Bunnik,
1999). Deforestation on the valley slopes led to the first
significant soil erosion and formation of colluvium in
the valleys.
A widespread deforestation phase in South-Limburg
took place during the first half of the Roman Age (53
BC­415 AD). A population expansion took place
resulting in the growth of several cities. Forest areas
were transformed into arable land and on the large
plateaus so-called Villae Rusticae (a Roman villa
associated with a large farm) were founded (Van de
Westeringh, 1980; Renes, 1988). The Roman agricultural
activities are expressed in pollen diagrams by the
presence of high pollen counts for cereals, sweet
chestnut and walnut pollen (Havinga and Van den
Berg van Saparoea, 1980; Bunnik, 1999; Bazelmans et
al., 2004). The deforestation led to severe soil erosion in
the catchment of the Geul River. During the second half
of the third century, the Roman Empire collapsed and
many Roman Villas fell into disrepair, resulting in a
population decline (Renes, 1988; Bazelmans et al.,
2004). The forest on the plateaus and valley slopes
regenerated. During this period (called the dark ages or
migration period, 220­500 AD), South-Limburg was
not very populated and only a few small settlements
were situated in the river valleys (Renes, 1988).
The first signs of the recovery of the arable lands date
from the Early Middle Ages (500­1000 AD), although
the rapid population expansion during the High Middle
Ages (1000­1500 AD) resulted in an almost complete
deforestation of the area (Renes, 1988; Bunnik, 1999).
Only the steepest slopes and poorest soils remained
covered with forest and the river valleys were in use as
grassland. The Medieval deforestation phase also marks
the second phase of massive soil erosion in the area (Van
de Westeringh, 1980). Soil erosion mainly took place on
the plateaus and valley slopes and sediment accumulated
in river valleys. Due to this soil erosion on the slopes,
measures were taken by local farmers to protect their
land of further being eroded. Erosion barriers or
lynchets were constructed. It is, however, not clear
when the first lynchets were constructed. During the last
centuries, large parts of arable land were converted into
grassland as farming innovations led to higher crop
yields (Renes, 1988). During the last 50 yr the scale and
intensity of agriculture has increased, while the removal
of lynchets and the change of plough direction from
parallel to perpendicular to the slope have increased soil
erosion. The arable land is exposed now for longer
periods during the year and this also increases the
vulnerability to erosion (Stam, 2002).
The Geul catchment had a long history of lead and
zinc mining, dating back to the Middle Ages with
maximum extraction between the middle and the end of
the 19th century (Stam, 2002). These mining activities
have contaminated floodplain sediments of the Geul
and, in turn, may be used as chemical markers to
reconstruct valley sedimentation, depositional rates and
fluvial dynamics. Leenaers (1989), Swennen et al.
(1994) and Stam (1999, 2002) have thoroughly studied
the presence, amounts and dispersal of the contaminated
sediments. Stam (2002) used the contaminated sediment
to reconstruct floodplain sedimentation rates over the
last 200 yr in relation to changes in climate and land use.
2. Methods
We cored several detailed cross-sectional profiles
across the valley floor at different locations in the Geul
catchment (Fig. 2a­f), using an Edelman hand-auger.
Occasionally we used a 6 cm diameter gouge to obtain
peat and organic samples. Two cut-bank sections were
also studied. Sediments were described every 10 cm and
classified using the USDA texture triangle (USDA,
2005). Samples from the Edelman cores were taken for
radiocarbon dating and grain size analysis. Samples for
321J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Optical Stimulated Luminescence dating (OSL) were
taken from the cut-bank exposures by hammering
opaque steel tubes into the wall. Grain sizes were
analyzed using a Fritsch laser particle sizer A22, with
the 8 m as the upper limit of the clay fraction (Konert
and Vandenberghe, 1997).
OSL dating determines the time since deposition and
burial of mineral grains; the OSL age is obtained by
dividing the absorbed radiation dose received by
mineral grains since burial (equivalent dose) by the
dose received by the grains per year (dose rate). OSL
dating for this research was carried out at the Netherlands
Centre for Luminescence Dating at Delft University
of Technology. Quartz grains in the fraction 63­
90 m were used for equivalent dose estimation. The
samples were sieved and then treated with HCl and
H2O2 to remove carbonates and organic material.
Subsequently the samples were treated with concentrated
HF to dissolve feldspars and etch the outer -exposed
skin of the quartz grains. Then the samples were sieved
again to remove grains that were severely damaged by
the HF treatment. The SAR protocol of Murray and
Wintle (2000, 2003) was used for measurement of the
equivalent dose (see Table 1); aliquots showing rogue
luminescence behaviour were discarded (see Table 2 for
rejection criteria). Based on investigations of the
dependence of equivalent dose on preheat temperature
a 10 s preheat at 225 °C was applied before measurement
of the natural and regenerative dose OSL
responses, and heating to 200 °C was applied before
measurement of the test dose OSL responses. Dose
recovery tests (Roberts et al., 1999) confirmed that the
adopted procedure could accurately recover a laboratory
dose (average dose-recovery ratio 1.020.02). Given
the fluvial nature of the samples, incomplete resetting of
the OSL signal of some grains prior to deposition and
burial is likely (e.g. Wallinga, 2002). To avoid bias of
results due to this heterogeneous bleaching, singlealiquot
equivalent doses of more than 2 standard
deviations from the sample mean were removed in an
iterative procedure. The dose rate was estimated using
high-resolution gamma-ray spectroscopy (Murray et al.,
1987). Water contents as measured on the samples
(ranging from 20­26% by weight) were used to estimate
attenuation of the effective dose rate by water; we
included a large uncertainty of 5% to allow for changes
in water content during geological burial.
For radiocarbon dating, peat and sediments containing
organic material (like twigs, leaves, seeds and other
macroscopic plant remains) were sieved and suitable
material (seeds, charcoal and leaves) was selected for
14
C AMS dating at the Centre for Isotope Research at
the University of Groningen. Seeds were identified prior
to submission for 14
C dating. One bulk sample was
dated using the conventional method. The 27 14
C dates
were calibrated with the CALIB radiocarbon calibration
program (Stuiver and Reimer, 1993 (version 5.0)), using
the calibration dataset from Reimer et al. (2004).
2.1. General sedimentation pattern and dating results
In this paper we present 5 different cross-valley
transects from the middle and downstream sections of
the Geul River (Figs. 3­7). In these cross-sections, 11
different lithogenetical units have been identified, based
on differences in grain size, lithology and morphology.
Table 3 provides a general overview of the sedimentary
Table 1
The SAR procedure (modified from Murray and Wintle, 2000, 2003)
Step Action Measured
1 Regenerative beta dose a
2 10 s Preheat at 225 °C b
3 40 s Blue stimulation at 125 °C Ln, Li
c
4 Fixed test beta dose d
5 Cutheat to 200 °C
6 40 s Blue stimulation at 125 °C Tn, Ti
e
7 40 s Blue bleaching at 245 °Cf
8 Repeat step 1­7 for number of regenerative doses
Extra 1 Fixed test beta dose
Extra 2 Cutheat to 200 °C
Extra 3 40 s IR stimulation at 50 °C g
IRe
Extra 4 40 s Blue stimulation at 125 °C Te
a
No beta dose is administered for measurement of the natural OSL
signal in the first cycle of the procedure Ln.
b
The preheat temperature is selected based on the preheat-plateau
test and dose-recovery test.
c
The signal used for analysis is the signal measured during the first
0.32 s of stimulation minus the average background signal determined
over the last 4 s of stimulation.
d
The test dose is chosen to be approximately 25% of the equivalent
dose.
e
Response to the fixed test dose is used to monitor sensitivity
changes of the material during the measurement routine.
f
The preheat temperature plus 20 °C is used as an added bleaching
step before the sample is dosed (step 1).
g
After completion of the standard SAR routine, we use IR
stimulation to check whether the sample is contaminated by feldspar.
Table 2
Applied thresholds for accepting data for analysis
Test Ideal case Accepted if 
1 -- Recycling test (L5/T5)/(L1/T1)=1 0.9b(L5/T5)/(L1/T1)b
1.1
2 -- Recuperation test (L4/T4)/(L1/T1)=0 (L4/T4)/(L1/T1)b0.1
3 -- Feldspar test IRe/Te =0,
Te/T5 =1
IRe/Te b0.2 or
Te/T5 N0.9
322 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
characteristics and environments of the lithogenetical
units. Fig. 8a­g show the typical grain-size distributions
for several units; the percentages of clay, silt and sand of
these units are based on several samples. The units have
been dated using 14
C and OSL dating techniques (dates
are shown in Figs. 3­7 and Table 4/5). All the 14
C dates
have been accepted and considered reliable, within the
context of alluvial systems. The radiocarbon samples all
consisted of carefully selected macroscopic (terrestrial)
plant remains (seeds) and the material was "fresh" (not
rounded by transportation). The dates are consistent
with stratigraphy with the younger dates close to the
present-day river. OSL ages at both the Partij and
Hommerich sites (Fig. 4) are in the correct stratigraphic
order. However, single-aliquot equivalent doses for
samples Partij 1, and Hommerich 5 and 6 showed large
scatter (equivalent dose histograms provided in Fig. 9).
The scatter is likely due to incorporation of grains for
which the OSL signal was not completely reset upon
burial. Bias due to these outliers was largely removed by
excluding outliers in an iterative procedure as described
in the previous section, but we consider it likely that the
OSL ages for these samples slightly overestimate the
true burial ages. This is confirmed by the slight reversal
between OSL and radiocarbon ages at the Hommerich
section. Equivalent dose histograms for samples Partij 3
and 4 show no indication of heterogeneous resetting
(Fig. 9); OSL ages on these samples are expected to be
accurate estimates of the burial age.
2.1.1. Unit 1
The base of the Holocene valley fill is denoted by a
gravel layer. The top of this layer is often mixed with fine
to coarse sand and detritus. The gravel is poorly sorted,
ranging in size from a few centimetres to more than 15 cm
in diameter. The gravel is of local origin, as it is dominated
by sub-angular flint, which originates from the surrounding
limestones. The depth (from the surface) of the top of
this gravel layer varies in all transects. Differences of up to
two meters occur (e.g. Figs. 3, 7) and the upper surface of
the gravel layer is very irregular. This might be due to the
buried (pre-Holocene) topography of a braided floodplain
with channels. Another option is that the differences are
the result of the lateral migration and the formation of
gravel bars of the Holocene river, like those in the presentday
Geul River.
The gravel has been deposited as a channel deposit
and on channel bars. Since the gravel is often found at
the base of a fining-up sequence, it is probable that the
top of the gravel unit has been reworked by a
meandering system. However, the major body of the
gravel may have been deposited under a different
climate (glacial) and fluvial regime (braided) during
earlier periods (Van de Westeringh, 1980). We do not
have dates to confirm this, but it is very likely that the
gravel unit ranges widely in age.
2.1.2. Unit 2
This unit consists of a fining-up sequence from
gravel with coarse sand to sand loam and loam. The
gravel has a maximum diameter of a few centimetres
and the grain size of the sand varies between 63 m and
841 m with a mode of 250 m (Fig. 8a). No
sedimentary structures have been found in the corings,
but in fresh cutbanks cross-stratification is often visible.
The lower parts of this unit often contain organic
detritus like small twigs, leaves and other macroscopic
plant remains. Deposition of such detritus also takes
place in the present-day river during the flood stage on
the lower parts of the point bar. The unit basically
consists of two dominating lithologies: the coarse
fraction (gravel and coarse sand) represents the lower
point bar, while the finer fraction (fine sand) represents
the middle point bar. This fining-up sequence is typical
for point-bars. Unit 2 is present across almost the whole
valley and it forms part of the lateral accreting
floodplain (cf. Wolman and Leopold, 1957; Nanson
and Croke, 1992). Unit 2 is especially well developed in
the relatively young parts of the floodplain where active
lateral migration takes place.
Fig. 3. Coring transect Schoutenhof.
323J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Fig. 4. Coring transect Hommerich.
324J.J.W.deMooretal./Geomorphology98(2008)316­339
Most dates of this unit are from the base, near the
contact with the gravel of unit 1 (Figs. 3­7), thereby
giving a maximum age estimate for this unit. The dates
show a wide range of ages for unit 2. The oldest (Late
Glacial) parts of this unit have been found in transects
Burggraaf (10188­9809 BC, Fig. 5), Genhoes (11134-
10926 BC, Fig. 6) and Vroenhof (11635-11330 BC,
Fig. 7). In all transects, the youngest dates occur close to
the present-day channel. The majority of dates in unit 2
have a Late Holocene age, indicating the young
character of this unit (see Figs. 3­7 and Table 4). As
most of the dates in this unit have been obtained just
above the top of unit 1, it also gives a minimum age
estimate for unit 1.
2.1.3. Units 3 and 4
The most widespread and in general thickest units are
3 and 4. These units have a nearly identical texture of silt
loam and are very homogeneous and often bioturbated.
The modal grain size for both units is 31­37 m. Unit 3
contains 11.5­15% clay, 65­72% silt and 9­17% sand
(Fig. 8b). Unit 4 contains 12­13.5% clay, 64­76% silt
and 6.5­17% sand (Fig. 8c). No sedimentary structures
are visible in the sediment. The lower part of unit 3
probably represents the upper part of a point bar
succession. This silt loam of the upper point bar
represents the transition zone between lateral and
vertical accreting sediments. The upper part of unit 3
is interpreted as a floodplain unit, formed by vertical
accretion of overbank sediments. In the cores it is
impossible to distinguish the upper point bar from the
overbank floodplain sediments. That is why in all
transects (Figs. 3­7) we combined these sediments into
one unit (unit 3). In transects Burggraaf West, Burggraaf
East and Vroenhof (Figs. 5, 7) the silt loam is deposited
purely as vertical accreting floodplain sediment, because
it overlies a silty clay loam or peat and is therefore not
part of a point bar fining-up sequence. This is unit 4. The
grain size distribution of both units strongly resembles
loess. The low sand content (Fig. 8b/c) indicates fluvial
reworking of the loess (cf. Vandenberghe et al., 1993).
Units 3 and 4 have the largest thickness in most transects
(Figs. 3, 5, 7) and therefore the vertical accreting
floodplain is an important sedimentary environment in
the Geul catchment. Together with the lateral accreting
point-bar, the vertical accreting overbank sediments
form the main components of the floodplain (cf. Nanson
and Croke, 1992).
We only have OSL dates from these units; 14
C dates
were not obtained as these units hardly contain any
organic material. Two OSL samples (Table 5) from a cutbank
exposure near transect Hommerich (Fig. 4) returned
Late Glacial ages (12.20.7 and 12.40.7 ka for Partij 3
and 4, respectively). Sample Hommerich 5, taken near the
same location but on the other side of the river (Fig. 4)
yielded a Late Holocene age (2.400.16 ka). Based on the
dates from unit 2, we can get a maximum age for these
units (see Figs. 3­7 and Table 4). We suggest that most of
the sediments in units 3 and 4 have been deposited since
Medieval times, although locally much older deposits are
preserved as is indicated by the Late Glacial OSL ages
obtained on the Partij samples.
2.1.4. Unit 5
This unit has been found in transects Burggraaf,
Genhoes and Vroenhof (Figs. 5­7) and consists of
(humic) silty clay loam. No sedimentary structures have
been observed. The sediments are very compact. The
fine-grained character (14.5­28.5% clay; 69­80% silt;
0.5­5% sand; mode 13­26 m, Fig. 8d) points to a low
energy depositional environment, relatively far away
from the active channel. Compared to unit 4, the
conditions were quieter with less sediment available. No
dates have been obtained for these sediments, age
determination can only be made using dates from other
units.
The lower 20­30 cm of this unit in transect Genhoes
(Fig. 6) are more humic and have a darker colour. The
Fig. 5. a. Coring transect Burggraaf West. b. Coring transect Burggraaf East.
325J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Fig. 5 (continued ).
326J.J.W.deMooretal./Geomorphology98(2008)316­339
Fig. 6. Coring transect Genhoes.
327J.J.W.deMooretal./Geomorphology98(2008)316­339
Fig. 7. Coring transect Vroenhof. General legend for Figs. 3­7.
328J.J.W.deMooretal./Geomorphology98(2008)316­339
humic character is caused by partial decomposition of
vegetation growing on relatively wet parts of the
floodplain. Similar humic clays have been found in
numerous river valleys in Germany and have been
called "Black Floodplain Soil" or "Black Meadow Soil"
(Lang and Nolte, 1999; Rittweger, 2000; Dambeck and
Thiemeyer, 2002; Houben, 2003; Heine and Niller,
2003; Kalis et al., 2003). Similar dark brown organic
clays of Preboreal age have been found by Pastre et al.
(2001) in several river catchments in the Paris Basin,
France. One date in transect Genhoes (Fig. 6) has been
obtained from this unit, revealing a Preboreal age
(8564­8228 BC).
2.1.5. Unit 6
This unit has been found in transect Burggraaf
(Fig. 5), transect Genhoes (Fig. 6) and transect Vroenhof
(Fig. 7). In transect Burggraaf (Fig. 5), this unit consists
predominantly of humic silt loam with some alternating
thin layers of silt loam. In transect Genhoes (Fig. 6), this
unit consists of alternating thin layers of peat (with
numerous wood fragments) and humic loam, also with
wood fragments. Towards the top of this unit, the
sediment becomes more clastic. The maximum thickness
of this unit is 1 m. In transect Vroenhof (Fig. 7), this unit
consists of a fining-up sequence of loam and (humic) silt
loam. Although the lithology differs slightly at these
three locations, the sediments were deposited in the same
depositional environment. The horizontal bedding and
the topographic position in the three locations point
towards deposition in an abandoned channel or temporarily
inactive parts of the channel. The base of this unit
has been dated at all three locations, revealing ages
between 1694­1528 BC (transect Genhoes, Fig. 6) and
400­350 BC (transect Burggraaf, Fig. 5b).
2.1.6. Unit 7
This unit was found only in transect Burggraaf
(Fig. 5) on both sides of the river. Horizontally layered
silt loams and fine sand (105­250 m) alternate (clay
10­11%; silt 53­55%; sand 29­30%; Fig. 8e). Fig. 8e
also shows the bi-modal grain-size distribution for this
unit. This implies that the sediment composition may
be based on two different sediment sources. The first
Table 3
Characteristics of river deposits and sedimentary environments for the Geul River catchment (modified after Schumm, 1977; Knighton, 1998)
Place of
deposition
Name Characteristics Lithology and sedimentary structures Unit Transect
Channel Lag
deposit/
channel
bars
Consist of the coarsest material transported
during floods
Gravel (2­15 cm) 1 All
Channel
fill
Accumulation of predominantly fine-grained
sediments and peat in abandoned channel
reaches and scour holes
Loam and silt loam, peat and humic
silt loam Mainly horizontally
layered
6 Genhoes,
Burggraaf,
Vroenhof
Channel
margin
Lateral
accretion
deposits
Point-bars formed by lateral accretion on the
convex bank of meanders. Preserved through
channel shifting
(lateral migration) and added to the floodplain
by vertical accretion deposits at the top
Fining-up sequence from gravel to sand
loam and loam to silt loam (unit 3)
2,3 All
Floodplain Vertical
accretion
deposits
Fine-grained sediment deposited from
suspended load of overbank floodwater;
In general horizontally layered
Natural levee Sand loam and loam 7 Burggraaf
Floodplain Silt loam 3,4 All
Distal floodplain (humic) silty clay loam 5 Burggraaf,
Genhoes,
Vroenhof
Backswamp Peat, humic silty clay loam 8 Burggraaf,
Vroenhof
Valley
margin
Colluvium Deposits derived mainly from unconcentrated
slope wash and soil creep on adjacent valley sides
Predominantly silt loam with
fragments of brick, tile, coal and
charcoal
10 Vroenhof
Alluvial
fan
Formed by ephemeral streams emerging from steeply
dissected terrain on to a river valley; sediments rapidly
decrease in grain size with distance from the fan apex
although grainsize in Geul catchment is rather uniform
Loam and silt loam with fragments
of limestone, gravel, brick, tile, coal
and charcoal
9a/b Hommerich,
Genhoes
329J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
source would be reworked loess, eroded from the
hillslopes (mode is 37 m). The second source is likely
to be eroded sand from early Cretaceous deposits (mode
is 297 m). On the west bank of the river the sediments
become finer and the unit becomes thinner farther away
from the present channel.
Fig. 8. a­g. Typical grain-size distributions for units 2, 3, 4, 5, 7, 9a and 9b.
330 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
This unit is interpreted as a natural levee, because of
the characteristic topographic position (slightly higher
than the rest of the floodplain), the horizontal lamination
and the coarser grain size and higher sand content than
the surrounding floodplain silt loams (units 3 and 4).
During flood stages, the river is capable of transporting
medium sized sand and it will deposit the sediments
close to the channel as a natural levee. The upper parts
of this unit have characteristic dark brown and dark grey
colours, pointing to contamination with lead and zinc,
giving a clear age indication for this unit (cf. unit 2 from
Stam, 2002).
2.1.7. Unit 8
This unit consists almost entirely of peat, humic silt
loam and humic silty clay loam and is present at the valley
sides in the Burggraaf and Vroenhof transects (Figs. 5, 7).
The peat in the Burggraaf transect (Fig. 5) contains large
fragments of wood, indicating a waterlogged woodland
environment. A few thin (several cm's) clastic layers were
found in the Burggraaf transect, indicating (short) periods
of increased river activity. The peat in transect Vroenhof
(Fig. 7) is predominantly sedge peat containing some
wood fragments. In this transect two phases of backswamp
development can be distinguished, separated by a
Table 4
Radiocarbon dating results
Location/sample name Lab number Material Elevation (m+NAP) Depth (m) Method Age (14
C yr BP) Cal 2
Schoutenhof 3 GrA-26957 Seeds 107.45­107.55 2.4­2.5 AMS 442540 3124­2920 BC
Schoutenhof 5 GrA-26960 Seeds/1 beechnut 107.71 2.7 AMS 137035 602­694 AD
Schoutenhof 8 GrA-26947 Seeds 107.76­107.86 2.6­2.7 AMS 216035 262­95 BC
Schoutenhof 18 GrA-26950 Seeds 108.09­108.26 1.7­1.9 AMS 82535 1157­1272 AD
Hommerich 5 GrA-26952 Seeds 93.24­93.34 5.0­5.1 AMS 293035 1222­1018 BC
Hommerich 15 GrA-26953 Seeds 93.98­94.08 5.6­5.7 AMS 519040 4068­3942 BC
Hommerich 21 GrA-26955 Seeds 94.01­94.11 2.6­2.7 AMS 97035 1013­1158 AD
Hommerich 23 GrA-26956 Seeds 94.08­94.18 3.4­3.5 AMS 167035 312­434 AD
Hommerich 46 GrA-27291 Seeds/leaves 93.65­93.75 3.05­3.15 AMS 223035 331­203 BC
Burggraaf 39 GrA-26942 Seeds 84.52­84.62 3.3­3.4 AMS 84535 1152­1265 AD
Burggraaf 43 GrA-26961 Seeds/leaves 85.13­85.28 2.8­2.95 AMS 98035 1064­1155 AD
Burggraaf 70 GrA-26946 Seeds 85.89­85.99 2.6­2.7 AMS 1023050 10188­9809 BC
Burggraaf 53 GrA-26943 Seeds 83.75­83.90 3.7­3.85 AMS 95535 1019­1160 AD
Burggraaf 62 GrA-26945 Charcoal/seeds 83.03­83.13 3.9­4.0 AMS 227535 400­350 BC
Genhoes 39 GrA-27289 Seeds 74.94­75.14 5.1­5.3 AMS 81035 1167­1273 AD
Genhoes 55 GrA-27290 Seeds 72.03­72.23 6.7­6.9 AMS 193535 3 BC­132 AD
Genhoes 16 GrA-27285 Seeds 73.93­74.03 3.8­3.9 AMS 88035 1039­1223 AD
Genhoes 22 GrA-27287 Seeds 70.43­70.63 7.2­7.4 AMS 334535 1694­1528 BC
Genhoes 9 GrA-27284 Seeds/charcoal 72.63­72.83 3.7­3.9 AMS 152035 432­610 AD
Genhoes 5 GrA-27283 Seeds/wood 70.03­70.43 4.8­5.2 AMS 1104060 11134­10926 BC
Genhoes 3 GrA-27305 Seeds/wood 71.6­71.8 2.8­3.0 AMS 64035 1337­1398 AD
Genhoes 7 GrN-29025 Humic silty clay 71.4­71.6 3.9­4.1 Conventional 913080 8564­8228 BC
Vroenhof 28 GrA-27296 Seeds 54.29­54.49 2.3­2.5 AMS 120535 763­895 AD
Vroenhof V9.4 GrA-27293 Seeds/wood 52.92­53.09 3.54­3.71 AMS 254035 800­727 BC
Vroenhof 47 GrA-27307 Seeds 54.75­54.85 1.1­1.2 AMS 96035 1017­1159 AD
Vroenhof 30 GrA-27308 Seeds/wood 52.63­52.83 3­3.2 AMS 1157060 11635­11330 BC
Vroenhof V25.1 GrA-27294 Seeds 55.2­55.35 0.85­1 AMS 130535 657­774 AD
Vroenhof V25.12 GrA-27295 Seeds/wood 53.41­53.56 2.64­2.79 AMS 319035 1523­1408 BC
Table 5
Quartz OSL dating results
Sample name NCL code Depth Elevation Water content Radionuclide concentration (Bq kg -1
) Total Dose rate Equivalent Optical
(m) (m+NAP) % 238
U 232
Th 40
K (Gy ka- 1
) Dose (Gyr) age (ka)
Partij 1 6404046 0.85 96.91 205 40.10.3 39.81.0 5266 2.730.12 1.480.07 0.540.04
Partij 3 6404047 2.40 95.36 215 52.50.3 45.90.9 6175 3.170.14 38.71.3 12.20.7
Partij 4 6404048 3.00 94.76 265 46.10.5 46.10.5 5856 2.900.12 36.11.2 12.40.7
Hommerich 6 6504050 1.10 95.70 235 37.50.4 37.60.7 4674 2.440.10 2.630.11 1.080.06
Hommerich 5 6504049 2.60 94.20 215 33.20.4 26.00.7 3265 1.890.08 4.540.22 2.400.16
331J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
thin clastic interval. This silt loam interval is probably
related to a period with more river activity. As both units
are situated at the edge of the floodplain, it is very likely
that they have been formed in a backswamp environment.
Wet conditions at these locations were favoured by the
occurrence of seepage, especially on the eastern side of
the Burggraaf transect (see also Van de Westeringh,
1979). No clear age estimate of this unit in transect
Burggraaf is available, only a relative age estimate. The
dates in the Vroenhof transect (Fig. 7) indicate that the end
Fig. 9. Histograms of equivalent dose obtained on single aliquots using the single-aliquot regenerative-dose procedure depicted in Table 1. To avoid
overestimation of the burial age due to incomplete resetting of the OSL signal in some grains at the time of burial, aliquots returning a value greater
than two standard deviations from the sample mean were discarded in an iterative procedure. These points are shown as open circles in the graphs.
The sample equivalent dose used for age calculation was obtained on the aliquots remaining after iteration. Note that uncertainties indicated on the
graph are only random errors and do not include beta source calibration uncertainties.
332 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
of peat formation took place during the Middle Ages
(657­774 AD and 1017­1159 AD).
2.1.8. Unit 9 a/b
This unit, which is present in transects Hommerich
and Genhoes (Figs. 4, 6), has been divided in a relatively
coarse-grained loam (clay 7­9%; silt 55­69%; sand 16­
30%; mode 37 m, Fig. 8f) and a finer-grained silt loam
(clay 7.5­14.5%; silt 69­80%; sand 4­9%; mode 31­
37 m, Fig. 8g). Both subunits are characterized by
numerous coal, charcoal, brick and tile fragments. Clear
sedimentary structures have not been found, but small
coarsening-upward sequences occur at the east side of
transect Hommerich. The grain size distribution of the
finer-grained subunit 9b (see Fig. 8g) is similar to units 3
and 4, which points to a common loess source. However,
the depositional environment is completely different.
This is also expressed by the morphology (Fig. 2c/e) of
this unit and the thickness, which increases towards the
valley sides. The sediments have been deposited on an
alluvial fan by ephemeral streams from relatively steep
dissected terrain. In transect Genhoes (Fig. 6) the
sediments rapidly decrease in grain size with distance
from the fan apex. However, this is clearly not the case in
transect Hommerich (Fig. 4), probably because the
catchment area of the alluvial fan is smaller and hence the
discharge and transport capacity is lower, leading to less
grain-size differentiation on the alluvial fan. We interpret
the finer sediments (unit 9b) as the more distal parts and
the coarser (loam) sediments (unit 9a) as the proximal
part of the alluvial fan. It is difficult to distinguish by
grain size the alluvial fan sediments from floodplain
sediments (units 3 and 4) in transects Genhoes and
Hommerich because of the same loess-derived sediment
source. We differentiated the alluvial fan sediments from
floodplain sediments by the occurrence of brick, coal and
charcoal fragments, the presence of coarsening-up
sequences in alluvial fan sediments and the slightly
finer grain size of the underlying overbank floodplain
sediments.
Dates (OSL) from the Hommerich transect (Fig. 4)
indicate that the alluvial fans on both sides of the river
were active during and after the Middle Ages (1.08
0.06 ka and 0.540.04 ka) and that initial alluvial fan
activity probably started before or during the Roman
Period (2.400.16 ka and 331­203 BC). Activity of the
alluvial fan in transect Genhoes (Fig. 6) was probably
also initiated during the Roman Period (3 BC­132 AD).
This alluvial fan had been very active during and after the
High Middle Ages (1167­1273 AD and 1039­1223
AD). The numerous tile, brick and coal fragments in the
top of the alluvial fans point to a young age. The presence
of charcoal may point to burning activities on the hillsides,
either natural of induced by man.
2.1.9. Unit 10
This unit is very similar to unit 9b and has been found at
the eastern side of the Vroenhof transect (Fig. 7). The
sediments have been deposited by unconcentrated slope
wash and soil creep from adjacent valley slopes (colluvium).
Such colluvial deposits are very common in the
western and central European loess area and are the result
of deforestation phases during the last several thousand
years (e.g. Brown, 1992; Lang, 2003). No dates have been
obtained from this unit, but a date of 657­774 AD in the
underlying unit 8 indicates that the sediments of unit 10
must have formed during and after the Middle Ages.
2.1.10. Unit 11
In the western- and easternmost cores in transect
Hommerich (Fig. 4) we found a mixture of loam, sand and
gravel (predominantly clasts of limestone). As the
locations of this unit are very close to the valley edge, it
is very likely that these sediments are an infilling of a
tributary or are caused by local valley/slope processes.
They have probably not been deposited by the Geul River.
2.1.11. Synthesis of sedimentation processes
Based on the lithogenetic units and their volumes we
conclude that Geul River sedimentation is dominated by
four main processes (Table 3):
­ lateral erosion and channel scour forming a basal gravel
unit (unit 1),
­ lateral accretion deposits of sand and loam forming
point bars across mostof theriver valley(units 2 and 3),
­ vertical accretion forming a thick floodplain consisting
of overbank silt loam and silty clay loam (units 4,
5, 7, 8),
­ loam and silt loam deposition on alluvial fans with
thicknesses up to seven metres (unit 9 a/b).
3. Holocene valley development -- alluvial
architecture through time
3.1. Overview
Sedimentation in the Geul River catchment throughout
the Holocene has been dominated by lateral accretion
of point bar sediments and by vertical accretion of
overbank silt loam. There are, however, differences
between the upstream and downstream areas of the
catchment. The thickness of the overbank fines (units 3
and 4) increases in a downstream direction, from about
333J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
2 m in transect Schoutenhof (Fig. 3) to 2.5 m in transect
Burggraaf (Fig. 5) and almost 3 m in transect Vroenhof
(Fig. 7). The width of the floodplain increases from
250 m upstream (near the Belgian­Dutch border) to
more than 600 m close to the confluence with the Maas
River. The gradient decreases downstream from 0.005 m
m-1
near the Belgian­Dutch border to 0.0015 m m- 1
near the confluence with the Maas. The wider floodplain
(see Fig. 2b­f) and the lower transport capacity resulted
in a larger variety of fluvial environments. The
occurrence of extensive backswamp deposits in the
downstream transect Vroenhof (Fig. 7) is especially
striking, compared to transect Schoutenhof (Fig. 3)
where only overbank fines and point bar deposits are
Table 6
Synthesis diagram of depositional environments, vegetation and land-use history and phases of decreased and increased sedimentation
Partly based on data from (Havinga and Van den Berg van Saparoea, 1980; Van de Westeringh, 1980; Renes, 1988; Bunnik, 1999; Bazelmans et al., 2004).
334 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
present. Apparently due to the narrower floodplain at
Schoutenhof (see also Fig. 2a, e) and the more energetic
conditions (because of the higher gradient), complete
reworking of the sediment across the valley occurred,
while in the other transects (with a wider valley) some
isolated Late Glacial sediments are still present. In
addition, a slight trend of downstream fining of the
sediments is visible: in the downstream transects
Genhoes and Vroenhof (Figs. 6, 7) thicker units of silty
clay loam occur and especially in transect Vroenhof peat
and humic sediments are present.
3.2. Fluvial architecture
Late Glacial sedimentation consisted of gravel and
sand and local fine-grained (clayey) floodplain formation
(e.g. transect Hommerich, Fig. 4). Large parts of the
Late Glacial sediments have been eroded and very few
Early and Middle Holocene deposits are preserved in the
catchment. The only Early Holocene unit still present is
the lower part of unit 5 (the "Black Floodplain Soil") in
transect Genhoes (8564­8228 BC, Fig. 6). Limited
sediment availability during the Early and Middle
Holocene resulted in predominantly low rates of finegrained
overbank sedimentation and local peat formation.
During this period, the Geul River was probably a
quiet and stable river. Later in the Holocene (the last
2000 yr) more sediment became available and the Geul
River changed into a more dynamic system with active
lateral migration. The increasing river dynamics clearly
had a limiting effect on the preservation potential of
older sediments in the floodplain (cf. Lewin and
Macklin, 2003), as much of the older sediment has
been removed by the laterally migrating river. Sedimentation
during the last 2000 yr was dominated by
overbank sedimentation and is reflected in the widespread
occurrence of units 3 and 4. Numerous
radiocarbon dates (Table 4) in all transects (Figs. 3­7)
illustrate that the main phase of fine overbank
sedimentation (units 3 and 4) took place since High
Middle Ages (1000­1500 AD). The fine-grained overbank
sediments (unit 5) in transect Burggraaf and
Vroenhof (Figs. 5, 7) represent a period with less
sediment availability and a less dynamic environment.
Bank stabilisation measures during the last few
centuries have resulted in local formation of natural
levees (unit 7). The Geul River near transect Burggraaf
(Fig. 5) has been stable for at least 200 yr (based on
analyses of historical maps). The age for the natural
levee at this location is supported by the presence of
sediment contaminated with lead and zinc, related to
19th century mining (e.g. Stam, 2002). Formation of this
unit is probably (partly) induced by the effect of the lead
and zinc mining that started at the beginning of the 19th
century.
An evaluation of the thickness of the fining-upward
sequences in the fluvial deposits shows clear changes in
fluvial dynamics through time. During the Early and
Middle Holocene, sedimentation was dominated by point
bar deposition and limited overbank sedimentation. The
fining-up sequences from this period have a maximum
thickness of about 1 m (Fig. 7, above the date of 800­727
BC), indicating that the (bankfull) channel depth by that
time was also about 1 m. This changed during the Roman
Time. Fining-up sequences from this period show a
thickness of up to two meters (Fig. 4, above the dates of
2.400.16 ka, 331­203 BC and 312­434 AD), giving a
bankfull channel depth of about 2 m. Fining-up sequences
from the High Middle Ages have a thickness of 2­3 m (for
example Fig. 4, above the date of 1013­1158 AD and
several dates in transect Burggraaf (Fig. 5)), giving
maximum bankfull channel depths of about 3 meters. The
increasing thickness of the fining-up sequences could
represent an increase in bankfull discharge as a result of
deforestation. Our field data indicate that increased
discharges have not resulted in incision of the river (the
level of the basal gravel (unit 1) is constant through the
whole catchment). However, alternative interpretations
for an increasing bankfull depth (discharge) can be (1) a
multiple channel system as an effect of high vegetation
density, cohesive banks and low sediment load, or (2) a
higher width-depth ratio of the channel forced by the near
surface presence of bedrock (Palaeozoic/Mesozoic basement).
According to Nota and Van de Weerd (1978),
bedrock is within 1 m below the present-day Gulp River
(the main tributary of the Geul River). For these two
hypotheses, the bankfull discharge could have been
invariable. Unfortunately, we have no data to constrain
the paleowidth of the Geul channel.
3.3. Alluvial fan architecture
Transects Genhoes and Hommerich (see Figs. 2c/e and
5/6) largely consist of thick alluvial fan units. The 14
C and
OSL dates illustrate that several periods of alluvial fan
activity can be recognized in the Genhoes transect. The
first phase of alluvial fan sedimentation took place during
the Roman Period (post 3 BC­132 AD, Fig. 6). After the
Roman Period activity of the alluvial fan was limited;
locally some organic deposits formed, indicating humid
and stable (vegetated) conditions. A second major
activation phase of the alluvial fan took place during the
High Middle Ages. This second phase is illustrated by two
14
C dates in the Genhoes transect (Fig. 6) (1167­1273 AD
335J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
and 1039­1223 AD). The alluvial fan sediments (units 9
a/b) cover older fluvial sediments (units 3 and 5). The
dates in the Hommerich transect (Fig. 4) also confirm the
Late Holocene age of the alluvial fans. Small variations in
grain size in the alluvial fan deposits represent the
proximal and distal parts of the alluvial fan. On the
proximal side of the alluvial fan on the east bank of the
river (Fig. 4) small coarsening-up sequences are present,
indicating the progradingcharacter of the fan during active
phases. The alluvial fan sediments in this transect (Fig. 4)
also cover older fluvial deposits, consisting of point bar
and overbank floodplain sediments (units 2 and 3). The
dates (OSL and 14
C, see Fig. 4) from the latter deposits
indicate that the overlying alluvial fans probably started to
form in the Roman Period or later and were active during
and after the High Middle Ages. The tile, brick and
charcoal fragments in the alluvial fan sediments also
indicate the young character of the alluvial fans.
In the alluvial fans a clear difference exists between a
coarser-grained unit (9a) and a finer-grained unit (9b),
related to the distance to the fan apex and the size of the
sub-catchment of the alluvial fan. The thick unit 9a in
transect Genhoes points to intensified soil erosion and
higher transport rates from the large catchment feeding the
Genhoes fan. The fan deposits have a maximum thickness
of about 4 m in transect Hommerich and about 7 m at the
apex in transect Genhoes. The dates in transect Genhoes
clearly show an increase in alluvial fan sedimentation
rates during the Roman Period (about 2.5 m of
sedimentation) and the High Middle Ages (locally about
5 m of sedimentation).
4. Discussion: humanand climate impact on catchment
development
The Geul River catchment has experienced considerable
changes in fluvial dynamics and sedimentation rates
during the Holocene. Knox (1995) identified five major
controls on fluvial systems (climate, vegetation, tectonics/
eustatic base-level change, human impact and intrinsic
factors). Climate has often been regarded as the major
control on fluvial systems in Western Europe during glacial
and interglacial timescales (Veldkamp and Van den Berg,
1993; Vandenberghe, 1995, 2003). But on shorter timescales,
other controls (like human activities) are likely to be
of more importance, especially on relatively small river
catchments (e.g. Macklin et al., 2005; Starkel, 2005).
Although several of these forcing mechanisms can be
responsible for theobserved changes, it appears that human
activities have had the most severe impact in the Geul
River catchment. Heine et al. (2005) state that under natural
conditions (a fully vegetated area) no significant soil
erosion would occur in this type of catchment. According
to Starkel (2005), the intensity of sediment load and slope
wash is up to 4 orders of magnitude higher in cultivated
small catchments (especially in hilly, mountainous regions)
than in natural, completely forested situations.
The land-use and vegetation history of the study area
(as described earlier in the paper, see also Table 6) is well
reflected in the fluvial and alluvial fan sediments.
Accelerated rates of overbank sedimentation and alluvial
fan sedimentation coincide with two major deforestation
phases in the area: the Roman Period (53 yr BC­415 yr
AD) and the High Middle Ages (1000­1500 yr AD),
although deforestation probably started at a smaller scale
during the Iron Age (Bunnik, 1999). Deforestation
caused hillslope erosion and sediment (mainly loess) was
transported to the river, resulting in a thick floodplain
deposit and in the formation of alluvial fans (Table 6). A
period of finer-grained sedimentation (unit 5) can be
related to a period when forest expansion took place and
the population declined (the dark ages, 220­500 AD).
The scarcity of older sediments in the Geul catchment
can be explained by the pollen and archaeological data
(Table 6), which indicate that during the Neolithic and
the Bronze Age only on small-scale agriculture was
practiced (Renes, 1988; Bunnik, 1999). The effect of
these agricultural activities was probably too small to
influence landscape development. Stam (2002) shows
that high sedimentation rates in the Geul catchment in the
nineteenth century were the result of deforestation and
mining activities and that only during the second half of
the twentieth century increased precipitation has contributed
to an increase in sedimentation and channelchange
rate.
The central-German loess area reflects similar
increased sedimentation patterns during the Roman and
Medieval Periods as the Geul River catchment. Heine
et al. (2005) describe Medieval alluvial fan activity in
southern Germany. Lang and Nolte (1999), Lang (2003)
and Zolitschka et al. (2003) found several phases of
increased alluvial fan activity (colluvium sedimentation)
and floodplain sedimentation related to Roman and
Medieval land-use change. Lang (2003) concludes that
the periods of increased sedimentation were triggered by
changes in land use and that climate change only played
a minor role. Lang (2003) and Zolitschka et al. (2003)
also report less alluviation and colluviation in the period
prior to the Iron Age and during the Migration period.
This resembles the pattern in the Geul River catchment.
These authors also state that Holocene flooding events
are hard to interpret as being the result of (short-term)
climate fluctuations or land-use triggered catastrophic
events. Zolitschka et al. (2003) found little evidence for a
336 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
connection between alluviation and climate history
during the late Holocene. Heine and Niller (2003) state
that climate hardly played a role in the landscape
development during the Holocene in Southern Germany
and that anthropogenic activities were responsible for
accelerated soil erosion, formation of alluvial fans and
increased floodplain sedimentation.
Some climatic fluctuations occurred during the Holocene,
although not as distinct as during the Weichselian.
Some of the most well known climate excursions in the
Holocene are the Medieval Warm Period and the Little
Ice Age. Several authors report the effects of the Little
Ice Age and the Medieval Warm period on fluvial
dynamics in Great Britain, Central Europe and Spain
(e.g. Rumsby and Macklin, 1996; Brown, 1998; Starkel,
2002; Brown, 2003; Macklin and Lewin, 2003;
Thorndycraft and Benito, 2006). This is in most cases
expressed by an increase in flood frequency as a result
of events of extreme precipitation, an increase in
floodplain sedimentation and an increase in channel
migration. In transects Burggraaf and Vroenhof
(Figs. 5, 7) clastic layers within the peat may indicate
phases of increased flooding, however our dating
resolution is not sufficient to relate these clastic
sediments with a specific period of increased wetness
during the Holocene. Our cluster of dates from the High
Middle Ages (in alluvial fan and overbank silt loam
units) coincides with three episodes (860, 660 and
570 cal. yr BP) of major flooding in Great Britain,
Poland and Spain (Macklin et al., 2006). It is, however,
very difficult to attribute with any certainty this
increased sedimentation to climate-induced, increased
flooding. It is possible that a combination of large-scale
deforestation and increased wetness during the High
Middle Ages resulted in a dramatic increase in fluvial
and alluvial fan sedimentation (cf. Brown, 1998;
Macklin et al., 2006).
Although the effect of climate on catchment development
during the last few thousands of years is not as
clearly documented as the effects of land-use change,
climate change could play a role in the future development
of the Geul River catchment. As present-day channel and
floodplain forming processes (lateral channel migration
and point bar formation) predominantly take place during
high discharges, it is very likely that a future increase of
high-intensity precipitation events (as projected by the
IPCC, 2001) will have an impact on the catchment. Due to
the short response time of the catchment to high
precipitation events, the discharge of the river can rise
very rapidly. With increasing high discharge events,
lateral migration rates and reworking of sediments would
be expected to increase.
5. Conclusions
The results of this study show that several periods of
decreased and increased sedimentation can be linked
with periods of distinct changes in land use (Table 6).
During the Early and Middle Holocene, sedimentation
rates were low and deposition was dominated by lateral
accretion of point-bar sediments and limited overbank
deposition. There was a continuous reworking of the
sediments due to the lateral migrating river. Deforestation
during the Roman Period and the High Middle Ages
resulted in severe soil erosion and consequently in
increased overbank floodplain sedimentation and the
formation of alluvial fans in the valley of the Geul.
Following the massive sediment input in the system,
vertical accretion dominated the sedimentation pattern
and the Geul was able to build up a floodplain of over
2 m above the present river bed.
This study shows that changes in land use have been
the main trigger for this increase in sedimentation,
although climate factors, like increased flooding due to
extreme precipitation events, cannot be ruled out and
might have intensified the erosion and sedimentation
processes. Our results indicate that a small river
catchment in a loess area is very sensitive to changes
in land use. This implies that future changes in land use
will have a profound effect on fluvial processes in the
Geul River catchment and can, in combination with a
future increase in extreme precipitation events, have
severe consequences regarding flood risk and ecological
value.
Acknowledgements
This study is part of the "Wege des Wassers" project
and is partly funded by the European Union Interreg
III -- A program of the Euregio Maas-Rhine (contract
number EMR.INT3 06.02­3.1.28). We thank Esther
Bootsma, Florian ten Horn and Boris Webbers for their
assistance in the field, and Candice Johns for processing
the OSL samples. Martin Konert and Maurice Hooyen
are thanked for their assistance with the grain-size
analyses. Facilities for grain-size analyses and 14
C
sample preparation were provided by the Vrije Universiteit
Amsterdam. Sjoerd Bohncke is thanked for
helping to identify the macroscopic plant remains.
Critical and constructive comments on this paper by
Andy Howard and Tim van der Schriek have been
greatly appreciated. Equipment at the Netherlands
Centre for Luminescence dating is partly financed by
the Netherlands Organisation for Scientific research
through investment grant 834.03.003.
337J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
References
Bazelmans, J., Bakels, C., Kocken, M., 2004. De Romeinse wachtpost
op de Goudsberg, Valkenburg aan de Geul (Limburg). Historische
en Heemkundige Studies in en rond het Geuldal 14, 61­86.
Berendsen, H.J.A., Stouthamer, E., 2001. Palaeogeographic development
of the Rhine­Meuse delta. Van Gorcum, Assen.
Brookes, A., 1988. Channelized rivers: Perspectives for environmental
management. Wiley, Chichester. 326 pp.
Brown, A.G., 1992. Slope erosion and colluviation at the floodplain
edge. In: Bell,M., Boardman, J. (Eds.), Past and Present Soil Erosion,
vol. 22. Oxbow books, Oxford, pp. 77­88.
Brown, A.G., 1998. Fluvial evidence of the medieval warm period and
the Late Medieval climatic deterioration in Europe. In: Benito, G.,
Baker, V.R., Gregory, K.J. (Eds.), Palaeohydrology and Environmental
Change. Wiley, Chichester, pp. 43­52.
Brown, A.G., 2003. Global environmental change and the palaeohydrology
of western Europe: a review. In: Gregory, K.J., Benito, G.
(Eds.), Palaeohydrology: Understanding Global Change. Wiley,
Chichester, pp. 105­121.
Bunnik, F.P.M., 1999. Vegetationsgeschichte der Löbörden zwischen
Rhein und Maas von der Bronzezeit bis in die frühe Neuzeit.
Laboratory of Palaeobotany and Palynology, Utrecht University,
Utrecht. 149 pp.
Dambeck, R., Thiemeyer, H., 2002. Fluvial history of the northern
Upper Rhine River (southwestern Germany) during the Late glacial
and Holocene times. Quaternary International 93­94, 53­63.
doi:10.1016/S1040-6182(02)00006-X.
De Laat, P.J.M., Agor, M.L.C., 2003. Geen toename van piekafvoer van
de Geul. H2O 36 (9), 25­27.
Havinga, A.J., Van den Berg van Saparoea,R.M., 1980. Former vegetation
and sedimentation in the valley of the river Geul. Mededelingen van de
Landbouwhogeschool te Wageningen 80, 47­59.
Heine, K., Niller, H.-P., 2003. Human and climate impacts on the
Holocene landscape development in Southern Germany. Geographia
Polonica 76, 109­122.
Heine, K., Niller, H.-P., Nuber, T., Scheibe, R., 2005. Slope and valley
sediments as evidence of deforestation and land-use in prehistoric
and historic Eastern Bavaria. Zeitschrift für Geomorphologie N.F.
Suppl. 139, 147­171.
Houben, P., 1997. Late-glacial and Holocene fluvial sedimentation in a
small upland catchment in Hesse (Germany). Zeitschrift für
Geomorphologie 41, 461­478.
Houben, P., 2002. Die räumlich-zeitlich veränderte Reaktion des
fluvialen Systems auf jungquartäre Klimaänderungen. Eine Fallstudie
aus der Hessischen Senke. Frankfurt am Main, Johann
Wolfgang Goethe-Universität. 189 pp.
Houben, P., 2003. Spatio-temporally variable response of fluvial systems
to Late Pleistocene climate change: a case study from central
Germany. Quaternary Science Reviews 22, 2125­2140. doi:10.1016/
S0277-3791(03)00181-1.
IPCC, 2001. Climate Change 2001: The Scientific Basis. In: Houghton,
J.T., Ding, Y., Griggs, D.J., Noguer, M., Van der Linden, P.J., Dai,
X., Maskell, K., Johnson, C.A. (Eds.), Contribution of Working
Group I to the Third Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge University Press, Cambridge.
881 pp.
Kalis, A.J., Merkt, J., Wunderlich, J., 2003. Environmental changes
during the Holocene climatic optimum in central Europe -- human
impact and natural causes. Quaternary Science Reviews 22, 33­79.
doi:10.1016/S0277-3791(02)00181-6.
Klimek, K., 2002. Human-induced overbank sedimentation in the
foreland of the Eastern Sudety Mountains. Earth Surface Processes
and Landforms 27, 391­402.
Knighton, D., 1998. Fluvial forms and processes -- a new perspective.
Edward Arnold, London. 383 pp.
Knox, J.C., 1995. Fluvial systems since 20000 years BP. In: Gregory,
K.J., Starkel, L., Baker, V.R. (Eds.), Global Continental Palaeohydrology.
Wiley, Chichester, pp. 87­108.
Konert, M., Vandenberghe, J., 1997. Comparison of laser grain size
analysis with pipette and sieve analysis: a solution for the
underestimation of the clay fraction. Sedimentology 44, 523­535.
Kukulak, J., 2003. Impact of mediaeval agriculture on the alluvium in
the San River headwaters (Polish Eastern Carpathians). Catena 51,
255­266. doi:10.1016/S0341-8162(02)00165-0.
Lang, A., 2003. Phases of soil erosion-derived colluviation in the loess
hills of South Germany. Catena 51, 209­221. doi:10.1016/S0341-
8162(02)00166-2.
Lang, A., Nolte, S., 1999. The chronology of Holocene alluvial
sediments from the Wetterau, Germany, provided by optical and
14
C dating. The Holocene 9, 207­214.
Lang, A., Bork, H.-R., Mäckel, R., Preston, N., Wunderlich, J., Dikau,
R., 2003. Changes in sediment flux and storage within a fluvial
system: some examples from the Rhine catchment. Hydrological
Processes 17, 3321­3334. doi:10.1002/hyp.1389.
Leenaers, H., 1989. The dispersal of metal mining wastes in the
catchment of the river Geul (Belgium­The Netherlands). Netherlands
Geographical Studies, vol. 102. Koninklijk Aardkundig
Genootschap/Geografisch Instituut Rijksuniversiteit Utrecht,
Amsterdam/Utrecht. 200 pp.
Lewin, J., Macklin, M.G., 2003. Preservation potential for Late
Quaternary river alluvium. Journal of Quaternary Science 18,
107­120. doi:10.1002/jqs.738.
Mäckel, R., Schneider, R., Seidel, J., 2003. Anthropogenic impact
on the landscape of southern Badenia (Germany) during the
Holocene -- documented by colluvial and alluvial sediments.
Archaeometry 45, 487­501.
Macklin, M.G., Lewin, J., 2003. River sediments, great floods and
centennial-scale Holocene climate change. Journal of Quaternary
Science 18, 101­105. doi:10.1002/jqs.751.
Macklin, M.G., Johnstone, E., Lewin, J., 2005. Pervasive and longterm
forcing of Holocene river instability and flooding in Great
Britain by centennial-scale climate change. The Holocene 15,
937­943. doi:10.1191/0959683605hl867ft.
Macklin, M.G., Benito, G., Gregory, K.J., Johnstone, E., Lewin, J.,
Michczyńska, D.J., Soja, R., Starkel, L., Thorndycraft, V.R., 2006.
Past hydrological events reflected in the Holocene fluvial record of
Europe. Catena 66, 145­154. doi:10.1016/j.catena.2005.07.015.
Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz
using an improved single-aliquot regenerative-dose protocol.
Radiation Measurements 32, 57­73.
Murray, A.S., Wintle, A.G., 2003. The single aliquot regenerative dose
protocol: potential for improvements in reliability. Radiation
Measurements 37, 377­381.
Murray, A.S., Marten, R., Johnston, A., Martin, P., 1987. Analysis for
naturally occurring radionuclides at environmental concentrations
by gamma spectrometry. Journal of Radioanalytical and Nuclear
Chemistry 115, 263­288.
Nanson, G.C., Croke, J.C., 1992. A genetic classification of floodplains.
Geomorphology 4, 459­486. doi:10.1016/0169-555X(92)90039-Q.
Nota, D.J.G., Van de Weerd, B., 1978. A hydrogeological study in the
basin of the Gulp creek -- a reconnaissance in a small catchment area.
Mededelingen van de Landbouwhogeschool te Wageningen 78, 1­26.
338 J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339
Pastre, J.-F., Limondin-Lozouet, L.N., Gebhardt, A., Leroyer, C.,
Fontugne, M., Krier, V., 2001. Lateglacial and Holocene fluvial
records from the central part of the Paris Basin (France). In:
Maddy, D., Macklin, M.G., Woodward, J.C. (Eds.), River basin
sediment systems; archives of environmental change. Balkema,
Rotterdam, pp. 357­373.
Raab, T., Völkel, J., 2005. Soil geomorphological studies on the
Prehistoric to Historic landscape change in the former mining area at
the Vils River (Bavaria, Germany). Zeitschrift für Geomorphologie
N.F. Suppl. 139, 129­145.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand,
C.J.H., Blackwell, P.G., Buck, C.E., Burr, G.S., Cutler, K.B., Damon,
P.E., Edwards, R.L., Fairbanks, R.G., Friedrich, M., Guilderson, T.P.,
Hogg, A.G., Hughen, K.A., Kromer, B., McCormac, F.G., Manning,
S.W., Ramsey, C.B., Reimer, R.W., Remmele, S., Southon, J.R.,
Stuiver, M., Talamo, S., Taylor, F.W., Van der Plicht, J.,
Weyhenmeyer, C.E., 2004. IntCal04 Terrestrial radiocarbon age
calibration, 26­0 ka BP. Radiocarbon 46, 1029­1058.
Renes, J., 1988. De geschiedenis van het Zuidlimburgse cultuurlandschap.
Van Gorkum, Maastricht. 265 pp.
Rittweger, H., 2000. The "Black Floodplain Soil" in the Amoneburger
Becken, Germany: a lower Holocene marker horizon and indicator
of an upper Atlantic to Subboreal dry period in Central Europe?
Catena 41, 143­164. doi:10.1016/S0341-8162(00)00113-2.
Roberts, R.G., Galbraith, R.F., Olley, J.M., Yoshida, H., Laslett, G.M.,
1999. Optical dating of single and multiple grains of quartz from
Jinmium rockshelter, northern Australia: Part II, results and
implications. Archaeometry 41, 365­395.
Rommens, T., Verstraeten, G., Bogman, P., Peeters, I., Poesen, J.,
Govers, G., Van Rompaey, A., Lang, A., 2006. Holocene alluvial
sediment storage in a small river catchment in the loess area of
central Belgium. Geomorphology 77, 187­201. doi:10.1016/j.
geomorph.2006.01.028.
Rumsby, B.T., Macklin, M.G., 1996. River response to the last
Neoglacial (the "Little Ice Age") in Northern, Western and Central
Europe. In: Branson, J., Brown, A.G., Gregory, K.G. (Eds.), Global
Continental Changes: the Context of Palaeohydrology. . Special
Publication, vol. 115. Geological Society, London, pp. 217­233.
Schumm, S.A., 1977. The fluvial system. Wiley, New York. 338 pp.
Stam, M.H., 1999. The dating of fluvial deposits with heavy metals,
210
Pb and 137
Cs in the Geul catchment (The Netherlands). Physics
and Chemistry of the Earth. Part B: Hydrology, Oceans and
Atmosphere 24, 155­160.
Stam, M.H., 2002. Effects of land-use and precipitation changes on
floodplain sedimentation in the nineteenth and twentieth centuries
(Geul River, The Netherlands). Special Publications of the
International Association of Sedimentologists, vol. 32, pp. 251­267.
Starkel, L., 2002. Change in the frequency of extreme events as the
indicator of climatic change in the Holocene (in fluvial systems).
Quaternary International 91, 25­32. doi:10.1016/S1040-6182(01)
00099-4.
Starkel, L., 2005. Role of climate and anthropogenic factors accelerating
soil erosion and fluvial activity in Central Europe. Studia Quaternaria
22, 27­33.
Starkel, L., Soja, R., Michczyńska, D.J., 2006. Past hydrological
events reflected in Holocene history of Polish rivers. Catena 66,
24­33. doi:10.1016/j.catena.2005.07.008.
Stuiver, M., Reimer, P.J., 1993. Extended 14
C database and revised
CALIB radiocarbon calibration program. Radiocarbon 35,
215­230.
Swennen, R., Vankeer, I., De Vos, W., 1994. Heavy-metal contamination
in overbank sediments of the Geul river (East Belgium) --
its relation to former Pb­Zn mining activities. Environmental
Geology 24, 12­21.
Thorndycraft, V.R., Benito, G., 2006. The Holocene fluvial chronology
of Spain: evidence from a newly compiled radiocarbon
database. Quaternary Science Reviews 25, 223­234. doi:10.1016/
j.quascirev.2005.07.003.
USDA, 2005. National Soil Survey Handbook, title 430-VI. U.S.
Department of Agriculture, Natural Resources Conservation
Service. http://soils.usda.gov/technical/handbook/.
Van den Berg, M.W., 1996. Fluvial sequences of the Maas, a 10 Ma
record of neotectonics and climate change at various timescales.
PhD Thesis. Wageningen Agricultural University: 181 pp.
Van de Westeringh, W., 1979. Bodemkundig onderzoek in het dal van
de Eyserbeek bij Cartiels. Natuurhistorisch Maandblad 68,
234­239.
Van de Westeringh, W., 1980. Soils and their geology in the Geul
Valley. Mededelingen van de Landbouwhogeschool te Wageningen
80, 1­25.
Vandenberghe, J., 1995. Timescales, climate and river development.
Quaternary Science Reviews 14, 631­638. doi:10.1016/0277-3791
(95)00043-O.
Vandenberghe, J., 2003. Climate forcing of fluvial system development:
an evolution of ideas. Quaternary Science Reviews 22,
2053­2060. doi:10.1016/S0277-3791(03)00213-0.
Vandenberghe, J., Mommersteeg, H., Edelman, D., 1993. Lithogenesis
and geomorphological processes of the Pleistocene deposits at
Maastricht­Belvédre. Mededelingen Rijks Geologische Dienst
47, 7­18.
Veldkamp, A., Van den Berg, M.W., 1993. Three-dimensional
modelling of Quaternary fluvial dynamics in a climo-tectonic
dependent system. A case study of the Maas record (Maastricht,
The Netherlands). Global and Planetary Change 8, 203­218.
doi:10.1016/0921-8181(93)90009-D.
Wallinga, J., 2002. Optically stimulated luminescence dating of fluvial
deposits: a review. Boreas 31, 303­322.
Wolfert, H.P., 2001. Geomorphological change and river rehabilitation:
case studies on lowland fluvial systems in the Netherlands.
Alterra Scientific Contributions, vol. 6. Alterra Green World
Research, Wageningen (the Netherlands). 200 pp.
Wolman, M.G., Leopold, L.B., 1957. River flood plains: some
observations on their formation. U.S. Geological Survey Professional
Paper, vol. 282-C. U.S. Government Printing Office,
Washington, DC, pp. 87­107.
Zolitschka, B., Behre, K.-E., Scheider, J., 2003. Human and climatic
impact on the environment as derived from colluvial, fluvial and
lacustrine archives-examples from the Bronze Age to the Migration
period, Germany. Quaternary Science Reviews 22, 81­100.
doi:10.1016/S0277-3791(02)00182-8.
Zonneveld, J.I.S., 1974. The terraces of the Maas (and the Rhine)
downstream of Maastricht. Centenaire de la société géologique de
Belgique. L'évolution quaternaire des basins fluviaux de la mer du
nord méridionale. Lige, pp. 133­157.
339J.J.W. de Moor et al. / Geomorphology 98 (2008) 316­339