Geological mapping by geobotanical and …publikacio.uni-miskolc.hu/data/ME-PUB-25593/fulltext.pdfGeological mapping by geobotanical and geophysical means: a case study from the Bükk

Cent. Eur. J. Geosci. • 1(1) • 2009 • 84-94DOI: 10.2478/v10085-009-0003-x

Central European Journal of Geosciences

Geological mapping by geobotanical and geophysicalmeans: a case study from the Bükk Mountains (NEHungary)

Research Article

Norbert Németh1∗, Gábor Petho2†

1 Department of Geology and Mineral Deposits, University of Miskolc, 3515 Miskolc-Egyetemváros, Hungary

2 Department of Geophysics, University of Miskolc, 3515 Miskolc-Egyetemváros, Hungary

Received 7 November 2008; accepted 31 January 2009

Abstract: Geological mapping of an unexposed area can be supported by indirect methods. Among these, the use ofmushrooms as geobotanical indicators and the shallow-penetration electromagnetic VLF method proved to beuseful in the Bükk Mountains. Mushrooms have not been applied to geological mapping before. Common specieslike Boletus edulis and Leccinum aurantiacum are correlated with siliciclastic and magmatic formations whileCalocybe gambosa is correlated with limestone. The validity of this correlation observed in the eastern part ofthe Bükk Mts. was controlled on a site where there was an indicated (by the mushrooms only) but unexposedoccurrence of siliciclastic rocks not mapped before. The extent and structure of this occurrence were exploredwith the VLF survey and a trial-and-error method was applied for the interpretation. This case study presentedhere demonstrates the effectiveness of the combination of these relatively simple and inexpensive methods.

Keywords: Bükk Mts • geobotany • VLF • elongated structures • fungi© Versita Warsaw

1. Introduction

Geological mapping is often hindered by the a lack ofexposure when the bedrock is covered by thick soil anddetritus. Making artificial exposures (e.g., drillholes ortrenches) on the unexposed area is an expensive and time-consuming process. Furthermore, it can cause environ-mental damage. To avoid this, there are indirect meth-ods to gain information about the structural pattern of thearea. By mapping the eastern part of the Bükk Moun-∗E-mail: [email protected]†E-mail: [email protected]

tains two methods proved useful which were not used herebefore: the application of geobotanical indicators (mush-rooms in particular) and resistivity measurements basedon radio waves. This paper presents a case study wherethe joint use of these is demonstrated.It is a known fact that vegetation is linked to the bedrockof its habitat through the soil cover. Therefore, we candraw some conclusions about the character of the rocksfrom observations of this vegetation. The science dealingwith this, among other topics, is called geobotany. Themethod has been used for a long time (e.g. [1]), but theapplied multidisciplinary knowledge can discourage re-searchers. On different bedrocks, or rather in the soilsformed on them with different physical and chemical char-

84

Norbert Németh, Gábor Petho

acter, there will develop different typical associations evenwhen geographic conditions (climate, exposure, etc.) arethe same. The mapping of the natural distribution area ofthese species or, in the case of planted vegetation, map-ping of patterns of vividness can substitute for the map-ping of exposures of geological bodies. Therefore, it canbe useful for the mapping geologist to collect floristicalobservations during their field work, because these couldprove worthwhile in judging the poorly exposed parts ofthe area. These observations do not require supplementalinvestments and may spare expensive exploration projects.When one wants to know about the continuation of sur-face border traces at depth, the solution is a geophysicalsurvey. The method applied by the geological mappingshould be simple and quickly executable with a suitableresolution for the zone under the detritus cover. There isno need of deeper penetration than about a dozen metres.Different physical parameters of the rocks (like density,magnetic permeability, radioactivity, thermal conductivity,resistivity, elastic wave velocity) can be investigated bygeophysical methods. Radioactive, magnetic and resistiv-ity surveys are more easily undertaken than most othergeophysical measurements. Radioactivity measurementscannot be applied to this case, because no informationcan usually be expected on the underlying formations be-low the detritus cover due to absorption. Magnetic surveycan only be an efficient tool if the variation of ferromag-netic mineral content is significant. However, there is fre-quently no significant magnetic susceptibility contrast. Ifthere are lateral small-scale near-surface variations, thenelectromagnetic methods with frequent spacings are pre-ferred to geoelectric ones in general. These conditionsare met by the VLF method. It only requires at least oneoperating radio transmitter, a portable selective receiverinstrument and a single operator.The combined use of geobotany and shallow-penetrationgeophysics seems to be promising for detailed mappingof covered geological structures. It is especially true instratigraphically varied mountains or hill-country with acomplicated structure where the lithologic boundaries cansometimes be sharply marked in the vegetation.This is also the case in the Bükk Mountains. These moun-tains are situated in North Hungary, between the GreatHungarian Plain and the River Sajó (see Figure 1), andalthough their area is rather small, there are a lot of un-solved problems in its stratigraphy and structure. In theeastern part of the Bükk Mts. it can be observed thatthe spatial distribution of some mushroom species is cor-related with certain rock types. This is true for everycase where the rocks crop out. The aim of this study wasto show that these species can be used as geobotanicalindicators, by selecting a site without obvious outcrops,

Figure 1. Sketch map of Hungary with the locality of the Bükk Mts.

where the geobotany conflicts with previous geologicalmapping, and to demonstrate the use and efficiency of theVLF method in a geological environment with variouslydipping lithological contacts.Few shallow geophysical explorations have been madeearlier in the Bükk Mountains. Long and medium fre-quency radio waves had already been used by Takács [2]for mapping covered near-surface limestone with chang-ing topography in the Great Plateau. A seismic refractionsurvey was performed to determine the thickness of thesurface clay filling a doline in the Bükk Mts. [3]. In bothcases two media were assumed: conductive clay with lowseismic velocity overlying a resistive limestone of high ve-locity.This investigation was also prompted by the proximity ofthe study area to a spring swallowed back after a shortsurface runoff in a vulnerable karst area. As the watersupply of the town Miskolc is based on the karst aquifersof these mountains, the opinion of the authors is that moreknowledge of the geology of karst springs’ vicinity andtheir catchment area should be required in the future. Thismethod can be a part of the solution.2. StratigraphyIn several areas of the Bükk Mts. the frequent alternationof carbonate and non-carbonate stratigraphic units is typ-ical. In several areas, the individual rock types occur in arelatively small area: in a band or in a spot. In general,the differences in rock type can be seen in the topography:the more resistant limestone forms steep slopes and cliffs,while the more eroded strata between them form gentleslopes and local depressions covered by recent deluvium.The former provides well exposed areas; the latter have,in most cases, not even an artifical exposure, as on a gen-tle slope there is no need of a road-cut. However, thegentle slope alone is not a proof of the presence of anon-competent strata, as it also can develop over stand-

85

Geological mapping by geobotanical and geophysical means: a case study from the Bükk Mountains (NE Hungary)

Figure 2. Geological map of the vicinity of Köpüs spring com-piled by M. Forián-Szabó [4] and known habitats of bo-letus species. s,mC2: Szilvásvárad and Mályinka For-mation (shale, sandstone and laminated dark limestone);sP2: Szentlélek Formation (sandstone, evaporitic mud-stone); nP2: Nagyvisnyó Limestone Formation (bitumi-nous limestone); gT1: Gerennavár Limestone Formation(oolitic limestone); ftT2−3: Felsotárkány Limestone For-mation (laminated cherty limestone); kfT3: KisfennsíkLimestone Formation (light-gray massive limestone); snT3:Szinva Metabasalt Formation (laminated metavolcanite).Crosses indicate habitats of aspen bolete (Leccinum au-rantiacum), filled circles indicate habitats of king bolete(Boletus edulis).

ing bedrock. Nevertheless, if this assumption is supportedby the presence of certain plant species, we cannot avoidexploration of the spot’s rock type.Figure 2 is extracted from the geological map of the west-ern part of the Kis-fennsík (= Small Plateau) compiled byM. Forián-Szabó (coloured version published in [4]) andrepresents the vicinity of Köpüs spring. The stratigraphyof this area comprises an Upper Carboniferous to LowerTriassic succession, which is folded, faulted, thrusted bytectonic events and generally dips steeply. Clastic andcarbonate sediments are both present. The Upper Car-boniferous Mályinka Formation is an interbedded succes-sion of shale, sandstone and black laminated limestonewhich is tectonically cut into lenticular blocks. After a dis-cordance, the Upper Permian succession comprises conti-nental sandstone, then evaporitic mudstone and dolomite(Szentlélek Formation) followed by bituminous limestone(Nagyvisnyó Limestone Formation). The sedimentationcontinued without interruption into the Lower Triassicwith oolitic limestone (Gerennavár Limestone Formation),then the Ablakoskovölgy Formation containing alternat-ing shale-siltstone-sandstone-marl and limestone mem-bers. Eastwards from the Köpüs spring starting atKöpüs-ko (Köpüs rock), there is a nappe (known as“Kisfennsík nappe”) above this succession comprising lam-inated cherty limestone (Felsotárkány Limestone Forma-

tion) and massive light-gray limestone (Kisfennsík Lime-stone Formation) with small metavolcanic bodies (SzinvaMetabasalt Formation?). Detailed description of the Pa-leozoic units can be found in the work of Fülöp [5] andthat of the Mezozoic units of Haas [6].While the Carboniferous limestone blocks, the oolitic lime-stone and the massive limestone, are well exposed asthey form prominent features, other stratigraphic units arenot so easy to differentiate. This is particularly true forthe stratigraphically adjacent Szentlélek and NagyvisnyóLimestone Formations. Due to the multiple folding andfracturing around Köpüs spring the units occur in a com-plicated pattern, or they can even be cut out from thesuccession, which is not easy to map because of a lack ofexposure.3. GeobotanyAlthough most forests are planted and non-domestic conif-erous monocultures are frequent, other species connectedto the trees continue to form associations typical for thehabitat. The connection between the forest types andthe soil together with bedrock was investigated with the1:50000 vegetation mapping made in the 1950s by Zó-lyomi and his co-workers [7]. More detailed observationswere undertaken in the Southeastern Bükk Mts. and someother rock-association connections (mainly on dolomiteand rough limestone) were discovered by N. Less [8]. Hiswork was a part of the 1:10000 vegetation remapping ofthe whole Bükk Mts. [9], but unfortunately, it was not com-pleted because of his early death. The soil types of themountains and the vegetation typically formed on themwill be presented according to these works.The natural occurrence of the forest types and the speciescomposition depends essentially on altitude or the steep-ness and exposure of the slopes, and the microclimate ofthe area. The dependence on the soil characteristics isbest observed in the underwood, but sometimes the treespecies of the association can also change with them.For example, the hornbeam is prolific over limestone andgrows rapidly on clearings and former pastures, but overmetavolcanic rocks it is significantly less prolific.The soil of the Bükk Mts. developed from the detritusof the bedrock and contain grains of their material. Ondeluvial slopes they can cover-up exposures of other rocksor mix with their detritus by creeping or slipping down-wards. Different covering materials can rarely be found(remnants of an eroded Cenozoic sediment cover: extrane-ous pebbles, red clay and loess-type sediments), mainlyin sediment-traps, for example in dolines [10], althoughthe red clay is frequently distributed in the Southeastern

86

Norbert Németh, Gábor Petho

Bükk Mts. and can form very thick layers above carbon-ates. There are two main soil types in the Bükk Mts. withseveral variations. The pure limestones are covered withrendzina with a transition to brown forest soil over redclay, while the typical cover of the clastic sediments andmetavolcanic rocks is the non-podzolised brown forest soilwith lessivage. The non-waterexigent species that needcalcareous soil prefer the first soil type, the species de-manding an acid soil with good water-bearing capacityprefer the second one. According to the observations ofthe authors in road cuts, on steep slopes the different soiltypes border on each other rather sharply, with a 1-2 mtransition.The beechwood with sweet woodruff (Asperula odorata) isthe typical forest type of the calcareous soils on thick soilof good quality, on thin soil of poor quality the beechwoodstands with wood melick (Melica uniflora). Towards thewarmer zones the beech is substituted with hornbeam andoak. On slopes with rocks and boulder’s dog’s mercury(Mercurialis perennis) and bishop’s weed (Aegopodiumpodagraria) are the typical underwood members. The lightacidity of the soil is indicated by the bulk appearance ofbeech sedge (Carex pilosa). The soil of carbonate bedrockis preferred by several spectacular flowers like (accord-ing to observations of the authors) martagon lily (Liliummartagon) and oregano (Oreganum vulgare).As the higher parts of the Bükk Mts. are composed mainlyfrom limestones, the percentage of metavolcanic rocks, ra-diolarites, shales or other silicate rocks is rather small,the beechwood types characteristic of acid soils are com-paratively rare in the inner part of the mountains. Be-yond the above-mentioned beech sedge, the main indica-tors are white wood-rush (Luzula albida), common speed-well (Veronica officinalis) and bilberry (Vaccinium myr-tillus), which is infrequent in the Bükk Mts. The indicatorof the water in local depressions and in clay soil is thebulk appearance of wood sorrel (Oxalis acetosella) andsome ferns, mainly on soils of non-carbonate rocks or al-luvium [7, 8].On the explored site, at the Köpüs spring there is ameadow with aspen and birch groves on its southern andeastern border. On the vegetation map ([9], Figure 3) thereis a meadow around the spring. The bulk of the forestis beechwood with underwood typical of soils formed onlimestone. The vegetation on the western part consistsof species characteristic of limestone, such as the forestsin the neighbourhood and moisture lovers which live onthe eastern part beside the brook. This vegetation pat-tern does not show the exposure of the rock differing fromlimestone. Still, there is a visible element indicating thenon-calcareous, acid soil: the occurrence of some mush-rooms, namely boletes.

Figure 3. Simplified cut-out of Less Nándor’s vegetation map [9].

4. Mushrooms used as geobotani-cal indicatorsMushrooms have not been the subject of vegetation map-ping. The mushroom genera and species mentioned beloware classified according to a specialized book on the fungiof Hungary [11], while the findings concerning their distri-bution and connection to rock types are based on decen-nial observations of the authors based on a much largerpart of the mountains than presented here.Fungi usually live hidden in the soil or in the tissue ofother living beings, so the only visible parts that can

87

Geological mapping by geobotanical and geophysical means: a case study from the Bükk Mountains (NE Hungary)

Figure 4. Young individuals of orange-cap or aspen bolete (Lec-cinum aurantiacum) from Szentlélek.

be used for mapping are the sporophores. Accordingly,mapping of their distribution is only possible during thetime interval of abundant growth. There are many edi-ble species among them growing in huge numbers duringfavourable weather, which are gathered preferentially byconnoisseurs. Certain fungi are saprophytes, others aresymbionts or parasites of living trees linked to certaintree species. For example, the death cap (Amanita phal-loides) is a micorrhizal mushroom of oaks, while milkyagaric (Lactarius deliciosus) grows only in coniferouswoods. Most of them are moisture lovers and some ofthem need light. Some fungi live in meadows and pasturesonly. From a geological point of view the most interest-ing species are the ones sensitive to soil characteristics(e.g., water-retaining capacity, acidity, Ca-content). Onthe Kis-fennsík the species listed below proved to be in-dicators of certain soil and rock types.In the Bükk Mts, St. George’s mushroom (Calocybe gam-bosa) lives only in rendzina soils developed on pure, non-clayey limestone. Here it is an indicator of the KisfennsíkLimestone Formation and Gerennavár Limestone Forma-tion eastwards and southwestwards from the area on Fig-ure 2.King bolete (Boletus edulis) prefers acid soils and avoidscalcareous soils, therefore it is an indicator of strata con-taining no carbonates, mainly of sandstones and metavol-canic rocks.Orange-cap or aspen bolete, (Leccinum aurantiacum)(Figure 4) is a micorrhizal mushroom of aspens as wellas other Leccinum species of birches and/or hornbeams.The aspen-birch groves generally grow on non-carbonaterocks, but the appearance of aspen boletes (like king bo-letes) always suggests eluvium of non-carbonate rocks ornon-calcareous covering sediment.In addition to these, the bulk appearance of some other

mushroom species can indicate rock types, although (be-cause they do not grow exclusively in this soil) they oughtto be assessed carefully. Clayey, acid soil is generallyindicated by the chanterelle (Cantharellus cibarius), Rus-sula and Amanita genera or Lactarius piperatus. As theappearance of mushrooms is also linked to factors otherthan soil (for example, the aspen bolete to the aspen trees)they are not necessarily distributed over the whole areaof a certain rock type so the lack of the mushrooms cannotbe interpreted as the lack of the rock under consideration.The habitats of the boletes around the Köpüs spring ap-pear to be the forests above sandstone, siltstone and shale(Figure 2). Boletes occur in the vicinity of the area onFigure 2 not only on soils above the Szentlélek Forma-tion, but also over other stratigraphic units comprisingnon-calcareous acid rocks. Mainly the ones that includesandstone, for example on shales and sandstones of theMályinka Formation, on siltstones and sandstones of theAblakoskovölgy Sandstone Member and on metavolcanicrocks of the “Kisfennsík nappe” at Barátság-kert. Theexposures of these formations are indicated with well-defined outlines in some places by the habitat of the bo-letes.However, there was an occurrence at the Köpüs springwhere the geological map indicated limestone only, al-though both king bolete and aspen bolete live there. Butthe small terrace formed on the slope and the spring aris-ing there and swallowed up below it suggested an outcropof the non-competent, watertight strata of the SzentlélekFormation in a core of an anticline or along a thrust faultover which the water falls. Without exposed rocks, basedon the scant debris on the surface, this was impossibleto prove, because all perceptible grains came from lime-stone (not counting quartzite pebbles dragged in duringconstruction of the spring). So, for the mapping of the in-dicated occurrence, it was necessary to find an explorationmethod which enables us to trace the detritus-covered for-mation boundaries without exposures.5. Features of the VLF method

The VLF (very low frequency) method is an electromag-netic geophysical method for detecting conductive and/orresistive zones located at depths of about a dozen metres,(e.g. [12, 13]) and therefore it can be applied for geolog-ical mapping. This method utilizes the carrier waves ofsome very powerful VLF stations located at several pointsaround the globe. They broadcast at frequencies between15 and 30 kHz for communicating with submarines. Theantenna is usually a grounded vertical electric monopolewith high power broadcast. At distances considerably88

Norbert Németh, Gábor Petho

greater than one wavelength the combination of groundwave (travelling along the Earth’s surface) and sky wave(propagating in the space between the Earth’s surface andthe first ionized layers of the upper atmosphere by re-flections) can be utilized. At large distances (more thanabout 800 km from the transmitter) the second mode is thedominant one. For the purpose of a survey, the incidentEM field can be considered to be uniform within a smallarea. The incident magnetic field only has a horizontalcomponent (HΦ), which is perpendicular to the transmitterbearing. The incident electric field has both vertical andhorizontal components, where the latter is in the directionof the transmitter bearing. Due to the almost infinite con-ductivity of the ground compared to air, the refraction ofthe VLF wave is irrespective of its angle of incidence andresults in a vertically downwards propagating wave withunchanged horizontal magnetic field component and a ra-dial electric field component (Er). The skin depth, whichis the distance in which the amplitude of the surface EMfield is reduced by 1/e, that is, to 37% of the surface value,expresses the attenuation of the EM field. This depth (p)depends on frequency (f) and on conductivity (σ ) as fol-lows [14]:p = 12π

√107fσ . (1)The skin depth can also be correlated with the penetrationof EM waves, however, it cannot be stated that the explo-ration depth is equivalent to the skin depth. The largerthe resistivity contrast between the underlying layer is,the greater is the probability of the detection for the lowerlayer situated at the proximity of the skin depth. Apartfrom the difference between the EM source of magnetotel-lurics (MT) and VLF, the same physical phenomena canbe observed, so the basic equation of MT [14] can be ap-plied to determine the resistivity of the homogeneous half-space. From the VLF wave impedance – which is the ra-tio of the radial electric and the horizontal magnetic fieldcomponent – the resistivity of the homogeneous half-space(ρ), or using the same relationship, the apparent resistivityof the inhomogeneous ground (ρa) can be defined as:

ρa = 12πfVLFµ0∣∣∣∣ ErHφ

∣∣∣∣2 , (2)where fVLF stands for the frequency and µ0 denotes the ab-solute permeability of vacuum. In the carrier wave trans-mitted into the homogeneous ground, the horizontal ra-dial electric field component leads the horizontal magneticfield perpendicular to the transmitter bearing in phase by45o. Practically, when the thickness of a homogeneouslayer on the surface is more than two or three skin depths,the situation is identical to a homogeneous half-space. It

means that phase difference between Er and HΦ can fur-nish information about changes of conductivity to a depthof some tens metres. The definition for phase [14]:φ = arctan[ Im (Er/Hφ

)Re (Er/Hφ

)] . (3)Assuming a horizontally stratified half-space consistingof two layers within the skin depth, with a lower layer tobe more conductive compared to the upper one, one findsthat the phase angle Φ(Er ,HΦ) is generally more than 45°,and it is less than 45°if the conductivity ratio is inverse.In these homogeneous situations the measured values areindependent of the transmitter bearing.In case of inhomogeneity – when the geology is differentfrom the homogeneous or horizontally stratified half-space– it does not hold: both the apparent resistivity and itsphase depend on the mutual position of transmitter bear-ing and the direction of structural elements. Besides ap-parent resistivity and phase yielded by VLF R method pa-rameters of the polarization ellipse of the resultant mag-netic field are also provided in tilt angle mode by VLFinstruments. This latter method can also help geologicalmapping, mainly in the case of significant near-surfaceinhomogeneity contrasts.If the geologic structures are elongated, like a steeply dip-ping fault plane or a folded bedding plane, and the for-mations on opposite sides of these planes can be charac-terized with different conductivities, then these structurescan be approximated with 2D (two-dimensional) models.Restricted to 2D conductivity inhomogeneities, there aretwo modes based upon the angle between the transmitterbearing and the structural strike. If they are parallel toeach other, the term of E polarization or TE mode is used.If the structural strike is perpendicular to the transmit-ter bearing, the case is named as H polarization or TMmode. In TE mode, current channelling in the conductivepart can be detected. In TM mode, the galvanic effect canbe observed. As a result, a secondary electric field fromthe oscillating electrical charge accumulation on the in-terfaces perpendicular to the primary electric field can bemeasured. For this reason TM mode is an effective toolto locate near-surface vertical or nearly vertical contacts.6. Exploration of the Köpüs springoutcrop6.1. Material types of the survey areaFor the sake of proving the existence of the SzentlélekFormation rocks at the spring, we took samples from theclayey detritus below the humus with hand-driven auger

89

Geological mapping by geobotanical and geophysical means: a case study from the Bükk Mountains (NE Hungary)

drilling (Figure 5). From the two samples on the right sideof the brook near the habitat of the boletes (site 1 0.6–1.4 m; site 2 0.3–1.7 m) we collected greenish, flat grainstypical of the Szentlélek Formation Garadnavölgy Evap-orite Member. The clay was yellowish-greenish colouredand partly rust stained. After decanting and sieving thesamples and examining them under a microscope, it be-came evident that the material comes from this memberas it hardly contained any carbonate grains, while greenmudstone and quartz fragments were apparent. In the0.25–0.5 mm fraction it contained a few grains of hexa-hedric pyrite, which is typical of no other stratigraphicunits in the vicinity. Rust stains of the clay may originatefrom the secondary iron-oxide minerals after oxidation ofpyrite.

Figure 5. Site map of the surroundings of Köpüs spring. The num-bered crosses indicate soil sampling sites. Numberedlines with teeth mark VLF profiles with measuring stations.

On sampling site 1 between 0.3–0.6 m there was a stra-tum of cream-coloured grains with lime accretion whichproved to be the alluvium of the brook with travertine pre-cipitation comprising mainly dark limestone grains. In thesample collected from the left side of the brook this ma-terial reaches a larger thickness (site 4 0.4–1.3 m) underthe humus layer. This explains why boletes do not livehere, despite the fact that the aspen grove is continued.We also collected a sample in the meadow westward abovethe aspen grove (site 3) but the drilling broke down afteronly 80 cm at a limestone rock. At that point it went inreddish brown clay with small grains of bituminous lime-stone which can be accounted for as material of the debriscover formed above the Nagyvisnyó Limestone Formation.

Figure 6. VLF apparent resistivity map of the survey area.

Figure 7. VLF phase map of the survey area.

6.2. VLF measurements in the surroundingsof Köpüs spring

All measurements were made with a Geonics EM16R in-strument. At the beginning of the geophysical surveythere were two operating VLF transmitters: GBR (GreatBritain, Rugby, f=16.0 kHz) and JXZ (Norway, Helge-land, f=16.4 kHz). From the first measurement lines (1, 2,4 in Figure 4) laid along 110°-290°(which corresponds tothe GBR bearing) sharp changes in the phase angle witha nearly NE-SW striking line were observed with bothtransmitter stations. So GBR was supposed to meet theTM mode condition better than JXZ in this survey becausethese changes predicted the strikes to be more nearly per-pendicular to GBR than to JXZ bearing. Further measure-ment results strengthened this assumption and this trans-mitter was generally used. Moreover, GBR was active(with short intermissions) during all measurement sessionswhile other stations transmitted only episodically.90

Norbert Németh, Gábor Petho

The surroundings of the spring were investigated along 6lines parallel to the transmitter bearing (Figure 5). Ef-fort was made to have continuous profiles, however, ow-ing to some natural obstacles it was not always possible.VLF resistivity and phase measurements were made us-ing a 10m station interval which was also equal to theinterprobe (or electrode) spacing. Figure 6 shows the ap-parent resistivity contour map in ohmm, Figure 7 presentsthe areal distribution of the phase difference between (Er)and (HΦ) in degrees. The apparent resistivity and phasedata show similarities for neighbouring profiles. This cor-relation is more pronounced in the case of the phase map,because apparent resistivity is sensitive to surface inho-mogeneities, while the phase is hardly influenced by it.On the basis of the two contour maps it can be statedthat the main structural strike is parallel to the NE-SWdirection, and although the explored geologic structurescannot be considered as a pure 2D situation, it can be agood approximation.6.3. Modelling and interpretation

In the course of geophysical EM data interpretation, thetask is to determine the conductivity structure that has thesame EM response as the measured one. For interpreta-tion, different inversion algorithms are used and amongthose, robust methods are preferred nowadays (e.g. [15]).Taking into consideration the number of measured EMparameters and the number of the unknowns it is obviousthat the problem of underestimation can not be avoided.This is why the trial-and-error method was applied.The essence of the trial and error method is to solvethe forward problem several times in order to get an ac-ceptable agreement between the measured and the modeldata. In the course of successive model modification ourintention is to minimize the error, i.e. the difference be-tween the model response and the measured data. Inthis case, for forward modelling, a finite difference twodimensional magnetotellurics (2DMT) code developed forH-polarization was applied. This forward modelling as-sumes that the vertical section perpendicular to the strikeis divided to rectangular elements by a grid [16]. Thesizes and conductivity values are defined for every singlecell as input data. For each grid point numerical solu-tions are achieved by approximating the relevant differ-ential equation (time independent Helmholtz equation forboth polarizations) by a finite-difference equation takinginto account the boundary conditions. The solution to theset of these equations is the strike-directional magneticfield component solution for each gridpoint and the out-put of modelling is the apparent resistivity and the phasebetween Er and HΦ in the surface gridpoints. The finite

difference solution is preferable for simulating complex ge-ology. In our case the applied grids had 39 columns and49 rows.For the sake of having the resistivities of the formationsneeded for the model determination, we made VLF-R mea-surements on known outcrops with sufficient thickness notfar from our survey site. We distinguished three forma-tions: bituminous limestone (Nagyvisnyó Limestone For-mation) with a resistivity of 360 ohmm, mudstone (Garad-navölgy Evaporite Member) with a resistivity of 40 ohmm,and the covering detritus. This detritus is very inhomoge-neous, so it was characterized with two resistivity values:60 ohmm for the detritus consisting mainly of clay, and120 ohmm for the other type with limestone boulders.The aim of the interpretation was to determine the thick-nesses of the two-to-four underlying formations, usuallywith dipping boundaries when only two data were mea-sured for each station. This task can be considered as anunderestimated inverse problem as was mentioned before,so the solution is not unique. However, not all solutionsare equally probable as geological models. The way toget the solution was a classical trial-and-error method:the parameters of the model were changed as long as thebest fit between the measuring data and computed datawas reached along the profiles. The start model for thelines was constructed with horizontal and vertical bound-aries only on the basis of the measured data and thepredicted formations. On sections with Φ < 45°a moreconductive layer was placed over a less conductive layer,and vice versa on sections with Φ > 45°. Where therewas limestone on the surface along the section, we putlimestone layer on the top; where the habitat of the bo-letes was, mudstone was chosen as the uppermost layer.When changing the model it was transformed toward ge-ologically possible situations, that is, to gradual changesin thickness and depth with continuous, dipping, but notsharply breaking boundaries. After some dozen trials weachieved an acceptable agreement between measured andcomputed data (see Figure 8).Further modification was not made on the model becauseof limits of measurements and those of the modelling pro-cess. The measurement itself contains some measuringerror, usually not more than 5%. Topographic effects canbe an additional source of VLF survey error (e.g. [17]).From this point of view of the survey site, the horizontalmagnetic field and the contours of topography were mainlyparallel to each other and there was no significant topo-graphical slope change. For limits of the computing, thequestion of dipping boundaries and the fixed values of for-mation conductivities should be mentioned. This programapplies rectangular grid elements with fixed conductivityvalues as input data. If there are two cells with different91

Geological mapping by geobotanical and geophysical means: a case study from the Bükk Mountains (NE Hungary)

conductivity in contact, the linear change of conductivitybetween the centre of grid elements is assumed in thecomputation. In this way, by using sufficient number offine grid elements, the gradual change of boundary slopecan be approximated. However, even this resolution is notenough to model the very small-scale near-surface inho-

mogeneities (like rock blocks in the detritus or cavities inthe weathered zone) which may influence the measuredapparent resistivity data. The resistivity of bituminouslimestone and that of mudstone were taken as constant,but actually they are average values, just like in the caseof the covering detritus.

Figure 8. Measured (circles) and computed (crosses) apparent resistivity and phase data of line 6 and line 7-8 (see Figure 4) with the interpretedgeological cross-sections.

The modelled and the measured ρa and Φ data are pre-sented for the two outermost lines in Figure 8. The data ofthe remaining lines form a transition between them. Theinterpreted cross-sections according to the final models(which are similar even for the outermost lines) show asyncline-anticline pair covered with an eluvium of chang-ing thickness. The divergence from the 2D model arisespartly from the differences of this thickness and partlyfrom the lateral change of the fold geometry. The syn-cline of limestone seems to deepen towards line 6. As thedeepest part of this modelled syncline is at 1.7 skin depththe measurement of this depth may not be accurate. TheNE-SW striking axes of these folds (see map view, Fig-ure 9) correspond to the typical fold axis direction statedby Forián-Szabó and Csontos [4] and their possible con-tinuation is mapped by them NE from the Köpüs springarea.The mudstone in the core of the anticline is covered bylimestone detritus except on the topographically lower

side at the spring. The spring flows out from the detri-tus but comes from the limestone of the syncline dammedby the watertight layers in the anticline. The water isswallowed back where the runoff reaches the uncoveredlimestone of the relatively steep hillslope on the SE flankof the anticline.7. Conclusions

Mushrooms have never been used as geobotanical indica-tors before in the Bükk Mountains, but the observationsshowed a clear correlation between the spatial distribu-tion of the habitats of certain fungi and the soil-formingrocks. In this case, Leccinum aurantiacum and Boletusedulis were correlated with sandstone- and mudstone-dominated formations. The connection was valid in allobserved cases in the eastern part of the Bükk Moun-tains. Moreover, it provided the opportunity of correcting

92

Norbert Németh, Gábor Petho

Figure 9. Interpreted geological map of the survey area (area insideof the ellipse) and its 100 m surroundings.

the geological map. The exploration of the exposure bothdemonstrates and gives reasons for using the mushrooms– in particular the boletes – as indicators for geologicalmapping. The information gained in this way was utilizedin the interpretation of VLF measurements.Besides proving the existence of the indicated structure,some knowledge about the extent of it was gained by ap-plying the relatively simple and inexpensive VLF method.Both the measurements and the interpretation were madewhile taking into consideration the geological data knownso far. Before measurements are taken, an estimation ofthe strike of the structures being explored is required forchoosing the transmitter bearing which is perpendicularto the strike. Before interpretation, the resistivities of thepredicted formations are needed for the start model. Thisstart model can be constructed on the basis of the mea-sured apparent resistivities and their phases. The steps ofthe trial-and-error method should not be chosen at ran-dom, but in accordance with the possible geological modelgeometries in order to converge to a realistic solution. The2D assumption has to be verified in the interpretation. Inthis case the combination of these methods proved to bean efficient tool for mapping shallow elongated structureswith covered contacts, but with sufficient contrast in con-ductivity in both the lateral and vertical sense. The map-ping results were utilized in compiling the new geologicalmap of the Bükk Mts. [18].AcknowledgementsThe survey was supported by the projects T 042686 andT 37619 of the OTKA Hungarian Research Fund. FerencMádai is acknowledged for his help at the collection andinvestigation of the soil samples.

References

[1] Brooks R. R., Geobotany and biochemistry in min-eral exploration. Harper’s Geoscience Series, Harper& Row, New York, 1972[2] Takács E., Tapasztalatok a radiokip módszeralkalmazásában (Observations in application of ra-diofrequency method) Magyar Geofizika, 1971, 13,148-160 (in Hungarian)[3] Balás L, Balogh I, Petho G., Szeizmikus refrakciósmérések a Bükk-fennsíkon (Seismic refraction mea-surements in Bükk Plateau), Magyar Geofizika, 1975,16, 161-166 (in Hungarian)[4] Forián Szabó M, Csontos L., Tectonic structure ofthe Kis-fennsík area (Bükk Mountains, NE Hungary),Geol. Carpath., 2002, 53, 223-234[5] Fülöp J., Magyarország geológiája (Geology of Hun-gary). Paleozoikum II, Akadémiai kiadó, Budapest,1994 (in Hungarian)[6] Haas J. (ed.), Magyarország litosztratigráfiai alape-gységei, Triász (Lithostratigraphical units of Hungary,Triassic), MÁFI, Budapest, 1993 (in Hungarian)[7] Zólyomi B, Jakucs P, Baráth Z, Horánszky A.,Forstwirtschaftliche Ergebnisse der geobotanischenKartierung im Bükkgebirge (Forestrial results of thegeobotanical mapping in the Bükk Mts.), Acta Bot.Hung., 1955, 1, 361-395 (in German)[8] Less N., A Délkeleti-Bükk vegetációja és xerothermerdotársulásainak fitocönológiája (Vegetation andphytocoenology of xerotherm forest associations of theSoutheastern Bükk), PhD dissertation, Department ofOecology, University of Debrecen, Hungary, 1991 (inHungarian)[9] Less N. 1:10000 vegetation map of the Bükk Mts., De-partment of Oecology, University of Debrecen, Hun-gary, 1991[10] Jámbor Á., A Bükk-fennsík pleisztocén „vályog”-képzodményei (Pleistocene “cob”-formations of theBükk Plateau). Földtani Közlöny, 1959, 89, 181-184(in Hungarian)[11] Kalmár Z, Makara Gy., Eheto és mérges gombák (Edi-ble and toxic mushrooms), Natura, Budapest, 1978 (inHungarian)[12] Paterson N. R, Ronka V., Five years of surveying withthe very low frequency EM method, Geoexploration,1971, 9, 7-26[13] McNeill J. D, Labson V., Geological mapping usingVLF radio waves. In: Nabighian M.N. (Ed.), Electro-magnetic methods in Applied Geophysics, Society ofExploration Geophysicists, Tulsa, 1993[14] Cagniard L., Basic theory of the magnetotelluric93

Geological mapping by geobotanical and geophysical means: a case study from the Bükk Mountains (NE Hungary)

method of geophysical prospecting, Geophysics, 1953,18, 605-635[15] Kaikkonen P, Sharma S. P., A comparison of perfor-mances of linearized and global nonlinear 2-D inver-sions of VLF and VLF-R electromagnetic data, Geo-physics, 2001, 66, 462-475[16] Takács E, Tevan Gy., Numerical method for the com-putation of magnetotelluric fields in inhomogeneousmedia, Acta Geodaetica, Geophysica et MontanisticaAcademiae Scientiarum Hungaricae, 1973, 8, 55-69[17] Parasnis D. S., Principles of Applied Geophysics.Chapman & Hall: London – New York, 1986[18] Less Gy, Gulácsi Z, Kovács S, Pelikán P, PentelényiL, Rezessy A, Sásdi L., 1:50000 geological map of theBükk Mountains, MÁFI, Budapest, 2002

94