tectonic inception in caledonide marbles

8/13/2019 Tectonic Inception in Caledonide Marbles

1/16

ECONIC INCEPION IN CALEDONIDE MARBLES

EKONSKA INCEPCIJA V KALEDONSKIH MARMORJIH

revor FAULKNER1

Izvleek: UDC: 551.24:551.44(48)Trevor Faulkner: Tektonska incepcija v kaledonskih marmor-jih

V centralni Skandinaviji je ve kot tiso izdankov marmorjav katerih je znanih preko tiso jam. Razlika med temi jamamiin tistimi v apnencih, je pogojena s procesi metamorfoze inzanemarljivo primarno poroznostjo prvih. S tem je povezanodrobno plastenje (foliacija) in posledina odsotnost lezik.Najgloblja jama je globoka 180 m, kljub temu, da vertikalni

razpon izdankov znaa do 900 m. Jame najvekrat najdemov skupinah in so v vertikalnem merilu precej nakljunoporazdeljene, pri emer jih redko najdemo ve kot 50 m podpovrjem. Kljub temu, da so nekateri izdanki pasastega kra-sa dolgi ve deset kilometrov, zelo dolgih jam ne poznamo,kraki hidroloki sistemi pa ne presegajo doline 3.5 km. Kerso jame kratke in blizu povrja, je kemijska incepcija manj ver-

jetna. Zato predlagam model tektonske incepcije, ki predvide-va, da so jame v kaledonskih marmorjih nastale zgolj vzdolsistemov odprtih razpok, pri emer je dimenzija in frekvencateh razpok povezana z magnitude tektonskih premikov, ki sonastali kot posledica izostatinega uravnoteenja ob umikihledenikov po ledenih dobah. Sistemi takih razpok so nasta-

jali vzdol incepcijskih povrin med marmorji in neprepust-nimi plastmi in vzdol incepcijskih razpok v marmorjih,ki so vzporedne ali pravokotne s plastenjem. Model gradimtudi na poroilih o delno odcepljenem tankem vrhnjem deluskorje v podobnih okoljih na kotskem. Model podpirajo tudiopaanja kasnejih neotektonskih premikov, na katere kaejoostri robovi in tektonska zrcala v jamah in na povrju.Kljune besede:Kaledonidi, plastenje, rob ledenikov, incepci-

jska razpoka, incepcijska povrina, marmor, plitvi vodonosnik,neotektonika, seizminost, tektonska incepcija, pasasti kras,Weichelij.

1Limestone Research Group, University of Huddersfield, Queensgate, Huddersfield, HD1 3DH, UKe-mail: [email protected], telephone : +44 (0)1625 531558post: Four Oaks, Wilmslow Park North, Wilmslow, Cheshire, SK9 2BD, UK

Received / Prejeto: 03.03.2006

COBISS: 1.01

ACA CARSOLOGICA 35/1, 721, LJUBLJANA 2006

Abstract: UDC: 551.24:551.44(48)Trevor Faulkner: Tectonic inception in Caledonide marblesA fundamental difference between caves in sedimentary lime-stones and those formed in a repeatedly-glaciated 40000 km2region in central Scandinavia that contains over 1000 individu-al marble outcrops and has nearly 1000 recorded karst caves isthe metamorphic grade of the karst bedrock and its negligibleprimary porosity. Allied to this is the fine-scale foliation andconsequent lack of bedding-plane partings. Indeed, the folia-

tion is commonly vertical in the western part of the study area,where sub-horizontal openings must be along joints or otherfractures. Te deepest cave is only 180 m deep, despite outcrop

vertical ranges reaching over 900 m. Caves tend to cluster to-gether and are positioned randomly in a vertical dimension,whilst commonly remaining within 50 m of the overlying sur-face. Additionally, despite some stripe karst outcrops beingseveral tens of kilometres in length, there are no regional scalecaves, and karst hydrological system distances are invariablyshorter than 3.5 km. Because the caves are relatively short andepigean and there is a complete absence of long, hypogean, cavesystems, speleogenesis by the (chemical) inception horizon hy-pothesis is unlikely.A tectonic inception model is derived that proposes that it isonly open fracture routes that could provide the opportunityfor dissolution and enlargement into cave passages in the Cale-donide marbles. It is hypothesised that the dimensions of thesefractures are related to the magnitude, and perhaps to the fre-quency, of local earthquakes and commonly-small tectonicmovements that arose mainly from the isostatic rebound thataccompanied deglaciation at the end of each major Pleistoceneglacial. Te openings formed along inception surfacesbetweenthe limestone and adjacent aquicludes and at inception frac-turesthat are entirely within the limestone and are commonly(though not universally) parallel to, or orthogonal to, the fo-liation. Te model builds on reports of a partially detachedthin upper crustal layer in similar settings in Scotland and is

supported by observations of later neotectonic movements, asindicated by sharp edges and slickensides in most present relictcave passages and sporadically on the surface.Keywords:Caledonide, epigean, foliation, ice margin, incep-tion fracture, inception surface, marble, near surface aquifer,neotectonics, seismicity, tectonic inception, stripe karst, Weich-selian.


2/16

ACA CARSOLOGICA 35/1 20068

INRODUCION

Central Scandinavia is a repeatedly-glaciated 40000km2region that contains over 1000 individual Caledo-nide marble outcrops and over 1000 karst caves with a

total passage length >72 km, within an area about thesize o Switzerland (Fig. 1). A actual review o dataassembled into karst and cave databases (Faulkner,2001 and 2005a) revealed that cave development hasbeen predominantly phreatic, so that, commonly, justa single vadose streamway underlies upper-level rel-ict phreatic passages with ew vadose elements, cre-

ating an upside-down morphology. Recharge to thekarst is primarily allogenic and discharge commonlyremains unsaturated with calcite; autogenic recharge

is relatively insigniicant, mainly occurring during thespring snowmelt. hese caves have their own morpho-logical style, recognisable right across the area, whichdierentiates them rom caves ormed in classicalkarsts in sedimentary limestones. A key question toaddress is Why do these caves exist at all?

HE INCEPION PROBLEM

Te Inception Horizon Hypothesis (IHH; Lowe, 1992;Lowe and Gunn, 1997) proposed that the initiation oproto-conduits occurs as a syngenetic cave ormationalprocess during diagenesis. Te long, slow, non-karstic,inception phase is driven by capillarity, earth tides orionic diffusion at great depth and over great distanceswithin stratigraphical partingsor adjacent porous or frac-tured rocks. Eventually, chemical dissolution increasesconduit sizes to explorable dimensions. How does thishypothesis stand in relation to the karsts and caves othe study area?

LACK OF PRIMARY POROSIYMost o the high (up to amphibolite) -grade metalime-stones o the study area exhibit little memory o theiroriginal diagenesis, afer their subduction and meta-morphism to marble at elevated temperatures and pres-sures: any proto-conduits ormed syngenetically duringdiagenesis were closed as the rock experienced chemi-cal and physical changes in lithology. Te recrystallisa-tion to metacalcite produced a rock with a fine-scaleoliation and a primary porosity that can be regardedas negligible, even over the long timescales available orconventional inception. Te same applies to any mica

schist, amphibolite, granite or gneiss lying adjacent tothe marble: these rocks could not have sufficient pri-mary porosity to act as aquiers carrying water to thelimestone surace.

LACK OF SRAIGRAPHICAL HORIZONSTe oliation is commonly vertical in the western Helge-land Nappe Complex (HNC; Fig. 1), but caves in such

vertical stripe karst (VSK; oliation dip 8190o) com-monly display morphologies similar to those in horizon-tally-bedded limestones, with many horizontal passages

orthogonal to the oliation (Faulkner, 2005a), despite thelack o inception horizons to guide their ormation alongparticular bedding plane partings. Tere are also no con-sistent systems o sills or other intrusions to act as incep-tion horizons, so that the horizontal openings must bealong joints or other ractures.

LACK OF REGIONALSCALE SYSEMSTe IHH suggests that inception takes place over ex-tremely long timescales, at great depths and over greatdistances. Tere is no evidence that such a mechanismhas taken place in the study area, despite some o the

stripe karstoutcrops exceeding 50 km in length. Tereare no regional-scale caves; there are no known allogenicor autogenic sink-to-rising hydrogeological drainagesystems longer than the 3.5 km that occurs at Vallerdalon the border between Norway and Sweden; and thereis no evidence o very deep cavities or wells in the meta-carbonates. Te steep oliation and metamorphic historyhave lef many completely separate stripe karsts. Teircontained caves are, by necessity, commonly short (meanlength = 85 m) and completely unrelated to each otherinternally, even i proximate in the field, so that regional-scale inception is not possible.

EXISENCE OF SHALLOW SYSEMSDespite the large vertical range (VR) o someo the me-talimestone outcrops (up to 956 m), the deepest cave isonly 180 m deep, and only our others are more than 100 mdeep. Te mean cave VR is only 8.8 m and it rarely ex-ceeds 15% o the local outcrop VR. Tus, the caves arecommonly extremely epigean and there is a total absenceo long, hypogean, cave systems. It is sel-evident when

visiting such systems (e.g. a short shallow through cavethat carries a stream along a vadose passage rom one

REVOR FAULKNER


3/16

ACTA CARSOLOGICA 35/1 2006 9

TECTONIC INCEPTION IN CALEDONIDE MARBLES

entrance to another) that such passages have no relation-ship to any deeper, regional-scale, hydrogeology, even iit existed. Whereas it could perhaps be considered as apossibility that all such short and shallow caves are thelowest remnants o much longer systems ormed deepbelow landscapes that have since been eroded away, thisseems most unlikely as the carbonate outcrops would nothave been consistently longer in the past than at present.

THE IMPLAUSIBILITY OF THE IHH TO EXPLAININCEPTION IN SOME METALIMESTONES

From the our arguments presented above, the IHH can-not explain the inception o the overwhelming major-ity o caves in the study area. However, elements o the

Hypothesis may explain parts o the inception process insome caves, or groups o caves. For example, inceptionthat isguidedalong sub-horizontal aquicludes within the

oliation o marbles in low angle karst (LAK; oliation dip030o) seems likely, as at Ytterlihulletin Bryggfelldal.Similarly, inception along-strike at lithological bounda-ries within lower grade metacarbonates in angled stripekarst (ASK; oliation dip 3180o), as at KorallgrottaninSweden, is also easible. However, even in these exam-ples, another mechanism is needed to explain an initialporosity.

Fig. 1: ectono-stratigraphic map o central Scandinavia, rom Gee and Sturt (1985). Te numbers indicate the areas various allochthonsand nappes. Most nappes contain metacarbonate outcrops that are commonly aligned NS and decrease in size (along with a commonreduction in metamorphic grade) in an easterly direction. Caves are only recorded in metalimestone outcrops in the UppermostAllochthon (7; i.e. in the Helgeland Nappe Complex, HNC and the Rdingsfell Nappe Complex, RNC) and in the Kli Nappes o theUpper Allochthon (6). Small marble outcrops occur in the Seve Nappes o the Upper Allochthon (5) and in the Lower Allochthon (3),without recorded caves. Te Middle Allochthon (4) does not contain metacarbonates.


4/16


Despite the difficulty in utilising the IHH to explain theinception of the studied caves, these caves exist and theirorigins must post-date the last phase of metamorphic

activity. Te consistent style of the caves suggests thata consistent set of processes guides their inception, de-velopment and eventual destruction. wo major clues tothe inception process were noted in analysing the cavemorphologies: externally, their epigean association withthe landscape, and internally, the dominance of relictphreatic passages.

ASSOCIAION OF CAVES WIH LANDSCAPE

All cave passages in both VSK and ASK lie within 50 mof the overlying surface. Even in Ytterlihullet(LAK), allparts of its streamway are 93 m below the surface. Its

stream resurges, flows along a short surface valley, andthen sinks again in the same limestone outcrop beforefinally resurging some 200 m above the base of the lime-stone, and some 300 m above the valley floor. Tus, thiscave and most other active caves act in harmony withlocal hydrology and have an intimate, epigean, associa-tion with their local landscape. It seems safe to assumethat these caves evolved in association with the shapingof their local topography, whose dominating process isthe cycle of glaciation and deglaciation that has beenrepeated many times since the late ertiary.

RELIC PHREAIC PASSAGESTe absence of relict vadose caves shows that all relictcaves in the area developed phreatically, as did nearly allthe higher-level abandoned passages in the active caves.However, it is not possible to imagine present circum-stances, even during spring melt, when most of these re-lict caves could be flooded to create phreatic conditionsfor their enlargement. It may be possible to envisageearlier landscapes where these passages were submergedunder meteoric conditions, but a much simpler explana-tion is that these passages enlarged subglacially or duringdeglaciation phases, when whole valleys could be inun-dated by glacial meltwater.

HE ECONIC INCEPION MODELTe development (and destruction) of the present suiteof karst caves can therefore be addressed by consider-ing the way that glaciation has eroded the land surface,and perhaps provided sufficiently aggressive meltwatersto enlarge passages by dissolution. But these processescannot explain the actual inception along proto-con-duits. Without such openings, glacial meltwaters wouldnot penetrate into high grade metalimestone, even underpressure.

Te Tectonic Inception Model hypothesises that,through several separate, but commonly related, mecha-nisms, the stress release arising from the isostatic re-

bound and surface erosion that accompanied deglacia-tion at the end of each glacial cycle, plus longer-timescaleplate tectonics, caused the formation of tectonic fracturesin the upper (epikarstic) part of the limestone (Fig. 2).Tus, openings are created along inception suraces be-tween the limestone and adjacent aquicludes (which mayinclude dolostones), and by inception ractures that areentirely within the limestone, but are commonly (thoughnot universally) parallel to, or orthogonal to, the folia-tion. Tis model builds on the observations that the con-tinuing seismic and tectonic activity (in similar settings)in Scotland may be best understood in terms o a partially

detached thin upper crustal layer (Davenport et al., 1989,p 191) and that near-surface limestones are not very duc-tile and produce brittle fractures during folding, faultingand removal of overburden stress by erosion (e.g. Dorand Jensen, 1996, pp 426427). It is also assumed thatthe maximum thickness of permafrost during glaciationis c. 100 m. Rock above this level is subjected to moresevere temperature cycling and freeze-thaw processesthan rock below it, and is therefore more likely to forminception fractures when triggered by seismicity. Tepractical expression of these processes was provided byBoulton et al. (1996, p 403), who noted from pumpingtests that the crystalline basement rocks of the Scandi-navian shield (primarily non-carbonates) havea suracehorizon o ractured bedrock about 100 m thick which hasa hydraulic conductivity o 10-6ms-1. Tis provides a nearsurace aquier that is commonly found in crystallinerocks worldwide (Gustafson and Krasny, 1994).

Te idea of tectonic speleogenesis in karst rocks hasa precedent, because Riggs et al.(1994) proposed this at

HE ECONIC SOLUION

Fig. 2: Marble at Indrsen quarry, Velford: Shattered nature othe epikarst in high-quality metalimestone altered by contactmetamorphism.

REVOR FAULKNER


5/16

ACA CARSOLOGICA 35/1 2006 11

ectonic inception (and indeed any inception hypothesis)is not easy toprove. ectonic ractures may be too narrowto observe visually and may no longer be recognisableaer karstic dissolution and enlargements to explorablepassages. Tus, the ectonic Inception Model is sup-ported by several lines o evidence or Caledonide tec-tonism and racture ormation in the ollowing sections.Te hydrogeologyo ractured rock, including racturedmetalimestone, was considered separately by Faulkner(2003 and 2005a).

CALEDONIDE EVIDENCE FOR ECONICACIVIY

Faulkner (1998) reviewed recent ideas on the importanceo tectonic activity to cave development in sedimentarylimestones. Te idea that tectonism sensu latohas influ-enced karst cave development in Norway has been sug-gested, or hinted at, by several authors. Tus, Hoel (1906,p 8) raised the possibility that Aunhattenhullet 1, 2 and3and Langskjellighattengrottain Velford in the studyarea were ormed by dislocations. Horn (1947: McGradytranslation, 1978, p 135) noted that the Norwegian coast-al area at the Arctic Circle is still unstable tectonically,which should avour joint ormation, or the widening oold joints. Kirkland (1958) thought that collapsed blockson the floors o chambers in the Svartisen area could haveresulted rom movements along aults and rom seismicdisturbances. Lauritzen (1989a, 1989b and 1991b, p 122)suggested that cave passages in Norway are almost alwaysguided by the line o intersection between two planes(but see section 4). His statistical analysis revealed thatcommonly shear ractures (aults and shear joints) andless commonly tension ractures are utilised as primaryguiding voids or speleogenesis. Onac (1991) noted cavesormed by gravitational mass movement near Narvik,

and the influence o tectonic aults in guiding subterra-nean streams.

Randall et al. (1988) reported on the hydrogeologi-cal ramework o the NE Appalachians (USA), a regionwith a comparable metamorphic Caledonide geology.Tey noted high hydraulic yields rom ractured non-porous bedrock, especially rom wells that intersectcontacts between different lithologies. Earlier work wasquoted that showed that ractures decrease in size andrequency some 5075 m below the surace. Te water-table configuration in uplands nearly replicates the to-pography throughout the region, so that inter-basin flowsystems involving significant flux have not been shownto exist, as in central Scandinavia.Carlsten and Strhle(2001) reported that open, and partly-open, fissures wereound in a borehole at Bodagrottorna in non-carbonaterock on the Swedish Baltic coast at depths at least downto 150 m, in an area that was very active seismically in theearly Holocene.

ECONIC MECHANISMSSeismic and aseismic tectonic processes that create rac-tures can arise rom several separate mechanisms. Teevidence or considerable isostatic upli during themelting o the 23 km-thick Weichselian icesheet is welldocumented. Tat part o the evidence or upli that isassociated with caves includes Sjberg (1981a and b),who discussed 50 elevated caves in east Sweden ormedby cobble abrasion at the coast o the Baltic, and Sjberg(1988) who discussed elevated coastal caves in centralNorway. Tat seismic tectonic activity accompanied theupli was documented by: Husebye et al. (1978); Mrner(1980); Stephansson and Carlsson (1980), who discusseda Caledonian Zone o seismicity; Anderson (1980), whosuggested that the maximum number o earthquakes a-

FORMAION OF ECONIC FRACURES

Devils Hole, Nevada, although without subsequent dis-solution. Te only known paper to discuss the impor-tance o racturing by stress release in the developmento cave passages in sedimentary limestones was by Sa-sowsky and White (1994), who anticipated some o theprocesses described herein, but or a non-glacial setting

in ennessee.

HE GLACIAL / ECONIC CYCLEBecause the tectonically-induced inception ractures arecommonly produced at the endo each glaciation, theremay not always be sufficient time or phreatic passages toenlarge to explorable dimensions during the remaining

time o that particular deglaciation. Hence, the cyclicprocesses o glaciation, deglaciation and tectonic open-ing combine together to develop cave passages: tectonicinception provides ractures that permit the circulationo meltwaters that can be chemically aggressive evenwithout dissolved CO

2(Faulkner, 2004a and b; 2005a),

both during that deglaciation and during the next glacialand deglacial phases. As the cycle repeats itsel, passagesnear the surace enlarge and become removed by gla-cial and fluvial erosion (as noted by Isacsson, 1994, atKorallgrottan), and new passages orm at geologicallylower levels.



6/16


7/16


the onset o glaciation, this style is overprinted by theglacial stress, and new horizontal crustal motions in-crease outwards rom the icesheet centre. Tey showedthat subglacial water penetrates into the crust below en-hanced icemelt in topographic hollows, increasing thepore-water pressure, and that large icesheets stabilise

underlying crustal aults, whereas deglaciation desta-bilises the aults. Periods o cover by maximal Scandi-navian icesheets represent times o seismic quiescence,due to the muffling effect o the weight o ice, as the landis gradually compressed and isostatically depressed(Johnston, 1987). In Fennoscandia, aulting is linked tozones with very steep ice gradients, or to the final stageso recession, when the bulk o seismic activity probablyoccurs within a ew hundred years. During the similardeglaciation o Scotland, local movements were causedby differential glacial load flexure stresses (Davenportet al., 1989; Ringrose et al., 1991), at places with the

steepest ice gradients (Stewart et al., 2000). Johnston(1987) also noted that artificial reservoirs can triggerearthquakes by increasing hydrostatic pressure. It oc-curs to this author that local deglacial earthquakes maysimilarly be triggered by the ormation o ice-dammedlakes. Fjeldskaar et al. (2000) suggested that stress-gen-erating mechanisms can be grouped into three classes:first-order stresses across Fennoscandiathat arise romthe longer-term plate tectonic NWSE compressionridge-push orces caused by oceanic spreading rom theAtlantic Ridge; second-order stresses that are limited toScandinavia; and third-order stresses that relate to lo-

caleatures (e.g. topography) and rarely extend beyond~100 km.

Any o the above mechanisms may result in racturesopen to the surace. Tey may fill with water in summer,so that any winter reezing would subject the rock toincreased stress. Te magnitude o any widening is pro-portional to the sub-zero (oC) temperature at the surace(Matsuoka, 2001). Although most widening is reversedon thawing, there is a tendency or the racture to be per-manently enlarged, and then to admit a higher volumeo water during the next reezing cycle. Te temperaturecycling o rocks o differing lithologies that have unequal

coefficients o thermal expansion would also promoteracture enlargement along contact zones. Indeed, Gud-mundsson et al. (2002, p 64) stated that stresses tend toconcentrate at the contact between dyke rock and the hostrock and generate fractures that may conduct groundwa-ter. Tus, tectonism commonly leads to a growth in thesize o the near-surace racture network, even withoutinvoking karstic processes. I ice-dammed lakes com-pletely roze in winter or during a period o local perma-rost, then submerged ractures would also be subjectedto urther stress and widening.

Another mechanism to increase racturisation ishydrofracturing (e.g. Gudmundsson et al., 2002). Tisprocess orces groundwater upwards through bedrock atgaps in permarost, which may apply to metacarbonatesduring parts o the glacial cycle. At the base o a 500 m-deep ice-dammed lake, the excess pressure would be 50

atm. Tus, water can be injected into ractures that mayoccur within any underlying metalimestones, and, ac-cording to Banks et al. (1996, p 230), such pressures in aborehole may be sufficient to stimulate already racturedbedrock and to create new ractures. Lubrication by wa-ter would also ampliy the effects o local seismicity.

Tere is no reason to suppose that the concentratedseismic creation o ractures during the Weichseliandeglaciation was unique: similar processes must haveoccurred during the demise o all previous Cenozoicglacials (and perhaps stadials). However, rom the spe-leothem chronozones proposed or Norway (Lauritzen,

1991a), there are long intervals o several 10 ka whenspeleothems did not grow, and ull glacial coverage canbe inerred. It thereore seems likely that the largest mag-nitude earthquakes only occurred once per 100 ka glacialcycle.

NEOECONICSIn addition to the postglacial uplif, there are two mainsources o evidence o neotectonics in Scandinavia: therecent earthquake record, and the observation o move-ment along aults (e.g. Husebye et al., 1978; Olesen,1988; Bungum, 1989; Olesen et al., 1992 and 1995). Lo-

cal instrumentation can now record small earthquakeso magnitude 2, as summarised on a neotectonics mapby Dehls et al.(2000a). Te seismic events tend to ol-low NS alignments at depths commonly ocused above15 km at the Atlantic Ridge, along the Continental Sheledge, along the Norwegian coast, rather randomly alongthe border and onto the Swedish shield, and along theSwedish Baltic coast.

Many earthquakes have occurred in northern Nor-way and along the coast o southern Norway since 1750AD, but lower requencies and magnitudes coincide withthe study area, which occupies a saddle position between

higher mountain ranges. Central Scandinavia probablyacted as a ocus or ice flow during late Cenozoic glacia-tions. With thinner icesheets, there was less stress relieand lower seismicity at each deglaciation. Addition-ally, increased ice flow increases glacial erosion, leadingto less surace relie and less differential stress, and theincreased sedimentation on the Vring Plateau, off thecoast o the study area, may have a dampening effect. Tehistorical record o significant, but comparatively smallerand less requent neotectonic earthquakes in the studyarea (Fig. 4) may be representative o relative seismic



8/16


activity during the whole Holocene, although, ollowingthe pulse o deglaciation seismotectonics, the style oseismicity does change, as noted by Stewart et al. (2000,p 1381): Whereas present-day seismicity is concentratedaround the margins o the ormer icesheet, on deglaciation,earthquakes predominated at the centre o the rebound

dome. However, neotectonic earthquakes do ollow theRana Fault Complexsouth along the coast o the studyarea, and the largest recorded Northern European near-shore earthquake, o magnitude 5.8, occurred on 31 Au-gust 1819 AD in Rana, just north o the study area. Some10000 micro earthquake shocks were recorded instru-mentally at Mely, 70 km north o the study area, during10 weeks in 1978 (Bungum et al., 1979). Tese were upto magnitude 3.2, were heard and elt locally, and causedcracks in walls and chimneys.

Te documented active postglacial aults are com-monly NESW-trending reverse aults that lie within a

400 km x 400 km area in northern Fennoscandia (e.g.Arvidsson, 1996). Teir lengths and maximum scarpheights vary rom 3150 km and rom 130 m. Fault off-sets range up to 13 m (Dehls et al., 2000b). A magnitude4 earthquake occurred near one o these aults in 1996,when large amounts o groundwater poured out o theescarpment. Te ault length to offset ratio indicates thatthe structure itsel resulted rom an earthquake with amagnitude above 7. Te work o Olesen et al. (2004, p 17)supports previous conclusions regarding a major seismicpulse (with several magnitude 78 earthquakes) which

ollowed immediately afer the deglaciation o northern

Fennoscandia.Te earthquakes may not just be caused by isostatic

rebound aer the removal o ice. Tey may also indicatethe opportunity or adjustment to glacial erosion aer themuffling effect o the ice cover has gone. Te 2 km-wideWE Bsmoen Fault zone is just north o the junctionbetween the HNC and the Rdingsfell Nappe Complex(RNC) and can be traced or 50 km along Ranaford (Fig.4). It has a maximum displacement o 10 m, escarpmentsup to 80 m, provides evidence o recent movements (3040 cm between 8780 and 3880a BP: Hicks et al., 2000),and was associated with the 1819 earthquake. Te Rana

area was the subject o an in-depth seismic study, NEO-NOR, rom 19971999, when some 267 local earthquakeswere recorded with magnitudes up to 2.8 by Hicks et al.,(2000), who stated (p 1431):Te Rana area has a signifi-cant amount o the total seismic activity in onshore north-ern Norwayand concluded that postglacial upli is themost likely cause or this continuing high level o seismicactivity.

Muir-Wood (2000) discussed postglacial very shal-low stress-relie phenomena, known as pop-ups, whichare prevalent along the margins o the Laurentian ice-

sheet, but relatively unknown in Scandinavia. However,Roberts (2000) reported offset structures in boreholes atroad-cuts that are regarded as stress-relie eatures initi-ated by blasting. Within road tunnels there is anecdotalevidence that civil engineers report the sounds o rockmoving, and rock bursts occur when rocks all rom the

roo, aer blasting is complete. At the surace, crushedrocks and slipped blocks and notches on skylines mayindicate postglacial movements along aults and nappeboundaries.

Olesen et al.(2004, Appendix A) included 54 clas-sified claims o neotectonic movements rom onshoremainland Norway, prior to new evidence discussed here(section 3.4). Te earthquakes and ault movements arecommonly parallel maniestations o neotectonic activitythat arise rom both glacial isostatic upli and the lon-ger-term plate tectonics. Olesen et al. (1992) reportedthat the earliest detectable displacement in Finnmark

(the northernmost county in Norway), is o Proterozoicage, indicating an extremely long-lived ault zone. Suchault zones and their adjacent sub-parallel accommoda-tion aults lie parallel to the strike o the oliation, andgive low resistivity readings due to ingress o water into

ractures. Whereas the plate tectonic processes constitutethe most important ault-generating mechanism in Finn-mark, stress relie could still have been triggeredduringthe deglaciation period.

Tere are no known extensiveaults wholly withinthe study area, which, as noted above, is less seismicallyactive, although Olesen et al. (1995) showed an earth-

quake zone that extends NE across the north o the studyarea, passing through Mosjen and Korgen (Fig. 4). Be-cause the Weichselian icesheet had melted by 850014CaBP, the present pattern o neotectonic seismic activitycorresponds more to the horizontal stress field. As wellas being concentrated at the centre o the rebound dome,the earthquake pattern rom 10000850014Ca BP wasprobably aligned along the mountain ranges, and repre-sented the verticalisostatic rebound.

A conclusion rom this review o neotectonic acti-vity is that the seismic and aseismic creation and enlarge-ment o near-surace ractures continued throughout

each interglacial, to supplement the more intense rac-ture sets produced at each deglaciation. Tese processesprobably combine to create a spectrum o racture aper-tures, lengths, requencies and interconnectivities withinthe metalimestones. Such racture systems may includesubsystems that vary rom being too small to transmitwater, to those that are great enough to permit turbulentflow (without requiring karstic dissolution) over pathlengths that in the study area reach up to 3.5 km.

REVOR FAULKNER


9/16


EVIDENCE FOR ECONIC ACIVIY FROM HESUDY AREA

None o the 54 claimed examples o Norwegian neotec-tonic movement (section 3.3) lie within the study area.Te lithologies o affected or adjacent rocks are rarelygiven, but there is no indication that any are in carbon-ate rocks. Tus, a list o 56 possible examples o tectonicmovements in metalimestones presented by Faulkner(2005a, Appendix D1) may be the first recorded or thestudy area, and the first observed in both exokarstic andendokarstic situations. Altitudes range rom near sea lev-el to 770 m. Elgfell provides many good examples. Most

underground observations are intended to provide directevidence o movement, aer ormation o the observingpassage, rather than direct evidence o tectonic incep-tion.

Only one observation concerns allen, broken orcurved stalactites and stalagmites, which can be diag-nostic o earthquakes and relative roo movement. A ewmore unrecorded examples probably do exist, but spe-leothems are rare in the study area anyway, and most othose that do exist are small and probably grew in the Ho-locene, aferthe large earthquakes occurred. Speleothems

that grew in earlier interglacial periods have commonlybeen removed by subsequent deglacial outflows. Te ewchambers with roo spans greater than c. 6 m commonlycontain allen blocks, which almost universally compriselimestones with clean, sharp, angular suraces. Tis sug-gests that they ell aer any deglacial deposition, and

are situated high enough above streamways not to havebeen eroded by Holocene flood waters. Only two o thechambers are lit by daylight rom nearby entrances, sothat only these two may experience severe, seasonal, rostaction. Te others are not in entrance areas, and theirdisturbance by seismic shock seems the best explana-tion (e.g. Fig. 13). Human intervention is most unlikely,because o the common inaccessibility. However, all thelarge chambers are within 30 m o the overlying surace,and most within 15 m, so that a second possible processis downward flexing o the roo by the weight o an over-riding icesheet (providing the cave was not filled by ice

or water), as proposed by Warwick (1971), and upwardflexing when the ice melted. A third explanation basedon the reezing to a total ice fill during glacial conditionsalso cannot be ruled out.

It is the authors opinion, made aer field trips tomarble caves in central Scandinavia, northern Americaand Scotland, that evidence o small Holocene tectonicmovements (e.g. bedrock movement that displays sharpedges or slickensides, without subsequent calcite dissolu-tion or deposition) can be ound in all relict passages inmetalimestones in the Caledonides. Movements in VSKseem to occur in either vertical or horizontal slabs that

are typically 13 m thick. Te movements, presumablycaused by WE compressive stress, are commonly hori-zontal, normal to the strike, and have typical moved dis-tances o only a ew centimetres (and, rarely, several tenso centimetres), as expressed at the surace and withincave passages. Te horizontal movement o vertical slabso limestone 13 m thick is compatible with the surveyleg length o many caves in the study area, suggesting that

joint systems (in, e.g., VSK) are produced by this process(Figs. 5 and 6). Longer straight passage elements, and

very wide, but low, passages, may arise rom the hori-zontal movement o horizontal slabs o limestone (Figs.

7, 8 and especially 9 and 10). Tese observations agreewith those o Olesen et al. (2004, p 13): the Norwegianbedrock consists o individual blocks that, to some degree,move independently o each other. According to Mrner(2003, p 72), a passing seismic wave can cause bedrockto li up and then sink back, whilst the ground is beingseverely shaken. Tis probably happened at Cliff CaveinJordbrudal (Figs. 9 and 10). Because most tectonic move-ments are o only a ew centimetres, explorable cave pas-sages are unlikely to be truncated along aults, and ewsuch blind passages are known in the area.

Fig. 4: Historical earthquakes in the study area (Various sources).



10/16


A possible alternative explanation is that there hasnot been any movement, but that differential erosionor corrosion has given the appearance o movement.

Tis could arise particularly i the apparent movementis aligned with the oliation. However, the photograph-ic evidence or tectonic movements discussed above iscompelling. Te evidence provided at Elgfellhola(Fig.5) is particularly convincing, because the movement isacrossthe oliation, is seen all the way around the passagewalls, has a 1 mm-thick ault gauge waer protruding upto 50 mm, and includes an 11cm step across a wall scal-lop. Te evidence o protruding ault gauge waers at sur-ace sites (e.g. Fig. 11) that appear to cross-cut karren andstream channels suggests that these movements occurredin the Holocene, afer the transport o ice across the area.Te waers could have been extruded beyond the aces othe limestone blocks by the seismic movements, or elseHolocene chemical dissolution o the surace has lef themore resistant waers exposed to a height that indicates

Fig. 6: Slickensides in Paradox Cave, Elgell: Tectonic movemento ~20 cm afer enlargement o passage to its present size.

Fig. 5: Scallop in Elgellhola: 11cm tectonic movement at scallop(highlighted), which occurred afer ormation o the passage,

probably synchronously with movement in the nearby ParadoxCave (Fig. 6).

Fig. 7: Kidney Lake Cave, Jordhuleell: Relict phreatic passage,with ~2 m diameter. A prominent horizontal tectonic movementbisects the passage, probably resulting rom seismic amplification,because the cave lies in a ridge. Although this movement occurredafer the passage enlarged to it present size, cave inceptionprobably took advantage o a similar movement at the end o the

Saalian deglaciation. Photo by P. Hann.

Fig. 8: Horizontal tectonic movement on Elgell: Massmovement outward (afer glacial smoothing) o 2 m-thick slabo metalimestone, with proto-conduits at upper racture.

REVOR FAULKNER


11/16


the extent o local surace lowering, or wall retreat in acave. Te waers are calcitic, with polished suraces andunknown dissolutional characteristics.

Te observed tectonic movements in karst cavescommonly ollow the plane o the supposed inceptionracture. Additionally, caves commonly display a highconcentration o joints and ractures (c.. the epikarst in

sedimentary limestones) that lie parallel to, or normal to,the plane o oliation, and in some cases at other angles.Tese openings may not show lateral movement, but the

variable degree o sharpness or smoothing by dissolu-tional water indicates that they probably represent a gen-eral settling upwards o large superficial carbonate blocksafer seismic shocks. Te sporadic lines o speleothemsbeneath roo joints indicate ailed vertical inceptionractures, which transmit water more readily in vadoserather than phreatic conditions.

Fig. 9: Cliff Cave entrances: Shattered cliffs and towers o lime-stone near the Rockbridge, Jordbrudal.

Fig. 10: Cliff Cave entrance rom inside: Horizontal opening oc. 1 m to both lef and right that split floor o phreatic passageto create a box-like profile. Tis is the largest known tectonicmovement in the study area.

Fig. 11: Diverging flow on Elgell: Diverging flow o coffee acrossvertical racture with ault gauge, suggesting that movementoccurred afer the surace flow was established.

Fig. 12: Fountain at Litl Hjortskar, Svenningdal: Spring at highstage rom metalimestone ractures 1 m above level o adjacentstream.



12/16


Fig. 13: Te Blockpile, Kvannlihola 2: Well-away rom reeze-thawinfluences, this collapse likely occurred during early Holoceneearthquakes.

Fig. 14: Secret Stream Cave, Elgfell: Primarily a tectonic cave,ormed at junction o mica schist and marble. Te mica schist hassplit and rotated upwards. Te pick-axe is a relic o small-scalemining activity.

It is sel-evident that i tectonic caves can orm in non-carbonate rocks, such as the entrancetoSecret StreamCavein mica schist (Fig. 14), then, despite metalimestoneperhaps being slightly more ductile than some other locallithologies, there almost certainly exist natural conditionsthat promote the creation o tectonic caves in marbles, aslisted by Faulkner (2005a). Such caves may be recognisedby their angular or triangular passage profiles, especiallyat roo level. (Sediments, clastic materials and allen rockmay provide a flatter, sub-horizontal, floor). Whereas themovements along ractures in caves primarily ormed bykarstic dissolution are commonly small (the c. 1 m move-ment in Cliff Cave, Figs. 10 and 11, is exceptional), themovements at purely tectoniccaves could be much greater.It is also sel-evident that i a limestone tectonic cave laterbecame part o a drainage route, under vadose or phreaticconditions, then normal karstic chemical and mechani-cal erosion processes would apply, and, over time, thepassage would enlarge. I the drainage was phreatic, theneventually the evidence o its tectonic inception could

dissolve away. Even in vadose conditions, the signs o anoriginal tectonic movement may be destroyed in all butthe highest, perhaps inaccessible, levels. Te only knownexamples in the study area o caves in metalimestone that

possibly enlarged tectonically to explorable dimensionsand later enlarged significantly by karstic processes arethe adjacent caves Nordlysgrotta and Marimyntgrottain Velford, which may also have passages truncated bytectonic movements (Faulkner, 2005b).

Whenever a Caledonide karst passage has beenstudied by the author, it has always been ound to olloweither the contact between metalimestone and another,non-carbonate, rock, or a narrow (commonly horizontalin VSK) racture plane in the limestone. Because there arelikely to be rheological differences between rocks o di-ering lithologies, tectonic ractures are particularly likelyto orm at lithological contacts, under all conditions oseismic and aseismic tectonic movement. It is not neces-sary to have intersecting ractures or tectonic inception:apertures are uneven, and channel flow ollows the wid-est part o the opening (Hanna and Rajaram, 1998). Noris movement along the plane o the racture necessary: aseparating aperture adjacent to, or within, the limestoneis sufficient. Such local rock splitting, especially vertical,

may arise near the surace rom deglacial and erosionalunloading, without necessarily being triggered by seis-mic or aseismic processes.

EVIDENCE FOR ECONIC INCEPION

REVOR FAULKNER


13/16


14/16


Dehls, J. F., O. Olesen, H. Bungum, E. C. Hicks, C. D.Lindholm & F. Riis, 2000a: 1:3000000 Neotectonicmap: Norway and adjacent areas.- Geological Sur-

vey o Norway.Dehls, J. F., O. Olesen, L. Olsen & L. H. Blikra, 2000b:

Neotectonic aulting in northern Norway; the Stu-

oragurra and Nordmannvikdalen postglacial aults.-Quaternary Science Reviews, 19, 14471460.

Dor, A. G. & L. N. Jensen, 1996: Te impact o late Ceno-zoic upli and erosion on hydrocarbon exploration:offshore Norway and some other uplied basins.-Global and Planetary Change, 12, 14, 415436.

Faulkner, ., 1998: Karst and ectonics Symposium Re-view.- Cave and Karst Science, 25, 3, 150152.

Faulkner, ., 2001: Cave development in central Scandi-navia.- Proceedings o the thirteenth InternationalSpeleological Congress, Paper 155, 4 pp. (Abstractp 106).

Faulkner, ., 2003: Te hydrogeology o crystalline rocks:pointers to tectonic inception mechanisms in karst.-Proceedings o the International Conerence onKarst Hydrogeology and Ecosystems, Bowling Green,Kentucky, USA, 36 June 2003. (Abstract p 20).

Faulkner, ., 2004a: Who needs carbonic acid?- Cave andKarst Science, 30, 3, p. 132.

Faulkner, ., 2004b: Te deglaciation o central Scandina-via and its implications or karst cave development.-Cave and Karst Science, 31, 2, p. 88.

Faulkner, ., 2005a: Cave inception and development inCaledonide metacarbonate rocks.- PhD thesis, Uni-

versity o Huddersfield.Faulkner, ., 2005b: Nordlysgrotta and Marimyntgrotta.-

Norsk Grotteblad, 45, 2328.Fjeldskaar, W., C. Lindholm, J. F. Dehls & I. Fjeldskaar,

2000: Postglacial upli, neotectonics and seismicityin Fennoscandia.- Quaternary Science Reviews, 19,14131422.

Ford, D. C., 1983: Effects o glaciations upon karst aqui-ers in Canada.- Journal o Hydrology, 61, 1/3, 149158.

Ford, D. C. & R. O. Ewers, 1978: Te development olimestone cave systems in the dimensions o length

and depth.- Canadian Journal o Earth Sciences, 15,17831798.

Gee, D. G. & B. A. Sturt (Eds.), 1985: Te CaledonideOrogen Scandinavia and Related Areas.- JohnWiley. 1250 pp.

Gudmundsson, A., I. Fjeldskaar & O. Gjesdal, 2002: Frac-ture-generated permeability and groundwater yieldin Norway.- Norges Geologiske Underskelse Bul-letin, 439, 6169.

Gustason, G. & J. Krasny, 1994: Crystalline rock aqui-ers: their occurrence, use and importance.- AppliedHydrogeology, 94, 2, 6475.

Hanna, R. B. & H. Rajaram, 1998: Influence o aperturevariability on dissolutional growth o fissures inkarst ormations.- Water Resources Research, 34,

11, 28432853.Hicks, E. C., H. Bungum & C. D. Lindholm, 2000: Seis-

mic activity, inerred crustal stresses and seismotec-tonics in the Rana region, northern Norway.- Qua-ternary Science Reviews, 19, 14231436.

Hoel, A., 1906: Den marine graense ved Velforden.-Christiana Videnskaps-Selskabs Forhandlinger, 4,115.

Horn, G., 1947: Karsthuler in Nordland.- Norges Ge-ologiske Underskelse, 165, 77pp. [Partial Englishtranslation by A. D. McCrady, 1978: LimestoneCaves in Nordland.- Cave Geology, 1, 5, 123138].

Husebye, E. S., H. Bungum, J. Fyen & H. Gjystdal, 1978:Earthquake activity in Fennoscandia between 1497and 1975 and intraplate tectonics.- Norsk Geologiskidsskri, 58, 5168.

Isacsson, G., 1994: Vad kan man se i Korallgrottan?-Grottan, 29, 2, 2123.

Johnston, A. C., 1987: Suppression o earthquakes bylarge continental icesheets.- Nature, 330, 467469.

Kejonen, A., 1997: On Finnish Caves.- Proceedings othe twelh International Speleological Congress, 4,9398.

Kirkland, R. J., 1958: Te Karst o south Svartisen.- Un-

dergraduate dissertation, 87 pp, rinity College,Cambridge.

Lauritzen, S-E., 1986: Kvithola at Fauske, northern Nor-way: an example o ice-contact speleogenesis.-NorskGeologisk idsskri, 66, 2, 153161.

Lauritzen, S-E., 1989a: Fracture control o caves in mar-ble in Norway.- Cave Science, 16, 3, p. 114.

Lauritzen, S-E., 1989b: Shear, tension or both a criti-cal view on the prediction potential or caves.- Pro-ceedings o the tenth International SpeleologicalCongress, 118120.

Lauritzen, S-E., 1991a: Uranium series dating o speleo-

thems. A glacial chronology or Nordland, Norwayor the last 600 Ka.- Striae, 34, 127132.

Lauritzen, S-E., 1991b: Karst resources and their conser-vation in Norway.- Norsk Geografisk idsskri, 45,119142.

Lowe, D. J., 1992: Te origin o limestone caverns: an in-ception horizon hypothesis.- PhD Tesis, Manches-ter Metropolitan University.

REVOR FAULKNER


15/16


Lowe, D. J. & J. Gunn, 1997: Carbonate speleogenesis: AnInception Horizon Hypothesis.- Acta Carsologica,26/2, 38, 457488.

Matsuoka, N., 2001: Direct observation o rost wedgingin Alpine bedrock.- Earth Surace Processes andLandorms, 26, 601614.

Muir-Wood, R., 2000: Deglaciation seismotectonics: aprincipal influence on intraplate seismogenesis athigh latitudes.- Quaternary Science Reviews, 19,13991411.

Mrner, N-A., 1979: Te Fennoscandian Upli and LateCenozoic Geodynamics: Geological Evidence.-Geojournal, 3, 3, 287318.

Mrner, N-A., 1980: Te Fennoscandian Upli: geologi-cal data and their geodynamical implication.- in N-A. Mrner (Ed.), Earth Rheology, Isostasy and Eu-stasy.- Wiley, 251284.

Mrner, N-A., P. E. ren, R. Sjberg, D. Grant, S.

Dawson, C. Bronge, O. Kvamsdal & A. Sidn, 2000:Deglacial paleoseismicity in Sweden: the 9663 BPIggesund event.- Quaternary Science Reviews, 19,14611468.

Mrner, N-A. (Ed.), 2003: Paleoseismicity o Sweden: anovel paradigm.- Stockholm University, 320 pp.

Olesen, O., 1988: Te Stuoragurra Fault, evidence o neo-tectonics in the Precambrian o Finnmark, northernNorway.- Norsk Geologisk idsskri, 68, 107118.

Olesen, O., H. Henkel, O. B. Lile, E. Mauring, J. S. Rn-ning & . H. orsvik, 1992: Neotectonics in the Pre-cambrian o Finnmark, northern Norway.- Norsk

Geologisk idsskri, 72, 301306.Olesen, O., S. Gjelle, H. Henkel, . A. Karlsen, L. Olsen &

. Skogseth, 1995: Neotectonics in the Ranafordenarea, northern Norway.- Norges Geologiske Under-skelse Bulletin, 427, 58.

Olesen, O., L. H. Blikra, A. Braathen, J. F. Dehls, L. Olsen,L. Rise, D. Roberts, F. Riis, J. I. Faleide & E. Anda,2004: Neotectonic deormation in Norway and itsimplications: a review.- Norwegian Journal o Geol-ogy, 84, 334.

Onac, B. P., 1991: Contributions to the knowledge o thenorth Norway karst.- Studia Univ. Babes-Bolya,

Geographia, 36, 2, 3542.Randall, A. D., R. M. Francis, M. H. Frimpter & J. M. Em-

ery, 1988: Chapter 22, Region 19, Northeastern Appa-lachians.- in W. Back, J. S. Rosenshein & P. R. Seaber(Eds.), Hydrogeology, Geological Society o NorthAmerica, Te Geology o North America, v. O-2.

Riggs, A. C., W. J. Carr, P. . Kolesar & R. J. Hoffman, 1994:ectonic speleogenesis o Devils Hole, Nevada, andimplications or hydrogeology and the developmento long, continuous paleoenvironmental records.-Quaternary Research, 42, 241254.

Ringrose, P. S., P. Hancock, C. Fenton & C. A. Davenport,1991: Quaternary tectonic activity in Scotland.- inA. Forster, M. G. Culshaw, J. C. Cripps, J. A. Little& C. F. Moon (Eds.), Quaternary Engineering Geol-ogy, Geological Society Engineering Special Publi-cation No. 7, 679686.

Roberts, D., 2000: Reverse-slip offsets and axial rac-tures in road-cut boreholes rom the Caledonidesin Finnmark, northern Norway: neotectonic stressorientation ractures.- Quaternary Science Reviews,19, 14371445.

Rohr-orp, E., 1994: Present upli rates and groundwa-ter potential in Norwegian hard rocks.- Norges Ge-ologiske Underskelse Bulletin, 426, 4752.

Sasowsky, I. D. & W. B. White, 1994: Te role o stressrelease racturing in the development o cavernousporosity in carbonate aquiers.- Water ResourcesResearch, 30, 12, 35233530.

Sjberg, R., 1981a: unnel caves in Swedish non calcar-eous rocks.- Proceedings o the eight InternationalSpeleological Congress, 2, 652656.

Sjberg, R., 1981b: unnel Caves in Swedish ArcheanRocks.- Cave Science, 8, 3, 159167.

Sjberg, R., 1987: Caves as indicators o neotectonics inSweden.- Zeitschri ur Geomorphologie, N.F. Sup-pl. Bd 63, 141148.

Sjberg, R., 1988: Coastal caves indicating preglacial mor-phology in Norway.- Cave Science, 15, 3, 99103.

Sjberg, R., 1996a: avelsjbergets blockbrant och hurden kan ha bildats.- Grottan, 31, 1, 3640.

Sjberg, R., 1996b: Seismotektoniken ntligen erknd !(?).-Grottan, 31, 2, 48.

Sjberg, R., 1996c: Lervarv, grottor och jordskalv.- Grot-tan, 31, 2, 2931.

Sjberg, R., 1999a: Forskningsproject kring Bodagrot-torna.- Grottan, 34, 2, 2834.

Sjberg, R., 1999b: De stora jordbvningarna i Ume ochINQUA:s Sverige-exkursion 1999.- Grottan, 34, 3,48.

Stephansson, O. & H. Carlsson, 1980: Seismo-tecton-ics in Fennoscandia.- in N-A. Mrner (Ed.), EarthRheology, Isostasy and Eustasy, 327338, Wiley.

Stewart, I. S., J. Sauber & J. Rose, 2000: Glacio-seismotec-tonics: ice sheets, crustal deormation and seismic-ity.- Quaternary Science Reviews, 19, 13671389.

Torson, R. M., 2000: Glacial tectonics: a deeper perspec-tive.- Quaternary Science Reviews, 19, 13911398.

Warwick, G. ., 1971: Caves and the Ice Age.- ransac-tions o the Cave Research Group o Great Britain,13, 2, 123130.



16/16

tectonic inception in caledonide marbles

Documents