Geo log i cal Quar terly, 2018, 62 (3): 553–562 DOI: http://dx.doi.org/10.7306/gq.1435
Coastal dune dy nam ics along the north ern Curonian Spit, Lith u a nia:
to ward an in te grated da ta base
Albertas BITINAS1, Nikita DOBROTIN2, 3, Ilya V. BUYNEVICH4, Anatoly MOLODKOV5, Aldona DAMUYT#6 and Donatas PUPIENIS7, *
1 KlaipÅda Uni ver sity, Ma rine Re search In sti tute, 92294 KlaipÅda, Lith u a nia
2 KlaipÅda Uni ver sity, Fac ulty of Ma rine Tech nol o gies and Nat u ral Sci ences, 92294 KlaipÅda, Lith u a nia
3 Vilnius Uni ver sity, In sti tute of Geosciences, 03101 Vilnius, Lith u a nia
4 Tem ple Uni ver sity, De part ment of Earth and En vi ron men tal Sci ence, Phil a del phia, PA 19122, USA
5 Tallinn Uni ver sity of Tech nol ogy In sti tute of Ge ol ogy, 19086 Tallinn, Es to nia
6 Lith u a nian Geo log i cal Sur vey, 03123 Vilnius, Lith u a nia
7 Na ture Re search Cen tre, In sti tute Ge ol ogy and Ge og ra phy, 08412 Vilnius, Lith u a nia
Bitinas, A., Dobrotin, N., Buynevich, I.V., Molodkov, A., DamuëytÅ, A., Pupienis, D., 2018. Coastal dune dy nam ics along the north ern Curonian Spit, Lith u a nia: to ward an in te grated da ta base. Geo log i cal Quar terly, 62 (3): 553–562, doi: 10.7306/gq.1435 As so ci ate ed i tor: Anna Wysocka
Sand dunes are the most prom i nent sub jects of geo log i cal and geomorphological in ter est along the Curonian Spit – a mega-bar rier that sep a rates the Curonian La goon from the Bal tic Sea. To date, an as sess ment of var i ous pa ram e ters of mi - grat ing dunes along the spit has been based on com par a tive anal y sis of old maps or ae rial and sat el lite im ages, as well as geo detic mea sure ments. These in ves ti ga tions have al lowed as sess ment of dune dy nam ics over a rel a tively short his tor i cal pe riod (~1700s to pres ent). The most re cent de tailed in ves ti ga tions of the Dead (Grey) Dunes along the Lith u a nian part of the spit us ing ground-pen e trat ing ra dar (GPR) and mag netic sus cep ti bil ity (MS) sur veys, sup ported by a ra dio car bon (14C) chro no log i cal frame work of palaeosols and in fra red op ti cally stim u lated lu mi nes cence (IR-OSL) ages of sand ho ri zons, have ad vanced our un der stand ing of ae olian land scape evo lu tion. The in ter pre ta tion of dune ac tiv ity and sta bil ity phases has been gen er ally based on IR-OSL dat ing re sults of the sand lay ers lo cated be tween ra dio car bon-dated palaeosols. How ever, the in flu ence of soil-form ing pro cesses on the IR-OSL dat ing re sults re lated to pos si ble mi gra tion of nat u ral ra dio ac tive iso - topes via ae olian sand lay ers has not been pre vi ously con sid ered. Hy poth e ses of dune re ac ti va tion and mi gra tion caused by abrupt re gional cli mate shifts, cat a strophic for est fires, anthropogenic in flu ence, and more lo cal forcings have been tested.
An in te grated ap proach to dune in ves ti ga tions has of fered an es ti mate of the rates of sand ac cu mu la tion and key phases of ae olian dy nam ics dur ing both stormy and calm pe ri ods, as well as helped to ex tend the re cord of dune evo lu tion to the mid-Ho lo cene. The palaeoenvironmental and palaeodynamic re con struc tions of the Dead Dunes sug gest that this mid-Ho lo - cene phase of dune ac tiv ity was of a lo cal char ac ter and likely did not ex ceed sev eral cen tu ries.
Key words: palaeosols, ground-pen e trat ing ra dar (GPR), IR-OSL, ra dio car bon, mag netic sus cep ti bil ity.
INTRODUCTION
Sand dunes are prom i nent geomorphic fea tures of most coastal bar ri ers along the south ern Bal tic Sea and serve as ar - chives of re gional cli mate shifts, veg e ta tion dy nam ics, sed i - ment avail abil ity, and hu man-land scape in ter ac tion. How ever,
the sheer size of some dunes makes their in ves ti ga tion a chal - lenge. The pres ent study is fo cused on the north ern Curonian Spit – a mas sive sandy mega-bar rier along the south east ern Bal tic that sep a rates the Curonian La goon from the Bal tic Sea (Fig. 1). This landform is split ad min is tra tively in half by the Re - pub lic of Lith u a nia (north) and the Rus sian Fed er a tion (south), with both parts hav ing the sta tus of Na tional Parks (Buèas, 2001) and in cluded in the UNESCO list of cul tural her i tage mon - u ments. From the geo log i cal point of view it is still an ac tive en - vi ron ment dom i nated by ae olian pro cesses that shape the most prom i nent land forms along the en tire ~100 km length of the spit.
The re search has been con cen trated along the west ern slope of the Great Dune Ridge (GDR) that stretches along the la goon coast of the spit. This ridge con tains the high est dunes in North -
* Corresponding author, e-mail: donatas.pupienis@gf.vu.lt Received: February 7, 2018; accepted: May 24, 2018; first published online: October 25, 2018
Fig. 1. Study area on the Curonian Spit
Lo ca tion of the dune sites in ves ti gated in de tail is marked as fol lows: V – Vingis dune site, N – Naglis dune site
ern Eu rope, that reach >60 m in many ar eas along the spit. This landform has re ceived spe cial at ten tion from re search ers be - gin ning in the sec ond half of the 19th cen tury (Schumann, 1861; see Wichdorff, 1919; Paul, 1944). These in ves ti ga tions pro posed that the an cient dune ridge stretched rel a tively evenly along the en tire area of the Curonian Spit and was only re cently remolded into the GDR. This pro cess started in the 16th cen - tury due to ex tremely high ae olian ac tiv ity in flu enced by de - struc tive hu man prac tices, for in stance af ter the near com plete clear-cut ting of for ests (Gudelis and MichaliukaitÅ, 1976; Gu - delis, 1989–1990, 1998a, b).
Un til re cently, the as sess ment of var i ous as pects of dune dy nam ics along the Curonian Spit has been based on com par - a tive anal y sis of old maps or aero- and sat el lite-based im ages, as well as geo detic mea sure ments that al lowed gen eral re con - struc tion of dune palaeo ge ogra phy and ac tiv ity for the rel a tively re cent (~1700s to pres ent) his tor i cal pe riod (Minkevièius, 1982;
MardosienÅ, 1988; Kazakevièius, 1989–1990; MorkñnaitÅ and Èesnulevièius, 2005; Povilanskas et al., 2009; Èesnulevièius et al., 2016, 2017). Sev eral chronosequences of bur ied palaeo - sols wide spread along the Great Dune Ridge served as im por - tant in di ca tors of rel a tive sta bil ity phases and formed the fo cus of a num ber of stud ies over the past de cades (e.g., Michaliu - kaitÅ, 1962, 1967; Gaigalas et al., 1991; Gudelis et al., 1993;
Moe et al., 2005; Gaigalas and Pazdur, 2008). Most of these in - ves ti ga tions in volved ra dio car bon (14C) dat ing or pol len anal y - sis of palaeosols. Over the past de cade, re newed geo log i cal re - search along the Curonian Spit added mod ern meth ods of in - ves ti ga tions: high-res o lu tion ground-pen e trat ing ra dar (GPR) im ag ing; im proved ra dio car bon dat ing of palaeosols by the AMS method; ab so lute age dat ing of sand lay ers us ing in fra red op ti cally stim u lated lu mi nes cence (IR-OSL); and in situ mea - sure ments of bulk low-field mag netic sus cep ti bil ity (MS). These in te grated in ves ti ga tions sub stan tially im proved the ac cu racy of subsurface map ping and age-dat ing, thereby greatly ad vanc ing our un der stand ing of ae olian land scape evo lu tion as far back as the mid-Ho lo cene (Buynevich et al., 2007a, b; Dobrotin et al., 2013). Ac cord ing to the re sults of 14C dat ing of key palaeo - sols, at least four soil-form ing gen er a tions have been dis tin - guished: 5800–4500 cal BP, 3900–3100 cal BP, 2600–2400 cal BP, and 1900 cal BP to pres ent (Dobrotin et al., 2013). Re cent work has re vealed that an cient for est fires re sulted in only lo cal re ac ti va tion of ae olian pro cesses and did not trig ger sub stan tial mi gra tion of the main dune ridge. Still, some is sues linked with the rate of sand ac cu mu la tion and pe cu liar i ties of dune palaeo - dynamics dur ing calm and stormy pe ri ods re mained un solved.
The aim of this pa per is to ad dress some of these im por tant as - pects of Curonian dune dy nam ics through anal y sis of geo - chron ol ogi cal datasets in their sedimentological and geo mor - pho logical con text.
REGIONAL SETTING
Two sites for de tailed in ves ti ga tions were cho sen in the cen tral part of the Curonian Spit, to the south of JuodkrantÅ set - tle ment (Fig. 1). At each site, the in ves ti ga tions were con cen - trated along transects with ex ist ing GPR cov er age (start – end co or di nates of pro file): the Vingis dune site: 55°27’35.07"N, 21°05’ 20.13"E–55°27’36.44"N, 21°05’23.98"E; and the Naglis dune site: 55°26’47.64"N, 21°05’1.51"E–55°26’48.88"N, 21°05’4.15"E.
The Vingis dune site is lo cated on the east ward slop ing flank of the GDR dune with mi nor re lief, at an el e va tion of
~10–11 m above mean sea level. The Naglis site has a sim i lar con text, but oc curs at a higher el e va tion of ~20–22 m, the geo - phys i cal pro files at both sites were es tab lished along the in ner sec tions of an cient dunes where the up per most parts were re - moved through sub se quent de fla tion. The ex posed sec tions of ae olian sed i ments are rep re sented by yel low or grey ish-yel low sand dom i nated by quartz (96–98%). The me dium-grained, in places coarse-grained, sand spo rad i cally con tains thin ho ri zons (typ i cally 0.5–3.0 cm thick) of black or dark grey fine-to-me - dium-grained heavy-min eral con cen tra tions (HMCs). Ac cord ing to re sults of pre vi ous min er al og i cal stud ies (Gudelis, 1989–1990), these HMCs are rep re sented by hornblende, gar - net, mag ne tite, il men ite, and epidote. In some in stances, litho - lo gical anom a lies are rep re sented by glauconite-coated quartz grains that can be rec og nized by their green ish-grey col our.
The 300 MHz GPR sur vey line in the Naglis dune site (Fig.
2) closely fol lowed an ear lier 200 MHz line (Buynevich et al., 2007a) and formed the ba sis for IR-OSL sam pling and MS mea sure ments be tween the two palaeosols P2 (older) and P1.
METHODS
The in ves ti ga tions into the re con struc tion of ae olian palaeo - dynamics con sisted of a suite of se quen tial field and lab o ra tory meth ods. The re search be gan with an in ter pre ta tion of air photo im ages and LIDAR data for iden ti fi ca tion of ex posed palaeosol ho ri zons through out the Dead Dunes. Sub se quently, it was com ple mented by ex ten sive field map ping, ground-pen e trat ing ra dar sur veys, and sam pling for ab so lute age de ter mi na tions.
Thus, both study sites were spe cif i cally se lected to span the sec tions be tween ad ja cent palaeosols that have been pre vi - ously dated by the ac cel er a tor mass-spec trom e try (AMS) ra dio - car bon method at the Na tional Ocean Sci ences fa cil ity of the Woods Hole Ocean o graphic In sti tu tion, USA (Buynevich et al., 2007a) and us ing a Tri-Carb0 3170TR/SL in the Ra dio iso tope Re search Lab o ra tory of the In sti tute of Ge ol ogy and Ge og ra phy of the Na ture Re search Cen tre in Vilnius, Lith u a nia (Dobrotin et al., 2013). This most re cent phase of in ves ti ga tions in cluded de tailed GPR im ag ing along mul ti ple dip and strike pro files, vi - sual lithological ex am i na tion and de scrip tion of ae olian sed i - ments, in situ mea sure ments of mag netic sus cep ti bly (MS) of the sed i ments, and sam pling. The lab o ra tory com po nent in - volved dat ing of ae olian sed i ments by the in fra red op ti cally stim u lated lu mi nes cence (IR-OSL) tech nique (Molodkov and Bitinas, 2008).
GROUND-PENETRATING RADAR (GPR) SURVEY
Two GPR pro files rang ing from 50 to 55 m in length were made at each lo ca tion be tween two palaeosols in transects per pen dic u lar to the strike of each soil layer (Fig. 2). The sur - vey setup in cluded a RADAR Sys tems Zond 12-e sys tem with a 300 MHz monostatic an tenna, 400 V pulse gen er a tor (op ti - mized for res o lu tion and depth), and a time win dow (range) of 100 ns. A rel a tive di elec tric permittivity of 5.33 was used for un sat u rated dune sand, achiev ing 6.0–6.5 m pen e tra tion at max i mum range. The EM sig nal ve loc ity of 13 m/ns was cor - rob o rated by hy per bola fit ting. All post-pro cess ing was per - formed with Halliburton “Geographix” soft ware. Palaeosols Coastal dune dynamics along the northern Curonian Spit, Lithuania: toward an integrated database 555
were traced from sur face ex po sures and in ter preted in radar - grams based on their strong sig nal re turns and char ac ter is tic ge om e try. Key subsurface re flec tors were iden ti fied us ing 1 m deep trenches (Fig. 3).
IR-OSL DATING
Po tas sium feld spar-based in fra red op ti cally stim u lated lu - mi nes cence (IR-OSL) age de ter mi na tions were car ried out at the Re search Lab o ra tory for Qua ter nary Geo chron ol ogy (RLQG), In sti tute of Ge ol ogy, Tallinn Uni ver sity of Tech nol ogy.
These palaeodosimetric dat ing tech niques have long been suc - cess fully used in ear lier in ves ti ga tions in the south east ern Bal -
tic re gion (Molodkov et al., 2010; Bitinas et al., 2011), in clud ing the Curonian Spit dunes and vi cin i ties of the Curonian La goon (Molodkov and Bitinas, 2008; Bitinas et al., 2017). The IR-OSL dat ing pro ce dure used at the RLQG is the sub ject of a spe cial meth od olog i cal pa per (Molodkov and Bitinas, 2008).
At the Vingis dune site, five bulk sam ples for IR-OSL dat ing were col lected be tween the two palaeosols (Fig. 4). The sam - ples were taken at 0.5 m depth at 5 m in ter vals be tween the sam ples. At the Naglis dune site, the five bulk sam ples were col lected from 0.5 m be low the dune sur face at 7 m spac ing (Fig. 5). At both sites, sam ples from the W or NW end of the pro - file were taken 10–15 cm above the older palaeosol, whereas those at the E or SE end were col lected ~15–20 cm be low the youn ger palaeosol.
Fig. 2. Dis tri bu tions of ex posed palaeosols along the west ern slope of the Great Dune Ridge – a brighter shade on the photo-im age in di cates ar eas free of veg e ta tion
The dot ted rect an gles (A) in di cate en larged in serts of the Naglis dune site (B) and the Vingis dune site (C) sur round ings: the GPR pro - files at both dune sites are marked by the let ters V and N re spec tively; the num bers un der the palaeosols show their age (in cal y BP)
es tab lished by 14C dat ing (af ter Buynevich et al., 2007a; Dobrotin et al., 2013)
MAGNETIC SUSCEPTIBILITY
Mea sure ments of low-field mag netic sus cep ti bil ity (MS) were col lected in situ in a 0.4 m deep trench ex ca vated along GPR pro files us ing a Bartington MS3 metre with a MS2K scan ning field sen sor (Buynevich et al., 2007a). MS val ues were ob tained on ex posed sec tions of dune slipfaces at all vi -
su ally dis tinct lithological changes. At quartz-dom i nated in ter - vals, mea sure ments were ob tained ev ery 10 cm. This al lowed di rect com par i son of mag netic prop er ties (pri mar ily mag ne tite con tent) of spe cific ho ri zons to their elec tro mag netic sig nal re - sponse as they were traced along the dip in GPR im ages. All mea sure ments are ex pressed as bulk sus cep ti bil ity (mSI).
Coastal dune dynamics along the northern Curonian Spit, Lithuania: toward an integrated database 557
Fig. 3. Seg ments of ex posed palaeosols on the Great Dune Ridge west ern slope in the vi cin ity of the Naglis dune site on the Dead (Grey) Dune ridge: a sin gle palaeosol ex posed by de fla tion (A); in many places, palaeosols
of dif fer ent age are sit u ated near each other (B)
Fig. 4. Pro file of de tailed in ves ti ga tions at the Vingis dune site
A – ground pen e trat ing ra dar (GPR) im age (radarogram); B – curve of mag netic sus cep - ti bil ity (MS); C – in ter pre ta tion of radargram; in the lat ter sec tion palaeosols are marked by solid lines and interlayers with high con cen tra tions of heavy min er als by dot ted lines;
the ra dio car bon age of the palaeosols (in cal y BP; af ter Buynevich et al., 2007a;
Dobrotin et al., 2013) is in cluded at the top; sam pling lo ca tions for IR-OSL dat ing are in - di cated by black cir cles; dis tance along the hor i zon tal scale is marked from the be gin - ning of the GPR pro file
RESULTS
GPR PROFILING
Geo phys i cal im ages from the Vingis dune site re veal two main types of subsurface re flec tor (Fig. 4). The most prom i nent re flec tors are the palaeosols that out crop along the pres ent dune sur face at hor i zon tal dis tances of ~8 and 32 m. Both re - flec tions dip east at ap prox i mately 6° (older) and 12° (youn ger palaeosol). Mean while the less well-ex pressed and shorter re - flec tors (oc cur ring at in ter vals from 2 to >20 cm) cor re spond to thin heavy-min eral con cen tra tions (HMCs) that also dip to the east. At the Naglis dune site (Fig. 5), the prom i nent re flec tions again be gin at palaeosol ex po sures at 9 and 24 m along the pro file. Both re flec tions dip to the SE at ~8° (older) and 6°
(youn ger palaeosol). The few sec ond ary re flec tors co in cide, as
they do at the Vingis dune site, with thin SE-dip ping HMC ho ri - zons. The in ter pre ta tion of GPR re flec tors is based on the ex po - sures of ae olian sec tions along GPR transects, as well as in the shal low trenches ex ca vated for IR-OSL sam pling.
DUNES AGE ACCORDING TO IR-OSL DATING RESULTS
The age of the dune seg ment lo cated be tween two palaeosols at the Vingis dune site gen er ally falls into the time span be tween 2.4 ±0.2 and 2.2 ±0.2 ka (Fig. 4; Ap pen dix 1*).
Only one sam ple (Fig. 4, KOPOS-6), lo cated di rectly be low the palaeosol dated at 2430 ±25 cal y BP, shows an older age – 2.8
±0.2 ka. Whereas, at the Naglis dune site, the IR-OSL dat ing re - sults dem on strate that the age of sand lay ers grad u ally in - creases from NW to SE – from 1.1 ±0.1 ka to 2.2 ±0.2 ka (Fig. 5, Ap pen dix 1).
Fig. 5. Pro file of de tailed in ves ti ga tions at the Naglis dune site For ex pla na tions see Fig ure 4
* Supplementary data associated with this article can be found, in the online version, at doi: 10.7306/gq.1435
MAGNETIC SUSCEPTIBILITY OF AEOLIAN SEDIMENTS
Mea sure ments of MS along both pro files show a range of val ues from 8–10 to 20–30 mSI at the Vingis dune site and from 5–10 to 500–1000 mSI (with one anom aly, up to 2027.5 mSI) at the Naglis site (Figs. 4 and 5). At the Vingis site, an in crease was largely as so ci ated with HMCs. At Naglis, rel a tively low val - ues of MS were char ac ter is tic of the NW (up wind) part of the pro file, whereas anom a lies >800–1000 mSI were com mon in the SE half of the pro file. The in ter vals with low val ues of MS are char ac ter is tic of sand lay ers com posed mostly of quartz and feld spar, with the high est val ues in vari ably as so ci ated with thin HMC interlayers (up to 4–5 cm) (Fig. 6).
DISCUSSION
The re sults of geo phys i cal im ag ing dem on strate the subsurface ex ten sion of ex posed palaeosols and HMC ho ri - zons, pro vid ing a means of their map ping and cor re la tion into ad ja cent dune sec tions. Mag netic sus cep ti bil ity val ues pro vide a use ful in de pend ent dataset of min er al og i cal anom a lies that are re spon si ble for spe cific GPR re flec tions. Whereas this as - pect of field in ves ti ga tions has been suc cess fully tested at the two sites, the bulk of at ten tion should be given to the chro no log - i cal con trol.
The re sults of in ves ti ga tions ob tained by dif fer ent meth ods of ab so lute chro nol ogy dem on strate the com ple men tary na ture of the 14C AMS dat ing of palaeosols and IR-OSl dat ing of in ter - ven ing sand units (Figs. 4 and 5). How ever, the par tic u lar na - ture of dat ing re sults at each site war rants fur ther treat ment.
Ra dio car bon dat ing sug gests that the dune sec tion in ves ti - gated in de tail at the Vingis site (Fig. 4) rep re sents ae olian ac - tiv ity that post-dates a sta ble phase that ended ~3400 ±35 cal y BP (older palaeosol). This dune se quence ac cu mu lated over a rel a tively short time span: be tween 2.4 and 2.2 ka ac cord ing to IR-OSL dat ing re sults (Fig. 7). Af ter ae olian ac tiv ity ceased, the
new palaeosol (dated as 2430 ±25 cal y BP) started to form.
Only one anom a lous IR-OSL sam ple (2.8 ±0.2 ka) lo cated di - rectly be low the youn ger palaeosol (Fig. 4, Kopos-6) is not con - sis tent with this in ter pre ta tion.
Re con struc tion of ae olian ac tiv ity at the Naglis dune site is more prob lem atic. The IR-OSL dat ing re sults show that sand ac - cu mu la tion took place over a rel a tively long time span that be gan be fore 2.4 ka and con tin ued un til 1000 years be fore pres ent (Fig.
7). Based on the suite of IR-OSL ages, three sam ples fall within the rel a tively short time pe riod be fore ~1.4–1.2 ka BP (Fig. 7, sam ples Kopos-3, 4, and 5). Thus, this time span likely re flects the ac tual age of ae olian sed i men ta tion. In this way, the in ter pre - ta tion of IR-OSL dat ing re sults does not con tra dict the re sults of
14C AMS dat ing of palaeosols, i.e. 1350 ±45 and 850 ±35 cal yBP, re spec tively (Buynevich et al., 2007a).
The most prob lem atic is sue of the IR-OSL data in ter pre ta - tion re lates to the fact that the old est dates at both sites (Kopos-1 and Kopos-6) were ob tained from the strati gra phically youn gest sand ho ri zon. One ex pla na tion re lates to a pos si ble in flu ence of pedogenic pro cesses, given that the old est dates come from sam ples col lected di rectly be neath the palaeosol ho ri zons. As shown in Ap pen dix 1, very low con tents of ura nium (0.01 and 0.08 ppm) are char ac ter is tic of both sam ples (Kopos-1 and 6). This could have in flu enced the re sults of IR-OSL dat ing. The prev a lence of ura nium in sandy sed i ments Coastal dune dynamics along the northern Curonian Spit, Lithuania: toward an integrated database 559
Fig. 6. Sec tion of ae olian sed i ments along the GPR pro file at the Naglis dune site
Above dot ted line – de fla tion sur face with mea sure ment places marked by black squares, be low – ver ti cal wall of trench, in situ val - ues of mag netic sus cep ti bil ity (mSI) dem on strate a sub stan tial con - trast be tween quartz sand (low est val ues) and a dark grey sand interlayer with HMCs (high est value)
Fig. 7. Re sults of IR-OSL dat ing from the Vingis and Naglis dune sites Pro posed pe ri ods of ae olian sed i men ta tion are in di cated
by grey shad ing
is as so ci ated with com mon sol u ble salts of (UO2)2 such as ni - trate, chlo ride, ac e tate, sul phate, and car bon ate that may be easy dis solved dur ing soil-form ing pro cesses (Boyle, 1982). As a re sult, ura nium can be trans ported (eluviated) into un der ly ing sed i ments (Wintle, 2008). Tak ing into ac count the mor phol ogy of the an cient dune, the sam ple Kopos-1 at the Naglis dune site (Fig. 5), may also have been partly in flu enced by Late Ho lo cene pedogenesis. How ever, as sum ing an av er age of eight ura nium con tent val ues be tween 0.20–0.96 ppm (av er age: 0.47 ppm), the cor rected age could be youn ger by only 200 years if no U mi gra tion dur ing soil-form ing pro cesses had oc curred. There - fore, pedogenesis is un likely to ex plain the old age of the stratigraphically youn gest sam ples in both ar eas. Thus, an al - ter na tive ex pla na tion is pro posed.
In the ini tial stage of dune move ment oc cur ring in calm en vi - ron men tal con di tions, low to mod er ate wind pref er en tially re - moves the fin est sand frac tion. There fore, the sand par ti cles, be fore cas cad ing down the steeper, lee ward side (slipface) of the dune, were trans ported for a long enough pe riod to cause full bleach ing. Thus, the IR-OSL age of sand interlayers de pos - ited dur ing the ini tial stage of sec ond ary dune for ma tion shows the real age of sed i men ta tion (~2.4–2.2 ka at the Vingis site and 1.4–1.2 ka at the Naglis site). Dur ing the sec ond half of the ae - olian phase (cor re spond ing to the an oma lously old age of Kopos-6 at the Vingis site and two ages, Kopos-1 and Kopos-2, at the Naglis one) the sed i men ta tion likely oc curred dur ing a rel - a tively stormy pe riod. Thus, be cause sand was de flated from the pri mary dune, quickly trans ported and de pos ited, there was lit tle or no time for bleach ing, thereby re tain ing the “old” IR-OSL sig nal (i.e. with progenetic palaeodose-re lated ef fects in the min eral lat tice). This may ex plain the an oma lously old dates at both sites co in cid ing with the stratigraphically youn ger dune sand. More over, the higher frac tion of heavy min er als, even in thin lay ers, may have in flu enced the palaeodose, re sult ing in older ages.
Us ing multi-dat ing meth ods of ab so lute geo chron ol ogy and mak ing intercomparison re quires con sid er ation of key fac tors that may in flu ence each tech nique. In our case, the 14C and IR-OSL meth ods have dif fer ent pre ci sion and an a lyt i cal er rors by de cades and cen tu ries, re spec tively. Thus, ac cu rate es tab lish ing of ab so lute chro nol ogy should pro ceed with cau - tion. For ex am ple, at the Naglis dune site, the age of the older palaeosol (1350 ±45 cal y BP) and the sand layer di rectly above it (1.1 ±0.1 ka) ap pear in re verse or der. But, tak ing into ac count the pre vi ously men tioned dif fer ences in mea sure - ment er rors and the ob served tem po ral se quence of IR-OSL dates at this lo cal ity, it is pos si ble to ac cept these dates as rel - a tively re li able. More over, for the 14C AMS sam pling we used the lumps of char coal (rem nants of burned trees or bushes), as the best dat ing ma te rial. Tak ing into ac count that the de vel - op ment of a nor mal soil ho ri zon in or ganic-poor sand of the Curonian Spit takes up to 1100–1600 years (Peyart, 2007), the dated char coal rep re sents only a few de cades of this pro - longed soil-form ing pe riod (Dobrotin et al., 2013). There fore, we sug gest that the dated char coal rep re sents the lat est part of the pedogenesis, when the palaeosol was thick enough and en riched in or ganic mat ter, i.e. dur ing con di tions more suit able for higher plant growth.
Fluc tu a tions of the MS val ues of the sandy sed i ments be - tween the two palaeosols at the Vingis dune site re flect the gen - eral dy namic con di tions above the dune sur face at the time of ae olian ac tiv ity phases. Whereas the slipface dip re flects the di - rec tion of trans port, HMCs (i.e., the min er al og i cal den sity lag) rep re sent pe ri ods of in creased near-sur face wind speed (Buynevich et al., 2007a). The sec tions of the MS dataset dom i -
nated by low val ues (10–30 mSI) sug gest that rel a tively calm con di tions ex isted on the Curonian Spit be fore 2.4–2.2 ka BP.
Dur ing the next phase, at ~1.4–1.2 ka BP, the palaeoclimatic con di tions were dif fer ent. MS mea sure ments sug gest the first half of this pe riod was rel a tively calm, as only two HMC anom a - lies oc cur dur ing the sand ac cu mu la tion pro cess (Fig. 5). The sec ond half of this pe riod (the ~20–34 m in ter val) ex pe ri enced more dy namic trans port (MS anom a lies >800 mSI). In this part of the pro file, many HMCs that yielded high MS val ues are very thin (<2 cm), so the ma jor ity of them are not re solved in GPR im ages. In ves ti ga tions of mid and Late Ho lo cene ae olian ac tiv - ity in north west ern Eu rope by a num ber of re search ers (Jelgersma et al., 1995; Clemmensen et al., 2009; Niel sen et al., 2016) in di cate that storm i ness max ima re curred at ca. 4300, 2800, 1400 and 400 cal BP (Sor rel et al., 2012). The HMCs and an oma lously old age of Kopos-6 at the Vingis site as well as the Kopos-1 and Kopos-2 ages at Naglis be long to Ho lo cene stor - m i ness phases III (3300–2400 cal BP) and IV (1900–1050 cal BP; Sor rel et al., 2012).
In sum mary, in ves ti ga tions dur ing the 20th cen tury has led to a dom i nant opin ion that ae olian re ac ti va tion and de po si tion were pro longed pro cesses, pos si bly linked with large-scale changes in Ho lo cene cli mate (e.g., Borówka, 1975). These pe - ri ods of un in ter rupted ae olian ac tiv ity (the so-called “aeolo - dynamic stages” of Gudelis, 1998b) were dom i nant dur ing the past two mil len nia. The re sults of our study cor rob o rate re cent find ings (Jelgersma et al., 1995; Wil son et al., 2004; Clemmen - sen et al., 2009; Niel sen et al., 2016), which sug gest that dune ac ti va tion was trig gered pri mar ily by global or hemi spheric fac - tors such as abrupt cli ma tic shifts and storm i ness (Bond et al., 1997; Mayewski et al., 2004; Sor rel et al., 2012). Sec ondly, re-ac ti va tion of ae olian pro cesses and re-de po si tion of dune sand was likely caused by lo cal fac tors: nat u ral (or hu man-in - duced) for est fires, as well as de for es ta tion (in his tor i cal times) that op er ated on de cad al to cen ten nial scales (Moe et al., 2005;
Buynevich et al., 2007a; Gaigalas and Pazdur, 2008; Dobrotin et al., 2013).
CONCLUSIONS
A sub stan tial in ten si fi ca tion of ae olian pro cesses along parts of a mas sive Curonian dune ridge be gan at least in the mid-Ho lo cene and was not linked to large-scale cli ma tic chan - ge. At the two study sites, phases of sed i ment remobilization over the past 3,000 years were lim ited in both space (e.g., trig - gered by lo cal for est fires and de for es ta tion) and du ra tion (last - ing no lon ger than one to two cen tu ries). At pres ent, ae olian ac - tiv ity is con fined to sev eral ex ten sive unvegetated sec tions, in - clud ing a con tin u ing slow ad vance by slipface mi gra tion into the Curonian La goon.
Ground-pen e trat ing ra dar im ages and mag netic sus cep ti bil - ity trends within ae olian sed i ments that sep a rate palaeosols of dif fer ent gen er a tions dem on strate a vari a tion in heavy-min eral con tent that can be re lated to near-sur face con di tions (rel a tively calm, windy, or stormy re gime), which char ac ter ized var i ous phases of dune ac tiv ity. When in te grat ing ab so lute geo chron ol - ogi cal datasets from ad ja cent ae olian sand in ter vals and palaeosols, it is nec es sary to con sider geo chem i cal pro cesses re lated to both pedogenesis and the back ground ra di a tion re - gime of the non-quartz frac tion. These may have an im por tant in flu ence on bulk-sam pled op ti cal dat ing re sults and should be ad dressed in the fu ture if an ac cu rate chro nol ogy of dune ac tiv - ity is to be es tab lished for this coastal re gion.
Ac knowl edge ments. This study was sup ported by the KlaipÅda Uni ver sity grant of the na tional pro ject “Lith u a nian Mar i time Sec tors Tech nol o gies and En vi ron men tal Re search De vel op ment” No. VP1-3.1-MM-08-K-01-019. We thank the Curonian Spit Na tional Park for ac cess and V. Gregorau skienÅ
for as sis tance in field work and ad vice on top ics re lated to soil evo lu tion. We thank two anon y mous re view ers for im prov ing the manu script.
REFERENCES
Bitinas, A., DamuëytÅ, A., Molodkov, A., 2011. Geo log i cal struc - ture of the Qua ter nary sed i men tary se quence in the KlaipÅda Strait, south east ern Bal tic. In: Bal tic Sea Ba sin Book Se ries:
Cen tral and East ern Eu ro pean De vel op ment Stud ies (eds. J.
Harff, S. Bjorck and P. Hoth): 133–146. Springer.
Bitinas, A., Mažeika, J., Buynevich, I.V., DamuëytÅ, A., Molodkov, A., GrigienÅ, A., 2017. Con straints of ra dio car bon dat ing in south east ern Bal tic la goons: as sess ing the vi tal ef - fects. In: Coast line Changes of the Bal tic Sea from South to East – Past and Fu ture Pro jec tions (eds. J. Harff, K. Furmañczyk and H. van Storch): 137–171. Springer.
Bond, G., Show ers, W., Cheseby, M., Lotti, R., Almasi, P., deMenocal, P., Priore, P., Cul len, H., Hajdas, I., Bonani, G., 1997. A per va sive mil len nial-scale cy cle in North At lan tic Ho lo - cene and Gla cial cli mates. Sci ence, 278: 1257–1266.
Borówka, R.K., 1975. Prob lem of the mor phol ogy of fos sil dune forms on the £eba bar. Quaestiones Geographicae, 2: 39–51.
Boyle, R.W., 1982. Geo chem i cal Pros pect ing for Tho rium and Ura - nium De pos its. Elsevier, Am ster dam and New York.
Buèas, J., 2001. Curonian Spit Na tional Park (in Lith u a nian).
Savastis, Vilnius.
Buynevich, I.V., Bitinas, A., Pupienis, D., 2007a. Lithological anom a lies in a rel ict coastal dune: geo phys i cal and paleo - environ mental mark ers. Geo phys i cal Re search Let ters, 34:
L09707, doi: 10.1029/2007GL029767
Buynevich, I.V., Bitinas, A., Pupienis, D., 2007b. Re ac ti va tion of coastal dunes doc u mented by subsurface im ag ing of the Great Dune Ridge, Lith u a nia. Jour nal of Coastal Re search, SI50:
226–230.
Èesnulevièius, A., BautrÅnas, A., MorkñnaitÅ, R., KavoliutÅ, F., 2016. In ten sity of ae olian pro cesses on Curonian Spit (the end of the XX cen tury – be gin ning the XXI cen tury). Se lected Is sues of Ae olian Geo mor phol ogy, Jagiellonian Uni ver sity, Cra cow, Po - land: 111–130.
Èesnulevièius, A., MorkñnaitÅ, R., BautrÅnas, A., Bevainis, L., Ovodas, D., 2017. In ten sity of geodynamic pro cesses in the Lith u a nian part of the Curonian Spit. Earth Dy namic Sys tem, 8:
419–428.
Clemmensen, L.B, Murray, A., Heinemeier, J., 2009. The evo lu - tion of Ho lo cene coastal dunefields, Jutland, Den mark: a re cord of cli mate change over the past 5000 years. Geo mor phol ogy, 105: 303–313.
Dobrotin, N., Bitinas, A., Michelevièius, D., DamuëytÅ, A., Maûeika, J., 2013. Re con struc tion of the Dead (Grey) Dune evo lu tion along the Curonian Spit, South east ern Bal tic. Bul le tin of the Geo log i cal So ci ety of Fin land, 85: 53–64.
Gaigalas, A., Banys, J., Gulbinskas, S., SavukynienÅ, N., 1991.
The ra dio car bon age of the bur ied soils in the dunes of the Kurëi÷ Nerija spit (in Rus sian). In: Geo chron ol ogi cal and iso to - pic-geo chem i cal in ves ti ga tions in the Qua ter nary ge ol ogy and ar chae ol ogy (ed. A. Gaigalas): 8–13, Vilnius.
Gaigalas, A., Pazdur, A., 2008. Chro nol ogy of bur ied soils, for est fires and ex treme mi gra tion of dunes on the Kurëi÷ nerija spit (Lith u a nian coast). Landform Anal y sis, 9: 187–191.
Goble, R.J., Ma son, J.A., Loope, D.B., Swinehart, J.B., 2004. Op - ti cal and ra dio car bon ages of stacked paleosols and dune sands
in the Ne braska Sand Hills, USA. Qua ter nary Sci ence Re views, 23: 1173–1182.
Gudelis, V., 1989–1990. Li thol ogy of old par a bolic dunes of the Curonian Spit and the coastal dy nam ics of the late Litorina Sea (in Lith u a nian). The Geo graph ical Year book, 25–26: 13–17.
Gudelis, V., 1998a. The Lith u a nian off shore and coast of the Bal tic Sea (in Lith u a nian). Lith u a nian Acad emy of Sci ences Mono - graph, Ac a de mia, Vilnius.
Gudelis, V., 1998b. A cat a strophic dune for est fire on the Kurëi÷
Nerija Spit (Lith u a nian Coast) and its im pact on the coastal pop - u la tion in the Late Neo lithic time. Pact, 54: 45–50.
Gudelis, V., KlimavièienÅ, V., SavukynienÅ, N., 1993. Old bur ied for est soils in the Kurëi÷ Nerija spit and their palynological pat - tern (in Lith u a nian). In: Bal tic Sea Coastal Dy nam ics and Paleogeography Is sues, 2 (ed. V. Gudelis): 64–93. Ac a de mia, Vilnius.
Gudelis, V., MichaliukaitÅ, E., 1976. The old par a bolic dunes of the Curonian Spit (in Rus sian). Geographia Lituanica: 59–63.
Jelgersma, S., Stive, M.J.F., van der Valk, L., 1995. Ho lo cene storm surge sig na tures in the coastal dunes of the west ern Neth - er lands. Ma rine Ge ol ogy, 125: 95–110.
Kazakevièius, S., 1989–1990. Dy nam ics of the Curonian Spit coastal de vel op ment (car to graphic anal y sis) (in Lith u a nian).
The Geo graph ical Year book, 25–26: 46–56.
MardosienÅ, D., 1988. Dy nam ics of High Dunes in the Curonian Spit (in Lith u a nian). The Geo graph ical Year book, 24: 114–119.
Mayewski, P.A., Rohling, E.E., Curt Stager, J., Karlén, W., Maasch, K.A., Da vid Meeker, L., Meyerson, E.A., Gasse, F., van Kreveld, S., Holmgren, K., Lee-Thorp, J., Rosqvist, G., Rack, F., Staubwasser, M., Schnei der, R.R., Steig, E.J., 2004.
Ho lo cene cli mate vari abil ity. Qua ter nary Re search, 62: 243–255.
MichaliukaitÅ, E., 1962. The old dunes and paleosols of the Curonian Spit (in Lith u a nian). The Geo graph ical Year book, 5:
377–390.
MichaliukaitÅ, E., 1967. Dy nam ics of Curonian Spit shores and dune by last 100 years (in Lith u a nian). The Geo graph ical Year - book, 8: 97–115.
Minkevièius, V., 1982. Eolian re lief of the Bal tic Sea cost (in Lith u a - nian). The Geo graph ical Year book, 20: 156–160.
Moe, D., SavukynienÅ, N., StanèikatÅ, M., 2005. A new 14C (AMS) date from for mer heath land soil ho ri zons at Kurëi÷ Nerija, Lith u - a nia. Baltica, 18: 23–28.
Molodkov, A., Bitinas, A., 2008. Sed i men tary re cord and lu mi nes - cence chro nol ogy of the Late gla cial and Ho lo cene ae olian sed i - ments in Lith u a nia. Boreas, 35: 244–254.
Molodkov, A., Bitinas, A., DamuëytÅ, A., 2010. IR-OSL stud ies of till and inter-till de pos its from the Lith u a nian Mar i time Re gion.
Qua ter nary Geochronoloy, 5: 263–268.
MorkñnaitÅ, R., Èesnulevièius, A., 2005. Changes in blow out seg - ments of the Main Ridge in the Curonian Spit in 1999–2003.
Acta Zoologica Lituanica, 15: 145–150.
Niel sen, P.R., Dahl, S.O., Jansen, H.L., 2016. Mid- to late Ho lo - cene ae olian ac tiv ity re corded in a coastal dunefield and lac us - trine sed i ments on AndÝya, north ern Nor way. The Ho lo cene, 26:
1–16.
Coastal dune dynamics along the northern Curonian Spit, Lithuania: toward an integrated database 561
Paul, K.H., 1944. Morphologie und Veg e ta tion der Kurischen Nehrung. Nova Acta Leopoldina, 132 (96), Halle (Sante).
Peyrat, J., 2007. De vel op ment, prop er ties and clas si fi ca tion of dune soils in the Curonian Spit Na tional Park, Rus sian part.
Geologija, 59: 59–64.
Povilanskas, R., Baghdasarian, H., Arakeylyan, S., Satkñnas, J., Taminskas, J., 2009. Sec u lar morphodynamic trends of the Ho - lo cene dune ridge on the Curonian Spit. Jour nal of Coastal Re - search, 25: 209–215.
Schumann, J., 1861. Die wandernden Dünen auf der Kurischen Nehrung. Mittheilungen aus Justus Perthes’ geographisher An - stalt über wichtige neue Erforshungen auf dem Gesammtge - biete der Geographie, 5: 196–197.
Sor rel, P., Debret, M., Billeaud, I., Jaccard, S.L., McManus, J.F., Tessier, B., 2012. Per sis tent non-so lar forc ing of Ho lo cene storm dy nam ics in coastal sed i men tary ar chives. Nat u ral Geo - sciences, 5: 892–896.
Wichdorff, von H., 1919. Geologie der Kurischen Nehrung. Abhan - dlungen der Preußischen Geologischen Landes mannschaft Neue Folge, 77: 1–179.
Wil son, P., McGourty, J.M., Bate man, M.D., 2004. Mid- to late-Ho - lo cene coastal dune event stra tig ra phy for the north coast of North ern Ire land. Ho lo cene, 14: 406–416.
Wintle, A.G., 2008. Lu mi nes cence dat ing: where it has been and where it is go ing. Boreas, 37: 471–482.
