Earth-Science Reviews 174 (2017) 53–83

Contents lists available at ScienceDirect

Earth-Science Reviews
journal homepage: www.elsevier.com/locate/earscirev

Interpreting drivers of change in fluvial archives of the Western MARK

Mediterranean - A critical view
Dominik Faust, Daniel Wolf⁎
Institute of Geography, Technische Universität Dresden, Helmholtzstr. 10, D-01062 Dresden, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Fluvial archives, in particular, late glacial to Holocene floodplain records may provide valuable information
Fluvial behavior regarding past environmental conditions and stages of landscape evolution. In view of the high significance of
Palaeoenvironmental reconstruction floodplain dynamics for the development of entire landscapes, the number of studies that have been performed
Climate forcing on floodplain sediments remains comparatively low, especially regarding the Western Mediterranean region.
Human impact
However, one of the reasons could be seen in the high complexity and diversity of processes and factors that
Terrestrial archives
Spain
control and influence fluvial activity that often hampers a straightforward interpretation of floodplain dynamics.
Tunisia Therefore, a basic demand on fluvial archive research is to address the complexity of the factors that control the
characteristics of fluvial sequences in order to provide a robust basis for their interpretation.
As a starting-point for discussion this paper aims to give an overview of fluvial dynamic patterns in the
Western Mediterranean for the last 15 ka in order to examine their relevance for palaeoenvironmental re-
constructions. The basis for this is provided by previous investigations on four different river systems in Spain
and northern Africa the results of which are herein synthesized in order to propose a regional Late Pleistocene
and Holocene fluvial history for the Western Mediterranean realm. Basic results are related to alluvial floodplain
deposits and visible features within them such as alluvial soils, incision marks and aggradation phenomena.
Since fluvial systems are open systems, we discuss our findings against the background of different influencing
factors that could modify fluvial architecture and may restrict palaeoenvironmental reconstructions. A more
comprehensive interpretation focuses on signals that are common to each of the studied river systems. In this
respect, we critically discuss the customary nature of cumulative probability functions for the identification of
regional flooding episodes and point out the benefits of a stratigraphy-supported approach for characterizing
regional floodplain dynamics. Finally, three alluvial soil formation periods were found in all settings: a first
phase after 15 ka, ending with the Younger-Dryas Event; a second one from 7 to 5 ka with a break between 6.5
and 5.5 ka, and a third phase between 3 and 2 ka. These soil formation periods were interrupted and framed by
fluvial dynamic phases accompanied by channel aggradation, floodplain deposition, floodplain erosion and/or
river incision. In particular, after 5 ka, around 1.6 and at 1 ka, and during the Little Ice Age (LIA) floodplain
aggradation affected river systems in Iberia and northern Africa as well. A cautious assessment of these results
and extensive comparisons with secondary archive information prompts us to relate the aggradation periods
with climate forcing by means of a supra-regional aridification that effected large areas of the Western
Mediterranean. In contrast, the three mentioned soil formation periods can be linked with landscape stability,
most probably triggered by favorable climate conditions in the Western Mediterranean.
Apart from these large-scale patterns we discuss the reliability of information emerging from floodplain re-
cords against the background of individualistic river behavior and self-organization. Regarding future work we
want to emphasize the great potential of yet rarely applied system-oriented approaches that also attach im-
portance to sub-catchment dynamics as a link between catchment slopes and the main river floodplain.

1. Introduction with open systems such as fluvial systems, all of which have unique
configurations of geo-parameters in their catchments and different
In 1991, Stanley A. Schumm published a book, in which he pointed catchment sizes, the interpretation of fluvial archives in terms of pa-
to pitfalls in the way that we interpret the earth. In particular, if we deal laeoenvironmental reconstruction may increase the risk of misleading

⁎
Corresponding author.
E-mail address: daniel_wolf@tu-dresden.de (D. Wolf).

http://dx.doi.org/10.1016/j.earscirev.2017.09.011
Received 19 April 2016; Received in revised form 6 September 2017; Accepted 14 September 2017
Available online 28 September 2017
0012-8252/ © 2017 Elsevier B.V. All rights reserved.
D. Faust, D. Wolf
conclusions. In this respect, key issues for the robust interpretation of
fluvial archives are the problems of scale and place, the problems of
cause and process, and the problems of system response and complexity
(see Brunsden, 2001; Harvey, 2002; Schumm, 1991; Thomas, 2001).
Following these concepts, we need to discuss phenomena of different
catchment sizes, problems of singular occurrences, or phenomena of
convergence and divergence (e.g. Phillips, 2014). On the one hand
convergence means that we see identical features in all river deposits
(e.g. incision in all river systems) that may have several causes and on
the other hand divergence means that we observe different results (e.g.
flooding in one and incision in the other river system) that could be
related to the same causes. This awareness makes analogies almost
impossible (see Phillips, 2007). A key prerequisite to approach these
problems and to attain the basis for any deeper interpretation is to
create an appropriate and reliable data base of fluvial response by
characterizing fluvial architectural patterns based on a large number of
individual profile sections over long distances. This enables single
phenomena (aggradation phases, incision phases etc.) to be placed in
the context of the overall stratigraphy of the catchment and, conse-
quently, to develop a secured and representative standard stratigraphy
(Dusar et al., 2012). As a simple rule of thumb, the more profiles that
are taken, the better the resolution and the more robust the inter-
pretation will be. Beyond a suitable data base, a reasonable inter-
pretation is perhaps the most challenging part when working with
fluvial archives, considering the multitude of factors and their multi-
faceted interrelationships.
With this in mind and based on the suggestions of Schumm (1991)
we once again critically considered the detailed findings on the fluvial
history of three different river systems in the Western Mediterranean,
namely the Jarama River in central Spain (Wolf et al., 2013; Wolf and
Faust, 2015; Wolf and Faust, 2016), the Guadalete River in south-
western Spain (Wolf et al., 2014; Wolf and Faust, 2015), and the
Medjerda River in northern Tunisia (Faust and Zielhofer, 2002; Faust
et al., 2004; Moldenhauer et al., 2008; Zielhofer et al., 2002; Zielhofer
et al., 2004; Zielhofer and Faust, 2008; Zielhofer et al., 2008; Zielhofer
et al., 2009). These results were completed by isolated findings from the
Guadiana River in central southern Spain, the Guadalquivir River in
Andalusia (Wolf and Faust, 2015), as well as findings from the Oued
Moulouya in northern Morocco (Zielhofer et al., 2010, see Fig. 1). Our
objective was to characterise fluvial dynamic features such as channel
incision, river bank erosion, channel aggradation, floodplain sedi-
mentation, and soil formation in all the studied river systems in order to
relate these features to possible influencing factors. Stratigraphic find-
ings were supported by detailed geochemical and granulometric ana-
lyses, clay mineral detection, provenience analyses of the river deposits,
heavy metal detection in order to trace the onset of human activity, and
thin section analyses to detect soil formation features, biospoors and
rootmarks. The chronological frame was confirmed by > 200 radio-
carbon datings and supported by palaeomagnetics for the last 3 ka.
The overall aim of this study is to interpret our findings against the
background of theoretical concepts and problems in order to present a
general and robust history of Holocene fluvial dynamics in Western
Mediterranean river catchments. In the pursuit of this objective, we are
focusing on three key issues that we consider essential for the work on
fluvial sedimentary archives.
(1) a representative reflection of the stratigraphic sequence, together
with an appropriate temporal assessment of sedimentation dy-
namics, paying particular attention to kind, quality, and un-
certainties of the datings, as well as dating positions in a strati-
graphic context;
(2) information must be extracted from the archive that are suitable for
an interregional comparison with other archives. This requires a
careful assessment, whether such a comparison is even permissible,
for example, whether compared phases and processes might be
assigned to similar causations, and whether they are strongly scale
54
Earth-Science Reviews 174 (2017) 53–83
dependant;
(3) the awareness of fluvial systems to be highly complex and non-
linear systems and the associated difficulties in assessing the sys-
tem's response on specific influences of certain factors.
Beyond these considerations issues, the treatment of age data seems
to be a crucial point in fluvial archive research. River deposits in gen-
eral contain a variety of information that can be interpreted from dif-
ferent perspectives and in different ways. Already type and arrange-
ment may provide valuable information regarding sedimentary
environments (e.g. Collinson, 1996; Miall, 1985; Reineck and Singh,
1980), given that the results are placed in an adequate spatial context
(Houben, 2007). However, the interpretation of fluvial features against
the background of climate and environmental changes, the comparison
with independent archive information, or even the comparison of dif-
ferent data sets in order to derive more complex and large-scale inter-
dependencies heavily depend on a responsible handling of the age data.
In this context, it is becoming increasingly common to make use of
meta- analyses on radiocarbon dates from fluvial archives in order to
develop time series of flooding activity, and to transfer them in a vast
regional context. We will demonstrate that there are a number of
challenges to these approaches that are mainly based on the facts that
meta-analyses (e.g. cumulative probability functions) are often ac-
companied by the release of the age data from their original (strati-
graphic) context, and that the data is often unreflectedly used and
correlated across a variety of scales.
In summary, this paper aims to review regional fluvial archive re-
search with a particular focus on our own studies in Western
Mediterranean river catchments. Here, we will indicate some difficul-
ties in data collection, handling, and interpretation, and we will discuss
our stratigraphy-supported approach to derive fluvial sedimentation
phases as well as a step-wise up-scaling approach to synthesize several
fluvial archives against the background of meta-analyses approaches.
2. Geographical frame
The main research work was conducted in three different river
catchments, all of them situated in the Western Mediterranean (Fig. 1),
and all of them characterized by different geo-parameters that make
them unique. However, each chosen catchment has an adequate size to
address the above mentioned issues and all rivers are meandering rivers
in each examined location. Here we follow the principle that climate
driven dynamics can better be separated from human induced geo-
morphodynamics if the chosen scale is big enough to “filter” the human
signal (Faust et al., 2004). This indicates that relatively small areas are
more susceptible to the effect of human activities, or irregular in-
dividual events in general, whether natural or man-made, compared to
larger areas (Fig. 2). Thus, single events may have a greater significance
for fluvial records of small catchments compared to large catchments,
where at least several events are required in order to generate corre-
sponding sedimentation patterns. This is a first principle that is some-
times ignored in case studies dealing with fluvial dynamics (e.g. Faust
et al., 2000) or in comparative studies (e.g. Benito et al., 2015).
The Jarama catchment in central Spain has an extension of about
11.500 km2 reaching up to the high mountain chain of the Sierra de
Guadarrama (2100 m). It is a tectonically active area where fluvial
dynamics are partly influenced by halokinetic movements (events)
originating from the underlying Tertiary marls (Wolf and Faust, 2016).
The present-day average annual precipitation in the Madrid Basin is
about 450 mm. This means, that under present-day semi-arid climate
conditions, regardless of the distinct human influence, the landscape is
particularly vulnerable to periods of higher aridity and increased pre-
cipitation variability that could initiate intense surface erosion, runoff,
and sediment shifting. By contrast, in the semi-humid Guadalete
catchment (see below) in Western Andalucía the mean annual pre-
cipitation is actually about 750 mm that leads to a more stable and less
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Fig. 1. (A): Location of studied river catchments in the Western Mediterranean. The main focus is on the dark grey colored catchment areas (1–3), with 1: Jarama River, 2: Guadalete
River, 3: Medjerda River. The light grey catchments (4 and 5) provide Supplementary information only; 4: Guadalquivir River, 5: Moulouya River. More detailed study areas are indicated
by yellow rectangles. (B–D): study areas in (B) the Medjerda River catchment (Tunisia), (C) the lower Jarama River valley (Spain), and (D) the lower Guadalete River valley (Spain),
together with location of studied profile sections.
Modified after Wolf and Faust (2015) and Faust et al. (2004).

55
D. Faust, D. Wolf
Fig. 2. Conceptual idea relating the significance of erosion and sedimentation dynamics to the size of a catchment area (right side). While catchment slopes and lower order sub-systems
belong to colluvial systems, higher order sub-systems as well as the main river of highest order belong to fluvial systems. Dependant on the respective order of a considered sedimentary
sink, the interpretability regarding different influencing factors as well as certain system characteristics may change (left side).
sensitive landscape. Here, we can expect a dense vegetation cover,
especially during times of non-significant human impact (like the early
Holocene) (e.g. Fletcher et al., 2007). The different hydrological con-
ditions in these two catchments may be able to exert a considerable
influence on the speed of fluvial responses to climate changes
(Brunsden and Thornes, 1979). In the case of wetter initial conditions
(Guadalete), changes towards more arid conditions will be character-
ized by longer fluvial response times, whereas the geomorphic reaction
of this system to changes towards more humid conditions could be even
completely neutralized by higher buffer capacities. In contrast, in case
of arid initial conditions, climate change will cause immediate fluvial
reaction.
With a size of 3.500 km2 the Guadalete catchment is relatively
small, however, in our opinion still big enough to distinguish sedi-
mentation triggers and to evaluate the influence of climate versus
human induced dynamics. In general, the interpretation of sedimenta-
tion features in small catchments with respect to climate forcing is
difficult because in smaller catchments single events such as wildfires
or other natural disasters as well as human impact may totally overprint
the climate signal.
The catchment of the Guadalete reaches back into the Sierra de
Grazalema with mean annual precipitations up to 2.400 mm that dis-
tinctively effects the discharge by increasing the power of river flow. In
the lower reaches of the Guadalete River floodplain, halokinetic activity
linked to Upper Triasic “Gypsum Keuper” marls took place during the
Pleistocene (Rodríguez Vidal et al., 1993) leading to abrupt changes in
the floodplain width that influenced sedimentation conditions respec-
tively.
The Medjerda catchment in northern Tunisia has a size of about
25.000 km2 with main tributaries reaching far back into the Atlas
Mountains of Algeria. Mean annual precipitation in the Ghardimaou
and Jendouba Basin is about 350 mm. The Medjerda Basin is a sub-
sidence area where sediments are well preserved.
In all studied river catchments the surrounding slopes are steep and
all river basins are characterized by strong meandering main river
courses. The width of every alluvial plain depends on the catchment
56
Earth-Science Reviews 174 (2017) 53–83
size and the tectonic framework. The Medjerda alluvial plain is up to
10 km wide, whereas the current meander belt does not exceed a width
of 3 km. The alluvial plain of the Jarama River in central Spain is best
developed within the lower reach before the Jarama River converges
with the Tajo River. Here, the alluvial plain including the meander belt
is about 3 km wide. By comparison, the floodplain of the lower reach
Guadalete River in Andalucía changes its width considerably because of
the doming activities originating from gypsum marls, thus width varies
between 300 m and 2 km.
In summary, we assume that the studied river stretches meet the
comparability requirements in terms of fluvial morphology, fluvial ar-
chitecture and process regimes, which is a fundamental prerequisite in
order to correlate stratigraphic sequences.
3. Approaches and methods
The results shown below are based on extensive fieldwork including
the study of > 58 natural and man-made outcrops in floodplain posi-
tions. For such studies it is an essential precondition to record all clear
sedimentological and pedological features in detail before being able to
verify the recurrence of these features in different exposures. However,
recurrent features may only be detected if a “critical mass” of profiles is
investigated, while at the same time it becomes possible to identify
singularities, and thus to avoid a generalization of isolated results (see
e.g. Dusar et al., 2012). Therefore, as many exposures as possible have
been examined in each catchment for establishing a secured strati-
graphy and deriving representative standard profiles. In certain cases, it
became necessary to designate more than one standard profile for dif-
ferent river sections, to take into account the fact that in large and
diverse drainage areas not all sections of the drainage network are
functioning in phase, that was referred to as the location problem by
Schumm (1991). Apart from the immediate work on sedimentary out-
crops, detailed geochemical and further specific analyses were also
realized in previous studies, however, they are of less relevance for the
study in hand.
The fieldwork on the Spanish river systems was realized during the
D. Faust, D. Wolf
Spanish building boom when a large number of gravel pits were open.
In the Jarama area we focused on the lower river sections where a
particularly high number of often very large gravel pits were in use, so
we had access to 15 gravel pits that were investigated in detail (Wolf
et al., 2013). Our results on the Guadalete River system are based on the
detailed record of 18 gravel pits and 8 drillings up to 7 m in the lower
river sections (Wolf et al., 2014). In the downstream reach of the
Guadalete River that is close to the coastline, fluvial dynamics are quite
varied which requires to treat different river sections separately. In
accordance therewith, a distinction can be made between an upper
downstream section that is characterized by a tendency towards valley
meanders and frequent changes between fluvial deposition and flood-
plain erosion, and a lower downstream section close to the estuary,
with free meander development during the Holocene and more even
sedimentation conditions. The Medjerda River is a meandering river
that flows through the Ghardimaou and Jendouba Basin in northern
Tunisia. In the middle reach of the Medjerda River we realized ex-
tensive studies on 22 selected exposures.
All the selected pits and exposures have been documented and
drawn exactly, and at least one sediment sample was taken from each
layer and soil horizon for the analyses of chemical and physical prop-
erties. In more complex settings more samples were taken to better
trace soil forming processes or fining-up sequences. In some doubtful
cases, thin sections were realized in order to differentiate between in
situ soil formation and soil sediments (Zielhofer et al., 2009). Standard
lab analyses such as the measurement of granulometry, organic content,
calcium carbonate content, total iron (Fet) and pedogenic iron (Fed),
pH-levels and magnetic susceptibility were performed on all samples.
Fig. 3 indicates the methodological steps that were used in all the set-
tings. In certain cases, floodplain sediments have been deposited under
very calm sedimentation conditions that recommended them for pa-
laeomagnetic investigations. Here, samples were taken orientated in
order to precise the sedimentation history of the last 3 ka (Faust et al.,
2004; see Fig. 10). The chronological framework was carried out by
radiocarbon dating of charcoal pieces, botanical macro-remains, pa-
laeo-fire sites and humic horizons. Ceramic remnants, bones, or flint
artifacts complement the radiocarbon dates that have been documented
in previous publications and are shown in Table 1. Ultimately, the
fluvial history of each river has been published already individually (to
know more about the methodological approach see Faust et al., 2004;
Wolf et al., 2013; Wolf et al., 2014; Wolf and Faust, 2015; Zielhofer
57
Earth-Science Reviews 174 (2017) 53–83
et al., 2004; Zielhofer and Faust, 2008).
In Section 5.1 we thoroughly discuss the application of Cumulative
Probability Functions (CPFs) for developing time series of fluvial ag-
gradation and floodplain stability. In order to enhance this discussion
we tested this kind of meta-analysis on the basis of our own data and
constructed a CPF for radiocarbon dates of the Jarama floodplain
(Fig. 12). From a total of 31 age dates, just Holocene dates were se-
lected, we took just one sample from one specific position, and we re-
jected dates in case of obvious age overestimation. Finally 16 age dates
were included to the CPF that was calculated by calibrating the grouped
radiocarbon data using OxCal v4.2.4 (Bronk Ramsey, 2009) and the
Intcal13 calibration curve (Reimer et al., 2013).
For interpreting the results in terms of comprehending the fluvial
history of each river system, and synthesizing the archive information,
we realized a step-wise upscaling approach. The principal objective
here was to evaluate to what extent stratigraphic information as well as
age data that were obtained from single sites (local scale) are relevant
to dynamics on local, regional, but also interregional scales. Therefore,
we tried to filter our data in order to reject sedimentation patterns of
just local significance, and to identify regional significant dynamics for
finally being able to distinguish between regionally and locally oper-
ating influences. The main work steps are summarized in a flow dia-
gram in Fig. 4.
In a first step, following a detailed characterization of all strati-
graphic sequences within one river catchment, we were moving to-
wards a catchment level, trying to separate the most robust features
from the less meaningful for developing a representative composite
profile. In this context, the term robust is assigned with the meaning of
features that are found repeatedly within the investigated profiles. By
doing so, we avoided to over-interpret singularities of only local sig-
nificance. Afterwards, we attempted to interpret the composite profiles
and thus, individual stages of floodplain development over time by
reviewing all available information about the catchment character-
istics, including the possible influence of tectonics, base-level changes,
human activity, climate, and eventually known information on intrinsic
forcing (the latter was usually not available). After a careful assessment
and discussion of the influencing factors and a review of, and correla-
tion with regional climate sensitive archives we were able to summarize
the palaeoenvironmental history of each river catchment. During the
whole process of data collection, data analyses and evaluation of re-
sults, our main working hypotheses regarding actual influencing factors
Fig. 3. Schematic structure of the applied spectrum of methods to
work out the fluvial history of floodplain developments.
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Table 1
Radiocarbon dates obtained from Jarama, Guadalete, Guadalquivir, and Medjerda River systems.

14
Profile Lab no. Material (*samples used for C age yrs. BP Calibrated age BP (2σ) Significance Sediment unit References
Fig. 11)

Jarama River
Aranjuez Beta-293515 Organic sediment 26,150 ± 120 30,852 ± 279 Soil formation JA-2d Wolf et al. (2013)
Aranjuez Beta-293516 Branch 16,140 ± 70 19,193 ± 254 Filling of sinkhole JA-2e Wolf et al. (2013)
Aranjuez KIA 39205 Trunk (Populus sp.)* 4475 ± 30 5132 ± 155 Start sedimentation JA-6a Wolf et al. (2013)
Aranjuez Beta-313498 Wood* 3860 ± 30 4284 ± 126 Stagnant sed. JA-6a Wolf et al. (2013)
Aranjuez Beta-294964 Branch 3750 ± 40 4109 ± 125 – (−) Wolf et al. (2013)
Aranjuez KIA 39206 Ex situ charcoal 3090 ± 30 3300 ± 78 Active sedimentation JA-9 Wolf et al. (2013)
Aranjuez KIA 39204 Charred plant material 2925 ± 25 3089 ± 114 Stagnant sed. JA-8a Wolf et al. (2013)
Aranjuez Beta-294974 Branch 2870 ± 30 3007 ± 128 Stagnant sed. JA-8a Wolf et al. (2013)
Aranjuez KIA 39210 In situ fire site* 2870 ± 25 2982 ± 93 Stagnant sed. JA-8b Wolf et al. (2013)
Aranjuez KIA 39209 Ex situ plant remain 2845 ± 25 2966 ± 95 Stagnant sed. JA-8a Wolf et al. (2013)
Aranjuez Beta-301970 Ex situ charcoal* 2730 ± 30 2838 ± 77 Active sedimentation JA-9 Wolf et al. (2013)
Aranjuez Erl-15157 Ex situ charcoal* 2094 ± 41 2120 ± 171 Active sedimentation JA-10 Wolf et al. (2013)
Aranjuez Beta-301971 Ex situ charcoal* 1610 ± 30 1484 ± 72 Active sedimentation JA-10 Wolf et al. (2013)
Aranjuez KIA 39211 Ex situ charcoal* 405 ± 25 423 ± 90 Active sedimentation JA-12 Wolf et al. (2013)
Cemex Beta-300818 Organic sediment 33,290 ± 220 37,972 ± 768 Stagnant sed. JA-2 Wolf et al. (2013)
Holcim Erl-15163 Ex situ charcoal > 53,500 Stagnant sed. JA-1a Wolf et al. (2013)
Holcim Erl-15164 Ex situ charcoal 39,762 ± 1053 43,789 ± 1502 Start soil formation JA-1a Wolf et al. (2013)
Holcim Erl-15162 Ex situ charcoal 13,654 ± 236 16,493 ± 898 Active sedimentation JA-3b Wolf et al. (2013)
Holcim Erl-15161 Ex situ charcoal 10,599 ± 45 12,537 ± 113 Stagnant sed. JA-4b Wolf et al. (2013)
Holcim Beta-301972 Ex situ charcoal 4460 ± 40 5093 ± 200 Colluvial layer (−) Wolf et al. (2013)
Peralta Erl-15167 Ex situ charcoal* 2044 ± 39 2009 ± 109 Active sedimentation JA-10c Wolf et al. (2013)
Peralta KIA 39218 Ex situ charcoal 2035 ± 30 2004 ± 104 Active sedimentation JA-10c Wolf et al. (2013)
Peralta Erl-15166 Ex situ charcoal* 1983 ± 39 1932 ± 104 channel aggradation (−) Wolf et al. (2013)
Peralta Beta-313499 Ex situ charcoal* 1960 ± 30 1908 ± 79 Active sedimentation JA-10a Wolf et al. (2013)
Peralta Beta-300822 Bone collagen 1810 ± 30 1726 ± 97 Channel aggradation (−) Wolf et al. (2013)
San Martin Erl-15169 Ex situ charcoal* 386 ± 37 413 ± 96 Active sedimentation JA-13 Wolf et al. (2013)
San Martin Erl-15168 Ex situ charcoal* 307 ± 38 385 ± 89 Start sedimentation JA-13 Wolf et al. (2013)
Seseña Erl-15159 Ex situ charcoal* 6595 ± 38 7498 ± 67 End sedimentation JA-5b Wolf et al. (2013)
Seseña Erl-15160 Ex situ charcoal* 3103 ± 33 3317 ± 75 Start soil formation JA-7c Wolf et al. (2013)
Titulcia Beta-300823 In situ fire site* 4490 ± 90 5152 ± 288 Soil formation JA-5 Wolf et al. (2013)
Titulcia Beta-300824 Ex situ charcoal* 1690 ± 30 1612 ± 81 Cultural layer JA-10 Wolf et al. (2013)

Guadalete River
Braza-II Beta-293518 Trunk (Fraxinus sp.) 4050 ± 40 4609 ± 188 Start sedimentation GE-9a Wolf et al. (2014)
Braza-I Erl-15179 In situ macrofossil 3936 ± 36 4381 ± 133 Stability GE-9a Wolf et al. (2014)
Braza-III KIA 39216 In situ macrofossil 3840 ± 30 4279 ± 128 Stability GE-9a Wolf et al. (2014)
Braza-II Beta-293517 In situ macrofossil 320 ± 40 391 ± 89 Channel aggradation GE-13a Wolf et al. (2014)
Braza-I Erl-15178 Ex situ charcoal 311 ± 39 387 ± 89 Active sedimentation GE-13b Wolf et al. (2014)
Braza-III KIA 39215 Ex situ charcoal 170 ± 25 143 ± 143 Start sedimentation GE-14 Wolf et al. (2014)
Braza-I Erl-15180 Ex situ charcoal 122 ± 38 137 ± 137 Active sedimentation GE-14 Wolf et al. (2014)
El Torno KIA 39191 Ex situ charcoal 1740 ± 25 1639 ± 70 Channel aggradation (−) Wolf et al. (2014)
El Torno KIA 39194 Ex situ charcoal 1655 ± 35 1554 ± 136 Channel aggradation (−) Wolf et al. (2014)
El Torno KIA 39193 Ex situ charcoal 1140 ± 65 1082 ± 152 Channel aggradation (−) Wolf et al. (2014)
El Torno KIA 39195 Ex situ charcoal 1075 ± 25 993 ± 61 Active sedimentation GE-11b Wolf et al. (2014)
MEDINA-3 Beta-294976 Organic sediment 7190 ± 40 8046 ± 108 End soil formation GE-5b Wolf et al. (2014)
INA-3 Beta-294975 Shell (mollusk) 7020 ± 40 7851 ± 93 Start sedimentation GE-6 Wolf et al. (2014)
INA-1 Beta-321543 Shell 5230 ± 30 6045 ± 129 Start sedimentation GE-8 Wolf et al. (2014)
INA-3 Beta-300819 Shell 4850 ± 30 5568 ± 83 Start soil formation GE-8 Wolf et al. (2014)
INA-1 Beta-300821 Shell 4460 ± 30 5128 ± 158 Start soil formation GE-8 Wolf et al. (2014)
INA-1 Beta-321544 Shell 3520 ± 30 3788 ± 87 Start soil formation GE-9c Wolf et al. (2014)
José Antonio KIA 39207 In situ macrofossil 911 ± 28 830 ± 86 Stability GE-11a Wolf et al. (2014)
José Antonio Erl-15176 Ex situ charcoal 542 ± 40 576 ± 66 Start sedimentation GE-12 Wolf et al. (2014)
José Antonio Erl-15175 Ex situ charcoal 510 ± 35 564 ± 61 Active sedimentation GE-15 Wolf et al. (2014)
José Antonio Erl-15174 Ex situ charcoal 348 ± 39 402 ± 90 Active sedimentation GE-13b Wolf et al. (2014)
La Barca Erl-15156 Ex situ charcoal 343 ± 37 398 ± 89 Active sedimentation GE-13b Wolf et al. (2014)
La Barca Erl-15158 Ex situ charcoal 342 ± 32 396 ± 86 Active sedimentation GE-13b Wolf et al. (2014)
Pozo Romano Erl-15182 Ex situ charcoal 11,869 ± 178 13,718 ± 385 Active channel filling GE-3 Wolf et al. (2014)
Pozo Romano Erl-15181 In situ fire ground 2234 ± 35 2245 ± 93 Start colluviation GE-10 Wolf et al. (2014)
Pozo Romano Beta-294966 In situ macrofossil 150 ± 30 141 ± 141 – GE-1 Wolf et al. (2014)
Rancho Romero KIA 39213 Organic sediment 5436 ± 41 6216 ± 90 – GE-9c Wolf et al. (2014)
Rancho Romero KIA 39214 Ex situ charcoal 3975 ± 30 4414 ± 111 Active sedimentation GE-9b Wolf et al. (2014)
Rancho Romero KIA 39212 Ex situ charcoal 3915 ± 30 4335 ± 87 Active sedimentation GE-9b Wolf et al. (2014)
Rancho Romero KIA 39213 Humic acid 3843 ± 29 4279 ± 126 – GE-9c Wolf et al. (2014)
Rancho Romero Erl-15177 Ex situ charcoal 936 ± 38 848 ± 81 Start channel filling GE-11b Wolf et al. (2014)
Spinola KIA 23951 In situ fire site 8205 ± 48 9159 ± 143 Soil formation GE-5b Wolf et al. (2014)
Spinola KIA 23950 Ex situ charcoal 2020 ± 25 1969 ± 72 Active sedimentation GE-10 Wolf et al. (2014)

Guadalquivir River
Alcolea Erl-15170 Ex situ charcoal 1812 ± 38 1742 ± 119 Start sedimentation GQ-g Wolf and Faust (2015)
Alcolea Erl-15171 Ex situ charcoal 4175 ± 63 4687 ± 157 Start soil formation GQ-f Wolf and Faust (2015)
Alcolea Erl-15172 Ex situ charcoal 4366 ± 34 4946 ± 92 Active sedimentation GQ-e Wolf and Faust (2015)
Posadas Erl-15173 Ex situ charcoal 11,665 ± 93 13,531 ± 210 Active channel filling GQ-b Wolf and Faust (2015)

(continued on next page)

58
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Table 1 (continued)

14
Profile Lab no. Material (*samples used for C age yrs. BP Calibrated age BP (2σ) Significance Sediment unit References
Fig. 11)

Medjerda River
GH Beta-135720 In situ fire site 440 ± 25 497 ± 32 Stability ME-5 Faust et al. (2004)
GH Beta-135722 Charcoal 2680 ± 46 2800 ± 60 End sedimentation ME-3 Faust et al. (2004)
G KIA 15295 Charcoal 1910 ± 25 1835 ± 95 Active sedimentation ME-4 Faust et al. (2004)
G KIA 15295 Humic acid 1935 ± 30 1886 ± 34 Active sedimentation ME-4 Faust et al. (2004)
G Beta-135723 Charcoal 2940 ± 45 3105 ± 155 Stability ME-3 Faust et al. (2004)
G KIA 15296 Humic acid 3740 ± 70 4115 ± 235 Active sedimentation ME-3 Faust et al. (2004)
G KIA 15296 Charcoal 5280 ± 35 6055 ± 125 – ME-3 Faust et al. (2004)
G KIA 15297 Charcoal 6540 ± 40 7450 ± 120 – ME-3 Faust et al. (2004)
G KIA 15298 Charcoal 4195 ± 36 4750 ± 120 Channel filling ME-2 Faust et al. (2004)
G KIA 15299 Charcoal 6384 ± 83 7245 ± 215 Stability ME-2 Faust et al. (2004)
OM KIA 15307 Charcoal 3510 ± 40 3765 ± 125 Active sedimentation ME-3 Faust et al. (2004)
OM KIA 15308 Charcoal 4055 ± 35 4610 ± 200 Start sedimentation ME-3 Faust et al. (2004)
OM KIA 15309 Charcoal 4074 ± 33 4625 ± 185 Start sedimentation ME-3 Faust et al. (2004)
OM KIA 15304 Humic acid 3580 ± 90 3835 ± 235 – ME-2 Faust et al. (2004)
OM KIA 15306 Charcoal 4219 ± 37 4735 ± 125 End soil formation ME-2 Faust et al. (2004)
OM KIA 15305 Charcoal 4519 ± 38 5175 ± 135 Soil formation ME-2 Faust et al. (2004)
OM KIA 15304 Charcoal 4838 ± 39 5565 ± 95 Soil formation ME-2 Faust et al. (2004)
OM Beta- In situ fire site 11,440 ± 50 13,356 ± 123 Stability ME-1 Zielhofer et al. (2004)
135,718
OY Beta- Charcoal 250 ± 20 235 ± 85 Active sedimentation ME-5 Faust et al. (2004)
135,719
OY Beta- Charcoal 260 ± 20 290 ± 140 Active sedimentation ME-5 Faust et al. (2004)
135,719
OY Beta- Charcoal 150 ± 15 135 ± 145 Active sedimentation ME-5 Faust et al. (2004)
135,719
OY Beta- Charcoal 220 ± 20 155 ± 155 Active sedimentation ME-5 Faust et al. (2004)
135,719
PJ KI5014 Charcoal 2490 ± 30 2550 ± 190 Channel filling ME-4 Faust et al. (2004)
PJ KIA 15286 Humic acid 4600 ± 80 5275 ± 325 Soil formation ME-2 Faust et al. (2004)
PJ KIA 15286 Charcoal 5340 ± 35 6135 ± 145 End sedimentation ME-2 Faust et al. (2004)
PJ KIA 15285 Charcoal 5400 ± 150 6125 ± 375 End sedimentation ME-2 Faust et al. (2004)
SA KIA 15074 In situ fire site 2880 ± 30 3010 ± 130 Stability ME-3 Faust et al. (2004)
SA KIA 15408 In situ fire site 3000 ± 25 3200 ± 130 Stability ME-3 Faust et al. (2004)
BB KIA 15303 Charcoal 3310 ± 25 3555 ± 80 Active sedimentation ME-3 Faust et al. (2004)
SS KIA 15320 Charcoal 3847 ± 28 4280 ± 130 – ME-3 Faust et al. (2004)
SS KIA 15072 Charcoal 3710 ± 30 4035 ± 115 End sedimentation ME-3 Faust et al. (2004)
SS KIA 15073 Charcoal 5325 ± 30 6040 ± 130 End sedimentation ME-2 Faust et al. (2004)

and control mechanisms was repeatedly discussed and tested by using a
falsification approach, in philosophy known as the ‘critical rationalist
approach’ by Popper (1968). Finally, we were able to exclude the less
important factors and to further limit the most significant driving
forces.
In a next step, we compared the results of different regional studies
that are within the same climatic region. In the case of consistent dy-
namics we expected the effect of large-scale influences, and in case of
inconsistency we expected effects of local to regional influences, in-
cluding e.g. tectonics or intrinsic forcing (for a broader discussion we

Fig. 4. Flow diagram that was used as a kind of decision tree in order to deduce the palaeoenvironmental history of each studied river system, and to compare the different river systems
for identifying dynamics of regional and interregional significance. This procedure was implemented very cautiously including a lot of discussion and careful weighing. Therefore,
although this diagram appears to represent a strict algorithm, finally the approach corresponds rather to a careful approximation process.

59
D. Faust, D. Wolf
Fig. 5. Key profiles and stratigraphic correlation of the lower Jarama River. For location see Fig. 1C. Dating are compiled in Table 1.
refer to Section 5.2.4).
In a last step, we expanded the approach to the Western
Mediterranean level, that is our highest scale of consideration and
compared dynamics in Iberia with Tunisia. At the end, our objective
was to propose a general fluvial model for the Western Mediterranean
realm, and to reveal a possible interlinking of large-scale climate fluc-
tuations and supra-regional phases of landscape development, or to
explain deviations and differences of geomorphological responses in the
different regions, respectively (Fig. 4).
4. Main stratigraphic sequences in regional palaeoenvironmental
contexts
For each river system, the stratigraphic description is based on a
large number of single profiles (Figs. 5, 7, and 9) that result in different
types of composite profiles (Figs. 6, 8, and 11). These composite profiles
reflect the often complex fluvial architecture of the investigated
floodplain sediments, including the extent and thickness of sedimentary
layers, the presence of erosion features, and the indication of floodplain
soils. Generally, all rivers permit conclusions regarding past fluvial
processes such as river incision and river bank erosion, channel ag-
gradation, floodplain aggradation linked to flooding periods, but also
soil forming processes. In this respect, an important issue is the dis-
tinction between in situ floodplain soils and redeposited soil material or
soil sediments, respectively. Our findings show that almost all of the
Holocene alluvial fine sediments in the studied sections originated from
60
Earth-Science Reviews 174 (2017) 53–83
soils that have been eroded within the catchments. Thus, for example,
the ratio of pedogenetic iron versus total iron that normally is a good
proxy for in situ soil forming processes, is less significant with regard to
alluvial fines (see also Zielhofer et al., 2009).
4.1. Jarama River (central Spain)
The sedimentation history of the lower Jarama River in central
Spain is illustrated in Fig. 6, while some key profile sections are shown
in Fig. 5. During the Late Pleistocene, between 44 and 17 ka cal. BP,
almost exclusively gravels (Unit JA-2) accumulated at the valley
bottom, followed by sandy material at about 16.5 ka cal. BP (Unit JA-
3). The gravels were affected by halokinetic activity originating from
underlying gypsum marls that led to marked tilting and deformation
structures (for more information see Wolf and Faust, 2016). After Ho-
locene fine sediments were deposited between 8 and 7 ka cal. BP (Unit
JA-5), another gravel body (Unit JA-6) without deformation structures
and showing horizontal bedding started to accumulate after 5.1 ka cal.
BP. From this time on (5.1 ka) the river showed a clear tendency to
aggradation beginning with Unit JA-9 (at about 2.8 ka) subsequent to
the accumulation of gravels of Unit JA-6. Most of the following sedi-
mentation units ended up with soil formation. More precisely, soil
formation phases have been detected following the sedimentation of
Units JA-2d, JA-3, JA-5, JA-7, JA-9, JA-10 and JA-11 (after 30.9, at
16.5–12.5, 7.5–5.1, after 3.3, at 2.8–2.1, 2.1–1.6, and until 0.4 ka cal.
BP, respectively), whereas the last soil formation phase was of
D. Faust, D. Wolf
Fig. 6. Schematic composite profile and respective stages of floodplain development of the Jarama River in central Spain (after Wolf et al., 2013). Sedimentary sequences and classi-
fication of sediment units result from profile sections shown in Fig. 5.
negligible intensity. Since the sedimentary units JA-6 and JA-9 re-
present a singular appearance within the lower Jarama Valley, we tend
to not unduly stress them during further evaluations. Subsequent to
Unit JA-9, exclusively cohesive sediments were deposited in the
floodplain showing higher humic contents and the incorporation of
artifacts, and leading to the assumption that human impact became
increasingly important. The source of these sediments can be seen in
soil erosion on the nearby slopes as indicated by relatively high con-
tents of soluble salts (see Fig. 5 in Wolf et al., 2013). These salts pre-
dominantly originate from evaporitic gypsum marls that change over
abruptly to calcareous marls further upstream. Therefore, we conclude
that short-distance transport outweighed long-distance sediment
transport during these respective time spans. A last notable phase of
fluvial dynamics occurred at about 0.4 ka cal. BP, when the whole
channel-belt was filled with several meters of sandy to silty deposits
that likewise covered more distal floodplain positions in form of a thin
sedimentary sheet. However, during the whole time gravel transport
still remained dominant within the channel itself. Periods of river
61
Earth-Science Reviews 174 (2017) 53–83
incision are difficult to capture from a temporal point of view, but our
findings indicate extensive incision between 16 and 8 as well as 7 and
5 ka cal. BP, albeit especially during the latter period, fluvial erosion
may have been achieved mainly through river bank erosion in the
course of intense meander shifting.
4.2. Guadalete River (southwestern Spain)
In southwestern Andalucía, the floodplain of the lower Guadalete
River is divided into two sections that are the upper downstream sec-
tion and the lower downstream section. Because of location dependent
fluvial conditions, the fluvial behavior varies considerably between the
upper and the lower downstream sections, which becomes particularly
apparent by comparing the composite cross-sections of each zone (see
Fig. 8). In the upper downstream section (Fig. 8A), apart from a Late
Pleistocene sediment succession that is dominated by coarse gravel
(Units GE-1, GE-2, and GE-3), the first Holocene sediments found were
deposited at about 4.5 ka cal. BP (Unit GE-9a and GE-9b) followed by
D. Faust, D. Wolf
Fig. 7. Key profiles and stratigraphic correlation of the lower Guadalete River. For location see Fig. 1D. Dating are compiled in Table 1.
soil formation. In a temporal sense, the fluvial record is strongly frag-
mented in this river section. Early Holocene sediments were pre-
sumably eroded and Mid-Holocene sediments have only been preserved
along short valley openings that are limited because of several nar-
rowings related to gypsum marls that are highly vulnerable to de-
formation and doming processes. Further depositions of cohesive
floodplain sediments have been identified at about 2 (Unit GE-10), 0.9
(Unit GE-11a and GE-11b), and 0.4 ka cal. BP (Unit GE-13a and GE-
13b), as well as until recent times, whereas no evidences of intense soil
formation could be found for the last 2 ka.
It is a different matter, however, when the lower downstream sec-
tion close to the coastline is considered (Fig. 8B). Here we found no
indications of extensive floodplain erosion during the last 8 ka. The
accumulation of fine sediments started already during the Early Holo-
cene, presumably at about 10–11 ka cal. BP (Unit GE-5a and GE-5b),
and each sedimentation period (at 8–7, 6–5.5, 4.5–3.8, and at 0.4 ka
cal. BP) was followed by soil formation. Furthermore, sediment se-
quences are not arranged side by side in form of terraces like in the
upper downstream section, but they are stacked on top of each other.
This is because the site is situated downstream of the terrace crossing
point, which marks the submerging of Pleistocene terraces below the
Holocene floodplain level along the formerly deeply incised estuary
(see Wolf et al., 2014).
A period of intense river incision of about 10 m has been detected
between 13.7 and 11 ka cal. BP. During this time the latest Pleistocene
terrace was partly eroded and the river incised into the underlying
marls while channels on top of the terrace fell abandoned (see channel
in the left part of Fig. 7A). This incision provided the accommodation
space for Holocene floodplain sediments. Another phase of floodplain
dissection by secondary floodplain channels could be dated to about
1 ka.
This is an excellent example to show that material, morphology and
energy flow may vary from place to place (Graf, 1982; Schumm, 1991)
62
Earth-Science Reviews 174 (2017) 53–83
even within the same river valley. This phenomenon is further illu-
strated in Fig. 14 by comparing the sedimentation records of the two
river sections over time. Additionally, this figure shows an attempt that
relates the alluvial history of the Guadalete River with other terrestrial
archives of climate relevance from the surrounding area in order to
depict a possible triggering of fluvial dynamics by specific climate
conditions.
According to this regional archive comparison (Fletcher et al., 2007;
Martín-Puertas et al., 2008; Reed et al., 2001), the majority of Guada-
lete River sedimentation phases seems to be related to periods of cli-
matic aridification. In particular, floodplain aggradation at ~10.5 (Unit
GE-5b), after 8.0 (Unit GE-6), between 4.6 and 4.3 (Unit GE-9b), around
2.0 (Unit GE-10), at 1.0 (Unit GE-11a and GE-11b), and at 0.4 ka cal. BP
(Unit GE-13a and GE-13b) could be linked to phases of increased aridity
that go ahead with highly unstable landscape conditions.
4.3. Medjerda River (northern Tunisia)
In northern Tunisia, the fluvial history of the Medjerda River is well
recorded in sediment sequences of well-stratified cohesive overbank
fines. Despite internal meander migration these overbank fines re-
present a rather stable sediment archive (see also Brown, 1997) due to
good preservation conditions of the cohesive sediments against lateral
erosion (Zielhofer and Faust, 2008). Today sediments are exposed in
undercut banks that cover almost the entire Late Pleistocene and Ho-
locene period. In total, 22 profile sections were investigated in detail
(see selected key sections in Figs. 9 and 10). Sediments were pre-
dominantly interpreted as floodplain deposits with intercalated soils of
various forming intensity. Soil formation indicates a slowdown of flu-
vial sedimentation dynamics under more humid environmental condi-
tions increased with vegetation cover causing stronger sediment re-
tention on the catchment slopes. This, in turn, changes the sediment
load ratio and favors river incision. Fig. 11 summarizes the main
D. Faust, D. Wolf
Fig. 8. (A): Composite profile of the upper Guadalete downstream section showing sediment units and most important 14C dating results. (B): Composite profile of the lower Guadalete
downstream section. Cross sections are based on 18 individual profile sections and coring transects (see Figs. 1 and 7).
Modified after Wolf et al. (2014).
findings (Faust et al., 2004; Zielhofer et al., 2004; Zielhofer and Faust,
2008) and proposes a synthetic profile of the Medjerda River dynamics
including sedimentation rates and fluvial process history.
In the Medjerda Basin, the end of the Pleistocene is represented by a
well-developed rubefied soil that was formed until at least 14 ka and
that indicates thermal conditions that were almost similar to the ones of
today, but related to higher humidity (Yaalon, 1997). For the early
Holocene (11.8–6.7 ka cal. BP) we assume very slow aggradation rates
with almost exclusively silty to clayey sediments being accumulated
(Unit ME-2b, Fig. 9). During this long time span, features of soil for-
mation such as bio spoors and polyedric soil aggregate structures were
formed. Therefore, we expect more or less stable conditions during the
Early Holocene. This assumption is in accordance to the findings of
Ramrath et al. (2000) who described humid and warm conditions over
a longer time span deduced from sediments in the Lago di Mezzano in
central Italy. It seems that these stable conditions could be traced even
in a supra-regional context as Sirocko et al. (1993) report of reduced
dust load in the Arabian Sea from 11.5 to 6.6 that is attributed to more
humid conditions.
From 6.7 ka onwards, fluvial dynamics of the Medjerda River were
intensified by means of increased sedimentation rates, especially be-
tween 6.7 and 6.0 (Unit ME-2c), and between 4.7 and 3.6 ka cal. BP
(Unit ME-3). It is the beginning of enhanced aridification and the end of
the Holocene “green Sahara” phase. This climate change is reported
from lake records in northern Africa (i.a. Callot and Fontugne, 2008;
Fletcher and Zielhofer, 2013; Zielhofer et al., 2017).
The sedimentation record of the Medjerda River for the last 3 ka is
documented in high resolution in Faust et al. (2004). Floodplain sedi-
ments during that time were accumulated as overbank fines (Units ME-
3, ME-4, and ME-5) and sedimentation took place under very calm
conditions (slow sedimentation). This situation enabled the use of
63
Earth-Science Reviews 174 (2017) 53–83
palaeomagnetic methods in order to trace the sedimentation rate
through time. Because of significant similarity with already published
secular variation curves from central Europe and the Mediterranean
(Bucur, 1994; Clark et al., 1988; Evans, 1996; Gallet et al., 2002;
Tanguy et al., 1985) the inclination and declination values from the
Medjerda fine-grained sediments give major chronological information
about the sedimentation history over the last 2.0 ka that is shown in
detail in Fig. 10 and adapted in Fig. 11. This illustrates that sedi-
mentation pulses occurred from 3.0 to 2.8 (Unit ME-3), and from 1.7 to
1.4 ka cal. BP (Unit ME-4). These sedimentation phases were inter-
rupted by a long lasting soil formation period from 2.8 to 1.7 ka cal. BP,
the so called “Roman Humid Optimum” with mild and more humid
conditions and prevailing geomorphic stability (in accordance to the
findings of Mensing et al., 2015). More recent sedimentation pulses
occurred from 1.1 to 0.9 (Unit ME-4), and at about 0.6 (Unit ME-4) and
0.4 ka cal. BP (Unit ME-5).
Finally, Medjerda sedimentation rates were compared with North
Atlantic climate behavior proposed by Bond et al. (2001) and point to
climatic driven sedimentation rates that were influenced by North
Atlantic cooling notably from 4.7 ka on (Faust et al., 2004).
5. Regional fluvial sediment records and interregional
comparison
In general, one main motivation of studies on fluvial records is to
reconstruct regional, but also larger-scale palaeoenvironmental condi-
tions based on the interlocking of palaeoclimate conditions and hy-
drological processes. However, in principle each river system shows
distinctive features according to local contexts including catchment
size, relief, certain climatic properties, and other aspects (e.g. Lane and
Richards, 1997). This means that all systems are ultimately
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Fig. 9. Key profiles and stratigraphic correlation of the Medjerda River. For location see Fig. 1B. Dating are compiled in Table 1.
Modified after Zielhofer et al. (2004) and Faust et al. (2004).

characterized by individualities at both local and regional levels, which discuss our stratigraphy-supported approach to derive and compare
corresponds to the meaning of ‘singularities’ as referred to by Schumm periods of fluvial activity and floodplain stability against the back-
(1991), or the ‘individualistic concept of landscape evolution’ (badass ground of meta-analyses approaches that are increasingly applied in
geomorphology) introduced by Phillips (2015). On this basis, it is a fluvial archive research.
major concern to verify whether the fluvial record reflects, first of all,
local to regional dynamic patterns, or whether there may be even in-
5.1. Chronological resolution of fluvial dynamics
terregional dynamic patterns embedded in the local patterns, which in
turn might permit to draw conclusions on large-scale control factors. As
State of the art dating techniques mean that a wide variety of
a critical point, in particular local fluvial dynamics appear to be subject
methods can be used to constrain the age of fluvial sediments, de-
to great complexity including issues such as self-organization (see
pending on the targeted time period and the availability of appropriate
Phillips, 1999) which makes it difficult to build up relations to certain
dating material. A comprehensive review of existing dating methods
framework conditions (Marston, 2010; Murray et al., 2014; Phillips,
including radiocarbon, luminescence, electron spin resonance, 230Th/U
2015). Thus, a main concern of the following sections will be to com-
and terrestrial cosmogenic nuclides dating, together with advantages
prehensibly demonstrate in which way we attempt to distinguish be-
and limitations of each method has recently been published by Rixhon
tween regional and interregional patterns of floodplain dynamics with a
et al. (2017), and further information have been summarized by Stokes
certain focus on establishing links between stratigraphic findings and
and Walling (2005), Fuchs and Owen (2008), or Rittenour (2008).
environmental indications. After an initial consideration of radiocarbon
Hitherto, the most commonly used dating method in fluvial archive
dating with associated sampling strategies and derived age models as a
research is radiocarbon dating as it can be applied to various organic
fundamental basis to establish fluvial sedimentation patterns, we will
material that is preserved within fluvial sequences. Thus, the 14C dating

64
D. Faust, D. Wolf
Fig. 10. Palaeomagnetics and chronostratigraphic correlation of GH, G and OY profiles.
Modified after Faust et al. (2004).
method may provide information on sedimentation ages by using e.g.
charcoal or macro remains, or on periods of soil formation e.g. by using
humic substances from upper soil horizons. An important constraint
that is often mentioned is the risk of reworking of older organic ma-
terial into younger sediments leading to the incorporation of non-con-
temporaneous material (Chiverrell et al., 2011a; Rixhon et al., 2017).
Another problem may be the contamination by more recent carbon, e.g.
in form of plant roots (Howard et al., 2009). Our own experiences with
fluvial archive research in Mediterranean settings revealed that radio-
carbon dating of small charcoal fragments and branches that are highly
susceptible to attrition during transport over longer distances may
provide the most reasonable age control for sedimentation phases, an
observation that was also made by Sirocko et al. (2013) within the
context of filled-in maar basins in Germany. To give an example, in the
Jarama floodplain (Aranjuez profile) a gravel layer was capped by a
greyish, clayey and humic material (Unit JA-8), which is considered to
reflect the termination of multiple channel aggradations with sub-
sequent sporadic river flooding and low energy sedimentation within
shallow depressions of the gravel plain. Out of this thin clayey layer we
sampled a small branch, plant remains, charred plant material frag-
ments, as well as charcoal from an in situ fire site. These samples
yielded 14C ages of 3007 ± 128, 2966 ± 95, 3089 ± 114, and
2982 ± 93 a cal. BP, all referring to the same depositional age within
the scope of the uncertainties, which proves the suitability for 14C
dating. Above JA-8, a 2 m thick sandy loam (Unit JA-9) revealed ages of
2838 ± 77 and 3300 ± 78 cal. BP based on 14C dating of ex situ
65
Earth-Science Reviews 174 (2017) 53–83
charcoal fragments. The younger age indicates a fast deposition that is
in line with sedimentary observations. The older age illustrates the
incorporation of reworked charcoal fragments. This date could be ex-
cluded based on the abundant dating of the clays below, but it generally
underlines the risk of age overestimation that arises from the use of
single dating to prepare chronologies (Howard et al., 2009). Therefore,
we recommend to give attention to the stratigraphic context when the
reliability of a dating is evaluated, and to realize multiple dating, if
possible.
A general problem encountered in the construction of age models
lies in the non-linear dynamics of fluvial systems. From that point of
view, it is often not straightforward to define beginning and end points
of sedimentation phases based on a given number of dating informa-
tion. An increasingly used method to define time periods of fluvial
activity is the compilation of ‘cumulative probability functions’ (CPFs)
(e.g. Benito et al., 2003; Macklin and Lewin, 2003; Dusar et al., 2012;
Zielhofer and Faust, 2008). Such CPFs are produced by stacking prob-
ability distributions for a range of calibrated radiocarbon measure-
ments assuming that either peaks or troughs in cumulative probability
should reflect a prevailing geomorphological regime such as flooding
episodes or floodplain stability (Chiverrell et al., 2011a). There is an
interesting debate in literature on whether these CPFs represent an
appropriate tool to evaluate the chronology of geomorphological events
and phases (e.g. Chiverrell et al., 2011a, 2011b; Jones et al., 2015;
Macklin et al., 2006, 2011). The main points of criticism of CPFs ex-
pressed by Chiverrell et al. (2011a) are (1) that the possibility of
D. Faust, D. Wolf
Fig. 11. Composite profile and average sedimentation rates of the mid-Medjerda floodplain.
Modified after Zielhofer et al. (2008).
incorporated non-contemporaneous materials in radiocarbon samples is
disregarded, (2) that directional lags between radiocarbon ages and the
related geomorphological events (pre-dating or post-dating) exist (see
also discussion on ‘sucking in’ and ‘smearing’ of proxy events in Blaauw,
2012), (3) that each measured radiocarbon age scatters around the true
age, and (4) that the CPF curve is influenced by the calibration curve of
the radiocarbon ages (see also Bleicher, 2013). Macklin et al. (2011), on
the other hand, refer to useful insights into fluvial temporal and re-
gional variety that have been achieved by fluvial archive comparison on
the basis of CPFs. However, we would like to express some concerns
about the increasing use of CPFs for defining regional phases of fluvial
activity. In particular we want to discuss whether this approach ensures
that ‘like is compared with like’, and if the resultant curves can be in-
terpreted without hesitation. In a first place, we see major limitations
particularly where, as in case of the western Mediterranean, just a
limited number of radiocarbon dating have been realized for individual
river systems. One of our concerns on the CPF approach at the catch-
ment scale relates to the non-standardized sampling strategies that are
inherent to fluvial archive research. CPFs are substantially determined
by the radiocarbon ages that are incorporated. For different reasons,
researchers are often forced to make a selection regarding the number
of samples for dating. The criteria for such a selection in terms of
number and position of such samples often vary greatly. Accordingly,
not every phase of fluvial sedimentation dynamics will be representa-
tively covered by numerical dating. As a consequence, a lot of studies
try to chronologically grasp a certain sediment unit, and to interpret
overlying and underlying sediment successions on the basis of the
stratigraphic context. But when such stratigraphic information is lost, as
is the case when compiling CPFs, the age will be detached from the
sedimentary context, and thus becomes hardly interpretable
(Thorndycraft et al., 2012). As an example, we converted 16 activity
66
Earth-Science Reviews 174 (2017) 53–83
dates from the lower Jarama floodplain into a CPF (Fig. 12). Regarding
the sediment units belonging to the mid-Holocene (Units JA-6 and JA-7,
see Fig. 5, Titulcia, Seseña and Aranjuez profiles), the underlying
sampling strategy was to date the beginning and the end of deposition
(see ‘change dates’, Macklin and Lewin, 2003). Without this strati-
graphic information one would probably tend to identify three sepa-
rated episodes of increased flooding activity, but with the stratigraphic
background we determined just one. Furthermore, we attempted to
consider dating uncertainties and to evaluate more realistically sedi-
mentation phases with the help of stratigraphic information. It seems,
that an unreflected use of a CPF may partly lead to incorrect conclu-
sions (Fig. 12).
Even more challenging is the handling of event-related records that
represent an immediate reaction on extreme climatic events. Here, slack
water deposits (SWD) are a prominent example (Baker, 1987) and there
exist a number of studies that deal with SWDs in the western Medi-
terranean (e.g. Benito et al., 2003; Benito et al., 2010; Ortega and
Garzón, 2009; Thorndycraft et al., 2005; Zielhofer et al., 2010). SWDs
provide important information regarding the magnitude, but also the
temporal sequence of extreme flood events that is also highly relevant
for flood prevention and risk assessment. But here in particular, a re-
presentative reflection of the timing of extreme flood events depends on
a high resolution dating of the different event-layers (Thorndycraft
et al., 2012). Without such a comprehensive dating strategy, the se-
lection of individual dating samples may run the risk of being arbitrary,
all the more if the resulting dates are presented in the form of CPFs with
the claim to give a representative image of flood periods. Another ap-
proach that may help to overcome the problem of missing age in-
formation as well as the loss of geomorphologic information related to
the analyses of SWD and other fluvial records is Bayesian age modelling
as suggested by Chiverrell et al. (2011a, 2011b) and Thorndycraft et al.
D. Faust, D. Wolf
Fig. 12. Cumulative probability function (CPF) for radiocarbon dates of the Jarama
River. From the 31 samples shown in Table 1 we selected 16 samples according to the
following criteria: dates belong to the Holocene; just one sample from one specific po-
sition; dates have been rejected in case of obvious age overestimation. Grey and white
dots mark radiocarbon samples with 2-sigma calibrated age range (grey: sedimentation
ages, white: ages that pre- or postdate sedimentation phases). Light grey vertical bars
show potential time periods in which sedimentation may have taken place according to
dating information and dating uncertainties. Black triangles indicate increasing likelihood
that sedimentation begins on the basis of dating information and stratigraphic features.
Horizontal yellow bars and dark grey vertical bars indicate final sedimentation phases
that result from an assessment on the basis of both, dating information and stratigraphic
findings. Question marks indicate uncertain start and end of sedimentation between 8 and
7 ka cal. BP and a possible sedimentation gap at about 1.8 ka cal. BP that cannot be
proved beyond doubt. Note that by consulting stratigraphic information, sedimentation
phases do not necessarily correspond with peaks in the CPF curve. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version of
this article.)
(2012). An advantage of Bayesian age models is that modeled ages are
calculated also for undated units that enables age-depth plots and flood
magnitude-frequency plots based on the number of individual floods,
the available radiocarbon dating information, and the depths in-
formation of each flood unit whether directly dated or not
(Thorndycraft et al., 2012).
However, the problem of adequate age determination for obtaining
chronological resolution of fluvial sedimentation history concerns not
just SWDs but almost all research on fluvial records, but it becomes
particularly evident in the use of CPFs.
Related to floodplain overbank sediments we thus recommend to
use a stratigraphy-supported approach in order to derive age models of
longer sedimentary sequences. In our own research we used strati-
graphic findings and palaeo-pedological results to compensate existing
dating gaps. For example, soil formation intensity may provide in-
dications of exposure times (Faust et al., 2015; Wolf and Faust, 2015;
Zielhofer et al., 2009). Sedimentation structures may show, whether
sedimentation dynamics have been highly or weakly energetic, or
whether they lasted for long or just short time periods. As an example
from the Jarama floodplain, the fine laminated loamy overbank de-
posits of unit JA-7 point to occasional flooding events with just low
sedimentation rates over longer time periods. Contrary, the relatively
coarse and unlayered sands of unit JA-9 indicate a crevasse splay with
high sedimentation rates and an expected short duration of sedi-
mentation processes. Such an integration of stratigraphic information
may, of course, not provide a statistically usable data set of sedi-
mentation ages, but in our opinion these assessments (see e.g. Fig. 12)
are more valuable for the interpretation of regional or local fluvial se-
dimentation history than the depiction of single radiocarbon ages.
5.2. Interregional comparison of fluvial dynamics
An increasingly common meta-analysis approach to identify epi-
sodes of increased flooding activity across regions and even continents
is based on the use of CPFs (e.g. Benito et al., 2008; Benito et al., 2015;
67
Earth-Science Reviews 174 (2017) 53–83
Hoffmann et al., 2008; Macklin et al., 2006; Thorndycraft and Benito,
2006). An outline about the methodological development of this ap-
proach during the last decades is given in Jones et al. (2015). However,
apart from specific advantages offered by the interregional compilation
of radiocarbon data, we see also a number of difficulties in interpreting
these CPF curves. In the following sections, we will discuss these dif-
ficulties and would like to give some suggestions that may improve the
expressiveness and facilitate the interpretation of CPFs, and of inter-
regional comparisons of fluvial dynamics in general.
Three main issues were identified that complicate a straightforward
interpretation of CPFs. These issues are (1) the compilation of regional
chronologies of flooding episodes on the basis of several sub-regions
that show very heterogeneous climate conditions, (2) the combination
of data from different kinds of fluvial records that are probably in-
dicative of different parameters, and (3) the integration of data from
different catchment areas with very different catchment characteristics.
5.2.1. The identification of regional flood clusters based on CPFs
Apart from the constraints addressed in Section 5.1 the combination
of age information on the occurrence of major flood periods on a re-
gional scale may provide important insights into the hydrological
conditions of a larger area. In order to ensure a certain degree of re-
presentativeness of such a regional data set, it should build upon the
assumption that simultaneous flood periods found in different river
systems are potentially assignable to a series of hydrological events
with similar causes. This requires that the areas under consideration are
subject to somewhat similar climatic conditions as well as flood pro-
ducing mechanisms. This prerequisite appears difficult to achieve in the
case of the Iberian Peninsula with an exceptionally high spatial varia-
tion of annual and daily precipitation (e.g. Hidalgo-Muñoz et al., 2011;
Martin-Vide, 2004; Queralt et al., 2009; Vicente-Serrano, 2006; see also
Glaser et al., 2010). An impressive example for a high spatiotemporal
variability of climatic and environmental conditions for the last 2 ka on
the Iberian Peninsula is given by Sánchez-López et al. (2016). Based on
analyses of the sedimentary infill of the Cimera Lake in the Iberian
Central Range and a comparison with other climatic records in Iberia
they were able to identify interactions of the North Atlantic Oscillation
and East Atlantic to be the main climatic driver for frequent changes of
temperature and humidity conditions in Iberia. Moreover, a changing
dominance of N - S and E - W humidity gradients prevented the
emergence of homogeneous climatic conditions over larger parts of the
Iberian Peninsula. This illustrates that the appearance of flooding
phases may vary widely in its spatial expansion, which might have
serious effects even on neighboring river catchments. However, a broad
division concerning hydroclimatic conditions on the Iberian Peninsula
can be made between Atlantic and Mediterranean river basins. This is
based on the orographic configuration of the Peninsula, which is bor-
dered by the Atlantic Ocean to the west and the Mediterranean Sea to
the east, and the resulting impact of different air masses as e.g. de-
scribed by Muñoz-Díaz and Rodrigo (2003). Those are either Atlantic-
influenced air masses that effect the Atlantic river basins, or convective
systems that effect the Mediterranean coast. Furthermore, a number of
studies show that even in the Atlantic-influenced part of Iberia there are
great disparities between different regions regarding mean moisture
availability and runoff generation (e.g. Nadal-Romero et al., 2015;
Queralt et al., 2009). These hydrological differences may have been
further intensified during phases of Holocene climate change. A review
of climate change projections by Giorgi and Lionello (2008) revealed
that in most scenarios of the recent warming and drying of the Medi-
terranean region, precipitation changes over Iberia exhibit a very het-
erogeneous pattern. This may lead to a partial increase of winter pre-
cipitation in the Atlantic river basins of northern Iberia and a strong
gradient of increasing aridity towards the Atlantic river basins in
southern Iberia. Also Sánchez-López et al. (2016) found evidences for
such humidity gradients in the course of past climate changes, whereas
these gradients may even have run in an opposing direction during
D. Faust, D. Wolf
specific Holocene time periods. Taking these aspects into consideration,
it may be possible to draw the conclusion that Holocene sedimentation
dynamics of river systems belonging to different sub-regions of Iberia
should be reflective of spatially highly variable hydrological conditions
and flood dynamics. The spatial distribution of flooding dynamics
across different sub-regions would be of great value for the re-
construction of the spheres of influence for certain atmospheric con-
ditions over the Holocene. However, any spatial information gets lost if
data is summarized across regions/sub-regions with the aim to generate
a representative CPF curve for a large area such as Iberia (e.g. Macklin
et al., 2006; Thorndycraft and Benito, 2006), not taking into con-
sideration that this may be highly problematic in view of the strongly
heterogenic climatic conditions. It is often argued that the combination
of radiocarbon chronologies at a regional scale may help to overcome
the problem of fragmented records of single-sites. However, we would
like to point out that such a procedure brings the risk of aggregating
divergent flood chronologies into one single probability plot suggesting
a sequence of region-wide flooding episodes, which actually were
spatially limited to certain sub-regions or single systems. Moreover, as
shown by Bleicher (2013) the shape of a CPF may simulate a continuous
phase, which in reality may put together individual events from a
variety of different river systems.
In a recent paper by Benito et al. (2015) a strict separation into
Atlantic and Mediterranean river basins has been implemented, but age
information from floodplain sediments as well as SWDs were included.
The integration of both floodplain sediments and SWDs will be another
point of discussion (see Section 5.2.2). As a suggestion to improve
comprehension, we propose to indicate the affiliation of radiocarbon
ages to specific river systems or sub-regions (also within Atlantic river
basins) that would enable the reader to spatially assign the emergence
of flood clusters. In this context, it would also be helpful to provide an
indication to verify which radiocarbon dates have been selected from
larger datasets from other studies and why. This would be a further
improvement in terms of transparency and plausibility of CPF curves.
5.2.2. The comparability between different depositional environments
Radiocarbon dates from fluvial records have been classified ac-
cording to different types of depositional environments with the main
classes being alluvial channel facies, alluvial overbank facies, and slack
water flood deposits (Thorndycraft and Benito, 2006). A comparison
regarding the indications of SWDs and overbank deposits suggests that
SWDs are suitable to reflect single hydrological events, whereas over-
bank deposits result from a combination of hydrological conditions and
sediment delivery (Benito et al., 2008). The latter implies that overbank
deposition may be linked to climate conditions via the formation of
runoff as well as vegetation changes, or alternatively to human influ-
ences by means of land-use practices. Finally, it is concluded that a
temporal overlap of floodplain aggradation and slack water deposition
can be seen as evidence for a dominant climatic impact, while flood-
plain aggradation without simultaneous slack water deposition may
indicate the predominant influence of land-use changes. We applied
this concept to the SWD record of the Tagus River (Benito et al., 2003)
and the floodplain record of the Jarama River (Wolf et al., 2013) that is
the main tributary to the Tagus. The agreement of deposition at 8–7 ka,
~ 1 ka, 0.5–0.4 ka, and during the last 0.3 ka cal. BP (Wolf and Faust,
2015) illustrates that both records complement each other with the
conformities serving as a hint for the impact of climate forcing. But the
lack of SWDs between 7 and 1.2 ka cal. BP shows that SWD records
alone are not always suited to give a complete picture of flooding
phases which may be related either to the prevalence of lower magni-
tude floods that are not recorded in the archive, or alternatively to post-
flood erosion (Benito et al., 2003). To further develop this concept we
raise the question regarding the significance of slack water deposition
occurring without simultaneous overbank deposition in floodplain po-
sitions. It should be possible that as SWDs are deposited during short
extreme precipitation events that they are not accompanied by high
68
Earth-Science Reviews 174 (2017) 53–83
geomorphic efficiency and that are detached from significant climatic
and environmental changes. SWD records indicate the occurrence of
extreme events with highest runoff magnitude, but such events do not
necessarily initiate considerable overbank deposition since for that, also
other general conditions must be met. Flood events leading to overbank
deposition have a high geomorphic efficiency and result from and are
representative of certain environmental conditions within a river
catchment. Accordingly, slack water deposition and overbank deposi-
tion can be, but do not have to be in chronological accord, and are thus
sometimes difficult to compare to each other. In order to provide an
example, we refer to the period between 2.8 and 2.3 ka BP. We con-
ducted research within three river systems belonging to the Atlantic-
influenced part of Spain, thereby the results are representative for river
catchments under semi-humid to semi-arid conditions in that region.
The findings of all three studied river systems indicate intense flood-
plain soil formation and lacking overbank deposition for the considered
time span. Contrary, the CPFs shown in Benito et al. (2008, 2015) in-
dicate a flood cluster for the same period. Apparently, this flood cluster
is not in agreement with information on floodplain sedimentation in
Atlantic river basins. The CPF for Atlantic rivers in Benito et al. (2015)
is based on radiocarbon dating with a share of 86% belonging to SWDs
and 14% belonging to floodplain sediments. In view of this minor im-
portance of overbank deposits in the course of developing the CPF, the
representativeness for sedimentation dynamics in floodplain positions
is questionable. We do not argue against meta-analytical processing of
information on SWDs in general. But we would like to point out that by
mixing together information from SWDs and floodplain sediments,
central issues may become blurred. Ultimately, therefore, the question
is whether the main focus of a study is on the representation of in-
dividual extreme flood events (via SWDs) or whether it is on hydro-
logical and geomorphological changes of a whole river catchment (via
floodplain deposits).
Returning to the interregional comparison by Benito et al. (2015),
the SWD dominated dataset of Iberia is compared with CPFs from other
Mediterranean regions whose chronologies are predominantly based on
floodplain sediments. Keeping in mind that different kinds of hydro-
logical events may be compared to each other, we recommend to in-
terpret such compilations of CPFs with caution.
5.2.3. The interregional comparison of flood clusters based on CPFs
An attempt to compare datasets of radiocarbon dated fluvial units
on a national scale across Europe has been undertaken by Macklin et al.
(2006) based on Probability Difference Curves (PDCs) that constitute a
further development of CPFs by correcting them for the influence of the
calibration curve, or by Benito et al. (2015) based on Cumulative
Probability Density functions (CPDs). Such a correlation is an ambitious
objective regarding the complexity of fluvial dynamics, particularly in
view of the variety of influencing factors and their variability in time
and space (Schumm, 2005). On the other hand, according to size, lo-
cation, altitude, or land-use history river catchments may show very
different reactions to the same influence of whatever kind or nature
(e.g. contingency, Phillips, 2007). Beside the importance of regionally
specific climate conditions (discussed in Section 5.2.1), a significant
influence on the type of fluvial response originates from catchment size
and mean altitude.
The size of a catchment significantly determines the buffering ca-
pacity to the effect of external variables. As a general pattern, in-
creasing catchment size will cause increasing buffering capacity that is
mainly due to the growth in the number of sediment sinks in form of
valley bottoms and floodplain areas (Harvey, 2002; de Vente and
Poesen, 2005; Smith and Wilcock, 2015; Walling and Webb, 1996;
Walling, 1999). A larger number of interconnected sediment sinks is
also considered to relate to an increase of response time between dy-
namics on catchment slopes and the main river floodplain that may
delay the sediment transfer to the lower river courses (Törnqvist, 2007).
In case of a small catchment area that ends up with a second-order
D. Faust, D. Wolf
stream, a particular external influence like an erosive precipitation
event, a specific land-use measure, or any small-scale disturbance may
significantly increase the sediment yield to and initiate heavy sedi-
mentation in the floodplain. Instead, if the same influence affects a
much larger catchment, likewise sediment may be supplied to the
drainage system, but there is a higher probability that it will be de-
posited and stored within the lower-order stream systems without
reaching the higher order streams (Fig. 2). This means that very small
catchments and related floodplain archives are vulnerable to local
disturbances, whose traces may be totally obliterated in large catch-
ments (Segond et al., 2007).
Higher altitude may favor a higher precipitation regime and a more
accentuated runoff regime. Together with a higher relief energy this
may cause stronger fluvial dynamics which would be reflected in large
sedimentation amounts, but also intensive floodplain erosion, particu-
larly in case of torrential streams.
The aforementioned points show that it is difficult to compare
floodplain records from river systems with very different catchment
characteristics due to different levels of sensitivity and conditions of
sediment preservation (Lewin and Macklin, 2003; Trimble, 2010). For
example, the compilation of Macklin et al. (2006) uses data from river
basins of very different catchment sizes. The dataset from Great Britain
belongs to catchment areas of < 1 km2 through to > 1000 km2. A si-
milar diversity of catchment sizes can be found for Spain and Poland,
although there is even a certain amount of dating from areas larger than
10.000 km2. The authors mention that catchments with a size of < 1
km2 are particularly sensitive to the effect of environmental changes,
however, such small catchments seem to be sensitive to influences of all
kinds. Thus, sedimentary events must not necessarily be attributable to
climate forcing or even to distinct flooding phases (local anomalies, see
Walling and Webb, 1996). We finally conclude that, when comparing
small with large river catchments there is a likelihood that sedi-
mentation dynamics and its timing vary widely for a number of reasons.
This makes it difficult to integrate all data into one CPF-curve with the
firm suggestion to provide a regional or national fluvial record that is
representative for climate forcing.
5.2.4. The interregional comparison of fluvial dynamic phases based on a
stratigraphy-supported and catchment-specific approach
As a way of approaching the comparison of fluvial records in an
interregional context we suggest a kind of step-wise up-scaling of fluvial
dynamic patterns (Fig. 4). Ultimately, stratigraphic information as well
as age data are obtained from single sites (local scale). This raises the
crucial question on the representativeness of these findings for a wider
region (e.g. Burt, 2003) since each river system has its own character-
istics and should first be interpreted from a local perspective (Glaser
et al., 2010). In turn, this opens up a new discussion on spatial and
temporal scales (e.g. Schumm and Lichty, 1965), the complexity of
geomorphic systems (e.g. Schumm, 1991), or the significance of ex-
ternal and internal factors as well as self-organizing behavior for the
control of fluvial system dynamics (e.g. Murray et al., 2014; Phillips,
1999; Thomas, 2001; Vandenberghe, 2003). We will touch upon these
issues in a later section in order to keep the description of our approach
simple for the moment.
For research that emphasizes large-scale patterns, Burt (2003) refer
to a kind of filtering of data in order to remove small-scale noises. If we
transfer this image to fluvial archive research, we need an option to
reject sedimentation patterns of just local occurrence in order to iden-
tify dynamics of regional significance. Such a differentiation would
mean to get nearer to possible drivers and to be able to distinguish
between locally and regionally operating influences. The final aim to
describe fluvial system dynamics against the background of several
altering parameters requires a comprehensive understanding of the
system. However, the assessment of impacts emanating from extrinsic
as well as intrinsic controls and affecting fluvial systems via cause and
effect relationships is sometimes heavily criticized and even rejected
69
Earth-Science Reviews 174 (2017) 53–83
(e.g. Veldkamp et al., 2017) with the demand for a more independent
validation of system dynamics and relevant drivers. In general, we
understand this concern even though we dismiss the absolute nature of
this objective. From our perspective, a common procedure of pa-
laeoenvironmental reconstructions means to compare and bring to-
gether records that are indicative of different kinds of influences such as
geomorphological records, vegetation records, and further climate re-
cords (e.g. Wolf and Faust, 2015). This already results in a comparison
of independent parameters. Of course, such comparison should be done
with the utmost care and caution. But finally, working with different
natural archives is the only way to achieve a more realistic idea of
palaeoenvironmental conditions and associated geomorphic system
dynamics. Finally, we agree with the view that an interpretation of a
sedimentary archive that is solely based on theoretical and conceptual
ideas may be of limited reliability. As a further approximation to the
objection mentioned above we suggest an adequate comparison of se-
dimentation patterns from spatially different river systems. This may
shed light on the appearance of local, regional, or even interregional
dynamics. In turn, this should allow drawing conclusions on certain
influencing factors.
First of all, with respect to individual river systems, fluvial dynamics
may be an expression of the interplay between locally, regionally, and
interregionally relevant variables. At this level, the influence of certain
factors such as climate or land-use changes can only be assumed based
on comparison with further archive information. Thereby, the selection
of archives with a close spatial relationship should guarantee maximum
relevance to the fluvial record concerned. In our case studies we were
able to identify the influence of Late Pleistocene sea level changes (Wolf
et al., 2014), traces of human land-use (Faust et al., 2004; Wolf et al.,
2013), or even the effect of so-called gypsum tectonics (Wolf and Faust,
2016) on floodplain dynamics. Although these are more or less isolated
findings, they help to better understand local landscape history. Fur-
thermore, the succession of fluvial sedimentation periods and flood-
plain stability phases matches very well with sub-regional phases of dry
climate conditions, and more humid climate conditions, respectively
(Figs. 13 and 14). It might be argued that such coincidences in time do
not constitute real evidence of actual relations, but they give valuable
hints of the environmental conditions under which certain floodplain
dynamics took place.
In a next step, we compared sedimentation patterns of river systems
from different regions (central Spain and southwestern Spain, Fig. 15)
that both belong to the broader region of the semi-humid to semi-arid
Atlantic-influenced part of Iberia (Wolf and Faust, 2015). We assumed
that sedimentation phases that do not occur simultaneously in different
river systems are reflective for local to regional influences due to tec-
tonics, sea-level changes, human land-use, intrinsic forcing, or even
small-scale climate patterns. A further characterization of such local
systemic relationships is indeed often challenging given the complexity
of sedimentary systems. Phases which overlap in time we consider to be
controlled by large scale influences that are, in particular, climate or
alternatively land-use during the late Holocene. As can be seen from
Fig. 15, interregionally consistent phases of floodplain dynamics are
rarely absolutely identical in all river systems. They rather show a de-
layed onset or end, or a differing duration. We see this as being an
expression of different levels of sensitivity to catchment dynamics and
in particular to slope erosion as well as to flood producing mechanisms
that in turn may be strongly caused by initial levels of moisture avail-
ability in the respective river systems (Harvey, 2007; Knox, 1972;
Langbein and Schumm, 1958; Schumm, 1979; Walling and Kleo, 1979)
(for more information see discussion in Wolf and Faust, 2015).
In a last step, we contrast interregional phases of floodplain dy-
namics that we consider representative for the Atlantic-influenced part
of Iberia (except the strongly humid areas of northwestern and northern
Iberia) with fluvial dynamics in Tunisia (Fig. 16). This is a first attempt
to link findings from comprehensive stratigraphic works on floodplain
records from large river basins in Iberia and Tunisia with the aim to
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Fig. 13. Late Pleistocene and Holocene floodplain dynamics of the lower Jarama River in central Spain in comparison with palaeoecological information of regional relevance. (a): Phases
of fluvial sedimentation (red bars) and floodplain stability with soil formation (green bars) for the last 14 ka. Temporal uncertainties are indicated by question marks. (b): Radiocarbon
dating as basis for the determination of fluvial dynamic phases; shown is the mean of the 2σ cal. BP age. (c): Characterization of the sediments including sediment texture (S…sand, U…
silt, C…clay, L…loam) and sediment color (black: blackish and humic flood loam, brown: brownish sediments, white: greyish laminated sediments). (d): Position of cultural remains like
bricks and ceramics within floodplain sediments. (e): Pollen percentage record of Mesophytes, Xerophytes, Pseudoschizaea indicating lake dessication and anthropogenic indicators from
lake Siles, south-central Spain (Carrión, 2002), (f): Pollen percentage record of Steppe taxa from Sierra de Cebollera, north-central Spain (Gil Garcia et al., 2002). (g): Pollen percentage
record of arboreal trees from Amblés Valley peat bog, central Spain (López-Sáez et al., 2009). (h): Pollen percentage record of Xerophytes and Mesophytes from Villaverde, south-central
Spain (Carrión, 2002). (i): Benthic diatoms record pointing to arid phases and pollen percentage record of Arboreal trees and land use indicators from Somolinos Lake, central Spain
(Currás et al., 2012). (j)–(k): Floodplain dynamics linked to certain climate conditions based on the comparison of the own findings with secondary archive information. (l): Evidences of
certain anthropogenic contribution to floodplain sedimentation based on the character of floodplain sediments (after Wolf et al., 2013, Wolf and Faust, 2015). (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)

reveal fluvial dynamic phases that endure over the Western Medi-
terranean realm.
Compared to probability based meta-analyses, this stepwise up-
scaling represents a kind of generalization process. It enables the fil-
tering of local specifics (noises, in the sense of Romans et al., 2016) and
puts emphasis on sedimentation patterns of interregional relevance,
which in this case are the primary signals of interest. As Romans et al.
(2016) state, one fundamental goal of stratigraphy is to distinguish
between signal and noise, while the authors point out that noises at one
scale may represent signals at another scale. Consistent therewith, such
a stepwise approach allows to switch between scales, and for example,
to juxtapose impacts of climate change on fluvial sedimentation pat-
terns on a regional and interregional basis. On the other hand, if these
steps are waived, all dynamics that take place at local and/or regional
level will be summarized and equally weighted. Moreover, there will be
no possibility to draw conclusions on spatial patterns or to assess the
impact of climatic versus non-climatic factors. Referring to the debate
on whether this kind of meta-analyses involves a loss or a gain of in-
formation (Chiverrell et al., 2011a; Jones et al., 2015), we would like to
point out that (disregarding other problems) CPFs could well supply
valuable information provided that the input datasets were not ren-
dered anonymous (McCarroll, 2015). It is for this reason that
differences and similarities between input data sets become blurred and
flood periods are not spatially relatable anymore.
6. The interpretation of interregional fluvial sedimentation
patterns as indicators of past environmental conditions in the
Western Mediterranean
6.1. Limitations of causal interpretation
According to the above-presented procedure, we determined local
to sub-regional, regional, and finally interregional phases of floodplain
dynamics. Since all results are based on stratigraphic sequences of
floodplain sediments we can relate certain stratigraphic features, ar-
chitectural elements, and sedimentation characteristics to dominant
process regimes (Houben, 2007; Lorz et al., 2008). These process re-
gimes are representative of certain types of river behavior (Bridge,
1993). River behavior in turn should relate to environmental and cli-
mate conditions. In order to facilitate palaeoenvironmental interpreta-
tions, we will solely consider features that are clearly attributable to the
processes of river incision, bank erosion, floodplain aggradation in
terms of overbank sedimentation, or floodplain soil formation. These
main features are compiled in Fig. 16 for Atlantic-influenced Iberian

70
D. Faust, D. Wolf Earth-Science Reviews 174 (2017) 53–83

Fig. 14. Late Pleistocene and Holocene fluvial dynamic patterns of two different sections within the lower reach of the Guadalete River floodplain in Southwestern Spain in comparison
with palaeoecological information of regional relevance. (a)–(b): Phases of fluvial sedimentation (red bars), fluvial incision (black bars) and floodplain stability with soil formation (green
bars) for the last 14 ka. Temporal uncertainties are indicated by question marks. (c): Radiocarbon dating as basis for the determination of fluvial dynamic phases; shown is the mean of the
2σ cal. BP age. (d): Characterization of the sediments including sediment texture (S…sand, U…silt, C…clay, L…loam) and sediment color (black: blackish and humic flood loam, brown:
brownish sediments). (e): Position of cultural remains like brick fragments and ceramics within floodplain sediments. (f): Lake level changes and desiccation events of the Laguna de
Medina, Southwestern Spain (Reed et al., 2001), (g): Pollen percentage record of Xerophytes and Quercus forest from the lower Guadiana Valley, Southern Portugal (Fletcher et al., 2007).
(h): Lake level changes of the Zoñar Lake, Southern Spain (Martin-Puertas et al. 2008). (i)–(j): Floodplain dynamics linked to certain climate conditions based on the comparison of the
own findings with secondary archive information. (k): Evidences of certain anthropogenic contribution to floodplain sedimentation based on the character of floodplain sediments (after
Wolf et al., 2014; Wolf and Faust, 2015). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

rivers and the Medjerda River.
Several circumstances considerably impede an environmental inter-
pretation of stratigraphic findings. For example, features being triggered by
factors other than climate that are basically human impact, tectonics/halo-
kinetics, base-level changes, or system-inherent factors (Vandenberghe et al.,
2010) need to be excluded. Furthermore, not all features can be reliably
linked to certain factors, even in the case of climate forcing (Schumm, 1991;
Törnqvist, 2007). In the following we take a critical look at the features
documented in Fig. 16 and attempt to evaluate the climatic significance of
fluvial sediments.
6.1.1. Incision features
All of the studied river systems reveal incision features that generally
are related to periods of floodplain soil formation (Medjerda system) as
well as bank erosion (Jarama and Guadalete systems) (Figs. 15 and 16). In
most of these cases, we assume environmental forcing in the sense that
stable geomorphic catchment conditions resulted in decreased slope ero-
sion and lower sediment supply to the river. Decreased sediment load and
more constant discharges prevented floodplain sedimentation that in turn
led to floodplain soil formation. A discharge with just low sediment load,
however, has more capacity to incise and to erode river banks (Bull, 1997;
Wolman and Miller, 1960). On the other hand, especially in the case of
large-scale failure of banks during major erosion events, bank-derived
material can account for large proportions of the suspended load (Bull,
1997). But if discharge volumes do not exceed bankfull capacities, most
sediment will be transported not having the possibility to accumulate in
distal floodplain positions. However, beside this scenario, incision is a
feature that may have several causes all leading to the same result. This is
often referred to as equifinality (Chorley, 1962) and therefore difficult to
interpret. For example, causes for river incision may be tectonic uplift (von
Suchodoletz et al., 2016; Wegmann and Pazzaglia, 2002), base level
lowering (Schumm, 1993; Vis et al., 2016), meander cut-offs (Schumm,
1991), decreasing sediment load, or others. But even if climate is con-
sidered to be an important aspect regarding river incision, causal links
between both are based on complex relationships (Vandenberghe, 2003).
Relating to our own results, most of the observed phases of river incision
did not occur synchronously. Furthermore, we found indications of intense
halokinetic activity in all Iberian study areas (Wolf and Faust, 2016) and a
strong influence of sea-level changes especially in the Guadalete system
(Wolf et al., 2014). Taking these aspects into consideration, it is concluded
that incision features in Holocene fluvial archives can hardly be inter-
preted straightforward by means of climate forcing (Törnqvist, 2007).
Thus, the significance of river incision in terms of specific climate condi-
tions during the Holocene is thoroughly limited. But the situation is dif-
ferent when time scales of 105 years or greater are considered, and when
global climate and sea-level fluctuations linked to glacial-interglacial
changes are involved (Vandenberghe, 2002).

71
D. Faust, D. Wolf
6.1.2. Aggradation features
To interpret aggradation features in terms of climate seems likewise
somehow speculative in the first place. One of the main issues to be
discussed is the possibility to distinguish between different influencing
factors that trigger floodplain sedimentation. Regarding climate control
we share the idea of a causal chain starting with increasing aridification
(with higher risk of heavy rain falls) that in turn reduces the vegetation
cover leading to an increase of runoff and slope erosion and to extensive
fluvial sediment transport and floodplain aggradation. Such relations
have been confirmed, for example, by Blum and Törnqvist (2000),
Desprat et al. (2003), Eybergen and Imeson (1989), Faust et al. (2004),
Knox (1972), Langbein and Schumm (1958), Lavee et al. (1998), Ruiz-
Sinoga and Romero Diaz (2010), Törnqvist (2007), Vicente-Serrano
et al. (2006), or Walling and Kleo (1979). Beside the factor of climate,
there are a number of other factors that may cause floodplain sedi-
mentation, such as base-level raising and creation of accommodation
space (Blum and Törnqvist, 2000; Vis et al., 2016; Wright and Marriott,
1993) due to sea-level fluctuations or tectonics (Huerta et al., 2011),
sediment supply as a result of human activities (Bellin et al., 2013;
Kirchner et al., 2015; Notebaert and Verstraeten, 2010), or intrinsic
forcing (Murray et al., 2009; Phillips, 1999). With the following steps
we approached the identification of aggradation features that may have
been influenced by non-climatic factors:
72
Earth-Science Reviews 174 (2017) 53–83
Fig. 15. Summary of late Pleistocene and Holocene flood-
plain dynamics in Atlantic river basins in Spain. (a):
Guadalete River floodplain. (b): Jarama River floodplain.
(c): Guadalquivir River floodplain. (d): Synthesized time
series of floodplain dynamics for the Atlantic-influenced
part of Iberia (in case of stability in all systems, or flood-
plain aggradation in at least two of the systems). Temporal
uncertainties are indicated by question marks.
(1) We considered physical properties of floodplain sediments in order
to identify human impact. This may be indicated by means of in-
creased proportions of organic carbon and clay reflecting a higher
admixture of eroded top soil material, or by the incorporation of
cultural artifacts (see Figs. 13, 14, and 16);
(2) We run a comparison with climate sensitive archives such as pollen
or lake records on catchment scale. In this way, we tried to identify
aggradation periods that were not in line with climatic develop-
ments (see also Wolf and Faust, 2015).
(3) We compared the timing of aggradation features between the
considered river catchments in order to identify local sedimentation
patterns. These local patterns may indicate the effect of intrinsic
phenomena or other locally significant impacts. Thus, we are just
willing to make firm statements on relations between climate con-
ditions and floodplain aggradation if it is suggested by both, re-
gional archive comparison and interregional comparison of river
dynamics. However, since divergent aggradation patterns may also
be triggered by small-scale climate differences, this step of inter-
pretation may be less straight forward.
Following the above mentioned points, we found some striking
differences regarding aggradation between the considered river systems
(see Fig. 16):
D. Faust, D. Wolf
Fig. 16. Comparison between Iberian river systems and the Medjerda River to identify
common fluvial dynamic phases in the Western Mediterranean. Channel signatures in-
dicate incision in different intensities. Upright arrows indicate aggradation periods. Soil
forming periods are marked by hachures, and waved lines symbolize morphodynamic
stability characterized by very slow sedimentation of fines. Bars highlighted in green
indicate supra-regional soil forming conditions, and bars highlighted in orange represent
phases of supra-regional floodplain aggradation. (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
(1) In the Medjerda, the aggradation around 12 ka BP was interpreted
as a fluvial response to the Younger Dryas event, however, in the
Spanish river systems no floodplain aggradation could be observed
during this time. Quite the contrary, we found indications of river
incision in the lower Guadalete and Guadalquivir floodplains that
we link to lower sea-level.
(2) Between 8 and 7 ka BP, aggradation could be observed in the
73
Earth-Science Reviews 174 (2017) 53–83
Spanish river systems, but in the Medjerda indications of stability
have been found.
(3) Between 6.7 and 5.5 ka BP, aggradation took place in the Guadalete
and the Medjerda floodplains with just short overlap, but not in the
Jarama floodplain.
(4) Strong aggradation shortly after 3 ka BP in the Medjerda is ac-
companied by stability in the Guadalete Valley and a brief and
vigorous aggradation at around 2.8 ka BP in the Jarama Valley.
However, based on sedimentary structures and an examination of
numerous other profile sections in the Jarama floodplain, we clas-
sified this aggradation to be a singular occurrence.
(5) Aggradation occurred between 2 and 1 ka BP in almost all ex-
posures in the Medjerda and Jarama floodplains. But in the
Guadalete River aggradation took place in a time before 2 ka BP
and then only stability or even channel incision was observed.
6.1.2.1. Climate variability. It could be argued that all these above
mentioned phases are indicative of non-climatic triggering. But on the
one hand, in view of the large distances between the considered river
catchments, regional variations in climate patterns must be assumed.
On the other hand, if we for example consider the Guadalete River
(Fig. 14) it appears that at about 6 ka BP aggradation began (see above)
without any hints of climate fluctuations. Then again at about 3 ka BP
pollen records indicate a tendency towards aridification (Fletcher et al.,
2007) but without causing any fluvial reaction. That means, when
thinking about climate forcing of fluvial dynamics we are probably
faced with a divergence phenomenon where similar causes (here:
climate change in terms of mean annual precipitation) produce
different fluvial reactions in different areas (Schumm, 1973). In
current times the Madrid Basin (Jarama River) and the Jendouba/
Ghardimaou Basin (Medjerda River) receive about 450 mm mean
annual precipitation whereas mean annual precipitation in the
Guadalete area is about 300 mm higher (~ 750 mm). Assuming that
all catchments under consideration are affected by a shift to more arid
climate conditions including an accentuation of precipitation, the
effects on fluvial dynamics may differ (see Schumm, 1991). Taking
for example the period starting at 1.7 ka BP, aridification events have
been documented for all the three areas (see literature review in Faust
et al., 2004; Wolf et al., 2013; Wolf et al., 2014). In the two river
systems that are influenced by a generally drier climate (Jarama and
Medjerda rivers) the change to more arid conditions caused
aggradation. But the Guadalete system that is principally subject to
more humid conditions was apparently less effected by aridification
keeping floodplain soil formation and channel bed aggradation due to
river meandering ongoing. In general, the more stable the ecologic
conditions are, the more well-developed the soils will be. Well-
developed soils and vegetation cover, in turn, will increase the
resilience of the entire catchment resulting in lower sediment yield
and lower runoff (Brunsden and Thornes, 1979; Jothityangkoon and
Sivapalan, 2001). Thus, while both systems are affected by the same
climate trend, we observe catchment erosion and floodplain
aggradation in the one system (Jarama and Medjerda), and higher
resilience of the landscape associated with floodplain stability in the
other system (Guadalete). A similar conclusion has been provided by
Törnqvist (2007), stating that the impact of climatic aridification may
be opposite in different geographic regions that depends to a large
extent on initial boundary conditions.
6.1.2.2. Human impact. In addition, anthropogenic influence may be
crucial regarding the triggering of sedimentation dynamics. According
to our findings, physical properties of floodplain sediments changed
ubiquitously at about 2.2 ka ago. More precisely, blackish colors, high
contents of clay and organic matter, as well as admixed clay shards and
brick fragments indicate the supply of top soil material from catchment
soils to the floodplains. Whether or not human activities may lead to
floodplain aggradation raises the question about size and dimension of
D. Faust, D. Wolf
the agricultural area, as well as the considered scale. For this purpose
we want to refer once more to the diagram shown in Fig. 2. Taking a
river basin that is subdivided into a number of sub-catchments, the
responsiveness of each sub-catchment to climate fluctuations is more or
less equally high. However, the higher the order within the discharge
network, the longer will be the response time of the fluvial system to
environmental changes with respect to connectivity and interposed
sediment sinks (Harvey, 2002). In turn, the responsiveness to human
impact is highest in the affected lowest order catchments that are
slopes, foot slopes and small-scale sub-catchments. The lower the
percentage of agricultural land in a catchment, the higher is the
probability that eroded soil material is collected in sediment sinks
that are interposed between catchment slopes and the main river
floodplain (Wolf and Faust, 2013). Thus, land-use induced
sedimentation in the downstream section of the main river (highest
order) is more likely if a large percentage of the catchment is used for
agriculture. Specific system thresholds in regard to sediment shifting
due to land-use seem to be strongly related to catchment size.
Furthermore, even in case of intense land-use a certain precipitation
and runoff regime is indispensable in order to mobilize and relocate
sediment to a greater extent and over longer distances (Faust and
Schmidt, 2009; Nadal-Romero et al., 2015; Puigdefabregas et al., 1999).
This becomes particularly apparent during the last 2 ka. During most of
the time, the environment was under considerable pressure due to
human activities, but in all three catchments the brief and partially
vigorous sedimentation phases correlate with periods of climatic ar-
idification (e.g. at about 1 ka, see Figs. 13, 14, and 16), while periods of
moister climate conditions coincide with lower sedimentation rates or
even with floodplain stability and soil formation. It remains open
whether or not sedimentation during the last 2 ka could have come up
without agricultural activity. Our results show that human activities
exceed a certain threshold in Iberia since 2.2 ka (see also Carrión, 2002;
Currás et al., 2012; Gutiérrez-Elorza and Peña-Monné, 1998; Martín-
Puertas et al., 2008; Moreno et al., 2008) and in Tunisia since 3 ka
(Faust et al., 2004). However, we think that against the background of
permanent land-use over time, changes in sedimentation rates could be
related to other factors such as climate fluctuations. Thus, the geo-
morphic effect of climate aridification and human land-use on flood-
plain sedimentation will add up and thus multiply or mutually reinforce
one another.
6.2. What remains? - interregional dynamic phases and further archive
information
6.2.1. Floodplain aggradation (4.6 to 3.8 ka, at 1.6 ka, at 1 ka, and at
0.4 ka cal. BP)
6.2.1.1. Mid-Holocene. Concerning the period from 4.6 to 3.8 ka cal. BP
we found aggradation in all three river catchments that encourages us
to infer statements related to climate. Studies of higher significance in
terms of climate change (see Fletcher and Zielhofer, 2013) such as
pollen archive studies (e.g. Carrión, 2002; Carrión et al., 2010; Fletcher
et al., 2007; Jalut et al., 2010; Jiménez-Moreno et al., 2015; Pérez-Obiol
et al., 2011) or lake records in northeastern and southeastern Spain
(Höbig et al., 2012; Morellón et al., 2009; Navarro Hervás et al., 2014;
Revelles et al., 2015) and northern Algeria (Gallet et al., 2002), but also
studies from fluvio-lacustrine records in Morocco (Limondin-Lozouet
et al., 2013) or tufa deposits in central Spain (Domínguez-Villar et al.,
2012) point to a change towards more arid conditions at about 5 ka BP
(see also Aranbarri et al., 2014; Bellin et al., 2013). For the same time, a
rapid climate change is reported in northern Africa by means of
terrigenious input into the sea indicating the termination of the North
African Humid Period (de Menocal et al., 2000; Shanahan et al., 2015).
It has to be mentioned that a precise statement about the time of
climate change is always problematic if it is solely based on terrestrial
proxy records. That is because changes in terrestrial system dynamics
(such as aeolian activity, pollen transport or fluvial response) in
74
Earth-Science Reviews 174 (2017) 53–83
consequence of climate change are considered to happen with a
certain time delay, which in turn is often difficult to estimate
(Feurdean et al., 2014; López-Merino et al., 2012; Marston, 2010).
This makes it all the more difficult to exclude forcing other than climate
(Murray et al., 2009). However, looking at the time span under
consideration we are operating on the assumption that climate
became drier at 5 ka BP. An explanation based on environmental
forcing could provide that, as a consequence to large-scale
aridification, precipitation anomalies such as rainfall intensity and
variability increased in all studied systems. In response, vegetation
cover decreased and processes of slope erosion intensified leading to
sedimentation within nearby river floodplains. At least in two river
systems we found that fluvial floodplain aggradation started with a
delay of about 0.5 ka (Figs. 15 and 16). This becomes even more clear
when we look at the Guadalete system, where aggradation started
exactly at the time of maximum proportions of xerophytic vegetation,
although the increase of xerophytes apparently started
already > 500 years before (Fig. 14g). To identify such time lags of
geomorphic response to rapid climate changes is a scientific challenge
of high relevance (c.f. Marston, 2010). Also Törnqvist (2007) stated that
changes in fluvial activity often occur in the short time interval that
follows abrupt climate shifts and justified this with a delayed response
of the vegetation that opens a window for increased geomorphic work
(i.e. sediment yield). A possible explanation for the delayed fluvial
response in the Mid-Holocene may be the behavior of Mediterranean
vegetation with regard to climate changes. Certain types of vegetation
tend to withstand drier conditions by acclimatization and long lasting
resistance strategies (Vicente-Serrano et al., 2012) that may temporally
compensate erosion sensitivity.
6.2.1.2. Late Holocene. Regarding the late Holocene, we detected
intense flooding with associated floodplain aggradation in all river
systems during the last millennium, especially at about 1 ka and around
0.4 ka, and with one exception also around 1.7 to 1.5 ka cal. BP. These
aggradation periods may partly be attributed to climatic downturns
such as Oort and Maunder climatic minima that are supposed to be
related to lower sunspot activity. According to Sánchez-López et al.
(2016) the Medieval Warm Period that started at about 1.05 ka BP was
a warm and dry period in the Iberian Peninsula (see also Moreno et al.,
2012; Trouet et al., 2009) that again fits together with our
environmental interpretation of fluvial floodplain aggradation during
generally arid periods. However, the most strongest flooding period
occurred at about 0.4 ka cal. BP and seems to coincide with the
northern hemispheric cooling event of the “Little Ice Age” (LIA)
(Mann et al., 2009). For the Iberian river systems vigorous
aggradations between 0.45 and 0.35 ka cal. BP are confirmed by a
large number of radiocarbon dates, while in the Medjerda system
aggradation between 0.4 and 0.3 ka cal. BP is additionally supported by
a high resolution palaeomagnetic record (Fig. 10) (Faust et al., 2004).
For this period, a high frequency of high magnitude flood events in the
Tajo river system is likewise documented based on SWD records (Benito
et al., 2003; Thorndycraft and Benito, 2006), lake deposits (Moreno
et al., 2008), and historical documents (Bullón, 2011; Uribelarrea et al.,
2003). In contrast to former aggradation periods, the LIA is generally
considered as a cold period with prevailing wet conditions (Fletcher
and Zielhofer, 2013). On the other hand, a reconstruction of
precipitation in southern Spain based on historical documents
(Rodrigo et al., 1999, 2000) shows a frequent change of dry and
more humid phases with a decadal oscillation (see also Camuffo et al.,
2003). Accordingly, after a dry period culminating around 1540 CE,
wetter conditions covered the period from 1590 to 1649 CE. Eight
radiocarbon samples that were taken from the corresponding
sedimentary layer in the Guadalete and Jarama river systems, yielded
sedimentation ages between 385 ± 89 (1476–1654 CE) and
423 ± 90 a cal. BP (1437–1617 CE) (Table 1). A relation between
this sedimentation phase and a large catastrophic event in 1545 CE as
D. Faust, D. Wolf
described in Barriendos and Rodrigo (2006) is a matter of speculation.
From this it follows that floodplain aggradation features linked to the
LIA cannot be simply assigned to the prevalence of wet conditions
without doubts.
6.2.2. Floodplain soil formation (end of the late Pleistocene, 7 to 5 ka, and
2.7 to 2.2 ka cal. BP)
We consider soil formation to be a clear and reliably interpretable
feature that is first of all related to long lasting exposure times (Faust
et al., 2015; Zielhofer et al., 2009). As a reflection of geomorphic sta-
bility, this is in turn linked to stable environmental and climatic con-
ditions. Regarding the climatic conditions that are required for soil
formation during the Holocene, we assume precipitation as being more
relevant for a greater soil formation intensity than temperature is.
While a higher edaphic humidity promotes alteration processes
(Birkeland, 1999; Engel et al., 2016), the temperature in the study areas
was always stable enough to allow soil formation during the considered
time span. From a geomorphological point of view a balanced and
homogenous precipitation favors dense vegetation cover, high infiltra-
tion rates and thus, lower runoff (Jothityangkoon and Sivapalan, 2001;
Lavee et al., 1998) that finally supports floodplain soil formation under
conditions of landscape stability.
During the first phase of simultaneous soil formation in all catch-
ments we detected strong soil formation features such as rubefaction,
clay illuviation, and the complete decalcification of formerly calcareous
alluvial deposits. This soil formation started at about 16 ka BP in the
Jarama system, at 13.7 ka BP at the latest in the Guadalete system, and
sometime around 17 ka BP in the Medjerda system. In all systems soil
formation ended during the Bølling/Allerød, or with the beginning of
the Younger Dryas, respectively. By means of pollen assemblages taken
from archives in central and southern Spain Carrión (2002), Carrión
et al. (2010), Gil Garcia et al. (2002), Jalut et al. (2010), and Pons and
Reille (1988) report on climatic amelioration in terms of an increase of
precipitation and temperature for the same time period. Moreover, a
high resolution pollen record from the Alboran Sea (Combourieu
Nebout et al., 2009) points to prevailing temperate deciduous forest and
“warm mixed forest” during the Bølling/Allerød interstadials that is
indicative of stable environmental conditions (see also Fletcher et al.,
2010). Since all studied river systems show intense soil formation
during this time span, we propose a widespread climatic amelioration
along with landscape stability in the Western Mediterranean realm.
Between 7 and 5 ka soil formation again took place in all our set-
tings, however, with different intensities. For this time span, more
humid conditions are reported for the Western Mediterranean (Carrión,
2002; Carrión et al., 2010; Gasse, 2002; Gil Garcia et al., 2002; Lamb
and van der Kaars, 1995; Morellón et al., 2009; Pons and Reille, 1988).
In the Guadalete and Medjerda floodplains this soil formation period
was interrupted by temporary aggradations beginning around 6.6 ka
and 6.1 ka, respectively. This interruption is characterized by a ground
water drop in northern Tunisia (Faust and Zielhofer, 2002) and de-
position of coarse material in some parts of the floodplain. In the
Guadalete Valley respective sediments were found in the lower down-
stream section, close to the estuary. However, an interpretation of this
short interruption in terms of climate forcing remains unclear so far.
A last phase of soil formation that affected all three systems si-
multaneously was found for a period between 2.7 and 2.2 ka cal. BP.
Martín-Puertas et al. (2008) report high lake levels of the Zoñar Lake in
southern Spain between 2.7 and 2 ka BP assuming more humid condi-
tions. Also López-Sáez et al. (2009) and Currás et al. (2012) deduce
more humid conditions during 2.8 and 2.1 ka BP in central Spain, and
Vis et al. (2010) found just low fluvial activity in the estuary of the
Tagus River. Based on marine pollen records obtained from the Gulf of
Gabes, also Brun and Rouvillois-Brigol (1985) observe a gradual in-
crease of oak pollen around 2.5 ka BP for the Tunisian region that is
interpreted as an indication of humidity. Even if these studies reflect
local climatic conditions they show that in that case climate
75
Earth-Science Reviews 174 (2017) 53–83
information may be transferable into a broader context.
All evidence suggests the importance of higher humidity to promote
floodplain soil formation. It is once again worth highlighting that soil
formation within floodplain deposits is not just a local phenomenon but
indicates morphodynamic stability with markedly reduced sediment
transport in the entire landscape. Sediment shifting will then be limited
to the channel itself. This information is of broader importance, because
from this perspective floodplain dynamics reflect regional and even
supra-regional landscape conditions that in turn may be largely driven
by climate and vegetation cover.
6.3. Wet versus dry - environmental indications of floodplain sedimentation
A multitude of studies attempt to demonstrate that fluvial dynamics
are largely driven by climate conditions. However, there is not even a
common sense about the way in which climate may control
Mediterranean floodplain and catchment dynamics. In a review paper
on the impact of rapid climate changes on Western Mediterranean
landscapes Fletcher and Zielhofer (2013) state that flood producing
mechanisms in Mediterranean fluvial systems remain an open debate
within the scientific community. One concept suggests that decreased
flooding is linked to more humid conditions associated with denser
vegetation cover and lower surface runoff, while arid conditions result
in reduced vegetation cover and increased flooding activity. Another
opinion is that increased flooding activity may be caused by higher
moisture influx that is linked to a higher winter precipitation (see dis-
cussion in Fletcher and Zielhofer, 2013 and references therein). In the
course of our research we obtained a number of findings that may
contribute to this debate. However, we would like to caution the
drawing of premature conclusions since large-scale atmospheric con-
ditions especially during the Early and Mid-Holocene are still far from
being understood, and the response of terrestrial systems on climate
changes may be a lot more complex than is assumed.
A number of studies tend to interpret the occurrence of high mag-
nitude floods in connection with North Atlantic Oscillation (NAO)
modes. It is argued that a negative NAO-index corresponds to humid
conditions over Iberia that are necessary to produce extreme runoff
events (e.g. Benito et al., 2008; Benito et al., 2015; Macklin et al., 2006;
Moreno et al., 2008; Ortega and Garzón, 2009; Thorndycraft and
Benito, 2006; Vaquero, 2004). In this context, high magnitude floods
are assumed to take place during anomalous wet winters linked to
persistent negative NAO phases, while prevailing positive NAO modes
are considered to cause smaller floods (Salgueiro et al., 2013). More-
over Hidalgo-Muñoz et al. (2011) document a strong regional difference
of extreme precipitation trends over southern Iberia in recent times and
conclude that heavy rainfall is not generally related to negative NAO-
index values. Instead they see heavy rainfall in relation to partly highly
negative values of the Western Mediterranean Oscillation that corre-
sponds to a central European anticyclone in the north of Italy and a low-
pressure centre over southwestern Iberia (Martin-Vide and Lopez-
Bustins, 2006). For the last 1000 years Camuffo et al. (2003) propose a
scenario where the Iberian Peninsula was mainly dry without really wet
periods. They assume that instead of wet periods that are characterized
by frequent rainfall, prevailing aridity with a handful of intense pre-
cipitation and flooding events are deemed more probable especially
with a view to the LIA. Therefore, they assume the prevalence of cli-
mate variability and a sequence of severe droughts and phases of in-
tense rainfall due to the passage of Atlantic cyclones. Lastly, one view,
based on observations and conceptual considerations, is that which
relates flood episodes and phases of river activity to climatic change
(e.g. Knox, 1972; Törnqvist, 2007; Vandenberghe, 1995). For example,
Knox (1972) places maximum geomorphic work of fluvial dynamics
(e.g. sediment yield) into a period of climatic change from arid to
humid conditions that is principally based on considerations by
Langbein and Schumm (1958). This may be caused by a fast change of
precipitation regimes, and a delayed response of the vegetation cover
D. Faust, D. Wolf
that transiently overlaps with a still high potential for slope erosion.
There are a number of studies that attempt to reconstruct pre-
cipitation patterns (e.g. Rodrigo et al., 1999, 2000), flood episodes
(Glaser et al., 2010; Uribelarrea et al., 2003), and even the variability of
the NAO (Jones et al., 1997; Luterbacher et al., 2002) based on in-
strumental time series and documentary proxy data for a period of the
last 500 years, or even longer (Luterbacher et al., 2012; Olsen et al.,
2012). According to Muñoz-Díaz and Rodrigo (2003, 2004), in the 20th
century high NAO index values can be related to low annual and sea-
sonal precipitation (drought). Low NAO index values can be related to
intense seasonal rainfall (Muñoz-Díaz and Rodrigo, 2004) and river
flow (Trigo et al., 2004). A further relationship between the NAO and
solar activity is suspected (Vaquero, 2004), but linear dependencies are
subject of discussion (Anet et al., 2014; Brugnara et al., 2013; Georgieva
et al., 2007; Ineson et al., 2011). Queralt et al. (2009) describe a strong
effect of the NAO on precipitation patterns over Iberia during winter.
Thus, during negative NAO phases they see a higher precipitation in-
tensity as well as frequency in central and western/southwestern Spain.
But an increase of extreme events is mainly restricted on western and
northernmost regions. Nadal-Romero et al. (2015) show that based on
measurements taken over the last 20 years, the highest percentage of
runoff, erosion, and sediment yield can be associated with westerly
airflows in the mid-western areas or cyclonic pattern farther inland. An
analysis of responsible weather types revealed that these runoff and
erosion events occur in a very small temporal frame linked to a small
number of atmospheric patterns. However, even if just the Atlantic part
of Iberia is considered, the highest runoff and erosion events do not
belong to the same weather types (Nadal-Romero et al., 2015) that may
be mainly caused by the orographical configuration of the Iberian Pe-
ninsula (Cortesi et al., 2014). Furthermore, there is a number of studies
about historical flood events in Spanish river systems based on doc-
umentary data of the last millennium (e.g. Barriendos and Rodrigo,
2006; Glaser et al., 2010; Uribelarrea et al., 2003). For example,
Barriendos and Rodrigo (2006) give a broad compilation of docu-
mented floods in almost all Iberian river catchments and show a
variability of flood occurrence on a decadal scale. Atlantic river basins
where mainly affected by floods during winter months. Moreover, the
authors found three main flooding periods in the middle and late 16th
century, the end of the 18th century and the second half of the 19th
century that fits with prolonged periods of above average amount of
rainfall as reconstructed by Rodrigo et al. (2000).
In summary, there seem to be several indications of certain con-
formity between a negative winter NAO, and more frequent precipita-
tion and flood occurrence in the majority of Atlantic river basins in
Iberia for the last century. However, a fundamental question is whether
these documented flood episodes do correspond to phases of floodplain
aggradation that are indicated by geomorphological floodplain records,
and whether, on that basis, it is possible at all to reconstruct Holocene
aggradation periods and atmospheric pattern beyond the timeframe of
historical documentations?
Our results indicate that these questions cannot be answered beyond
doubt. Above all, there is a serious dating problem related to youngest
phases of floodplain dynamics. Radiocarbon ages covering the last
centuries are characterized by considerable uncertainties that generally
prevents exact chronological classifications. In this context, there is just
one strong sedimentation period that falls within the range of instru-
mental and documentary time-series. This phase of exceptional sedi-
mentation is documented in the Jarama, Guadalete, and Medjerda
systems at about 400 a cal. BP. In Iberia, this aggradation phase falls
within a period of frequent hydrological fluctuations. Bearing the
dating uncertainties in mind, this phase may be related to changing
climate with moderate flooding that followed a prolonged dry period
(Wanner et al., 2011), but eventually predating a phase of maximum
flooding intensities (Barriendos and Rodrigo, 2006; Rodrigo et al.,
2000). Beyond that, we found no correlation between flooding episodes
based on documentary data and aggradation periods based on
76
Earth-Science Reviews 174 (2017) 53–83
floodplain records. Recalling once again the timing of floodplain ag-
gradation, our results suggest interregional aggradation between 4.6
and 3.8 ka, at 1.6, at 1 ka, and at 0.4 ka cal. BP, while aggradation in
Atlantic river basins in Iberia took place from 8 to 7 ka, 5 to 3.8 ka, 2.2
to 1.9 ka, 1.8 to 1.5 ka, at 1.0 ka, at 0.4 ka cal. BP, and during the last
200 years (without temporal reliability for this last phase). From this
we conclude that phases of floodplain aggradation reveal a centennial
to millennial-scale variability that is about one magnitude higher than
expected fluctuations of the NAO. As a general pattern, the NAO ex-
hibits a variability at decadal to multi-decadal time scale (Hurrell,
1995). However, there are some studies suggesting that certain phases
with a concentration of positive or negative NAO modes lasted from
several decades up to more than one century (Luterbacher et al., 2002).
For example, Trouet et al. (2009) relate the warm and dry period of the
Medieval Climate Anomaly with persistent positive NAO signals. On the
other hand, numerous studies revealed indications of longer-term cli-
mate fluctuations during the Holocene. Mayewski et al. (2004) de-
monstrate the presence of rapid climate changes (RCC) at intervals of
approximately 2.8–2.0 and 1.5 ka with a global footprint all over the
Holocene. Also Wanner et al. (2011) show that on a multi-decadal to
multi-centennial scale, periods of more stable and warmer climate were
interrupted by cold spells at 8.2, 6.3, 4.7, 2.7, 1.55 and 0.55 ka BP in
the northern hemisphere. Relating to the origin of these fluctuations
they attach major importance to melt water fluxes into the North
Atlantic, low solar activity, explosive volcanic eruptions, and fluctua-
tions of the thermohaline circulation (Wanner et al., 2011). With re-
spect to timing and duration of floodplain sedimentation periods, we
suggest that Holocene floodplain dynamics in the Western Mediterra-
nean may be mainly controlled by such low-frequency factors (e.g.
RCCs), rather than by decadal-scale NAO modes. However, in view of
the LIA, it seems that floodplain dynamics are not predominantly linked
to, but partly influenced by prevailing NAO modes.
Apart from this scaling aspect, it is still an important question,
whether centennial to millennial-scale floodplain dynamics in the
Mediterranean can be characterized by dry or wet climatic/environ-
mental conditions. Based on our results and a comparison with further
climate and environmental archive information we developed the fol-
lowing conclusions. Holocene phases of floodplain aggradation in
Atlantic River Basins in Spain are generally linked to (1) a decline of
mesophytic vegetation, (2) an increase of xerophytic vegetation, (3)
lake desiccation phases by tendency, and (4) North Atlantic coolings
(Figs. 13 and 14, Wolf and Faust, 2015). That means that these phases
do not go hand in hand with pronounced wet periods. In this context,
we want to introduce a scenario, where floodplain dynamics may be
controlled by overlapping drivers that operate on different temporal
scales. Based on previous considerations, we assume a centennial to
millennial-scale shifting between prevailing warm or cold, and humid
or dry environmental conditions. The variability of this shifting corre-
sponds largely to the variability between periods of floodplain ag-
gradation and floodplain stability, respectively. However, for the last
millennium, the importance of drivers showing decadal-scale varia-
bility has been demonstrated with respect to temperature, droughts,
and precipitation/runoff patterns (e.g. Hurrell, 1995; Trigo et al.,
2004). From this we conclude that low-frequency environmental fluc-
tuations fundamentally determine whether geomorphic work takes
place in a landscape or not. However, depending on these general en-
vironmental conditions, high-frequency fluctuations may on the other
hand show either a reinforcing effect on the geomorphic work (e.g.
expressed by strong floodplain aggradation), or may be completely
buffered in case of a stable system.
This point can also be considered on the basis of precipitation
variability. Given a certain amount of evenly distributed rainfall, at
least during the rainy season, we may assume that the corresponding
vegetation, infiltration, and runoff characteristics caused lowest values
of soil erosion, decreasing flooding activity, and thus the systems lack
floodplain aggradations under stable catchment conditions (as observed
D. Faust, D. Wolf
at the end of the Late Pleistocene, from 7 to 5 ka, and from 2.7 to 2.2 ka
cal. BP, see chap. 6.2.2). If now the concept of the NAO (as re-
presentative for the strength and paths of North Atlantic storm) is ap-
plied, a negative NAO mode may increase the frequency and intensity
of rainfall events, but the erosive character of these events may be
buffered by more resilient catchment slopes and intense subsurface
runoff (based on the concept of climatic variation of sediment yield by
Langbein and Schumm, 1958). With a slight delay such events may lead
to high magnitude flood events, albeit the geomorphic efficiency of
which should be strongly reduced. If now precipitation will be reduced,
or the temporal distribution of precipitation will be modified leading to
higher rainfall variability, this may result in increased soil erosion and a
higher probability of extreme flood events due to disturbances of the
vegetation cover (Desprat et al., 2003; Eybergen and Imeson, 1989;
Ruiz-Sinoga and Romero Diaz, 2010; Vicente-Serrano et al., 2006) and
increasing runoff production (Blum and Törnqvist, 2000; Lavee et al.,
1998; Walling and Kleo, 1979) (as observed from 4.6 to 3.8 ka, at
1.6 ka, at 1 ka, and at 0.4 ka cal. BP, see chap. 6.2.1). In such a case,
both negative and positive NAO modes may cause floodplain sedi-
mentation since almost every serious rainfall event may initiate slope
erosion and floodplain sedimentation. But a higher frequency of heavy
rainfall may considerably increase sedimentation dynamics. Looking at
it the other way round, a higher frequency of extreme flood events is
not inevitably accompanied by wetter conditions. This may likewise be
caused by a shift in the rainfall distribution in terms of decreasing
rainfall frequency and increasing rainfall intensity that is a character-
istic of landscape aridity (Bullón, 2011; Lavee et al., 1998). As Camuffo
et al. (2003) note, periods are often defined as wet or dry because of
documented anomalies like extreme events but these events do not
reflect the everyday rainfall patterns of the whole period considered. In
areas like Spain, monthly and yearly amounts of rainfall are determined
by highly different daily amounts of rain, and already one high daily
amount may be decisive when classifying any given month, season or
year as dry or wet (rainy) (Martin-Vide, 2004). Furthermore, if periods
of floodplain sedimentation in the Western Mediterranean were asso-
ciated with wet periods in terms of persistent rainfall, various concepts
regarding climate sensitivity of slope dynamics, including buffer capa-
city and landscape connectivity, or concepts on morphodynamic ac-
tivity (accompanied by flooding, sediment accumulation) and mor-
phodynamic stability (accompanied by soil formation, river incision)
(Rohdenburg, 1989; Rose et al., 1999) could not help to explain land-
scape dynamics. Thus, we suggest to clearly differentiate between
flooding periods occurring under generally arid climate conditions that
may cause a high geomorphic effect, and flooding under generally
humid climate conditions that may produce peak runoff but may not
cause considerable floodplain aggradation. In both scenarios, geo-
morphic efficiency that strongly affects landscapes and ecosystems
differs significantly.
In summary, we have identified centennial to millennial scale
floodplain sedimentation in the Western Mediterranean that may be a
response to low-frequency climate forcing. The decadal variability of
the NAO that was reconstructed for the time period of historical doc-
umentation, is not reflected in floodplain dynamics apart from one
phase of brief and vigorous sedimentation at about 0.4 ka cal. BP.
Because of dating uncertainties it cannot be stated whether this sedi-
mentation took place during a negative NAO mode with highly vari-
able/intense rainfall and flooding patterns, or whether it took place
during a short period with less rainfall but even higher rainfall varia-
bility in the course of a progressive change from the warm and dry MCA
(Moreno et al., 2012; Sánchez-López et al., 2016) towards the LIA.
Regarding atmospheric circulation patterns in the time prior to the LIA,
it is still uncertain in which way climatic aridity and related floodplain
sedimentation with centennial to millennial-scale variability have been
controlled, and whether it can be related to persistent positive NAO
modes as suggested by Olsen et al. (2012).
77
Earth-Science Reviews 174 (2017) 53–83
7. About the reliability of fluvial system information
Compared to the overall research on terrestrial archives for the
purpose of palaeoenvironmental reconstructions, studies on fluvial
dynamics are still relatively rare, especially in the Western
Mediterranean. One reason could be that many researchers have doubts
on whether fluvial records really have the explanatory power that is
attributed to them. This is demonstrated by ongoing debates on the
actual triggers of fluvial dynamics, and particularly on the role climate
actually plays in the build-up of floodplain records. Our intention is to
show that, in fact, there are a number of imponderabilities, but if we are
aware of the limitations involved in the interpretation of floodplain
dynamics, floodplain deposits may express a multitude of information
on river behavior and palaeoenvironmental conditions. At the begin-
ning we emphasized the book by Schumm (1991) who pointed out a
series of problems in the interpretation of fluvial systems related to
scale and place, cause and process, as well as system response (time,
space, location, convergence, divergence, singularity, sensitivity, com-
plexity, efficiency, multiplicity, see Schumm, 1991) in order to raise
awareness of the complexity of natural systems and the necessity to
develop rational scientific approaches. All these considerations refer to
the principle that on different spatial and temporal scales geomorpho-
logical systems move within the frame of natural laws, but nevertheless
will create variable process structures and phenomena due to the
complex interplay between a number of parameters (see also Lane and
Richards, 1997). From this it follows that certain aspects of landscape
evolution are difficult to understand and assess in detail, which also
implies that the prediction of future developments is only possible to a
limited extent. As shown by Phillips (2015) these ideas were continued
and singularities were described to be linked to non-linear complexity
and historical contingency, with contingency referring to geo-
graphically and temporally contingent factors that make landscapes
unique. Phillips (2015) emphasizes the individualistic behavior of
geomorphic systems, and introduces the concept of ‘badass geomor-
phology’ that is based on individualism, together with non-conformity
of geomorphic systems with conventional wisdom and the existence of
amplifiers of climate change effects. This perspective is undeniably a
reasonable counter pole to the constant search for universal rules in the
context of geomorphic system dynamics. This is not to say that certain
regularities do not exist, but the awareness that sometimes landscape
evolution does not comply with conceptual frameworks of geomor-
phology may prevent from jumping to conclusions about causal re-
lationships between controlling factors and geomorphic system beha-
vior. Therefore, particularly in regard to fluvial archives, an open-
ended approach in research that takes into account all possibilities and
uncertainties of system control is indispensable for delivering solid re-
sults that remain well comprehensible. A very recent paper by Veld-
kamp et al. (2017) notes, that based on numerical modelling studies,
autogenic system behavior may cause very individualistic fluvial record
properties that, in the end, lead to a situation that does not allow any
statements on system controls at all. However, we do not agree with
this perspective. Moreover, we want to challenge the conclusion that it
is not possible to identify relations between floodplain dynamics and
external controls based on a potential equifinality that arises from nu-
merical modelling. To speak with the words of Phillips (1999): “The
determination of the self-organization properties of a landscape should
be a starting point rather than a goal of geographic explanation.” For
example, the confrontation with concepts of singularity or in-
dividualistic landscape evolution tends to create new impulses in re-
search approaches and scientific viewpoints that should be considered
constructive and stimulating for geomorphic archive research. A com-
parable stimulus can and should arise from numerical modelling ap-
proaches. There exist already important results from modelling, al-
lowing us to assess the effect of intrinsic forcing on fluvial systems. For
example, Van De Wiel and Coulthard (2010) give an example of self-
organized criticality in bedload sediment output based on simulations
D. Faust, D. Wolf
with the computational landscape evolution model Caesar. But this
brings up the question whether such examples should be treated as an
indication that external drivers are not necessary for the formation of
sedimentary records, or that intrinsic forcing can be one factor among
others that complicates the identification of causative effects resulting
from external drivers. There has been major progress regarding the
refinement of models or the incorporation of different external factors
(Coulthard and Van De Wiel, 2012; Notebaert et al., 2011). However, in
many cases simplifications must be made, making it partly difficult to
consider models as a reflection of processes that can be found in nature,
e.g. when soil erosion is simulated irrespective of a vegetation cover or
processes of interrill erosion (Coulthard and Van de Wiel, 2013).
Especially, vegetation cover and the event character of highly variable
rainfall/runoff in the Mediterranean are key issues to interpret en-
vironmental floodplain records, and it would be crucial to include these
aspects in the modelling process. Without detracting from the im-
portance of numerical modelling, which we consider a key competence
for the review of actual cause-and-effect relationships in future re-
search, we think it is an improper way to negate, out of hand, a
dominant effect of external factors. This is particularly the case since
there is an infinite number of case studies that provide evidences for
effects of external drivers on geomorphic systems. Besides, it is hardly
conceivable (and certainly not the aim) that models will provide a
faithful representation of the complex reality, which ultimately would
be necessary for providing definitive answers about the impacts of a
mixture of factors. Lane and Richards (1997) state that without an
appropriate contextualization models will not comply with the crucial
importance of configuration (e.g. morphological conditioning) that is
indispensable for the determination of river system behavior. Thus, the
agreement between model predictions and reality will remain uncertain
(Lane and Richards, 1997). At this point, we have come full circle back
to the issues of singularity and individualistic behavior of geomorphic
systems that constitutes a major challenge to the idea of modelling
landscape evolution.
But what remains of the significance of floodplain dynamics? There
is an interesting review on the ‘Fluvial record of climate change’ by
Macklin et al. (2012) debating the climatic significance of fluvial re-
cords (as well as the meaningfulness of numerical models) that we are
not going to repeat here. Referring to our own findings, we have shown
that there are numerous features of floodplain dynamics, such as river
incision or aggradation in certain cases, that cannot be linked
straightforward to specific causes since involved interrelationships may
be manifold and not referable to a sole influence of climate, humans,
tectonics, or intrinsic forcing. But we are confident that the approach of
extensive archive comparisons between fluvial systems on different
scales, but also between fluvial systems and climate archives may in-
itially lead to the detection of local, regional, and also interregional
patterns on the basis of temporal coincidence. Moreover, as soon as
some interregional patterns are identified that are embedded in locally
specific patterns, it appears that there is a large-scale control of flood-
plain dynamics. From our viewpoint, such large-scale patterns may
serve as a kind of exclusion criterion for the relevance of intrinsic for-
cing. But the question remains whether or not a similar fluvial response
in spatially separated areas is actually caused by the same parameter
constellation. In that case, the comparison with archives that are sen-
sitive to external influences such as pollen or lake records may shed
light on interdependencies. A further approximation with the identifi-
cation of system controls may be achieved through considering a scale
dependent influence of parameters. When we evaluate our field-based
findings, we observe conformity but also high variability on (site-spe-
cific) profile-scale and we see variability, but also conformity on
catchment-scale as well as on interregional scale. On a site-specific
scale, floodplain dynamics are influenced by manifold factors leading to
partly variable time periods of sedimentation and sedimentation rates.
We assume that on this scale intrinsic behavior may show particularly
great effects, thus interpretation is usually limited to identifying river
78
Earth-Science Reviews 174 (2017) 53–83
behavior e.g. based on architectural elements. On catchment-scale,
generally sedimentation patterns become obvious that allow to derive
stages of floodplain evolution. Here fluvial dynamics are still de-
termined by numerous factors, but it should be possible to assess the
relevance of regionally effective controls, such as sea-level fluctuations,
large-scale tectonics, or intensive human land-use in more recent times.
The actual relevance of these factors emerges in the interregional
comparison given that one factor affects one river catchment but not
another one. For instance, Pleistocene sea-level lowering plays a pivotal
role for river incision in continental margins (e.g. Guadalete River),
whereas continental interiors (e.g. Jarama River) might be more or less
unaffected (upstream vs. downstream controls, see Blum and Törnqvist,
2000). As already mentioned, congruent Holocene sedimentation pat-
terns on an interregional scale prompt us to rule out a dominant impact
of sea-level changes, tectonics (as in case if Iberia), and intrinsic system
behavior. Thus, at the end, climate conditions and land-use changes
turn out to be the main triggers of large-scale floodplain dynamics.
Repeatedly indications have been specified that permit a climatic in-
terpretation of fluvial dynamics (e.g. Fletcher and Zielhofer, 2013;
Macklin et al., 2012). For the Western Mediterranean there exist dif-
ferent views and perspectives that attribute a dominant influence either
to climate (e.g. Bellin et al., 2013) or to land-use (e.g. Wainwright and
Thornes, 2004), respectively. In central Europe, as an area where pre-
cipitation patterns are more predictable and much effort has gone into
the reconstruction of land-use history, attempts have been made to
model catchment-scale sedimentation dynamics influenced by climate
and land-use changes (Notebaert et al., 2011). Based on different sce-
narios and with the limitation of certain assumptions, Notebaert et al.
(2011) were able to demonstrate that a dominant influence on geo-
morphodynamics can be attributed to land-use changes. However, apart
from such modelling approaches we have not heard, so far, of any field-
based approach to reliably differentiate between the impacts of climate
and land-use on floodplain dynamics. Such a differentiation may be a
main challenge for future research.
A further point of criticism against causal explanations of floodplain
records is the spatio-temporal delay of erosion and sedimentation
events along a river profile (e.g. Veldkamp et al., 2017). In Section 6.1.2
we discussed the issue of response times, but we related delayed reac-
tions mainly to slope-vegetation-erosion dynamics, rather than to
temporary floodplain storage. Our assumption is based on two ob-
servations. First, overbank deposits in the lower Jarama Valley contain
certain percentages of gypsum that indicates a pronounced connectivity
between floodplain and slopes of the immediate catchment area (Wolf
et al., 2013). That is why we exclude pure long-distance transport that
is a main mechanism in the course of response delay and temporary
floodplain storage. In general, we expect a strong connectivity between
catchment and main stream in the semi-arid Mediterranean according
to commonly strong intensities of rainfall and runoff events especially
during climate anomalies (Faust and Schmidt, 2009; Imeson and Lavee,
1998; Nadal-Romero et al., 2015). Individual events with just selective
effects may be buffered by the catchment, but sub-regional to regional
climatic events should show more extensive effects (Fig. 2). The level of
connectivity, however, is subject of debates and intensive research ac-
tivities (e.g. Fryirs, 2013; Hoffmann, 2015; Notebaert and Verstraeten,
2010). Therefore, in order to evaluate the actual level of connectivity,
detailed and systematic research in the entire catchments will be ne-
cessary. As a consequence, for future research there is a high demand
for studying spatio-temporal floodplain dynamics of large river systems.
But at least equally important is the investigation of tributary stream
networks that finally transfer geomorphic reactions of catchment slopes
to the main river floodplain. In this context, also Notebaert and
Verstraeten (2010) underline the need for system-oriented approaches
that attach importance to erosion and colluviation, sediment supply to
the discharge network, and floodplain deposition within streams of
different orders.
D. Faust, D. Wolf
8. Future perspective
With regard to debates and concerns that are frequently mentioned
in the literature, one could conclude that some obstacles have devel-
oped over time that hamper progress in the characterization of fluvial
records. On the one hand, palaeoenvironmental reconstructions that
rely on field-based research on sedimentary archives are frequently
accused of being an urgent search for causative relations. It is often
argued that in these kind of studies certain interrelations and de-
pendencies should be discussed more controversially. e.g. with regard
to elusive control factors. In fact, it is extremely difficult to detect and
quantify amplifier effects, intrinsic thresholds, or non-linear dynamics
in general. On the other hand, research based on numerical modelling
but also laboratory experimentation is often confronted with the re-
proach to not correspond to reality or to not be sensitive to indirect
controls that are not quantifiable (e.g. contingency). Furthermore,
models are strongly dependent on the nature and the quality of input
parameters that is often a limiting factor especially for times long past.
An advantageous option for future developments would be a greater
convergence and combination of both field-based and modelling ap-
proaches, even if necessary methodological approaches may currently
be still a kind of intangible. Considering the points made in Section 7,
one may comprehend fluvial dynamics of highest order streams as a
product of innumerable individualities all over the catchment, which
constitutes a thankless starting point for model-based facsimile. One
possible option includes to draw more abstract conclusions with a focus
on the importance of dominant factors and features that is endeavored
by the study in hand. Furthermore, Murray et al. (2014) propose to
starting models directly on a relatively large scale that could provide a
more reliable way to understand and predict the way landscapes in
general evolve and react to critical changes in external forcing. Another
option would be to go into more detail opening the possibility to resolve
relevant subprocesses and to comprehend singular appearances. For
example, Lane and Richards (1997) derive the importance of shorter
time-scales and smaller space-scale processes in controlling longer-term
system behavior. Based on a number of examples also Coulthard and
Van de Wiel (2013) discuss the question as to what level of process
detail, spatial and temporal resolution is required to simulate landscape
dynamics, while referring to the problem of lacking input data for long-
term landscape modelling.
A number of arguments indicate that more system-oriented ap-
proaches including a stronger focus on subsystems and tributary
catchments could drive fluvial archive research forward. Main river
floodplains represent the terminal link of a chain of sub-catchments.
Thus, an environmental interpretation is considerably more challenging
if main river floodplains are considered separately from the dynamics of
the sometimes highly divers sub-catchments. Moreover, an integrated
approach would allow for statements to be made about landscape
connectivity, temporary floodplain storage, or even the differentiation
between climate and land-use changes as controls of floodplain dy-
namics.
9. Conclusion
Fluvial archive research is first of all an attempt to reconstruct river
behavior and floodplain development under certain framework condi-
tions by means of fluvial architecture, sediment characteristics, and
recurring features of palaeoenvironmental significance. The great
challenge is to associate sediment structures with system conditions in
terms of the main triggering/controlling factors. Fluvial archive studies
enable us to deduce process combinations that are responsible for the
formation of certain sedimentary or fluvial architectural patterns.
However, direct causes that led to this formation remain uncertain and
therefore a critical evaluation can only be achieved by using the ex-
clusion principle. For this purpose a wide range of information is
needed to exclude less important factors. Independent climate
79
Earth-Science Reviews 174 (2017) 53–83
information can be obtained from climate-relevant archives such as
pollen records, speleothems, lake records or comparable sediment re-
cords that enable the use of biomarker, mollusk, or isotopic analyses,
etc. The evaluation of these secondary archives shows that the closer
the archives are situated to the respective river catchment, the better
are the possibilities to compare information and to depict relationships
(Wolf et al., 2013). This points to the fact that each river catchment
possesses an individual sedimentation history and that fluvial dynamics
are related to regional or even local environmental conditions that are
constituted, among others, by climate, human impact, and environ-
mental resilience. Accordingly, the examination of just one river
catchment does not necessarily serve to work out conditions that are
valid over large distances. In the study in hand we tried to face this
problem by identifying singularities and deviating patterns in order to
assess supra-regional dynamic patterns that may be related to supra-
regional climate forcing. However, after extensive data collections and
comprehensive evaluations only a modest number of clear statements
can be proposed. First of all we assume that soil formation in all settings
from 15 to 13 ka, 7 to 5 ka, and between 3 and 2 ka cal. BP is a clear
signal of landscape stability in the Western Mediterranean realm trig-
gered by favorable climate conditions that are characterized by ba-
lanced and homogeneous rainfall.
Aggradation as a result of climate forcing is not entirely un-
ambiguous. However, as all studied settings show increased floodplain
sedimentation after 5 ka BP, we assume climate forcing to be the main
trigger. Therefore, a supra-regional climatic deterioration characterized
by an aridification trend affected the Western Mediterranean during the
mid-Holocene. Our results show a last strong sedimentation phase in all
settings that coincides with a period of the LIA. Here again, we interpret
this feature to be climate driven in the entire region.
Finally, we are convinced that this kind of detailed studies is useful
and indispensable, not only for palaeoenvironmental reconstruction or
even palaeoclimate research, but also for an understanding of fluvial
dynamic responses to changes in catchment conditions. The promptness
of a landscape or catchment response to changes in vegetation cover,
land use dynamics, or climate changes is strongly correlated to catch-
ment size and the respective state of resilience. River behavior is
therefore dependent on landscape conditions that are mirrored in
floodplain architecture, fluvial sediment characteristics, and the ap-
pearance of floodplain soils. Sensu stricto, we reconstruct fluvial be-
havior and system response under certain framework conditions (cli-
mate and others). This could be a key-competence for future scenarios
and might be able to offer an outlook on expected reactions of river
systems in catchments that are faced with changes in future. Thus,
considering a better combination of field-based and modelling ap-
proaches in future, fluvial archive research should be able to predict
fluvial dynamics and geomorphic changes in long terms provided that
forecast models of landscape and climate change are reliable. This in
turn can be essential for regional planning purposes.
Acknowledgements
We gratefully acknowledge the German Research Foundation (DFG)
for financing these projects (FA 239/2-1 and FA 239/2-2; FA 239/14-1
and FA 239/14-2). Special thanks to Christoph Zielhofer for fruitful
discussions and for providing the original illustrations. We also wish to
thank two anonymous reviewers for their constructive comments that
considerably improved this manuscript.
References
Anet, J.G., Muthers, S., Rozanov, E.V., et al., 2014. Impact of solar versus volcanic activity
variations on tropospheric temperatures and precipitation during the Dalton
Minimum. Clim. Past 10, 921–938.
Aranbarri, J., González-Sampériz, P., Valero-Garcés, B., Moreno, A., Gil-Romera, G.,
Sevilla-Callejo, M., García-Prieto, E., Di Rita, F., Pilar Mata, M., Morellón, M., Magri,
D., Rodríguez-Lázaro, J., Carrión, J.S., 2014. Rapid climatic changes and resilient
D. Faust, D. Wolf
vegetation during the Lateglacial and Holocene in a continental region of south-
western Europe. Glob. Planet. Chang. 114, 50–65.
Baker, V.R., 1987. Paleoflood hydrology and extreme flood events. J. Hydrol. 96, 79–99.
Barriendos, M., Rodrigo, F.S., 2006. Study of historical flood events on Spanish rivers
using documentary data. Hydrol. Sci. J. 51, 765–783.
Bellin, N., Vanacker, V., De Baets, S., 2013. Anthropogenic and climatic impact on
Holocene sediment dynamics in SE Spain: a review. Quat. Int. 308-309, 112–129.
Benito, G., Sopeña, A., Sánchez-Moya, Y., Machado, M.J., Pérez-González, A., 2003.
Palaeoflood record of the Tagus River (Central Spain) during the Late Pleistocene and
Holocene. Quat. Sci. Rev. 22, 1737–1756.
Benito, G., Thorndycraft, V., Rico, M., Sánchez-Moya, Y., Sopeña, A., 2008. Palaeoflood
and floodplain records from Spain: evidence for long-term climate variability and
environmental changes. Geomorphology 101, 68–77.
Benito, G., Rico, M., Sánchez-Moya, Y., Sopeña, A., Thorndycraft, V.R., Barriendos, M.,
2010. The impact of late Holocene climatic variability and land use change on the
flood hydrology of the Guadalentín River, southeast Spain. Glob. Planet. Chang. 70,
53–63.
Benito, G., Macklin, M.G., Zielhofer, C., Jones, A.F., Machado, M.J., 2015. Holocene
flooding and climate change in the Mediterranean. Catena 130, 13–33.
Birkeland, P.W., 1999. Soils and Geomorphology. Oxford University Press (430 pp.).
Blaauw, M., 2012. Out of tune: the dangers of aligning proxy archives. Quat. Sci. Rev. 36,
38–49.
Bleicher, N., 2013. Summed radiocarbon probability density functions cannot prove solar
forcing of Central European lake-level changes. The Holocene 23, 755–765.
Blum, M.D., Törnqvist, T.E., 2000. Fluvial responses to climate and sea-level change: a
review and look forward. Sedimentology 47, 2–48.
Bond, G.C., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W., Hoffmann, S.,
Lotti-Bond, R., Haidas, I., Bonani, G., 2001. Persistant solar influence on north
Atlantic climate during the Holocene. Science 294, 2130–2136.
Bridge, J.S., 1993. Description and interpretation of fluvial deposits: a critical perpective.
Sedimentology 40, 801–810.
Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51,
337–360.
Brown, A.G., 1997. Alluvial Geoarchaeology: Floodplain Archaeology and Environmental
Change. Cambridge Manuals in Archaeology. Cambridge University Press (377 pp.).
Brugnara, Y., Bronnimann, S., Luterbacher, J., Rozanov, E., 2013. Influence of the sunspot
cycle on the Northern Hemisphere wintertime circulation from long upper-air data
sets. Atmos. Chem. Phys. 13, 6275–6288.
Brun, A., Rouvillois-Brigol, M., 1985. Apport de la Palnologie a l'Histoire du peuplement
en Tunisie. (ONRS) Palynol. Archéol. II. pp. 215–226.
Brunsden, D., 2001. A critical assessment of the sensitivity concept in geomorphology.
Catena 42, 99–123.
Brunsden, D., Thornes, J.B., 1979. Landscape sensitivity and change. Trans. Inst. Br.
Geogr. 4 (4), 463–484.
Bucur, I., 1994. The direction of the terrestrial magnetic field in France during the last 21
centuries – recent progress. Phys. Earth Planet. Inter. 87, 95–109.
Bull, L.J., 1997. Magnitude and variation in the contribution of bank erosion to the
suspended sediment load of the River Severn, UK. Earth Surf. Process. Landf. 22,
1109–1123.
Bullón, T., 2011. Relationships between precipitation and floods in the fluvial basins of
Central Spain based on documentary sources from the end of the 16th century. Nat.
Hazards Earth Syst. Sci. 11, 2215–2225.
Burt, T., 2003. Scale: resolution, analysis and synthesis in physical geography. In: Clifford,
N.J., Holloway, S.L., Stephen, P.R., Valentine, G. (Eds.), Key Concepts in Geography.
SAGE Publication, pp. 199–216.
Callot, Y., Fontugne, M., 2008. Les sites lacustres d'âge holocène dans l'est du Grand Erg
occidental (nord-ouest du Sahara algérien): interpretation géomorphologique et pa-
leoclimatique. Géomorphol. Relief Process. Environ. 3, 187–200.
Camuffo, D., Sturaro, G., Benito, G., 2003. An opposite flood pattern teleconnection be-
tween the Tagus (Iberian Peninsula) and Tiber (Italy) rivers during the last
1000 years. In: Thorndycraft, V.R., Benito, G., Barriendos, M., Llasat, M.C. (Eds.),
Palaeofloods and Climatic Variability: Applications in Flood Risk Management. CSIC-
CCM, Madrid, pp. 295–300.
Carrión, J.S., 2002. Patterns and processes of Late Quaternary environmental change in a
montane region of southwestern Europe. Quat. Sci. Rev. 21 (18–19), 2047–2066.
Carrión, J.S., Fernández, S., Jiménez-Moreno, G., Fauquette, S., Gil-Romera, G., González-
Sampériz, P., Finlayson, C., 2010. The historical origins of aridity and vegetation
degradation in southeastern Spain. J. Arid Environ. 74, 731–736.
Chiverrell, R.C., Thorndycraft, V.R., Hoffmann, T.O., 2011a. Cumulative probability
functions and their role in evaluating the chronology of geomorphological events
during the Holocene. J. Quat. Sci. 26, 76–85.
Chiverrell, R.C., Thorndycraft, V.R., Hoffmann, T.O., 2011b. Reply to comment: cumu-
lative probability functions and their role in evaluating the chronology of geomor-
phological events during the Holocene. J. Quat. Sci. 26, 241–244.
Chorley, R.J., 1962. Geomorphology and general systems theory. US Geological Survey
Professional Paper 500-B. (10 pp.).
Clark, A.J., Tarling, D.H., Noël, M., 1988. Developments in archaeomagnetic dating in
Britain. J. Archaeol. Sci. 15, 645–667.
Collinson, J.D., 1996. Alluvial sediments. In: Reading, H.G. (Ed.), Sedimentary
Environments: Processes Facies and Stratigraphy. Blackwell, Oxford, pp. 5–36.
Combourieu Nebout, N., Peyron, O., Dormoy, I., Desprat, S., Beaudouin, C., Kotthoff, U.,
Marret, F., 2009. Rapid climatic variability in the west Mediterranean during the last
25,000 years from high resolution pollen data. Clim. Past 5, 503–521.
Cortesi, N., Gonzalez-Hidalgo, J.C., Trigo, R.M., Ramos, A.M., 2014. Weather types and
spatial variability of precipitation in the Iberian Peninsula. Int. J. Climatol. 34,
2661–2677.
80
Earth-Science Reviews 174 (2017) 53–83
Coulthard, T.J., Van De Wiel, M.J., 2012. Modelling river history and evolution. Philos.
Trans. R. Soc. London, Ser. A 370, 2123–2142.
Coulthard, T.J., Van de Wiel, M.J., 2013. Climate, tectonics or morphology: what signals
can we see in drainage basin sediment yields? Earth Surf. Dyn. 1, 13–27.
Currás, A., Zamora, L., Reed, J., García-Soto, E., Ferrero, S., Armengol, X., Mezquita
Joanes, F., Marqués, M., Riera, S., Julià, R., 2012. Climate change and human impact
in central Spain during Roman times: high-resolution multi-proxy analysis of a tufa
lake record (Somolinos, 1280 m asl). Catena 89 (1), 31–53.
de Menocal, P., Ortiz, J., Guilderson, T., Adkins, J., Sarnthein, M., Baker, L., Yarusinsky,
M., 2000. Abrupt onset and termination of the African Humid Period: rapid climate
responses to gradual insolation forcing. Quat. Sci. Rev. 19, 347–361.
de Vente, J., Poesen, J., 2005. Predicting soil erosion and sediment yield at the basin
scale: scale issues and semi-quantitative models. Earth Sci. Rev. 71, 95–125.
Desprat, S., Sánchez Goñi, M.F., Loutre, M.-F., 2003. Revealing climatic variability of the
last three millennia in northwestern Iberia using pollen influx data. Earth Planet. Sci.
Lett. 213 (1–2), 63–78.
Domínguez-Villar, D., Vázquez-Navarro, J.A., Carrasco, R.M., 2012. Mid-Holocene ero-
sive episodes in tufa deposits from Trabaque Canyon, central Spain, as a result of
abrupt arid climate transitions. Geomorphology 161–162, 15–25.
Dusar, B., Verstraeten, G., D'haen, K., Bakker, J., Kaptijn, E., Waelkens, M., 2012.
Sensitivity of the Eastern Mediterranean geomorphic system towards environmental
change during the Late Holocene: a chronological perspective. J. Quat. Sci. 27,
371–382.
Engel, J.M., Ma, L., Sak, P.B., Gaillardet, J., Ren, M., Engle, M.A., Brantley, S.L., 2016.
Quantifying chemical weathering rates along a precipitation gradient on Basse-Terre
Island, French Guadeloupe: new insight from U-series isotopes in weathering rinds.
Geochim. Cosmochim. Acta 195, 29–67.
Evans, M.E., 1996. Archeomagnetic results from the Mediterranean region – an overview.
Geol. Soc. Spec. Publ. 105, 373–384.
Eybergen, F.A., Imeson, A.C., 1989. Geomorphological processes and climatic change.
Catena 16, 307–319.
Faust, D., Schmidt, M., 2009. Soil erosion processes and sediment fluxes in a
Mediterranean marl landscape, Campiña de Cádiz, SW Spain. Z. Geomorphol. 52 (2),
247–265.
Faust, D., Zielhofer, C., 2002. Reconstruction of the Holocene water level amplitude of
Oued Medjerda as an indicator for changes of the environmental conditions in
Northern Tunisia. Z. Geomorphol. Suppl. 128, 161–175.
Faust, D., Dı́az del Olmo, F., Baena Escudero, R., 2000. Soils in the Holocene alluvial
sediments of the Rio Fraja Valley, Spain: in situ or soil-sediments? CATENA 41,
133–142.
Faust, D., Zielhofer, C., Baena, R., Diaz del Olmo, F., 2004. High resolution fluvial record
of late Holocene geomorphic changes in Northern Tunisia: climatic or human impact?
Quat. Sci. Rev. 23, 1757–1775.
Faust, D., Yurena Yanes, Y., Willkommen, T., Roettig, Ch., Richter, D., Richter, D., Hans,
V., von Suchodoletz, H., Zöller, L., 2015. A contribution to the understanding of late
Pleistocene Dune sand-Paleosol-Sequences in Fuerteventura (Canary Islands).
Geomorphology 246, 290–304.
Feurdean, A., Perşoiu, A., Tanţău, I., Stevens, T., Magyari, E.K., Onac, B.P., Marković, S.,
Andrič, M., Connor, S., Fărcaş, S., Gałka, M., Gaudeny, T., Hoek, W., Kolaczek, P.,
Kuneš, P., Lamentowicz, M., Marinova, E., Michczyńska, D.J., Perşoiu, I., Płóciennik,
M., Słowiński, M., Stancikaite, M., Sumegi, P., Svensson, A., Tămaş, T., Timar, A.,
Tonkov, S., Toth, M., Veski, S., Willis, K.J., Zernitskaya, V., 2014. Climate variability
and associated vegetation response throughout Central and Eastern Europe (CEE)
between 60 and 8 ka. Quat. Sci. Rev. 106, 206–224.
Fletcher, W., Zielhofer, C., 2013. Fragility of Western Mediterranean landscapes during
Holocene rapid climate changes. Catena 103, 16–29.
Fletcher, W.J., Boski, T., Moura, D., 2007. Palynological evidence for environmental and
climatic change in the lower Guadiana valley, Portugal, during the last 13,000 years.
The Holocene 17, 481–494.
Fletcher, W.J., Sanchez Goñi, M.F., Peyron, O., Dormoy, I., 2010. Abrupt climate changes
of the last deglaciation detected in a Western Mediterranean forest record. Clim. Past
6, 245–264.
Fryirs, K., 2013. (Dis)Connectivity in catchment sediment cascades: a fresh look at the
sediment delivery problem. Earth Surf. Process. Landf. 38, 30–46.
Fuchs, M., Owen, L.A., 2008. Luminescence dating of glacial and associated sediments:
review, recommendations and future directions. Boreas 37, 636–659.
Gallet, Y., Genevey, A., Le Goff, M., 2002. Three millennia of directional variation of the
Earth's magnetic field in western Europe as revealed by archeological artefacts. Phys.
Earth Planet. Inter. 131, 81–89.
Gasse, F., 2002. Diatom-inferred salinity and carbonate oxygen isotopes in Holocene
waterbodies of the western Sahara and Sahel (Africa). Quat. Sci. Rev. 21, 737–767.
Georgieva, K., Kirov, B., Tonev, P., Guineva, V., Atanasov, D., 2007. Long-term variations
in the correlation between NAO and solar activity: the importance of north-south
solar activity asymmetry for atmospheric circulation. Adv. Space Res. 40, 1152–1166.
Gil Garcia, M., Valiño, M.D., Rodriguez, A.V., Zapata, M.R., 2002. Late-glacial and
Holocene palaeoclimatic record from Sierra de Cebollera (northern Iberian Range,
Spain). Quat. Int. 93–94, 13–18.
Giorgi, F., Lionello, P., 2008. Climate change projections for the Mediterranean region.
Glob. Planet. Chang. 63, 90–104.
Glaser, R., Riemann, D., Schönbein, J., Barriendos, M., Brázdil, R., Bertolin, C., Camuffo,
D., Deutsch, M., Dobrovolný, P., van Engelen, A., Enzi, S., Halícková, M., Koenig, S.J.,
Kotyza, O., Limanówka, D., Macková, J., Sghedoni, M., Martin, B., Himmelsbach, I.,
2010. The variability of European floods since AD 1500. Clim. Chang. 101, 235–256.
Graf, W.L., 1982. Spatial variation of fluvial processes in semi-arid lands. In: Thorne, C.E.
(Ed.), Space and Time in Geomorphology, pp. 193–217 (London).
Gutiérrez-Elorza, M., Peña-Monné, J., 1998. Geomorphology and late Holocene climatic
D. Faust, D. Wolf
change in Northeastern Spain. Geomorphology 23, 205–217.
Harvey, A.M., 2002. Effective timescales of coupling within fluvial systems.
Geomorphology 44, 175–201.
Harvey, A.M., 2007. Geomorphic instability and change–Introduction: Implications of
temporal and spatial scales. Geomorphology 84, 153–158.
Hidalgo-Muñoz, J., Argüeso, D., Gámiz-Fortis, S., Esteban-Parra, M., Castro-Díez, Y.,
2011. Trends of extreme precipitation and associated synoptic patterns over the
southern Iberian Peninsula. J. Hydrol. 409, 497–511.
Höbig, N., Weber, M.E., Kehl, M., Weniger, G.-C., Julià, R., Melles, M., Fülöp, R.H., Vogel,
H., Reicherter, K., 2012. Lake Banyoles (northeastern Spain): a Last Glacial to
Holocene multi-proxy study with regard to environmental variability and human
occupation. Quat. Int. 274, 205–218.
Hoffmann, T., 2015. Sediment residence time and connectivity in non-equilibrium and
transient geomorphic systems. Earth Sci. Rev. 150, 609–627.
Houben, P., 2007. Geomorphological facies reconstruction of Late Quaternary alluvia by
the application of fluvial architecture concepts. Geomorphology 86, 94–114.
Howard, A.J., Gearey, B.R., Hill, T., Fletcher, W., Marshall, P., 2009. Fluvial sediments,
correlations and palaeoenvironmental reconstruction: The development of robust
radiocarbon chronologies. J. Archaeol. Sci. 36, 2680–2688.
Huerta, P., Armenteros, I., Silva, P.G., 2011. Large-scale architecture in non-marine ba-
sins: the response to the interplay between accommodation space and sediment
supply. Sedimentology 58, 1716–1736.
Hurrell, J.W., 1995. Decadal trends in the North Atlantic Oscillation: regional tempera-
tures and precipitation. Science 269, 676–679.
Imeson, A.C., Lavee, H., 1998. Soil erosion and climate change: the transect approach and
the influence of scale. Geomorphology 23, 219–227.
Ineson, S., Scaife, A.A., Knight, J.R., Manners, J.C., Dunstone, N.J., Gray, L.J., Haigh, J.D.,
2011. Solar forcing of winter climate variability in the Northern Hemisphere. Nat.
Geosci. 4, 753–757.
Jalut, G., Turu I Michels, V., Dedoubat, J.-J., Otto, T., Ezquerra, J., Fontugne, M., Belet,
J.M., Bonnet, L., de Celis, A.G., Redondo-Vega, J.M., Vidal-Romani, J.R., Santos, L.,
2010. Palaeoenvironmental studies in NW Iberia (Cantabrian range): vegetation
history and synthetic approach of the last deglaciation phases in the western
Mediterranean. Palaeogeogr. Palaeoclimatol. Palaeoecol. 297 (2), 330–350.
Jiménez-Moreno, G., Rodríguez-Ramírez, A., Pérez-Asensio, J.N., Carrión, J.S., López-
Sáez, J.A., Villarías-Robles, J.J.R., Celestino-Pérez, S., Cerrillo-Cuenca, E., León, A.,
Contreras, C., 2015. Impact of late-Holocene aridification trend, climate variability
and geodynamic control on the environment from a coastal area in SW Spain. The
Holocene 1–11.
Jones, P.D., Jonsson, T., Wheeler, D., 1997. Extension to the North Atlantic oscillation
using instrumental pressure observations from Gibraltar and south-west Iceland. Int.
J. Climatol. 17, 1433–1450.
Jones, A.F., Macklin, M.G., Benito, G., 2015. Meta-analysis of Holocene fluvial sedi-
mentary archives: a methodological primer. Catena 130, 3–12.
Jothityangkoon, C., Sivapalan, M., 2001. Temporal scales of rainfall-runoff processes and
spatial scaling of flood peaks: Space-time connection through catchment water bal-
ance. Adv. Water Resour. 24, 1015–1036.
Kirchner, A., Nehren, U., Behling, H., Heinrich, J., 2015. Mid- and late Holocene fluvial
dynamics in the tropical Guapi-Macacu catchment, Southeast Brazil: the role of cli-
mate change and human impact. Palaeogeogr. Palaeoclimatol. Palaeoecol. 426,
308–318.
Knox, J.C., 1972. Valley alluviation in Southwestern Wisconsin. Ann. Assoc. Am. Geogr.
62 (3), 401–410.
Lamb, H., van der Kaars, S., 1995. Vegetational response to Holocene climatic change:
pollen and palaeolimnological data from the Middle Atlas, Morocco. The Holocene 5,
400–408.
Lane, S.N., Richards, K.S., 1997. Linking river channel form and process: time, space and
causality revisited. Earth Surf. Process. Landf. 22, 249–260.
Langbein, W.B., Schumm, S.A., 1958. Yield of sediment in relation to mean annual pre-
cipitation. Trans. Am. Geophys. Union 39 (6), 1076–1084.
Lavee, H., Imeson, A.C., Sarah, P., 1998. The impact of climate change on geomorphology
and desertification along a mediterranean-arid transect. Land Degrad. Dev. 9,
407–422.
Lewin, J., Macklin, M.G., 2003. Preservation potential for Late Quaternary river alluvium.
J. Quat. Sci. 18, 107–120.
Limondin-Lozouet, N., Haddoumi, H., Lefèvre, D., Ghamizi, M., Aouraghe, H., Salel, T.,
2013. Holocene molluscan succession from NE Morocco: palaeoenvironmental re-
construction and biogeographical implications. Quat. Int. 302, 61–76.
López-Merino, L., Silva Sánchez, N., Kaal, J., López-Sáez, J.A., Martínez Cortizas, A.,
2012. Post-disturbance vegetation dynamics during the Late Pleistocene and the
Holocene: an example from NW Iberia. Glob. Planet. Chang. 92–93, 58–70.
López-Sáez, J.A., Blanco-González, A., López-Merino, L., Ruiz-Zapata, M.B., Dorado-
Valiño, M., Pérez-Díaz, S., Valdeolmillos, A., Burjachs, F., 2009. Landscape and cli-
matic changes during the end of the Late Prehistory in the Amblés Valley (Ávila,
central Spain), from 1200 to 400 cal BC. Quat. Int. 200 (1–2), 90–101.
Lorz, C., Kleber, A., Faust, D., 2008. From archive to process-editorial. Z. Geomorphol.
(Suppl. 52), V–VI.
Luterbacher, J., Xoplaki, E., Dietrich, D., Jones, P.D., Davies, T.D., Portis, D., et al., 2002.
Extending North Atlantic Oscillation reconstructions back to 1500. Atmos. Sci. Lett.
2, 114–124.
Luterbacher, J., García-Herrera, R., Akcer-On, S., et al., 2012. A review of 2000 years of
paleoclimatic evidence in the Mediterranean. In: Lionello, P. (Ed.), The Climate of the
Mediterranean region: From the past to the future. Elsevierer, Amsterdam, pp.
87–185.
Macklin, M.G., Lewin, J., 2003. River sediments, great floods and centennial-scale
Holocene climate change. J. Quat. Sci. 18, 101–105.
81
Earth-Science Reviews 174 (2017) 53–83
Macklin, M.G., Benito, G., Gregory, K.J., Johnstone, E., Lewin, J., Soja, R., Starkel, L.,
Thorndycraft, V.R., 2006. Past hydrological events reflected in the Holocene fluvial
history of Europe. Catena 66, 145–154.
Macklin, M.G., Jones, A.F., Lewin, J., 2011. Comment: cumulative probability functions
and their role in evaluating the chronology of geomorphological events during the
Holocene. J. Quat. Sci. 26, 238–240.
Macklin, M.G., Lewin, J., Woodward, J.C., 2012. The fluvial record of climate change.
Phil. Trans. R. Soc. A 370, 2143–2172.
Mann, M.E., Zhang, Z., Rutherford, S., Bradley, R.S., Hughes, M.K., Shindell, D., Ammann,
C., Faluvegi, G., Ni, F., 2009. Global signatures and dynamical origins of the little ice
age and medieval climate anomaly. Science 27, 1256–1260.
Marston, R.A., 2010. Geomorphology and vegetation on hillslopes: interactions, de-
pendencies, and feedback loops. Geomorphology 116, 206–217.
Martín-Puertas, C., Valero-Garcés, B., Pilar Mata, M., Gónzalez-Sampériz, P., Bao, R.,
Moreno, A., Stefanova, V., 2008. Arid and humid phases in southern Spain during the
last 4000 years: the Zoñar Lake record, Córdoba. The Holocene 18, 907–921.
Martin-Vide, J., 2004. Spatial distribution of a daily precipitation concentration index in
peninsular Spain. Int. J. Climatol. 24, 959–971.
Martin-Vide, J., Lopez-Bustins, J.-A., 2006. The Western Mediterranean Oscillation and
rainfall in the Iberian Peninsula. Int. J. Climatol. 26, 1455–1475.
Mayewski, P.A., Rohling, E.E., Curt Stager, J., Karlén, W., Maasch, K.A., David Meeker, L.,
Meyerson, E.A., Gasse, F., van Kreveld, S., Holmgren, K., Lee-Thorp, J., Rosqvist, G.,
Rack, F., Staubwasser, M., Schneider, R.R., Steig, E.J., 2004. Holocene climate
variability. Quat. Res. 62, 243–255.
McCarroll, D., 2015. ‘Study the past, if you would divine the future’: a retrospective on
measuring and understanding Quaternary climate change. J. Quat. Sci. 30, 154–187.
Mensing, S.A., Tunno, I., Sagnotti, L., Florindo, F., Noble, P., Archer, C., Zimmerman, S.,
Pavón-Carrasco, F.J., Cifani, G., Passigli, S., Piovesan, G., 2015. 2700 years of
Mediterranean environmental change in central Italy: a synthesis of sedimentary and
cultural records to interpret past impacts of climate on society. Quat. Sci. Rev. 116,
72–94.
Miall, A.D., 1985. Architectural-element analysis: a new method of facies analysis applied
to fluvial deposits. Earth Sci. Rev. 22, 261–308.
Moldenhauer, K.-M., Zielhofer, C., Faust, D., 2008. Heavy metals in alluvial soils as in-
dicators for sediment provenance in a semiarid catchment. Quat. Int. 189, 129–134.
Morellón, M., Valero-Garcés, B., Vegas-Vilarrúbia, T., González-Sampériz, P., Romero, O.,
Delgado-Huertas, A., Mata, P., Moreno, A., Rico, M., Corella, J.P., 2009. Lateglacial
and Holocene palaeohydrology in the western Mediterranean region: the Lake
Estanya record (NE Spain). Quat. Sci. Rev. 28, 2582–2599.
Moreno, A., Valero-Garcés, B.L., González-Sampériz, P., Rico, M., 2008. Flood response to
rainfall variability during the last 2000 years inferred from the Taravilla Lake record
(Central Iberian Range, Spain). J. Paleolimnol. 40, 943–961.
Moreno, A., González-Sampériz, P., Morellón, M., Valero-Garcés, B.L., Fletcher, W.J.,
2012. Northern Iberian abrupt climate change dynamics during the last glacial cycle:
A view from lacustrine sediments. Quat. Sci. Rev. 36, 139–153.
Muñoz-Díaz, D., Rodrigo, F.S., 2003. Effects of the North Atlantic Oscillation on the
probability for climatic categories of local monthly rainfall in Southern Spain. Int. J.
Climatol. 23, 381–397.
Muñoz-Díaz, D., Rodrigo, F.S., 2004. Impacts of the North Atlantic Oscillation on the
probability of dry and wet winters in Spain. Clim. Res. 27, 33–43.
Murray, A.B., Lazarus, E., Ashton, A., Baas, A., Coco, G., Coulthard, T., Fonstad, M., Haff,
P., McNamara, D., Paola, C., Pelletier, J., Reinhardt, L., 2009. Geomorphology,
complexity, and the emerging science of the Earth's surface. Geomorphology 103,
496–505.
Murray, A.B., Coco, G., Goldstein, B., 2014. Cause and effect in geomorphic systems:
complex systems perspectives. Geomorphology 214, 1–9.
Nadal-Romero, E., González-Hidalgo, J.C., Cortesi, N., Desir, G., Gómez, J.A., Lasanta, T.,
Lucía, A., Marín, C., Martínez-Murillo, J.F., Pacheco, E., Rodríguez-Blanco, M.L.,
Romero Díaz, A., Ruiz-Sinoga, J.D., Taguas, E.V., Taboada-Castro, M.M., Taboada-
Castro, M.T., Úbeda, X., Zabaleta, A., 2015. Relationship of runoff, erosion and se-
diment yield to weather types in the Iberian Peninsula. Geomorphology 228,
372–381.
Navarro Hervás, F., Ros Salas, M.M., Rodríguez Estrella, T., Fierro Enrique, E., Carrión,
J.S., García Veigas, J., Flores, J.A., Bárcena, M.A., García, M.S., 2014. Evaporite
evidence of a mid-Holocene (c. 4550–4400 cal. yr BP) aridity crisis in southwestern
Europe and palaeoenvironmental consequences. The Holocene 24, 489–502.
Notebaert, B., Verstraeten, G., 2010. Sensitivity of West and Central European river
systems to environmental changes during the Holocene: a review. Earth Sci. Rev. 103,
163–182.
Notebaert, B., Verstraeten, G., Ward, P., Renssen, H., Van Rompaey, A., 2011. Modeling
the sensitivity of sediment and water runoff dynamics to Holocene climate and land
use changes at the catchment scale. Geomorphology 126, 18–31.
Olsen, J., Anderson, N.J., Knudsen, M.F., 2012. Variability of the North Atlantic
Oscillation over the past 5200 years. Nat. Geosci. 5, 808–812.
Ortega, J.A., Garzón, G., 2009. A contribution to improved flood magnitude estimation in
base of palaeoflood record and climatic implications - Guadiana River (Iberian
Peninsula). Nat. Hazards Earth Syst. Sci. 9, 229–239.
Pérez-Obiol, R., Jalut, G., Julía, R., Pelachs, A., Iriarte, M.J., Otto, T., Hernandez-Beloqui,
B., 2011. Mid-Holocene vegetation and climatic history of the Iberian Peninsula. The
Holocene 21, 75–94.
Phillips, J.D., 1999. Divergence, convergence, and self-organization in landscapes. Ann.
Assoc. Am. Geogr. 89, 466–488.
Phillips, J.D., 2007. The perfect landscape. Geomorphology 84, 159–169.
Phillips, J.D., 2014. Thresholds, mode switching, and emergent equilibrium in geo-
morphic systems. Earth Surf. Process. Landf. 39, 1096–9837.
Phillips, J.D., 2015. Badass geomorphology. Earth Surf. Process. Landf. 40, 22–33.
D. Faust, D. Wolf
Pons, A., Reille, M., 1988. The Holocene- and upper Pleistocene pollen record from Padul
(Granada, Spain): a new study. Palaeogeogr. Palaeoclimatol. Palaeoecol. 66 (3–4),
243–263.
Popper, K.R., 1968. The Logic of Scientific Discovery.- New York. (480 pp.).
Puigdefabregas, J., Sole, A., Gutierrez, L., del Barrio, G., Boer, M., 1999. Scales and
processes of water and sediment redistribution in drylands: results from the Rambla
Honda field site in Southeast Spain. Earth Sci. Rev. 48, 39–70.
Queralt, S., Hernández, E., Barriopedro, D., Gallego, D., Ribera, P., Casanova, C., 2009.
North Atlantic Oscillation influence and weather types associated with winter total
and extreme precipitation events in Spain. Atmos. Res. 94, 675–683.
Ramrath, A., Sadori, L., Negendank, F.W., 2000. Sediments from Lago die Mezzano,
central Italy: a record of Lateglacial/Holocene climatic variations and anthropogenic
impact. The Holocene 10, 87–95.
Reed, J., Stevenson, A., Juggins, S., 2001. A multi-proxy record of Holocene climatic
change in southwestern Spain: the Laguna de Medina, Cádiz. The Holocene 11,
707–719.
Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Grootes,
P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., Hatté, C., Heaton, T.J., Hoffmann,
D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M.,
Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M.,
van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves
0–50,000 years cal BP. Radiocarbon 55, 1869–1887.
Reineck, H.E., Singh, I.B., 1980. Depositional Sedimentary Environments With Reference
to Terrigenous Clastics. Springer-Verlag, Berlin-Heidelberg-New York, pp. 549.
Revelles, J., Cho, S., Iriarte, E., Burjachs, F., van Geel, B., Palomo, A., Piqué, R., Peña-
Chocarro, L., Terradas, X., 2015. Mid-Holocene vegetation history and Neolithic land-
use in the Lake Banyoles area (Girona, Spain). Palaeogeogr. Palaeoclimatol.
Palaeoecol. 435, 70–85.
Rittenour, T.M., 2008. Luminescence dating of fluvial deposits: applications to geo-
morphic, palaeoseismic and archaeological research. Boreas 37, 613–635.
Rixhon, G., Briant, R.M., Cordier, S., Duval, M., Jones, A., Scholz, D., 2017. Revealing the
pace of river landscape evolution during the Quaternary: recent developments in
numerical dating methods. Quat. Sci. Rev. 166, 91–113.
Rodrigo, F.S., Estenan-Parra, M.J., Pozo-Vázquez, D., Castro-Diez, Y., 1999. A 500-year
precipitation record in Southern Spain. Int. J. Climatol. 19, 1233–1253.
Rodrigo, F.S., Estenan-Parra, M.J., Pozo-Vázquez, D., Castro-Diez, Y., 2000. Rainfall
variability in southern Spain on decadal to centennial time scales. Int. J. Climatol. 20,
721–732.
Rodríguez Vidal, J., Gracia, F.J., Giles, F., 1993. Deformaciones cuaternarias en la terraza
fluvial de la Laguna de Medina (Rio Guadalete, Jerez de la Frontera). Rev. Soc. Geol.
Esp. 6, 47–55.
Rohdenburg, H., 1989. Landscape Ecology, Geomorphology. Catena, Reiskirchen
(177 pp.).
Romans, B.W., Castelltort, S., Covault, J.A., Fildani, A., Walsh, J.P., 2016. Environmental
signal propagation in sedimentary systems across timescales. Earth Sci. Rev. 153,
7–29.
Rose, J., Meng, X., Watson, C., 1999. Palaeoclimate and palaeoenvironmental responses
in the western Mediterranean over the last 140 ka: evidence from Mallorca, Spain. J.
Geol. Soc. 156, 435–448.
Ruiz-Sinoga, J.D., Romero Diaz, A., 2010. Soil degradation factors along a Mediterranean
pluviometric gradient in Southern Spain. Geomorphology 118 (3–4), 359–368.
Salgueiro, A.R., Machado, M.J., Barriendos, M., Garcia Pereira, H., Benito, G., 2013.
Flood magnitudes in the Tagus River (Iberian Peninsula) and its stochastic relation-
ship with daily North Atlantic Oscillation since mid-19th Century. J. Hydrol. 502,
191–201.
Sánchez-López, G., Hernández, A., Pla-Rabes, S., Trigo, R.M., Toro, M., Granados, I., Sáez,
A., Masqué, P., Pueyo, J.J., Rubio-Inglés, M.J., Giralt, S., 2016. Climate reconstruc-
tion for the last two millennia in central Iberia: the role of East Atlantic (EA), North
Atlantic Oscillation (NAO) and their interplay over the Iberian Peninsula. Quat. Sci.
Rev. 149, 135–150.
Schumm, S.A., 1979. Geomorphic thresholds: the concept and its applications.
Transactions Institute of British Geographers 485–515.
Schumm, S.A., 1991. To Interpret the Earth. Ten Ways to be Wrong. Cambridge
University Press (133 pp.).
Schumm, S.A., 1993. River response to base level change: implications for sequence
stratigraphy. J. Geol. 101, 279–294.
Schumm, S.A., 2005. River Variability and Complexity. Cambridge University Press,
Cambridge (220 pp.).
Schumm, S.A., Lichty, R.W., 1965. Time, space, and causality in geomorphology. Am. J.
Sci. 263, 110–119.
Segond, M.-L., Wheater, H.S., Onof, C., 2007. The significance of spatial rainfall re-
presentation for flood runoff estimation: a numerical evaluation based on the Lee
catchment, UK. J. Hydrol. 347, 116–131.
Shanahan, T.M., McKay, N.P., Hughen, K.A., Overpeck, J.T., Otto-Bliesner, B., Heil, C.W.,
King, J., Scholz, C.A., Peck, J., 2015. The time-transgressive termination of the
African Humid Period. Nat. Geosci. 8, 140–144.
Sirocko, F., Sarnthein, M., Erlenkeuser, H., Lange, H., Arnold, M., Duplessy, J.C., 1993.
Century scale events in monsoonal climate over the past 24,000 years. Nature 364,
322–324.
Sirocko, F., Dietrich, S., Veres, D., Grootes, P.M., Schaber-Mohr, K., Seelos, K., Nadeau,
M.-J., Kromer, B., Rothacker, L., Röhner, M., Krbetschek, M., Appleby, P., Hambach,
U., Rolf, C., Sudo, M., Grim, S., 2013. Multi-proxy dating of Holocene maar lakes and
Pleistocene dry maar sediments in the Eifel, Germany. Quat. Sci. Rev. 62, 56–76.
Smith, S.M.C., Wilcock, P.R., 2015. Upland sediment supply and its relation to watershed
sediment delivery in the contemporary mid-Atlantic Piedmont (U.S.A.).
Geomorphology 232, 33–46.
82
Earth-Science Reviews 174 (2017) 53–83
Stokes, S., Walling, D.E., 2005. Radiogenic and isotopic methods for the direct dating of
fluvial sediments. In: Tools in Fluvial Geomorphology, pp. 231–267.
Tanguy, J.C., Bucur, I., Thompson, J.F.C., 1985. Geomagnetic secular variation in Sicily
and revised ages of historic lavas from Mount Etna. Nature 318, 453–455.
Thomas, M.F., 2001. Landscape sensitivity in time and space — an introduction. Catena
42, 83–98.
Thorndycraft, V.R., Benito, G., 2006. The Holocene fluvial chronology of Spain: evidence
from a newly compiled radiocarbon database. Quat. Sci. Rev. 25, 223–234.
Thorndycraft, V.R., Benito, G., Rico, M., Sopeña, A., Sánchez-Moya, Y., Casas, A., 2005. A
long-term flood discharge record derived from slackwater flood deposits of the
Llobregat River, NE Spain. J. Hydrol. 313, 16–31.
Thorndycraft, V.R., Benito, G., Sánchez-Moya, Y., Sopeña, A., 2012. Bayesian age mod-
elling applied to palaeoflood geochronologies and the investigation of Holocene flood
magnitude and frequency. The Holocene 22, 13–22.
Törnqvist, T.E., 2007. Fluvial environments/responses to rapid environmental change. In:
Elias, S.A. (Ed.), Encyclopedia of Quaternary Science. Elsevier, Amsterdam, pp.
686–694.
Trigo, R.M., Pozo-Vázquez, D., Osborn, T.J., Castro-Díez, Y., Gámiz-Fortis, S., Esteban-
Parra, M.J., 2004. North Atlantic Oscillation influence on precipitation, river flow
and water resources in the Iberian Peninsula. Int. J. Climatol. 24, 925–944.
Trimble, S.W., 2010. Streams, valleys and floodplains in the sediment cascade. In: Burt,
T., Allison, R.J. (Eds.), Sediment Cascades: An Integrated Approach. John
Wiley & Sons, Chichester, UK, pp. 307–344.
Trouet, V., Esper, J., Graham, N.E., Baker, A., Scourse, J.D., Frank, D.C., 2009.
Persistentpositive North Atlantic Oscillation mode dominated the Medieval climate
anomaly. Science 324, 78e80.
Uribelarrea, D., Pérez-González, A., Benito, G., 2003. Channel changes in the Jarama and
Tagus rivers (central Spain) over the past 500 years. Quat. Sci. Rev. 22, 2209–2221.
Van De Wiel, M.J., Coulthard, T.J., 2010. Self-organized criticality in river basins: chal-
lenging sedimentary records of environmental change. Geology 38, 87–90.
Vandenberghe, J., 1995. Timescales, climate and river development. Quat. Sci. Rev. 14,
631–638.
Vandenberghe, J., 2002. The relation between climate and river processes, landforms and
deposits during the Quaternary. Quat. Int. 91 (1), 17–23.
Vandenberghe, J., 2003. Climate forcing of fluvial system development: an evolution of
ideas. Quat. Sci. Rev. 22 (20), 2053–2060.
Vandenberghe, J., Cordier, S., Bridgland, D., 2010. Extrinsic and intrinsic forcing of
fluvial development: understanding natural and anthropogenic influences. In:
Proceedings of the Geologist's Association. 121. pp. 107–112.
Vaquero, J.M., 2004. Solar signal in the number of floods recorded for the Tagus river
basin over the last millennium. Clim. Chang. 66, 23–26.
Veldkamp, A., Baartman, J.E.M., Coulthard, T.J., Maddy, D., Schoorl, J.M., Storms, J.E.A.,
Temme, A.J.A.M., van Balen, R., van De Wiel, M.J., van Gorp, W., Viveen, W.,
Westaway, R., Whittaker, A.C., 2017. Two decades of numerical modelling to un-
derstand long term fluvial archives: advances and future perspectives. Quat. Sci. Rev.
166, 177–187 (in press).
Vicente-Serrano, S.M., 2006. Spatial and temporal analysis of droughts in the Iberian
Peninsula (1910–2000). Hydrol. Sci. J. 51, 83–97.
Vicente-Serrano, S.M., Cuadrat-Prats, J.M., Romo, A., 2006. Aridity influence on vege-
tation patterns in the middle Ebro Valley (Spain): evaluation by means of AVHRR
images and climate interpolation techniques. J. Arid Environ. 66 (2), 353–375.
Vicente-Serrano, S.M., Gouveia, C., Camarero, J.J., Beguería, S., Trigo, R., López-Moreno,
J.I., Azorín-Molina, C., Pasho, E., Lorenzo-Lacruz, J., Revuelto, J., Morán-Tejeda, E.,
Sanchez-Lorenzo, A., 2012. Response of vegetation to drought time-scales across
global land biomes. PNAS 1–6.
Vis, G.-J., Bohncke, S.J.P., Schneider, H., Kasse, C., Coenraads-Nederveen, S., Zuurbier,
K., Rozema, J., 2010. Holocene flooding history of the Lower Tagus Valley (Portugal).
J. Quat. Sci. 25, 1222–1238.
Vis, G.-J., Kasse, C., Kroon, D., Vandenberghe, J., Jung, S., Lebreiro, S.M., Rodrigues, T.,
2016. Time-integrated 3D approach of late Quaternary sediment-depocenter migra-
tion in the Tagus depositional system: from river valley to abyssal plain. Earth Sci.
Rev. 153, 192–211.
von Suchodoletz, H., Gärtner, A., Hoth, S., Umlauft, J., Sukhishvili, L., Faust, D., 2016.
Late Pleistocene river migrations in response to thrust belt advance and sediment-flux
steering — the Kura River (southern Caucasus). Geomorphology 266, 53–65.
Wainwright, J., Thornes, J.B., 2004. Environmental Issues in the Mediterranean: Process
and Perspectives From the Past to the Present. Routledge, London, U.K.
Walling, D.E., 1999. Linking land use, erosion and sediment yields in river basins.
Hydrobiologia 410, 223–240.
Walling, D.E., Kleo, A.H.A., 1979. Sediment Yields of Rivers in Areas of Low Precipitation:
A Global View. International Association of Hydrological Sciences.
Walling, D.E., Webb, B.W., 1996. Erosion and sediment yield: a global overview. In:
Walling, D.E., Webb, B.W. (Eds.), Erosion and Sediment Yield: Global and Regional
Perspectives, (Proceedings of the Exeter Symposium, July, 1996), IAHS Publication
No. 236. IAHS Press, Wallingford, UK, pp. 3–19.
Wanner, H., Solomina, O., Grosjean, M., Ritz, S.P., Jetel, M., 2011. Structure and origin of
Holocene cold events. Quat. Sci. Rev. 30, 3109–3123.
Wegmann, K.W., Pazzaglia, F.J., 2002. Holocene strath terraces, climate change, and
active tectonics; the Clearwater River basin, Olympic Peninsula, Washington State.
Geol. Soc. Am. Bull. 114, 731–744.
Wolf, D., Faust, D., 2013. Holocene sediment fluxes in a fragile loess landscape (Saxony,
Germany). Catena 103, 87–102.
Wolf, D., Faust, D., 2015. Western Mediterranean environmental changes: evidences from
fluvial archives. Quat. Sci. Rev. 122, 30–50.
Wolf, D., Faust, D., 2016. River braiding caused by rapid floodplain deformation - insights
from Holocene dynamics of the Jarama River in Central Spain. Quat. Int. 407,
D. Faust, D. Wolf
126–139.
Wolf, D., Seim, A., Diaz del Olmo, F., Faust, D., 2013. Late Quaternary fluvial dynamics of
the Jarama River in central Spain. Quat. Int. 302, 20–41.
Wolf, D., Seim, A., Faust, D., 2014. Fluvial system response to external forcing and human
impact - Late Pleistocene and Holocene fluvial dynamics of the lower Guadalete River
in western Andalucía (Spain). Boreas 43, 422–449.
Wolman, M.G., Miller, J.P., 1960. Magnitude and frequency of forces in geomorphic
processes. J. Geol. 68, 54–74.
Wright, V.P., Marriott, S.B., 1993. The sequence stratigraphy of fluvial depositional
systems: the role of floodplain sediment storage. Sediment. Geol. 86, 203–210.
Yaalon, D.H., 1997. Soils in the Mediterranean region: what makes them different?
Catena 28, 157–169.
Zielhofer, C., Faust, D., 2008. Mid- and Late Holocene fluvial chronology of Tunisia. Quat.
Sci. Rev. 27, 580–588.
Zielhofer, C., Faust, D., Diaz del Olmo, F., Baena, R., 2002. Sedimentation and soil for-
mation phases in the Ghardimaou Basin (northern Tunisia) during the Holocene.
Quat. Int. 93-94, 109–127.
83
Earth-Science Reviews 174 (2017) 53–83
Zielhofer, C., Faust, D., Baena, R., Diaz del Olmo, F., Kadereit, A., Moldenhauer, K.-M.,
Porras, A., 2004. Centennial-scale late Pleistocene to mid-Holocene synthetic profile
of the Medjerda floodplain (Northern Tunisia). The Holocene 14, 851–861.
Zielhofer, C., Faust, D., Linstädter, J., 2008. Late Pleistocene and Holocene alluvial ar-
chives in the Southwestern Mediterranean: Changes in fluvial dynamics and past
human response. Quat. Int. 181, 39–54.
Zielhofer, C., Recio Espejo, J.M., Nunez Granadoz, M.A., Faust, D., 2009. Durations of soil
formation and soil development indices in a Holocene Mediterranean floodplain.
Quat. Int. 209, 44–65.
Zielhofer, C., Bussmann, J., Ibouhouten, H., Fenech, K., 2010. Flood frequencies reveal
Holocene Rapid Climate Changes (Lower Moulouya River, NE Morocco). J. Quat. Sci.
25, 700–714.
Zielhofer, C., Fletcher, W.J., Mischke, S., De Batist, M., Campbell, J.F.G., Joannin, S.,
Tjallingii, R., El Hamouti, N., Junginger, A., Stele, A., Bussmann, J., Schneider, B.,
Lauer, T., Spitzer, K., Strupler, M., Brachert, T., Mikdad, A., 2017. Atlantic forcing of
Western Mediterranean winter rain minima during the last 12,000 years. Quat. Sci.
Rev. 157, 29–51.