551.31:551.311.2:552.54(492.945) MEDEDELINGEN LANDBOUWHOGESCHOOL WAGENINGEN • NEDERLAND • 80-8(1980)
SOIL CONDITIONS, SOIL CARBONATES AND FORMER VEGETATION IN THE GEUL VALLEY FROM GULPEN TO MEERSSEN (SOUTH LIMBURG, THE NETHERLANDS)
W.VA N DE WESTERINGH etal.
Departmentof Soil Science andGeology, Agricultural University, Wageningen, The Netherlands (received3-IV-80 )
H.VEENMA N & ZONEN B.V.-WAGENINGEN-1980
iüOL '•:.,:i TT
CONTENTS
1. PREFACE
2. W. VAN DE WESTERINGH. Soils and their geology in the Geul valley 1 Summary 20 Samenvatting 21
3. R. MIEDEMA. Incidence and distribution ofcarbonate s inth esoil so f the meandering river Geul 27 Summary 43 Samenvatting 44
4. A. J. HAVINGA and R. M. VAN DEN BERG VAN SAPAROEA. Former vegetation and sedimentation in the Geul valley 47 Summary 58 Samenvatting 58
ACKNOWLEDGEMENTS 60 PREFACE
During fiveyear s(1974-78 )a soi lsurve ywa sundertake n byth e Department ofSoi lScienc ean dGeolog yo fth eAgricultura lUniversit yo fWageninge ni nth e valley of the river Geul between Gulpen and Meerssen (South Limburg, The Netherlands).Th emai npurpos ewa st ointroduc eundergraduat estudent st osoi l surveypractice .Th ewor kconsiste d insurveying ,describin gan dmappin gsoils , making cross-sections and estimating the fluctuation of the water table with reference to soilcharacteristics . The results form a sound basis for a better understanding of the geology and soils in the Geul valley. Because of the very detailed field work (boring density 25 x 50m ornarrower) , much more information wasobtaine d than in thecas eo fa norma lsoi lsurvey .T ocomplet eth esurve ysom especia lstudie san d laboratory research were undertaken by the authors in co-operation with graduate students.Th e results are presented in the following three papers. Thefirst pape routline sth eHolocen estratigraph y ofth evalle yfill i nrelatio n to soil conditions. Five sedimentation stages have been distinguished which determined thesoils .Henc e thelegen d of thesoi lma p hasa marke d geological character.Th esoi lpatter nreflect s theyounge rsedimentatio n historyo fa smal l river,tha t occupies afairl y narrow valley. Theincidenc ean dpatter n ofdistributio n ofcarbonate s showsmarke d differ encesbetwee nth evariou ssoils .Ver ydetaile dfield studie swer eundertake nt oex plain the relationship between the presence of carbonates, the sedimentation regimeo f the river Geul, time-dependent decalcification and to the occurrence of fans along the valley border. The results of this investigation are given in thesecon d paper. During the soil survey it was assumed that a relationship exists between former sedimentation in the valley and former agricultural practices (deforest ation) in the surrounding upland, as the texture of the valley soils is often similar to that of the upland loess soils. Extensive and often thick colluvial deposits along the valley borders are evidence of another effect of human ac tivity. To obtain some insight into the age of the fluvial and colluvial deposits and theirrelationshi pt oagricultura l history,tw osoi lprofile s wereinvestigate d bymean s ofpolle n analysis.Th eresult sar e setou t inth e third paper. SOILS AND THEIR GEOLOGY IN THE GEUL VALLEY
W.va n deWestering h
CONTENTS
1. INTRODUCTION 1
2. GEOLOGY 4 2.1 Stage I 4 2.2 Stage II 4 2.3 Stage III 6 2.4 Stage IV 7 2.5 Stage V 8
3. SOIL CONDITION 9 3.1 Soils 9 3.2 Calcareous material 10 3.3 Hydrology 10 3.4 Soil map legend 11 3.5 Soil map 12 3.6Fan s 17
4. THE PRESENT LANDSCAPE 18
5. SUMMARY 20
6. SAMENVATTING 21
7. REFERENCES 24
1. INTRODUCTION
Therive rGeu risesl nea rEynatte n inBelgium .I tenter sTh eNetherland snea r Cottessenan dflows int oth eMeus enea rBunde .B ythe ni tha scovere dabou t5 6 km, two-thirds of this distance being in The Netherlands (Fig. 1). From the Dutch-Belgian frontier to Mechelen it flows in a south-north direction, from Mechelent oSchi no pGeu li trun snorth-wes t andfro m Schino pGeu lt oBund e in a more westerly direction. It receives many small tributaries. Together they form an essential part ofth epleasan t South Limburg landscape (Fig.2) . By Dutch standards the Geul is a fast flowing river. The slope varies from about 7.62m/k m near Epen to 1.33m/k m near Houthem. Theaverag ei sabou t 3m/km . Itsdischarg e isver yirregular . Near Meerssen theaverag edischarg ei s
Meded. Landbouwhogeschool Wageningen 80-8(1980) 1 K.j- /^""de / X boundary of HEERLEN vïf / s. Meerssen ^ ->(the river system # ' Rothem-iÇ^. Vroenhof , ^-_j., (^St.Gerlach / _^ Schaloe n f vauenDurg ^fC-Schin op Geul MAASTRICHT Genhoes/ >.Etenaken _,'' •i GERMANY
AKEN
BELGIUM \
HENRI-CHAPELLE
FIG. 1.Th e river system of the Geul.
Meded. Landbouwhogeschool Wageningen 80-8(1980) Meded. Landbouwhogeschool Wageningen 80-8(1980) about 2,3001/ s; th eminimu m discharge inrecen t yearswa s95 01/ s (September 1971),wit ha maximu m of24,2001/ si nOctobe r 1974.I nth ever ywe tperio do f December 1966th edischarg ea tValkenbur g wasabou t 65,0001/s (Information supplied byth e LimburgPubli cWork s Department). Underpresen tcondition s severefloodin g isrelativel yrare .Th elas tfloo d was inth esprin go f1974 . Ona naverag eflood s occurever ytw oyears .Th edischarg e by means ofregulation , hasbee n improved sotha t floods canb eprevente do r theirduratio nshortened .Locall ysmal larea sar emor efrequentl y flooded.Whe n floods occur in the late autumn, winter or early spring, there isusuall y little damage to agriculture, high water levels being generally of short duration. Beforeth emoder nregulatio no fth eGeu lth erive rha da natura lcharacter .Bu ti t should beadde d that for centuries thewate r discharges andlevel s have been influenced byman ,e.g .b ylan dreclamation ,th econstructio n ofwatermills ,etc . The geology and soil conditions of the Geul valley also reflect agricultural practicesi nth eSout h Limburgloes slandscape .
2. GEOLOGY
Fivedifferen t stagescoul db edistinguishe di nth ehistor yo fth evalle yfil li nth e Geulbasin .Th efirs t stagei so fpre-Holocen eage ,th eothe r four areHolocene .
2.1. STAGEi (FIG. 3a)
Theshap ean dth esiz eo fth eGeu lvalle yindicat ea Pleistocen eorigin .Durin g theformatio n ofth edissecte dplatea u landscapeth evalle yo fth eGeul ,a major tributary of the Meuse, was deepened and widened. As a result of various geological processes thecros s section ofth evalle yha sbecom e asymmetricali n manyplaces ,wit hon estee psid ean don egentl eslope .Smal lterrace-lik eirregula rities areoccasionall yfoun d alongth evalle ywalls .A tth een do fth ePleistocen e thedissecte d plateau landscape hasbee ncovere d byaeolia nloess . The valley was largely incised inUpper-Cretaceou s limestones. Inth e flood plain ofth epresen t valleysom emetre so fPleistocen e gravelwer elai d downb y braidedriver soverlyin gth eCretaceou smaterial .Holocen esediment shav ebee n deposited onto p ofit .Thes elas t depositsgiv eth evalle yit spresen t character.
2.2. STAGEi i (FIG. 3b)
The temperature rose atth ebeginnin g ofth e Holocene andinitiate d anew , more thermophilic type of vegetation (JANSSEN, 1960). This led to a forest vegetation inth e wholeSout h Limburg loessarea . Thenatura l situation ofth e
4 Meded. Landbouwhogeschool Wageningen 80-8(1980) cretaceous limestone a. Stage I. Pleistocene; deposits of a braided river, gravel. aeolian loess
|»°e°»| gravel
I« ». «I Peat
«] clay
T fluvial material , '-J fine-silty ffijgffl fluvial material, b. Stage II. Holocene; oldest deposits, tiöiä coarse-silty (loess-like) clay and peat. Ij-.-'.-çj colluvial material \:°'':i (loess-like) I' * *,| fluvial material, sandy
c. Stage III. Holocene; increased activity, fine-s1lty deposits.
i •i •I r r -r r 'T i •i •i -r
d. Stage IV. Holocene; increased activity, e. Stage V. Holocene; coarser-silty deposits. recent deposits, sandy.
FIG. 3.Stage si nth efilling o f thevalle yo fth eGeul .
Geulvalle yi nwoode dSout hLimbur gwa swette rtha na tpresen ta sa resul to fit s lowelevatio nwit hrespec tt oth esurroundin ghill san dth eoccurrenc eo fspring s along the valley at points where water rises over heavy, impermeable Lower- Cretaceouscla ylayer so rflow si nCretaceou slimestones .Th evegetatio nbecam e a marsh forest (HAVINGAan d VANDE N BERG VAN SAPAROEA,1980) . At the beginning of the Holocene, in such a densely forested landscape, the river mainly carried groundwater base flow that reached the valley through springs and seepages. In a wooded landscape when precipitation exceeds the évapotranspirationth esurplu swate rdoe sno treac hth erive rvi ath esurface ,bu t Meded. Landbouwhogeschool Wageningen80-8 (1980) through the soilan d after deeppercolatio n viath egroundwater . At thisperio d thedifferenc e between base flow and peak dischargeo fth erive rwa sno t great. Asth elandscap ewa scovere db yvegetatio nth esuspende dloa d ofth erive rwa s small.Ther ewa slittl eo rn osedimentation .Onl ysom ecla ywa sdeposited ,o rels e peat was formed.
2.3. STAGEm (FIG. 3C)
Aslon ga sth eentir eloes sare awa sunde rpermanen tvegetatio n thedischarg e and activity of the Geul remained more or lessunchanged . It wasonl y slightly affected byclimati cchange si n the Holocene.Th e situation was again changed whenth evegetatio nwa scleare db yma nan dbar esoi lbecam esubjec t toerosion . The first human settlers in South Limburg to clear some parts of the forest wereth e Dutch Bandkeramik people.The ylive do n theboundarie s ofth eleve l loessplateaus .The ywer efarmer s andthei rarabl elan dwa so nth eplateaus .Th e coarse-siltyto player so fth eloes ssoil sha da lowe rcla yconten ttha nth efine-silt y subsoils and were weakly acidic. This soil condition was conducive to arable farming, as sparse woodland was growing on loamy and weakly acidic soil (PONS, 1973). For reclamation and farming such situations were attractive because: - it wasrelativel y easy to clear the sparsewoodlan d and reclaim the land, - the original vegetation waseasie r to control than onclaye yan d richsoils , - tillage ofcoarse-silt y soilswa seasy . Owing to the situation of the arable land on plateaus and the sparse forests around the fields on the boundaries of the plateaus and on the hills, the hy- drologicalregim ewa sonl ysubjec t tosligh tchange .Erosio nwa sno tseriou san d mosto fth esurplu swate rstil lreache d therive rvi ath egroundwater .I nthi spar t ofSout hLimbur gn osettlement so fBandkerami k agehav ebee nfoun d (BAKELS, 1978),s otha t little if any silt reached the river and the sedimentation rate was low.Th esedimentatio n pattern inth eGeu lvalle yonl ydiffere d slightlyfro m the situation described in stage II.Thi swa schange d when larger parts of theloes s area,bot ho nth eplateau san do ngentl eslopes ,wer ecleare db yman .Th enatura l vegetationdisappeare d anderosio nstarted .I ti sno tknow nwhethe rerosio nwil l have become serious as early as from the beginning of intensive agricultural practise or later on (at theend?) . Itwa sdurin gth eRoma n eratha tlarg epart so fSout h Limburgwer euse d for arable farming. The clearance of the natural vegetation and resulting erosion then had amarke d effect on theregime so fth eGeu l and other rivers.I t caused that the hydrological regime of theloes sare a wasdrasticall y changed and that thesuspende d load increasedconsiderably . Upt o thisperio d thesurplu swate r entered the soil and when the soil was saturated, the water percolated to the groundwatervi awhic hi teventuall yreache dth eriver. Bu towin gt oth eclearanc e ofth evegetatio n inhill yarea sles swate r penetrated the soil;i n stead it flowed laterally down the slopes.Thi s meant that the surplus water took less time to
6 Meded. Landbouwhogeschool Wageningen 80-8(1980) reach the river, and that thewate r levelsan d discharges became higher and more irregular with great differences between base flow and peak discharge. Owing to the greater discharges the current velocity, turbulence and sediment load of the river Geul increased substantially. Soil particles eroded from the arable land to the lower parts of the landscape (colluvium) or to the riverstha t deposited themateria l in thevalley s (alluvium). Loess material, especially the coarse-silty top layers of the loess soils, is very liable to erosion. During heavy rainfall or thaw greater quantities of water and much silto floes sorigi n(i.e .wit ha hig hconten t ofsil tparticles )entere dth e river. This was transported and deposited elsewhere. The river's sedimentation pattern changed compared with that of stage II. Instead of clay and peat the river deposits now consisted of silty loess-like material, with a differentiation of texture from thelevee st o the basins. VAN DEN BROEKan d VAN DER MAREL(1964 )als orefe r toth eloes sorigi n ofth efluvia l soils of the Geul.
2.4. STAGE IV (FIG. 3d)
Later, when more vegetation had been cleared and more land was used for agriculture, esp. for arable, occasionally more water entered the rivers with greater frequency. As a result more water and silt flowed into the Geul; thus its current velocity and sediment load increased. It isth e same process as described in stage III but more intensively. Our borings and the work of TEUNISSEN VAN MANEN (1958)clearl y showed that most of thesoil shav e ahighe r clayconten t in thesubsoi ltha n inth euppe rpar t ofth esoil .Apparentl y twoimportan t sedimen tation stages should be distinguished. The latest stage must be the result of reclamations even more extensive than those which occurred in the Roman era (seeabove) .I t undoubtedly datesfro m the MiddleAge s (HAVINGAan d VAN DEN BERG VAN SAPAROEA, 1980). RIEZEBOS and SLOTBOOM (1978) also point to a relationship between soil erosion, sedimentation and human activity. In Luxemburg they found a change from peat to clayey material that could bedate d to the 15thcentury . Although a climatic reason also might be an important factor, the use of land for farming, especially the increase of the arable/pasture ratio,cause d an increased supply of sediment, combined with a rise in the water table or more frequent and extreme flooding byriver water.Thi s isi ncomplet e agreement withwha t wasfoun d in the Geulvalley .Th e occurrence ofman y soilswit h a finer-silty subsoilma y reflect a further deforestation or an increased arable/pasture ratio. The change from stage III to stage IV might be post-medieval. According to BRUNNACKER (1977) about 1000 years ago distinct changes occurred in sedimentation patterns and soil conditions in the valleys owing to a substantional change in river movements. Although slight climatic changes and oscillations of the glaciers in the Alps and of sea level might have been of importance to large rivers, small rivers and brooks were mainly influenced by
Meded. Landbouwhogeschool Wageningen 80-8(1980) 1 humaninterferenc ei nth elandscape .Deforestatio n ledt oerosio nan dth eerode d material accumulated within the smallvalleys .
2.5. STAGEv (FIG. 3e)
After the MiddleAge sth eentir eloes sare a wascleare d and cultivated except for the steepest hillsan d poorest soils,wher e the forests remained. Forests still exist on sites where General Tranchot mapped forests in the beginning of the 19thcentury .Thi si sa good illustration of thefac t that large-scale reclamation hasbecom eimpossibl ei nrecen tcenturies .Nor , ofcours edi d theriver schang e theircharacte r during this period. In recent times, farming practices have greatly improved (e.g. better weed control,ro wculture ,reallotment) . Otherwise thismean s that the arable land is exposed for longer periods, thus increasing erosion danger (BOLTe t al., 1980). Another important factor is that in the 20th century many roads have been paved, so that overland flow increased. Moreover households use more water thanbefor ean dal leffluen t waterreache sth eriver sthroug hsewerage .Thes etw o non-agricultural causesma yb eth echie ffactor s creatingth eirregula r discharge of theGeul . The net result is that the Geul sometimes receives more water than before. There is a possibility, however, that a different hydrological regime of the Meusei nforme r timesals oinfluence d theslop ean dcurren tvelocit yo fth eGeul . Under present conditions the river cuts into its former deposits and the recent deposits are at a lower level. It is not the silt but the sand fraction that pre dominatesi nth egrain-siz edistribution .Small ,distinc tterrace sar eforme d (Fig. 4).Th esand scontai n carbonates and areintercalate d withorgani cmatter . The
FIG. 4.Formatio n of smallterrace s and sandy deposits inrecen t times(cf . Fig.3e) . 8 Meded. Landbouwhogeschool Wageningen 80-8(1980) present stream has cut so deeply into its own deposits that the river channel has cutint oth ePleistocen e substratum ofth ebraide d river.It scurren tvelocit yi sno t so great that it has cut into Pleistocene coarse gravel bed. Most of the material (clay and silt) that reaches the river flows to the mouth of the Geul and comes into the Meuse. The recent deposits in the valley are sandy and no longer loess like.
3. SOIL CONDITION
3.1 SOILS
Most of the deposits in the Geul valley belong to stages HI and IV, both in horizontal extent and in vertical depth or soil profile. As described in stages III and IV, the more recent loamy deposits of stage IV have a lower clay content than the loamy deposits of stage III. In many soils the texture trend increases withdepth . TEUNISSEN VANMANE N(1958) , BRETELER(1967) , DAMOISEAUX (1968) and VAN DE WESTERINGH (1979, 1980)recor d the same trend for the deposits of theGeu lan d itstributarie s theGul p and Eyserbeek. Buta differenc e existsi n the thickness of thecoarse-silt ypar t (e.g.textur e L)overlyin g the fine-silty part (e.g. M) of the soil profiles. In the Lower Rhine area in Germany it was also ascertained that the older Holocene river deposits have a finer texture than the younger river deposits. In contrast with the situation in the Geul valley several river terraces, all dating from the Holocene, are found at various levels in the Lower Rhine area (BRUNNACKER, 1978). In general themor e recent deposits of stageI Var e thicker than the older ones. In the levee position the later stage is thicker than in the basin position. The texture of recent deposits also differs in the leveet o basin direction; in the basin position the texture is finer. Three classes of soil texture have been distinguished, viz. L, M and Z. These three classes record the clay content (lutum %) of the loamy (loess-like) parent material. They correspond to the soil textural classes of the SOIL TAXONOMY (1975): L : approx. 10-20% lutum ('coarse-silty') M: approx. 20-30% lutum ('fine-silty') Z :approx . more than 30% lutum and V: pea t or peaty material. Ingenera lth eleve esoil shav ea textura lclas sL ove rgrea tdepth ,viz .exceedin g 120c m and sometimes 220cm . Soilssituate d at the boundary between levee and basin havetextura l classL ove r M (coded LM),viz .withi n a depth of 120c m the texture increases from L to M. Such soils show the two stages of geogenesis. An increasing textural trend was also found in many other situations. In the basin
Meded. Landbouwhogeschool Wageningen 80-8(1980) 9 positions soils are found with textures of M, MZ, Z or ZV. These not only indicate the basin positions but the two stages of geogenesis.
3.2. CALCAREOUS MATERIAL
Hitherto no reference has been made to calcareous material in the deposits of the Geul. As stated above, the river sedimentation had been correlated with the clearance of the vegetation and the reclamation of the loess area. When such activities led to soil erosion, loess-like material came into the river and was deposited. Hence sedimentation is a function of erosion and the carbonate content of deposits must be related to erosion. Thenatura l soili nth eloes sare acovere d byfores t wasa 'brik ' soil(D E BAKKER and SCHELLING, 1966) or alfisol (SOIL TAXONOMY, 1975). Such a soil had no carbonates and wasweakl yacidi ct o adept h of2. 5t o 3.0m .I n other words, non- calcareous loess material was the first to erode. The carbonate content of the deposits was determined by changes in the sources of eroded material and by the sedimentation pattern in the valley. Decalcification isanothe r process affecting the carbonate content of the soils of the Geul valley. Further information on the carbonates in these soils has been provided by MIEDEMA (1980).
3.3. HYDROLOGY
The Dutch system of 'Gt' (= water-table classes) for characterizing the fluc tuations of the groundwater is relatively unimportant for such a small valley as that of the Geul. The groundwater level in the valley largely depends on the height of the river water and this height may vary extensively. Mottling in the soils are evidences of the water fluctuations (reduction spots of a more grayish tint and brown or orange-brown oxidation spots). The groundwater fluctuations of profiles with a deep homogeneous brown colour without mottles are characterized by a high Gt class (VII), i.e. the mean groundwater is always deeper than 1.20 m. Soils mottled at shallow depth but with a mean lowest groundwater table deeper than 1.20 m have wide ground water fluctuations and are characterized by Gt V. Soils with higher ground water levels (both the mean highest and the mean lowest) are characterized by Gt classes lower than V. They are mottled at shallower depth. The simplified map of groundwater fluctuations in the Geul valley (Fig. 5) showsthre emai n zoneso fG t classes: asmal lzon ewit hhig hG t classesalon g the Geul,a transitiona lzon ewit hG tclasse so fwid egroundwate r fluctuations, and a zone with low Gt classes.Th e first zonecorrespond s to levees, the last to basins. Table I shows the Dutch system of water-table classes (Gt classes).
10 Meded. Landbouwhogeschool Wageningen 80-8 (1980) Valkenburg
VIZ> 40-80 vit cm b.s. >120 cm below surface
1) non-present
Z) together
boundary between fluvial and colluvial soils
disturbed Stokhem * • Wij Ire FIG. 5.Simplifie d estimated groundwaterfluctuation ('Gt' )ma po fth evalle yo fth erive rGeu l from Wijlre-Stokhem to Valkenburg.
TABLEi
Gt1 MHW2 MLW'
I _ <50 II - 50-80 III <40 80-120 IV >40 80-120 V <40 > 120 1 Gt: water-table class VI 40-80 > 120 1MHW : mean highest water table VII >80 >120 'MLW: mean lowest water table (inc mbelo w surface) (cf. VAN HEESEN,1970 )
3.4. SOILMA P LEGEND
Threemai nclasse shav ebee ndistinguishe d inth etextur eo fth efluvial soil si n theGeu lvalley ,viz .L ,M an d Z,indicatin g thelutu m orcla ypercentag e (= % < 2micron) ,i.e .respectivel yc a 10-20% , ca20-3 0% an dmor etha n3 0% lutu m (see 3.1). There is also a class Vindicatin g organic matter contents of over ca 15% (peat ymateria lan dpeat) .Generall y speaking,th etextur eo fth esoil si ssil t loami nwhic hth ecla yconten tvaries .I ti sderive dfro m loesssoil swit ha sil tloa m texture.Soi ltexture scoarse r than L,i.e .wit hles stha n 10%clay , onlyoccu ri n Meded. Landbouwhogeschool Wageningen 80-8(1980) 11 themos trecen tdeposits .Thes ehav ea sand ytextur einstea do fa loam yone . As thesedeposit scove r sosmal l an area theycoul d not beshow n on the soilmap . Thema punit s (Fig.6 )indicat eth esoi l texturalclasse san d combinations of classest oa dept ho f12 0c mo fth esoils .Figur e7 showsth etextur eprofile s toa deptho f22 0c mo fsom epart so fth eGeu lvalle ydownstrea mfro m Valkenburg. Although a change in textural class is shown onth efield map s bymean so f numeralsindicatin git sbeginnin g(e.g .1,2 ,3,4, 5an d6 respectivel ydenotin git s beginningu pt o4 0cm ,fro m 40t o8 0cm ,fro m 80t o 120cm ,fro m 120t o 160cm , from 160t o20 0c man d from 200t o22 0cm) ,thes enumeral sha d tob eomitte d from Figure 6becaus eo fth escale .
3.5. SOILMA P
(western part)
L. coarse-si 1ty soils to 1.20m
LMg+LMj. coarse-silty soils with a fine-silty subsoil, beginning deeper than 0.40m
LML. item LM„+LM3, but with a coarse-silty subsoil, beginning deeper than 0.80m LMZ+MZ. item LM.+LM,, but with a clayey subsoil; the topsoil is occasionally fine-sil ty V. fine-silty soils with a subsoil of peat or peaty clay, beginning within 1.20m .••••• boundary between fluvial and colluvial soils 1 disturbed n.s. not surveyed FIG. 6.Simplifie d soilma pt o 1.20m o f the Geulvalle yfro m Gulpen toMeerssen .
12 Meded. Landbouwhogeschool Wageningen 80-8(1980) (eastern part)
Meded. Landbouwhogeschool Wageningen 80-8(1980) 13 Meerssen Fig. 6. location map
Valkenburg
Partij
Figure 6 shows the soil map of the Geul valley from Gulpen to Meerssen, (borings to 120cm) . Figure 7present s the soil map, based on borings to 220 cm, for part ofth evalle ydownstrea m from Valkenburg. Somegenera linformatio n is given first. Soilsi nth eGeu lvalle ywit ha leve lo r nearlyleve lsurfac e are regarded asfluvia l soils.Soil sa t the boundary ofth evalle y beneath slopes,an d at a higher level than the surrounding fluvial soils, are regarded as colluvial soils. As both fluvial and colluvial soils consist of loess-like material it isdifficul t differentiate them.Wher eth eGeu lha sdeposite d fluvialmateria l thevalle yi snarrow , varying from about 150m nea r Etenaken and Strucht toabou t 600m nea r Meerssen. The boundary between fluvial and colluvial soils isver y irregular, especially in front of tributary-valleys (cf. 3.6). Oneo fth emos t noticeable features ofth esoi lma p (Fig.6 )i stha t at practically every point ofth e valley therive r has only formed onemeande r belt correspond ing to coarse-silty soils (unit L).I n general Figure 7show s that coarse-silty soils (unit L) occur at the same places in the subsoil as in the upper part of the soils. Thelevee so n both sideso fth e Geul arenarrow . Thetota l width of soilso funi t L variesfro m aminimu m of 50m t o about 350m wher e somebranche s ofth e Geul run side-by-side near the castles of Genhoes and Schaloen and the Meerssen watermill. Another noticeable feature isth e comparative rarity of clayey basin soils and peaty soils,th e main reason being the narrowness of the vally.Claye y and peaty
14 Meded. Landbouwhogeschool Wageningen 80-8(1980) Valkenburg
LH. coarse-silty soils with a fine-silty subsoil, mostly beginning deeper than 0.80 m and extending to 2.20 m LHL. item LH, but with a coarse-silty subsoil, beginning within 2.20 m
LMZ.. item LM, but with a clayey subsoil beginning deeper than 1.20 m & v. fine-silty soils, sometimes with a coarse-silty toplayer, with a subsoil of peaty clay, beginning within 1.20 m v . fine-silty soils, sometimes with a coarse-silty toplayer, with a subsoil of peaty clay, beginning deeper than 1.20 m 500 m
•••••• boundary between fluvial and colluvial soils n.s. not surveyed
FIG. 7. Simplified soil map to 2.20 m of the Geul valley between Valkenburg and Vroenhof. soilsoccurrin ga tth esurfac e (unitswit htextura lclasse s(M) ,Z an dV )ar ealway s situated on the other sideo f thevalle ya swher e therive r flows. A typical feature of most map units is that the clay content increases with depth(Tabl eII) .I ti sonl yth ecoarse-silt ysoil so fth elevee stha td ono tsho wthi s increase(Tabl eII) .Th eincreas ei ncla ywit hdept hcorrespond st oth egeogenesi s of the Geul deposits. The subsoil is formed by material of the first stages, the upperpar to fth esoi lb ymateria lo flate rdeposit swit ha coarse rtextur e(se e3.2) . It was shown by cross-sections made in various parts of the valley that the surface level of a peaty soil in the basin position is often higher than that ofa levee.A t such sites river sidimentation wasclearl y impossible. In the basins or backswamp sfluvia l sedimentation mightonl yhav ebee npossibl eafte r thepea t haderoded ,o rels ea sa resul to fimprove dnatura lwate rdischarg eo fth erive ro r artificialdrainag eo fth evalley .I twa sonl ya tcertai npoint so nth eborder so fth e valleynea rslope swher espring soccurre dtha tgoo ddrainag ewa simpossibl ean d peat wasunabl e to settle. With theexceptio n ofth ema puni t Lo fth elevees ,al lother s aremor eo rles s complex,dependin go ntextura lchange si nth esoi lprofile s andth ema pscale . In somearea sma punit sa s(M) Zan d (M)ZVar erelated ,a sthe yindicat eth ebasi n position ofth evalley . Mapunit s as LM, LML,etc. ,denot e thetransitio n from levee to basin. These soils show the geogenesis of the two main stages of the fluvial soils.Th emos trecen tan duppe rpar to fth elevee s(L )ha sbee ndeposite d on a former basin situation (M), or with the same genesis the basin was early situated in aleve esituatio n (L),a sfoun d inth e oldest leveesubsoil . The soilma p of Figure 6show s a difference in soilcondition s upstream and Meded. Landbouwhogeschool Wageningen 80-8(1980) 15 TABLE II. Soil textureso f two profiles. Profile Schin op Geul II, a basin soil (unit MV). horizon depth particlesiz e distribution pH organic % CaCo3 (cm) < 2 um 2-15 urn 16-50 um > 50 um KCl H20 matter (Wese- %c mael)
Al 0-8 22.7 22.3 48.4 6.6 6.90 7.1 4.56 2.5 Al 8-16 22.9 21.9 48.6 6.6 6.97 7.1 3.36 3.1 A3 16-26 20.8 20.4 53.1 5.7 6.95 7.2 1.55 4.2 A3 26-37 20.4 23.1 54.2 2.3 7.11 7.3 0.81 0.7 B2 37-50 20.5 21.3 55.5 2.7 7.03 7.2 0.95 1.4 B2 50-62 29.4 29.6 39.9 1.1 6.77 7.2 2.05 0.4 HCl 62-80 44.6 29.9 24.3 1.2 6.35 7.0 4.31 0.3 IIC2 + 80 30.2 28.9 34.6 6.3 5.87 6.0 22.7 1.8
Profile Schin op Geul III, a levee soil (unit L). horizon depth particle sizedistributio n pH organic 7„CaC03 (cm) < 2 um 2-16 um 16-50 um > 50 um KCl H20 matter (Wese- %c mael)
Al 0- 5 12.8 10.7 51.9 24.6 6.34 6.2 4.36 0.2 Al 5-10 12.4 10.2 52.1 25.3 6.53 6.3 3.46 0.1 Al 10-20 12.0 10.1 51.0 26.9 6.73 7.0 3.33 0.4 A3 20-35 13.0 10.7 58.1 18.2 6.10 6.6 1.32 0 B2 35-50 13.4 12.0 58.8 15.8 6.57 6.9 0.90 0 B2 50-65 13.6 12.0 59.4 15.0 7.05 7.4 0.47 0.3 B2 65-80 12.3 11.3 60.7 15.7 7.06 7.4 0.53 0.2 B3 80-95 11.4 9.9 61.4 17.3 6.67 7.1 0.43 0 B3 95-110 10.2 8.4 63.0 18.4 7.37 7.6 0.29 1.0
downstream from Valkenburg. Basinswit h Fine-siltyan d clayey soils(M ,Z )o r peat(V) ,wit ho rwithou tsom ecoarse-silt ymateria li nth euppe rpar to fth esoil , are only welldevelope d upstream from Valkenburg. Downstream they do not occuro rar esmall .Coarse-silt ysoil spredominate ,e.g .mor eL an d LM2 ofLM 3 than upstream where more LMi and LM2 occur. The reason for this pheno menoni sno tclear .I tcoul db ea geologica lon ea sa differenc e inth evalle yfill o n bothside so fValkenburg ,o rth eeffec t ofth eMeus eo nth esedimentatio npatter n of the Geul may not have extended beyond Valkenburg. It is not known how changes in the course of the Meuse and Geul may have affected the latter's sedimentation pattern. BECKERS(1929 )reporte da numbe ro fconsiderabl echan ges.I twoul dsee munlikel ytha t thisdifferenc e insoi lcondition swa sconnecte d witha differen t deforestation or occupation pattern upstream and downstream from Valkenburg. Although theGeu lha s onlyforme d asingl emeande r belt,downstrea m from Valkenburg it waspossibl et osurve ya nol drive rsystem . Particularlyo nth elef t of the valley between Geulhem and Rothem an old system is present in the 16 Meded. Landbouwhogeschool Wageningen 80-8(1980) subsoil (Fig. 7).Thi s system is also clearly indicated by soil carbonates (MIE- DEMA, 1980). Near Meerssen the Geul has three branches which must be connected with watermills. Thesoi l conditions and soil carbonates of the branches, the Grote (Great) Geul andth eKlein e (Small) Geul show that fluvial sedimentation must befairl y recent. The southern branch, the Geulke(littl e Geul)i sa millrac e which has scarcely ledt oan yfluvia l sedimentation (cf. thesoi l map).
3.6. FANS
The border between thefluvia l soils of the Geul andth ecolluvia l soils along the valley slopes bends towards thevalle y axis at points where tributary-valleys join themai n valley(Fig . 6). Infron t ofthes etributary-valley s fanshape d bodies protrude into themai n valley. The deposits ofwhic h thefan s aremade , usually aredifferen t from those ofth eadjacen t Geul sediments.Clearl y the sediments of these fans have been supplied through tributary-valleys. Although the deposits have aloess-lik e texture, they are different from the fluvial deposits ofth e Geul. Thisi smos t striking where clayey soilso fth e main valley arei ncontac t withth e outer fringes ofth e fans. Insuc hcase sth efa ndeposit s havea coarse r texture due to the sudden changei nth eslop eo fth etributary-valley , wherei tenter sth e plain of the main valley. The different origin of the parent material is also revealed when masses of calcareous material have been eroded in the catchment area of the tributary- valley; there thefa n deposits have calcareous soils. Carbonate detritus derived from the tributary-valleys has also affected the fluvial soils of the Geul (MIE- DEMA, 1980). It isclea r from the shape of some fans that man wasresponsible . They dono t have usual round shape, buta straight projection onth eroundin g (Fig. 8). The morphology ofsom e ofth efan s givesevidenc e of human activity. Instead ofth e
colluvium or other non-fluvial material
straight projection on the base
colluvium or other non-fluvial material FIG. 8.Sketche so fa fa ni nth eGeu lvalley . Meded. Landbouwhogeschool Wageningen 80-8(1980) 17 usual semi-circular form, a straight extension is present, that probably is a last remnant ofa wate rcours e over the fan. The function ofsuc ha ditc hwa san d still ist o drain the farm land. During most of the year, however, the ditch isdry . The sediment, aggraded in the ditch is removed by the farmers and thrown over the sides.Thi sspoi lform s akin d ofman-mad e artificial levee,preventin gth e surplus waterrunnin g out ofth editc h toofast . Soa sa resul to fth editch ,th elas t remnant of a small tributary in the triburaty-valley, and by farmers' activities in cleaning the ditch, some fans have a straight projection instead of a round shape. Exam ples are found near Stokhem, Etenaken, in front of the Gerendal and near Vroenhof. The fans are of recent date and of the same origin" as the Geul deposits, viz. clearance of the forests for farming. The centre of the fan may be old where it connects with thepoin t at which a tributary enters the Geul,bu t theextensio n of the fan is recent. At some sites the border of the fan overlies peat which means that the genesiso fth efans , liketha t ofth eGeu l deposits,i srelate d to the erosion occurring after the deforestation of the loess area (HAVINGA and VAN DEN BERG VAN SAPAROEA, 1980).
FIG. 9.Fa n infron t ofa dr yvalley .
4. THE PRESENT LANDSCAPE
Only small and altered remnants have been left of the original marsh forest in the Geul valley. At present the valley consists of farmland, mainly pasture. In 18 Meded. Landbouwhogeschool Wageningen 80-8(1980) earlier centuries, when the drainage wasles sefficient , a wide area of land was used as meadows. Names with the suffix 'beemd' ('meadow') are a relico f this agriculturallan duse .Marsh ylan dadjoine d themeadows ,a sindicate db yplace - nameswit h the suffix 'broek' ('brook'). The difference between the levees and the basins is hardly visible in the landscape. In some places differences in kinds of grasses and herbs show the differences in water-table classes (Gt) of thelevee s and the basins. This givesa goodindicatio no fth ehydrologica lcondition ,a tleas twher epasture sar eno tto o intensively exploited.
FIG. 10.Th eattractiv elandscap eo fth eGeu lvalle y: pastures ,trees ,pollards ,th erive rGeu l(1. )an da millrace(r.) .
TheGeu lvalle ylandscap ei sver yattractiv e(Fig . 10).I ti sa pastora l landscape andth eman ypoplar san dwillow salon gth eGeu lan di nlanes ,pollard s(willow , alders, ash, hornbeam, etc.), overgrown alder and hawthorn hedges along the meadowsan dpath sgiv ei ta specia lcharacter .Severa lspecie so fbird sfind a goo d biotopei nthi slandscape . Therecen terosio na tth eoute rbend san dsedimentatio ni nth einne rbend sca n besee n along the borders of the Geul.Th emos t recent deposits arelowe r than the older river banks. The vegetation clearly responds to this habitat which is richi ncarbonate san dnitrogen .An di ti sonl yfai rt oad dtha trive rpollutio nan d pollutiono fth eenvironmen t andth erive rbank sb yhuma nrefus e createa grea t problem for landscape management. Meded. Landbouwhogeschool Wageningen 80-8(1980) 19 Villagesan d otherbuilding sar efoun d onth ehill salon gth evalley .Ther ear e fewbuilding s in theactua l valley,e.g . afe w farms, castles,water-mill san d tea houses Unfortunately recenthousebuildin g (Schino pGeul )an d ane wmotor way(nea r Meerssen)hav eupse t thecharacte r ofth evalley . Asth e valley isimportan t for farming and recreation in South Limburg,w e must be on guard against any undesirable development that could destroy the valley landscape. It should also be remembered that the pleasant valley land scape can only be saved for the future provided agriculture continues to remain itsessentia l feature.
5. SUMMARY
ByDutc hstandard sth eGeul ,a majo r tributaryo fth eMeuse ,i sa fast-flowing river.Unde r present conditions thesedimentatio n byth eGeu li sdifferen t from the sedimentation under conditions when the loess landscape was under per manent vegetation. Fivestage si nth egeogenesi so fth evalle yfil lcoul db edistinguished . The first stage is of pre-Holocene age, the other four are Holocene. The Pleistocene braidedriver sha dlai ddow ngrave llayer soverlyin gCretaceou smaterial .I nth e beginning of the Holocene the Geul had a low activity. In the densely forested loesslandscap e therive rmainl ycarrie d groundwater baseflo w; th e suspended loadwa ssmall .Onl ysom ecla ywa sdeposited , or elsepea t was formed. This situation changed when the vegetation wascleare d by man for agricul ture. Bare soilbecam e subject to erosion. Instead of penetrating the soil,mor e surplus water flowed laterally down the slopes. It caused higher and more irregular water levels and discharges of the Geul. The current velocity and sedimentloa d ofth erive rsubstantiall y increased. Owingt oth ecoarse-silt yan d weak acidic top layers the loess soils were very liable to erosion. The river deposits nowconsiste d of siltyloess-lik ematerial . Mostsoil shav ea textura ltren dtha tincreas ewit hdepth .Thi sincreas ei ncla y contentcorrespond st otw omai nstage si ndeforestatio n andresultin gerosion .I n thispar to fSout hLimbur gn osettlement so fBandkerami k agehav ebee n found. Severeerosio n occurred in or after the Roman era and the Middle Ages,whe n largepart s ofth e South Limburgloes slandscap e weredeforeste d and used for arable farming. Inrecen ttime sth eGeu lsometime sreceive smor ewate rtha n before, asresul t ofimprove dfarmin g practices,pave droad san deffluen t waterfro m households through sewerage.Unde r present conditions theGeu lha sa nincrease d current velocity; th erive rcut sint oit sforme rdeposits .Th erecen tdeposit sar ea ta lowe r level.San d predominates. Most soils in the Geul valley have a coarse-silty top layer and a fine-silty subsoil.Th elegen do fth esoi lma pi sbase do nthre etextura lclasse s: L,M an dZ
20 Meded. Landbouwhogeschool IVageningen 80-8 (1980) (respectivelyc a 10-20% ,c a20-3 0% an dmor etha n 30% lutu mo rclay) ,an dV (peatymateria l and peat, i.e.mor e than ca 15% organicmatter) . There is a relation between soil types and soil carbonates. Groundwater fluctuations havebee ndistinguishe di nG tclasses . Leveesoil shav ehig hG t(VII) , basin soilslowe r Gt's (Van dlower) . The soil maps show a single meander belt with coarse-silty levee soils (L), clayeyan dpeat ybasi nsoil s(M ,Z an dV )an dsoil si ntransitio n(LM ,LML) .Th e surfaceleve lo fpeat ysoil ssometime si shighe rtha ntha to fa levee .A tsite swher e springs occurred good drainage wasimpossibl e and peat wasunabl et o settle. Thesoi lma pshow sa differenc e insoi lcondition supstrea m and downstream from Valkenburg. Upstream from Valkenburg basin soils are well developed, downstream from Valkenburg a coarse-silty layer is overlying deposits with other textural classes.Th e reason for thisphenomeno n isno t clear. Anoticeabl efeatur e ofth evalle yar eth efan si nfron t oftributary-valleys .Th e border between thefluvia l and thecolluvia l soilsbend stoward s thevalle yaxis . Fromth eshap eo fsom efan s iti sclea rtha tma nwa sresponsible .Instea d ofth e usualsemi-circula rfor mo fa fan ,a straigh tprojectio ni spresent ,tha tprobabl yi s thelas tremnan to fa wate rcours eove rth efan . Thefunctio n ofsuc ha ditc hwa s andstil li st odrai nth efar mland .Th efarmer sremove dth eaggrade dsedimen ti n theditc han d throwed overth esides . Thepresen t landscape isa landscape ofpastures , treesan d pollards,hedges , etc.I nsom eplace ssevera lkind so fgrasse san dherb sstil lindicat eth ehydrologi - calconditions . Birdsfin d agoo d biotope in thislandscape . Thevalle yi simportan t not onlyfo r farming, but alsofo r recreation. Unfor tunately recentdevelopment sa shousebuilding ,a ne wmotorwa yan d pollution, haveupse t the character of the valley.Th e pleasant character ofth e valley has mainly been created byagricultura l use and it can only be saved for the future provided agriculture continues to remain itsessentia l feature.
6. SAMENVATTING
GEOLOGIE EN BODEMGESTELDHEID VAN HET GEULDAL
De Geul ontspringt bij Eynatten in België. De Geul heeft een lengte van ongeveer5 6 km,waarva n 2/3dee li nNederlan d (fig. 1).D eGeu li see nsnelstro menderivier ,me tee ngemiddel d verhangva n3 m/k m (minimum 1,33m/k m bij Houthem,maximu m7,6 2m/k mbi jEpen) . Hetgemiddel ddebie tva nd eGeu lbi j Meerssen bedraagt ongeveer 2300l/se c(minimu m 960l/sec , maximum 65.000 l/sec).1 Overstromingen van lage terreinen komen nog regelmatig voor. Grote overstromingen van het geheleda lzij n zeldzamer.
1 Gegevens van de Provinciale Waterstaat van Limburg.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 21 Uit de bodemkartering en geologische doorsneden is een goed beeld van de Holocene opvulling van het Geuldal verkregen. De geologische opbouw en de bodemgesteldheid weerspiegelenduidelij k degevolge nva nd eontginnin ge nhe t agrarisch landgebruik (ontbossing, erosie) in dit deel van het Zuidlimburgse lössgebied. Ind egeologisch eopbou w kunnen vijf fasen worden onderscheiden. Infas e I (fig. 3a) valt de laat-Pleistocene dalopvulling met grof materiaal, afgezet door eenverwilderd e rivier. Fase II (fig. 3b)val t in het oudere Holoceen, toen Zuid-Limburg nog geheel metbo swa sbedekt .I ndi etij dha dd eGeu lee nvee lkleine re nregelmatige rdebie t envoerd evoora lbronwate r af. Erwer dweini gmineraa lmateriaa lmeegevoer d; desedimentati e beperkte zichto t zwareklei .Oo kvee n hoopte zichop . Infas e III (fig. 3c)tra d eree nveranderin g ind eaktivitei t vand eGeu lop ,di e eengevol gwa sva nee nveranderen d leefpatroon vand emens .D emen swa sva n jagen en vissen en verzamelen van voedsel overgegaan op landbouw, wat ge paard gingme t ontbossing. Hiervan waserosi e het gevolg. De eerste landbou wers,d eBandkeramisch e bewonersva n Zuid-Limburg, troffen opd elössgron - denme thu noorspronkelijk e lichtee nzwakzur ebovengronde n een betrekkelijk lichtbo saan .Zowe ld eaar dva nd ebegroeiin gal sd ebodemkundig eeigenschap penva nd elössgronde nbetekende nee ngunstig esituati evoo rd evestigin gva nd e landbouw. Doordat de agrarische aktiviteiten van de Bandkeramische bewoners zich vooral op de plateaus afspeelden, terwijl de hellingen met bos bedekt bleven, bleefd e(geringe )erosi ee ndaarme e ook desedimentati e door deGeu l beperkt. Degrot eBandkeramisch e woonplaatsen warenve rbuite n het Geuldalgelegen . In die tijd zal het oorspronkelijke, zeer rustige karakter van de Geul dan ook maarweini g kunnen zijn veranderd. Dit werd anders in de Romeinse tijd toen grote delen van het Zuidlimburgse lössgebied werden ontbost. Erosie trad in versterkte mate op,d eriviere n - ook deGeu l- kregenmee r water ensli ba f te voerene ne rvon dvee lsedimentati eplaats .He ttransporteren d vermogenva nd e Geulwa szozee r toegenomen, dat er in plaats van zware kleie n veen voortaan afzettingen vanee nlichter etextuur ,voornamelij k lössachtigva nsamenstelling , sedimenteerden. Het isi n het algemeen nog onzeker ofd esterkst eerosi e inhe t begin of aan het eind van eenbelangrijk e bewoningsperiode plaatsvindt. Ook in de Middeleeuwen trad er een sterke erosie op. De Geul kreeg nu periodiek zelfsno gmee rwate r aft evoere ne nha d ookhoger e stroomsnelheden dan voorheen. Dit leidde tot een versterkte opslibbing met lichtere sedimenten (faseIV fig., 3d) . Veelbodemprofiele n bezittenee ntextuuropbou wva nlicht-op - zwaar, zodat er twee sedimentatiefasen moeten zijn geweest. Dit moet weer in verbandstaa nme ttwe ebelangrijk ebewoningsperiode nwaari nlandbouwaktivi - teitenme the toptrede nva nerosi eplaatsvonden .I nd etij dn ad eMiddeleeuwe ni s deintensitei t vansedimentati e nogverde rtoegenome ntengevolg eva nd evoort gaandeontbossing ,d etoegenome nontwaterin ge nd evergrot ewaterafvoe r (fase V, fig. 3e). De Geul sneed zich nu langzamerhand in tot in de Pleistocene ondergrond. Dejongst e afzettingen zijn zandige nkalkhouden d enligge nlage r
22 Meded. Landbouwhogeschool Wageningen 80-8( 1980) dan voorheen (fig.4) . In het lössachtige sediment zijn drie textuurklassen onderscheiden metresp . 10-20%, 20-30% en meer dan 30%lutum , die op de kaart en hieronder zijn aangeduid met L, M en Z. Venig materiaal of veen is weergegeven als V. De oeverwallen hebben als textuurprofiel L tot 120 cm, vaak ook tot 220 cm. Bodemprofielen buiten de oeverwallen vertonen naar onderen meestal eentoe name in textuur. In de komgebieden komt in het textuurprofiel de tweeledige opbouw alsgevol gva n deverschillend e erosie-e nsedimentatie-intensitei t ind e opelkaa r volgendekultuurperiode n fraai tot uiting. Voor zover de bodemprofielen kalk bevatten, zijn dit voornamelijk bodem profielen in oeverwalpositie met een lichte textuur (L). Fig. 5geef t voor een gedeelte van het Geuldal een vereenvoudigde grondwaterfluktuatiekaart ('Gt'- kaart) weer. Het beeld op deze kaart vertoont een nauwe samenhang met de aanwezigheid van oeverwallen enkomme n opd ebodemkaar t (fig.6) . Debodemkaar ti sgebaseer do pd eonderscheidin gva nd etextuurklasse n L, M enZ e no pd eaanwezighei d vanvee n(V) ,di talle sto t opee ndiept eva n 120c m (Voorfig. 7 to t opee ndiept e van 220cm) .Zowe ld efluviatiel e alsd ecolluvial e gronden hebben een lössachtige textuur. Alsgren s tussen beide isgenome n de overgang van een vlak gelegen naar een hellende bodem. Deze grens verloopt onregelmatig, vooral bij diezi jdale n waarvoor eencolluvial e lobligt . Sommige vandez elobben ,zoal sdi ebi jVroenhof , bezittenee n merkwaardig smaluitsteekse l (fig. 8).Di t isontstaa n vooral alsgevol gva n agrarisch gebruik om het uiteinde open te houden van de zijbeek, die alleen periodiek, namelijk gedurende tijden van veel neerslag, snel invallende dooi en zware regenval bij onweersbuien,i nkort etij dvee lwate rafvoerd ee ndi ei ne nvoo rd euitmondin gi n het Geuldal veelsli bsedimenteerde .Daaro m werddi tdee lva nd ezijbee k regel matig uitgediept en verbreed. Het grondoppervlak ter weerszijden hiervan is daardoor voortdurend hoger geworden.1 DeGeu lheef t slechtséé nstroomgorde lgevormd .D ebreedt eva nd eoeverwa l (eenheid L) is gering, soms maar 50 m. De plaatsen waar zware klei en veen voorkomen, in dekommen , liggen altijd aan detegenovergesteld e zijde vanhe t dalal sdi ewaa rd estroomgorde lligt . Veelbodemprofiele n hebbenee nkleigehal - teda t toeneemt bij groterediepte . Erbestaa tee nverschi ltusse nd ebodemgesteldhei dstroomopwaart se nstroom afwaarts van Valkenburg. Kommen met zware klei en veen zijn stroomop waarts van Valkenburg beter ontwikkeld; stroomafwaart s ise r overal een dek van licht materiaal aanwezig. De oorzaak van dit verschil is vooralsnog onduidelijk. Vermeldenswaardi sno gda ttusse nGeulhe me nRothe mi nd eondergron dee n oud riviersysteem uit het Holoceen aanwezig is en dat stroomafwaarts van Meerssen deGeu l drietakke n heeft. Uit debodemgesteldhei d komt naar voren dat tweeervan ,d eGrot eGeu le nd eKlein eGeul ,betrekkelij kjon gmoete n zijn.
1 Een dergelijke beek voert vaak de naam van vloedgraaf.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 23 Het Geulkelijk t eenecht e molenbeek. Vanhe toorspronkelijk e broekbosi nhe tGeulda li sniet sover .Allee ni nveld enplaatsname n komtd erelati eme the toorspronkelijk e 'broek'no guit . Hetda l is geheel een agrarisch gebied. Vroeger washe t voornamelijk als hooiland ('beemd') in gebruik. Opvee lplaatse n ise rno gee nbetrekkelij k kleinschaligzee rafwisselen d land schap metwei - en hooilanden. De grassen- en kruidenvegetatie vertoont op verschillende plaatsen nog duidelijk verband metd ehydrologisch e bodemge steldheid.Populiere ne nwilge nlang sd eGeu le n inlane no fal sbossen ,geknott e bomenva n velerlei soort, heggenlang sd epade n enpercele n verhogen deland schappelijke waarde en zijn tevens een goede biotoop voor vogelse ninsekten . Helaasplege n vervuilingen van velerlei soort, bebouwing (o.a. Schin opGeul) , campings,wege n(autosnelweg ) enbepaald e agrarische aktiviteiten eenaansla g ophe t landschap. HetGeulda lheef t zowelvoo rd e landbouwal s voord e rekreatieee n belangrij ke funktie. Verdere ongewenste ontwikkelingen moeten worden voorkomen, waarbij bedacht moet worden datd elandbou w eenessentiee l element in het landschap vanhe t Geuldalis .
7. REFERENCES
BAKELS, C. C. (1978).Fou r Linearbandkeramik settlements and their environments. Thesis Leiden, 248 pp. BAKKER, H. DE and J. SCHELLING (1966). Systeem van bodemclassificatie voor Nederland. Wagen ingen, 217 pp. BECKERS,H .J .(1929) .Wijzigin g ind eloo p derMaa s tusschen Itteren enStein . Jaarverslag vanhe t Geologisch Bureau vanhe tNederlandsch e Mijngebied, pp. 67-69. BOLT, A. J. J., H. J. MÜCHER, J. SEVINK and J. M. VERSTRATEN (1980). A study on loess derived colluvia in Southern Limburg (The Netherlands). Neth. J. agric. Sei. 28, 1,(i npress) . BRETELER, H. G. M. (1967). De bodemgesteldheid van het ruilverkavelingsgebied Mergelland. Rapport 692, Stichting voor Bodemkartering, Wageningen, 55 pp. BROEK,J .M .M . VAN DEN and H. W. VAN DER MAREL (1964). De alluvialegronde n van deMaas ,d e Roer end eGeul . Bodemkundige Studies, Vu, 83 pp. BRUNNACKER, K. (1977). DasHolozä n imBinnenlan d - die geologische Gegenwart. Geol. Rund schau. 66, pp. 755-770. BRUNNACKER, K. (1978). DerNiederrhei n imHolozän . Fortschr. Geol. Rheinld. u.Westf . 28, pp. 399-440. DAMOISEAUX,J .H .(1968) .Aanvullen d rapport betreffende de bodemgesteldheid van hetruilverka velingsgebied Mergelland (uitbreiding). Rapport 796, Stichtingvoo r Bodemkartering, Wagenin gen, 16 pp. HAVINGA, A.J .an dR.«M . VAN DEN BERG VAN SAPAROEA (1980). Former vegetation and sedimen tation inth e valley of therive r Geul. In: W. van deWestering h et al. (1980).Soi lconditions , soil carbonates and former vegetation inth e Geul valley from Gulpen toMeersse n (South Limburg, The Netherlands). Mededelingen Landbouwhogeschool, Wageningen (Communication Agricul tural University), 80-8,pp . 47-58.
24 Meded. Landbouwhogeschool Wageningen 80-8(1980) HEESEN,H . C. VAN(1970) .Presentatio n of the seasonal fluctuations of thewate r table on soil maps. Geoderma, 4, 3, pp. 257-278. JANSSEN, C. R. (1960). On the late-glacial and post-glacial vegetation of South Limburg (Nether lands). Wentia, 4, 112pp. . MEERMAN, M. (1975). De Geul, zijrivier van de Maas; bijdrage tot de hydrografie van een uniek riviertje. Kerkrade, 192pp . MffiDEMA, R. (1980). Incidence and distribution of carbonates in the soils of the meandering river Geul. In: W. van de Westeringh et al. (1980). Soil conditions, soil carbonates and former vegetation in the Geul valley from Gulpen to Meerssen (South Limburg, The Netherlands). Mededelingen Landbouwhogeschool, Wageningen (Communication Agricultural University), 80-8, pp. 27-46. PONS, L. J. (1973). Enkele aspecten van de bodemgesteldheid van Zuid-Limburg in verband met de plantengroei. Natuurhistorisch Maandblad, 62, 11, pp. 135-143. RiEZEBOS,P .A .an d R. T. SLOTBOOM(1978) .Polle n analysis ofth eHusterbac h peat (Luxemburg);it s significance for the study of subrecent geomorphological events. Boreas, 7, 2, pp. 75-82. SOIL TAXONOMY(1975) .A basicsyste m ofsoi lclassificatio n for making and interpreting soilsurveys . Agriculture Handbook No 436, Soil Survey Staff, Washington, 754 pp. TEUNISSEN VAN MANEN, T. G. (1958). Het riviersysteem van de Geul. Boor en Spade, 9, pp. 53-61. WESTERINGH, W. VAN DE(1979) . Bodemkundig onderzoek in het dal van de Eyserbeek bij Cartiels. Natuurhistorisch Maandblad, 68, 12,pp . 234-239. WESTERINGH, W. VAN DE (1980). Een bodemkartering in het Gulpdal tussen Waterop en Groenen- daal. Natuurhistorisch Maandblad, 69.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 25 INCIDENCE AND DISTRIBUTION OF CARBONATES IN THE SOILS OF THE MEANDERING RIVER GEUL
R. Miedema
CONTENTS
1. INTRODUCTION 27
2. OBSERVATIONS,INTERPRETATIO N ANDDISCUSSIO N 28 2.1. Carbonates deposited aspar t ofth esedimen tloa d 28 2.2. Incidence and distribution of carbonates associated with the fluvial sedimen tation regime 30 2.2.1. Thesimpl emode l 30 2.2.2. Theinterferenc e model 32 2.2.3. Sedimentation due tofloodin g 35 2.2.4. Subsoilcarbonate s 35 2.2.5. Distribution of carbonate incidence associated with the sedimentation of the Geul 36 2.3. Carbonates associated withfan s 41 2.4. Othercarbonat eincidenc e 41
3. ACKNOWLEDGEMENTS 43
4. SUMMARY 43
5. SAMENVATTING 44
6. REFERENCES 45
1. INTRODUCTION
The published evidence gives little detailed information on the incidence and distribution of carbonates in fluvial soils of the large meandering rivers. Levee soils are predominantly calcareous and a description has been given of their progressively deeper decalcification with age of the levee, leading to complete decalcification of some of the oldest soils (PONS, 1957; VAN DEN BROEK and VAN DER MAREL, 1964; SOIL SURVEY INSTITUTE, 1977). Practically all basin soils are non-calcareous, chiefly asa result of decalcification during sedimentation (BEN- NEMA, 1953; PONS, 1957). This proces has been described in detail for Dutch estuarine and marine sediments (BENNEMA, 1953; ZONNEVELD, 1960; VAN DER SLUYS, 1970). But thecomple x nature of the incidence and distribution of carbonates in the Meded. Landbouwhogeschool Wageningen 80-8(1980) 27 soils of the meander belts of the large Dutch rivers is imperfectly understood. The fluvial soils of the Geul are a small-scale model of the sedimentation systemo fth emeanderin g river (VAN DEWESTERINGH , 1980). During asoi lsurve y of this area the presence of carbonates, as determined by the reaction of the soil with 2 M HCl, was one of the characteristics checked at every 10 cm of each augering. At anumbe r of sitesth e presence ofcarbonate s in the upper 1.2 or 2.2 m of the soils along the Geul was investigated in detail by augerings 1-20 m apart. Micromorphological and chemical observations were made of samples of two profiles, and water samples of the river, two tributaries fed by wells and a well discharging into the river were analysed. From these data two models werederive d with the objective ofexplainin g the incidence and distribution of carbonates in the soils of the meander belt of the Geul. These models are discussed in this paper and it is hoped they may throw moreligh t on thecomple x incidencean d distribution ofcarbonate s inth e fluvial soils of the meander belts of the large rivers of the Netherlands. Some help will also be provided in distinguishing sedimentation phases and former river systems according to the incidence of carbonates in the deeper subsoil. The presence of fans along the valley border in front of dry valleys (VAN DE WESTERINGH, 1980)i s a characteristic feature of this part of the Geul valley and the associated carbonate incidence will be discussed.
2. OBSERVATIONS, INTERPRETATION AND DISCUSSION
2.1. CARBONATES DEPOSITED AS PART OF THE SEDIMENT LOAD
VAN DENBROE Kan d VAN DERMARE L(1964 )hav eshow ntha t thecarbonate s in the alluvial soils of the Geul are almost exclusively CaC03. To check thepossibilit y ofchemica l precipitation ofCaC0 3 from thewate r of the Geul, analyses were made of water samples of the river, a well and two tributaries fed by wells. The results are presented in Fig. 1. The well and tri butaries fed by wells (from Cretaceous limestone) have the highest concen trations of Ca, HC03 and H4Si04 (owing to the solution of Cretaceous lime stone and silica leaching from overlying Pleistocene terrace deposits) and have distinctly lower Mg, K, Na, CI and S04 contents than the river samples. Pol lution by inorganic fertilizers, organicmanur e and other human influences may be responsible for the higher concentrations of these ions in the river samples. Intermediate contents are found downstream of the confluence of the Geul with water from wellsan d tributaries.Th ep H isfairl y constant and rangesfro m 7.1t o 7.3. Calculations were made of the ionic strength and the ion activity product (I.A.P. at 25°C , 20°C , 15°C, 10°C and 5°C). The former is highest in the wells
28 Meded. Landbouwhogeschool Wageningen 80-8(1980) O^^^H-H^-^^H^^H — U o O ööööööööööö O 00000--00000
ööööööööööö
O t-0 0^ H o m n n t t © © < " O O O O O O O
oooooo-^oooo ooooooooooo e X o U cd 3 3 o -1 o cd «/-> ooo so o^ of - m O COÎS'H) ööööööööööö < 0) 0 •
(+TÖ ooooooooooo O
(+*N) ooooooooooo
o (+l»Wf) ööööööööööö I O cd v^^H^Hi^i^Tto^-*nOt^ '> Ü TtT tT fT f^ tv- j vS
o O cd w u 0 Î 0>
Meded. Landbouwhogeschool Wageningen 80-8(1980) 29 and tributaries. Dividing theI.A.P . byth e solubility product (Kso) of CaC03 gives thedegre e of CaC03 saturation at a given temperature of the water. No supersaturation with CaC03 occurs at normal water temperatures ofth e Geul ranging from 3°C to 15°C (MRS. BAKKER, pers. comm., 1979). Chemical pre cipitation ofCaC0 3 can therefore berule d out, norwa s it observed. The carbonate particles inth e Geul sediments must bederive d from sedimen tation aspar t ofth e sediment load of the river. This conclusion isborn e out bya micromorphological study oftw oleve esoil s(Schi n op Geul and Kapolder). The sizeo fth eobserve dcarbonat e skeleton grainsfoun d scattered inth e groundmass is comparable to that of the other skeleton grains; it is larger in the sandier Kapolder than in the siltier Schin op Geul profile. The morphology of the carbonate particles observed ranges from coarse to fine crystalline. The car bonate particles are derived from Cretaceous limestone and calcareous loess. Erosion may bring this material down toth e Geul valley where iti stake n upb y the Geul anddeposite d downstream. Carbonate particles areals o produced by erosion of the river's fluvial calcareous deposits and resedimentation down stream. The coarse crystalline type of carbonate particles originates from limestone fragments and calcareous loess. The fine crystalline carbonate par ticles are probably secondary carbonates derived from eroded loess profiles redeposited by the Geul andno w found scattered inth e groundmass. The most recent sediment of the Geul in still unhomogenized sandy loam terraces on a lower elevation than the levee deposits (VAN DE WESTERINGH, 1980) hasa n invariably high CaC03 content (8%—10% w/w). Determinations of material showing effervescence with 2M HC l show that the CaC03 content inth e older, completelyhomogenize d brown siltloa mleve esoil salon gth erive rcours e ranges from 2-3 % (w/w),whic h isi nagreemen t with the contents given by TEUNISSEN VAN MANEN (1958) and VAN DEN BROEK and VAN DER MAREL (1964). The latter refer toa simila r difference inCaC0 3 content between fresh sediment and levee soils ofth e Meuse.
2.2. INCIDENCE ANDDISTRIBUTIO N OFCARBONATE SASSOCIATE DWIT H THEFLUVIA L SEDIMENTATION REGIME
The sedimentation pattern along themeanderin g Geul isdu e to accretiono f the inner and erosion of the outer bends. During short periods of flooding sediment is deposited over the entire floodplain. Thecarbonat e distribution pattern was studied indetai l insevera l meander belt sectionsb yaugering s 1-20 m apart. The presence of carbonates inth e groundmass inlayer s ofa minimum thicknesso f 20 cm, irrespectiveo fth erelativ eamount s ofCaC0 3, wasmappe d in 7 legend units. These are based ona division ofth e upper 1.2m int o three parts by arbitrary boundaries at 40 and8 0 cm. The legend units areshow n inFig .2 .
2.2.1. The simple model Fig.2 show ssom erepresentativ e innerbend si nwhic h we observea succession
30 Meded. Landbouwhogeschool Wageningen 80-8(1980) distance from river 20 40 60 80 100 m Gulpen
50 m
Schoonbron
lach
Legend units I II III IV VI VII IV 0 calcareous +\ 40 + + non calcareous + + +\ _ » ^ 80 augerhole + + + + + -t- + -f flow direction 120 depth(cm) FIG. 2. Carbonate distribution pattern following the simple model in some representative inner bends,an ddept h ofcarbonat e incidencea sa function ofth edistanc e from the river. of arcuate bodies representing legend units I, II and III at an increasing distance from theriver .Th e location ofthes ebend si sindicate d in Fig. 6.A t a still greater distance from the river no carbonates are found within a depth of 1.2 m (legend unit IV). The accretion of the sediment in such inner bends shows corresponding time- dependent arcuate bodies, thus indicating a difference in agei n the sediments in the direction mentioned. The outer bends are in an erosive position, so that the sediment isolde r than that found in inner bends. It istherefor e conceivable that the outer bends contain no carbonates or units with carbonates starting at a greater depth.Th emicromorpholog y ofth eKapolde r (unitI )an d Schino p Geul (unit III) profiles shows that the primary carbonate skeleton grains are severely corroded near the boundary between calcareous and non-calcareous material. Secondarycarbonate swer eoccasionall y observed inan d around voids (calcitans and neocalcitans) (Fig. 3),thi s being due to decalcification. The fact that secon dary carbonates wereonl yoccasionall y observed inth efield o r inthi n sectionsi s due to the discharge of the dissolved Ca(HC03)2 by the groundwater which fluctuates with the level of the river. The shape of the graphs showing the depth of carbonate incidence as a function of the distance from the river (Fig. 2) is determined by the fact that the rate of accretion of an inner bend increases with increasing excentricity ofth e meander combined with decalcification. The latest depositssho wshallowe r decalcification and weredeposite d over ashorte r period of time. Hence the carbonate distribution pattern is caused by decalcification as a function of the age of the sediment, as determined by the sedimentation pattern in.the accreting inner bend. The oldest deposits are farthest from the river in an inneran di noute rben dpositio n andhav ebee ndecalcifie d toth egreates tdepth .I n the older, deeply decalcified profiles, weak illuvation features were observed in thefor m ofchanne lmatri-ferri-argillan sa sdescribe d byVA N SCHUYLENBORGHet al.(1970 ) in a non-calcareous Geul profile.
2.2.2. The interference model The pattern found in inner bends may be more complex. Fig. 4 shows the observations in two such bends. Their location is also shown in Fig. 6. The pattern observed cannot beexplaine d with theus e ofth e simplemode l discussed above. In Fig.4 A legend unit Ioccur s along the river in the form of awide r strip in theeaster n part of the bend with an east-west oriented extension, and a fairly narrow strip in the western part. A zone with legend unit II accompanies the zone with unit I, and an island with legend unit III occurs in the western part of thebend . To thenort h no carbonates werefoun d within a depth of 1.2m (legend unit IV), but this lies outside the accretion area of the bend. This asymmetrical pattern of carbonate incidence does not correspond to the accretion asymmetry of the present shape of the meander bend, as one would expecta wide rstri po frecen t sedimentsi nth ewester npar t(cf . alsoth epattern si n Fig. 2).Th e east-west extension ofuni t Ii stopographicall y somewhat lower and scattered coarse, gravelly bedding material was found within a depth of 1.2 m.
32 Meded. Landbouwhogeschool Wageningen 80-8(1980) !;;.;.;.;;;.j secondary carbonates ^'••'•''•'••'•.'••'j (Neocalci tan/cal ci tan]
groundmass m voids FIG. 3.Secondar y carbonates inan d around voidsi nth e subsoil of theSchoonbro n inner bend.
It would seem likely that changes in the river bed inside the meander belt interfered withth epatter n ofth esimpl emodel .Th eforme r channelswer e filled withmor erecen tcalcareou smateria lt oalmos tth esam eleve la sth esurroundin g soils,s otha tthei rprofile s areno tdecalcifie d toth esam edepth .Thi sca nals ob e seeni nth egrap h showingth edept h ofcarbonat e incidencea sa functio n ofth e distancefro mth erive r(Fig .4B) .Th eTrancho ttopographica lma po f1803-182 0 shows a meander of the same shape; hence the assumed former meander must antedate 1800. Theexampl ei nFig . 4Ccome sfro m apar t ofth eGeu lvalle ywher e TEUNISSEN Meded. Landbouwhogeschool Wageningen 80-8(1980) 33 distance from river 40 80 120 160 200 m kenburg
cal careous non calcareous • augerhole — fl ow di recti on
160 m
legend units II III IV V VI VII V +\ \ +\ + + V + + \ + + A + + A + + + + V + + + + Kapolder 120 A depth(cm)
FIG. 4.Carbonat e distribution pattern following theinterferenc e model in some representative inner bends, and depth of carbonate incidence as a function of the distance from the river.
34 Meded. Landbouwhogeschool Wageningen 80-8( 1980) VAN MANEN (1958) carried out a very detailed soil survey. He found distinct former channelswhic har eresponsibl efo r thepatter n ofcarbonat e incidence and theassociate d graph ofdept h ofcarbonat e incidencea sa functio n ofth e distance from the river (Fig. 4D). Here too we find an island of decalcified soils sur rounded by an area of shallower decalcification. Itca n alsob esee n from Fig. 6tha t changesi n therive r bed must betake n into account to explain the distribution pattern of carbonates; here the old channels nearOu d Valkenburgan dcut-of f meandersnea rEtenake n arestil lclearl yvisibl e in the field and the distribution pattern of carbonates is in agreement with the sedimentation pattern of the former course of the Geul according to the simple model.
2.2.3. Sedimentation due to flooding Sedimentation due to flooding is responsible for the more heavily textured soils in the transition and basin position (texture classes M and Z; VAN DE WESTERINGH, 1980). These soils are usually non-calcareous as a result of com plete syn-sedimentary decalcification under conditions of poor drainage (BEN- NEMA, 1953; PONS, 1957; ZONNEVELD, 1960; VAN DER SLUYS, 1970). Minor amounts of pyrites and spherical iron hydroxides as oxidation products were observed in the peaty subsoil in thin sections of a basin profile near Schin op Geul and these may have speeded up the process of syn-sedimentary decalcifi cation. The thin layer of L material covering more heavily textured deposits upstream of Valkenburg belongs to a more recent sedimentation stage (VAN DE WESTERINGH, 1980). This layer having been deposited in areas of poor drainage remote from the river, it may have been completely decalcified by a syn- sedimentary process. Alternatively, its age may be such that it has lost its carbonates by the decalcification process described for the levee deposits. But downstream of Valkenburg this deposit is more wide-spread and thicker and is still not completely decalcified. For this reason the extent of carbonates as sociated with the sedimentation regime of the Geul increases downstream of Valkenburg and calcareous soils are no longer exclusively found in inner bends.
2.2.4. Subsoil carbonates The incidence and distribution of subsoil carbonates (between 1.2 and 2.2m) , on which data are available from 1976, 1977 and 1978 and from deep augering sequences, may provide evidence to support VAN DE WESTERINGH'S (1980) theo ries of important changes in the course of the Geul during its sedimentation history (Fig. 5).I n this area the subsoil carbonates are found in lightly textured deposits underlying more heavily textured transition and basin deposits. In the present basin area former filled-in channels can be identified in the field by their somewhat lower topography, pointing to changes in the river bed. This has also been found in other areas now in a basin position with respect to the Geul. But in extensive areas along the Geul the carbonate incidence in the subsoil is merely the continuation of this incidence in the upper 1.2m . In such cases the decalcification sequence is extended by types with carbonates starting pro-
Meded. Landbouwhogeschool Wageningen 80-8(1980) 35 MEERSSEN 120
++ ++ K I IX X XI XII legend units
\\\ boundary V^\ of alluvial soils
Vroenhof
n.s.not surveyed Fig. 5.Incidenc e and distribution pattern of subsoil carbonates (between 1.2 and 2.2meters ) ina selected area ofth e Geulvalley . gressively deeper in the subsoil between 1.2 and 2.2 m only, indicating a still deeper decalcification in older deposits at a larger distance from the Geul (sequence I-II-IH-IX-X-XI-IV + XII). The texture of these soils remains L throughout the profile.
2.2.5.Distribution of carbonate incidenceassociated with thesedimentation of the Geul Inspection of the carbonate distribution map (Fig. 6), bearing in mind the theories outlined above, shows that in the south-eastern part of the survey area between Partij and Gulpen only very limited amounts of carbonates along the Geul are found in clearly distinguishable former channels. This agrees with the observations made by VAN DEN BROEK and VAN DER MAREL (1964). The small amount ofcarbonate s isdu e to the almost complete absence ofstee p Cretaceous limestone slopes south of Partij. The loess plateaus are not highly dissected and are mainly grassland and forest. Wells feeding the tributaries are derived from the Vaals formation (glauconitic green sand and clay of Cretaceous age; MEER MAN, 1975). Between Gulpen and Wijlre the carbonates increase along the Geul. Near Wijlre in particular this part of the valley has suffered from fairly frequent flooding (MEERMAN, 1975),s otha t mosto fth eleve esoil sma yb erelativel y recent 36 Meded. Landbouwhogeschool Wageningen 80-8(1980) and not completely decalcified. Both the simple model (Fig. 2A) and the in terference model are needed to explain the carbonate distribution. TEUNISSEN VAN MANEN (1958) surveyed part of this area and identified distinct former channels within the levee deposits where the interference model has to be used (Fig. 4C, D). Steep Cretaceous limestones slopes occur with very active wells discharging into the Geul and its Eyserbeek tributary (MEERMAN, 1975). The plateaus are more heavily dissected and mainly arable. Between Wijlre and Valkenburg the carbonates are chiefly found in inner bends, following the simple model (Fig. 2B). Near Oud Valkenburg the car bonate incidence in former channels is very clear. One of the many active wells from Cretaceous limestone in this section discharging into the Geul wassample d and analysed (Fig. 1).Th eloes splateau s are fairly heavily dissected and are mainly arable. Downstream of Valkenburg to the division of the Geul in two branches (near Vroenhof) carbonates are also more frequent around outer bends, although in a narrower stripan d more deeply decalcified than ininne r bends.Th e distribution canusuall y beexplaine d byth e simplemode l(Fig .2C) ,althoug h the interference model is occasionally needed (Fig. 4A, B). Steep Cretaceous limestones slopes border the valley to the south, but tributaries of the Geul from the north are fed by wells from the Oligocène Cerithien clay overlying Cretaceous limestone (MEERMAN, 1975). The plateaus are less highly dissected and mainly arable. From therive rfor k nearVroenho f toth econfluenc e ofth ethre e branches west of Meerssen carbonates are found in extensive areas between Grote Geul and Geulke with relatively shallow decalcification. Islands ofuni t IV surrounded by calcareous soils are a striking feature (Fig. 7A).The y are usually surrounded by soilsdecalcifie d to40-8 0 cm,an d surrounded intur n bysoil sdecalcifie d to 0-40 cm. Many former filled-in channels were observed in the field in the neigh bourhood of the unit IV islands. The latest sedimentation phase of L material is widespread in this part of the valley; it is still calcareous and overlies more heavily textured deposits within a depth of 1.2 m (VAN DE WESTERINGH, 1980). This explains why units VI and VII are frequently found as shallower repre sentatives of I and II.Th e extent and thickness of thislates t deposit may be read from Fig. 7B (depth of carbonate incidence in this area). A clear picture of the sedimentation history of this area isobtaine d bycombinin g Figs. 7A and Bwit h Fig. 5an d comparing the result with the soil map of this part of the valley (VAN DE WESTERINGH, 1980). The carbonate distribution pattern results from the application ofth einterferenc e modelt olarge rarea s; thi swil ln odoub t havet ob e done even more frequently when carbonates are distributed in the soils of the largerivers .Alon g thebranche s of theGeu l the simplemode lca n be successfully used.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 37 (western part)
Ber
see Fig. 2
carbonates associated with the Geul * detail seeFig .2 (units I,II,III,VI,VII) carbonates associated with fans (units V,VI, sometimes 1,11,111) * detail see Fig. 4 other carbonate incidence (mainly unit V) T disturbed area
| | detail seeFig .5,7,8, 9 n.s. not surveyed ,.,.•• boundary of alluvial soils
FIG. 6. Distribution of carbonate incidence in the soils of the river Geul (western part and eastern part).
1ocation map
38 Meded. Landbouwhogeschool Wageningen 80-8(1980) (eastern part)
I*----« Partij
Meded. Landbouwhogeschool Wageningen 80-8(1980) 39 depth of decalcification 0 - 40 cm (units I and VI) depth of decalcification 40 - 80 cm MEERSSEN (units II and VII) depth of decalcification 80 - 120 cm 'unit III)
thickness latest deposit at least 120 cm MEERSSEN (units I, II and III) thickness latest deposit mostly around 80 cm HI (units VI and VII) thickness latest deposit unknown (unit IV)
n.s. not surveyed V\boundary of alluvial soils
Vroenhof
FIG. 7A.Carbonat eincidenc ean ddept h ofdecalcificatio n ina n areawit hextensiv edeposit so fth e latestsedimentatio n phase. FIG. 7B. Thickness of the deposits of the latest sedimentation phase based upon the depth of carbonate incidence,ignorin g thedept h ofdecalcificatio n and anthropogenic influences.
40 Meded. Landbouwhogeschool Wageningen 80-8(1980) 2.3. CARBONATES ASSOCIATED WITH FANS
Fans are acharacteristi c feature of the Geul valleyan d occur along the valley border in front of dry valleys cutting into Cretaceous limestone or Tertiary deposits.Th eformatio n ofthes efan s isdescribe d by VAND EWESTERING H( 1980) . Wherethes efan s originatefro m valleyscuttin gint o Cretaceouslimestone , the deposited material is calcareous and large areas of carbonates are found near Gulpen, near Stokkem, between Etenaken and Schin op Geul, between Schin op Geul and Oud Valkenburg. Fig. 8 shows that the Stokkem fan (for location cf. Fig. 6)i scompletel y calcareous (unit I) and the calcareous material wedges out over the alluvial soils further away from the fan (succession of units VI and V). The two connections from the fan to the Geul are also distinctly visible.Th e fan between Schin op Geul and Oud Valkenburg in front of the 'Gerendal' (Fig. 9) shows a succession of units I, II and III from the road to the Geul. In this case progressively deeper decalcification has occurred in the older part of the sedi ments. Detailed augerings in the fan north-west of Etenaken (KWANTES and SNEL, 1976) revealed successive stages of deposition, separated by decalcified intervals overlying peat in the deeper subsoil. HAVINGA and VAN DEN BERG VAN SAPAROEA (1980) have presented palynological data on the age of the sediment overlying peat in the fan south of Etenaken. Distinct fans are also found downstream of Valkenburg, but here they origi nate from side valleys cutting into non-calcareous Tertiary deposits, so that the actual fan is non-calcareous and not surrounded by calcareous deposits. Fig. 6 shows the carbonate incidence associated with fans, and the difference noted above can be clearly seen.
2.4. OTHER CARBONATE INCIDENCE
This includes calcareous soils at the foot of steep Cretaceous limestone slopes which are due to colluviation and normally excluded from alluvial soils. Scattered carbonates, often in a basin position and almost exclusively cor responding to legend unit V, very often show sharp, angular boundaries on the map (Fig. 6). In most cases human influence is responsible. Farmers use lime stone chips to improve the bearing capacity of the soil, as is known from infor mation obtained in the area between Gulpen and Wijlre and between Valken burg and St. Gerlach. Near Vroenhof the topsoil of large areas has been con taminated by transport of excavated limestone. It must, however, be admitted that in some cases no satisfactory explanation can be given. Summarizing, the incidence and distribution of carbonates linked to the sedimentation regime of the Geul can usually-be satisfactorily explained by the useo fth etw omodel sdiscussed . VANDE NBROE Kan d VANDE RMAREL' S statement (1964)tha t the leveedeposit s of the Geul downstream of Gulpen are calcareous needs modification in view of the variation in extent of the calcareous
Meded. Landbouwhogeschool Wageningen 80-8(1980) 41 to e
CU O u vvj
••- "<- TT s G -o Ol J3 m ui S o i— •o O ^ o '*Z (L>
c •— > T3 i-H •— - t-l *-t > 'I •-i »—i u (/) *ö >> c o CU => . > 3 -t-J urn ^ 3 O CU T3 & Si. 0) 01 to .— o ° O - Q C
o
a
CM 00 o +3 3
o =3 O O t 'm I/1M <" m o (ö t—« o (O o c CU " <° (0 1— o o (Ö c u "BIO t—i •— l__l « >- 'o "O > > +•> -p -p E ° c c c CU 3 3 3 =J u-o ; -o •o "O c c c c o CU CU CU CU .c Dl en u CU CU CU ca 2 .= ^ ,= O :arbo n [lege n bound c U o
42 Meded. Landbouwhogeschool Wageningen 80-8(1980) part of the levee soils along the valley. Subsoil carbonates in more lightly textured deposits underlying more heavily textured deposits in thepresen t basin positionsma yhel pt oidentif y pastchange si nth erive r bed. Studieso n the dating ofth e sedimentation phases (HAVINGA and VAN DEN BERG VAN SAPAROEA, 1980) are needed to ascertain the decalcification rate with the use of such modern methods as those described by LEEDER (1975), SALOMONS(1975 )an d SALOMONS and MOOK (1976). Another important part of the carbonate incidence can be linked to the presence of fans in front of dry valleys cutting into Cretaceous limestone. Most of the other carbonate occurrence can be attributed to the human influence responsible for the calcareous topsoil.
3. ACKNOWLEDGEMENTS
I gratefully acknowledge the contributions made by the students KWANTES- DORREPAAL, SNEL, BROEKHUIZEN, VAN HEMMEN, BALTISSEN, DE BOO, OOMEN and VAN DERZEE ,wh ocarrie d out thedetaile d investigations needed tochec k the models described. I would alsolik et o thank mycolleague s for the stimulating discussions on the proposed models and their constructive criticism of earlier drafts of this paper.
4. SUMMARY
Very detailed field research showed that the incidence and distribution of carbonates in the fluvial soils of the meander belt of the river Geul could be explained by the use of two models. In the simplemode l thepatter n of distribution ofcarbonate s ininne r bends of arcuate bodies with a progressively deeper incidence ofcarbonate s according to their distance from the river corresponds to the accretion of an inner bend followed by time-dependent decalcification. In the interference model the pattern of distribution of carbonates is com plicated by changes in the river bed inside the meander belt. The distribution patterns observed can be explained by reconstructing the sedimentation se quence and using the simple model. The more heavily textured transitional and basin deposits are non-calcareous as a result of synsedimentary decalcification. This enables us to distinguish sedimentation phases, likewise on the basis of carbonate incidence. Evidence of former river systems, sometimes situated in a basin position with respect to the present course of the river (buried levees), is provided by the incidence of
Meded. Landbouwhogeschool Wageningen 80-8(1980) 43 carbonates in the deeper subsoil in lightly textured deposits underlying non- calcareous,mor e heavily textured deposits. Onepeculia rfeatur ei sth elarg earea so fcarbonate sfoun d irrespectiveo fthei r leveeo rbasi npositio n and associatedwit h fans alongth evalle yborde r in front ofdr yvalley scuttin g into Cretaceouslimestone . Iti shope dtha tth eidea spu tforwar d inthi spape rma ythro wmor eligh to nth e littleunderstoo d complexincidenc ean ddistributio n ofcarbonate si nth emean der beltso fth elarg e riverso f theNetherlands .
5. SAMENVATTING
VOORKOMEN EN VERBREIDING VAN CARBONATEN IN DEGRONDE N VAN DE MEANDERENDE GEUL
Het voorkomen en de verbreiding van carbonaten in de fluviatiele gronden vand emeandergorde lva nd eGeu lka nverklaar dworde ndoo rgebrui kt emake n vantwe emodellen ,zoal si sgebleke nui tzee rgedetailleer d veldonderzoek. In het eenvoudige model laat het distributiepatroon van de carbonaten in binnenbochten een schilvormige opeenvolging zien, waarbij de diepte van voorkomenva nd ecarbonate ntoeneem tme td eafstan d totd erivier .Di tpatroo n is in overeenstemming met de laterale sedimentatie van een binnenbocht, gekoppeld aan een in detij d voortschrijdende ontkalking. In het interactiemodel wordt het patroon ingewikkelder door stroomver- leggingen binnen de meandergordel. Niettemin kan men de waargenomen patronen verklaren door een reconstructie van de sedimentatiegeschiedenis, waarna het eenvoudige modelwee rka n worden toegepast. De in kompositie en in overgangspositie afgezette lutumrijkere gronden zijn kalkloostengevolg eva n synsedimentaire ontkalking. Dit biedtd e mogelijkheid omoo ko pbasi sva nkalkhoudendhei d sedimentatiefasen teonderscheiden . Het voorkomen van carbonaten in de diepere ondergrond in lichtere afzettingen gelegen onder zwaardere kalkloze afzettingen (begraven oeverwallen), is een aanwijzing voorvoormalig estroomsysteme n opplaatse ndi esom si nkompositi e liggente n opzichte van dehuidig e rivierloop. Een bijzonderheid zijn de grote oppervlakken kalkhoudende gronden, die onafhankelijk van stroomrug- of kompositie voorkomen, en die samenhangen met colluviumtongen voor droge dalen, ingesneden in de kalksteenafzettingen uit het Krijt. Hopelijk kunnen dehie r gepresenteerde modellenee nbijdrag e leveren tot de ontrafeling van de momenteel nog weinig begrepen complexiteit van het kalkvoorkomen ind emeandergorde l van degrot e rivieren van Nederland.
44 Meded. Landbouwhogeschool Wageningen 80-8(1980) 6. REFERENCES
BENNEMA,J . (1953).Decalcificatio n during accretion ofalluvia l soilsi nth eNetherland s (Dutch with English summary). Auger and Spade VI, pp. 30-40. BROEK, J.J . M.VA N DEN and H. W. VAN DER MAREL (1964).Th e alluvial soils ofth e rivers Meuse, Roer and Geul in the province of Limburg (Dutch with English summary). Agricultural Research Reports 645, Soil Science Studies 7, Soil Survey Institute. 83 pp. GARRELS, R. M.an dC . L. CHRIST (1965). Solutions, minerals andequilibria . Harper, Row and Weatherhill, New York, 450 pp. HAVINGA, A. J. and R. M. VANDEN BERG VANSAPAROEA(1980) .Forme r vegetation and sedimenta tion in the valley of the river Geul. In: W. van deWestering h et al (1980). Soil conditions, soil carbonates and former vegetation in the Geul valleyfro m Gulpen to Meerssen (South Limburg, Netherlands). Mededelingen Landbouwhogeschool. 80-8, pp. 47-58. KWANTES-DORREPAAL, M.E .an dA .R . SNEL (1976). Verslag onderzoek alluviale gronden in het Geuldal (Dutch). M. Sc.-thesis Department of Soil Science and Geology. Agricultural Uni versity, Wageningen, 21 pp. LEEDER, M. R. (1975). Pedogenic carbonates and flood sediment accretion rates: a quantitative model foralluvia l arid-zone lithofacies. Geol. Mag. 112 (3),pp . 257-270. MEERMAN, M.(1975) . DeGeul , zij-rivier van deMaas . Bijdrage totd ehydrografi e vanee n uniek riviertje (Dutch), 192pp . PONS, L.J .(1957) .Geology , soil formation and history ofth e drainage conditions inth e 'Land van Maas en Waal'an d part of the 'Rijk vanNijmegen' . Ph.D.-thesis,Agricultura l Research reports 63.11,156 pp. SALOMONS,W .(1975) .Chemica lan disotopi ccompositio n ofcarbonate si nrecen tsediment san d soils from Western Europe. J. Sed. Petr., 45 (2), pp. 440-449. SALOMONS, W. and W. G. MOOK (1976). Isotope geochemistry of carbonate dissolution and repre- cipitation insoils . Soil Sei., 122 (1),pp . 15-24. SCHUYLENBORGH,J . VAN, S. SLAGER and A. G. JONGMANS(1970) .O n soilgenesi si ntemperat e humid climate. VIII. The formation ofa 'Udalfic' Eutrochrept. Neth. J.Agric . Sei., 18,pp . 207-214. SLUYS,P . VAN DER (1970).Decalcificatio n ofmarin e clay soilsconnecte d with decalcification during silting. Geoderma 4, pp. 209-227. SOIL SURVEY INSTITUTE (1975). Report to 1:50.000 map sheets 40E & W (Arnhem) , 197pp . TEUNISSEN VAN MANEN, T. C.(1958) . The river system ofth e Geul (Dutch with English summary). Auger and Spade IX, pp. 53-62. WESTERINGH,W . VAND E(1980) .Soil san d their geologyi n theGeu l valley.I n: W.va n de Westeringh et al (1980). Soil conditions, soil carbonates and former vegetation in the Geul valley from Gulpen toMeersse n(Sout h Limburg,Netherlands) .Mededelinge n Landbouwhogeschool. 80-8, pp. 1-26. ZONNEVELD, I.S .(1960) .D eBrabants e Biesbosch; a stud y ofsoi lan d vegetation ofa fres h water tidal delta. Soil Science Studies4 , Soil Survey Institute,Agricultura l Research Reports 65.20,396pp .
Meded. Landbouwhogeschool Wageningen 80-8(1980) 45 FORMER VEGETATION AND SEDIMENTATION IN THE VALLEY OF THE RIVER GEUL
A.J . Havinga and R. M.va n den Bergva n Saparoea
CONTENTS
1. AIMO FTH EINVESTIGATIO N 47
2. THEPROFILE S STUDIED 48
3. SOMEPOLLE N ANALYSISPROBLEM S 49
4. DESCRIPTION AND INTERPRETATION OFTH E DIAGRAMS 52 4.1. Late-glacial zone 52 4.2. Pre-borealzon e 53 4.3. Borealzon e 53 4.4. Atlanticzon e 53 4.5. Sub-boreal zone 55 4.6. Sub-atlanticzon e 56 5. SOMECONCLUSION S 57
6. SUMMARY 58
7. SAMENVATTING 58
8. REFERENCES 59
1. AIM OF THE INVESTIGATION
Atentativ einvestigatio n wasmad eint oth eag eo fth eloes sdeposit so f fluvial or colluvial origin in the valley of the river Geul. Our aim was to establish whether these deposits could have been the indirect result of the deforestation which began with the introduction of agriculture in the hilly region of South Limburg. If so, theywoul d beo f relatively youngdate . Being of a tentative kind, the investigation was kept simple and limited in scope. Only two profiles were subjected to pollen analysis, one consisting of fluvial loesso npea tan dth eothe ro fcolluvia lloes so npeat .Th epea tpar t ofth e twoprofile swa sanalyse dove rit sentir edepth ,th ecoverin gminera lpar tnea rth e contact zone only.
Meded. Landbouwhogeschool Wageningen 80-8(1980) Al 2. THE PROFILES STUDIED
Meerssen
v St.Qerlach Valkenburg ^^^Z^Geuj
Geul va ley profile "Vroenhof"
.Wi Ire profi e"Etenaken >
THE NETHERLANDS
Location *
Fig. 1. Location of the two 'Vroenhof and 'Etenaken' profiles.
TheVroenho fprofil ei slocate di na smal lbasi nare ao fth eriver cla ylandscap e in the Geul valley at a relatively great distance from the river course with its accompanyinglevees ,an d out ofreac h ofa colluvium .I tconsist so fa napprox imately 1.40 m layer of fluvial loess resting on the peat in an undisturbed position. TheEtenake n profile isi na comparabl esituatio n asregard sth ecours eo fth e riverbu tlocate dwithi nreac ho fa colluviu malon gth emargi no fth evalley .Th e colluvialmateria l doesno t resto n alaye r offluvia l loessbu t isi ndirec t contact with the peaty subsoil. This situation suggests that deposition of the colluvial loessma yhav ebee nprecede d bysom eerosion ,leadin gt oth edisappearanc e of the former fluvial loess layer and the top of the peat. Near the bottom of the colluvium itcontain s peatyparticles .
48 Meded, Landbouwhogeschool Wageningen80-8 (1980) Profile descriptions. Vroenhof profile 0-141cm .Fluvia lloess . 141-160cm . Clayey l Carex peat with fairly extensive decay. 160-170cm .Humi cclay 1. 170-200cm .Clayey 1 Carexpea t with fairly extensivedecay . 200-250cm .Carex pea t withmoderat e decay. 250- cm.Humi c sand. Etenaken profile 0-220 cm. Colluvialloes swit hsmal lpiece so fbake dloam ,charcoal ,potsherd s and somegravel .Fro m 0-60 cmth esoi lappeare d to beraise d by human agency. 220-227 cm. Colluvial loess mixed with peaty particles which increase in a downward direction. 227-330 cm. Carexpea t alternately with fairly extensive decay. Fragments of limestone arepresen t at variousdepths . 330-360cm . Carexpea t mixed with fluvial loess which increases in a down ward direction. Locally iti scalcareous . 360- cm. Fluvial loessmixe d with somepeat y material. It may be inferred from the occurrence of clayey material in the peat in the Vroenhofprofil ean dth efragment s oflimeston ei nth eothe rprofil etha tth epea t was periodically flooded during formation. The flood water may have eroded peaty material whichcoul d have been deposited elsewhere. Local decay of the peatshow stha tperiod swit ha lo wwate rtabl emus tals ohav eoccurred .I twa sa t thosetime stha t thepolle ngrain si nth epea twer eattacke d bymicro-organisms , giving thepolle n wallit sgenerall y fairly severelycorrode d appearance.
3. SOME POLLEN ANALYSIS PROBLEMS
Owing to selective pollen corrosion the two pollen diagrams may give a somewhat distorted picture.Thi swoul d explain whyCarpinus i srepresente d in the Vroenhof diagram by one pollen grain only viz. in the topmost Subboreal spectrum and is completely absent from the Etenaken diagram, even in the Sub-atlanticzon ewher ei tmigh t havebee nexpecte d(se eals op . 56).Fraxinus i s also absent from the latter diagram. According to the literature both species often showlittl e resistancet o corrosion. Inth eminera l subsoilo fth eEtenake n profile thepolle nconcentratio n below 3.65 m was so slight that the samples analysed contained too little pollen to afford a complete spectrum. To obtain a better insight the pollen counts were added to a singlespectrum , viz. 3.85m . 1Th egrain-siz edistributio no fth e'clay 'componen twa sno tanalysed ,bu ti tma yb eassume dtha ti t mainlyconsist so floess . Meded. Landbouwhogeschool Wageningen 80-8(1980) 49 3 >. C
ï ï .C .1- «4- Û..O -C UI +j m >i
DO»* • •444
-f< 1 OL.O < 1 o o
I > I» I » I I > I » I • > > > > Î > I - Î: 'î> Î : I » I > I ' I lis!;
-^
§£ S ï £ £=£ S S OT< eO «o < 1- « 0 A _i eO — -J I « _i mW
r«. «0 00
> > 4. DESCRIPTION AND INTERPRETATION OF THE DIAGRAMS
In the diagrams the first column shows the lithology of the soil profile, the second thepolle n zones,an d the third the tree-pollen percentages based on the total ofal ltre especie s(includin g Corylus) (SA P = 100%).Th efourt h column shows on a larger scale the proportion of each of the four components of the Quercetum - mixtum(Q.M.) , Quercus, Tilia, Ulmus and Fraxinus and also the representation ofFagus and Carpinus, neither ofwhic happear s before the later Holocene. The fifth column is an overall picture of the vegetation, the per centages being calculated on the basis of combined arboreal pollen and non- arborealpolle n(EA P + ENAP = 100%). Thevalue sfo rArtemisia, cereal san d agriculturalweed sar egive no nth eextrem eright. Artemisia i sa commo nspecie s in Late-glacial pollen spectra and isevidenc e of open terrain conditions. The ratio of all arboreal pollen to all non-arboreal pollen (SAP: E NAP ) is usually a good index of the extent to which the surrounding area wascovere d withforest . Ahig hrati o indicatesa dens eforest , alo won ea n open vegetation. To obtain the best impression one should omit from the pollen assemblage considered plant specieswhic hgre wi nth eneighbourhoo d ofth esamplin gsite , asotherwis eth epolle no fthes especie swoul db eover-represented .T oasses sth e extentt owhic hth emor eelevated ,drie rsoil si nth earea saroun dth etw oprofile s analysedwer ecovere db yfores t duringth eperiod srepresente db yth ediagrams , wecoul domi tth eferns ,cype rgrasses ,par to fth egrasse san dherbs ,an dals oth e waterplants ,al lo fwhic hha dthei rhabita to nth emois tsoil so fth evalle yfloor .Th e localcharacte ro fpar to fth enon-arborea lpolle ni seviden tfro mth efac ttha tth e non-arborealspecie ssho ws omuc hhighe rpercentage si nth eVroenho fdiagra m than inth e Etenaken diagram. Butthi smetho d ofcalculatio n wasfoun d tob eineffective ; after omissiono f theferns ,cype rgrasse san dwate rplant sfo rwhic ha mois thabita tma yb etake n for granted,th ecurv efo r theZA P: ZNA Prati owa sye terrati ci nth eVroenho f diagram. Apparently much of the remaining NAP pollen also came from the moist vegetation near the profile although the actual proportion cannot be determined. As a result we did not construct a diagram giving reduced percentages. Thetw odiagram swil lb ediscusse d togetherbelow ,t obegi nwit hth ediagra m inwhic hth epolle nzon eunde rdiscussio nshow sth ecleares to rsimples tpattern .
4.1. LATE-GLACIALZON E
The arboreal section of the Late-glacial spectra in the Vroenhof diagram is highly dominated by Betula. The low ZAP:SNAP ratio and the fairly high Artemisia percentages indicate an open vegetation. Here and there some birch occurred.Cory lus, transitor yrepresente db y1 %,wa sno tye tgrowin gi nth evicinity . 52 Meded. Landbouwhogeschool Wageningen 80-8(1980) 4.2. PRE-BOREAL ZONE
Inth ePre-borea lzon eo fth eVroenho fdiagra m Pinusshow sa rapi d advance to high values to the detriment of Betula. This is a common feature of diagrams from South Limburg and adjacent areas. Corylus puts in a definite appearance towards the end of this period. TheZA P :S NA P ratio islo w as a result of local growth of ferns and cyper grasses. At the top of this zone the ratio increases somewhat. During the Pre-boreal period Artemisia disappears. The landscape became covered with a uniform fir forest which afforded little space for herb growth. The only spectrum of Pre-boreal age in the Etenaken diagram has much the samecompositio n asth e topmost Pre-boreal spectrum in theVroenho f diagram although the Filicinae percentage is much lower. Unlike the latter diagram, the Pre-boreal zone of the Etenaken diagram falls within a deposit of fluvial loess.Thi scoincidenc e may explain the relatively high Alnus valuean d therepresentatio n ofPicea i nthi szon ei nth eEtenake n diagram. Both pollen types may have originated from older soil layers and have been redeposited together with the loess material in which they are found.
4.3. BOREAL ZONE
The Borealzon e ofth eVroenho f diagram starts witha n increase in Corylus to thedetrimen t of Pinus and the appearance of the Q.M. Thereafter the latter also increases to the detriment of Pinus. The increase is chiefly in respect of a single component of the Q.M., viz. Ulmus.TheZA P :ZNAP ratio reaches a very high figure. Thefores t iso fa les sunifor m charactertha n duringth ePre-borea l period. Broadley speaking the Etenaken diagram isa compresse d picture ofth e above Boreal diagram section. One striking difference is the Quercus peak at 3.45 m.
4.4. ATLANTIC ZONE
The very high ZAP :ZNA P ratio in all Atlantic spectra in the Etenaken diagram indicates a dense forest vegetation. Pinus shows a marked recession at the beginning of the Atlantic period whereas the Q.M. makes a further advance. Corylus provisionally maintains its ground. Thebeginnin g ofth eAtlanti czon ei spu t at thecrossin go fth ePinus and the Q.M. curve (cf. JANSSEN, 1960). At a much later stage Alnus comes to the fore. This phenomenon is also found in a diagram from the more southerly country of Luxemburg (RIEZEBOS and SLOTBOOM 1974), whereas JANSSENS'S diagramsfro m South Limburgsho wAlnus dominancedurin gth ewhol eAtlanti c period aswel la sdurin g thelate rstag eonly .I n thisconnectio n weals orefe r to the pollen diagrams from sand soils in the neighbouring part of Belgium (MUNAUT 1967) in which Alnus is of minor importance during the whole Atlantic period.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 53 The Q.M. isfirs t dominated by Ulmusan d later by Tilia.Highe r Ulmusvalue s followed by higher Tiliavalue s are also found in several diagrams from adjacent regions in Belgium and Germany (see JANSSEN 1960,p . 67 and WOILLARD 1975, p. 98). But this succession is not a general rule. For instance, two of JANSSEN'S diagrams in which the older part of the Atlantic period is represented, show a dominance of Quercus or Tilia, whereas in both cases Ulmus is last. As in the earlierpar t eacho fth ethre especie sma ydominat e inth elate rpar t ofth e Atlantic zone. Apparently local forest composition had a marked effect on the fossil pollen floras. Thismus t bedu e to the limited extent ofth e smallpea t moors from which the experimental material was taken. Between 310 and 305 cm the diagram shows an erratic pattern. Pinus and Corylus occasionall y attain high values whereas Tiliai smissin g from two of the spectra. In the Atlantic zone of two of JANSSEN'S (1960) diagrams Pinus is also temporarily represented withfairl y highpercentages . Heexplain sthi sb y assum ing that the tree still formed part of the forest vegetation of South Limburg during the Atlantic period, or that the high representation is due to certain vegetation conditions as an indirect result of which the proportion of pollen conveyed over a long distance became over-represented in the pollen rain. We prefer a different explanation and assume that peaty material originating from an olderpea tlaye r containing Boreal spectra isinserte d inth e soilprofil e at thedept h in question. Thiscoul d beth e result of one or more ofth e inundations to which the peat moor was so often subjected during formation. Peaty material from elsewherema yhav ebee nerode d andredeposite da t thesit eo fth e Etenaken profile (cf. VAN ZEIST 1958/59, p. 24). Intercalation of an older peat layer in a later peat deposit is a fairly common feature. It is occasionally found near the coast and near or in (former) lakes. Obviously it may also occur in the flood area of a river. In thisconnectio n wewoul d point out that someo f JANSSEN'Sdiagram s show a gap, also presumably as a result of erosion. It isclea rfro m theAtlanti cspectr a belowan d aboveth esectio ni nth e Atlantic zonejus t discussed that a mixed oak forest occurred in the fairly close vicinity. Its most important tree was Tilia, particularly in the later Atlantic period. But during the earlier part dominance of this tree may also be assumed, despite the fact that the spectra suggest a moderate dominance of Ulmus. It should be remembered, that unlike the other tree species, Tilia is insect-pollinated and consequently has a relatively low pollen dispersion capacity, which is why it is usually very much under-represented in pollen diagrams. During the later part of the Atlantic period the Alnetum gained ground in the surroundings of the profile. As increasing Alnus percentages are a common feature during this period (see above, p. 53) the most plausible explanation is that near the valley floor the mixed oak forest was superseded by the Alnetum owing to the increasing humidity of the soil. But JANSSEN'S diagrams show that local alder growth on carr peat may also have occasionally contributed to a higher proportion of Alnus pollen in the pollen rain.
54 Meded. Landbouwhogeschool Wageningen 80-8(1980) The two Atlantic spectra in the Vroenhof diagram, representing the older and later stage respectively, give a fragmentary picture of a similar vegetation de velopment during the Atlantic period. But unlike the Etenaken diagram they show relatively high Pinus and Corylus values. The same can be seen in the Sub- boreal spectra above. Now these spectra are all present in clayey peat or fluvial loess,unlik eth ecorrespondin g spectrai nth eothe r profile wherefluvia l sedimen tation did not occur in any form. This iswh y we assume that the relatively high Pinus and Corylus percentages are largely due to secondary pollen deposition, i.e. from the flood water. It is striking that the Atlantic spectra rich in Pinus in JANSSEN'S diagrams (see above) also originate from soil layers deposited in an aquaticenvironment , viz.peat y gyttja and alderpea t mixed withcalcareou s lake sediment. In this connection we would also point out that relatively high Pinus percentages are more commonly found in fluvial sediments in the Netherlands.
4.5. SUB-BOREAL ZONE
In contrast to the common picture the Etenaken and Vroenhof diagrams do not show aclea r retreat of Ulmusand/o r Tiliaa t the transition from the Atlantic to the Sub-boreal zone.Th e most conspicuous feature of the Sub-boreal zone of the Etenaken diagram is that non-arboreal pollen again dominates arboreal pollen, a position it had lost since the Early Holocene. Fagus now appears, but otherwise thezon e closely resembles thearborea l section of theAtlanti c spectra. The later part of the Sub-boreal period is absent, tallying with the assumed erosion of the fluvial loess and the upper peat layer (seep . 48).Late r Sub-boreal spectra are in fact found in the Vroenhof diagram in which Quereus move su p to higher percentages, showing a peak at 1.30 m. Tilia and Ulmus recede to low values,althoug h thelatte rmove su pagai na tth eto po fth ezone .Fagus i spresent . Carpinusi sonl y found once,viz .i n the topmost Sub-boreal spectrum (cf. p. 56). The high representation of the non-arboreal pollen in the Vroenhof diagram is accompanied by the representation of cultivation indicators, viz. cereals and someagricultura l weeds.Thes e occur from the bottom spectrum of thezon ejus t below the fluvial loess deposit. The coincidence isgoo d evidence that the forest vegetation declined as a result of reclamation and cultivation of the forested soil in the vicinity. The percentages for the Compositae-Liguhflorae are high in the fluvial loess. This may partly be due to the fact that pollen of this sub-family originated from a secondary pollen source, viz. the eroding cultivated soil, in which it may have occurred in high concentrations (seebelow , p. 57). The same may also be true of the pollen of the other cultivation indicators. The above theories do no apply to the Etenaken diagram. The absence of cultivation indicators could mean that during the earlier part of the Sub-boreal farming was only carried out either at a great distance or on a very small scalei n the profile environment. For therelativel y highPinus an d Coryluspercentage si nth eVroenho f diagram we refer to the discussion of the Atlantic zone in the same diagram.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 55 The Etenaken diagram shows that the change in the composition of the Q.M. continued during the Sub-atlantic. According to FIRBAS (1949, p. 170) a final advance of the oak may be ascertained in landscapes during the Middle Ages (Zone Xa), where, having regard to other observations, an intensified use or destruction ofth efores t hast ob eassumed . Heassume stha t suchactivitie smigh t also have occurred, now and then, during early-historic and prehistoric eras. To some extent this may explain the advance of Quercus during the Sub-boreal period shown in the Vroenhof diagram.
4.6. SUB-ATLANTIC ZONE
The Sub-atlantic zonei srepresente d in theEtenake n diagram only. It entirely coincides with the colluvium. In addition to pollen sedimented from the air this deposit naturally containspolle n redeposited together withth eerode d loess. The bottom 7 cm may also contain pollen displaced secondarily with the peaty material at that depth in the colluvial loess. Despite these complications the pollen curves are normal. Fagus shows an advance but subsequently recedes again. Carpinus is com pletelyabsent ,a featur e alsosee ni nman y Sub-atlanticspectr ai nJANSSEN' S(1960 ) diagrams. He believes this to be due to a marked human influence on the forest composition but, as mentioned above (p. 49) differential decay of Carpinus pollen in the colluvial material may also have been a factor. Corylusan despeciall yth eQ.M . becomemor eimportan t than inth e preceding zone. Quercus retains its great dominance in the Q.M. The Alnus and Pinus curves show a marked shift :th e first recedes and the other advances. The non- arboreal pollen, whichinclude sa greatdea l ofcultur epollen , isno w represented with very high percentages. Itma y beinferre d from thedeclin ei nAlnus valuestha t duringth e Sub-atlantic period the lower humid soils near or on the valley floor were also used for farming, the alder carr being cleared to make room for grassland. The advance of the Pinus curve must be an indirect effect of the decline of the forest vegetation in the region in the course of the Sub-atlantic period (cf. JANSSEN 1960), declining production and dispersion of the regional tree pollen resultingi na relativeris ei ntre epolle nconveye d overa lon gdistance .W ed o not think itcoul d reflect an absolute increase inth eprecipitatio n of Pinuspolle n due to more recent pine plantations because it is hardly likely that the 2.20 layer of colluvium at the bottom of which the Sub-atlantic spectra were found, would have been deposited over its entire depth since such a late date.
56 Meded. Landbouwhogeschool Wageningen 80-8(1980) 5. SOME CONCLUSIONS
Thelimite d scope of theinvestigatio n prevents usfro m drawing any definite conclusions.W ebelieve ,however ,tha ta clos econnectio nbetwee nth etyp eo fth e Holoceneminera lvalle yfil lan d theag eo fth epolle n spectranea rth ebotto mo f thisdeposit ,a sfoun d duringthi sinvestigation , mayb emor egenerall yfoun d in the Geul valley. Relatively late spectra can always beexpecte d in the colluvial loess,an d in thefluvia l loesssuc h spectra willi nan ycas eb efoun d in thebasi n areaswher ei t coversa mor e orles speat y substratum. Theformatio n ofthi sfluvia llaye rbega ndurin gth eperio dwhe nth elandscap e was only cultivated to a very limited extent, i.e. the Sub-boreal period, the colluvialdeposit sbein gforme dlate rdurin gth eSub-atlanti cperio dan dprobabl y mainlysinc eth eRoma ner awhe nth eregio nwa sextensivel yfarmed .I nthi scon nectionw erefe r toth eresult so f RIEZEBOSan d SLOTBOOM'S(1974 )palynologica l investigationo fsoi lmateria lfro m slopescovere dwit hloes si nLuxemburg .The y concludedtha tther eals ocolluviatio ndi dno tbecom eimportan t beforeth eSub - atlantic period. Although the Tiliapercentage si nth eSub-atlanti can d later Sub-borealspec tra areofte n low,the yma ynevertheles sindicat etha t limeplaye d acertai n part inth edevelopmen to fth efores t vegetationdurin gth emor erecen tHolocene .I n thisconnectio n wemus tagai n remember thetree' srelativel ylo wpolle n disper sioncapacity . Butinformatio n on theexistenc eo fforme r limeforest s mayals ob eobtaine d from aquit edifferen t sourcetha n palynologicalstudies ,viz . from ancientdocu ments and place-names (S. VAN DERWERF ,ora l comm. 1979).Som e mediaeval documents show that remnants of limeforest s must haveexiste d as late as the MiddleAges .Th e sameconclusio n may bedraw n from certain place-namesi n South Limburg, e.g. the village name 'Terlinden'1 and the farm-road name 'Mheerlindje'1 found near the village of Mheer. Place-names containing the element 'linde' are also found in the Netherlands outside the South Limburg loessarea ,i n regions whereth e soili sforme d bycove r sand. It isinterestin g to notetha thig h Tiliapercentage sar efairl ycommo ni nth eAtlanti co r Sub-boreal zone ofpolle n diagrams from soilprofile s whichdevelope d incove r sand. The high Compositae-Liguflorae percentages in the upper mineral part of bothprofile s deserveparticula r attention. Thesehig hvalue sshoul d not leadu s todra wan yconclusion sa st oth eproportio no fthi ssub-famil yi nth evegetation s represented. HAVINGA(1962 , 1971) and BOTTEMA(1975 )showe d that highcon centrationso fLiguliflora e pollenar efoun d fairly frequently inminera lsoi la sa resulto fdifferentia l pollenpreservatio no rdepositio nb ydigger-bees .Th elatte r author notes that environments created or influenced by man (and farmers in particular) often favour theexistenc eo f the bees.I t may therefore beassume d that part of the Liguliflorae pollen in question originates from cultivated soil
'Linde'i sth eDutc hfo r 'limetree' . Medea. Landbouwhogeschool Wageningen 80-8(1980) 57 rich in Liguliflorae pollen. It must have beencarrie d and redeposited together with the eroded soil material which now constitutes the fluvial and colluvial deposits.
6. SUMMARY
A tentative investigation was carried out into the age of the fluvial and colluvialloes sdeposit si nth evalle yo fth erive rGeu li nSout hLimbur g (Nether lands).Tw osoi lprofile swer estudie dpalynologically ,on econsistin go fa laye ro f fluvial loesso npea tan dth eothe r ofcolluvia lloes so npeat .I twa sfoun d that at the siteo f theprofile s the depositsbot h date from the era at which agriculture had already made its mark on the landscape. The fluvial sedimentation began during the earlier part of the Sub-boreal period, when the landscape was only cultivated to a very limited extent. The colluvial deposit is of later date, being formed during the Sub-atlantic period, probably mainly since the Roman era whenth e soilbecam e intensively cultivated. Special attention ispai d to the significance of the Tilia percentages and the highCompositae-Liguliflora e percentages in thediagrams .
7. SAMENVATTING
EERTIJDSE VEGETATIE EN SEDIMENTATIE IN HET DAL VAN DE GEUL
Eri see nvoorlopi gonderzoe kuitgevoer dnaa rd eouderdo mva nd efluviatiel e encolluvial elössafzettinge n in het dal van deGeul .Daarbi j zijn twee profielen palynologisch onderzocht, respektievelijk bestaande uit fluviatiele losso pvee n encolluvial elos so pveen .He t bleek dat debeid elössafzettinge n terplaats eva n deonderzocht e profielen stammen uit detijd , waarind elandbou w zijn stempel reedso phe tlandscha pha dgedrukt .D efluviatiel esedimentati ebego ntijden she t ouderedee lva nd esubboreal etijd ,toe nhe tlandscha pno gmaa ri nzee rbeperkt e matei ncultuu rwas .D ecolluvial eafzettin g vondplaat si nd esubatlantisch e tijd, waarschijnlijk vooralseder td eRomeins etijd ,toe nd ebode mreed sintensie fwer d bebouwd. De aandacht wordt gevestigd op de aanwezigheid van verschillende toponymeni nd estreek ,waari nd enaa mva nd elindeboo mteru gi st evinden .Di t wijsterop ,da td elind eno gi nd eMiddeleeuwe nee nnie tt everwaarloze nro li nd e vegetatie speelde. Delag e 77/z'a-percentagesi n de subboreale en subatlantische pollenspektra van de beide diagrammen zijn mede een gevolg van de reedsz o vaak vermeldeslecht everspreidin gva nhe t stuifmeel van dezeboom .
58 Meded. Landbouwhogeschool Wageningen80-8 (1980) De hogeCômpositen-waarde n kunnen op indirekte wijze met de landbouw- cultuur inverban d worden gebracht.
REFERENCES
BOTTEMA, S. 1975. The interpretation of pollen spectra from prehistoric settlements (with special attention to Liguliflorae). Palaeohist., 17: 17-35. FIRBAS, F. 1949.Spät -un d nacheiszeitlicheWaldgeschicht e Mitteleuropas nördlich derAlpen . Erster Band: Allgemeine Waldgeschichte, Jena: 480 pp. HAVINGA, A. J. 1962. Een palynologisch onderzoek van in dekzand ontwikkelde bodemprofielen. Diss. Wageningen: 165pp . — 1963. A palynological investigation of soil profiles developed in cover sand. Meded. Land bouwhogeschool, Wageningen, 63 (1): 93 pp. (Abstr. of diss.). —1971. An experimental investigation into thedeca yo fpolle nan d sporesi nvariou s soiltypes .I n: J . BROOKS,P .R . GRANT, M.D . MUIR, P. VANGIJZE L and G. SHAW (Editors),Sporopollenin . Acad. Press, London: 446-479. JANSSEN, C. R.1960 . On the Late-Glacial and Post-Glacial vegetation of South Limburg (Nether lands). Diss. Utrecht: 112pp . RIEZEBOS, P.A . and SLOTBOOM, R. T. 1974.Palynolog y in the study ofpresentda y hillslope develop ment. Geol. Mijnb., 53(6) :436-448 . WOILLARD, G. 1975.Recherche spalynologique s sur lePleistocen edan sl'es td ela Belgiqu ee tdan sle s Vosges Lorraines. Acta geogr. Lovaniensia, 14: 118pp . ZEIST, W. VAN. 1958/1959. Palynologische Untersuchung eines Torprofils bei Sittard. Palaeohist., 6/7: 19-24.
Meded. Landbouwhogeschool Wageningen 80-8(1980) 59 ACKNOWLEDGEMENTS
Wear egratefu l toal lwh ohelpe du si nth esoi lsurve yan dlaborator ywor k: to ourcolleague so fth eDepartmen t ofSoi lScienc ean dGeolog yfo r theirstimula ting discussions, in particular Prof. J. D. de Jong and Dr. D. Nota; to Prof. Brunnacker of the Geological Institute of the University of Cologne for the discussion ofsedimentatio n processes innarro w valleyso fsmal lrivers . Toth egraduat e studentswh ostudie d specialproblems . Toth etechnica lan d administrative staff ofou rdepartment ,i nparticula r Mr. P.G . M.Verstee gwh omad eth edrawing san d pollendiagram s and Mrs.C .B . Beemster and Mrs. M.H .J . Miedema for typing the manuscripts. To Mr.J . E.Wilso n (Luxemburg) for correctingth e English and last but not least to all those students without whosefield wor k thepaper s could not have been written.
60 Meded. Landbouwhogeschool Wageningen 80-8 (1980)