late quaternary vegetation and fire history in the northernmost nothofagus forest region: mallín...

JOURNAL OF QUATERNARY SCIENCE (2009) 24(3) 248–258Copyright � 2008 John Wiley & Sons, Ltd.Published online 27 November 2008 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/jqs.1233
Late Quaternary vegetation and fire history in thenorthernmost Nothofagus forest region: MallınVaca Lauquen, Neuquen Province, ArgentinaVERA MARKGRAF,1,2* CATHY WHITLOCK,3 R. SCOTT ANDERSON2 and ADRIANA GARCIA41 Institute of Arctic and Alpine Research, University of Colorado, Boulder, Colorado, USA2 Center for Environmental Sciences and Education, and Quaternary Sciences Program, Northern Arizona University, Flagstaff,Arizona, USA3 Department of Earth Sciences, Montana State University, Bozeman, Montana, USA4 School of Earth and Environmental Science, University of Wollongong, Wollongong, New South Wales, Australia

Markgraf, V., Whitlock, C., Anderson, R. S. and Garcıa, A. 2009. Late Quaternary vegetation and fire history in the northernmost Nothofagus forest region: Mallın VacaLauquen, Neuquen Province, Argentina. J. Quaternary Sci., Vol. 24 pp. 248–258. ISSN 0267-8179.

Received 28 December 2007; Revised 7 August 2008; Accepted 7 August 2008

ABSTRACT: The last 16 000 cal. a of vegetation, fire and limnological history are described from thesteppe-forest ecotone in the northernmost Nothofagus forest region east of the Andes (Mallın VacaLauquen, Neuquen Province, Argentina, latitude 368 51.3360 S, longitude 718 02.5380 W). Between16 000 and 14 800 cal. a BP, scrub steppewith substantial open ground expanded in formerly glaciatedvalleys, whereas Nothofagus–Prumnopitys andina woodland covered mountain slopes. The site was arelatively deep and unproductive small lake at this time. After 14 800 cal. a BP, both steppe andwoodland vegetation became denser, indicating increased moisture and temperatures, although not to
present levels. The lake was still relatively deep and dystrophic, but became more alkaline by10 000 cal. a BP. Between 8900 and 5500 cal. a BP, conditions were markedly drier than before;a Cyperaceae marsh developed and disturbance taxa increased. After 5500 cal. a BP, moistureincreased but varied greatly, as evidenced by fluctuating water levels and high fire activity from5500 to 4400 cal. a BP and from 2300 to 1000 cal. a BP. Human activity, in terms of forest clearanceand livestock grazing, is documented in the uppermost levels. The evidence of high environmentalvariability in the middle and late Holocene is consistent with the onset or strengthening of the El Nino–Southern Oscillation, but differences in the timing of fire activity among sites on the west and east sidesof the Andes suggest that fuel conditions were important in determining the local occurrence of fire.Copyright # 2008 John Wiley & Sons, Ltd.
KEYWORDS: Lateglacial; Holocene vegetation; fire and climate history; mid-latitude Argentine Andes.

Introduction

The Mallın Vaca Lauquen site is a �200m diameter seasonallyinundated wet meadow, on the east side of the Andes ofArgentina at the steppe forest ecotone (latitude 368 51.3360 S,longitude 718 02.5380 W, elevation 1567m). The site waspreviously the focus of a reconstruction of late Pleistocene/Holocene vegetation and climate change (Markgraf, 1987).With the more recent focus on the role of past fire activity inshaping structure and distribution of temperate forests inPatagonia and fire-producing climate conditions (Whitlocket al., 2007), re-evaluation of this site, including new pollen,charcoal and plant macrofossil analyses, seemed warranted.Research on documentary evidence, fire-scarred tree rings

and sedimentary charcoal records has shown the importance of

* Correspondence to: V. Markgraf, 763 N. Pine Cliff Dr., Flagstaff, AZ 86001,USA.E-mail: [email protected]

past fires in Patagonia (Veblen et al., 1999; Huber andMarkgraf, 2003; Kitzberger and Veblen, 2003; Veblen et al.,2003; Whitlock et al., 2006, 2007). These records documentlatitudinal differences in the timing and duration of past fireintervals, including those occurring in recent centuries(Bianchi, 2000; Huber and Markgraf, 2003; Huber et al.,2004; Whitlock et al., 2007). Fire activity has been attributed toclimate anomalies linked to the latitudinal position of theSouthern Westerlies and the strength and location of thesoutheastern Pacific subtropical high-pressure system (Whit-lock et al., 2007; Garreaud et al., 2008). In the mid-latitudinalregion of Patagonia (latitude 38–438 S), periods of weakenedand poleward-shifted Westerlies and a stronger than normalsubtropical high-pressure system produce drought conditionsand large areas burned as a result of higher than normal summertemperatures and lower than normal summer precipitation(Villalba, 1994; Kitzberger and Veblen, 2003; Whitlock et al.,2007). At interannual and interdecadal scales, seasonalmoisture deficits and high fire occurrence are also related toEl Nino–Southern Oscillation (ENSO) and decadal climate

LATE QUATERNARY VEGETATION AND FIRE HISTORY 249

variability. North of latitude 388 S, including the study site,winter and spring precipitation is somewhat enhanced during ElNino events (Compagnucci and Vargas, 1998; Montecinos andAceituno, 2003), although it does not outweigh the moisturedeficit resulting from higher than normal summer and autumntemperatures (Garreaud et al., 2008).Our research objective was to reconstruct the environmental

history on the east side of the Andes at latitude 378 S at thenorthern limit of mixed Nothofagus obliqua/N. pumilio forestand the southern limit of the Monte desert scrub vegetation tobetter understand the climate history of this transitional region.We present pollen, charcoal and limnological data fromMallınVaca Lauquen in northern Patagonia. Two time periods are ofspecial interest: the Lateglacial and early Holocene interval(16 000–8000 cal. a BP) when summer insolation was lowerthan at present and ENSO variability was less pronounced (e.g.Markgraf and Diaz, 2000; Moy et al., 2002), and the lateHolocene interval (last 3000 a) when ENSO variability wasstrengthened.

Site description

Mallın Vaca Lauquen is located near the confluence of thevalleys of Lagunas Epulauquen and Vaca Lauquen about 60 km

Figure 1 Location of the Mallın Vaca Lauquen site, showing major topograpsteppe areas are in light brown, Andean tundra in grey

Copyright � 2008 John Wiley & Sons, Ltd.

north-west of the town of Andacollo, northwestern NeuquenProvince (Fig. 1). The lakes are dammed by recessionalmoraines, the lowest of which lies at about 1500m elevation.The surrounding peaks are about 2500m elevation. Theabundance of shrubs and introduced herbs in the localvegetation attests to heavy livestock grazing. The present-dayvegetation consists of impoverished high-elevation bunchgrasssteppe (estepa graminosa; Leon et al., 1998) with Festucapallescens and shrubs in the families Asteraceae (Baccharisssp., Chiliotrichium rosmarinifolium, Chuquiraga ssp., Mutisiassp., Perezia ssp.) and Rhamnaceae (Colletia spinosissima), aswell as Berberis rosmarinifolia, Ephedra frustillata andEryngium paniculatum (Apiaceae). Native herbs includeOsmorhiza berteroi (Apiaceae), Calceolaria biflora (Scrophu-lariaceae), Quinchamalium chilense (Santalaceae) and Phace-lia sp. (Hydrophyllaceae), among others. Introduced herbsinclude Rumex acetosella and Plantago lanceolata. Themountain slopes are covered by the northernmost extent ofsouthern beech (Nothofagus) forest, which is found up totreeline at 1700m elevation. Nothofagus obliqua dominateslower elevation forests, while N. pumilio and N. antarcticagrow at higher elevations, forming a krummholz belt above thetreeline in the transition zone to the Andean tundra. N.antarctica also grows on poor soils at all elevations. FewAustrocedrus chilensis trees grow at lower elevations, and burntand dead trees of Nothofagus obliqua in the watershed offerabundant evidence of fires in recent times.

hic features and vegetation cover. Nothofagus forests are in green; the

J. Quaternary Sci., Vol. 24(3) 248–258 (2009)DOI: 10.1002/jqs

250 JOURNAL OF QUATERNARY SCIENCE

Given the absence of meteorological stations in thismountainous region, mean annual precipitation and tempera-ture were estimated at 700mm and 48C, respectively, based ondata from the Climatologic Atlas of South America (J. A.Hoffman, ed.), UNESCO, 1975. Precipitation occurs primarilyduring autumm and winter months (April through July). Areview of the climate of Patagonia and its relation to majorvegetation zones lists mean annual precipitation at the steppe–forest ecotone of about 600mm and mean annual temperatureof 68C, based on 62 stations between latitudes 378 and 558 S(Paruelo et al., 1998).

Methods

In 2002, three sediment cores, MVL-02A, MVL-02B, and MVL-02C, were taken with a modified Livingstone corer from thecentre of the wetland. The most complete core, MLV-02C, was342 cm long. In the laboratory, cores were split and thesediments described. The core MVL-02A was analysed forthe upper 50 cm (this segment was not compacted compared tothe same segment from coreMVL-02C), and coreMVL-02Cwasanalysed between 50 and 342 cm depth. Magnetic suscepti-bility (MS; Sandgren and Snowball, 2001), was measured usinga Bartington MS meter with a hand-held MS2E sensor.Measurements were taken every 0.5 cm along the entire lengthof core.For pollen analysis, 1 cm3 volumetric samples were taken at

2.5 or 5 cm intervals and prepared with standard techniques(Faegri and Iversen, 1989). Pollen grains were identified at400� and 1000�magnification, with counts ranging from 150to 400 grains. Terrestrial pollen percentages were based on thetotal sum of pollen grains from trees, shrubs and herbs and usedto calculate the percentages of terrestrial and aquaticpalynomorphs. Excluded from the terrestrial pollen sum wereCyperaceae, fern allies (mostly undifferentiated Polypodiaceaeand a few Hymenophyllum and Pteris), aquatic taxa (Myr-iophyllum) and the algae Pediastrum boryanum var. longicorneand Botryococcus. Escallonia and Caryophyllaceae were alsoexcluded from the pollen sum, given their erratic and probablylocal occurrences in high numbers in some intervals. Escalloniais a shrub that often grows at the margin of wetlands andCaryophyllaceae (locally present taxa are within the generaStellaria and Arenaria pollen type) grow primarily on disturbedsandy–gravelly soils. Lycopodium tracer spores were added toeach sample for calculation of pollen concentration(grains cm�3). Changes in pollen percentage and concentrationwere used to interpret past vegetation changes supported alsoby CONISS cluster analysis (Grimm, 1987). Pollen data wereplotted using TILIA programs (Grimm, 1992).A total of 72 pollen, spore and algal types were identified.

Local tree taxa included Nothofagus dombeyi type (this pollentype includes N. dombeyi, N. betuloides, N. pumilio and N.antarctica; however, N. dombeyi and N. betuloides do notoccur north of latitude 378 450 S). In this record, this pollen typeprobably reflects N. pumilio and N. antarctica. Other local treepollen taxa are Nothofagus obliqua type (N. obliqua, N.procera) and Prumnopitys andina; rainforest taxa from thelowlands west of the Andean crest included Hydrangea,Myrtaceae and Weinmannia; and local steppe shrub taxaconsisted of Lomatia/Gevuina type, Rhamnaceae, Berberis,Ribes and Verbena. Common steppe herbs were Acaena,Chenopodiaceae, Wendtia, Plantago and Phacelia. Wetlandherbs (Ranunculaceae, Valeriana, Lamiaceae, Liliaceae, Gen-tiana) and introduced European taxa (Rumex, Plantagolanceolata) were also present. Macrofossil remains were


identified (Table 1) and total number of oospores ofcharophytes (Charales) plotted in the pollen diagram.

Microscopic charcoal particles (larger than 20mm) werecounted on the pollen slides and converted to charcoalconcentration (particles cm�3). These data were convertedcharcoal accumulation rates (CHAR; particles cm�2 a�1) bydividing concentration by the deposition time of each sample(cm a�1). In addition, samples of 1 cm�3 volume were taken at0.5 cm intervals for macroscopic charcoal and plant macro-fossil analyses. These samples were washed through nestedmetal screens of mesh sizes of 250 and 125mm. The residueswere identified and counted under a stereo microscope.Because both macroscopic charcoal size fractions showedsimilar trends through time, charcoal counts were combinedand divided by sample volume to calculate charcoalconcentration (particles cm�3). Charcoal concentrations wereresampled into contiguous 25 a bins (the median resolution ofthe record) in order to sample over equally spaced timeintervals through the record. CHARwas determined by dividingconcentrations (particles cm�3) by deposition time (a cm�1)using Char Analysis software (Higuera et al., 2008). The slowlyvarying trend often referred to as background charcoal(BCHAR) was determined with a 700 a lowess smoother,robust to outliers. Throughout the paper, we distinguishbetween microscopic CHAR from discontinuous pollen slidesand macroscopic CHAR from the high-resolution analysis ofcontiguous samples. Charcoal and terrestrial plant remainsused for AMS radiocarbon dating were washed thoroughly withdistilled water.

Results

Sediment description and chronology

From the base at 342–220 cm depth, the sediment wascomposed of laminated inorganic silty clay that above250 cm depth became more organic. Between 220 cm depthand the surface, the sediment consisted of compact peats withdarker and lighter horizons. The magnetic susceptibility record(Fig. 3) indicates high magnetic susceptibility associated withinorganic sediment below 250 cm depth and low magneticsusceptibility above that depth in organic sediments. Peaks inmagnetic susceptibility corresponded to volcanic ash layers,including three thick layers at 106.5–112 cm, 265.5–272.5 cmand 315.5–320.5 cm depth.

Core depth was initially adjusted for minor sedimentcompaction (of 2–4 cm/50 cm core segment) by multiplyingthe compacted length by a correction factor (core length/compacted length) to reconstruct the actual length of each coresegment. For the purpose of developing an age model, coredepth was further adjusted by excluding the three thick andpumiceous volcanic ash layers (106.5–112 cm, 265.5–272.5 cm and 315.5–320.5 cm original depth, Fig. 3), assumedto represent rapid deposition. Hereafter, depth refers toadjusted depth.

Eight AMS radiocarbon dates were obtained on sievedcharcoal and organic material (Table 2 and Fig. 4). Two of theradiocarbon dates (AA-57612: 7790� 65 14CaBP, 309–311cmoriginal depth; and AA-65184: 3751� 39 14C aBP, 145–145.5 cm original depth) were out of sequence and not usedin the age model. Two other dates overlapped (AA-57613:5162 14C aBP at 80.5 cm and AA-65183: 4505 14C aBP at99 cm original depth), and probability calculations using BCal(Buck et al., 1999) indicated that the older date was an outlier.


Table 1 List and counts of terrestrial and aquatic plant macrophytes and Cladocera remains identified (numbers cm�3 volume)

Depth (cm) Nitella opaca Nitella hyalina Characf. braunii Hypericum sp. Bryophyte spores Utricularia sp. Cladocera Zones

2 6 0 0 1 0 0 0 54 10 5 0 6 5 0 06 8 0 0 4 6 0 7

20 12 18 0 0 15 0 0 4c20.5 22 13 0 0 0 0 228.5 40 34 0 0 15 0 629.5 18 13 0 0 18 0 0

34.5 10 9 0 0 15 0 047 18 0 0 0 15 0 058.5 10 8 0 0 10 0 0 4b59.5 10 4 0 0 10 0 060 23 44 0 0 12 0 0

76.5 12 17 0 0 0 0 0 4a80.5 21 36 1 0 0 0 181.5 2 31 0 0 0 0 291.5 22 46 0 0 0 0 895 3 34 0 1 15 0 596 15 59 0 1 0 0 096.5 22 31 0 0 0 0 1

115 0 13 0 9 0 0 0 3

214 28 0 0 3 0 0 0 2216 42 0 0 0 0 0 0235 31 0 0 0 0 0 0239 43 0 0 0 0 0 0239.5 84 0 0 0 0 0 0

321 16 0 0 18 0 3 0 1


Hence, only five dates were used to develop the agemodel afterconversion to calendar years before present (cal. a BP) usingCALIB 5.0 (Stuiver et al., 2005). The dates were calibrated usingSouthern Hemisphere calibration to the date 4504 cal. a BP andNorthern Hemisphere calibration for the dates of 8171, 10 250and 13 820 cal. a BP. The basal age of the core was extrapolatedto ca 16 000 cal. a BP.A range of possible calibrated dates and probability

distributions were determined for every radiocarbon date

Figure 2 Map showing sites of charcoal records (see Fig. 7)


using CALIB 5.0.1 (Stuiver et al., 2005). Monte Carlo samplingwas used to generate a cubic smoothing spline through all thedates 2000 times, and the final age–depth model was based onthe median of all the runs (Higuera, 2008).Based on this chronology, deposition times were about 30–

40 a cm�1 between ca. 17 000 and 7000 cal. a BP, 10–20 a cm�1 between 7000 and 4000 cal. a BP, 10 a cm�1

between 4000 and 2200 cal. a BP, 30–40 a cm�1 between 2200and 400 a cm�1 and 10 a cm�1 during the last 400 a.


Table 2 List of radiocarbon dates, corresponding calibrated ages and information on materials dated

Lab code Depth (cm) 14C age SD� d13C% cal age BP(median probability)

Upper Lower Material dated

13 400 400 European taxaAA65182 30.5–31 2740 130 �25 2 873 2 990 2 700 Sedge fragmentsAA57613 80.5–81.5 5 162a 77 �28.44 5 831 5 925 5 837 Sedge fragments/charcoalAA65183 99–100 4 505 60 �22 5 155 5 135 4 966 Sedge fragments/charcoalAA65184 145–145.5 3 751a 39 �27.5 4 113 4 130 3 924 Sedge fragmentsAA57611 228.5–229 8 171 72 �28.18 9 133 9 142 8 984 Organic fragmentsAA65185 254.5–256 10 250 58 �21.8 12 002 12 192 11 754 Bulk sedimentAA57612 309–311 7 790a 65 �28.53 8 568 8 636 8 397 Organic fragmentsAA65186 331–332 13 820 76 �14.4 16 465 16 669 16 247 Bulk sediment

aDates not used in chronology.


Pollen and charcoal record

Throughout the record, the pollen assemblages are co-dominated by steppe and forest taxa, suggesting that, like atpresent, steppe grew in the valleys and Nothofagus forest grewon the mountain slopes. The pollen stratigraphy was dividedinto five zones using a constrained cluster analysis (Grimm,1987) of terrestrial and aquatic taxa (Fig. 5). Microscopic CHARprovided a reconstruction of regional fire conditions, andmacroscopic CHAR was used to infer local fire conditions(Fig. 6). The variable sedimentation rates suggest a changingdepositional environment and, for that reason, charcoal peaksor particular fire episodes were not identified. Instead, totalmacroscopic CHAR was used to infer general fire activityduring the time represented by different pollen zones.

Figure 3 Magnetic susceptibility of the Mallın Vaca Lauquen sedi-ment core (original depth)


Zone 1 (320–288 cm depth; 17 000–14 900 cal. a BP)featured 60–80% non-arboreal taxa, composed of Poaceae(40%); steppe shrubs, including Asteraceae (10–15%), Ephedra(2%), Rhamnaceae (5%) and Ericaceae (5–10%); and steppeherbs (15–20%). Arboreal taxa Nothofagus dombeyi type andPrumnopitys andina accounted for 10–20% each. N. obliquatype and rainforest taxa were present in trace amounts.Escallonia accounted for 5%. Pollen concentration was low.with a mean of 400 grains cm�3. Pediastrum boryanum var.longicorne was abundant but fluctuated markedly (200–1000%). At levels where Pediastrum boryanum var. longicornewas less abundant, traces of Myriophyllum, rare Utriculariaseeds and oospores of charophytes were found. The charophytewas identified as Nitella opaca (Table 1). Microscopic andmacroscopic CHAR values were negligible, suggesting that fireswere virtually absent.

The primary pollen constituents of this zone suggestherbaceous grassland with open ground. A considerablenumber of taxa were present that grow on sandy and rockysubstrates, such as alluvial fans and cobble shorelines (Ephedra,Ericaceae (Pernettya), Plantago, Colobanthus (Caryophylla-ceae)). These taxa are often found in Lateglacial basal samples,reflecting colonisation of disturbed terrain (e.g. M. Pollux,Markgraf et al., 2007) and thus the assemblage has no directanalogue with modern steppe assemblages (Paez et al., 2001;Markgraf et al., 2002). Despite the relatively high inorganiccontent the deposition time was slow (30 a cm�1).

Zone 2 (288–212.9 cm depth; 14 900–8600 cal. a BP) hadslightly higher values of Poaceae than before (up to 50%).Nothofagus dombeyi type increased to 30%, whereas Prum-nopitys andina decreased to 10%, and N. obliqua typecontinued to be present in traces, reaching 2–5% atsome levels. Rainforest taxa (Hydrangea, Myrtaceae, etc.)were more prominent during this interval. Pollen abundance ofsteppe shrub taxa continued as before with a mean of 5%.Pollen of open ground taxa (e.g. Plantago) decreased tonegligible amounts. Wetland herb types increased to 5%.Pollen concentration continued to be low with 300–400 grains cm�3. Pediastrum boryanum var. longicornefluctuated around a mean of 200% and Nitella opaca appearedin the upper part of this zone (ca. 10 800 cal. a BP) withnumbers between 20 and 50 cm�3. At those levels, Pediastrumboryanum var. longicorne was rare or absent. Microscopic andmacroscopic CHAR values were low in this zone but higherthan before, implying some fires in the region. The depositiontime was even slightly slower than in Zone 1 (40 a cm�1). Theassemblage indicates a more diverse herbaceous steppe thanbefore, with less open and disturbed ground, comparable to thepalynologically defined ’mid-grass steppe’ of Paez et al. (2001)


00.0250100150200250300350

0

2000

4000

6000

8000

10000

12000

14000

16000

depth below mud-water interface (cm)

Age

(ca

l yr

BP

)

cubic spline interpolant, 95% CI

0.0810

sed. rate resolution(cm yr-1) (yr sample

30 50

-1)

Figure 4 Age–depth model and sedimentation rates for the Mallın Vaca Lauquen record


or estepa graminosa (Leon et al., 1998). Reduced erosion,inferred from the decreased magnetic susceptibility of thesediments, could explain the markedly lower sedimentationrates than in the previous zone. Nothofagus forest along themountain slopes, probably composed of both N. pumilio andN. antarctica, became denser, limiting the habitat forPrumnopitys andina.Zone 3 (212.9–107.5 cm depth; 8600–5300 cal. a BP)

showed an increase of Nothofagus dombeyi type to 40%, onaccount of a decrease of Poaceae (40–50%). Prumnopitysandina decreased to trace amounts. Nothofagus obliqua typeappeared continuously, reaching values of 5% in the upper partof the zone; steppe shrub pollen types continued with lowvalues; and rainforest pollen percentages decreased. Wetland

Figure 5 Pollen record of Mallın Vaca Lauquen


herb taxa slightly increased to 10% and Escallonia showedseveral peaks of 15%. Pollen concentration continued lowwith400 grains cm�3, except for one high value at 110 cm depth.High percentages of Cyperaceae (500–600%) indicateinitiation of wetland conditions, although some open watermust have existed intermittently to support the presence ofMyriophyllum, oospores of Nitella opaca and Botryococcus insome levels. In the upper portion of this zone, Caryophyllaceaepollen (Stellaria/Arenaria pollen type) appeared with valuesbetween 100% and 500%, suggesting shoreline disturbance byseasonal water-level fluctuations. Bothmacroscopic CHAR andmicroscopic charcoal values increased in this zone. Macro-scopic CHAR reached 0.80 particles cm�2 a�1, suggesting localfires, andmicroscopicCHARwasmoderate (100particles cm�2 a�1),


Figure 6 Charcoal record of Mallın Vaca Lauquen


consistent with more regional fire activity. Deposition time wastwice as fast (10–20 a cm�1) as during zones 1 and 2, reflectingthe onset of peat growth.Zone 4 (107.5–13 cm depth; 5300–400 cal. a BP) is

characterised by the onset of continuing presence of Notho-fagus obliqua type together with high values of N. dombeyitype. Rainforest and steppe shrub pollen types were sporadi-cally present. Pollen of steppe herbs and wetland herbs wereless abundant overall compared with values in Zone 3. Thiszone can be divided into three subzones based on changes inthe proportions of arboreal and non-arboreal taxa. The initialsubzone (4a: 107.5–60.2 cm depth; 5300–4200 cal. a BP) had50% Poaceae, 30% Nothofagus dombeyi type, as well asslightly higher percentages of Acaena and lower Asteraceae.Pollen concentration continued low with 400 grains cm�3,although deposition time was high, as in Zone 3 (10–20 a cm�1). The middle subzone (4b: 60.2–30.8 cm depth;4200–2200 cal. a BP) showed slightly lower values for Poaceae(40% mean) and steppe taxa and slightly higher values forNothofagus dombeyi type (45%). Pollen concentrationdoubled with values of 800 grains cm�3, reaching a maximumof 1200 grains cm�3, while deposition time continues high (10a cm�1). The upper subzone (4c: 30.8–13 cm depth; 2200–400 cal. a BP) featured an increase in Poaceae percentages to50% and a decrease of Nothofagus dombeyi type to 40%.Pollen concentration was slightly lower with 700 grains cm�3

and deposition time was slower as well (30–40 a cm�1).Cyperaceae decreased to a mean of 200%, whereas aquatictaxa increased. Among the aquatic taxa, Charophyte oosporeswere persistently high, although highly variable, especially insubzones 4a and 4c and one level in 4b. Charophytes weredominated byNitella hyalina (Table 1), a shallow-water speciesthat currently grows in 0.15–0.8m water depth (Garcıa, 1987;Caceres and Garcıa, 1989); N. opaca, a species toleratingdeeper waters, was also present in low amounts. At one level in


subzone 4a, Chara braunii, a taxon commonly associated withNitella hyalina (Corillion, 1957), was found. Pediastrumboryanum var. longicorne appeared intermittently with lowvalues in subzones 4a and 4c; Myriophyllum was essentiallyabsent except for occurrences in subzone 4a. Botryococcus hadrare occurrences. Subzones 4a and 4c indicate shallow,fluctuating water levels; subzone 4b had high amounts ofbryophyte spores, indicating intermittently dry conditions.

During 4a, macroscopic and microscopic CHAR reachedtheir highest levels of the record (macroscopic CHAR was up to12.0 particles cm�2 a�1 and microscopic CHAR reached 1516particles cm�2 a�1). These high levels decreased dramaticallyin subzone 4b, and after 3500 cal. a BP very few fires wereregistered. In subzone 4c, CHAR levels were again elevated(macroscopic CHAR was up to 14.50 particles cm�2 a�1 andmicroscopic CHAR reached 6650 particles cm�2 a�1), indi-cating significant local and regional fire activity. The pollen andcharcoal data suggest that moisture variability must have beenhigher overall, especially in the lower and upper subzones (4aand 4c) than in the middle subzone to create conditions forNothofagus obliqua type to occur, and higher fire activity.Overall conditions, however, were wetter than in Zone 3,especially in subzones 4a and 4c.

Zone 5 (13–0 cm depth; 400–0 cal. a BP) showed a markeddecrease in Poaceae, coupled with an increase in Nothofagusdombeyi type, steppe shrub taxa and mesic herbs. IntroducedEuropean weed taxa appeared (Plantago lanceolata, Rumexacetosella), indicative of grazing and logging. During the upperpart of this zone, Nothofagus dombeyi type also decreased,documenting more intense forest clearance in more recenttimes. Pollen concentration continued high with 800grains cm�3 and deposition time increased (10 a cm�1),suggesting higher erosional input. The charophytes, Nitellahyalina and N. opaca, dominated during this interval, whilePediastrum boryanum var. longicorne and Myriophyllum werepresent intermittently, indicating seasonally fluctuating waterlevels, as observed during fieldwork in 1981. CHAR values inthe last 1000 a are relatively low (macroscopic CHARwas <0.90 particles cm�2 a�1 and microscopic CHAR was<530 particles cm�2 a�1). These data suggest a limited role forfire in the local watershed as well as regionally in recentcenturies.

Palaeoenvironmental and palaeoclimatehistory

The general vegetation history inferred from this core iscomparable to the early study (Markgraf, 1987). However, theadditional detailed record of limnological changes and firehistory provide a more detailed palaeoenvironmental andpalaeoclimatic picture. The different proportions of aquatictaxa in the two records are related to the fact that the presentrecord was taken in the deepest, wettest part of the basin, whichwas dry in 2002 but inundated in 1981. The earlier core, takenat a marginal location, showed high percentages of Botryo-coccus and low values of Pediastrum boryanum var. long-icorne, indicating shallow and fluctuating water levels. Thepresent record has high amounts of Pediastrum boryanum var.longicorne and only a few levels with Botryococcus, reflectingits deeper water location.

During the last 17 000 cal. a, open-water conditions musthave existed in the shallow basin, allowing for the continuouspresence of aquatic taxa. Between 16 000 and 14 800 cal. a BP,high amounts of aquatic taxa were present: Pediastrum



boryanum var. longicorne, an algae common in small, high-elevation lakes (Komarek and Fott, 1983; Jankovska andKomarek, 2000); the charophyte Nitella opaca, a cosmopolitanspecies that thrives in water depths of 0.5–2.5m, but also growsin clear waters up to 40m depth (Corillion, 1957); Utricularia, amontane to subalpine, shallow-water plant; and shallow-waterdiatoms growing in low-productivity environments (Fragilariaconstruens, var. venter and Cyclotella stelligera in the basalsamples of the original core; Markgraf, 1987). These taxasuggest clear, dystrophic and slightly acidic waters. Such low-productivity environments were apparently common in theLateglacial period during the early stages of lake development(Bradbury and Whiteside, 1980).The initial terrestrial vegetation at Mallın Vaca Lauquen

between 17 000 and 14 900 cal. a BP was a shrub–steppe withconsiderable open, sandy and gravely ground. This patchylandscape suggests cool and dry conditions. An interestingaspect of this interval is the occurrence of Prumnopitys andinapollen, which is also documented at the Tagua Tagua site westof the Andes at latitude 34.58 S (Heusser, 1990). Prumnopitysdoes not occur at either site today, but scattered individualsgrow on the Chilean side of the Andes from latitude 35.58 S to438 S at 600 and 1000m elevation, in association withAustrocedrus chilensis, Nothofagus obliqua, N. pumilio andLomatia hirsuta (Veblen et al., 1995). At latitude 388 S to thesouth of Mallın Vaca Lauquen, Prumnopitys andina crosses theAndes into Argentina as individual trees near upper treeline(Donoso, 1974). This limited geographic and elevationaldistribution is confined to areas of low temperatures and drysummers (Veblen et al., 1995). Apparently, Prumnopitys andinawas more widespread than today on both sides of the Andesduring the initial period of deglaciation (Markgraf et al., 1992,2002; Whitlock et al., 2001). Its presence at Mallın VacaLauquen is consistent with about 68C cooler Pacific sea surfacetemperatures (latitude 418 S, Lamy et al., 2004; latitude 348S,Kim et al., 2002), a strengthened southeastern Pacific high-pressure system, and weakened Southern Westerlies (Markgrafet al., 1992; Lamy et al., 1999).Between 14 900 and 8600 cal. a BP, the steppe environment

was more diverse than before and the high proportion of herbpollen is characteristic of modern pollen assemblages from thepresent-day grass steppe (Paez et al., 2001), where precipi-tation is about 500mm (Paruelo et al., 1998). Forests on theslopes around the site continued openwith Prumnopitys andinaandNothofagus pumilio/N. antarctica. Higher amounts of long-distance pollen from rainforest taxa are attributed to anexpansion of Valdivian rainforest in the Chilean lowlands(Moreno, 1997, 2004). Initially the lake was still dystrophic butincreased abundance of the charophyte Nitella opaca by10 000 cal. a BP suggests increased alkalinity. Conditions werewarmer and wetter than before, although the openness of theforest at Mallın Vaca Lauquen suggests that precipitation wasless than at present. Fires were not a significant component ofthe ecosystem, judging from low CHAR values.Drier-than-present conditions have also been suggested from

marine and terrestrial records at latitude 348 S (Lamy et al.,1999; Villa-Martınez et al., 2004). In contrast, records betweenlatitudes 40 and 458 S show rainforest expansion by 14 800 cal.a BP, suggesting that moisture had reached present levels in thatregion (e.g. Moreno, 1997, 2004; Villagran, 2001; Markgrafet al., 2002; Haberle and Bennett, 2004). Westerly winterstorms had apparently become established south of latitude408 S at this time (Markgraf et al., 2002; Whitlock et al., 2006),but the southeastern Pacific High continued to be stronger thantoday. This combination resulted in a less pronouncedprecipitation increase to the north of latitude 408 S than to


the south (Markgraf et al., 1992; Lamy et al., 1999; Romeroet al., 2006).After 12 000 cal. a BP, however, aridity returned to mid

latitudes. This early Holocene (12 000–8000 cal. a BP) aridityhas been widely documented in marine and terrestrial recordsbetween latitudes 54 and 338 S (Villagran, 1991, 2001;Markgraf et al., 1992, 2007; Lamy et al., 1999, 2001, 2004;Huber et al., 2004; Moreno, 2004). In the rainforests west of theAndes, drought- and disturbance-adapted rainforest elements(especially Nothofagus obliqua and Weinmannia trichos-perma) became dominant at that time (Villagran, 1991,2001; Moreno and Leon, 2003; Abarzua and Moreno, 2008)and high-latitude Nothofagus forests continued quite open(Huber et al., 2004). The duration of this interval is longer andmore pronounced at high latitudes than at mid latitudes andeast of the Andes, as illustrated by a comparison of charcoalrecords between latitudes 52 and 358 S (Fig. 7). Highest fireactivity at latitude 528 S occurred between 12 000 and5500 cal. a BP and at 458 S between 11 000 and 7000 cal. a BP(Rio Rubens: Huber et al., 2004; Mallin Pollux: Markgraf et al.,2007). At latitude 42–358 S, west of the Andes, highest fireactivity occurred between 11 000 and 9000 cal. a BP and after3000 cal. a BP (Lago Melli: Abarzua and Moreno, 2008; TaguaTagua: Heusser, 1990; Puren-Lumaco: Abarzua, pers. comm.).East of the Andes at lat 418 S (Laguna el Trebol: Whitlock et al.,2006), high activity is registered after 8000 cal. a BP with 1000a long maxima centred at 6500, 4500, 3500 and 1000 cal. a BP.Arid conditions during the early Holocene have been

attributed to substantially weaker Westerlies or latitudinallyshifted westerly storm tracks, related to a strengthenedsoutheastern Pacific High (Markgraf et al., 1992; Whitlocket al., 2007). However, a poleward shift in the Westerlies doesnot fully explain these east–west differences in fire activity. Ananomalous component of upper-level flow associated with thewest–east dipole may have resulted in opposite moistureregimes west and east of the Andes, but this hypothesis bearsfurther study.Mallın Vaca Lauquen did not show this early Holocene arid

and warm interval with high fire activity; instead conditionscontinued cool and dry until 8600 cal. a BP. Only between8600 and 5300 cal. a BP did conditions became markedlywarmer at Mallın Vaca Lauquen, based on the development ofCyperaceae wetland in the shallow basin, only brieflyinterrupted by periods with open water. Precipitation becamemore variable, as evidenced by fluctuations in local shorelinetaxa (Escallonia, Arenaria/Stellaria pollen type) and betweensteppe and forest taxa. The steppe assemblage compares withpresent-day shrub steppe (estepa arbustivo-graminosa; Leonet al., 1998, Paez et al., 2001), where precipitation is<400mm(Paruelo et al., 1998). Elevated levels of microscopic CHAR atMallın Vaca Lauquen, especially between 5000 and 4500 cal. aBP, may have come from regional fires in the Chilean lowlands.The Tagua Tagua site features high levels of charcoal at thistime (Heusser, 1990) (Fig. 7). Higher levels of charcoal and lowlake levels are also observed between ca 10 000 and 6000 cal. aBP in a record from the Puren-Lumaco valley (latitude 388 300 S,longitude 738 W), located on the southeastern slope of theCoastal Chilean Range (Nahuelbuta Range, Abarzua, pers.comm.). The macroscopic charcoal record suggests some localfires at Mallın Vaca Lauquen at about 7500 cal. a BP.By 5300 cal. a BP, the present-day mixed Nothofagus

obliqua/N. pumilio forest had become established in theregion, indicating a shift to present-day winter rain/summerdrought conditions. High abundance of charophytes, Nitellaopaca and N. hyalina at Mallın Vaca Lauquen suggests morepermanent water than before.Nitella hyalina is a shallow-waterspecies growing at 0.15–0.8m water depth (Garcıa, 1987;


Ag

e (c

al y

r BP)

Abarzua & Moreno 2008Heusser 1990

Whitlock et al. 2006 this study

Lago Melli

la

t 42.76S, lo

ng 73.55W, e

lev. 70 m

Laguna el Tre

bol

la

t 45.08S, lo

ng 71.84W, e

lev. 758 m

Mallin

Vaca Lauquen

lat 3

6.86S, long 71.05W

, elev.

1567 m

Particles cm-2 yr-1

Particles x 103 gm-1

Particles cm-2 yr-1 Particles cm-2 yr-1

Particles cm-2 yr-1

Particles cm-2 yr-1

Tagua Ta

gua

l

at 35.50S, lo

ng 71.16W, e

lev. 200 m

50 100 150 10 204 80 0 0 0

0

16.2

Mallin

Pollux

la

t 45.08S, lo

ng 71.84W, e

lev. 640 m

Markgraf et al. 2007

Rio Rubens

la

t 52.13S, lo

ng 71.88 W, e

lev. 220 m

Huber et al. 2003

Figure 7 Charcoal records for sites in southern South America (for location of sites see Fig. 2)


Caceres and Garcıa, 1989), whereas N. opaca tolerates deeperwaters, generally around 2m, but occasionally to 40m depth(Corrillion, 1957). The fluctuations between these two taxaindicate variable conditions, with shallower water between5300 and 4200 cal. a BP and after 2200 cal. a BP and slightlydeeper water between 4200 and 2200 cal. a BP. Thelimnological changes seen at Mallın Vaca Lauquen are moreor less synchronous with fluctuations in the abundance of thewetland taxon Typha at the Tagua Tagua site during the last5000 cal. a (Heusser, 1990). Freshwater algae at Laguna deCuleo during the last 2500 cal. a (Chilean lowlands at 348 S;Jenny et al., 2002; Villa-Martınez et al., 2004) are alsointerpreted to reflect increased moisture coupled with highhydrologic variability.Very high CHAR values between 5300 and 4200 cal. a BP

and to a lesser degree after 2200 cal. a BP indicate increased fireactivity locally and regionally, as a result of more fuel biomassand more suitable fire climate. Reduction of Nothofaguspumilio/N. antarctica forest cover and the expansion ofgrassland and disturbance taxa are consistent with more fireactivity at this time. High charcoal abundance during thisinterval is also seen in records from the Chilean lowlands atTagua Tagua (Heusser, 1990), the Puren-Lumaco valley(Abarzua, pers. comm.) and Lago Melli (Abarzua and Moreno,2008), beginning by 3000 cal. a BP.Higher-than-present moisture coupled with increased multi-

centennial variability by 5000 cal. a BP has been inferred frommarine and terrestrial records (McGlone et al., 1992; Lamyet al., 1999; Jenny et al., 2002; Villa-Martinez et al., 2004). Atlatitudes north of 408 S, it is likely that ENSO variability andassociated increased convective storm activity during El Ninoyears increased fire activity in the last 5500 cal. a (Kitzbergerand Veblen, 2003) (e.g. Markgraf and Diaz, 2000; Moy et al.,2002).Multi-centennial records, such as those presented here,

cannot document ENSO’s inter-annual and inter-decadal


variability. However, the present-day correlation between ElNino-related drought conditions (Montecinos and Aceituno,2003) and historical fire years (Kitzberger and Veblen, 2003)suggests that ENSO variability is a factor in mid-latitude fireoccurrence. Palaeoenvironmental records have indicated thatENSO variability was strengthened during the mid and lateHolocene (McGlone et al., 1992; Moy et al., 2002), and strongENSO variationsmay account for high fire activity at 5000 cal. aBP (primarily east of the Andes), 3000 cal. a BP and between2000 and 1000 cal. a BP. One could speculate that high fireactivity in the late Holocene seen similarly in the mid latitudeson both sides of the Andes reflects the influence of ENSO andthe build-up of fuels.

The last 400 cal. a show the influence of humans on thevegetation and fire regimes of the Mallın Vaca Lauquen region.A decrease in Poaceae and Nothofagus and an expansion ofEuropean weeds (Rumex acetosella, Plantago lanceolata) arenoted in the pollen record and attributed to a decrease ingrassland from livestock grazing and a decrease of Nothofagusdue to forest clearance. The impact began in the 16th centurywith the arrival of Spanish colonists. The markedly fasterdeposition time after 400 a (30 a cm�1) is probably not theresult of greater in-wash of clastic sediments into the basinbecause magnetic susceptibility values are low. Instead, it maybe the result of greater water run-off leading to more persistentmoisture in the basin and wetland development. Low levels ofcharcoal at this time, implying negligible or small-scaleburning, support the local wet site conditions.

Conclusion

High relief and steep climatic gradients in the southern Andesproduce diverse environments that are also reflected by a



diverse environmental history. Mallın Vaca Lauquen representsthe only high-elevation site so far studied with mixedNothofagus obliqua/N. pumilio forest located near thetransition to the Monte desert scrub vegetation. One uniquefeature of the past history is the Lateglacial and early Holocene(17 000–8600 cal. a BP) presence of Prumnopitys andina,indicating cooler and drier summers than today. This mightperhaps explain the absence of fires in that area at this time. Thelate appearance of Nothofagus obliqua after 5000 cal. a BPrepresents another environmental change not seen in otherrecords. Together with more permanent water in the lake, thepresence of N. obliqua marks the establishment of the present-day winter rain/summer drought climate regime in the lateHolocene. Although Austrocedrus chilensis is locally present inthe forests in small numbers, it never played a major role in thisregion’s past history, in contrast to records farther south andeast of the Andes.Unlike most sites in southern Patagonia, which show high

fire activity in the early Holocene, Mallın Vaca Lauquen andLaguna el Trebol at latitude 41.58 S record high charcoal levelsin themid and late Holocene. At Mallın Vaca Lauquen, inferredfire activity is highest between 5300 and 4200 cal. a BP andafter 2200 cal. a BP. The latitudinal differences in timing maybe related to fundamentally different climatic controls on fireactivity, in particular the stronger influence of ENSO variabilityat mid latitudes. Additional records are needed in this region todescribe the geographic extent of this pattern and betterunderstand the underlying drivers of past fire activity.

Acknowledgements NSF grants to CW (ATM 0117160 and ATM0714061) are acknowledged. We also thank Allison Bair, CaitlinMacCracken and Jaime Toney for assistance with laboratory analyses,and Victor Leshyk for help with the maps. Christy Briles helped with theage model. Northern Arizona University, Laboratory of PaleoecologyContribution 112.

References

Abarzua AM, Moreno PI. 2008. Changing fire regimes in the temperaterainforest region of southern Chile. Quaternary Research 69: 62–71.

Bianchi MM. 2000. Historia de fuego en Patagonia: registro de carbonvegetal sedimentario durante el Post-glacial y el Holoceno en el LagoEscondido (41- S, 72- W). Revista Cuaternario y Ciencias Ambien-tales, Publ. Especial 4: 23–29.

Bradbury JP, Whiteside MC. 1980. Paleolimnology of two lakes in theKlutlan Glacier region, Yukon Territory, Canada. QuaternaryResearch 14: 149–168.

Buck CE, Christen JA, James GM. 1999. BCal: an on-line Bayesianradiocarbon calibration tool. Internet Archaeology Vol. 7.

Caceres EJ, Garcıa A. 1989. Nitella hyalina (DC) Ag. (Characeae,Charophyta) in Argentina. Nova Hedwigia 48: 111–117.

Compagnucci RH, Vargas WM. 1998. Inter-annual variability of theCuyo rivers’ streamflow in the Argentinean Andes Mountains andENSO events. International Journal of Climatology 18: 1593–1609.

Corillion R. 1957. Les Charophycees de France et d’Europe Occiden-tale, (reprinted 1972). Koeltz: Koenigstein-Taunus.

Donoso C. 1974. Dendrologıa: Arboles y Arbustos Chilenos. ManualNo. 2, Facultad de Ciencias Forestales, Universidad de Chile.

Faegri K, Iversen J. 1989. Textbook of Pollen Analysis, (4th edn). Alden:London.

Garcıa A. 1987. Estudio del gametangio femenino de las Charophytaactuales de Argentina. Analisis comparado con el registro fosilcorrespondiente. PhD thesis, Facultad de Ciencias Naturales yMuseo, La Plata, Argentina.


Garreaud RD, Vuille M, Compagnucci R, Marengo J. 2008. Present-daySouth American climate. Palaeogeography, Palaeoclimatology,Palaeoecology (in press).

Grimm EC. 1987. CONISS: a FORTRAN 77 program for stratigraphi-cally constrained cluster analysis by the method of incremental sumof squares. Computer and Geosciences 13: 13–35.

GrimmEC. 1992. Tilia and Tilia-graph: pollen spreadsheet and graphicsprograms. In 8th International Palynological Congress, Aix-en-Prov-ence, 6–12 September 1992.

Haberle SG, Bennett KD. 2004. Postglacial formation and dynamics ofNorth Patagonian rainforest in the Chonos Archipelago, SouthernChile. Quaternary Science Reviews 23: 2433–2452.

Heusser CJ. 1990. Ice age vegetation and climate of subtropical Chile.Palaeogeography, Palaeoclimatology, Palaeoecology 80: 107–127.

Higuera P. 2008. CharAnalysis and Monte Carlo Age-Depth programs.http://charanalysis.googlepages.com/ [28 August 2008].

Huber U, Markgraf V. 2003. Holocene fire frequency and climatechange at Rio Rubens Bog, Southern Patagonia. In Fire and ClimaticChange in Temperate Ecosystems of the Western Americas, VeblenTT, Baker WL, Montenegro G, Swetnam TW (eds). EcologicalStudies 160. Springer: Berlin; 357–380.

Huber U, Markgraf V, Schaebitz F. 2004. Geographical and temporaltrends in Late Quaternary fire histories of Fuego-Patagonia, SouthAmerica. Quaternary Science Reviews 23: 1079–1097.

Jankosvka V, Komarek J. 2000. Indicative value of Pediastrum and othercoccal green algae in palaeoecology. Folia Geobotanica 35: 59–82.

Jenny B, Valero-Garces BL, Urrutia R, Kelts K, Veit H, Appleby PG,Geyh M. 2002. Moisture changes and fluctuations of the WesterliesinMediterraneanCentral Chile during the last 2000 years: the LagunaAculeo record (33- 50( S). Quaternary International 87: 3–18.

Kim J-H, Schneider RR, Hebbeln D, Muller PJ, Wefer G. 2002. Lastdeglacial sea-surface temperature evolution in the Southeast Pacificcompared to climate changes on the South American continent.Quaternary Science Reviews 21: 2085–2097.

Kitzberger T, Veblen TT. 2003. Influences of climate on fire in northernPatagonia, Argentina. In Fire and Climatic Change in TemperateEcosystems of the Western Americas, Veblen TT, Baker WL,Montenegro G, Swetnam TW (eds). Ecological Studies 160. Springer:Berlin; 296–321.

Komarek J, Fott B. 1983. Chlorophyceae (Gru}nalgen) Ordnung: Chlor-ococcales. Gattung Pediastrum Meyen. 1829. In Das Phytoplanktondes Su}sswassers, Huber-Pestalozzi G (ed.). Systematik und Biologie,Part 7(1). Schweizerbart: Stuttgart; 283–308.

Lamy F, Hebbeln D, Wefer G. 1999. High-resolution marine record ofclimatic change in mid-latitude Chile during the last 28,000 yearsbased on terrigenous sediment parameters. Quaternary Research 51:83–93.

Lamy F, Hebbeln D, Rohl U, Wefer G. 2001. Holocene rainfallvariability in southern Chile: a marine record of latitudinal shiftsof the Southern Westerlies. Earth and Planetary Science Letters 185:369–382.

Lamy F, Kaiser J, Ninnemann U, Hebbeln D, Arz HW, Stoner J. 2004.Antarctic timing of surface water changes off Chile and Patagonianice sheet response. Science 304: 1959–1962.

Leon RJC, Bran D, Collantes M, Paruelo JM, Soriano A. 1998. Grandesunidades de vegetacion de la Patagonia extra andina. EcologiaAustral 8: 125–144.

Markgraf V. 1987. Paleoenvironmental changes at the northern limit ofthe subantarctic Nothofagus forest, lat 37- S, Argentina. QuaternaryResearch 28: 119–129.

Markgraf V, Diaz HF. 2000. The past ENSO record: a synthesis. InEl Nino and the Southern Oscillation; Multiscale Variability andGlobal and Regional Impacts, Diaz HF, Markgraf V (eds). Cam-bridge University Press: Cambridge, UK; 465–488.

Markgraf V, Dodson JR, Kershaw PA, McGlone M, Nicholls N. 1992.Evolution of late Pleistocene and Holocene climates in circum SouthPacific land areas. Climate Dynamics 6: 193–211.

Markgraf V, Webb RS, Anderson KH, Anderson L. 2002. Modernpollen/climate calibration for southern South America. Palaeogeo-graphy, Palaeoclimatology, Palaeoecology 181: 375–397.

Markgraf V, Whitlock C, Haberle SG. 2007. Vegetation and fire historyduring the last 18,000cal yr B.P. in southern Patagonia: Mallin



Pollux, Coyhaique, Province Aisen (45- 41(30((; S, 71- 50( 30((; W,640m elevation. Palaeogeography, Palaeoclimatology, Palaeoecol-ogy 254: 492–507.

McGlone MS, Kershaw AP, Markgraf V. 1992. El Nino/Southern Oscil-lation climatic variability in Australasian and South Americanpaleoenvironmental records. In El Nino: Historical and PaleoclimaticAspects of the Southern Oscillation, Diaz HF, Markgraf V (eds).Cambridge University Press: Cambridge, UK; 435–462.

Montecinos A, Aceituno P. 2003. Seasonality of the ENSO-relatedrainfall variability in Central Chile and associated circulationanomalies. Journal of Climate 16: 281–296.

Moreno P. 1997. Vegetation and climate near Lago Llanquihue in theChilean Lake District between 20 200 and 9500 14C yr BP. Journal ofQuaternary Science 12: 485–500.

Moreno P. 2004. Millennial-scale climate variability in northwestPatagonia over the last 15 000 yr. Journal of Quaternary Science19: 35–47.

Moreno P, Leon AL. 2003. Abrupt vegetation changes during the lastGlacial–Holocene transition in mid-latitude South America. Journalof Quaternary Science 18: 787–800.

Moy CM, Seltzer GO, Rodbell DT, Anderson DJ. 2002. Variability of ElNino/Southern Oscillation activity at millennial timescale during theHolocene epoch. Nature 420: 162–165.

Paez MM, Schabitz F, Stutz S. 2001. Modern pollen-vegetation andisopoll maps in southern Argentina. Journal of Biogeography 28:997–1021.

Paruelo JM, Beltran A, Jobbagy E, Sala OE, Golluscio RA. 1998. Theclimate of Patagonia: general patterns and controls on biotic pro-cesses. Ecologia Austral 8: 85–101.

Romero OW, Kim J-H, Hebbeln D. 2006. Paleoproductivity evolutionoff central Chile from the last glacial maximum to the early Holo-cene. Quaternary Research 65: 519–525.

Sandgren P, Snowball I. 2001. Application of mineral magnetic tech-niques to paleolimnology. In Tracking Environmental Change UsingLake Sediments, Vol. 2: Physical and Geochemical Methods,Last WM, Smol JP (eds). Kluwer Academic: Dordrecht; 217–238.


Stuiver M, Reimer PJ, Reimer RW. 2005. CALIB 5.0 (WWW programand documentation).

Veblen TT, Burns BR, Kitzberger T, Lara A, Villalba R. 1995. Theecology of the conifers of southern South America. In Ecology ofthe Southern Conifers, Enright NJ, Hill RS (eds). Melbourne Uni-versity Press: Melbourne; 120–155.

Veblen TT, Kitzberger T, Villalba R, Donnegan J. 1999. Fire history innorthern Patagonia: the roles of humans and climatic variation.Ecological Monographs 69: 1–47.

Veblen TT, Kitzberger T, Raffaele E, Lorenz DC. 2003. Fire history andvegetation changes in northern Patagonia, Argentina. In Fire andClimatic Change in Temperate Ecosystems of the Western Americas,Veblen TT, Baker WL, Montenegro G, Swetnam TW (eds).Ecological Studies 160. Springer: Berlin; 265–295.

Villagran C. 1991. Historia de los bosques templados del sur de Chiledurante el Tardiglacial y Postglacial. Revista Chilena de HistoriaNatural 64: 447–460.

Villagran C. 2001. Un modelo de la historia de la vegetation de laCordillera de La Costa de Chile central-sur: la hipotesis glacial deDarwin. Revista Chilena de Historia Natural 74: 1–15.

Villalba R. 1994. Tree-ring and glacial evidence for theMedieval Epochand the Little Ice Age in southern South America. Climate Change 26:183–198.

Villa-Martinez R, Villagran C, Jenny B. 2004. Pollen evidence for late-Holocene climatic variability at Laguna de Aculeo, Central Chile (lat.34- S). The Holocene 14: 361–367.

Whitlock C, Bartlein PJ, Markgraf V, Ashworth AC. 2001. The mid-latitudes of North and South America during the last glacial maxi-mum and early Holocene: similar paleoclimatic sequences despitedifferent large-scale controls. In Interhemispheric Climate Linkages,Markgraf V (ed.). Academic Press: San Diego, CA; 391–416.

Whitlock C, Bianchi MM, Bartlein PJ, Markgraf V, Marlon J, Walsh M,McCoy N. 2006. Postglacial vegetation, climate, and fire historyalong the east side of the Andes (lat 41–42.5- S) Argentina. Qua-ternary Research 66: 187–201.

Whitlock C,Moreno P, Bartlein PJ. 2007. Climatic controls of Holocenefire patterns in southern South America. Quaternary Research 68:28–36.


late quaternary vegetation and fire history in the northernmost nothofagus forest region: mallín...

Documents