Geochronology of Soils and Landforms in Cultural Landscapes on Aeolian Sandy Substrates, Based on Radiocarbon and Optically Stimulated Luminescence Dating (Weert, SE-Netherlands)

The landscape of the study area (fig. 1,2) is underlain by coversand, deposited during the Late Glacial of the Weichselian. In the Preboreal, aeolian processes reduced soil formation (Stichting voor Bodemkaratering, 1972) and from the Preboreal to the Atlantic a deciduous climax forest developed (Janssen, 1974). The geomorphology was a coversand landscape, composed of ridges (umbric podzols), coversand plains (gleyic podzols), coversand depressions (histic podzols) and small valleys (gleysols). The area was used by hunting people during the Late Paleolithic and Mesolithic (Nies, 1999). Analysis of the urnfield ‘Boshoverheide’, indicated that the population increased during the Bronze Age between 1000 and 400 BC to a community of several hundreds of people, living from forest grazing, shifting cultivation and trade (Bloemers, 1988). The natural deciduous forests gradually degraded into heath land. The deforestation accelerated soil acidification and affected the hydrology, which is reflected in drying out of ridges and wetting of depressions, promoting the development of histosols and histic podzols. Sustainable productivity on chemically poor sandy substrates required application of organic fertilizers, composed of a mixture of organic litter with animal manure with a very low mineral compound (Van Mourik et al., 2011a), produced in shallow stables (Vera, 2011). The unit plaggic anthrosol on the soil map of 1950 AD identifies the land surface, which was used for plaggen agriculture. At least since 1000 AD, heath management was regulated by a series of rules that aimed to protect the valuable heat lands against degradation (Vera, 2011). During the 11th, 12th and 13th centuries there was an increasing demand for wood and clear cutting transformed the majority of the forests in driftsand landscapes (Vera, 2011). The exposed landscape was subjected to wind erosion and accumulation which endangered heath, arable land and even farmhouses. As a consequence, umbric podzols, the natural climax soil under deciduous forests on coversand, degraded into larger scale driftsand landscapes, characterized by deflation plains (gleyic arenosols) and complexes of inland dunes (haplic arenosols) (Van Mourik et al., 2011b). In such driftsand landscapes, the majority of the podzolic soils in

coversand has been truncated by aeolian erosion. Only on scattered sheltered sites in the landscape, palaeopodzols were buried under mono or polycyclic driftsand deposits. They are now the valuable soil archives for palaeoecological research.
The city of Weert was founded at the end of the 13 th century on a deforested topographic ridge of dry sandy soils, surrounded by swampy heath lands (Nies, 1999). Around 1300 AD the citizens ensured the supply of fresh water for the city moat by the creation of the 'Weerter Beek', a canal to connect the moat with a wetland area near the present Belgium border (Salmans and Tillemans, 1994). The topographical map of 1550 AD ( fig.1) shows a deforested landscape surrounding the city, with distinct zones of arable land and heath; the natural forest had already completely been transferred into a cultural landscape.
During the 18 th century, the population growth and regional economical activity stimulated the agricultural productivity. Farmers introduced the innovative 'deep stable' technique to increase the production of fertilizers (Vera, 2011). Additional to mowed biomass, farmers collected heath sods, including the top of the Ah horizon of the humus forms. This consequently promoted heath degradation and sand drifting, resulting in the extension of driftsand landscapes. During the 19 th century, farmers tried to find alternative fertilizers and authorities initiated reforestation projects. The invention of chemical fertilizers at the end of the 19 th century marked the end of the period of heath management and plaggen agriculture (Spek, 2004;Van Mourik at al., 2011b;Vera, 2011). The heath was not longer used for the harvesting of plaggic matter and new land management practices were introduced. Heath was reclaimed to new arable land or reforested with Scotch pine. During the 20 th century the landscape dramatically changed again through a shift towards industrialization and bioindustry. Geomorphological features belonging to the historical sand drifting and plaggen agriculture survived in the landscape and are now included in the geological inheritance.
During recent decades the interest and need for restoration ecology and geoconservation has increased on global and regional scale (Bal et al., 2001). In the Netherlands, a national ecological master structure was designed to recover the ecological quality and biodiversity and in this context attention was paid to the preservation and restoration of driftsand habitats and landscapes (Koster, 2009(Koster, , 2010.
Soil maps often serve as abiotical archives for ecosystem restoration management. However, soil classifications are normally based on actual diagnostic properties and therefore neglect relicts of former phases in soil and landscape development. Consequently, soil maps only show the distribution of recent soil types and are thus useless to fully understand the interaction of natural and human processes in time and space. To overcome this gap in knowledge of long term impact of human land use on the development of landforms and soils, the results of three innovative methods, applied to a selection of formerly investigated palaeosols, are presented in this paper. Firstly, the application of OSL dating on formerly investigated and 14 C dated palaeosols, to improve the geochronology of the phases in landscape evolution. Secondly the application of biomarker analysis to select the plant species, responsible for the production of organic carbon, stored in humic soil horizons. Finally it is shown how the complete package of palaeoecological information can be processed into soil maps of paleo-landscapes using a geographical information system.

Profile selection
Since 1988 several pilot studies have been dedicated to the analysis of histosols, buried podzols and plaggic anthrosols around the city of Weert). Palynology, soil micromorphology and radiocarbon dating were the analytical tools to unlock the palaeoecological information from these valuable soil archives. For the reconstruction of the Late Holocene landscape evolution around the City of Weert, we selected several previously investigated key profiles. This selection comprises a histosol (Kruispeel), 3 buried histic podzols (IJzerenman, Tungelerwallen, Weerter Bergen), 3 buried (polycyclic) podzols (profiles Defensiedijk-1 and -2, Boshoverheide) and 3 plaggic anthrosols (profiles Tungelerakker, Dijkerakker and Valenakker).

Pollen analysis
Peat deposits are considered to contain syn-sedimentary pollen records and the diagrams reflect the characteristics of the local and regional vegetation development. Palaeosols are considered to contain a post-sedimentary pollen content as a result of bio-infiltration of pollen and the diagrams reflect ecological fingerprints of the soil ecological evolution (Van Mourik, 1999-b, 2001. Palynological reference of the Holocene vegetation development of SE-Netherlands, based on pollen analysis of extensive peat bogs in the Peel region have been published by Eshuis (1946) and Janssen (1974). The pollen zoning of the reference diagram is based on the zones of Firbas (1949). To understand the human impact on soils and landforms, local palaeoecological data was collected from a selection of palaeosols. From the profiles Kruispeel, IJzerenman and Weerter Bergen, samples were collected with an auger, vertical sampling distance 2.5 cm. From the other profiles, samples were collected in 10 ml tubes in profile pits, vertical sampling distance 2.5 cm. Pollen extractions were carried out using the tufa extraction method (Moore et al., 1991, p. 50). The exotic marker grain method was applied (Moore et al., 1991, p. 53) for estimation of the pollen densities in the mineral soils. For the identification of pollen grains the pollen key of Moore et al. (1991, p. 83-166) was applied. Pollen scores were based on the total pollen sum of arboreal and non arboreal plant species with the exception the aquatic species in pollen diagram Kruispeel. The pollen extractions were performed in the palynological laboratory IBED, University of Amsterdam. Pollen densities are represented in kilo grains / ml, or in the logarithmic value (log D)of the total amount of pollen grains / ml.

Soil micromorphology
Micromorphological observations are needed to assess the validity of soil ecological fingerprints, based on pollen spectra of drained soils. . In thin sections of soil horizons, pollen grains are detectable but not determinable. The quality of the micro environment of pollen grains provide evidence of infiltration and preservation processes. Micromorphological observations are also relevant to understand the presence and distribution of soil organic carbon. Secondary soil processes can affect the quality of the original soil organic matrix and alter the organic chemistry composition and consequently the results of radiocarbon dating (Van Mourik et al., 2010). Undisturbed soil samples were collected in Kubiena boxes.
Undisturbed soil samples of buried humic horizons were collected in Kubiena boxes. Thin sections of undisturbed soil samples were produced, using the method of Jongerius and Heintzberger (1976) in the soil laboratory of IBED, University of Amsterdam. Thin sections were used to study the occurrence and preservation.

14 C-dating
In traditional palynological research, radiocarbon dating was applied for absolute dating of pollen zones, as found in peat and limnic deposits. In palaeopedology, radiocarbon datings also have been used for dating purposes. The interpretation of radiocarbon datings of extracted soil organic carbon from soil samples is complicated. Due to the complexity of the provenance of soil organic carbon in humic soil horizons, conventional radiocarbon ages of bulk samples (BULK) are not very reliable. It is preferable to apply radiocarbon dating on extractions of fulvic acids (FUL), humic acids (HAC) and humin (HUM) (Goh and Molloy, 1978;, 2010. These fractions are based on extractability behavior. Fulvic acids are soluble in acid and in lye, the humic acids are insoluble in acid and soluble in lye and the humin fraction is insoluble in acid and in lye. The biological decomposition rate of FUL is relatively high; they migrate easily through weak acid soil profiles or leach completely. Therefore, they are unreliable for any dating purposes. The biological decomposition rate of HAC is medium high. Compared with FUL, they are immobile in the soil profiles and more reliable for dating purposes. The 14 C age of HAC is therefore expected to reflect the moment of burying of the soil. HUM (including pollen grains and charcoal) will accumulate in humic topsoil during an active period of soil development; therefore, 14 C ages of this fraction will definitely overestimate the age of burial. From these characteristics, we infer that the radiocarbon age difference between the HUM and HAC fractions of the same level will be greater during longer periods of active soil formation. To create a 14 C based geochronology, conventional 14 C www.intechopen.com dating was firstly applied on bulk samples of soil organic matter extracted from soil samples of the profiles Tungeler Wallen, Defensiedijk 1,2 and Boshoverheide; additional on FUL, HAC and HUM fractions of soil organic matter, extracted from samples of the profiles Defensiedijk 1 and 2 and Boshoverheide, and finally on HAC and HUM fractions of soil organic matter, extracted from samples of the profiles Kruispeel, Tungelruy, Dijkerakker,Weerter Bergen and IJzerenman. Radiocarbon datings have been performed in the Centrum voor Isotopen Onderzoek (CIO) of the University of Groningen, Netherlands (table 1). For calibration of the radiocarbon ages to calendar years, the program OxCal 4.1 (1 sigma confidence interval) was used.

Luminescence dating
For a general description of the OSL dating method is referred to Wintle (2008). Applications of OSL dating to plaggic deposits have been discussed earlier by Bokhorst et al. (2005) and Van Mourik et al. (2011b), while applications to polycyclic driftsand sequences were publishe in Van Mourik et al. (2010). The OSL samples were collected in standard pFrings. Luminescence measurements used an automated Risø TL/OSL reader (DA 15) equipped with an internal Sr/Y beta source, and blue and IR diodes (Bøtter-Jensen et al., 2000). A single-aliquot regenerative dose (SAR) procedure was used for equivalent dose estimation (Murray and Wintle, 2003). OSL ages were calculated by dividing the sample burial dose by the dose rate and can be expressed in years relative to the year of sampling as well as in AD / BC. Quoted uncertainties contain all random and systematic errors in both the dose rate and burial dose assessment, and reflect the 1-sigma confidence interval. Luminescence dating of profile Defensiedijk 1 was performed at the Department of Geology and Soil Science, Laboratory of Mineralogy and Petrology (Luminescence Research Group), Ghent University (UG); of profile Dijkerakker at the Institute of Geography and Earth Sciences, University of Wales, Aberystwyth (UW); of profile Valenakker and Boshoverheide at the Netherlands Centre of Luminescence dating of the Delft University of Technology (NCL).

Biomarker analysis
Two pilots were chosen for the application of biomarker analysis. The first pilot was a polycyclic driftsand sequence. The research question was: Does the combination of pollen and biomarker analysis allow for a selection of the responsible plant species for the production of biomass and sequestration of soil organic carbon in buried humic horizons? We selected n-alkanes and n-alcohols with carbon numbers 20 -36, which are exclusive to the epicuticular wax layers on leaves and roots of higher plants (Kolattukudy et al., 1976), as biomarkers of past vegetation. Recently, analysis of n-alkane and n-alcohol patterns preserved in Ecuadorian soils enabled a reconstruction of past vegetation dynamics in the area (Jansen et al., 2008). To explore the applicability of biomarker analysis for vegetation reconstructions in a wider range of soils we applied it as an additional proxy in a selected profile in the Weert setting: Defensiedijk-1. To this end the same A horizon samples were used as for the analysis of fossil pollen in the profile. In addition, leaves and roots from species expected to have been responsible for the dominant biomass input in the profile were sampled from the present day vegetation. These consisted of Polytrichum piliferum, www.intechopen.com Cladonia rangiferina, Calluna vulgaris, Molinia caerulea, Corynephorus canescens, Deschampsia flexuosa, Pinus sylvestris, Betula pendula and Quercus robur. Approximately 0.1 g of each of the freeze-dried and ground vegetation and soil samples was extracted by Accelerated Solvent Extraction (ASE) using a Dionex 200 ASE extractor. The extraction temperature was 75ºC and the extraction pressure 17 × 10 6 Pa, employing a heating phase of 5 min and a static extraction time of 20 min. Dichloromethane/methanol (DCM/MeOH) (93:7 v/v) was used as the extractant (Jansen et al., 2006a). The extracts were pre-treated and derivatized with BSTFA (N,O-bis(trimethylsilyl) trifluoroacetamide) containing 1% TMCS (trimethylchlorosilane) following a previously described protocol (Jansen et al., 2006b). Sample analysis took place on a ThermoQuest Trace GC 2000 gas chromatograph connected to a Finnigan Trace quadrupole mass spectrometer (MS), Separation took place by oncolumn injection of 1.0 µl of the derivatized extracts on a 30 m Rtx-5Sil MS column (Restek) with an internal diameter of 0.25 mm and film thickness of 0.1 µm, using He as a carrier gas. Temperature programming was: 50ºC (hold 2 min); 40ºC/min to 80ºC (hold 2 min); 20ºC/min to 130ºC; 4ºC/min to 350ºC (hold 10 min). Subsequent MS detection in full scan mode used a mass-to-charge ratio (m/z) of 50-650 with a cycle time of 0.65 s and followed electron impact ionization (70 eV). The n-alkanes and n-alcohols were identified by their mass spectra and retention times and quantified using a deuterated internal standard (d 42 -n-C 20 alkane and d 41 -n-C 20 alcohol) (Jansen et al., 2006b). The concentrations of n-alkanes and n-alcohols with carbon numbers 20-36 in vegetation and soil samples were subsequently used as input for the VERHIB model that was specifically designed to translate such biomarker patterns into the most likely past vegetations patterns (Jansen et al. 2010). As required boundary conditions for the model we assumed a leaf biomarker vs. root biomarker input ratio in the soil of 1:10. For the tree species we assumed the root input to be equally distributed with depth, while for the grass and heath species we assumed all root input to have taken place within the first 36 cm, with 75% of that within the first 2 cm. The lichen and moss species were assumed to have given input only at the surface. The second pilot was a plaggic anthrosol. The research question was: Does the combination of pollen and biomarker analysis enable to detect the origin of the collected biomass for the preparation of plaggic manure? Three samples were analyzed from the plaggic horizons of profile Valenakker. Because of the limited number of horizons and the number of samples was too small to apply the VERHIB model.

Processing in ArcGIS
Three soil maps for the Weert area have been prepared in a digital environment in ArcGIS 10.0 (http://www.esri.com), following the method described by van Mourik et al.  The translation of local (Dutch) soil type names in international labels is based on the World Reference Base (ISRIC/FAO, 2006). The vector-based soil map was clipped to the extent of the Weert study area by using a rectangular mask. The approximately 300 original legend categories were thus reduced to 17 legend units. These were further aggregated, based on similarities in soil texture properties, to 10 categories (table 1).
For the reconstruction of the soil maps of 1500AD and 2000 BC, detailed information on time development was used, which was obtained from the 10 key soil profiles described in detail in this chapter. This refers both to the age and to the palaeoecological information of buried soil horizons, which is indicative for renewed landscape dynamics, that is driven by either natural processes or man-induced interference. Crucial in this step is to define soil sequences for a time-span of approximately 4000 years, which is based on properties of the parent material, position in the landscape, local groundwater conditions over time and historical land cover and land use changes and, to a lesser extent, climate change. In the attribute table of the clipped digital soil geodatabase three additional columns were added for the 1950 AD, 1500 AD and the 2000 BC situations. Detailed, 5m resolution topographical information was used from the digital 'Algemeen Hoogtemodel Nederland' (AHN5) which contains heights information expressed in cm. Land cover and land use data has been extracted from historical maps around 1550 AD, which shows an almost total deforested landscape with 'islands' of sand around the initial settlement of Weert and locations of former swampy areas. Visualization of the newly generated soil attributes leads to so-called 'interpretative historical soil maps' for the 1500 AC and 2000 BC situations. It was supposed that boundaries between the soil units did not significantly change over this time span. The soil boundaries that underlie the currently urbanized areas were reconstructed according to their likely fit with neighboring soil boundaries of the digital soil map of 2006, and by using historical map information.

Palaeoecological information from a histosol and buried histic podzols
14 C Datings of (buried) histic horizons indicate the start of peat accumulation in coversand depressions around Weert between 1000 and 500 BC, due to wetting of depressions caused by deforestation of the surrounding higher coversand ridges. Similar datings were found in the study area Maashorst, 55 km north of Weert (Van Mourik et al., 2011b). In the more extensive depression of Kruispeel, the accumulation of peat continued during the Late Subboreal and Subatlantic; in small scale depressions gleyic podzols just transferred in histic podzols before they got buried under driftsand deposits around 1000 AD.

Profile Kruispeel, terric histosol (fig. 3-4; table 2; ); pollen diagram first published in Van Mourik (1988)
Kruispeel used to be a shallow peat bog, situated in a depression in the coversand landscape ( fig. 2). The formation of the histosols started in the Early Preboreal (Pinus, Cyperaceaee) with the deposition of humus gyttja, representing a shallow lake bottom soil, followed by a phase of terrestrialization with peat accumulation (Betula, Salix, Artemisia, Helianthemum, Juniperus). Compared with the Firbas pollen zoning of the Peel references (Janssen 1974), the composition of the pollen spectra is indicative for pollen zone IV. In the Boreal (zone V) the peat accumulation slowed down. The Atlantic (zones VI, VII) is absent. The accumulation of peat accelerated in the Late Subboreal (zone VIII) around 3300 BC (table 2) and continued in the Subatlantic (zones IX, X). The pollen spectra reflect a mix of species from the disappearing forest (Corylus, Alnus, Quercus, Fagus) and the emerging cultural landscape (Ericaceae, Cerealia, Fagopyrum, Plantago). Cerealia pollen is present in pollen spectra since the Bronze Age, Fagopyrum was introduced around 1350 AD (Leenders, 1987).
Diagram Kruispeel shows a clear expression of the impact of the vegetation development of the topographic higher surroundings on the hydrology of the depression. Deforestation on ridges (lower evapotranspiration, higher soil water infiltration) resulted in accumulation of gyttja or peat in the depression. Forested ridges (higher evapotranspiration and lower soil water infiltration) as a contrast resulted in a reduction of the accumulation rate or even in erosion by bio-oxidaten. The youngest spectra reflect the start of the period of reforestation (Pinus) since 1850 AD.   Based on radiocarbon ages, the histic horizon of the palaeopodzol in coversand developed between 500 BC and 1000 AD. The post-sedimentary pollen spectra in the mineral horizon reflect decreasing scores of deciduous trees (Corylus, Tilia, Quercus), indicative for deforestation and high scores of Ericaceae, marking the extension of heath. Moist conditions during the development of the 3H or reflected by the scores of Sphagnum. Shortly after 1000 AD the histic podzol was buried by driftsand deposits. This age correlates with the period of forest clear cutting (Vera, 2011 Based on 14 C datings, the histic horizon developed between 400 BC and 600 AD. The pollen spectra of the palaeopodzol show, just as Tungeler Wallen and IJzerenman, high scores of elements of the former deciduous forest (Corylus, Quercus, Tilia), but also Ericacea show high scores, indicating heath in the surroundings. The development of the 3Hh took place between ± 200 BC and ± 700 AD. Special attention was paid in this profile on the determination of Ericacae, using the special pollen key of Moore et al. (1992)

Palaeoecological information from buried carbic podzols
Especially in well drained sandy landscapes, polycyclic sequences with buried podzols are unique parts of the soil archives. Pollen spectra provide information about the vegetation during stable and instable period. Application of biomarker analysis allow the selection of species, rooting during stable periods and responsible for the sequestration of soil organic carbon. The combination of 14 C and OSL datings indicates that the oldest palaeopodzols have been buried by Pre-Mediaeval small scale driftsand deposits, probably related to natural causes as hurricanes and forest fires or early shifting cultivation. Around 1000 AD a more extensive land surface with podzols got buried by larger scale driftsand deposits, probably related to the period of forest clear cutting. A generation of micropodzols, developed in stabilized driftsands, was buried by younger driftsand deposits after the 17 th century, probably related to heath degradation due to the application of deep stable management. Application of radiocarbon dating in paleopedology is supposed to provide information for the establishment of the geochronology of the landscape evolution, but the interpretation of radiocarbon datings from paleosols is more complicated than in surveys of peat bogs and limnic deposits. This is caused by differential biological decomposition of soil organic litter, resulting in various fractions with different chemical compositions and turnover rates. For that reason, fractionated radiocarbon dating was introduced in paleopedology. Goh and Molloy (1978) investigated the suitability of radiocarbon datings of soil organic matter in quaternary geology and demonstrated the important role of the methods of extraction of organic matter. Ellis and Matthews (1984) established the differences in radiocarbon datings of FUL and HAC in palaeopodzols. The interpretations of the results of fractionated 14 C dating are heterogeneous. Mattheuws and Quentin (1983) selected the HAC fraction of the HF horizon for dating of a buried podzol in Norway. Hammond et al. (1991) established the importance of fractionated 14 C dating. In their research of peats and organic silts, FUL and HAC were considered as contaminates, leaching from podzolic environments. Dansgaard and Odgaard (2001)    To establish the geochronology of the polycyclic sequence, 14 C dating was applied (     The composition of pollen, precipitating on and infiltrating in a soil, is a mixture of pollen, dispersed by species rooting on the site and by species, present on distance. Consequently it is impossible to select the species, responsible for soil formation and humus sequestration from pollen spectra alone. Plant leaves are dispersed over much shorter distances by wind than pollen whereas plant root material normally enters a soil record in-situ, except for such cases where a large scale human induced deposition takes place e.g. through plaggen agriculture. As a result, in contrast to pollen records, biomarker records are expected to reflect better the local plant species responsible for soil formation and humus sequestration. For this purpose, Defensiedijk 1 was resampled again in 2008. Samples were taken from the mineral humic 1(A), 2Ae, 3Ah and 4Ah horizons and the humic driftsand layers 1C3, 1C5, 2C3 to compare pollen and biomarker spectra. Also a reference base was created for biomarkers of species, possibly involved in the carbon sequestration during soil formation (Pinus, Betula, Quercus, Calluna, Molinia, Corynephorus, Polytrichum, Cladonia) Fig .16 summarizes the combined results of pollen and biomarker analyses from profile Defensiedijk 1, 2008. Micromorphologically, the intertextic modexal organic aggregates in buried Ah and Ah horizons seem undisturbed, but biomarkers indicate that the original tissue derived compounds can be overruled by younger root derived compounds. This observation is confirmed by a comparison of the pollen and biomarker results.  The pollen spectrum of the 4Ah horizon is dominated by Ericaceae, Corylus and Alnus and micromorphological observations indicate an undisturbed soil matrix. Pollen grains can be extracted from organic aggregates with a modexal intern fabric and an intertextic distribution pattern. However, no biomarkers derived from Ericaceae or Corylus are present. Instead in the biomarker based reconstruction, Pinus and Poaceae are dominant. Pine trees were not introduced in the area until the 19 th century and therefore are unlikely to represent the onsite vegetation at the time that the 4Ah horizon was at the surface, given the results of the dating of the horizon (Table 7). Instead, the dominance of Pinus biomarkers most likely represents 'contamination' of the soil organic carbon in this horizon with younger decomposition products of the roots of this deep rooting species. At the same time, the low abundance of Ericaceae in the biomarker based reconstruction most likely indicates that this species was absent at the site in significant numbers. Its abundance in the pollen records has been caused by windblown dispersal of pollen from surrounding areas. A cover dominated by grass and moss species as inferred by the biomarker reconstruction seems more likely.

Profile
The pollen spectrum of the 3A horizon is dominated by Ericaceae. The micromorphological structure is similar to the 4Ah, while the biomarker spectrum is now dominated by Ericaceae and Pinus. Assuming the abundance of Pinus once more to represent younger root input, heath would appear to have been the dominant vegetation at the site at the time that the 3Ah horizon was at the surface. Most likely the Ericaceae that were already present in the vicinity of the site earlier, as indicated by their presence in the pollen spectra of the 4Ah horizon, by now had reached the sampling site, making them show up in the more local biomarker based reconstruction .When the 2Ah horizon was at the surface, the site was most likely covered by heath and lichens, the former being present in the pollen spectra as well as the biomarker reconstruction, the latter showing up only in the biomarker reconstruction since it does not produce pollen. Only some fungal spores are present in the pollen extractions but they do not allow the identification of the fungi, associated with lichens that occur on driftsand and heath (Cladena and Cladonia) (Domsch and Gams, 1970;Aptroot and Van Herk,1994). The set of fractionated 14 C datings show similar features as the set of Defensiedijk 1. The 14 C ages of the oldest palaeosols (1S) are in line with the composition of the pollen spectra. Dominance of Alnus instead of Corylus, due to continuous infiltration of younger pollen grains and decay of fossil grains, points to a Middle Subatlantic palynological age; the development of the carbic podzol of the oldest cycle (1S) could continue from the Preboreal till ±1200 AD.
The burial of the palaeopodzol took place after 1200 AD. This is still in line with the period of forest clear cutting. Time, available for the development of the micropodzol (2S), de stable period between the depositions of S2 and S3, was (based on OSL datings) maximal 130. The radiocarbon ages of the carbon fractions, extracted from the 2Ah (with the exception of FUL) look 'too old', but the OSL age of the burial of the micropodzol is fits with the heat degradation after the introduction of deep stable management.

Profile Defensiedijk 2, mono-cyclic haplic arenosol, overlying a bi-cyclic carbic podzol (fig.20-21, table 9), pollen diagram first published in Van Mourik, 1988
Profile Defensiedijk 2 is located 300 m NNE of Defensiedijk 1 near the border of a deflation plane and inland dunes. The palynological ages (post sedimentary pollen spectra with relatively high scores of Corylus in the oldest buried podzol, the decreasing scores of Corylus and the increasing scores of Alnus in the youngest buried podzol) and the 14 C ages (Bronze Age and Middle Ages) of the burial of the carbic podzols are comparable to Profile Defensiedijk 1. 3D is expressed by a 20 cm thick driftsand layer. The well preserved 14 C age of the 2S indicates that the driftsand deposits have been truncated. It is therefore impossible to distinguish between a mono or polycyclic driftsand package  www.intechopen.com

Palaeoecological information from the 14 C and OSL dated plaggic anthrosols
The introduction and extension of plaggen agriculture on chemical poor sandy soils has long been controversial in historical geography and soil science (Spek, 2004;Horsten, 1994, 1995). A complication was the interpretation of the validity of radiocarbon 14 C datings of organic extractions of plaggic deposits, due to the complexity of the composition of the soil organic matter , 2010. This contradiction is very well illustrated by the observation that Fagopyrum pollen is already present in the lowest samples of plaggic deposits with radiocarbon age over 1000 year.
Recently, the application of OSL dating in palaeopedology could clarify the genesis of plaggic anthrosols (Bokhorst et al., 2005, Van Mourik 2007Van Mourik et al., 2011a). Based on radiocarbon datings agricultural land use is registered from approximately 1000 BC but OSL datings show that the accumulation of true plaggic deposits did not start before approximately 1500 AC. An older organic matrix is suspended in the younger mineral skeleton of plaggic deposits (Van Mourik et al., 2011a,b). The organic matrix consists of aggregates with an internal fabric, related to the internal fabric of modeled excrements of humus inhabiting soil fauna, consisting of organic plasma and micro particles, charcoal and pollen grains (fig.24-25, Van Mourik, 1999b. Consequently, palynological information and radiocarbon datings from plaggic anthrosols may contribute to the reconstruction of the evolution of vegetation and land use, but not to the absolute dating of plaggic deposits.

Profile Tungelerakker, plaggic anthrosol, overlying a ploughed (plaggic) podzol (fig. 22-23; table 10), Pollen diagram first published in Van Mourik, 1992
The first investigated plaggic anthrosol near Weert was profile Tungelerakker, situated 1050 m east of the Tungeler Wallen site, offering an optimal possibility to compare the pollen zoning in the buried histic podzol and the zoning in a plaggic anthrosol. The scores of arboreal pollen in the post-sedimentary pollen spectra from the ploughed palaeopodzol (S2) are low, the scores of Ericaceae are maximal and pollen of Cerealia is present. Based on the 14 C HUM and HAC ages, this indicates agricultural activity on the site in the period ± 230-770 AD. This fits with the 14 C BULK ages of the buried histic horizon in profile Tungeler Wallen, ± 300 BC-1100 AD.  www.intechopen.com The 14 C ages of the plaggic deposits (2D) suggest plaggic deposition from ± 250-1400 AD. Application of OSL dating in more recently investigated plaggic deposits (Dijkerakker and Valenakker) indicate a contradiction between 14 C and OSL ages. It is proven that 14 C datings are valuable for the reconstruction of the agricultural development. Accurate dating of the plaggic deposits must be based on OLS ages. Palynologically, the Aan can be divided in two zones (2D1 and 2D2). 2D1 is characterized by high pollen densities, decreasing medium scores of Ericaceae and medium scores of Cerealia. The 2D2 is characterized by lower pollen densities, lower scores of Ericaceae and higher scores of Cerealia. This reflects probably the increasing crop production and the increasing content of mineral grains of the plaggic manure, related to the introduction of deep stable economy (Vera, 2011).   Table 11. 14 C and OSL datings of profile Dijkerakker

Profile
The skewness of the histogram of sample 47/2 indicates that the calculated age is too high, caused by palaeodose overestimation. The true palaeodose cannot be accurately deducted because it is not known whether the skewness is caused by a large amount of partly bleached grains, by a smaller amount of much older unbleached grains or by a combination of these two factors. The aliquot measurements can be a mix of the age of mineral (2D) www.intechopen.com grains, never again bleached after sedimentation and grains, bleached at the surface after ploughing. Additional, bioturbation can be responsible for vertical transport of 'older' grains from below and younger grains from above.  The histogram of sample 47/2 in fig. 5b indicates a reliable age, 394 ± 40 before 2001 (343 ± 40 BP). The majority of the mineral grains was completely bleached, promoted by the improvement of plough techniques, the increase of the mineral fraction of the plaggic manure and consequently the increase of the sedimentation rate.

Profile Valenakker, plaggic anthrosol, overlying a ploughed (plaggic) podzol (fig.27-31; table 12), pollen diagram first published in Van Mourik and Horsten, 1995
Profile Valenakker is well preserved in the urban environment of Weert on the sport field of a college and for more than 100 years never been ploughed.    Micromorphological observations show the complexity of soil organic matter in plaggic deposits. There are various sources of organic carbon as decomposing tissues of rooting plants, sods. The same is true for pollen spectra, a mix of the regular aeolian pollen influx and pollen, released from sods. The result is a plaggic deposit consisting of an older soil organic matrix, suspended in a younger mineral skeleton.
Traditionally, the origin of sods, used in plaggen agriculture, was reconstructed on the base of pollen spectra. The spectra of the Aan and the 2Abp show very low scores of arboreal trees but reasonable scores of Ericaceae and Poaceae. Ectorganic matter from forest soils is www.intechopen.com unlikely, Ericaceae pollen may indicate heat sods, Poaceae pollen on grassland sods, the combination on sods from degrading heath. It is very probably that during the formation of the 2Abp the farmers used mainly mowed heath shallow stables to produce manure for the arable land. During the formation of the Aan the collected sods with a higher mineral fraction to produce plaggic manure in deep stables (Vera, 2011). In fact, the origin of litter or sods cannot be satisfactorily detected with pollen diagrams. For that reason we applied biomarker analysis on two samples of the Aan and one sample of the 2Abp. In fig. 31 the n-alkane distribution patterns is shown in soil organic matter, extracted from soil samples, as well as in the leaves of Caluna vulgaris. While a complete biomarker reconstruction with the VERHIB model was not possible because of the limited samples numbers, a visual comparison of the patterns does yield some interesting information. The C 31 and C 33 n-alkanes dominated all three soil samples. Of the nine plant species considered (f 2.6) The leaves of Caluna vulgaris were the only to also have C 31 and C 33 n-alkanes as the dominant n-alkanes present. This, together with the uniformity of the n-alkane patterns at all depths sampled, is consistent with the history of plaggen agriculture on the profile and points to a dominant use of heath sods therein. It is also consistent with the importance of Ericaceae in the pollen spectra. At the same time, visual assessment of the n-alcohol patterns (data not shown) does not show a clear link to a single plant species. In this respect the limited number of samples with respect to the soil profile and the plant species considered are a limiting factor in these exploratory analyses. Nevertheless, the results have great potential for the use of biomarkers to help reconstruct organic matter input in plaggen soils in combination with other proxies.

Soil map around 2000 BC
The surface parent material exists of coversands which were deposited during Late Glacial and Preboreal time. Initial soil formation started without human interference and continued until Late Subboreal time. As a consequence, the soil map shows an undisturbed pattern of climax soils: umbric and gleyic podzols on coversand ridges, Siltic, umbric and histic gleysols in depressions and valley's. From the Bronze Age on, the effect of human land use on soil forming processes increases. The deciduous forest degraded partially into Calluna heath as the result of forest grazing, small scale wood cutting and shifting cultivation. In a further phase, deforestation accelerated and shifting cultivation was replaced by sedentary forms of agriculture during the Iron Age. The soil map reconstructed for 2000 BC therefore shows a soil distribution of only few soil types which is strongly adapted to local coversand topography.

Soil map around 1500 AC
People learned to collect ectorganic and mowed biomass to produce in the last remains of the forest and on the heath. Plaggen agriculture was probably locally introduced in the Early Middle Ages (Spek, 2004) and became the regular land use system on mineral poor sandy soils in the medieval period. As an effect, the removal of ectorganic matter triggered soil acidification, and replacement of umbric podzols by carbic podzols. Another effect in the landscape was the disturbance of surface and soil water hydrology. Less water was intercepted by forests, which enhanced soil water infiltration. In lower terrain, in-between coversand ridges such as small valleys and local depressions, umbric gleysols developed into histosols. The intense application of plaggic manure on arable land prevented extensive acidification, so that plaggic podzols became a common soil types. The soil map reconstructed for 1500AC shows a much more fragmented pattern as a consequence of this early human interference in the previous natural soil conditions.

Soil map around 1950 AC
An increasing population demanded intensification of food production. During the 11 th , 12 th and 13 th century clear cutting of the forested areas created an environmental catastrophe: extensive sand drifting (Vera, 2011). Deflation plains with gleyic arenosols and inland dune complexes with haplic arenosols were the results. After the introduction of the deep stable economy in the 18 th century, farmers collected additional to mowed biomass more and more heath sod, including the endorganic humic mineral horizon. The mineral fraction of the plaggic manure increased. Sod digging resulted in a first local degradation of the Calluna heath, and over time regional effects were noticeable, that even prevented the regeneration www.intechopen.com of the vegetation cover. In parts of the Late Glacial coversand landscape this triggered the extension of the typical Late Holocene driftsands. The consequence for soil formation was that plaggic podzols were replaced by plaggic anthrosols, mainly as the result of accumulation of the mineral compound of the plaggic manure. Degradation of the Calluna heath initiated the degradation of carbic podzols to xeromorphic and hydromorphic arenosols. The invention of chemical fertilizers around 1900 AD took away the need to collect sod manure for crop production. The information which is nowadays preserved in many polygenetic soil profiles proves that the resulting soil patterns are strongly controlled by human interference. The legend of the soil map of 1950 AD is based on the Dutch Soil Map, sheet 57-Escale 1:50,000 (Stichting voor Bodemkartering, 1972). The Dutch soil units are translated to the system of the World Reference Base (ISRIC/FAO, 2006).

Recent developments
Recent land management is further affecting soil development, which will continue to control the transformation of existing soil patterns. The former Calluna heath is partly reclaimed for arable land and partly turned into forest (dominated by Scotch pine). There is also a trend towards stabilization of driftsand areas by invasion and succession of vegetation. The reclamation of the Calluna Heath into arable land changed carbic podzols into anthric podzols. Under Scotch pine, mormoder humus forms developed (Sevink and De Waal, 2010) and haplic arenosols develop to albic arenosols. Peat accumulation was interrupted in the depressions, because soil water infiltration in the forest stands decreased and due to bio oxidation, histosols shift towards umbric gleysol.

Conclusions
Paleosols can be considered as important geo-ecological records, but due to the complexity of soil organic carbon, extracted from buried humic soil horizons, an accurate geochronology of such records cannot be based only on 14 C datings. The combination of OSL and radiocarbon datings enables the correlation of paleo-ecological data derived from histosols and mineral paleosols and improves the geochronology of the evolution of polycyclic soils and landforms in aeolian sands and plaggic deposits.
In palaeopodzols contradictions appear between OSL and 14 C ages. Soil micromorphological observations and the results of biomarker analysis show that the composition of soil organic carbon in buried humic horizons can be affected by secondary soil formation (buried podzols) and sedimentation (micropodzols).
In plaggic deposits radiocarbon datings reflect the development of an older organic matrix, suspended in a younger mineral skeleton. 14 C datings of the organic matrix are relevant for the reconstruction of the agricultural history, OSL datings of the mineral skeleton are relevant for the destination of the age of plaggic deposits.
Multi method analyses of polycyclic soil profiles provide the detailed knowledge which is necessary to fully understand time development of soil patterns in areas which are strongly affected by human land use. The combination of traditional soil survey techniques (soil classification, soil mapping), pollen analyses, micromorphology and soil dating techniques ( 14 C, OSL) makes it possible to date major changes in geo-ecological evolution. www.intechopen.com