Catena 34 Ž1999. 277–291

Thermoluminescence dates and palaeoenvironmental information of the late Quaternary sand deposits, Tierra de Pinares, Central Spain Mark D. Bateman

a,)

, Andres ´ Dıez ´ Herrero

b

a

Sheffield Centre for International Drylands Research: UniÕersity of Sheffield, Winter St., Sheffield S10 2TN, UK b Departamento de Medio Ambiente, Facultad de Ciencias, UniÕersidad Europea de Madrid, VillaÕiciosa de Odon 28.670 Madrid, Spain Received 2 April 1997; revised 21 April 1998; accepted 10 September 1998

Abstract A large amount of Quaternary palaeo-environmental evidence lies within the terrestrial aeolian sediments filling the Duero basin in central Spain. Previous work has identified a number of stratigraphical units although the absolute age and origin of some of these is equivocal. The uppermost unit, the focus of this paper, is a widely distributed sand which covers most of the Tierra de Pinares, north of Segovia. Detailed investigations of an exposure at Burgomillodo reveal over 35 m of laminated and cross-bedded sand from which seven thermoluminescence dates were obtained. The bedding structures indicate a northwesterly palaeowind direction, different from the present dominating westerly winds. Two phases of sand deposition are identified at the site with evidence of erosion and reactivation phases. An upper unit of ca. 7 ka during the warm-aridity of the Atlantic period and a lower unit between 12.5–11 ka coinciding with the cold-aridity of the Younger Dryas event. The latter, in which at least 25 m of sand were deposited at the site, is the dominant phase and is coincident with what has been identified elsewhere in Spain from biological data as a period which was both cold and dry. The Burgomillodos site, therefore, represents the first sedimentological evidence for this arid phase in the Younger Dryas in central Spain. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Younger Dryas; Aeolian sand; Arid environment; Iberian peninsular

)

Corresponding author. Tel.: q44-114-222-7929; Fax: q44-114-279-7912; E-mail: [email protected] 0341-8162r99r$19.00 q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 4 1 - 8 1 6 2 Ž 9 8 . 0 0 1 1 3 - 1

278

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

1. Introduction To date very little has been published in the English language literature about the extensive Quaternary deposits found in central Spain with the notable exceptions of Arche Ž1983., Alonso and Garzon ˜ Ž1994. and Rendell et al. Ž1994, 1996.. The central Spanish basins contain a terrestrial sedimentary record which spans large parts of the Quaternary period. This source of palaeo-environmental and palaeo-climatic information is yet to be utilised. As part of an ongoing investigation of the Quaternary evolution of the southern Duero basin, this paper focuses on the uppermost sedimentary unit which is a widespread surficial sand. A review of previous research shows that both its origin and age are contentious with chronological assignments ranging from early Quaternary ŽHernandez Pacheco, 1923; Perez-Gonzalez, 1982. through to late Pleistocene or ´ ´ ´ Holocene ŽFernandez, 1987, 1988; Perez-Gonzalez ´ ´ ´ et al., 1994.. The sand’s stability at present infers that different climaticrenvironmental conditions were responsible for its deposition. In this paper we report and interpret results on an investigation into the extensive sand deposits around Burgomillodo. Its object is to interpret the timing and palaeowind direction of the aeolian phases of sand deposition. 2. Geological and geographical context The Tierra de Pinares is positioned on the southern edge of the Duero sedimentary basin close to the northern part of the Variscian Massif which forms the Sierra de Guadarrama mountains. Approximately 1150 km2 in size, it is bounded on the north, east and west by the Duero, Zapardiel and Duraton ´ rivers respectively with its meridional edge lying approximately between Arevalo and the city of Cantalejo ŽFig. 1.. ´ The pine forest, from where it derives its name, is primarily of Pinus pinaster with minor amounts of Pinus pinea and this extends, excluding settlements and river valleys, over the entire area. The Tierra de Pinares is part of a plateau lying between 750–950 m a.s.l. which is gently tilted towards the northwest. This meseta has been dissected by tributaries of the Duero river ŽAdaja, Eresma–Voltoya and Cega–Piron ´ . which flow northeast and west northwest. Over time the Duero basin has acted as a sedimentary repository for detrital material and a number of units have been distinguished. The basal units are conglomerates, sands, silts and marls with gypsum and calco-dolomite formed as the middlerdistal deposits of Paleogene and Neogene alluvial fans. Above this are Pliocene arkosic facies of coarse to medium sand ŽArevalo facies. which are comprised of a complex series of ´ Mid-Pleistocene ŽInstituto Tecnologico Geominero de Espana, ´ ˜ 1989. channel sands, generally fining upwards, floodplain muds and intercalations of aeolian sand ŽPerez´ Gonzalez ´ et al., 1994.. Next in the stratigraphy are the southern Paramos sandy facies which are composed of sands with a high proportion of clay; and finally the uppermost fine to medium arkosic sand unit which is the focus of the present study. This sand is of great economic importance for the area as it is exploited as a silica or feldspar rich sand at a number of quarries. Also the soils formed on the sand are ideal for pine plantations which are grown not only for timber but also for resin production.

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

279

Fig. 1. Location map of the study area showing the distribution of the upper sand unit of the Tierra de Pinares and the Burgomillodo site.

The present-day climate is of the continental Mediterranean type, with a mean annual temperature of 11.88C and a mean annual rainfall of 463 mm at Segovia ŽInstituto Nacional de Meteorologıa, ´ 1995.. Winds are dominantly from the west.

3. The Tierra de Pinares sand unit The superficial sand unit of the Tierra de Pinares is formed of fine to medium arkosic sands. It is preserved as spatially irregular patches ŽFig. 1. with subhorizontal beds and is clearly discordant to the underlying Palaeozoic, Mesozoic and Paleogene materials described above.

280

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

Typically the particle size distribution shows between 45–70% medium sand, 11–42% mostly fine sand Ž6–26%. and small amounts of siltrclays Ž2–7%. ŽAleixandre et al., 1971.. The sand is well sorted and homogeneous, with subrounded to rounded grains which exhibit a matt surface appearance. The grain rounding, size sorting and grain surface abrasion are all indicative of sand that has been transported by aeolian processes. The mineralogy of the sand is dominated by quartz Ž; 60%., potassium feldspars Ž25%. and sodium feldspars Ž4%.. Heavy minerals which are present are primarily tourmaline, garnet and andalusite but their relative proportions vary from sector to sector ŽAleixandre et al., 1971; Alcala´ del Olmo, 1972; Casas et al., 1972; Olive´ et al., 1982.. The sand has an average thickness of ca. 3–4 m but it is variable and at Navalilla it is 1 m or less whilst elsewhere in the Segovia province thicknesses of up to 20 m have been reported and at Burgomillodo between 50–60 m of sand has accumulated ŽCortazar, ´ 1891; Bravard, 1965; Calonge, 1987.. The latter’s above-average thickness is due to the negative palaeo-relief of the underlying Cretaceous dolomites and limestones in this part of the Duero basin formed by a monocline fold within the syncline. The provenance of the sand has been widely reported upon. The consensus of opinion is that it is primarily derived from aeolian reworking of the Cretaceous and Tertiary detrital sands described above ŽCortazar, 1891; Hernandez Pacheco, 1923; Casas et al., ´ ´ 1972.. Locally it includes additional inputs of igneous and metamorphic materials from the south Guadarrama mountains and the St. Marıa ´ de Nieva Massif ŽAleixandre et al., 1971; Alcala´ del Olmo, 1972. and also material from river terraces ŽOlive´ et al., 1982; Perez-Gonzalez, 1982.. ´ ´ The age of the sand unit is contentious probably due to the incorporation of older material Žsee above. but it is definitely of the Quaternary era. Hernandez Pacheco Ž1923. ´ proposed a model of multiple phases of emplacement throughout the Quaternary to the present day. This view was supported by Perez-Gonzalez ´ ´ Ž1982. who thought intermittent emplacement took place up to the mid-Pleistocene although he did recognise that dunes on the Portillo plateau, the middle and lowest terraces of the Duero at Castronuno ˜ and around Coca were the result of Holocene aeolian activity. Bravard Ž1965. thought that the early and mid Quaternary was a period of erosion in this region thus making the aeolian movement more recent. Aleixandre et al. Ž1971. placed it in the mid Quaternary whilst Garcıa ´ Abbad and Rey Ž1973. correlated this unit with the Duero terraces TD11 Ž1987, 1988. stated that and TD13 making it of mid to late Pleistocene age. Fernandez ´ the superficial deposits around Coca were of Late Pleistocene or Holocene age. This interpretation was supported by Perez-Gonzalez ´ ´ et al. Ž1994. on the basis that the sand unit overlies the Mid-Upper Pleistocene Arevalo facies in which fossils of Equus ´ mostbachensis have been found. 3.1. Burgomillodo The area around Burgomillodo ŽFig. 1. was chosen for this initial investigation for several reasons. Firstly, as the aeolian sand unit is thickest here it potentially maximises the amount of palaeo-environmental information, either in terms of resolution or duration, that can be gained. Secondly, since 1961 Industrias del Cuarzo ŽINCUSA. has

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291 281

Fig. 2. Photograph and interpretative line drawing of the 35-m vertical section revealed at the INCUSA mine, Burgomillodo.

282

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

been extracting sand from Burgomillodo ŽCarrascal del Rio. for both high-purity silica sand and feldspar-rich sand. Quarrying has revealed a 35-m vertical exposure of bedded fine sand providing a good opportunity for detailed stratigraphical analysis and sampling ŽFig. 2.. At the study site the boundary at the base of the sand is planar showing evidence of a stratigraphical hiatus with the underlying deposits. Within the sand laterally tapering subhorizontal bedding structures extend for up to tens of metres. These are interpreted as erosion scars and reactivation pseudo-beds noticeable for their relative higher proportion of fines. From the exposure no appreciable lateral changes in the facies were evident. Detailed examination of the sand revealed a number of different sedimentary structures which have been divided into nine units on the basis of structure and grain size distribution ŽFig. 3.. Unit 1 Ža,b,c,d. is characterised by an alternation of sand with and without a silty matrix. It contains subhorizontal laminae, current ripples, relatively few scour and fill structures and some convoluted laminae. This sand unit is interpreted as belonging to the floodplain sandflats associated with a braided river system. Unit 2 is a finer sand with a silt matrix and subhorizontal laminae. The lack of structure indicative of fluvial movement indicates a trend toward subaerial exposure of the sand on the floodplain and aeolian redistribution. Unit 3 is a fine sand with high angle cross-bedding and internal reactivation structures. It represents trains of dunes Žlocally lobate or lunate. moving over a thin sand sheet. Similar dunes have been reported found on the present-day surface of the sand sheet and are dominantly parabolic ŽTemino ˜ et al., 1997.. Units 4 and 5 are similar in composition to unit 3 but with a trend away from high-angle cross-bedding to low-angle cross-bedding representing a change in depositional context from large dunes to sand sheet deposition. Unit 6 Ža,b,c. is similar to unit 1 with alternation of sand with silt matrix and sand without a silt matrix although convoluted laminae are absent. Unit 7 is a fine structureless homogeneous sand interpreted as a sand sheet deposit. Unit 8 is a palaeosol comprising of a sand with an abundant dark clay matrix and some charcoal. Unit 9 is a thin sand unit into which the present-day soil has developed. Overall the stratigraphy is dominated by the cross-bedding Žmedium angle. and subhorizontal laminations. The cross-bedding is formed into co-sets with a tendency of fining upwards. These are found amongst the erosion and reactivation surfaces with their angles varying from 58 up to 308. The laminations are clearly defined Žup to 1 cm thick. with a sequence of sand laminae and fines-rich matrix laminae. Palaeo-currents measured in the high-angle laminations as part of this study were found to be 1458 and 1508. Ž1988. who reported a direction of These agree with the findings of Fernandez ´ southwest to northeast from the dune morphology and a direction of northwest to southeast for the high-angle laminations over 2 m in size. Neither palaeo-current directions are concordant with the present day prevailing westerly winds or with the direction of flow of the nearby river Duraton ´ which flows northwest.

Fig. 3. Site stratigraphy, location of samples for luminescence dating and palaeoenvironmental interpretation based on the sedimentary structures.

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

283

284

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

The sedimentological interpretation of structure and sequences seen at Burgomillodo reveals two palaeo-environments: one a large sand-sheet unit with a braided fluvial system ŽSaskatchewan type—see Cant and Walker, 1978; Miall, 1985., the other dunes reworking and moving over thin sand-sheet deposits. Aeolian activity is present in both palaeo-environments. 4. Luminescence dating Luminescence dating of sediments is now a well established technique Žsee Wintle, 1993.. It relies on the fact that ionising radiation, from naturally occurring radionuclides Žuranium, thorium and potassium., causes displacement of electrons into trap defects within a quartz or feldspar crystal lattice. The number of trapped electrons increases with time until stimulated with heat in the laboratory or exposed to sunlight. When this happens luminescence is produced which can be measured and used to calculated the equivalent dose absorbed since last exposed to light Ž De .. Sample age is determined by dividing the De by the measured annual dose rate. 4.1. Field procedures All samples collected for luminescence dating must not be exposed to light. To avoid this a vertical section was cleaned revealing sand which had not been exposed to sunlight for any length of time. Into this vertical section, opaque plastic tubes 5 cm in diameter, fitted with a light-tight black plastic cap on one end were pushed until the section was totally filled. On extraction the open end was capped and the sample immediately wrapped in opaque black plastic. Bulk samples were also collected and placed in airtight plastic bags to be used for moisture and chemical analysis. The luminescence samples from the 35-m exposure at Burgomillodo mine ŽFigs. 2 and 3. were collected in order to ascertain whether sand deposition had been in a number of distinct phases, was the result of a slow accretion over a long period of time or a single rapid phase of emplacement. Access to the entire face was not possible so in order to meet the above objectives four samples were collected from the lower 11 m ŽShfd95093–95096. and a further three samples from the upper 15 m ŽShfd95097–95099. ŽFig. 3.. 4.2. Laboratory procedures All samples were prepared in the Sheffield Centre for Drylands Research ŽSCIDR. luminescence laboratory under controlled red-light conditions. Full details of the procedures employed can be found in Bateman and Catt Ž1996.. In brief the samples were treated with hydrochloric acid and hydrogen peroxide to remove any carbonate and organic material from the sand. They were then wet sieved and a size fraction of 180 to 250 mm selected to reduce grain size variability whilst reflecting the mean grain size of the samples. Heavy minerals were removed from the fraction using sodium polytungstate ŽSG 2.7.. Finally the samples were etched in 40% hydrofluoric acid for 60 min and sieved through a 150-mm sieve to leave a pure, clean quartz. Feldspars, whilst prevalent in these arkosic sands, were avoided due to reported problems associated with

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

285

anomalous fading ŽWintle, 1973, 1977.. For luminescence measurement, a monolayer of prepared sample was mounted on 1-cm-diameter aluminium discs using silicone spray. Optically stimulated luminescence ŽOSL. is a relatively new advancement in luminescence dating allowing stimulation Žwith light. and measurement of predominantly light-sensitive traps Žsee Wintle, 1993.. It has the advantage of rapid resetting of the luminescence signal when sediment is exposed to light and also completely resets to zero unlike thermoluminescence ŽTL. where a residual signal level has to be calculated. However, at the time of luminescence measurement Ž1995., practical reasons prevented the adoption of OSL for most of the samples collected. Luminescence measurement was therefore by TL. TL has been proven to work well with well-bleached aeolian sand and has the advantage of providing crucial information on where the signal is stable and unstable and the degree of bleaching the sample has undergone prior to burial. De estimation used the total bleach additive dose method ŽAitken, 1985; Mejdahl, 1987. with all samples split into at least five groups each containing a minimum of three discs. One group was used for determining the residual unbleachable TL component, one for the level of naturally acquired TL and the remaining groups had a range of laboratory doses on top of their naturally acquired dose. The maximum dose in each case was selected to be around 10 times greater than the expected De so as to allow extrapolation of the growth curve back to the x-axis intercept to be carried out with confidence. Luminescence measurements were carried out in a Risø automated luminescence reader fitted with a Hoya U-340 filter in front of the photomultiplier tube. Laboratory irradiation was undertaken with a Strontium-90 beta source whilst bleaching was carried out using extended periods of natural sunlight Ž4 days.. All samples were heated to 4508C at 28C per minute after having been preheated for 5 min at 2208C. Interdisc normalisation was carried out using data from a second equal dose to all discs. Rainer Grun’s ¨ software was used to fit simplex growth curves to the data ŽGrun ¨ and MacDonald, 1989.. As can be seen from the examples given in Fig. 4 the TL glow curves responded well to additional laboratory dose with only a slight peak intensity shift with dose. Natural TL peak intensities, filtered through a Hoya U-340, range from 5000 Ctsr8C to 10,000 Ctsr8C showing strong emission in this wavelength band as found by Townsend et al. Ž1993. for etched quartz. All peak TL intensities of the glow curves are centred around the light-sensitive 3258C trap region. As the samples might have been composed of poorly bleached sediment, the stability of De vs. temperature was examined ŽFig. 4.. In both the examples given in Fig. 4 the plateau regions extend, with only a slight increase in ED for sample Shfd95099, beyond 3308C indicating that there had been complete bleaching prior to deposition. All samples had linear or just sublinear growth curves showing no signs of reaching saturation. De’s were calculated from data in the 2908C to 3408C plateau region where interference from unstable signal and black body radiation is minimal. Interdisc scatter within dose points was low allowing good growth curve characterisation for all seven samples. This is reflected in the errors on the calculated De’s which range from 6% to 10% ŽTable 1.. As an additional check on whether full bleaching had taken place prior to deposition sample Shfd95099 underwent the single aliquot additive dose approach ŽDuller, 1994. using optically stimulated luminescence ŽOSL.. Each aliquot was preheated to 2208C for 120 s before being stimulated with light filtered through a Schott GG-420 for 0.1 s

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

286

Fig. 4. TL glow curves, De vs. temperature and additive dose growth curves for the basal Ža. and uppermost Žb. samples from Burgomillodo. TL glow curves are the mean of six aliquots normalised using a second equal dose.

whilst held at 508C. OSL was measured through a Hoya U-340 filter placed in front of the photomultiplier tube. Loss of OSL signal from repeated preheating and stimulation was corrected for by the preheat correction procedure using an additional four discs of Table 1 Dosimetry and TL age determinations for Burgomillodo mine Sample

Depth Žm.

K Ž%.

Th Žppm.

U Žppm.

Dcosmic ŽmGyra.

Dose rate ŽmGyra.

De ŽGy.

Age Žka.

Shfd95093 Shfd 95094 Shfd 95095 Shfd 95096 Shfd 95097 Shfd 95098 Shfd 95099

31.60 28.80 27.65 24.05 14.30 7.30 2.50

3.25 3.36 3.24 3.49 3.49 2.61 3.19

4.81 4.75 4.03 2.19 2.58 3.93 1.14

0.93 0.65 0.73 0.45 0.59 0.47 0.23

23"1 25"1 26"1 31"2 50"3 99"5 176"9

3676"208 3713"216 3566"205 3609"213 3689"214 2973"167 3341"200

49.1"4.9 45.7"4 43.1"2.5 41.6"2.5 42.2"3.1 32.6"2.8 21.6"1.8

12.5"1.0 12.3"1.3 12.1"1.0 11.5"1.0 11.4"1.1 11.0"1.1 6.5"0.7

ICP-AES and ICP-MS results for potassium, uranium and thorium, calculated cosmic contributions ŽPrescott and Hutton, 1994. and dose rates calculated from these results, with 5% errors to account for natural sample variability, using conversion data given by Aitken Ž1985. and attenuated for density and moisture. De determinations were measured using the total bleach additive dose method.

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

287

Table 2 De determination for sample Shfd95099 derived using optically stimulated luminescence and the single aliquot method ŽDuller, 1994. Aliquot

Natural OSL ŽCtsr0.5 s.

De ŽGy.

"

1 2 3 4 5 6 7 Mean

7910 5006 17587 10755 9929 6396 11938

21.6 18.8 22.7 21.9 17.2 22.1 18.0 20.3

3.3 3.8 1.8 1.3 1.9 1.7 3.7 2.3

sample ŽDuller, 1994.. Results of the seven unnormalised replicate De’s can be seen in Table 2. The mean single aliquot De of 20.3 " 2.3 compares very favourably with the TL derived De of 21.6 " 1.8. As OSL is bleached from quartz far more rapidly than TL ŽGodfrey-Smith et al., 1988. the concordance of both TL and OSL results coupled with the low interdisc variability of the single aliquot replicates all indicate that the sample was well-bleached prior to deposition. Samples were also analysed by inductively coupled plasma mass spectroscopy ŽICP-MS. to obtain concentrations of uranium ŽU., thorium ŽTh. and inductively coupled plasma atomic emission spectroscopy ŽICP-AES. for concentrations of potassium ŽK.. The potassium values, the largest contributor to the annual dose rate, were cross-checked by atomic absorption spectroscopy. The cosmic contribution was calculated using the equation given by Prescott and Hutton Ž1994.. From these data, annual dose rates were calculated using published conversion data ŽAitken, 1985, p. 67. ŽTable 1, columns 3–5. carefully attenuating for a palaeomoisture content of 3 " 3% and for the grain sizes used ŽTable 1.. 4.3. Results Table 1 shows the elemental concentrations, the dose rates calculated from them and the measured equivalent dose Ž De .. The final age determinations are shown in calendar years before present Ž1996. and are expressed with one standard deviation errors. Age consistently increases with depth agreeing with the law of super-position. Also, despite disparate De’s and annual dose rates, the ages calculated for six of the seven samples are extremely close to one another. These TL ages quite unequivocally indicate that the majority of sand deposited at this site did so in a single phase over a maximum of 3500 years but probably less than 2000 years between ca. 12,500 and 11,000 years. This main sand emplacement phase coincides with the Younger Dryas event which Fairbanks Ž1990. dated to between 13,000 and 11,700 calendar years from the present. The last sample ŽShfd95099. corresponds to the transition between the Boreal and Atlantic periods in the Holocene.

288

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

5. Results and discussion The TL dates show that at least 30 m of sand were deposited during the Younger Dryas at Burgomillodo. As the basal contact was not dated, remnants of previous sand-deposition phases, relating to sedimentation during previous deglaciations, may be represented in the 4–5 m of sand below sample Shfd95093 and above the solid geology. However, for such a large accumulation of sand to have been deposited in a such a short period of time, climatic conditions for aeolian deposition must have been optimal. Recent work on pollen, ostracods and molluscan assemblages, mainly from lake sites in Northern and Central Spain, have shown that the Younger Dryas climatic signal is clearly identifiable Že.g., Preece, 1991; De Beaulieu et al., 1994; Allen et al., 1996; Gonzalez ´ Dıez ´ et al., 1996; Wansard, 1996; Penalba et al., 1997; Sobrino et al., 1997.. The Younger Dryas saw much colder Žup to y88C cooler in summer—Wansard, 1996. and drier climatic conditions in Spain than the present ŽPerez-Obiol and Julia, ´ ` 1994; Penalba et al., 1997.. Sparse open vegetation coupled with the drier conditions would have enhanced environmental instability and led to increased erosion of sand from the older Tertiary and Quaternary sediments in the Duero basin. A similarly timed aeolian period was identified on the Manchega plain Žcentral southern Spain. where clay and sand dunes have been luminescence dated ŽRendell et al., 1996.. These new TL dates also fall within a period of deflation identified by Temino ˜ et al. Ž1997.. Interestingly the Burgomillodo dates also coincide with a period of widespread aeolian activity which formed coversand across much of northwestern and central Europe Že.g., Bateman, 1995; Zeeberg, 1995.. Difficulty arises when extrapolating these preliminary dates from a single site to suggest the chronological assignment of such a widespread unit as the Tierra de Pinares sand. However, the TL dates do disprove the suggestion that sand deposition occurred solely during the mid-Pleistocene ŽAleixandre et al., 1971.. Additionally multiple-phase sand emplacement throughout the entire Quaternary to the present day ŽHernandez ´ Pacheco, 1923. seems unlikely if at least 30 m of sand were deposited in less than 3 ka. The dates suggest a minimum average sand accretion in the order of 9–15 mmryear although, if sand accumulation was strictly confined to the Younger Dryas event, this could have been far greater. The sample nearest the surface ŽShfd95099. with an age of 6.5 ka is mid Holocene ŽAtlantic period.. Given the apparent full bleaching of the samples prior to burial ŽFig. 4. this date indicates a much later secondary phase of deposition in which a further few metres of sand were deposited at this site. Evidence of Holocene aeolian activity Ž1987, elsewhere in the area has been reported by Perez-Gonzalez ´ ´ Ž1982. and Fernandez ´ 1988. although there is no firm data to suggest synchroneity between sites. Rendell et al. Ž1996. on the Manchega plain have dated an aeolian activity episode to a similar time and Temino ˜ et al. Ž1997. identified this period as one when dunes were active in the area around Segovia. It is perhaps significant that the age of Shfd95099 coincides with what is thought to be the period of maximum post-glacial temperature and minimum rainfall based on palaeo-climatic reconstructions elsewhere in Spain ŽGonzalez ´ Dıez ´ et al., 1996.. It also coincides with the start of the Neolithic period, a time of land clearance and development of sedentary agriculture. At a recently discovered Bronze age burial

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

289

site at Sanchonuno, ˜ Segovia, aeolian sand has been excavated to form a subterranean burial chamber sealed by only 30 cm of sand ŽD. Vega, personal communication.. Other Bronze age and Roman remains throughout the Tierra de Pinares Že.g., at Navalmanzano and Cuellar ´ . also have not been covered by aeolian sand ŽD. Vega, personal communication., thus the sand must have remained stable at least since the Bronze age period. 6. Conclusions At least 30 m of aeolian derived sand are deposited at Burgomillodo. From bedding structures much of this seems to have come from the northwest. At least one phase of sand deposition at this site has been identified between ca. 12.5–11 ka probably coinciding with the dry and cold climatic conditions identified from biological evidence elsewhere in Spain. This would have provided conditions ideal for aeolian remobilisation of sand by inhibiting vegetation and keeping water-tables low. Interruptions in aeolian deposition as indicated by the erosion scars and reactivation horizons must have been of short duration based on the thermoluminescence results. The Burgomillodo site, therefore, represents the first sedimentological evidence for Younger Dryas aridity in central Spain. It is tentatively proposed that the Tierra de Pinares sand unit is primarily associated with the Younger Dryas event although only further dates will categorically prove whether or not older aeolian phases are also represented within it. If this chronological assignment is correct then the Tierra de Pinares may represent an Iberian peninsular equivalent to the widely reported coversands of northwestern Europe. Acknowledgements Fieldwork was financially supported by a small grant from the British Geomorphological Research Group’s research fund. Thanks must be extended to Mr. Perez ´ Cincunegui at Industrias del Cuarzo for providing access to the Burgomillodo quarry. Collection of the samples for luminescence dating was accomplished with the help of Marion Holmes whilst sample preparation and cartography was assisted by William Crowe and Paul Coles respectively. The authors would like to thank Drs. A Murray, Dr. H. Yaalon and an anonymous person for critically reviewing and improving this manuscript. References Aitken, M.J. 1985. Thermoluminescence Dating. Academic Press, London. Alcala´ del Olmo, L., 1972. Estudio sedimentologico de los arenales de Cuellar ´ ´ ŽSegovia.. Estudios Geologicos ´ 28, 345–358. Aleixandre, T., Benayas, J., Guerra, A., 1971. Procesos de movilizacion ´ del hierro en algunos suelos de la region Anales de Edafologıa ´ central espanola. ˜ ´ y Agrobiologıa ´ 30, 1095–1111. Alonso, A., Garzon, ˜ G., 1994. Quaternary evolution of a meandering gravel bed river in Central Spain. Terra Nova 6 Ž5., 465–475. Allen, J.R.M., Huntley, B., Watts, W.A., 1996. The vegetation and climate of northwest Iberia over the last 14 000 yr. Journal of Quaternary Science 11, 125–147.

290

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

Arche, A., 1983. Coarse-grained meander lobe deposits in the Jarama River, Madrid ŽSpain.. In: Collinson, J.D., Lewing, U. ŽEds.., Modern and Ancient Fluvial Systems. Special Publıcation, International Associa´ tion of Sedimentologists, 6, pp. 313–321. Bateman, M.D., 1995. Thermoluminescence dating of the British coversand deposits. Quaternary Science Reviews 14, 791–798. Bateman, M.D., Catt, J.A., 1996. An absolute chronology for the raised beach and associated deposits at Sewerby, East Yorkshire, England. Journal of Quaternary Science 11, 389–396. Bravard, Y., 1965. Notas morfologicas sobre la Tierra de Pinares Segoviana. Estudios Geograficos 27, ´ ´ 107–124. Calonge, G., 1987. El complejo ecologico y la organizacion ´ ´ de la explotacion ´ forestal en la Tierra de Pinares segoviana. Excma. Diputacion ´ Provincial de Segovia, Segovia, 347 pp. Cant, D.J., Walker, R.G., 1978. Fluvial processes and facies sequences in the sandy braided South Saskatchewan River, Canada. Sedimentology 25, 625–648. Casas, J., Leguey, S., Rodrıguez, J., 1972. Mineralogıa ´ ´ y sedimentologıa ´ de los arenales que recubren el Terciario entre los rıos 28, 287–296. ´ Piron ´ y Voltoya ŽSegovia.. Estudios Geologicos ´ Cortazar, D., 1891. Descripcion y geologica de la provincia de Segovia. Boletın ´ ´ fısica ´ ´ ´ de la Comision ´ del Mapa Geologico de Espana ´ ˜ 17, 190–198. De Beaulieu, J.-L., Andrieu, V., Ponel, P., Reille, M., 1994. The Weichselian late-glacial in southwestern Europe ŽIberian Peninsula, Pyrenees, Massif Central, northern Apennines.. Journal of Quaternary Science 9, 101–107. Duller, G.A.T., 1994. Luminescence dating of sediments using single aliquots: new procedures. Quaternary Geochronology ŽQuaternary Science Reviews. 13, 149–156. Fairbanks, R.G., 1990. The age and origin of the ‘Younger Dryas climatic event’ in greenland cores. Palaeoceanography 5 Ž6., 937–948. Fernandez, P., 1987. Geomorfologıa ´ ´ del sector comprendido entre el Sistema Central y el Macizo de Santa Marıa ´ la Real de Nieva. Tesis doctoral, Universidad Complutense de Madrid, 336 pp. Fernandez, P., 1988. Evolucion ´ ´ Cuaternaria y sistemas de terrazas en la subfosa Terciaria de Valverde del Majano y el Macizo de Sta Ma Real de Nieva ŽSegovia.. Boletın ´ de la Real Sociedad Espanola ˜ de Historia Natural ŽGeol. 84 Ž1–2., 69–83. Garcıa del Terciario y Cuaternario de Valladolid. Boletın ´ Abbad, F., Rey, G., 1973. Cartografıa ´ geologica ´ ´ Geologico y Minero 84, 213–227. ´ Godfrey-Smith, D.I., Huntley, D.J., Chen, W.-H., 1988. Optical dating studies of quartz and feldspar sediment extracts. Quaternary Science Reviews 7, 373–380. Gonzalez ´ Dıez, ´ A., Salas, L., Dıaz ´ de Teran, ´ J.R., Cendrero, A., 1996. Late Quaternary climate changes and mass movement frequency and magnitude in the Cantabrian region, Spain. Geomorphology 15, 291–309. Grun, ¨ R., MacDonald, P.D.M., 1989. Non-linear fitting of TL ESR dose response curves. Applied radiation and isotopes. International Journal of Radiation Applications and Instruments Part A 40, 1077–1080. Hernandez Pacheco, F., 1923. Las arenas voladoras de la provincia de Segovia. Boletın ´ ´ de la Real Sociedad Espanola ˜ de Historia Natural 23, 211–217. Instituto Nacional de Meteorologıa, principales Ž1961–1990.. Observato´ 1995. Valores normales y estadıstico ´ rio meteorologico de Segovia. Ministerio de Obras Publicas, Transportes y Medio Ambiente, Madrid. ´ ´ Instituto Tecnologico Geominero de Espana, ´ ˜ 1989. Mapa del Cuaternario de Espana, ˜ escala 1:1,000,000. Mapa y Memoria, Madrid. Mejdahl, V., 1987. Thermoluminescence dating of sediments. Radiation Protection Dosimetry 17, 219–227. Miall, A.D., 1985. Architectural-element analysis. A new method of facies analysis applied to fluvial deposits. Earth Science Reviews 22, 261–308. Olive, Elorza, M., Molina, E., 1982. Memoria del Mapa Geologico Nacional a ´ A., Portero, J.M., Gutierrez ´ ´ escala 1:50,000, hoja 445 ŽArevalo y Minero de Espana, ´ .. Segunda Serie, Instituto Geologico ´ ˜ Madrid. Penalba, M.C., Arnold, M., Guiot, J., Duplessy, J.-C., de Beaulieu, J.-L., 1997. Termination of the last glaciation in the Iberian peninsula inferred from the pollen sequence of Quintanr de la Sierra. Quaternary Research 48, 205–214. Perez-Gonzalez, A., 1982. El Cuaternario de la region ´ ´ ´ central de la cuenca del Duero y sus principales rasgos geomorfologicos, ‘1a Reunion y Minero de ´ ´ sobre la Geologıa ´ de la cuenca del Duero’. Instituto Geologico ´ Espana, ˜ Salamanca 1979, 717–740.

M.D. Bateman, A.D. Herreror Catena 34 (1999) 277–291

291

Perez-Gonzalez, A., Martın-Serrano, A., Pol, C., 1994. Depresion M. ŽEd.., ´ ´ ´ ´ del Duero. In: Gutierrez, ´ Geomorfologıa ´ de Espana, ˜ Rueda, Madrid, pp. 351–388. Perez-Obiol, R., Julia, ´ ` R., 1994. Climatic change on the Iberian Peninsula recorded in a 30,000-yr pollen record from Lake Banyoles. Quaternary Research 41, 91–98. Preece, R.C., 1991. Radiocarbon dated molluscan successions from the Holocene of Central Spain. Journal of Biogeography 18, 409–426. Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long-term variations. Radiation Measurements 2r3, 497–500. Rendell, H.M., Calderon, A., Gallardo, J., Millan, ´ T., Perez-Gonzalez, ´ ´ ´ A., Townsend, P.D., 1994. Thermoluminescence and optically stimulated luminescence dating of Spanish dunes. Quaternary Science Reviews 13, 429–432. Rendell, H.M., Perez-Gonzalez, A., Calderon, ´ ´ ´ T., Beneitez, P., 1996. Late Quaternary aeolian activity in the Manchega Plain, Central Spain. In Abstracts of the 8th International Conference on Luminescence and Electron Spin Resonance Dating, Canberra, Australia. Sobrino, C.M., Ramil-Rego, P., Guitian, M.R., 1997. Upland vegetation in the north-west Iberian peninsula after the last glaciation: forest history and deforestation dynamics. Vegetation History and Archaeobotany 6, 215–233. Temino, ˜ J., Garcia-Hidalgo, J.F., Segura, Y.M., 1997. Caracterizacion y evolution geologica del sistema dunas-humedales de Cantalejo ŽSegovia.. Estudios Geologico 53, 135–143. ´ Townsend, P.D., Rendell, H.M., Luff, B.J., 1993. High sensitivity TL spectra of quartz and feldspar. Ancient TL 11, 36–39. Wansard, G., 1996. Quantification of Palaeotemperature changes during isotopic Stage 2 in the La Draga continental sequence ŽNE Spain. based on the MgrCa ratio of freshwater ostracods. Quaternary Science Reviews 15, 237–245. Wintle, A.G., 1973. Anomalous fading of thermoluminescence in mineral samples. Nature 245, 143–144. Wintle, A.G., 1977. Detailed study of a thermoluminescence mineral exhibiting anomalous fading. Journal of Luminescence 15, 385–393. Wintle, A.G., 1993. Luminescence dating of aeolian sands: an overview. In: Pye, K. ŽEd.., The Dynamics and Environmental Context of Aeolian Sedimentary Systems. Geological Society Special Publication No. 72, pp. 49–58. Zeeberg, J.J., 1995. The nature and distribution of Late Pleistocene inland dunes in the European lowlands and on the Russian Platform. Report for The Netherlands Centre for Geo-ecological Research ICG, pp. 1–28.