See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/222678338

Climate variability over the last 9900 cal yr BP from a swamp forest pollen
record along the semiarid coast of Chile

Article in Quaternary Research · September 2006

DOI: 10.1016/j.yqres.2006.04.003 · Source: OAI

CITATIONS READS

123 308

2 authors, including:

Antonio Maldonado
Centro de Estudios Avanzados en Zonas Áridas University of La Serena
179 PUBLICATIONS 3,216 CITATIONS

SEE PROFILE

All content following this page was uploaded by Antonio Maldonado on 26 February 2018.

The user has requested enhancement of the downloaded file.

 Quaternary Research 66 (2006) 246 – 258
www.elsevier.com/locate/yqres

Climate variability over the last 9900 cal yr BP from a swamp forest
pollen record along the semiarid coast of Chile
Antonio Maldonado a,⁎, Carolina Villagrán b
a
Centro de Estudios Avanzados en Zonas Aridas (CEAZA), Universidad de La Serena, Casilla 599, La Serena, Chile
b
Laboratorio de Sistematica y Ecología Vegetal, Departamento de Ciencias Ecológicas, Facultad de Ciencias,
Universidad de Chile, Casilla 653, Ñuñoa, Santiago, Chile
Received 20 May 2005
Available online 9 June 2006

Abstract

We present a fossil pollen analysis from a swamp forest in the semiarid coast of Chile (32°05′S; 71°30′W), at the northern influence zone of
southern westerly wind belt. A ∼10,000 cal yr BP (calendar years before 1950) palynological sequence indicates a humid phase characterized by
dense swamp forest taxa dated between ∼9900 and 8700 cal yr BP. The presence of pollen-starved sediments with only scant evidence for
semiarid vegetation indicates that extreme aridity ensued until ∼5700 cal yr BP. The swamp forest recovered slowly afterwards, helped by a
significant increase in moisture at ∼4200 cal yr BP. A new swamp forest contraction suggests that another slightly less intense drought occurred
between ∼3000 and 2200 cal yr BP. The swamp forest expansion begins again at ∼2200 cal yr BP, punctuated by a highly variable climate.
Comparisons between the record presented here with other records across the region imply major variations in the extent of the southern westerlies
during the Holocene. This variability could have been caused either by latitudinal displacements from the present mean position of southern
westerlies wind belt or by changes in the intensity of the South Pacific Subtropical Anticyclone, both of which affect winter precipitation in the
region.
© 2006 University of Washington. All rights reserved.

Keywords: Northern Chile; Holocene; Pollen record; Paleoenvironmental change; Paleoclimate; Southern westerlies; Semiarid; Swamp forests

Introduction The Chilean Mediterranean zone is highly sensitive to

interannual variations associated with the El Niño-Southern
Coastal semiarid Chile extends from the central Mediterra- Oscillation (ENSO). Negative phases of ENSO (associated with
nean region to the south up to the Atacama Desert in the north. El Niño events) are associated with abnormally hot and wet winters
Dry summers and variable winter rains, accompanied by whereas abnormally dry and cold winters prevail during the
occasional winter drought, especially at its northern limit, positive phase (associated with La Niña events) (Aceituno, 1988).
characterize this portion of Chile (VanHusen, 1967). The winter- Seasonal latitudinal changes in the SPA, reduced frequency
wet climate is generated from the interaction of two large systems of westerlies related stormfronts, and ENSO-related effects all
that govern the extra-topical region of Chile: the southern combine to produce a highly variable precipitation regime in
westerlies (center of influence situated at 47°) and the South north-central Chile that is sensitive to climate fluctuations. It is
Pacific Anticyclone (SPA). During summers the SPA is displaced thus ideally suited for studying paleoclimatic conditions and
southwards, effectively blocking the westerly storm fronts north reconstruction of past variations of the different climatic
of 34°S. In contrast, during the winter the SPA weakens and systems governing the region. Very few high-resolution
migrates northwards, allowing frequent stormfronts associated paleoclimatic studies exist for central Chile (33–36°S) and
with the southern westerlies to reach into the semiarid portion of adjacent Norte Chico (33–27°S) and most are not older than
the country (Garreaud and Aceituno, 2002; Miller, 1976). 6000 cal yr BP. All published records show alternations
between wet and dry periods (Earle et al., 2003; Grosjean et
⁎ Corresponding author. Fax: +56 51 334741. al., 1997a, 1998; Jenny et al., 2002b; Maldonado and Villagrán,
E-mail address: [email protected] (A. Maldonado). 2002; Villagrán, 1982; Villa-Martínez and Villagrán, 1997),
0033-5894/$ - see front matter © 2006 University of Washington. All rights reserved.
doi:10.1016/j.yqres.2006.04.003
 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258 247

although there is little agreement on how long these phases central Chilean coast, between 36°30′S and 30°00′S
actually lasted. Most records agree, however, on the presence of (Fuenzalida, 1965; San Martín et al., 1988). These isolated,
a prolonged drought during the early and middle Holocene discontinous forests are commonly groundwater-fed systems
(Heusser, 1990; Jenny et al., 2002b; Kim et al., 2002; Lamy et restricted to particular geomorphic settings with periodic or
al., 1999; Marchant et al., 1999; Veit, 1996; Villagrán and permanent flooding (Ramírez et al., 1983, 1995; San
Varela, 1990; Villa-Martínez et al., 2003), although the timing Martín et al., 1988; Varela, 1981). Seven swamp forests
and duration vary between different proxy records. can be found in small ravines between the coastal villages
Detailed middle to late Holocene records have been obtained of Los Vilos and Pichidangui (from 31°50′ to 32°05′S;
in the past from the coastal swamp forests of north-central Chile ∼71°30′W). Most of these are very close to the coast and
(Villa-Martinez and Villagrán 1997; Maldonado and Villagrán, have an east–west orientation (Maldonado, 1999). These
2002). Here, we present a record of paleoecological change from ravines are located on the coastal plain associated with
a small swamp forest record characterized by thick accumula- small hydrographic basins where local groundwater re-
tions of organic sediments intercalated with soils. Our objective charge is determined by the infiltration along the adjacent
was to obtain a detailed and diverse fossil pollen record of past coastal mountain range (which reaches up to 1000 m in
climatic variations that spanned the entire Holocene. elevation). Thus, forest hydrology depends chiefly upon local
geomorphology, regional winter precipitation and groundwater
Study area recharge.
The coastal plains and adjacent mountain slopes are
Swamp forests dominated by Myrtaceae (Myrceugenia dominated by zonal vegetation characterized by plants from
exsucca and Luma chequen) are common along the north- the family Asteraceae (Baccharis vernalis and Bahia

Figure 1. Study area map showing the locality and topography of Palo Colorado swamp forest (contour interval 100 m).
248 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

Table 1
Radiocarbon dates for Palo Colorado record: Laboratory number, depth, radiocarbon dating, calendar year calibration (maximum and minimum year), calibrated age
(mid point), δ13C and material dated
Laboratory no. Depth, cm Age 14C Age, max and min range cal yr BP Age, mid point δ13C ‰ Material
yr BP (2 sigma; 95,4%) cal yr BP
AA-63302 26–27 510 ± 37 502–628 565 −25.97 Charcoal
AA56375 29–30 90 ± 31 11–141 76 −27.5 Plant Macro-remains
AA-63303 61–62 903 ± 38 738–915 826 −27.32 Charcoal
AA-56374 96–98 1581 ± 33 1398–1538 1468 −27.58 Bulk
AA-63304 129–130 1958 ± 39 1825–1990 1908 −27.69 Bulk
AA-56376 159–160 2151 ± 34 2007–2306 2156 −28.2 Charcoal
AA-56373 205–206 3165 ± 54 3254–3553 3404 −27.81 Plant Macro-remains
AA-53359 226–228 3851 ± 43 4151–4413 4282 −28.28 Charcoal
Beta-204526 254–255 5550 ± 40 6287–6406 6346 −25.9 Bulk
AA-58073 277–279 7578 ± 59 8211–8536 8374 −26.3 Charcoal
AA-53358 285–286 8466 ± 84 9265–9595 9353 −27 Bulk

ambrosioides) associated with Puya chilensis, Haplopappus Methods

foliosus, Fuchsia lycioides and Lithrea caustica. The
transition between zonal and azonal vegetation can be We sampled sediment soils from the Palo Colorado swamp
abrupt, depending on local geomorphology. Within a few forest using a 5-cm-diameter Wright square-rod piston corer
meters, surverys of coastal matorral can shift into a swamp (Fig. 1; 32°05′S; 71°29′W). To describe the stratigraphic
forest. These swamp forests are floristically very distinctive column we used X-ray images, texture analysis and sediment
and dominated by myrtles (mainly Myrceugenia exsucca and color. Samples were extracted at regular intervals from the
Luma chequen) associated with other arboreal species such sediment core for loss-on-ignition (LOI) and fossil pollen
as Drimys winteri, Escallonia revoluta, Rhaphithamnus analyses.
spinosus and the vine Cissus striata. The LOI analyses were performed by drying and
At Palo Colorado (Fig. 1), the transition between swamp burning sediment samples at 550°C and 925°C to eliminate
forest and the coastal matorral is represented by a vegetation organic matter and carbonates, respectively. We compared
belt dominated by arboreal sclerophyllous species such as sediment weight before and after treatment to calculate
Maytenus boaria, Quillaja saponaria, Schinus latifolius, percent and density of organic and inorganic matter as well
Pouteria splendens, and associated shrubs such as Eupator- as carbonates.
ium salvia, E. glechonophyllum and Fuchsia lycioides. The We extracted pollen from the stratigraphic cores using
sclerophyllous forest is also abundant in central Chile, but standard methods (Faegri and Iversen, 1989), including
along the semiarid coast of the Norte Chico, it is restricted acetolysis. Pollen content from the samples was concentrated
to creek bottoms and wetter south-facing slopes of the with ultrasound and samples were mounted using glycerine-
coastal mountains. jelly and glycerine.

Figure 2. Graph showing radiocarbon (circles) and calendar (crosses) dates, and sixth-order polynomial regression against sediment depth.
 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258 249

excluded owing to probable contamination from modern root

systems (Table 1; Fig. 2).

Results

Stratigraphy and chronology

The Palo Colorado record reaches a depth of 292 cm with a

maximum date of ∼9400 cal yr BP close to the base of the
column. Ten radiocarbon dates are given in Table 1 and plotted
in Figure 2 against core depth. Sediments from the middle to
late Holocene (from 4200 cal yr BP) accumulated at a much
higher rate than those from the early to middle Holocene (9900–
4200 cal yr BP).
Basal core sediment is composed of a 12-cm-thick organic
silt layer with a few sand intercalations (Fig. 3). Gray silts with
low organic content (10%, from LOI analysis) are prevalent
from 280 cm to 231 cm. These sediments were pollen-starved
and counting was infeasible. A depositional hiatus in this
section is, however, not very probable as indicated by a 14C date
at 254 cm (Figs. 2 and 3). This is followed by a section of
organic-rich sediments (50% organic matter content) to 204 cm
depth. From this point upwards, gray silt sediments with low
organic content reappear up to 160 cm. The upper section is
dominated by high organic content in the sediments (∼70%).
The last 30 cm exhibits a moderate decrease in organic content,
which reverses towards the top (Fig. 3).
Although organic matter density displayed similar trends as
the lithology, the changes are less distinctive. Inorganic matter
density agrees more with the data from percent organic matter
content (Fig. 3). Carbonates exhibited very low percentages
throughout the entire sequence (<7%, data not shown but
considered within the error measure range from the LOI
method).
Figure 3. Results from loss on ignition analysis in Palo Colorado record showing
Percent pollen diagram
radiocarbon dates (cal yr BP), sediment depth, lithology, % organic matter,
organic matter density (g/cc) and inorganic matter density (g/cc).
The percentage diagrams (Fig. 4) document the vegetation
history from Palo Colorado over the last ∼10,000 cal yr BP.
The basic pollen count for each sample includes at least The diagram is divided into three pollen zones, based on
300 terrestrial pollen grains (excluding aquatics and ferns conspicuous changes in the pollen stratigraphy according to
taxa). The results are expressed in percent per taxon. stratigraphically constrained CONISS ordination, using
Lycopodium spore tablets (2 tablets per cm3 sediment) square-root transformed data and the Edward and Cavalli-
were added to calculate pollen concentration and accumula- Sforza's chord distance as a dissimilarity coefficient (Grimm,
tion rates (influx) (Stockmarr, 1971). Pollen zones were 1987).
defined with the aid of a pollen affinity dendrogram
generated by the Contrained Incremental Sum of Squares Zone PC-1 (∼10,000 to ∼8700 cal yr BP)
(CONISS) program (Grimm, 1987) and also considering Swamp forest indicators, mainly pollen of Myrtaceae
changes in lithology. (probably Myrceugenia exsucca and Luma chequen; ∼35%)
Major breaks in stratigraphy were sampled and the and Escallonia (probably Escallonia revoluta; ∼30%), are
sediments radiocarbon dated at the Accelerator Mass Spec- dominant. Other swamp forest taxa are present in smaller
trometry (AMS) Laboratory of the University of Arizona proportions, such as Drimys winteri (<6%) and the vine
(USA) and Beta Analityc INC. (USA). Results were calibrated Cissus striata (<7%). Of note are the trees Maytenus and
to calendar years using CALIB 5.1 (Table 1; (Reimer et al., Anacardiaceae (Schinus and Lithrea), both forest-edge taxa.
2004). A sixth-order polynomial regression (R2 = 0.9997) was Non-arboreal taxa are present in smaller proportions: Poaceae
used to interpolate between radiocarbon dates along the age- (<17%), Asteraceae-Tubuliflorae and Asteraceae-Liguliflorae
depth profile (Fig. 2). Only one date (90 ± 31 14C yr BP) was (<22% and <25% respectively). Aquatic taxa and fern spores
 250
A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258
Figure 4. Palo Colorado percentages pollen diagram showing dates, depth, age (cal yr BP), lithology, zones and CONISS dendrogram.
 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258
Figure 5. Influx diagram of selected taxa from Palo Colorado record, showing depth, age (cal yr BP), lithology, deposition times (yr/cm), total accumulation rates (grains/cm2/year) and zones.

251
252 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

exhibit low but constant proportions (<12%). The middle of Phase PC-3E (∼600 cal yr BP to the present)
this zone displays a decrease in arboreal taxa and concomitant A new decrease in swamp forest pollen indicators,
increase in Poaceae and Asteraceae. Sediments during this particularly in the Myrtaceae, which reach minimum values at
interval display a mostly sandy texture. around 470 cal yr BP Maytenus boaria percentages (<11%)
remain high and constant as well as Asteraceae-Tubuliflorae
Zone PC-2 (∼8700 to 5700 cal yr BP) (<10%). Other non-arboreal taxa, such as Poaceae (<8%) and
Characterized by the absence of arboreal pollen, with Brassicaceae (<9%), increase slightly.
dominance of shrubs taxa at the base, but sediments are
pollen-starved above. Influx diagram

Phase PC-2A (∼8700 to 7800 cal yr BP) The influx diagram (total pollen grains per cm2 per year)
Abrupt decrease in arboreal taxa and replacement by shrub shows a good correspondence between pollen frequencies and
indicators (Asteraceae <50% both Liguliflorae and Tubuli- lithology (Fig. 5).
florae and Poaceae <15%). Aquatic and fern spores are at low Zone PC-1: very low pollen influx. Total pollen accumula-
and constant levels throughout (∼6%). This phase displays tion reaches only 600 grains/cm2/yr and sediment deposition is
the lowest pollen influx values throughout the entire column. very slow (ca. 100 yr/cm). Best-represented taxa are the swamp
forest indicators Myrtaceae (<175 grains/cm2/yr) and Escallo-
Phase PC-2B (7800 to 5700 yr cal BP) nia (<175 grains/cm2/yr). Poaceae is also present (<65 grains/
Pollen-starved sediments. Only traces of open area vegeta- cm2/yr).
tion (Asteraceae-Liguliflorae) were observed, probably due to Zone PC-2: nil or very low pollen influx. Total pollen
pronounced droughts which exposed and oxidized sediments accumulation rates are very low, fluctuating between 20 and
leading to poor pollen preservation. 70 grains/cm2/yr. Deposition times are as low as the previous
zone. Non-arboreal taxa are best represented; Asteraceae,
Zone PC-3 (5700 cal yr BP to the present) Tubuliflorae and Liguliflorae jointly attain 30 grains/cm2/yr.
Characterized by a series of pollen oscillations indicative of Zone PC-3: The first phase is also characterized by low
current swamp forests. From these changes and from the pollen influx, increasing only at the end. Total accumulation and
sediment sequence we distinguished five different phases. deposition times at the base are similar to that of previous zones.
Escallonia shows maximum values of <130 grains/cm2/yr
Phase PC-3A (∼5700 to 4200 cal yr BP) together with Anacardiaceae (∼60 grains/cm2/yr) and non-
Swamp forest indicators reappear and are dominated by arboreal Poaceae and Asteraceae (<60 grains/cm2/yr).
Escallonia (<25%), Myrtaceae (<40%), Maytenus (<9%) and Although pollen influx in the second phase begins with
Anacardiaceae (<8%). Non-arboreal taxa, Brassicaceae (<16%), similar values as the end of the first phase, almost all taxa
Poaceae (<9%) and Liliflorae (<11%, mainly Amaryllidaceae increase their influx values in the upper portion. Higher total
and Liliaceae) are also present. Aquatic taxa and spores display accumulation rates are reached at the end of this phase.
little variation. Depositional times per cm are still high although they tend to
decrease to ∼17 yr/cm.
Phase PC-3B (∼4200 to ∼2200 cal yr BP) The third phase shows a trend towards increasing influx
Swamp forest taxa increase in abundance, initially Myrta- throughout the whole period. Arboreal taxa showed the highest
ceae (<43%) and then Escallonia (<48%). The second half of values during this phase. Non-arboreal taxa (Poaceae, Fabaceae
the phase shows an increase of a sclerophyllous arboreal and Brassicaceae) showed higher values than during the
species: Maytenus boaria (<30%) along with a decrease in previous phase with maximum values of Liliflorae, while
swamp forest species. Asteraceae-Liguliflorae (<14%) also Apiaceae increase towards the end. Depositional times decrease
increased towards the second half of the phase. to very low values, ∼13 yr/cm.
The two next phases display high pollen influx for almost all
Phase PC-3C (∼2200 to 1300 cal yr BP) taxa at their initiation. From 470 cal yr BP onwards, influx
Widespread recovery of swamp forest indicators, chiefly values decrease in all taxa. Depositional times fluctuate between
Myrtaceae (<74%), Escallonia (<56%) and Drimys winteri 28 and 11.8 yr/cm.
(<7%). In contrast, sclerophyllous species Maytenus boaria and
Azara celastrina together with non-arboreal indicators (with Discussion
exception of Fabaceae and Liliflorae) decrease substantially.
Paleoclimatic interpretation
Phase PC-3D (1300 to 600 cal yr BP)
Characterized by mostly high but variable percentages of The basal section of the Palo Colorado record (zone PC-1,
swamp forest taxa. This phase begins with a decrease in ∼10,000 cal yr BP) displays dark silt with high organic matter
Myrtaceae and high percentages of Escallonia, followed by content and a pollen record dominated by swamp forest
maximum values (<11%) of Drimys winteri. Towards the end, indicators. Clearly, wet conditions prevailed during the early
high values of Maytenus boaria are indicated. Holocene. These conditions changed in zone PC-2. The
 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258 253

prevalence of non-arboreal taxa (mostly Asteraceae) together same cores indicate that sea surface temperatures between
with overall low pollen influx and the abundance of inorganic 10,000 and 8000 cal yr BP were similar to modern conditions
sediments all indicate that an open landscape with very arid (Kim et al., 2002). At the continental scale, an increase in
conditions existed between ∼8700 and ∼7800 cal yr BP. This humidity is inferred from sites along the south coast of Peru,
was followed by deposition of increasingly pollen-starved associated with onset of an ENSO-like signal (Carré et al.,
inorganic sediments (insufficient for counting), which together 2005; Sandweiss, 2003) (Fig. 6, Table 2).
with traces of Asteraceae pollen suggest that aridity became Aridity intensified from 8700 to 5700 cal yr BP, as inferred by
even more intense between 7800 and 5700 cal yr BP. As the predominance of non-arboreal taxa followed by pollen sterile
previously stated, we rule out the possibility of any significant sediments at Palo Colorado. Other records for Mediterranean
depositional hiatuses based on the stratigraphy and chronology Chile and the Norte Chico also indicate hot and dry conditions.
of the record. For example, paleopedological evidence reveals an overall >3°C
This pronounced drought ended at ∼5700 cal yr BP, when temperature increase between ∼8000 and ∼5900 cal yr BP (Veit,
swamp forest indicators at Palo Colorado began to reappear. 1996). An ephemeral shallow lake that existed at Laguna Aculeo
Climate changed only gradually until 4200 cal yr BP, when until 8500 cal yr BP was followed by a saline lake with carbonate
both the percent pollen diagram and the influx pollen precipitation until 5700 cal yr BP (Jenny et al., 2002b). The
diagram show important increases. This shift is also dominance of Chenopodiaceae pollen beginning at 7500 cal yr
associated with a marked increase in organic matter content BP is indicative of high evaporation rates at Laguna Aculeo,
in sediments (zone PC-3). concordant with increased salinity and a hot and dry climate
Between 3000 and 2200 cal yr BP, the decline in swamp (Villa-Martínez et al., 2003). Arid stable conditions between
forest taxa and arboreal taxa in general, together with the 8000 and 4000 cal yr BP associated with a sea surface
increase of sclerophyllous tree Maytenus boaria and in the temperature increase of 2.5°C within the span of 500 yr,
relative proportion of Asteraceae-Liguliflorae, again suggest beginning at 8000 cal yr BP, have been estimated from offshore
onset of (albeit milder) drought conditions, with a maximum marine cores (Lamy et al., 1999). According to Kim et al. (2002),
peak in aridity at ∼2750 cal yr BP. Beginning at 2200 cal yr sea surface temperatures may have been at their warmest at 6500
BP (PC-3C), Myrtaceae and Drimys winteri slowly recovered, cal yr BP Geomorphological data at Algarrobo (33°22′S) also
suggesting optimal swamp forest conditions most likely suggest a sea-level rise of 3.8 m at 6500 cal yr BP with respect to
associated with a return to wetter conditions. The high degree the modern shoreline (Hervé et al., 2003). Furthermore, the
of variability displayed by swamp forest pollen, taken together maximum marine transgression for the region (∼30°S) is
with the variations in the proportion of arboreal/non-arboreal registered at 6800 cal yr BP (Ota and Paskoff, 1993). In the
taxa and the relative increase in Liliflorae (monocot bulb high Andes, the Laguna del Negro Francisco record (4000 m,
plants adapted to arid regions with variable precipitation 27°S), also located in the influence zone of the westerlies,
regimes), suggest a highly variable climate regime within an indicates that a saline lake persisted in an arid climate between
overall humid general context. Influx values and deposition 6800 and 4200 cal yr BP (Grosjean et al., 1997a). Arid
rates suggest that this humid phase may have been more conditions are also indicated by swamp forest pollen records in
intense than the earliest humid period recorded at the bottom Central Chile and the Norte Chico. At Ñague, the pollen record
of sequence. A new decline in swamp arboreal taxa occurs (∼32°S) shows a dominance of matorral at 6100 cal yr BP, later
during the last phase of the record (PC-3E, from ∼620 cal yr replaced by wet indicators at 4200 cal yr BP (Maldonado and
BP) with an increase in Maytenus boaria and Asteraceae Villagrán, 2002). At Quintero, (33°S, site Quintero II) hot and
Tubuliflorae and the almost complete disappearance of Drimys dry conditions are suggested by the dominance of Chenopodia-
winteri. These changes may be attributable to recent ceae between 6900 and 4500 cal yr BP (Villa-Martínez and
anthropogenic disturbance. Villagrán, 1997). Even though the precise chronologies may
vary, the evidence for an arid phase during the early to middle
Regional correlations Holocene is pervasive throughout all the available records.
Differences in timing and signal are most likely due to the
Similar evidence for an early Holocene wet phase is also sensitivity of the different proxy-data used. The trends indicated
indicated in other paleoclimate records from Central Chile and in the sea surface temperature proxy records suggest an abrupt
Norte Chico. Using paleopedological methods between 27 and beginning followed by a gradual demise of the early to mid-
33°S, Veit (1996) suggest wet conditions at the beginning of Holocene arid phase (Kim et al., 2002). One mechanism for
Holocene at ∼8300 cal yr BP. In Central Chile, the Laguna creating this arid phase is through an enfeebling of the westerly
Aculeo record (33°50′S) indicates arid conditions from the winds.
early to mid-Holocene (>5700 cal yr BP); however, a gyttja Consensus has yet to be reached, however, on whether this
sediment layer present in this record at ∼9000 cal yr BP is arid phase had any impact in the Chilean central Andes and
interpreted as a brief lake beach phase followed by fluvial adjacent Atacama Desert (Betancourt et al., 2000; Grosjean,
sedimentation (Jenny et al., 2002b). Evidence from marine 2001; Grosjean et al., 2003; Latorre et al., 2003; Quade et al.,
sediments of the coast of Central Chile (33°S) have also been 2001). If the central Andes was dry during this period, as the
interpreted in terms of increasing aridity culminating at ∼8000 majority of the paleoclimatic evidence demonstrates (Abbott et
cal yr BP (Lamy et al., 1999). Alkenone analyses from the al., 2003; Baker et al., 2001; Grosjean, 1994, 2001; Grosjean
254 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

Figure 6. Map showing paleoclimate sites discussed in the text for the central and semiarid region of Chile (circles, and triangles for Veit's localities) and the Palo
Colorado site (square).

et al., 1997b, 2003; Grove et al., 2003; Maldonado et al., by a period of high variability from 2200 cal yr BP to the
2005; Paduano et al., 2003; Rigsby et al., 2003; Servant and present. Other records across the region also indicate an
Servant-Vildary, 2003; Tapia et al., 2003). The combined increase in climatic variability during the late Holocene (Jenny
evidence would suggest a weakening of both the tropical and et al., 2002b; Marchant et al., 1999; Riedinger et al., 2002;
extra-tropical circulation. The first would be withdrawing Villa-Martínez et al., 2003). At Laguna Aculeo, a wet and
towards the equator and the second towards the pole, thus variable period began at 1800 cal yr BP and continues to the
significantly widening the South American “arid diagonal” present (Jenny et al., 2002a,b). The Quintero swamp forest
during this period. became established at ∼1600 cal yr BP (Villagrán and Varela,
The Palo Colorado record indicates that the climate became 1990) and at 1950 cal yr BP in Quintero II (Villa-Martínez and
increasingly wetter between 5700 and 3000 cal yr BP, followed Villagrán, 1997). Arid phases at the coastal Palo Colorado site,
Table 2
Summary of paleoclimatics sites and reconstructed climatic conditions discussed in the text for the central and semiarid region of Chile
Geographics location Proxi-data and temporal range Interpretation References
Chile Between 27 and 33°S Paleosols (11,000–0 Increased influence of the westerlies (more humidity) between Veit, 1996
cal yr BP) 11,000–8160, 5800–4100, 3200–1800 300–0 cal yr BP.
Lake Negro Francisco (27°28′S; Lacustrine sediments 6800–4200 cal yr BP arid conditions Grosjean et al.,
69°14′W; 4125 m) (6800–0 cal yr BP) After 4200 humid increase with 2 peaks, between 3200–2800 1997a
and 2400–1800 cal yr BP.
Encierro Valley (29°05′S; Glacial geomorphology 20–30% more humidity on the area, between 3200 and 1800 cal yr BP Grosjean et al.,
69°55′W; 4500 m) (4000–1900 cal yr BP) Glacial advance younger than 2700 cal yr BP. 1998
La Herradura and Tongoy beach Shorelines (6800 cal yr BP) High marine level at 6800 cal yr BP. Ota and Paskoff,

A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

(29°58′S; 71°20′W; 0 m and 1993
30°17′S; 71°32′W; 0 m)
Ñague (31°50′S; 71°28′W; 50 m) Pollen record 6200–4200 cal yr BP relatively arid condition Maldonado and
(6200–0 cal yr BP) Villagrán, 2002
4200–3200 cal yr BP relatively humid conditions Maldonado and Jackson, 2005a
3200–1300 cal yr BP relatively arid conditions
1300–0 cal yr BP relatively humid conditions
a recent date show this start at 1700 cal yr BP.
Quereo (31°55′S; 71°30′W; 10 m) Pollen and sedimentology 13,000–11,300 cal yr BP humid conditions Villagrán and
(13.000–0 cal yr BP) 11,300–3000 cal yr BP arid conditions Varela, 1990
3000–0 cal yr BP humid conditions
Quintero (32°47′S; 71°32′W; 10 m) Pollen (6700–0 cal yr BP) 6700–5800 cal yr BP arid conditions Villa-Martínez
5800–4500 cal yr BP arid conditions, with mash taxa and Villagrán,
4500–2000 cal yr BP humid conditions 1997
2000–0 cal yr BP more humid conditions
Marine Core offshore of Valparaiso Sediments and Alkenones (33.000–0 cal yr BP) Sediments: trend toward arid condition Lamy et al., 1999
(GIK-17748-2, 32°45′S; 72°02′W; between deglaciation and 8000 cal yr BP; 8000–4000 cal yr BP, stable conditions, Kim et al., 2002
−2545 m and GeoB-3302-1, with increased aridity in the Coastal Range. 4000–0 cal yr BP more variability and
33°13′S; 72°06′W; −1498 m) more humid conditions. Alkenones: between 10.000 and 8000 cal yr BP, sea surface
temperature (SST) was similar to modern conditions. Between 8000 and 7500 cal yr BP
SST increased 2.5°C, high SST are maintained between 7500 and 5000 cal yr BP, after
that SST disminuye gradualmente hasta lo actual.
Algarrobo beach (33°22′S; Sedimetology (6500 cal yr BP) High marine level at 6500 cal yr BP. Hervé et al., 2003
71°45′W; 6 m)
Lake Aculeo (33°50′S; Pollen and Sedimetology Sediments: 9500–5700 cal yr BP arid period Jenny et al., 2002a,b
70°54′W; 350 m) (9500 cal yr BP) 5700–0 cal yr BP humid conditions increasing, with modern levels at 3200 cal yr BP Villa-Martínez et al., 2003
and little arid phase between 2200 and 1800 cal yr BP.
Fluvial units were deposited before 9000 cal yr BP an after 5700 cal yr BP Villa-Martínez et al., 2004
and more frequent since 3200 cal yr BP.
Flood events, possibly ENOS-related occurs between 1800–1600,
1500–1300, 700–300, 150–0 cal yr BP.
Pollen: 7500–5700 cal yr BP drying indicators
5700–3200 cal yr BP humidity indicators
a
Holocene climate change and human settlement on the semiarid coast of Chile (32°S).

255
256 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

however, occur between 3000 and 2200 cal yr BP, whereas cal yr BP is correlated with periods of higher occurrence of
records from the High Andes indicate a humid phase with ENSO events (particularly El Niño events). Although modern
maximum moisture at 2600 cal yr BP (Grosjean et al., 1998). ENSO events have an average recurrence of between 3 and 6 yr
This suggests that past changes during the late Holocene may (Trenberth, 1976), and despite the fact that the Palo Colorado
not occur synchronously along the Andes and the coast. One record reaches a maximum resolution of 14–15 yr/cm of
possible explanation is that the Andean highlands receive part sediment during this period, the accumulation of several El
of their moisture during the summer wet season (i.e. through a Niño events in a sample can be expressed by relatively high
tropical source). These sources must have been of considerable values of pollen of Myrtaceae and geophytes (bulb plants).
magnitude for glaciers advance to occur during episodes of Conversely, less frequent El Nino events should be expressed
drought in the lowlands to the west of the divide. Another by a relative decrease in the pollen production of the
possibility is that glacier equilibrium lines may in fact also hygrophilous indicator, i.e., Myrtaceae, which would explain
depend on lower temperatures in this region. This would the large influx and percentage variations displayed by these
explain why these changes might go completely undetected in taxa during the period in question.
the coastal areas, where slightly to moderately decreased Records of La Niña events in this system are difficult to
temperatures would have very little impact on vegetation observe because a pollen sample takes several years to record
distributions. (in this case up to 15 yr). It is possible that recurrent la Niña
Paleorecords from Central Chile (Jenny et al., 2002a,b; events that lack a direct pollen indicator may remain obscure
Maldonado and Villagrán, 2002; Villagrán, 1982; Villa- in our record, even assuming that concomitant decreases in
Martínez and Villagrán, 1997; Villa-Martínez et al., 2003) are pollen production and absence of certain taxa, such as the
in good agreement with the inferences made from the last geophytes, did occur. Nevertheless, both pollen types, as well
portion of the Palo Colorado pollen record, all of which indicate as the majority of the other records previously cited (Jenny et
some degree of anthropogenic disturbance. Such influence is al., 2002a,b; Moy et al., 2002; Riedinger et al., 2002; Rodbell
suggested here by the decrease in swamp arboreal taxa and an et al., 1999; Villa-Martínez et al., 2003), are indicative
increase in Maytenus boaria and non-arboreal taxa (i.e., exclusively of El Niño and not La Niña events, which
Poaceae, Asteraceae and Brassicaceae). questions whether both events ( El Niño and La Niña ) had a
Many authors have recently emphasized the importance of high frequency during this period, or if only El Niño events
ENSO events in the Holocene (Jenny et al., 2002a,b; Marchant increased in frequency.
et al., 1999; Moy et al., 2002; Riedinger et al., 2002; Rodbell et
al., 1999; Sandweiss et al., 1996, 2001b). Most authors agree Conclusions
that little or no ENSO influence was present during the early to
middle Holocene (Jenny et al., 2002b; Marchant et al., 1999; The palynological record from the Palo Colorado swamp
Moy et al., 2002; Riedinger et al., 2002; Rodbell et al., 1999). forest enables us to infer the following paleoclimatic event
These authors disagree, however, when ENSO actually began; sequence for Central Chile:
some propose 5700–5800 cal yr BP (Jenny et al., 2002b; A brief wet phase during the early Holocene (between
Sandweiss et al., 2001), whereas others suggest a slow gradual ∼10,000 and ∼8700 cal yr BP), characterized by swamp
increase in ENSO periodicity from 7000 cal yr BP (Moy et al., forest indicators, was followed by a prolonged arid phase
2002; Riedinger et al., 2002; Rodbell et al., 1999). Neverthe- when arboreal taxa disappeared and open vegetation taxa
less, authors concur in finding an increase in frequency of dominated (between ∼8700 and 7800 cal yr BP). This
ENSO events beginning at 3200 cal yr BP. Many authors also culminated in an extremely arid phase characterized by
postulate that current ENSO dynamics began at this time absence of pollen and high inorganic sediment content.
(Jenny et al., 2002b; Marchant et al., 1999; Riedinger et al., These climatic conditions agree with previous paleoclimate
2002; Sandweiss et al., 2001; Villa-Martínez et al., 2003). In a studies from Central Chile and Norte Chico, suggesting that
more detailed analysis, Moy et al. (2002) suggested a gradual the westerly wind belt was in all probability displaced
increase in ENSO events with a superimposed variability southwards during times of strengthened anticylconic circula-
signal at millennial time scales during the Holocene. They tion in our study area.
propose that this trend culminated at ∼1200 cal yr BP and has Swamp forest recovery at Palo Colorado began at 5700 cal yr
since declined. Sedimentary records from the Galapagos BP with a non-linear increase in moisture, peaking at 4200 cal
Islands also show a higher frequency of ENSO events between yr BP. Between ∼3000 and 2200 cal yr BP, a less pronounced
∼1950 and ∼930 cal yr BP (Riedinger et al., 2002). Results episode of drought occurred (i.e. less intense than the previous
from climate simulation models studying the evolution of one). Afterwards, highly variable moisture conditions begin-
ENSO during the Holocene also suggest an increase in ning at ∼2200 cal yr BP prevailed. These conditions suggest
frequency between 2000 and 1000 yr BP (Clement et al., that the influence of the westerly wind belt was similar to the
2000). This agrees with our record from Palo Colorado, which present. Between 2200 and 1300 cal yr BP, however,
indicates increased variability in Myrtaceae pollen frequencies precipitation was more variable than the present, suggesting
during the period between 2200 and 1300 cal yr BP. greater ENSO frequency and prevailing wet conditions.
From the evidence presented above, we conclude that the The climatic events described here from the Palo Colorado
variability recorded at Palo Colorado between ∼2200 and 1300 pollen record are consistent with both physical and biological
 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258 257

records of paleoclimate across the region. We therefore propose Laguna del Negro Francisco, northern Chile, and its palaeoclimatic
that the local precipitation record inferred from our dataset is in implications. The Holocene 7, 151–159.
Grosjean, M., Nuñez, L., Cartajena, I., Messerli, B., 1997b. Mid-Holocene
all probability reflecting a regional climatic signal, associated climate and culture change in the Atacama Desert, Northern Chile.
with changes in both the westerly wind belt influence and the Quaternary Research 48, 239–246.
SPA intensity. Grosjean, M., Geyh, M.A., Messerli, B., Schreier, H., Veit, H., 1998. A late-
Holocene (<2600 BP) glacial advance in the south-central Andes (29°S),
northern Chile. The Holocene 8, 473–479.
Acknowledgments
Grosjean, M., Cartajena, I., Geyh, M.A., Nuñez, L., 2003. From proxy data to
paleoclimate interpretation: the mid-Holocene paradoxa of the Atacama
The authors thank Juan Sparwater for his good will and Desert, Northern Chile. Palaeogeography, Palaeoclimatology, Palaeoecol-
facilities to work at Fundo Palo Colorado. Donald Jackson was ogy 194, 247–258.
very supportive and interested during the whole investigation. Grove, M.J., Baker, P.A., Cross, S.L., Rigsby, C.A., Seltzer, G.O., 2003.
We thank Patricio Moreno, Claudio Latorre, Jean Pierre Application of strontium isotopes to understanding the hydrology and
paleohydrology of the Altiplano, Bolivia-Peru. Palaeogeography, Palaeo-
François and Anita Abarzua for assistance with fieldwork. Ek climatology, Palaeoecology 194, 281–297.
del Val, Ruth McBride and Claudio Latorre for assistance with Hervé, F., Encinas, A., Villa-Martínez, R., Sven, N., Finger, K., Peterson, D.,
the English translation. Funding was provided by Grants 2003. Registro sedimentológico de la transgresión marina del Holoceno
Fondecyt No. 3040032; Fondecyt No. 1030585; Beca Doctoral Medio en el área de Algarrobo (33°22′S), Chile Central. Implicaciones
de Conicyt; Beca de Apoyo a la Realización de Tesis Doctoral tectonicas y paleoecológicas. 10° Congreso Geológico Chileno. Universidad
de Concepción, Concepción, Chile.
de Conicyt; Proyecto PG/01/02-Universidad de Chile. Finally, Heusser, C.J., 1990. Ice age vegetation and climate of subtropical Chile.
we thank Dan Sandweiss, Gerard Roe and an anonymous Palaeogeography, Palaeoclimatology, Palaeoecology 80, 107–127.
reviewer for helping to significantly improve this manuscript. Jenny, B., Valero-Garcés, B.L., Urrutia, R., Kelts, K., Veit, H., Appleby, P.G.,
Geyh, M., 2002a. Moisture changes and fluctuations of the Westerlies in
Mediterranean Central Chile during the last 2000 years: the Laguna Aculeo
References record (33°50′S). Quaternary International 87, 3–18.
Jenny, B., Valero-Garcés, B.L., Villa-Martínez, R., Urrutia, R., Geyh, M.A.,
Abbott, M.B., Wolfe, B.B., Wolfe, A.P., Seltzer, G.O., Aravena, R., Mark, B.G., Veit, H., 2002b. Early to mid-holocene aridity in Central Chile and the
Polissar, P.J., Rodbell, D.T., Rowe, H.D., Vuille, M., 2003. Holocene Southern Westerlies: The Laguna Aculeo record (34°S). Quaternary Reseach
paleohydrology and glacial history of the central Andes using multiproxy 58, 160–170.
lake sediment studies. Palaeogeography, Palaeoclimatology, Palaeoecology Kim, J.-H., Schneider, R.R., Hebbeln, D., Müller, P.J., Wefer, G., 2002. Last
194, 123–138. deglacial sea-surface temperature evolution in the Southeast Pacific
Aceituno, P., 1988. On the functioning of the Southern oscillation in the South compared to climate changes on the South American continent. Quaternary
American sector. Part I. Surface climate. Monthly Weather Review 116, Science Reviews 21, 2085–2097.
505–523. Lamy, F., Hebbeln, D., Röhl, U., Wefer, G., 2001. Holocene rainfall variability
Baker, P.A., Seltzer, G.O., Fritz, S.C., Dunbar, R.B., Grove, M.J., Tapia, P.M., in southern Chile: a marine record of latitudinal shifts of the Southern
Cross, S.L., Rowe, H.D., Broda, J.P., 2001. The history of South American Westerlies. Earth and Planetary Science Letters 185, 369–382.
tropical precipitation for the past 25,000 years. Science 291, 640–643. Latorre, C., Betancourt, J.L., Rylander, K.A., Quade, J., Matthei, O., 2003. A
Betancourt, J.L., Latorre, C., Rech, J.A., Quade, J., Rylander, K.A., 2000. A vegetation history from the arid prepuna of northern Chile (22–23°S) over
22,000-year record of monsoon precipitation from Northern Chile's the last 13,500 years. Palaeogeography, Palaeoclimatology, Palaeoecology
Atacama Desert. Science 289, 1542–1546. 194, 223–246.
Carré, M., Bentaleb, I., Fontugne, M., Lavallé, D., 2005. Strong El Niño events Maldonado, A., 1999. “Historia de los bosques pantanosos de la costa de Los
during the early Holocene: stable isotope evidence from Peruvian sea shells. Vilos (IV Region, Chile) durante el Holoceno medio y tardío.” Unpublished
The Holocene 15, 42–47. Magíster thesis, Facultad de Ciencias, Universidad de Chile.
Clement, A.C., Seager, R., Cane, M.A., 2000. Suppression of El Niño during the Maldonado, A., Villagrán, C., 2002. Paleoenvironmental changes in the
mid-Holocene by changes in the Earth's orbit. Paleoceanography 15, semiarid coast of Chile (∼32°S) during the last 6200 cal years inferred
731–737. from a swamp-forest pollen record. Quaternary Research 58, 130–138.
Earle, L.R., Warner, B.G., Aravena, R., 2003. Rapid development of an unusual Maldonado, A., Jackson, D., 2005. Holocene climate change and human
peat-accumulating ecosystem in the Chilean Altiplano. Quaternary Research settlement on the semiarid coast of Chile (32°S). Holocene environmental
59, 2–11. catastrophes in South America: from the lowlands to the Andes. Abstract
Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis. Blackwell Scientific Volume and field guide. Miramar-Córdoba, 11–17 March 2005. Third Joint
Publication. Meeting of ICSU Dark Nature and IGCP 490.
Fuenzalida, P., 1965. Biogeografía. In: CORFO (Ed.), Geografía Económica de Maldonado, A.J., Betancourt, J.L., Latorre, C., Villagrán, C., 2005. Pollen
Chile. Editorial Universitaria, S.A., Santiago, Chile. analyses from a 50,000-Yr rodent midden series in the Southern Atacama
Garreaud, R., Aceituno, P., 2002. Atmospheric circulation over South America: Desert (25°39′S). Journal of Quaternary Science 20, 493–507.
mean features and variability. In: Young, K. (Ed.), The Physical Geography Marchant, M., Hebbeln, D., Wefer, G., 1999. High resolution planktic
of South America. Oxford University Press. foraminiferal record of the last 13,300 years from the upwelling area off
Grimm, E., 1987. CONISS: a fortran 77 program for stratigraphically Chile. Marine Geology 161, 115–128.
constrained cluster analysis by the method of incremental sum of squares. Miller, A., 1976. The climate of Chile. In: Schwerdtfeger, W. (Ed.), Climates of
Computers and Geociences 13–35. Central and South America. Elsevier Scientific Publishing Company,
Grosjean, M., 1994. Paleohydrology of the Laguna Lejia (North Chilean Amsterdam, pp. 113–145.
Altiplano) and climatic implications for late-glacial times. Palaeogeoraphy, Moy, C.M., Seltzer, G.O., Rodbell, D.T., Anderson, D.M., 2002. Variability of
Palaeclimatology, Palaeoecology 109, 89–100. El Niño/Southern oscillation activity at millennial timescales during the
Grosjean, M., 2001. Mid-Holocene climate in the South-Central Andes: humid Holocene epoch. Nature 420, 162–165.
or dry? Science 292, 2391a. Ota, Y., Paskoff, R., 1993. Holocene deposits on the Coast of North-Central
Grosjean, M., Valero-Garcés, B.L., Geyh, M.A., Messerli, B., Schotterer, U., Chile: radiocarbon ages and implications for coastal changes. Revista
Schreier, H., Kelts, K., 1997a. Mid- and late-Holocene limnogeology of Geológica de Chile 20, 25–32.
258 A. Maldonado, C. Villagrán / Quaternary Research 66 (2006) 246–258

Paduano, G.M., Bush, M.B., Baker, P.A., Fritz, S.C., Seltzer, G.O., 2003. A Servant, M., Servant-Vildary, S., 2003. Holocene precipitation and
vegetation and fire history of Lake Titicaca since the Last Glacial atmospheric changes inferred from river paleowetlands in the Bolivian
Maximum. Palaeogeography, Palaeoclimatology, Palaeoecology 194, Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 194,
259–279. 187–206.
Quade, J., Rech, J.A., Betancourt, J.L., Latorre, C., 2001. Mid-Holocene climate Stockmarr, J., 1971. Tablets with spores used in absolute pollen analysis. Pollen
in the South-Central Andes: humid or dry? Response. Science 292, 2391a. et Spores 13, 615–621.
Ramírez, C., Ferriere, F., Figueroa, H., 1983. Estudio fitosociológico de los Tapia, P.M., Fritz, S.C., Baker, P.A., Seltzer, G.O., Dunbar, R.B., 2003. A Late
bosques pantanosos templados del sur de Chile. Revista Chilena de Historia Quaternary diatom record of tropical climatic history from Lake Titicaca
Natural 56, 11–26. (Peru and Bolivia). Palaeogeography, Palaeoclimatology, Palaeoecology
Ramírez, C., San Martín, C., San Martín, J., 1995. Estructura florística de los 194, 139–164.
bosques de Chile sur-central. In: Armesto, J., Villagrán, C., Kalin, M. (Eds.), Trenberth, K.E., 1976. Spatial and temporal variations of the Southern
Ecología de los bosques nativos de Chile. Editorial Universitaria, Santiago, Oscillation. Quarterly Journal of the Royal Meteorological Society 102,
pp. 215–234. 639–653.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C., VanHusen, C., 1967. “Klimagliederung in Chile auf der Basis von Häufigkeits-
Blackwell, P.G., Buck, C.E., Burr, G., Cutler, K.B., Damon, P.E., Edwars, R. verteilungen der Niederschlagssummen”. Freiburger Geographischa Hefte 4,
L., Fairbanks, R.G., Friedrich, M., Guilderson, T.P., Hughen, K.A., Kromer, 0–113. Universitat Freiburg I. BR.
B., McCormac, F.G., Manning, S., Bronk Ramsey, C., Reimer, R.W., Varela, J., 1981. Geología del Cuaternario del área de los Vilos-Ensenada El
Remmele, S., Southon, J.R., Stuiver, M., Talamo, S., Taylor, F.W., van der Negro (IV Región) y su relación con la existencia del bosque ''relicto'' de
Plicht, J., Weyhenmeyer, C.E., 2004. IntCal04 terrestrial radiocarbon age Quebrada Quereo. Congreso Internacional de Zonas Aridas y Semiáridas,
calibration, 0–26 cal kyr BP. Radiocarbon 46, 1029–1058. pp. 17–30. 33, La Serena.
Riedinger, M.A., Steinitz-Kannan, M., Last, W.M., Brenner, M., 2002. A 6100 Veit, H., 1996. Southern Westerlies during the Holocene deduced from
14C yr record of El Niño activity from the Galápagos Islands. Journal of geomorphological and pedological studies in the Norte Chico, Northern
Paleolimnology 27, 1–7. Chile (27–33°S). Palaeogeography, Palaeoclimatology, Palaeoecology 123,
Rigsby, C.A., Baker, P.A., Aldenderfer, M.S., 2003. Fluvial history of the Rio 107–119.
Ilave valley, Peru, and its relationship to climate and human history. Villagrán, C., 1982. Estructura floristica e historia del bosque pantanoso de
Palaeogeography, Palaeoclimatology, Palaeoecology 194, 165–185. Quintero (Chile, V Región) y su relación con las comunidades relictuales
Rodbell, D.T., Seltzer, G.O., Anderson, D.M., Abbott, M.B., Enfield, D.B., de Chile Central y Norte Chico. Actas del III Congreso Geológico
Newman, J.H., 1999. An ∼ 15,000-year record of El Niño-Driven 377–402.
Alluviation in Southwestern Ecuador. Science 283, 516–520. Villagrán, C., Varela, J., 1990. Palynological evidence for increased aridity on
San Martín, J., Troncoso, A., Ramírez, C., 1988. Estudio fitosociológico de los the Central Chilean Coast during the Holocene. Quaternary Research 34,
bosques pantanosos nativos de la Cordillera de la Costa en Chile Central. 198–207.
Bosque 9, 17–33. Villa-Martínez, R., Villagrán, C., 1997. Historia de la vegetación de bosques
Sandweiss, D.H., 2003. Terminal Pleistocene through Mid-Holocene archaeo- pantanosos de la costa de Chile central durante el Holoceno medio y tardío.
logical sites as paleoclimatic archives for the Peruvian coast. Palaeogeo- Revista Chilena de Historia Natural 70, 391–401.
graphy, Palaeoclimatology, Palaeoecology 194, 23–40. Villa-Martínez, R., Villagrán, C., Jenny, B., 2003. The last 7500 cal yr BP
Sandweiss, D.H., Richardson III, J.B., Reitz, E.J., Rollins, H.B., Maasch, K.A., of westerly rainfall in Central Chile inferred from a high-resolution
1996. Geoarchaeological evidence from Peru for a 5000 Years BP Onset of pollen record from Laguna Aculeo (34°S). Quaternary Reseach 60,
El Niño. Science 273, 1531–1533. 284–293.
Sandweiss, D.H., Maasch, K.A., Burger, R.L., Richardson III, J.B., Rollins, H.B., Villa-Martínez, R., Villagrán, C., Jenny, B., 2004. Pollen evidence for late-
Clement, A., 2001. Variation in Holocene El Niño frequencies: Climate Holocene climatic variability at Laguna de Aculeo, Central Chile (lat. 34°S).
records and cultural consequences in ancient Peru. Geology 29, 603–606. The Holocene 14, 361–367.

View publication stats