Dendrochronology

Dendroclimatology
Progress and Prospects
Edited by
Malcolm K. Hughes
University of Arizona, Tucson, AZ, USA
Thomas W. Swetnam
University of Arizona, Tucson, AZ, USA
Henry F. Diaz
University of Colorado, Boulder, CO, USA
13
Part III
Reconstruction of Climate Patterns and
Values Relative to Today’s Climate
Chapter 7
Dendroclimatology from Regional to
Continental Scales: Understanding Regional
Processes to Reconstruct Large-Scale Climatic
Variations Across the Western Americas
Ricardo Villalba, Brian H. Luckman, Jose Boninsegna, Rosanne D. D’Arrigo,

Antonio Lara, Jose Villanueva-Diaz, Mariano Masiokas, Jaime Argollo,
Claudia Soliz, Carlos LeQuesne, David W. Stahle, Fidel Roig, Juan Carlos
Aravena, Malcolm K. Hughes, Gregory Wiles, Gordon Jacoby, Peter
Hartsough, Robert J.S. Wilson, Emma Watson, Edward R. Cook, Julian
Cerano-Paredes, Matthew Therrell, Malcolm Cleaveland, Mariano S.
Morales, Nicholas E. Graham, Jorge Moya, Jeanette Pacajes, Guillermina
Massacchesi, Franco Biondi, Rocio Urrutia, and Guillermo Martinez Pastur
Abstract Common patterns of climatic variability across the Western Americas

are modulated by tropical and extra-tropical oscillatory modes operating at different
temporal scales. Interannual climatic variations in the tropics and subtropics of the
Western Americas are largely regulated by El Niño-Southern Oscillation (ENSO),
whereas decadal-scale variations are induced by long-term Pacific modes of cli-
mate variability such as the Pacific Decadal Oscillation (PDO). At higher latitudes,
climate variations are dominated by oscillations in the Annular Modes (the Arctic
and Antarctic Oscillations) which show both interannual and longer-scale temporal
oscillations. Here we use a recently-developed network of tree-ring chronologies
to document past climatic variations along the length of the Western Cordilleras.
The local and regional characterization of the relationships between climate and
tree-growth provide the basis to compare climatic variations in temperature- and
precipitation-sensitive records in the Western Americas over the past 3–4 centuries.
Upper-elevation records from tree-ring sites in the Gulf of Alaska and Patagonia
reveal the occurrence of concurrent decade-scale oscillations in temperature dur-
ing the last 400 years modulated by PDO. The most recent fluctuation from the
cold- to the warm-phase of the PDO in the mid 1970s induced marked changes in
R. Villalba (B)
Departamento de Dendrocronología e Historia Ambiental, Instituto Argentino de Nivología,
Glaciología y Ciencias Ambientales (IANIGLA), CONICET, CC 330, 5500, Mendoza, Argentina
e-mail: [email protected]
M.K. Hughes et al. (eds.), Dendroclimatology, Developments in Paleoenvironmental 175

Research 11, DOI 10.1007/978-1-4020-5725-0_7,

C Springer Science+Business Media B.V. 2011
176 R. Villalba et al.
tree growth in most extratropical temperature-sensitive chronologies in the Western

Cordilleras of both Hemispheres. Common patterns of interannual variations in
tree-ring chronologies from the relatively-dry subtropics in western North and
South America are largely modulated by ENSO. We used an independent recon-
struction of Niño-3 sea surface temperature (SST) to document relationships to
tree growth in the southwestern US, the Bolivian Altiplano and Central Chile and
also to show strong correlations between these regions. These results further doc-
ument the strong influence of SSTs in the tropical Pacific as a common forcing
of precipitation variations in the subtropical Western America during the past 3–4
centuries. Common patterns of interdecadal or longer-scale variability in tree-ring
chronologies from the subarctic and subantarctic regions also suggest common forc-
ings for the annular modes of high-latitude climate variability. A clear separation
of the relative influence of tropical versus high-latitude modes of variability is
currently difficult to establish: discriminating between tropical and extra-tropical
influences on tree growth still remains elusive, particularly in subtropical and tem-
perate regions along our transect. We still need independent reconstructions of
tropical and polar modes of climate variability to gain insight into past forcing inter-
actions and the combined effect on climates of the Western Americas. Finally, we
also include a series of brief examples (as ‘boxes’) illustrating some of the major
regional developments in dendrochronology over this global transect in the last
10 years.
Keywords Dendrochronology · Regional scale · Continental scale · Climate

variations · Americas
7.1 Introduction
Instrumental records show that the climate system is characterized by low- and high-
latitude patterns or modes of variability such as the El Niño/Southern Oscillation
(ENSO) in the equatorial Pacific and the Arctic (AO) and Antarctic (AAO)
Oscillations in the extratropics. The Pacific and high-latitude atmospheric circu-
lation features associated with interannual to decadal variability of climate over
the Americas exhibit large spatial and temporal variance that remains poorly doc-
umented. The resulting regional climate variability has enormous socioeconomic
impacts, as was vividly demonstrated by the disastrous flooding in Paraguay and
eastern Argentina, and the extended drought and massive wildfires in the south-
western United States and Mexico during the 1997–1998 El Niño event. At decadal
scales, the prolonged shift in sea surface temperature (SST) patterns over the north
and south Pacific Ocean after 1976 (Graham 1994) has resulted in ocean and atmo-
spheric changes that have caused costly changes in commercial fish populations
in the eastern north Pacific (Mantua and Hare 2002; Chavez et al. 2003; Beamish
et al. 2004) and a greatly reduced carrying capacity for commercially important
7 Dendroclimatology from Regional to Continental Scales 177
Patagonian grasslands. These coherent interhemispheric changes in annual and

decadal climate patterns associated with the Pacific Decadal Oscillation (PDO)
appear to have been driven by fundamental changes in the hydrologic cycle of the
tropical Pacific Ocean (Graham 1994; Evans et al. 2001a, b; Villalba et al. 2001).
Hemispheric-scale networks of instrumental and proxy climate data are needed
to document and help understand these changes in the ocean-atmosphere system and
their impact on the Americas.
Substantial recent effort has been devoted to the development of ocean-
atmospheric monitoring arrays in the tropical Pacific (e.g., TOGA/TAO,
TOPEX/POSEIDON; Wallace et al. 1998). The cost of these arrays has already
been justified by the economic benefits provided by the long-lead climate forecast-
ing associated with recent ENSO warm events. However, there is clear instrumental
and paleoclimatic evidence that, for example, the frequency of warm and cold ENSO
events has been subject to substantial changes over the past several centuries. The
available instrumental meteorological records are simply too short to clearly define
the important temporal and spatial modes inherent in the low-frequency dynamics
of the Pacific and high-latitude major circulation systems. As these decade-scale
changes in atmospheric circulation have strong impacts on regional climates and
society, understanding these phenomena will improve the skill of long-range cli-
mate forecasting. There is increasing evidence (e.g., Gershunov and Barnett 1998)
that they modulate the character of high-frequency ENSO teleconnections, produc-
ing more extreme and more predictable anomaly patterns when the two systems are
in phase.
The Western America Cordilleras provide a contiguous latitudinal transect of
mountainous terrain flanking the world’s largest ocean that invites comparative stud-
ies of climate variations along the Americas. The American Cordilleras can provide
high-quality proxy climate records over most of their lengths. Tree rings provide the
most broadly distributed, annually resolved source of proxy climate data through-
out the Cordillera and thereby supply the comprehensive baseline data necessary to
evaluate natural climate variability on different temporal and spatial scales.
Progress in dendroclimatology across the Americas has been concerned with the
geographical expansion of the research from the local to regional and continen-
tal scales. The work of Harold Fritts (1976, 1991) and co-workers in the 1970s
represented the first attempt to reconstruct the patterns of spatial variation in tem-
perature, precipitation, and atmospheric pressure across North America and the
Pacific Ocean, based on 65 ring width chronologies from the western United States.
Collaborative work between several research groups during the past 10 years (Meko
et al. 1993; Cook et al. 1999, 2004) have extended this methodology by compiling
835 chronologies from Canada, the United States, and Mexico to reconstruct a grid-
ded network of 297 summer Palmer Drought Severity Indices (PDSIs) across North
America that spans the past 500–600 years (or longer) over much of the grid (Cook
et al. 2004). Following these initiatives, new projects have continued to develop
databases of tree-ring chronologies that cover large areas in the Western Americas
(Fig. 7.1).
Fig. 7.1 Cross-section from the South to North Pole across the Western American Cordilleras
showing changes in elevation with latitude, the approximate location of the mean annual zero
degree (0◦ C) line, and the distribution of upper-elevation tree-ring chronologies. Major tree taxa
used for developing the chronologies are also indicated
One of the Collaborative Research Networks supported by the Inter-American

Institute for Global Change Research (IAI) was focused around the develop-
ment of tree line chronologies from Alaska to Tierra del Fuego (Luckman and
Boninsegna 2001). Using this research as a framework, we discuss in this chapter
some of the most significant developments in tree-ring research across the western
Americas, reviewing local- and regional-scale studies and how they contribute to
our understanding of present and past variations in the circulation modes of climate
variability at continental and interhemispheric scales. The Western Cordilleras of
the Americas runs transverse to the generally latitudinal organization of the major
climate-ocean circulation systems, and therefore past variations in the major modes
of general circulation dynamics linked to El Niño/Southern Oscillation, Pacific
Decadal Oscillation, the Arctic and Antarctic Oscillations can be investigated by
using tree-ring records from this global-scale transect.
7.2 Oscillatory Modes of Climate Variability Across the Western

Cordilleras
Instrumental records show that the climate system is characterized by low- and
high-latitude patterns or modes of variability. These dominant modes of climate
variability fluctuate at many different temporal scales. The best known is the El
Niño-Southern Oscillation phenomenon in the tropical Pacific, which dominates
global climate variations on interannual timescales, mostly ranging from 3 to 6
years (Wallace et al. 1998). On longer than interannual timescales, the dominant
climate pattern in the Pacific Ocean has an ENSO-like spatial distribution of sur-
face temperature and atmospheric circulation and has been identified as the Pacific
Decadal Oscillation in the extratropical north Pacific, the Pacific Interdecadal Mode
in the whole Pacific basin, and as the Global Residual (GR) index on a global scale
(Mantua et al. 1997; Garreaud and Battisti 1999; Enfield and Mestas-Nuñez 2000).
Decadal variability in the climate of the Atlantic basin has also been identified
(Deser and Blackmon 1993), but its interhemispheric climate effects on the Western
Cordilleras are less well known.
The Arctic and Antarctic Oscillations are the dominant modes of climate vari-
ability at the highest latitudes in both hemispheres. The positive state of these
annular modes is associated with intensified subtropical highs and strong polar lows,
which drive a strong extratropical circulation. They also exhibit short- and long-term
modes of variability.
7.2.1 El Niño/Southern Oscillation (ENSO)

Modern interannual variations in the Pacific basin and their interhemispheric effects
on the Western Cordilleras have been extensively documented. The pattern of sea
surface temperature associated with ENSO, measured as the SST anomalies from
6◦ N to 6◦ S, 180◦ to 90◦ W (the Cold Tongue [CT] index of Deser and Wallace
1990), are indicated by correlations mapped in Fig. 7.2. Positive correlations indi-
cate regions where the ocean is warmer when the index is positive. As can be
expected by the nature of the CT index, the strongest correlation with SST occurs in
the tropical Pacific. The subtropical north and south Pacific Oceans are dominated
by anomalies out of phase with the ones occurring in the tropical Pacific, forming a
symmetric pattern about the equator.
The continental effects of the ENSO-related climate variations in terms of sur-
face air temperature (SAT) and precipitation along the Western Cordilleras are
remarkably symmetric about the equator (Fig. 7.3a,b). Positive temperatures in
the tropical Pacific are associated with deeper than normal Aleutian lows, and
the resulting steep north-south gradient in the middle latitudes brings storms and
precipitation to the southwestern and southeastern parts of North America. In a
rough parallel to the circulation changes in the Northern Hemisphere, the steeper
gradient in pressure over the southeastern Pacific related to El Niño events, corre-
sponds to deflections of the low-pressure systems and the associated storms towards
the subtropical belt of South America, increasing precipitation in central Chile
(Fig. 7.3b). Positive CT indices are associated with warmer than normal surface
conditions all along the American Cordilleras (Fig. 7.3a). El Niño brings cool
temperatures to the southeastern United States and to the eastern Amazon basin
(Dettinger et al. 2001).
7.2.2 Pacific Interdecadal Mode
Using instrumental records, several studies have reported Pacific decadal-scale

oscillatory modes (Trenberth and Hurrell 1994; Mantua et al. 1997; Zhang et al.
Fig. 7.2 Correlation coefficients between annual-averaged sea surface temperatures and (upper)
Cold Tongue (CT) index (1903–1990) and (lower) Global Residual (GR) index (1903–1990). The
contour interval is 0.2, dashed where negative. Correlations greater than +0.2, or less than –0.2,
pass a two-tail Student’s t-test of being different from zero at 95% significance levels (modified
from Dettinger et al. 2001)
1997; Garreaud and Battisti 1999; Deser et al. 2004). The physical processes
responsible for the decadal variability across the Pacific remain uncertain, but are
connected to well-documented pan-Pacific changes in the atmosphere and ocean
in recent decades. For example, the 1976–1977 climatic shift influenced climatic
conditions all along the western Americas and is a remarkable manifestation of
this Pacific decade-scale climatic variability (Ebbesmeyer et al. 1991). Sea surface
temperatures along the equatorial belt and along the coast of the Americas become
warmer, while further west at temperate latitudes the sea surface becomes cooler
(Fig. 7.2). The array of atmospheric and oceanic changes that have been linked
to these basin-wide regime shifts is collectively referred to as the Pacific Decadal
Oscillation or the Pacific Interdecadal Mode (Mantua et al. 1997; Enfield and
Mestas-Nuñez 2000). Warm and wet decades in the equatorial Pacific tend to be
marked by extratropical circulation patterns that bring mild weather conditions to
coastal Alaska and northern Patagonia. In contrast to the interannual mode of ENSO
variability, the decadal mode is characterized by less pronounced anomalies in the
Fig. 7.3 Regression coefficients (B) estimated during the interval 1904–1990, relating Cold
Tongue (CT) index to October–September (a) surface air temperatures, and (b) precipitation.
Figures (c) and (d) are same as (a) and (b) but for the Global Residual (GR) index. Radii of circles
are proportional to the magnitude of regression coefficients: red and light blue, respectively, for
positive and negative relations with surface air temperatures; green and light brown, respectively,
for positive and negative relations with precipitation. The circles, lower left in each diagram, indi-
cate the scale of influences. Temperatures from an updated version of the monthly, 5◦ × 5◦ -gridded
temperature anomaly set of Jones et al. (1986a, b), and land precipitation anomalies on a similar
grid from Eischeid et al. (1991) were compared with CT and GR. Regression coefficients may be
affected by the magnitude of the variable used in the analysis. The lack of significant regression
coefficients between CT and precipitation in the central Andes along the South American Pacific
coastline is likely due to the reduced precipitation across this region (modified from Dettinger et al.
2001)
eastern Pacific (the classic key ENSO region) and is not narrowly confined along the
equator. The documented decadal oscillatory mode of Pacific SST shows anomalies
in the western Pacific that extend to the northeast and southeast into the American
subtropics.
Overall, the atmospheric expressions of the ENSO-like climate variations on
both interannual and decadal timescales are remarkably symmetric about the equa-
tor, especially on the Pacific coast of western Americas (Fig. 7.3c,d). Positive
variations in the CT and GR are associated with equatorward diversions of the west-
erlies, enhancement of the low-pressure systems, and storms from the midlatitude
Pacific basin toward North and South America subtropical latitudes (Dettinger et al.
2001).
7.2.3 Annular Modes
The Northern Hemisphere (NAM) and Southern Hemisphere (SAM) Annular

Modes dominate extratropical climate variability throughout their respective hemi-
spheres (Thompson and Wallace 2000). The NAM is alternatively referred to
as the North Atlantic Oscillation (NAO) (Hurrell and van Loon 1997) and the
Arctic Oscillation (Thompson and Wallace 2000); whereas the SAM is alternatively
referred to as the High-Latitude Mode (Karoly 1990) and the Antarctic Oscillation
(Gong and Wang 1999; Thompson and Wallace 2000).The structures of the Northern
Hemisphere and Southern Hemisphere Annular Modes are shown to be remarkably
similar, not only in the zonally averaged geopotential height and zonal wind fields,
but in the mean meridional circulations as well. Both annular modes are associated
with equivalent barotropic vacillations in the strength of the zonal flow between cen-
ters of action located at ~35◦ −40◦ and ~55◦ −60◦ latitude. The SAM is moderately
symmetric about the pole, but due to the more complex distribution of the northern
continents, the NAM is more evident over the north Atlantic and the north Pacific
Oceans (Fig. 7.4).
Periods when the zonal flow along ~55◦ −60◦ latitude is anomalously westerly
(the so-called high-index polarity of the annular modes) are characterized by lower
than normal geopotential heights and temperatures over the polar cap, and by higher
than normal geopotential heights and temperatures in the middle latitudes centered
at ~45◦ . As atmospheric variability in the Northern Hemisphere is largest in winter,
the spatial pattern of the conventional NAM mostly reflects the winter variability
(Thompson and Wallace 2000). The positive polarity of the winter AO is associ-
ated with positive surface air temperature anomalies throughout the high latitudes
of Eurasia and much of North America, and negative anomalies over extreme eastern
Canada, North Africa, and the Middle East. This zonally asymmetric pattern of SAT
anomalies is evident throughout the year except during the Northern Hemisphere
summer months (Thompson and Wallace 2000). The leading mode of the Empirical
Orthogonal Function (EOF) in summer months has a smaller meridional scale than
the conventional NAM. Associated summertime low-level temperature anomalies
show more extended warm anomalies over the midlatitudes than the winter NAM
Fig. 7.4 High-latitude annular modes. The Southern Annular Mode (SAM) or Antarctic
Oscillation (AAO) mode is moderately symmetric about the pole (left), but due to the more
complex distribution of the northern continents, the Northern Annular Mode (NAM) or Arctic
Oscillation (AO) is more evident over the North Atlantic and the north Pacific Oceans. Zonal-mean
geopotential height fields are represented with contour intervals of 5 m (modified from Thompson
and Wallace 2000)
counterpart, especially over Europe, the Sea of Okhotsk, and northern America. For
example, the summer NAM pattern accounts for many of the anomalous weather
features observed during the summer of 2003. Temperature anomalies over north-
western Eurasia, northeastern Siberia, and Canada during that period exceeded 3◦ C
(Ogi et al. 2004).
The positive polarity of the Southern Hemisphere Annular Mode is associated
with cold anomalies over most of Antarctica. The one notable exception is the
Antarctic Peninsula and southern South America, where the enhanced westerlies
related to the high SAM polarity increase the advection of relatively warm oceanic
air over the lands (Thompson and Solomon 2002). The observed trend in the SAM
toward stronger circumpolar flow is in the same sense as the trends that have domi-
nated the Northern Hemisphere extratropical circulation over the past few decades.
The occurrence of positive trends in both the NAM and SAM suggests that the trends
reflect processes that transcend the high-latitude climate of a particular hemisphere.
7.3 Tree-Ring Records Across the Western Americas
A major result of the Collaborative Research Network has been the consolidation
and expansion of tree-ring collections across the traditional research regions of
North and South America, the focusing on key areas, and the start of many devel-
opments in new regions of Canada, Mexico, Peru, Bolivia, Chile, and Argentina
(Boxes 7.1, 7.2, 7.3, 7.6 and 7.8). Along the western coasts of North and South
Box 7.1 Climate signals in Gulf of Alaska tree-ring records
A network of climatically sensitive tree-ring records has been compiled

for the Gulf of Alaska (GOA) region, and the records have been used to
develop time series of temperatures over the past one to two millennia (Box
Fig. 7.1). This region is strongly sensitive to the climatic effects of the
Pacific Decadal Oscillation (PDO), and the GOA chronologies show evidence
for decadal-scale regime shifts, including the noteworthy 1976 transition in
Pacific climate. These records have been linked to sea surface temperature
(SST) variations in the Pacific and have been included in reconstructions of
the PDO (e.g., D’Arrigo et al. 2001). On longer timescales, century to mil-
lennial temperature variations are evident and are linked to glacial changes in
southern Alaska.
Box Fig. 7.1 Map of North Pacific region showing locations of tree-ring sites (yellow)
used as candidate predictors of the tropical Indo-Pacific climate index (NPI). Sites in red
are those included in the regression model used to reconstruct the NPI. Some dots represent
more than one site (D’Arrigo et al. 2005)
Analyses of instrumental data demonstrate robust linkages between

decadal-scale North Pacific and tropical Indo-Pacific climatic variability, yet
information on the tropical–high-latitude climate connection is limited prior
to the twentieth century. Gulf of Alaska and western Canadian tree-ring
records were used to reconstruct the December–May North Pacific index
(NPI—an index of the atmospheric circulation related to the Aleutian low-
pressure cell) from 1600 to 1983 (D’Arrigo et al. 2005; Box Fig. 7.2). This
NPI reconstruction shows evidence for the climatic regime shifts seen in the
instrumental NPI data, and for additional events in prior centuries. It correlates
significantly with both instrumental tropical climate indices and a coral-based
reconstruction of an optimal tropical Indo-Pacific climate index (OTI), sup-

porting evidence for a tropical/North Pacific link extending as far west as the
western Indian Ocean. The coral-based reconstruction (1781–1993) shows the
twentieth-century regime shifts evident in the instrumental NPI and OTI, as
well as previous shifts. Changes in the strength of the correlation between
the NPI and OTI reconstructions over time, and the timing of regime shifts in
both series prior to the twentieth century, suggest a varying tropical influence
on North Pacific climate, with greater influence in the twentieth century. One
likely mechanism is the low-frequency variability of the El Niño/Southern
Oscillation (ENSO) and its varying impact on Indo-Pacific climate.
—R. D’Arrigo, G. Wiles, and R. Wilson
Box Fig. 7.2 Tree-ring-based reconstruction of the tropical Indo-Pacific climate index
(NPI): (a) actual and estimated December–May NPI for the 1900–1983 calibration period,
adf = adjusted degrees of freedom; (b) reconstruction of the December–May NPI from AD
1600 through 1983 based on North Pacific tree-ring data. The highlighted phase shifts were
identified by using intervention analysis (significant at the 90% confidence level; D’Arrigo
et al. 2005)
America, there is a gradual environmental gradient from the relatively dry-warm

subtropics to the wet-cold high latitudes. Tree-ring records from subtropical regions,
such as the southwestern United States and central Chile, are remarkably sensitive
to precipitation variations (Boninsegna 1988; Cook et al. 2004; LeQuesne et al.
2006). In the transitional zones to higher latitudes, tree-ring responses to climate
are largely determined by site conditions. Depending on elevation, aspect, slope,
and soil characteristics, tree growth can be influenced by temperature, precipita-
tion, or more commonly by a combination of both. In the extreme wet and cold
environments at high-elevation or high-latitude upper tree lines, temperature is the
major limiting factor controlling tree growth (Wiles et al. 1996; Luckman et al.
1997; Wiles et al. 1998; Aravena et al. 2002; Villalba et al. 2003; Lara et al.
2005). These changes in tree response with latitude were instrumental in setting the
strategies for selecting tree-ring records sensitive to temperature and precipitation
variations along the western Americas. Temperature reconstructions based on upper-
elevation chronologies on mountains near the coasts around the Gulf of Alaska and
northern Patagonia were selected as proxy records of temperature for North and
South America, respectively. Tree-ring records from mesic to dry environments in
the southern-central United States, the Bolivian Altiplano, and central Chile were
used for the interhemispheric comparison of precipitation-sensitive records across
the American Cordilleras (Fig. 7.1). The available data and maturity of dendrocli-
matological research differ considerably between regions, and therefore the kind
of comparison between regional records and forcings will be different across the
north-south transect.
7.3.1 Temperature-Sensitive Records
The strong cross-equatorial symmetries of SST, continental temperature, and con-

tinental precipitation patterns documented from instrumental records motivated
the search in high-resolution proxy records for common spatial patterns of cli-
mate variability across the western Americas during the past centuries. Have the
patterns of climate variability documented during the instrumental period been
recurrent in previous centuries? Were these patterns different during the domi-
nantly cooler conditions during the first half of the nineteenth century? Answers
to these questions can provide useful information on the stationary nature of cli-
mate variations and how they could change under different global atmospheric
conditions.
Box 7.2 Studies from the Canadian Cordillera
During the last 5 years, significant progress has been made in dendrochrono-
logical and dendroclimatic studies in the Canadian Cordillera (49◦ −65◦ N),
building on limited earlier collections. Sampling has targeted temperature-
sensitive sites at altitudinal tree line (Picea engelmanni, P. glauca, Larix
lyallii, Pinus albicaulis, and Abies lasiocarpa) and moisture-sensitive sites
at the lower forest border (Pseudotsuga menziesii and Pinus ponderosa),
mainly using a network approach to isolate regional rather than local signals.
Although initially focused on the southern Cordillera (ca. 125 chronologies),
over 100 new sites have been sampled in the Yukon over the last 5 years.
Studies at tree line in the Coast Ranges of British Columbia and Vancouver
Island have developed several single- and multiple-species chronology net-
works (Tsuga mertensiana, T. heterophylla, and Chamaecyparis nootkatensis)
that include sites with the potential for millennial-length reconstructions
(see http://geog.uvic.ca/dept/uvtrl/uvtrl.htm). In the southern Cordillera, the
network of low-elevation, moisture-sensitive sites has been used to recon-
struct spatial patterns of precipitation and drought over the last three to four
centuries (Watson and Luckman 2004a, 2005), and these data have been incor-
porated into the new gridded Palmer Drought Severity Index (PDSI) network
developed by Cook et al. (2004).
New ring width and density data have been used to revise and extend a
millennial-length (950–1994) summer temperature record from the Canadian
Rockies (Box Fig. 7.3). Comparison with adjacent areas (e.g., Wiles et al.
2004) and global Northern Hemisphere curves suggests this is a regionally
representative record. The influence of Pacific-forced decadal-scale variabil-
ity in this record is more subtle, but the low-frequency signal suggests solar
forcing has been an important control of summer temperature patterns in this
region.
The use of multispecies networks allows the combination of temperature-
and/or precipitation-sensitive chronologies to investigate climate-related phe-
nomena that are influenced by the combined variation of temperature and
precipitation. Box Figure 7.4 shows a reconstruction of glacier mass balance
using independent tree-ring-derived summer and winter balances. Although
winter balance (precipitation input) is strongly controlled by atmospheric
circulation patterns from the Pacific, summer balance (mass loss through
melt) is driven primarily by solar radiation. The major periods of positive
net balance reflect a combination of higher winter inputs and cooler sum-
mers rather than summer temperatures alone. Future work can adopt similar
approaches to the reconstruction of streamflow and other climate-related
variables.
—B.H. Luckman, R.J.S. Wilson, and E. Watson
Box Fig. 7.3 Maximum May–August temperatures at the Columbia Icefield, Canadian
Rockies. Temperatures are anomalies based on 1901–1980 means smoothed with a 30-year
filter. Vertical bars represent significant sunspot minima (Luckman and Wilson 2005)
Box Fig. 7.4 Reconstructed net mass balance for Peyto Glacier, Alberta 1673–1994
(Watson and Luckman 2004b)
7.3.1.1 Extratropical Pacific Ocean

We use temperature reconstructions from coastal Gulf of Alaska and northern
Patagonia to investigate past changes in the decadal oscillatory modes across the
Pacific domain (Boxes 7.1 and 7.3). Wiles et al. (1998) presented a well-verified
reconstruction of spring (MAM, March–May) temperature variations, based on
three ring-width chronologies from coastal sites along the Gulf of Alaska, dat-
ing from 1600 to 1988. This reconstruction explains 34% of the variance in the
instrumental temperature data. The decade-long variations in this reconstruction
are consistent with changes in the Aleutian low-pressure system, which in turn
is affected by ENSO (Dettinger et al. 2001). Spectral analysis of the temperature
reconstruction shows significant peaks consistent with the ENSO-like bandwidth
(7, 8, 11, and 19 years). The reconstructed spring temperature series suggests that
the recent warming exceeds temperature levels of prior centuries, extending back
to AD 1600 (Wiles at al. 1998). The three coldest intervals in the spring series
occurred in the seventeenth century. This cooling is consistent with the glacial
record from coastal Alaska, which shows a strong advance during the late seven-
teenth to mid-eighteenth centuries (Wiles and Calkin 1994; Wiles 1997; Wiles et al.
2004).
A critical appraisal of surface air temperature from station records has recently
been presented for southern South America (Villalba et al. 2003). Two different
spatial temperature patterns were recognized in the southern Andes during the twen-
tieth century: (1) surface cooling from 1930 to 1976 at the stations located in the
northern sector of the southern Andes by the Pacific Coast (37◦ −42◦ S), and (2) a
remarkable surface warming in the southern stations (south of 46◦ S), which inten-
sifies at higher latitudes. Changes in the Pacific Decadal Mode around 1976 were
seen in summer temperature records at most stations in the Pacific domain, start-
ing a period with increased temperature across the southern Andes and at higher
latitudes. Tree-ring records from upper tree line were used to reconstruct past tem-
perature fluctuations for the two dominant patterns over the southern Andes. The
resulting reconstructions for the northern and southern sectors of the southern Andes
explain 55% and 45%, respectively, of the temperature variance over the interval
1930–1989. Cross-spectral analysis of actual and reconstructed temperatures over
the common interval 1930–1989, indicates that most of the explained variance is
at periods >10 years in length. Consequently, these reconstructions are especially
useful for studying multidecadal temperature variations in the South American sec-
tor of the Southern Hemisphere over the past 360 years. These reconstructions show
that temperatures during the twentieth century have been anomalously warm across
the southern Andes. The mean annual temperatures for the northern and southern
sectors during the interval 1900–1990 are 0.53◦ C and 0.86◦ C above the 1640–1899
means, respectively (Villalba et al. 2003).
Box 7.3 Climate signals in Patagonian upper-elevation

tree-ring records
A great deal of progress has been made in increasing the number of upper-
elevation tree-ring chronologies across the southern Andes during the past
decade. This work has involved the development of more than 90 chronolo-
gies from collections of Nothofagus pumilio, the dominant subalpine tree in
the Andes of Chile and Argentina (Villalba et al. 1997; Lara et al. 2001;
Aravena et al. 2002; Villalba et al. 2003; Lara et al. 2005). These new col-
lections have increased both the spatial coverage (ca. 35◦ 35 to 55◦ S) and
the temporal span of upper-elevation records across the southern Andes (Box
Fig. 7.5). The broad latitudinal distribution of N. pumilio across 2000 km in
a north-south direction provides the opportunity to examine the relationships

between N. pumilio growth and climate along a temperature gradient from the
subtropical central Andes to the sub-Antarctic Tierra del Fuego.
Box Fig. 7.5 Nothofagus pumilio chronologies in the Patagonian Andes
Nothofagus chronologies from upper tree line have been used to recon-
struct past temperature fluctuations for the northern and southern sectors of
the southern Andes. The reconstructions describe a well-defined cold interval
from ~1640 to 1850, which conforms with the consensus view of the ‘Little
Ice Age’ (LIA), a term commonly used to describe these cold episodes on a
global scale (Bradley and Jones 1992).
Relationships between temperature reconstructions in southern South
America and sea surface temperatures (SSTs) in the South Pacific and South
Atlantic Oceans clearly show that the temperature reconstructions contain

information about climate variability extending over much of the tropical-
subtropical Pacific and over the south Atlantic to Africa. The correlation
fields between the reconstructions and SST (Box Fig. 7.6) are reminiscent of
some of the global modes of SST recently derived from instrumental records.
The spatial amplitudes obtained by correlating the northern Patagonian recon-
struction with SSTs closely resemble the Southern Hemisphere counterpart of
the interdecadal mode of the Pacific SST variability identified by Garreaud
and Batistti (1999) and Enfield and Mestas-Nuñez (2000). Consistent with
the documented decadal oscillatory mode of Pacific SST, the spatial field
of correlations is characterized by anomalies in the western Pacific that
extent to the southeast into subtropical South America. The spatial pattern
that results from comparing the southern Andes reconstruction and SSTs
resembles the ‘global warming’ mode identified by Enfield and Mestas-
Nuñez (2000). According to these authors, the ‘global warming’ mode is
the ocean counterpart to the global warming seen in surface air temperatures
(SATs).
—R. Villalba, A. Lara, and M. Masiokas
Box Fig. 7.6 Spatial correlation patterns (1857–1989) between sea surface temperature
(SST) anomalies over the south Pacific and south Atlantic Oceans and the temperature
reconstructions for the northern and southern sectors of the southern Patagonian Andes
The Gulf of Alaska and southern Andes reconstructions clearly show the well-
documented transition from cold to warm conditions over the tropical Pacific in
1976 and are consistent with regional temperature compilations. This result reflects
a comparable sensitivity of the temperature records to SST changes in the Pacific
Ocean during recent decades. If decadal timescale variations in climate forced by
the tropical Pacific had also affected temperature changes in the past, tree-ring-based
reconstructions of temperature along the coast of North and South America should
present similar oscillatory patterns. Indeed, for the common interval 1640–1989,
reconstructed temperature variations from the Gulf of Alaska are significantly cor-
related with those of northern (r = 0.42, p < 0.01; Fig. 7.5) and southern Patagonia
(r = 0.38, p < 0.01). Spatial patterns obtained by correlating the Alaska and north-
ern Patagonia temperature reconstructions with SSTs across the Pacific and Atlantic
Fig. 7.5 This figure compares temperature-sensitive tree-ring records (red triangles) from high-
latitude, western North and South America with a geochemical coral record (yellow triangle) from
Raratonga, in the tropical South Pacific during the past three to four centuries. The series shown
from top to bottom are: spring/summer Gulf of Alaska temperature reconstruction (1600–1994;
Wiles et al. 1998), Sr/Ca coral record from Rarotonga (1726–1996; Linsley et al. 2004), and annual
northern Patagonia temperature reconstruction (1641–1989; Villalba et al. 2003). Correlation coef-
ficients between records are indicated. To facilitate the comparison, the Sr/Ca coral record is shown
inverted
Oceans (Villalba et al. 2001) closely resemble those observed for the decadal mode
of Pacific SST variability identified by Zhang et al. (1997) and Garreaud and Battisti
(1999). Temperature anomalies related to ENSO-like variations are larger and more
spatially consistent in northern than in southern Patagonia (Fig. 7.3c), reflecting the
decrease in correlation between Alaskan and Patagonian records with increasing
southern latitudes.
The Gulf of Alaska and northern Patagonia temperature reconstructions are dis-
played in Fig. 7.6, along with the waveforms of the two oscillatory modes that are
Fig. 7.6 Comparison of temperature reconstructions from northern Patagonia (blue line) and
coastal Alaska (red line) and their dominant oscillations isolated by using singular spectrum anal-
ysis (SSA; panel a; Vautard 1995). Common oscillatory modes in both records have periods of (b)
> 30 years, and (c) 9–10 years. Percentages of the original variance contributed by Patagonian and
Alaskan waveforms are indicated in the upper and lower left corners, respectively. The Pearson’s
correlation coefficient, r, between the series, is shown in the lower far right. Time series included
in (d) represent the sum of the oscillations shown in (b) and (c)
the major contributors to the common variance between these records. Waveforms
were extracted from the original reconstructions by using singular spectrum anal-
ysis (SSA), basically a statistical technique related to EOF analysis, to determine
oscillatory modes in the time domain (Vautard and Ghil 1989). The reconstructed
waveforms, representing oscillations >30 years and approximately 10 years, reveal
interesting changes in amplitude during the past 350 years. As was previously noted
(Villalba et al. 2001), the temporal evolution of these components is more closely
related in amplitude and intensity from 1640 to approximately 1850. After 1850,
relationships between waveforms are weaker. The most remarkable feature in the
long-term oscillations is the positive amplitudes during the past 100 years, reflecting
the warming in the twentieth century.
7.3.1.2 Tropical Pacific Ocean

Although the temperature reconstructions from the Gulf of Alaska and northern
Patagonia provide insight into the temporal evolution of the relationships between
the tropical ocean and higher latitudes in the Americas, it is important to note that
the variability in the records is related to tropical teleconnections along the west-
ern coasts of the Americas and not to direct forcing from the equatorial Pacific.
In a first attempt to connect the extratropical tree-ring records from North and
South America with climate variability in the tropical Pacific, we compared the
temperature reconstructions with high-resolution coral records in the Pacific Ocean.
Long-lived corals provide continuous, high-resolution records of tropical Pacific
climate that supplement the instrumental record of climate from this key region.
Modern coral records from the central tropical Pacific are several centuries in length
and have yielded insights into the recent history of tropical Pacific climate variabil-
ity on a variety of timescales. The δ18 O and Sr/Ca time series from corals in the
subtropical Pacific at Rarotonga (21◦ 14 S and 159◦ 49 W), have recently been com-
pared with indices of climate variability in the north Pacific, suggesting some degree
of cross-hemispheric symmetry of interdecadal oceanographic variability in the past
centuries (Linsley et al. 2004). A tree-ring reconstruction of the north Pacific index,
a measure of the intensity of the large-scale atmospheric circulation related to the
Aleutian low-pressure cell, correlates significantly during the twentieth century with
both instrumental tropical climate indices and a coral-based reconstruction of an
optimal tropical index for the Indian and Pacific Oceans, supporting evidence for
a tropical/north Pacific link that extends as far west as the western Indian Ocean
(D’Arrigo et al. 2005).
The coral skeletal Sr/Ca at Rarotonga appears to be related to SST variability on
annual through at least decadal timescales based on correlation with instrumental
SST (Linsley et al. 2004). The coral record, which covers the period 1726–1997,
is significantly correlated with both the Gulf of Alaska (r = −0.32, p < 0.01)
and northern Patagonia (r = −0.34, p < 0.01) temperature reconstructions
(Fig. 7.5). The low-pass fraction for each time series was isolated by using SSA
(Vautard 1995), and all reconstructed components with mean frequencies longer
than 20 years were summed. The results are shown in Fig. 7.7. The subtropical
Fig. 7.7 Interdecadal to centennial variability in temperature-sensitive series from Gulf of Alaska
(red line), northern Patagonia (blue line), and Raratonga (brown line), isolated by using singular
spectrum analysis (SSA; Vautard 1995). For each record, all SSA-reconstructed components with
mean frequencies longer than 20 years were summed. Thin and thick arrows indicate coincidences
in oscillations between the Raratonga and one or two high-latitude records, respectively
Pacific records indicate that some of the interdecadal transitions in coral Sr/Ca
temporally align with comparable transitions in the Gulf of Alaska and north-
ern Patagonia temperature reconstructions. The remarkable shift in tropical Pacific
climate during the mid-1970s is clearly captured by all three records. However,
some differences are observed between interdecadal oscillations in the subtropical
coral and the North and South American tree-ring records. Interdecadal tempera-
ture oscillations in northern Patagonia closely align with transitions in the Pacific
coral Sr/Ca records from the 1850s to the beginning of the twentieth century,
whereas the Gulf of Alaska oscillations align better with Rarotonga Sr/Ca during
the second half of the twentieth century.
7.3.1.3 High-Latitude Oscillations

As was indicated in Section 7.2.3, temperature variations in high latitudes of
the Northern and Southern Hemispheres are also related to changes in the NAM
and SAM, respectively (Thompson and Wallace 2000; Thompson and Solomon
2002). We search for common patterns in temperature variations in the sub-Artic
and sub-Antarctic regions, which in turn might provide insight on common forc-
ings of high-latitude past climates in both hemispheres. Boreal tree-ring records
from high latitudes in the Northern Hemisphere were used to provide a long-term
perspective of Arctic annual temperatures (D’Arrigo and Jacoby 1993). The recon-
struction was based on 12 chronologies from North America: 3 in Alaska north
of 67◦ N, 4 in northwestern-central Canada from the Yukon to Churchill, and 5
in eastern Canada. This sub-Arctic network was complemented with five boreal
Fig. 7.8 Temperature reconstructions from Arctic and sub-Antarctic regions. The geographical
locations of tree-ring chronologies (red triangles) used for developing the temperature reconstruc-
tions for the Arctic (left) and sub-Antarctic (right) regions are shown. See text for reconstruction
details
tree-ring chronologies from Scandinavia (67◦ −69◦ N) and three from the northern
Ural Mountains (Fig. 7.8). The total variance in temperature variations explained
by the tree-ring chronologies during the 1880–1969 calibration period is 66%. The
major low-frequency trends in the reconstructed Arctic temperatures include a cool-
ing in the late 1600s to early 1700s, a relative warming in the 1700s, an abrupt
decline in temperature in the early 1800s, a gradual warming since the middle to
late 1800s, and unprecedented warming during the twentieth century. Recently, a
new reconstruction of temperature variability for the Arctic has been developed
with significantly improved geographical coverage and replication than previously
(Gordon Jacoby, in preparation). The new temperature record reproduces most
climatic events previously reconstructed, reinforcing the occurrence of major tem-
perature changes in the sub-Arctic during the past four centuries. For comparison
with the sub-Antarctic temperatures, the two reconstructions were averaged in a
single Arctic temperature record.
The northern latitude record was compared with the temperature reconstructions
for northern and southern Patagonia (Villalba et al. 2003, Fig. 8). For the common
interval 1670–1987, the correlation coefficient between the Arctic and sub-Antarctic
(average of the two southern reconstructions) is r = 0.55 (p < 0.001). For the
past 400 years, striking similarities in temperature fluctuations are observed in both
regions. The records exhibit their largest common variances at low frequencies
Fig. 7.9 Comparison of the amplitudes from the first principal components of the temperature
reconstructions from Patagonia (blue line) and the Arctic (red line). The Patagonian and the Arctic
records were obtained by averaging the temperature reconstructions shown in Fig. 7.8. Common
oscillatory modes in both records have periods of (b) > 100 years, and (c) around 36 years. Time
series included in (d) represent the sum of the oscillations shown in (b) and (c) (for explanation of
the data in each panel of this figure see Fig. 7.6)
(Fig. 7.9). In both records, positive levels during twentieth-century periods exceed
values back to 1670. An abrupt decrease in temperature in both regions is recorded
in the 1810s, quite likely related to a series of large tropical volcanic eruptions,
including an unknown source in 1809, Soufriere in 1812, and Tambora in 1815,
among others (Zielinski 2000). A notable feature of temperature change revealed
by the high-latitude records is the continuous transition from anomalous cold
conditions in the mid-nineteenth century to anomalous warm conditions in the
mid-twentieth century. In contrast, the global and hemispheric mean instrumental
temperatures show almost no trend between the late 1850s and the 1910s (Jones and
Moberg 2003), suggesting that high latitudes in both hemispheres share common
patterns of temperature changes that are not seen at global scales.
7.3.2 Precipitation-Sensitive Records

Tree-ring chronologies from precipitation-sensitive regions across the western
Americas, such as the southern United States, the Bolivian Altiplano, and central
Chile, reveal common interannual to decadal-scale oscillations in precipitation vari-
ations during the past centuries (Boxes 7.4, 7.5, 7.6, 7.7 and 7.8). Spatial correlation
patterns between precipitation-sensitive records and SST also show that variations
in these records are strongly connected with SST anomalies in the equatorial Pacific
and off the western coast of the subtropical Americas (Villalba et al. 2001).
Box 7.4 Spatial patterns of drought and wetness regimes

over western North America
The network of moisture-sensitive tree-ring chronologies now available for

North America has been used to reconstruct the summer (June–July–August,
JJA) Palmer Drought Severity Index (PDSI) for 1200 years on 286 grid
points extending from southern Mexico across the United States into southern
Canada (Cook et al. 2004). These reconstructions have high temporal and spa-
tial fidelity when compared with instrumental PDSIs on annual and decadal
timescales. The average of all 286 grid points for North America indicates
that the driest single year in the past 500 years occurred in 1864, and the
wettest single year occurred in 1833 (Box Figs. 7.7 and 7.8). The reconstruc-
tions indicate that the twentieth century was relatively moist compared with
the past 500 years, the severe Dust Bowl and 1950s droughts notwithstanding
(Box Fig. 7.7).
The long-range climate influence of the El Niño/Southern Oscillation
(ENSO) over western North America during the late nineteenth and twen-
tieth centuries has been strongest in the Texas-Mexican sector of northern
Mexico and the southwestern United States, with drought during La Niña
events and wetness during El Niño events. The epicenter of reconstructed
decadal drought was often located in the ENSO teleconnection province over
the Southwest, implicating the tropical Pacific in these decadal dry regimes.
The pluvials of the past 500 years were spatially heterogeneous and did not
tend to recur in the ENSO teleconnection region. The notable exception was
the early twentieth-century pluvial (Box Fig. 7.7), one of most extremely
wet decades in 500 years, and which was concentrated in the drainage basin
of the Colorado River. This period of exceptional wetness inflated expec-
tations of surface water supplies in the Southwest, and provides a modern
demonstration of the significant environmental and socioeconomic impacts
associated with these decadal droughts and pluvials. The seventeenth-century
Pueblo Drought lasted at least six years over the same region impacted by
the twentieth-century pluvial (Box Fig. 7.7), and provides a compelling con-
trast to the pluvial and a strong analog for the recent multiyear drought
that has severely impacted surface water supplies in the Southwest (1999–
2006). Socioeconomically, the seventeenth-century Pueblo Drought caused
starvation, death, and the permanent abandonment of five Pueblo communities
and other villages in New Mexico.
—D.W. Stahle and E.R. Cook
Box Fig. 7.7 North American summer Palmer Drought Severity Index (PDSI), a time series
average of reconstructed PDSI from all 286 grid points over North America
Box Fig. 7.8 Reconstructed summer Palmer Drought Severity Index (PDSI) for the wettest
(1833) and driest (1864) single years in the ‘North America summer PDSI’ series (Box
Fig. 7.7), showing the continental scale of these record moisture anomalies
Box 7.5 Western United States droughts in medieval times

linked to changes over the Pacific basin
Box Fig. 7.9 Reconstructed Nevada division 3 July–June precipitation (smoothed with 50-
year Gaussian filter). Gray line marks the transition at ~AD 1400
Several multicentury shifts in precipitation over the central Great Basin

in the United States are seen in an ~8000-year reconstruction from the
bristlecone pine chronology at Methuselah Walk, and in an ~1800-year recon-
struction based on this and five other chronologies (Hughes and Funkhouser
1998). A remarkable, but not unique, transition from drier to wet conditions
is reconstructed between the periods AD 400–1400 and AD 1400–2000 (Box
Fig. 7.9). We set out to find an explanation for this transition (Graham et al.
2007). Proxy evidence from tree-ring and pollen-based reconstructions, and
ocean core isotopic data suggest that the circa 1400 transition was marked by
warming sea surface temperatures (SSTs) along the central California coast
(Kennett and Kennett 2000; Box Fig. 7.10, plot labeled Santa Barbara Basin
SST), and increasing winter precipitation with cooler summer temperatures
from southern and central California into the Great Basin (Box Fig. 7.10, top
plot). In today’s climate, such changes are associated with El Niño episodes,
suggesting the possibility that El Niño-like changes in tropical Pacific SSTs
may have played a causal role in producing the mid-latitude changes sug-
gested by the proxy records. A coral-based Niño-3.4 SST reconstruction from
Palmyra Atoll in the central tropical Pacific (Box Fig. 7.10, plot 5; Cobb et al.
2003 and a foram Mg/Ca based SST reconstruction from near Mindanao in
the northwest equatorial Pacific (Box Fig. 7.10, middle plot; Stott et al. 2004)
support the idea of a trend towards more El Niño–like conditions at the circa
1400 transition. Analysis of the proportion of terrestrial material in a marine
core taken off the coast of central Peru indicates a contemporaneous increase
in river discharge associated with high-flow events (Rein et al. 2004). This
picture would be consistent with the Pacific-wide pattern proposed here.

Hence, it is suggested that the proxy-inferred warming of California coastal
SSTs and increasing western US precipitation around, roughly, AD 1400
resulted (at least in part) from increasing tropical Pacific SSTs and resulting
changes in tropical precipitation patterns. The question now arises whether
other transitions in the Nevada record (for example, around 400 BC, and in
the converse direction around AD 400) have similar causes.
—Nicholas E. Graham and Malcolm K. Hughes
Box Fig. 7.10 Proxy records from areas in the equatorial Pacific Ocean and its eastern mar-
gin (western North America) since 500 BC. Top Plot: extent of Mono Lake low stands (Stine
1994) green horizontal lines, and a long tree-ring chronology from the White Mountains of
California. Middle plot: Mg/Ca-based SST reconstruction from foraminifera near Mindanao
in the northwest equatorial Pacific. Scale inverted for comparison. Bottom Plot: yearly
values and period means (horizontal blue lines) of sea surface temperature (SST) (After
Graham et al. 2007)
In 1999, a gridded network of 154 Palmer Drought Severity Index reconstruc-

tions for the continental United States was generated from a set of 388 tree-ring
chronologies (Cook et al. 1999). More recently, the spatial and temporal coverage
of the PDSI reconstructions was expanded, including 286 points in a 2.5◦ × 2.5◦
grid covering most of North America (Cook et al. 2004; Box 7.4). The new PDSI
reconstructions are based on an expanded network of 835 tree-ring chronolo-
gies. The temporal coverage was also expanded to the maximum permitted by
the available tree-ring data, extending back nearly 2000 years for some locations.
Finally, the process of variance restoration applied to the grid point reconstruc-
tions allows for updates of those records to AD 2003 with instrumental PDSI
data. In a previous work (Villalba et al. 2001), the temporal evolution of the
PDSI in four cells located in the midwestern-southwestern United States were
compared with precipitation-sensitive chronologies from central Chile. For the com-
mon interval 1700–1978, the correlation coefficient between the first Principal
Component (PC) from the four PDSI reconstructions and tree-ring variations at
El Asiento, central Chile, is r = 0.32 (p < 0.001), which was considered as
an indication of common modes of variations in these series. With the increas-
ing number of PDSI reconstructions across North America, the use of a spatial
approach to study large-scale atmospheric variations connecting mid- to high-
latitude precipitation changes in North and South America now appears to be
feasible.
A major advance in the effort to expand the spatial coverage of tree-ring
records across the Americas has been the recent development of Polylepis tara-
pacana chronologies in the Bolivian Altiplano (Argollo et al. 2004). These records,
located between 17◦ and 20◦ S and above 4500 m elevation, represent the closest-to-
equator tree rings in the Andes and the highest-elevation chronologies worldwide
(Box 7.7). A careful examination of interannual variations in ring width and
climate in the Altiplano indicate that the growth of Polylepis is remarkably asso-
ciated with summer water balance. Most Polylepis records cover the past three
to four centuries, but some of them extend over seven centuries. Two of the
longest chronologies (Caquella and Soniquera) were merged in a single record and
used for comparison with precipitation-sensitive records in other regions of the
western Americas. For central Chile, a ring-width chronology from Austrocedrus
chilensis D. Don at El Asiento (32º40 S), which represents the northernmost
extent of this species in central Chile, was used for comparison with North
American records (Box 7.9). Recently, the site was revisited, and series from the
new cores were merged with the original data collected in 1974 by LaMarche
(1975).
Box 7.6 A network of tree-ring chronologies for northern

and central Mexico
The mixed conifer forests and riparian areas of northern and central Mexico
contain some of the most climate-sensitive species in the North American
region (Box Fig. 7.11). Douglas-fir (Pseudotsuga menziesii) is one of these
species that has been widely recognized due to its sensitivity to climate (Fritts
1976). Douglas-fir has a native latitudinal range covering at least 38◦ in the
Northern Hemisphere and extending well into southern Mexico at latitudes
below 17◦ N. In Mexico, Douglas-fir occurs in scattered insolated popula-
tions of mixed conifer forests located at high elevations, thriving in cool
microenvironments and scarped terrains of the Sierras Madre Occidental
and Oriental (Martínez 1963). The annual ring of this species is anatom-
ically divided into two distinct layers: the earlywood (EW) is composed
of low-density, light-colored cells, whereas the latewood (LW) has smaller,
darker cells with thicker walls. The development of separate EW and LW
chronologies provides more information on the influence of intra-annual cli-
mate variability than using total ring width data (Cleaveland 1986; Stahle et al.
1998).
Box Fig. 7.11 Tree-ring chronologies in Mexico

Ahuehuete or Montezuma bald cypress (Taxodium mucronatum), a widely

distributed riparian species that is considered to be the national tree of Mexico,
has been used to develop the longest tree-ring chronologies in Mexico. In
recent years, over 30 new tree-ring chronologies have been developed or
are in process as part of the CRN03-IAI project. The Douglas-fir chronolo-
gies range between 129 and 604 years, bald cypress between 117 and 1550
years, and pinyon pine (Pinus cembroides) over 400 years. Some available
Douglas-fir and bald cypress chronologies are being extended with the use
of cross sections from subfossil wood or logged material. The EW growth
of Douglas-fir trees in Mexico is influenced by dominant climatic condi-
tions in the winter–spring period previous to growth (November to current
June; Box Fig. 7.12), explaining around 70% of the variance in growth. Cool
season precipitation in northern Mexico is increased by the warm phase of
El Niño/Southern Oscillation (ENSO). On the other hand, the LW width is
affected by the summer precipitation and the monsoon system. Annual ring
width of bald cypress is influenced by the seasonal late spring to summer pre-
cipitation (June–September), explaining 52% of the variance in growth. The
regions of northern and central Mexico have highly limited water resources,
and paleoclimatic reconstructions are essential to understand the hydrocli-
matic variability that characterizes these regions. Determining past climatic
variability is essential to planning proper management strategies for water use,
and tree-ring studies offer the best opportunity to understand this variability
over the last 1000 years.
—J. Villanueva-Diaz, J. Cerano-Paredes, D. Stahle, M. Therrell, M.
Cleaveland, and B.H. Luckman
Box Fig. 7.12 The winter–spring precipitation series reconstructed from an earlywood
Douglas-fir chronology in Chihuahua, Mexico. The reconstruction covers the period
1472–2002
Box 7.7 The Polylepis tarapacana chronologies: The highest

elevation tree-ring records worldwide
Polylepis, a genus from the Rosaceae family, includes several woody species
of small- to middle-sized trees that grow at very high altitudes in the tropi-
cal Andes of South America (Kessler 1995). Polylepis tarapacana, adapted
to drier and colder conditions than other species of the same genus, reaches
the highest elevation of tree growth in the world. On the slopes of the high
volcanoes in Bolivia and along the Bolivian-Chilean-Argentinean border, P.
tarapacana grows between 4100 and 5200 m elevation.
Box Fig. 7.13 Polylepis tarapacana chronologies in the Bolivian Altiplano and adja-
cent areas of Chile and Argentina. HUA: Huarinka; SER: Serke; NIC: Cerro Nicolás;
ANA: Analasjchi; NAS: Nasahuento; SAJ: Sajama; GUA: Guallatire; TUN: Tunupa; CAQ:
Caquella; TAP: Tapachilca; SON: Soniquera; UTU: Uturun-co; GRA: Cerro Granadas
Polylepis stands are almost exclusively restricted to volcanic slopes, with a

strong preference for well-insolated north-facing slopes. Extensive collections
of P. tarapacana were conducted in Bolivia, Chile, and Argentina (16◦ –
22◦ S) as part of the IAI-CRN program to reconstruct climate variations from
upper-elevation tree rings along the Americas (Box Fig. 7.13). Presently, the
chronologies range between 98 and 705 years in length, and represent the
highest tree-ring records worldwide (Box Fig. 7.14). In order to determine the
climatic variables controlling P. tarapacana growth, interannual variations in
tree growth were compared with regional records of precipitation and temper-
ature. Correlation functions indicate that the radial growth of P. tarapacana is
influenced by water balance during the summer previous to the ring formation.
At the sampling sites, precipitation explains around 50% of the total vari-
ance in growth. Summer temperatures, which increase evapotranspiration and
reduce soil water supply, are negatively correlated with tree growth (Argollo
et al. 2004).
1.6
Tree-ring Index
1.4
1.2
1.0
0.8
0.6
0.4
1400 1500 1600 1700 1800 1900 2000
Box Fig. 7.14 Composite chronology resulting from merging the Polylepis tarapacana
ring width series from Caquella and Soniquera in the Bolivian Altiplano
Traditionally, the wood of Polylepis tarapacana has been used by local

populations in the Bolivian Andes for construction, particularly for house and
church roofs. Wood from old buildings offers the possibility of extending the
upper-elevation records of P. tarapacana back in time for the past millennium.
These records offer the unique opportunity for reconstructing precipita-
tion variations across the altiplano during the past five to seven centuries
or more.
—Jaime Argollo, Claudia Soliz, Jorge Moya, Janette Pacajes, Mariano S.

Morales, and Ricardo Villalba
7.3.2.1 Subtropical Precipitation and ENSO

Precipitation variations in the United States–Mexico, the Bolivian Altiplano, and
central Chile are related to climatic changes in the tropical Pacific. It is well
known that there is a strong teleconnection between SST changes in the tropi-
cal Pacific and precipitation anomalies in the southern United States and northern
Mexico (Ropelewski and Halpert 1986; Kiladis and Diaz 1989; Cole and Cook
1998). Warmer SSTs in the tropical Pacific typically result in increased precipi-
tation anomalies in this region. Spatial correlations between different indices of
tropical Pacific circulation and the grid point PDSI series over the conterminous
United States show that the geographic location of the highest correlation field is
the southwestern United States (Cook et al. 2000), a finding that is consistent with
the patterns identified by using instrumental records.
Interannual variability in precipitation over the altiplano is primarily related
to changes in the mean zonal flow, reflecting changes in meridional baroclinicity
between tropical and subtropical latitudes, which in turn is a response to sea surface
temperature changes in the tropical Pacific (Garreaud et al. 2003). There is a general
agreement that a significant fraction of interannual variability in summer precipita-
tion is related to ENSO (e.g., Aceituno 1988; Lenters and Cook 1999; Vuille 1999;
Vuille et al. 2000; Garreaud et al. 2003). All of these studies concluded that El Niño
years tend to be dry, whereas La Niña years are often associated with wet conditions
on the altiplano. However, dry La Niña years and wet El Niño years are not com-
pletely uncommon, which indicates that the relationship between SST anomalies in
the tropical Pacific and precipitation in the central Andes is not simple (Garreaud
et al. 2003). Finally, relationships between SSTs in the equatorial Pacific and precip-
itation anomalies in central Chile (30º−35ºS) have been reported by several authors
(Quinn and Neal 1983; Aceituno 1988; Rutlland and Fuenzalida 1991; Aceituno
and Montecinos 1996; Montecinos and Aceituno 2003). Positive rainfall anomalies
in central Chile are associated with warmer SSTs in the tropical Pacific. Conversely,
cold SSTs correspond quite closely to dry conditions in the area.
To gain insights into the long-term relationships between SST in the tropical
Pacific and precipitation in the southwestern United States, the Bolivian Altiplano,
and central Chile, we compared precipitation-sensitive records from these three
regions with a multiproxy-based reconstruction of SST for the El Niño-3 region
(Mann et al. 2000). Through exploiting the complementary information shared by
a wide network of different types of proxy climate indicators, the multiproxy El
Niño-3 reconstruction reduces the weaknesses in any individual type or location
of indicator and makes use of the mutual strength of the diversity in the records.
The reconstructed eastern equatorial Pacific Niño-3 areal-mean SST index has been
previously used as a direct indication of ENSO itself for the past 400 years (Mann
et al. 2000). A large proportion of the tree-ring chronologies from the southwestern
United States and Mexico have been used as predictors of both the Niño-3 index and
PDSI reconstructions, which make the reconstructions not statistically independent.
In contrast, neither the El Asiento nor the Bolivian Altiplano chronologies have been
included in the Niño-3 index reconstruction.
Box 7.8 Tree-ring chronologies from Austrocedrus chilensis

in central Chile
Austrocedrus chilensis (D. Don, Serr et Bizz.) is the most northerly-distributed
conifer species of the Andean Patagonian forests. The species occurs within a
wide latitudinal range between 32º39 and 43º40 S. The northernmost pop-
ulations of the species also occupy the tree line at high elevations in the
Andes of central Chile. These populations are growing on steep, rocky slopes
under severe water stress in low-density, scattered stands. The trees from
these marginal stands exceed 1200 years in age and exhibited typical fea-
tures of long-lived species, like strip bark growth, twisted branches, and crown
dieback. According to Edmund Schulman (1956), Austrocedrus ‘was the most
suitable dendrochronologic species in the southern Andes. Its ring record is as
well defined as any in the drought conifers of the Rocky Mountains, and it
possesses the type of cambial growth regime which leads to good crossdating
quality in the ring series.’
Box Fig. 7.15 Austrocedrus chilensis chronologies in central Chile. ELA: El Asiento;
SGB: San Gabriel; RCL: Río Clarillo; URO: Urriloa Oeste; URE: Urriola Este; ELB: El
Baule; AMU: Agua de la Muerte
Several tree-ring collections of Austrocedrus chilensis were taken in cen-

tral Chile during the early years of dendrochronological studies in South
America. The El Asiento site (ELA, Box Fig. 7.15) was first visited by
Valmore LaMarche in 1972. After that, several collections were conducted in
San Gabriel, Río Clarillo, Urriola Oeste, Urriola Este, El Baule, and Agua de
la Muerte. In recent years, new collections have been carried out in these sites
to update and extend these chronologies, especially by collecting preserved
relict wood (Box Fig. 7.15). Correlation function analyses show a strong cli-
matic signal related to winter–spring precipitation during the previous and
current growing seasons. A composite record consisting of El Asiento and El
Baule chronologies has been used by LeQuesne et al. (2006) to develop new
estimates of June–December precipitation for central Chile extending from
AD 1200 to 2000 (Box Fig. 7.16).
The reconstruction suggests that the decadal variability of precipitation
in central Chile was greater before the twentieth century, with more intense
and prolonged dry and wet episodes. Multiyear drought episodes in the eigh-
teenth, seventeenth, sixteenth, and fourteenth centuries exceed the estimates
of decadal drought during the twentieth century. The reconstruction also indi-
cates an increase in interannual variability after 1850. In fact, the risk of
drought exceeding all thresholds increases dramatically in the reconstructed
precipitation series after 1850, consistent with the drying trends indicated by
selected long instrumental precipitation records.
—Carlos LeQuesne and David Stahle
Box Fig. 7.16 Tree-ring-reconstructed precipitation for central Chile from AD 1200 to
2000. A cubic smoothing spline highlighting multidecadal variability (ca. 25 years, fit for
the period 1205–1995) and the ±1.0 standard deviation thresholds are also plotted
It is important to note that based on instrumental records, the strongest telecon-

nections between precipitation in the southern United States–northern Mexico and
SST in the tropical Pacific have been identified for the winter months (Kiladis and
Diaz 1989; Stahle et al. 1998; Cleaveland et al. 2003), whereas the PSDI reconstruc-
tions are reflecting drought conditions during the summer months. On the other
hand, the Polylepis chronologies in the Bolivian highlands and the Austrocedrus
chronologies in central Chile are sensitive to summer and winter precipitation,
respectively. Despite the limitation imposed by the differences in seasonal-
window responses of trees from different regions to precipitation and the fact
Fig. 7.10 Comparison of precipitation-sensitive tree-ring records across western America and El
Niño/Southern Oscillation. The reconstruction of Niño-3 region temperature (Mann et al. 2000)
was used as a proxy for tropical forcing of precipitation variations. The series used for comparison,
from top to bottom, are: Tree-ring-based Palmer Drought Severity Index (PDSI) reconstructions
from the southwestern United States (SW US) from Cook et al. (2004); a composite Polylepis
record, including the Caquella and Soniquera chronologies from the Bolivian Altiplano (Argollo
et al. 2004); and the El Asiento chronology in central Chile (LeQuesne et al. 2006). Correlation
coefficients between records are indicated. The number of years for the comparisons is 331. The
PDSI and the Niño-3 region reconstructions are not statistically independent. Some Texas-Mexican
chronologies were used as predictors in both reconstructions
that we are making comparisons with an annual Niño-3 index reconstruction, the
precipitation-sensitive records from the three extratropical regions in North and
South America are significantly correlated with the Niño-3 index reconstruction
during the 1650–1980 interval used for comparison (Fig. 7.10). The correlation
coefficients between the PDSI reconstructions in the midwestern-southwestern
United States and the Niño-3 index oscillate between r = 0.20 and r = 0.65.
Although most correlation coefficients are remarkably high, the lack of inde-
pendence between these records makes it difficult to determine the statistical
significance of these relationships.
For the common interval 1650–1980, the first PC of the El Asiento chronology in
year t and t+1 is significantly correlated with the Niño-3 index (r = 0.34, p < 0.01),
which may be considered as a first indication of persistence in the influence of trop-
ical SSTs on precipitation in central Chile. Long-term relationships between SST in
the tropical Pacific and precipitation in the Bolivian Altiplano are also inferred from
the statistically significant correlation between the two independent estimates (r =
−0.34, p < 0.01).
7.3.2.2 Dominant Oscillations in Precipitation Variations

Cross-spectral analysis was used to identify coherent oscillation modes in tropical
Pacific SST and precipitation variations in central Chile and the Bolivian Altiplano.
Coherent oscillations between Niño-3 index and precipitation in central Chile are
observed at 3.5 years, the classic El Niño frequency domain, but also at 20–28 years,
an oscillation likely related to the Pacific Interdecadal Mode (Fig. 7.11). Cross-
spectral analysis of the Niño-3 index and Polylepis records in Bolivia indicates
significant coherence at 2.9, 3.2, 3.8, 8.5–10, and 19 years, a cycle also identified in
the Gulf of Alaska temperature reconstructions (Fig. 7.11).
Following the previous cross-spectral analysis, we proceed to isolate the major
waveforms in the precipitation-sensitive records using singular spectrum analysis
(Vautard and Ghil 1989; Vautard 1995). Two dominant oscillations, representing
modes of common variance at 3.6 and 28 years, were isolated from the Niño-3 and
Fig. 7.11 Coherency spectra between the Niño-3 temperature reconstruction (Mann et al. 2000),
and precipitation-sensitive chronologies in central Chile (upper) and the Bolivian Altiplano (lower)
during the interval 1650–1981. Records are highly coherent at 3–4 years, the classic El Niño
oscillations, but also at decennial-scale wavelengths longer than 10 years. Horizontal broken lines
represent the 95% confidence level for the squared-coherency analyses. The periods are given in
years for each significant coherency peak
Fig. 7.12 Comparison of sea surface temperature (SST) in the tropical Pacific, represented by
Niño-3 temperature reconstruction (Mann et al. 2000; red line), and precipitation variations in
central Chile, inferred from El Asiento chronology (blue line), and significant correlated oscillatory
modes extracted by singular spectrum analysis (SSA). Common oscillatory modes have periods
of (b) 27–28 years, and (c) 3.5–3.6 years. Percentages of the original variance contributed by
each of Niño-3 and El Asiento waveforms are indicated in parentheses at the upper and lower
left corners of the figures, respectively. In the lower right corners, r is the Pearson’s correlation
coefficient between Niño-3 and El Asiento series. Marked increases in the interannual oscillatory
modes centered at 3.6 years are observed in both series since about 1850 (c)
El Asiento chronology (Fig. 7.12). In general, the temporal evolution of the 28-
year component shows similar fluctuations in amplitude and intensity from 1650 to
1850. After that, relationships between the 28-year waveforms are weaker. Starting
around 1850, a marked increase in amplitude is observed in the 3.6-year waveforms
from both the Niño-3 index and El Asiento series. Contrasting patterns in El Niño–
related oscillations before and after 1850 have been noted previously by several
authors (Stahle et al. 1998; Villalba et al. 2001; D’Arrigo et al. 2005).
Singular spectrum analysis of the precipitation-sensitive records from the
Bolivian Altiplano also reveals common oscillatory modes with Niño-3 SST. Two
major temporal patterns, centered at 3 and 4 years, are coherent with similar oscilla-
tions in tropical SSTs during the past four centuries. Similar to previous waveform
comparisons, the common oscillation between the Bolivian and Niño-3 records, cen-
tered at 9.7 years, are more consistent in time and amplitude before 1850 (Fig. 7.13).
Fig. 7.13 Comparison of sea surface temperature (SST) in the tropical Pacific, represented by the
Niño-3 temperature reconstruction (Mann et al. 2000; red line), and precipitation variations in the
Bolivian Altiplano, inferred from Caquella-Soniquera composite chronology (blue line) lagged 1
year (t + 1), and significant correlated oscillatory modes extracted by singular spectrum analysis
(SSA). Common oscillatory modes have periods of (b) 9–10 years, (c) 4 years, and (d) 3 years.
Percentages of the original variance contributed by each of Niño-3 and the Bolivian waveforms are
indicated in parentheses at the upper and lower left corners of the figures, respectively. In the lower
right corners, r is the Pearson’s correlation coefficient between Niño-3 and the Bolivian composite
series
These spectral analyses basically reaffirm the existence of common oscillatory

modes in tropical Pacific SST and precipitation-sensitive chronologies in subtropical
North and South America, but also reveal changes in the stability of teleconnec-
tions over time. Indeed, moving correlations (using a 50-year window) between
the precipitation-sensitive chronologies from South America and the El Niño-3 SST
reconstruction revealed such changes in the temporal stability of the teleconnections
(Fig. 7.14). Significant correlations occur for most of the sub-periods compared;
however, correlation coefficients were not statistically significant during the second
part of the eighteenth century between the Niño-3 SST and both the Bolivian and
Fig. 7.14 Relationships between reconstructed Niño-3 temperature variations and interannual pre-
cipitation variability in the Bolivian Altiplano and Central Chile since 1650. Tropical influences
on precipitation in both regions are evaluated by changes in the moving Pearson correlation coeffi-
cients (blue lines) between reconstructed Niño-3 temperatures and tree-ring width variations from
Polylepis (above) and Austrocedrus (below) plotted on centroids of 50-year intervals. Lines at the
95 and 99% confidence intervals are indicated
central Chile records. This interval is characterized by low amplitudes of the inter-
decadal (27–28 years) oscillations in the Niño-3 and central Chile, and out-of-phase
relationships in the 3- to 4-year oscillations between Niño-3 and both the central
Chile and Bolivian records (Figs. 7.12 and 7.13).
The El Asiento precipitation-sensitive record from central Chile is positively
correlated (r = 0.24 for the 350-year interval 1650–1998) with the first PC from
28 PDSI reconstructions in the midwestern-southwestern United States, the area
most consistently affected by variations in tropical Pacific SST. Warm ENSO events
are related to abundant precipitation in both regions. In contrast, droughts in the
Bolivian Altiplano occur in years of warm ENSO events. Overall, the upper-
elevation chronologies from Bolivia are consistently negatively correlated (r =
−0.16, for the 346-year interval 1650–1998) with the first PC from 28 PDSI
reconstructions from the southwestern United States and northern Mexico.
An additional indication of common ENSO influences on precipitation variations
across the western Americas is provided by similarities in spatial correlation pat-
terns between PDSI and Niño-3 index reconstructions and those resulting from the
comparison between the PDSI reconstructions and precipitation-sensitive records
in South America during the past 400 years. Figure 7.15 shows the teleconnec-
tion map for the Niño-3 index with reconstructed PDSI over the 1650–1980 time
period common to all series. The geographical location of the highest correlation
field is where one would expect based on instrumental and past proxy record anal-
yses (Ropelewski and Halpert 1986; Kiladis and Diaz 1989; Cole and Cook 1998).
Fig. 7.15 Spatial correlation patterns between 297 Palmer Drought Severity Index (PDSI) recon-
structions across North America (Cook et al. 2004) and (a) El Niño-3 region temperature
reconstruction (Mann et al. 2000); (b) tree-ring variations from the precipitation-sensitive El
Asiento Austrocedrus chronology, Central Chile; and (c) composite Polylepis chronology from
Caquella and Soniquera in the Bolivian Altiplano
As was mentioned above, the tree-ring records used in the Niño-3 reconstruc-
tion are not completely independent of those used in the PDSI reconstructions,
but the resulting spatial correlation pattern provides the basis for the search for
commonalities with spatial patterns from comparison with the central Chile and
the Bolivian Altiplano chronologies.
Figure 7.15b,c shows the spatial correlation patterns for the central Chile and
altiplano chronologies with reconstructed PDSI over the 1650–1980 time period
common to Niño-3 index reconstruction. Significant correlations are observed in the
southwestern United States–northern Mexico for both patterns, with the sign being
positive for El Asiento and negative for the Polylepis records. Overall, the spatial
correlation patterns produced by Niño-3 index reconstructions and the precipitation-
sensitive records from South America are similar, indicating that the tree-ring
estimates are capturing the interhemispheric ENSO teleconnection pattern. The sig-
nificant correlations between El Asiento and PDSI are more extensive and penetrate
northeastward into the central and upper Midwest of the US. This extension in the
correlation pattern also has been observed by using instrumental records, particu-
larly during the first decades of the twentieth century (Cole and Cook 1998). In
contrast, the spatial pattern with the Bolivian chronologies is highly concentrated
to the southwestern United States–northern Mexico region, making it remarkably
similar to the pattern based on the Niño-3 index (Fig. 7.15).
7.4 Future Research

The previous results highlight the potentiality of tree rings as reliable sources
of information concerning past climate variations. However, dendroclimatological
studies are confronted by many challenges in the western Americas. New tree-
ring chronologies in subtropical Mexico (Pseudotsuga menziesii, Pinus hartwegii,
Taxodium mucronatum, Pinus cembroides, Pinus lumholtzii), in the high-elevation
tropics of Bolivia and adjacent areas of Chile and Argentina (Polylepis tarapacana),
and subtropical Argentina (Prosopis ferox) allow, for the first time, studies of
annual variability in the subtropics. However, these records are still sparse and
difficult to develop (Box 7.9). On the other hand, the development of climate-
sensitive chronologies from the seasonally dry areas of the high- and lowland
tropics remains a major task in the Americas and the world. It was only during
the first years of the twenty-first century that a number of exploratory studies,
mainly in coastal Peru, have established the basis for building reliable and calendar-
dated chronologies from Bursera graveolens. The chronology, which covers the
interval 1950–2002, shows an ENSO signal with an average recurrence period of
about 5 years (Rodriguez et al. 2005). The search for longer records from tropi-
cal species with clear, identifiable annual rings is still a major challenge in modern
dendroclimatology.
An alternative to the reliance on the presence of annual rings in tropical trees is
provided by isotopic studies. An improved mechanistic understanding of controls
on the oxygen isotope ratio (δ18 O) of alpha cellulose (Roden et al. 2000) and rapid
processing techniques (Brendel et al. 2000) make possible the construction of high-
resolution isotope series from tropical trees that can be analyzed to provide both a
chronology and paleoclimatic information, even in trees lacking annual rings (Evans
and Schrag 2004). The seasonal cycle in δ18 O of tropical montane precipitation,
primarily controlled by the difference in the amount of precipitation between wet
and dry seasons, is reflected in δ18 O of the cellulose of trees. This seasonal rhythm
can then be used for chronological control. Pilot applications of these techniques in
Costa Rica, Peru, and the Amazon show annual isotope cycles as great as 4−6‰,
which permit high-resolution chronological control even in the absence of annual
rings. Interannual variability up to 8‰ in δ18 O cellulose is associated with year-to

year differences in precipitation amounts (Evans and Schrag 2004).
The seasonality in climate and tree growth is an important feature to be con-
sidered in the processes of interhemispheric comparisons. For example, ENSO
matures during the boreal winter (austral summer) when the climatic anomalies set
by the extratropical teleconnections are more marked. However, tree responses vary
according to their location and biology, leading to wide variability between species
in the total amount of the ENSO signal captured by trees during the growing season
concurrent or following the event. There is a need for further investigation of the
basic processes that bring about year-to-year variations in tree-ring properties, cell
dimensions, wood chemistry, wood density, or ring width (Hartsough and Biondi
2003; Roig et al. 2003; see Box 7.9). Long-term monitoring that employs a combi-
nation of weather stations and dendrometers will better characterize both the climate
to which the tree is responding and the season of wood formation, particularly in
tropical trees. A better understanding of the relationship between tree growth and
climate will facilitate the interpretation of the temporal climatic window recorded
by trees, which in turn will result in better, ecophysiologically sound reconstructions
of past climate variations.
Box 7.9 Monitoring of tree growth dynamics to improve

dendroclimatic models
Box Fig. 7.17 Seasonal course of climate related to variations in ring morphology and leaf
phenology of Nothofagus pumilio trees in Tierra del Fuego, as illustrated by the 2002–2003
observational period. Air temperatures are at 1.2 m above ground and soil temperatures are
at 0.25 m underground (the layer with major development of the Nothofagus root system)
Fine-scale monitoring of tree growth dynamics is essential to improve our

understanding of the climatic controls that influence both tree-ring mor-
phology and size. Timing of the ring formation can bring clues to improve
interpretation of the tree-ring growth/climate models. In recent years, we set
up monitoring networks to follow tree growth and obtain in situ weather data
from forest ecosystems in Argentina and Mexico. At Cerro Krund, southern
Tierra del Fuego, Argentina, a network of permanent plots was established
along the altitudinal gradient of Nothofagus pumilio, from the valley bottom
to the tree line. At a given plot, the initial surge of earlywood occurs after
the first formed leaves are fully expanded; later, the cambium becomes most
active and then finally ceases by producing latewood cells toward the end of
summer. The influence of altitude is expressed in a delay of the onset of the
initial growth stages (up to 2 weeks between lower and higher plots) and in
the shortening of the vegetative growth period. At lower elevations the cam-
bium remains active for 3–4 months, compared to only 2–3 months at the
tree line. Rapid changes in temperature act as the environmental catalyst to
force the changes of the leaf phenological phases and the wood morphologi-
cal changes through the ring. Box Figure 7.17 shows a synchrony between a
sustained increase of the maximum air temperatures to values above a thresh-
old of 10◦ C and the beginning of the growth phases. Toward the end of austral
summer, trees slowly end their growth cycle as maximum temperatures fall
below this threshold line.
Box Fig. 7.18 Two seasons of growth are shown in the monitored Pinus hartwegii
stand in southwestern Mexico. Onset of growth is coincident with the increase of spring
temperatures above 3.6ºC. Continuation of growth is dependent on monsoon moisture
In Nevado de Colima, a 3900 m elevation site in central western México,

we have been co-monitoring stem growth and weather on two plots of Pinus
hartwegii. The monitoring network consists of a meteorological station and
two sets of dendrometers installed on pure stands of P. hartwegii located
~100 m below tree line. This species has previously shown promise for pale-
oclimate analysis because of sensitivity to the North American Monsoon.
Co-measuring tree growth and climate over a 3-year period has allowed a
better understanding of the complex response of this species to changing envi-
ronmental conditions. Box Figure 7.18 shows that the onset of spring growth
is coincident with soil temperatures higher than a threshold of around 3.6ºC.
Summer growth, however, is dependent on monsoon precipitation, with high
correlations to precipitation and relative humidity. The pattern of growth mir-
rors the lull in the monsoon in late summer. We also have confirmed the
cessation of growth and shutdown of cambial activity in the winter months at
this high-elevation tropical site. Long-term ecological monitoring has allowed
direct correlation at daily and monthly timescales between tree growth and
weather. It is desirable that the monitoring previously described be continued
over long time periods to better understand the radial growth behaviors.
—F. Roig, P. Hartsough, G Massacchesi, F. Biondi, and G. Martinez Pastur
In South America, the only chronologies that presently extend into the mid-
Holocene are for Fitzroya cupressoides from southern Chile. The Volcan Apagado
(41º35 S, 72º30 W) chronology is now the longest continuous chronology in the
Southern Hemisphere, with a total of 5666 years (Wolodarsky-Franke et al. 2005).
However, much more work is needed to develop multimillennial length chronolo-
gies encompassing most of the Holocene and employing other species in more arid
regions of South America.
As in other areas of the world, it is also possible that anthropogenic activities
may be subtly changing climate-growth relationships in these trees, compounding
the difficulties of isolating a clear climate signal in these records. The recent high
growth rates of Nothofagus at the upper tree line across Patagonia provide a major
piece of evidence to assemble a case for anomalous regional warming in response
to anthropogenic activities (Villalba et al. 2003). While this conclusion may prove
to be a valid interpretation of the data, changes in the efficiency with which water is
used in relation to increased atmospheric CO2 content (fertilization), may also exert
some influences on tree growth.
7.5 Discussion and Conclusions

Precisely dated, annually resolved tree-ring records from numerous sites through-
out the Americas provide the basis to evaluate the changing signatures of tropical
and high-latitude modes of climate variability, their time-evolving patterns, and their
interactions in the past. Based on our present knowledge of climate variations during
the twentieth century, we intended this contribution to provide an interhemispheric
view of interactions between large-scale modes of variability and climate along
the American Cordilleras over several centuries prior to the period of instrumen-
tal observations. The striking symmetries in SSTs across the equatorial Pacific and
in sea level pressure (SLP) patterns at high latitudes in both hemispheres induce
similar patterns of climate variability in widely separated regions of the western
Americas (Dettinger et al. 2001).
Comparisons of fire histories between the southwestern United States and
northern Patagonia over the past several centuries provide additional evidence for
similarities in past climate variations across the extratropical Americas (Kitzberger
et al. 2001). The synchrony of fire regimes in these two distant regions has tenta-
tively been interpreted as a response to decadal-scale changes in ENSO and PDO
activities. For example, a period of decreased fire occurrence in both regions from
about 1780 to 1830 was attributed to decreased amplitude and/or frequency of
ENSO events.
Strong interhemispheric symmetries of ENSO and ENSO-like variations of
Pacific climate on decadal timescales produce similar patterns of temperature vari-
ations in Patagonia and the Gulf of Alaska. However, the comparison of these
extratropical temperature reconstructions with the Raratonga temperature-sensitive
coral record from the tropical Pacific indicates that over the last 200 years,
interdecadal SST variations in the Pacific alternated between times of more geo-
graphically widespread interdecadal changes—such as the shift in the mid-1970s
that was recorded across the entire Pacific basin—and times of less geographically
organized interdecadal changes shared by the tropics and the north or south Pacific
Ocean. Our observations that interdecadal variations in SST in the tropical Pacific
were more strongly connected to the north Pacific in the twentieth century than in
the mid-1800s agree with previous studies by Evans et al. (2001a) and Labeyrie
et al. (2003).
In addition to tropical forcings, similarities in decadal- to century-scale climate
variations also result from changes in high-latitude modes of climate variability
that simultaneously affect the extratropical regions of North and South America.
Instrumental records show a simultaneous intensification of the AO and AAO during
recent decades (Thompson and Solomon 2002).
As was shown in Section 7.3.3, the annular modes are strongly coupled with
surface air temperatures over high latitudes in both hemispheres (Thompson and
Wallace 2000). Sustained positive trends in both modes in recent decades may
be linked to large-scale warming, particularly in Eurasia and northern Canada
in the Northern Hemisphere and across Patagonia in the Southern Hemisphere
(Thompson and Wallace 2000; Thompson and Solomon 2002; Ogi et al. 2004).
D’Arrigo et al. (2003) presented a first reconstruction of a warm season AO tem-
perature index during the interval 1650–1975. Values during the middle twentieth
century, overlapping with the anthropogenic increase in trace gases, equal or exceed
those in the prior record. Lower values are reconstructed for several colder periods,
including the early nineteenth-century interval. Trends in the AO temperature index
resemble those of the tree-ring reconstructions of Arctic mean annual temperatures

(r = 0.47, n = 305). According to D’Arrigo et al. (2003), the similarities between
these records reflect some data overlap, but also the strong linkage between the AO
and Arctic temperatures. Similarly, instrumental observations support a strong rela-
tionship between AAO and temperature in southern South America during the past
50 years, suggesting that long-term trends in temperature across Patagonia during
the past centuries have also been influenced by changes in the AAO (Thompson and
Solomon 2002).
As the reconstructions from high latitudes in both hemispheres indicate large
similarities in temperature changes for at least the past 300 years, we could infer
some common forcings of annular modes of climate variability in both hemispheres
during the past centuries. Presumably, any climate change mechanism that projects
onto the meridional temperature gradient between the middle and high latitudes
may affect the polarity of the annular modes. Recent trends in tropical sea surface
temperatures have been shown to affect the NAM (Hoerling et al. 2001), and it has
been hypothesized (but not yet demonstrated) that a similar link may exist for the
SAM (Hurrell and van Loon 1994).
Interactions between low- and high-latitude circulation modes are difficult to
document based on the current array of climate-sensitive chronologies and forcing
reconstructions. For example, interannual and decadal modes of tropical climatic
variability such as ENSO and PDO strongly affect weather conditions in low lati-
tudes of the western Americas, but their influences on the climate of the extratropics
have been noted here and widely documented elsewhere. On the other hand, the
annular modes have large amplitudes at extratropical latitudes, but several studies
reveal that they have a substantial signature at lower latitudes as well (Thompson
and Lorenz 2004).
Reflection symmetries about the equator in interdecadal ocean temperature dur-
ing the past four centuries suggest that the tropical Pacific has played a pivotal
role in linking westerly winds in both hemispheres through meridional telecon-
nections (Zhang et al. 1997; White and Cayan 1998). As has been indicated here,
temperature-sensitive records from the Gulf of Alaska and northern Patagonian
are coherent on long-term (>10 years) oscillatory modes, largely in response to
ENSO-like decadal to interdecadal modes of variability in the Pacific Ocean.
However, instrumental records show that high-latitude forcings of climate variabil-
ity also affect both regions. For example, recent trends in Alaskan climate are better
explained by the juxtaposition of ENSO-like and AO-related SLP variability over
the north Pacific.
The ‘ENSO-like’ interdecadal variability, as documented in Trenberth and
Hurrell (1994) and Zhang et al. (1997), has contributed to SLP declines over the
north Pacific in conjunction with warming over western Canada and Alaska. Apart
from that feature, SLP over the north Pacific has risen slightly during the past 30
year, consistent with the trend toward the ‘high-index’ state of the AO during this
period. That the trend in the SAM accounts for approximately 50% of the warm-
ing over the Antarctic Peninsula and southern Patagonia attests to the importance
of climate mechanisms other than ENSO over this region. In consequence, the
additive effects of low- and high-latitude forcings on climates along the American
Cordilleras hamper a current estimation of past interactions between tropical and
high-latitude forcings. Independent reconstructions of tropical and polar modes of
variability are needed to gain insight on past forcing interactions and the combined
effect on climates of the western Americas.
Teleconnections between precipitation estimates and climatic forcings evolve
over time, suggesting a changing global signature of ENSO and other forcings.
Our results indicate that the timing of interdecadal transitions in temperature- and
precipitation-sensitive records has not always been consistent across the region. In
addition, the degree of correlation between records has varied over time. Over the
last 400 years, interdecadal climate variations in the Pacific alternated between times
of larger amplitude and more geographically widespread interdecadal changes and
times of lower amplitude and less geographically organized interdecadal changes.
Most tree-ring records suggest that the interdecadal variability in the Pacific
region was particularly more organized before the mid-1800s, whereas interannual
variability increased after that period. These observations point to a major reorgani-
zation of the climate modes of variability in the Pacific around 1850, the generally
accepted time of the end of a particularly cool period in the 1800s, identified in some
regions as the end of the Little Ice Age (LIA; Grove 1988; Bradley and Jones 1992;
Luckman and Villalba 2001).
Detailed analysis of the influences of tropical and high-latitude modes of cli-
mate variability across the American Cordilleras is somewhat constrained by the
current length of the proxy records. Additional insight can be gained by extend-
ing the records to cover different climatic intervals, such as the ‘warmer’ medieval
period. Jones et al. (2001), Mann and Jones (2003), and many other authors have
stressed the critical need for developing longer proxy records of both regional cli-
mate variations and hemispheric climatic forcings. Longer proxy series can be used
to evaluate the long-term natural behavior of these modes of variation and their
regional impacts more comprehensively. In so doing, these extended series can pro-
vide a long-term context for variability during the period of increasing trace gases,
and for testing climate prediction models. Proxy climate records stand as our only
means of assessing the long-term variability associated with large-scale modes of
climate variability and their global influences.
Acknowledgements We gratefully acknowledge the support of the Inter-American Institute for
Global Change Research through its Collaborative Research Network Program for the past 6 years.
Parts of this project have also been funded by a range of national, international, and nonprofit
organizations.
References
Aceituno P (1988) On the functioning of the Southern Oscillation in the South American sector.
Part I: surface climate. Mon Weather Rev 116:505–524
Aceituno P, Montecinos A (1996) Assessing upper limits of seasonal predictability of rainfall in
central Chile based on SST in the equatorial Pacific. Exp Long-Lead Forecast Bull 5:37–40
Aravena JC, Lara A, Wolodarsky-Franke A, Villalba R, Cuq E (2002) Tree-ring growth patterns
and temperature reconstruction from Nothofagus pumilio (Fagaceae) forests at the upper tree
line of southern Chilean Patagonia. Rev Chil Hist Nat 75:361–376
Argollo J, Soliz C, Villalba R (2004) Potencialidad dendrocronológica de Polylepis tarapacana en
los Andes Centrales de Bolivia. Ecología en Bolivia 39:5–24
Beamish RJ, Benson AJ, Sweeting RM, Neville CM (2004) Regimes and the history of the major
fisheries off Canada’s west coast. Prog Oceanogr 60:355–385
Boninsegna JA (1988) Santiago de Chile winter rainfall since 1220 as being reconstructed by tree
rings. Quaternary South Am Antarctic Peninsula 4:67–87
Bradley RS, Jones PD (1992) When was the ‘Little Ice Age’? In: Mikami T (ed) Proceedings of
the international symposium on the Little Ice Age climate. Department of Geography, Tokyo
Metropolitan University, Japan, pp 1–4
Brendel O, Iannetta PPM, Stewart D (2000) A rapid and simple method to isolate pure cellulose.
Phytochem Anal 11:7–10
Chavez FP, Ryan J, Llucha-Cota SE, Niquen M (2003) From anchovies to sardines and back:
multidecadal change in the Pacific Ocean. Science 299:217–221
Cleaveland MK (1986) Climatic response of densitometric properties in semiarid site tree rings.
Tree-Ring Bull 46:13–29
Cleaveland MK, Stahle DW, Therrell MD, Villanueva-Diaz J, Burns BT (2003) Tree-ring recon-
structed winter precipitation and tropical teleconnections in Durango, Mexico. Clim Change
59:369–388
Cobb KM, Charles CD, Cheng H, Edwards RL (2003) El Niño/Southern Oscillation and tropical
Pacific climate during the last millennium. Nature 424:271–276
Cole JE, Cook ER (1998) The changing relationship between ENSO variability and moisture
balance in the continental United States. Geophys Res Lett 25:4529–4532
Cook ER, Meko DM, Stahle DW, Cleaveland MK (1999) Drought reconstructions for the
continental United States. J Clim 12:1145–1162
Cook ER, Cole JE, D’Arrigo RD, Stahle DM, Villalba R (2000) Tree-ring records of past ENSO
variability and forcing. In: Diaz HF, Markgraf V (eds) El Niño and the Southern Oscillation:
multiscale variability, global and regional impacts. Cambridge University Press, Cambridge,
UK, pp 297–323
Cook ER, Woodhouse CA, Eakin CM, Meko DM, Stahle DW (2004) Long term aridity changes in
the western United States. Science 306:1015–1018
D’Arrigo RD, Jacoby G (1993) Secular trends in high northern latitude temperature reconstructions
based on tree rings. Clim Change 25:163–177
D’Arrigo RD, Villalba R, Wiles G (2001) Tree-ring estimates of Pacific decadal climate variability.
Clim Dynam 18:219–224
D’Arrigo RD, Cook ER, Mann ME, Jacoby GC (2003) Tree-ring reconstructions of temperature
and sea-level pressure variability associated with the warm-season Arctic Oscillation since AD
1650. Geophys Res Lett 30(11):1549. doi:10.1029/2003GL017250
D’Arrigo R, Wilson R, Deser C, Wiles G, Cook E, Villalba R, Tudhope S, Cole J, Linsley B (2005)
Tropical-north Pacific climate linkages over the past four centuries. J Clim 18:5253–5265
Deser C, Blackmon M (1993) Surface climate variations over the North Atlantic Ocean during
winter: 1900–1989. J Clim 6:1743–1753
Deser C, Wallace JM (1990) Large-scale atmospheric circulation features of warm and cold
episodes in the tropical Pacific. J Clim 3:1254–1281
Deser S, Phillips A, Hurrell J (2004) Pacific interdecadal climate variability: linkages between the
tropics and north Pacific during boreal winter since 1900. J Clim 17:3109–3124
Dettinger MD, Battisti DS, McCabe GJ, Bitz CM, Garreaud RD (2001) Interhemispheric effects
of interannual and decadal ENSO-like climate variations on the Americas. In: Markgraf V (ed)
Interhemispheric climate linkages. Academic, San Diego, pp 119–140
Ebbesmeyer CC, Cayan DR, McClain DR, Nichols FH, Peterson DH, Redmond KT (1991) 1976
step in the Pacific climate: forty environmental changes between 1968–75 and 1977–84. In:
Proceedings of the seventh annual Pacific climate (PACLIM) workshop. California Department
of Water Resources, pp 115–126
Eischeid JK, Diaz HF, Bradley RS, Jones JD (1991) A comprehensive precipitation data set for
global land areas. DOE/ER-69017T-H1, TR051, United States Department of Energy, Carbon
Dioxide Research Program, Washington, DC
Enfield DB, Mestas-Nuñez AM (2000) Global modes of ENSO and non-ENSO sea surface temper-
ature variability and their associations with climate. In: Diaz H, Markgraf V (eds) El Niño and
the Southern Oscillation: multiscale variability and global and regional impacts. Cambridge
University Press, Cambridge, UK, pp 89–112
Evans MN, Schrag DP (2004) A stable isotope-based approach to tropical dendroclimatology.
Geochim Cosmochim Ac 68:3295–3305
Evans MN, Cane MA, Schrag DP, Kaplan A, Linsley BK, Villalba R, Wellington GM (2001a)
Support for tropically-driven Pacific decadal variability based on paleoproxy evidence.
Geophys Res Lett 28:3869–3692
Evans MN, Kaplan A, Cane MA, Villalba R (2001b) Globality and optimality in climate field
reconstructions from proxy data. In: Markgraf V (ed) Interhemispheric climate linkages.
Academic, San Diego, pp 53–72
Fritts H (1976) Tree rings and climate. Academic, London
Fritts H (1991) Reconstructing large-scale climatic patterns from tree-ring data. University of
Arizona Press, Tucson
Garreaud R, Battisti DS (1999) Interannual and interdecadal (ENSO-like) variability in the
Southern Hemisphere tropospheric circulation. J Clim 12:2113–2123
Garreaud R, Vuille M, Clement A (2003) The climate of the Altiplano: observed current conditions
and mechanisms of past changes. Palaeogeogr Palaeocl 194:5–22
Gershunov A, Barnett TP (1998) Interdecadal modulation of ENSO teleconnections. B Am
Meteorol Soc 79:2715–2725
Gong D, Wang S (1999). Definition of Antarctic Oscillation index. Geophys Res Lett 26:459–462
Graham NE (1994) Decadal-scale climate variability in the 1970s and 1980s: observations and
model results. Clim Dynam 10:135–162
Graham NE, Hughes MK, Ammann CM, Cobb KM, Hoerling MP, Xu T, Kennett DJ, Kennett JP,
Rein B, Stott L, Wigand PE (2007) Tropical Pacific-mid-latitude teleconnections in Medieval
times. Clim Change 83:241–285
Grove JM (1988) The Little Ice Age. Methuen, London, 498 pp
Hartsough P, Biondi F (2003) The importance of high elevation monitoring in a tropical treeline
environment: a project update from southwestern Mexico. In: Fourth annual science meeting,
IAI CRN03, Abstracts, p 18
Hoerling MP, Hurrell JW, Xu T (2001) Tropical origins for recent North Atlantic climate change.
Science 292:90–92
Hughes MK, Funkhouser G (1998) Extremes of moisture availability reconstructed from tree rings
for recent millennia in the Great Basin of western North America. In: Beniston M, Innes JL
(eds) The impacts of climate variability on forests. Springer, Berlin, pp 99–107
Hurrell JW, van Loon H (1994) A modulation of the atmospheric annual cycle in the Southern
Hemisphere. Tellus 46A:325–338
Hurrell JW, van Loon H (1997) Decadal variations associated with the North Atlantic Oscillation.
Clim Change 36:301–326
Jones PD, Moberg A (2003) Hemispheric and large-scale surface air temperature variations: an
extensive revision and update to 2001. J Clim 16:206–223
Jones PD, Raper SCB, Bradley RS, Diaz HF, Kelly PM, Wigley TML (1986a) Northern
Hemisphere surface air temperature variations: 1851–1984. J Clim Appl Meteorol 25:161–179
Jones PD, Raper SCB, Wigley TML (1986b) Southern Hemisphere surface air temperature
variations: 1851–1984. J Clim Appl Meteorol 25:1213–1230
Jones PD, Osborn TJ, Briffa KR (2001) The evolution of climate over the last millennium. Science
292:662–667
Karoly DJ (1990) The role of transient eddies in low-frequency zonal variations of the Southern
Hemisphere circulation. Tellus 42A:41–50
Kennett DJ, Kennett JP (2000) Competitive and cooperative responses to climatic instability in
coastal southern California. Am Antiquity 65(2):379–395
Kessler M (1995) The genus Polylepis (Rosaceae) in Bolivia. Candollea 50:1–172
Kiladis GN, Diaz HF (1989) Global climatic anomalies associated with extremes of the Southern
Oscillation. J Clim 2:1069–1090
Kitzberger T, Swetnam TW, Veblen TT (2001) Inter-hemispheric synchrony of forest fires and the
El Niño-Southern Oscillation. Global Ecol Biogeogr 10:315–326
Labeyrie L, Cole J, Alverson K, Stocker T (2003) The history of climate dynamics in the late
quaternary. In: Alverson KD, Bradley RS, Pedersen TF (eds) Paleoclimate, global change and
the future. IGBP Series, Springer, Berlin, pp 33–62
LaMarche VC (1975) Potential of tree rings for reconstruction of past climate variations in the
Southern Hemisphere. In: Proceedings of the WMO/IAMAP symposium on long term climatic
fluctuations, Norwich, England. World Meteorological Organization, WMO No. 421, Geneva,
pp 21–30
Lara A, Aravena JC, Villalba R, Wolodarsky-Franke A, Luckman B, Wilson R (2001)
Dendroclimatology of high-elevation Nothofagus pumilio forests at their northern distribution
limit in the central Andes of Chile. Can J Forest Res 31:925–936
Lara A, Villalba R, Wolodarsky-Franke A, Aravena JC, Luckman BH, Cuq E (2005) Spatial and
temporal variation in Nothofagus pumilio growth at treeline along its latitudinal range (35º40
to 55ºS) in the Chilean Andes. J Biogeogr 32:879–893
Lenters JD, Cook K (1999) Summer-time precipitation variability over South America: role of the
large scale circulation. Mon Weather Rev 127:409–431
LeQuesne C, Stahle DW, Cleaveland MK, Therrell MD, Aravena JC, Barichivich J (2006) Ancient
ciprés tree-ring chronologies used to reconstruct central Chile precipitation variability from AD
1200–2000. J Clim 19:5731–5744
Linsley B, Wellington G, Schrag D, Ren L, Salinger M, Tudhope A (2004) Geochemical evi-
dence from corals for changes in the amplitude and spatial pattern of South Pacific interdecadal
climate variability over the last 300 years. Clim Dynam 22:1–11
Luckman BH, Boninsegna JA (2001) The assessment of present, past and future climate variability
in the Americas from treeline environments. PAGES Newsletter 9:17–19
Luckman BH, Villalba R (2001) Assessing the synchroneity of glacier fluctuations in the Western
Cordillera of the Americas during the last millennium. In: Markgraf V (ed) Interhemispheric
climate linkages. Academic, San Diego, pp 119–140
Luckman BH, Wilson RJS (2005) Summer temperature in the Canadian Rockies during the last
millennium: a revised record. Clim Dynam 24:131–144
Luckman BH, Briffa KR, Jones PD, Schweingruber FH (1997) Summer temperatures at the
Columbia Icefield, Alberta, Canada, 1073–1983. Holocene 7:375–389
Mann ME, Jones PD (2003) Global surface temperatures over the past two millennia. Geophys Res
Lett 30, art 1820
Mann ME, Bradley RS, Hughes MK (2000) Long-term variability in El Niño/Southern Oscillation
and associated teleconnections. In: Diaz HF, Markgraf V (eds) El Niño and the Southern
Oscillation: multiscale variability, global and regional impacts. Cambridge University Press,
Cambridge, UK, pp 357–412
Mantua JN, Hare SR (2002) The Pacific Decadal Oscillation. J Oceanogr (Japan) 58:35–44
Mantua JN, Hare SR, Zhang Y, Wallace JM, Francis RC (1997) A Pacific interdecadal climate
oscillation with impacts on salmon production. B Am Meteorol Soc 78:1069–1080
Martínez M (1963) Las pináceas mexicanas, 3rd edn. Instituto de Biología, Universidad Nacional
Autónoma de México, México
Meko D, Cook ER, Stahle DW, Stockton CW, Hughes MK (1993) Spatial patterns of tree-growth
anomalies in the United States and southeastern Canada. J Clim 6:1773–1786
Montecinos A, Aceituno P (2003) Seasonality of the ENSO-related rainfall variability in central

Chile and associated circulation anomalies. J Clim 16:281–296
Ogi M, Yamazaki K, Tachibana Y (2004) The summertime annular mode in the
Northern Hemisphere and its linkage to the winter mode. J Geophys Res 109:D20114.
doi:10.1029/2004JD004514
Quinn WH, Neal VT (1983) Long-term variations in the Southern Oscillation, El Niño, and the
Chilean subtropical rainfall. Fish B 81:363–374
Rein B, Lueckge A, Sirocko F (2004) A major Holocene ENSO anomaly in the medieval period.
Geophys Res Lett 31, L17211. doi:10.1029/2004GL020161
Rein B, Lueckge A, Reinhardt L, Sirocko F, Wolf A, Dullo W-C (2005) El Niño variability off
Peru during the last 20,000 years. Paleoceanography 20. doi:10.1029/2004PA001099
Roden JS, Lin G, Ehleringer JR (2000) A mechanistic model for interpretation of hydrogen and
oxygen ratios in tree-ring cellulose. Geochim Cosmochim Ac 64:21–35
Rodriguez R, Mabres A, Luckman B, Evans M, Masiokas M, Ektvedt T (2005) ‘El Niño’ events
recorded in dry-forest species of the lowlands of northwest Peru. Dendrochronologia 22:181–
187
Roig FA, Martínez Pastur G, Massaccesi G, Pinedo L (2003) Toward an interpretation of the inter-
actions between cambial activity, phenology and climate in Nothofagus pumilio forests of Tierra
del Fuego, Argentina. In: Fourth annual science meeting, IAI CRN 03, Abstracts, pp 33–34
Ropelewski CF, Halpert MS (1986) North American precipitation and temperature patterns
associated with El Niño/Southern Oscillation (ENSO). Mon Weather Rev 115: 1606–1626
Rutlland J, Fuenzalida H (1991) Synoptic aspects of the central Chile rainfall variability associated
with the Southern Oscillation. Int J Climatol 11:63–76
Stahle DW, D’Arrigo RD, Krusic PJ, Cleaveland MK, Cook ER, Allan RJ, Cole JE, Dunbar RB,
Therrell MD, Gay DA, Moore MD, Stokes MA, Burns BT, Villanueva-Diaz J, Thompson
LG (1998) Experimental dendroclimatic reconstruction of the Southern Oscillation. B Am
Meteorol Soc 79(10):2137–2152
Stine S (1994) Extreme and persistent drought in California and Patagonia during mediaeval time.
Nature 369:546–549
Stott LD, Cannariato KG, Thunell R, Haug GH, Koutavas A, Lund S (2004) Decline of surface
temperature and salinity in the western tropical Pacific Ocean in the Holocene epoch. Nature
431:56–59
Thompson DWJ, Lorenz DJ (2004) The signature of the annular modes in the tropical troposphere.
J Clim 17:4330–4342
Thompson DWJ, Solomon S (2002) Interpretation of recent Southern Hemisphere climate change.
Science 296:895–899
Thompson DWJ, Wallace JM (2000) Annular modes in the extratropical circulation. Part I: Month-
to-month variability. J Clim 13:1000–1016
Trenberth K, Hurrell J (1994) Decadal atmospheric-ocean variations in the Pacific. Clim Dynam
9:303–19
Vautard R (1995) Patterns in time: SSA and MSSA. In: von Storch H, Navarra A (eds) Analysis of
climate variability, applications of statistical techniques. Springer, Berlin, pp 259–279
Vautard R, Ghil M (1989) Singular spectrum analysis in nonlinear dynamics, with applications to
paleoclimatic time series. Physica D35:395–424
Villalba R, Boninsegna JA, Veblen TT, Schmelter A, Rubulis S (1997) Recent trends in tree-ring
records from high elevation sites in the Andes of northern Patagonia. Clim Change 36:425–454
Villalba R, D’Arrigo RD, Cook ER, Wiles G, Jacoby GC (2001) Decadal-scale climatic variability
along the extratropical western coast of the Americas: evidences from tree-ring records. In:
Markgraf V (ed) Interhemispheric climate linkages. Academic, San Diego, pp 155–172
Villalba R, Lara A, Boninsegna JA, Masiokas M, Delgado S, Aravena JC, Roig FA, Schmelter
A, Wolodarsky A, Ripalta A (2003) Large-scale temperature changes across the southern
Andes: twentieth-century variations in the context of the past 400 years. Clim Change 59:
177–232
Vuille M (1999) Atmospheric circulation over the Bolivian Altiplano during dry and wet periods
and extreme phases of the Southern Oscillation. Int J Climatol 19:1579–1600
Vuille M, Bradley RS, Keimig F (2000) Interannual climate variability in the central Andes and its
relation to tropical Pacific and Atlantic forcing. J Geophys Res 105:12447–12460
Wallace JM, Rasmusson EM, Mitchell TP, Kousky VE, Sarachik ES, von Storch H (1998). On the
structure and evolution of ENSO-related climate variability in the tropical Pacific: lessons from
TOGA. J Geophys Res 103:12241–14259
Watson E, Luckman BH (2004a) Tree-ring based reconstructions of precipitation for the southern
Canadian Cordillera. Clim Change 65:209–241
Watson E, Luckman BH (2004b) Tree-ring estimates of mass balance at Peyto Glacier for the last
three centuries. Quaternary Res 62:9–18
Watson E, Luckman BH (2005) Spatial patterns of pre-instrumental moisture variability in the
southern Canadian Cordillera. J Clim 18:2847–2863
White W, Cayan DR (1998) Quasi-periodicity and global symmetries in interdecadal upper ocean
temperature variability. J Geophys Res 103:21335–21354
Wiles GC (1997) North Pacific atmosphere-ocean variability over the past millennium inferred
from coastal glaciers and tree rings. In: Preprint volume, Eighth symposium on global change
studies. American Meteorological Society, 2–7 February 1997, Long Beach, California. pp
218–220
Wiles GC, Calkin PE (1994) Late Holocene, high-resolution glacial chronologies and climate,
Kenai Mountains, Alaska. Geol Soc Am Bull 106:281–303
Wiles GC, D’Arrigo RD, Jacoby GC (1996) Temperature changes along the Gulf of Alaska and
the Pacific Northwest coast modeled from coastal tree rings. Can J Forest Res 26:474–481
Wiles GC, D’Arrigo RD, Jacoby GC (1998) Gulf of Alaska atmosphere-ocean variability over
recent centuries inferred from coastal tree-ring records. Clim Change 38:289–306
Wiles GC, D’Arrigo RD, Villalba R, Calkin PE, Barclay DJ (2004) Century-scale solar variabil-
ity and Alaskan temperature change over the past millennium. Geophys Res Lett 31:L15203.
doi:1029/2004GL020050
Wolodarsky-Franke A, Moreno P, Lara A, Pino M, Villarosa G (2005) Tree-ring, stratigraphic, and
palynological evidence for volcanic and climatic controls on Fitzroya cupressoides forests in
southern Chile, over the last 5700 years. In: Piovano EL, Leroy S (eds) Holocene environmental
catastrophes in South America: from the lowlands to the Andes. Abstract volume and field
guide. 11–17 Third joint meeting of the International Council for Science, Dark Nature and the
International Geological Correlation Programme 490, March 2005, Miramar-Córdoba
Zhang Y, Wallace JM, Battisti DS (1997) ENSO-like interdecadal variability: 1900–93. J Clim
10:1004–1020
Zielinski GA (2000) Use of paleo-records in determining variability within the volcanism-climate
systems. Quaternary Sci Rev 19:417–438

Dendrochronology

Uploaded by

Copyright:

Available Formats

Dendrochronology

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dendrochronology

Uploaded by

Copyright:

Available Formats

Dendroclimatology

Progress and Prospects

Ricardo Villalba, Brian H. Luckman, Jose Boninsegna, Rosanne D. D’Arrigo,

Abstract Common patterns of climatic variability across the Western Americas

M.K. Hughes et al. (eds.), Dendroclimatology, Developments in Paleoenvironmental 175

tree growth in most extratropical temperature-sensitive chronologies in the Western

Keywords Dendrochronology · Regional scale · Continental scale · Climate

Patagonian grasslands. These coherent interhemispheric changes in annual and

One of the Collaborative Research Networks supported by the Inter-American

7.2 Oscillatory Modes of Climate Variability Across the Western

7.2.1 El Niño/Southern Oscillation (ENSO)

7.2.2 Pacific Interdecadal Mode

Using instrumental records, several studies have reported Pacific decadal-scale

7.2.3 Annular Modes

The Northern Hemisphere (NAM) and Southern Hemisphere (SAM) Annular

7.3 Tree-Ring Records Across the Western Americas

Box 7.1 Climate signals in Gulf of Alaska tree-ring records

A network of climatically sensitive tree-ring records has been compiled

Analyses of instrumental data demonstrate robust linkages between

reconstruction of an optimal tropical Indo-Pacific climate index (OTI), sup-

America, there is a gradual environmental gradient from the relatively dry-warm

7.3.1 Temperature-Sensitive Records

The strong cross-equatorial symmetries of SST, continental temperature, and con-

Box 7.2 Studies from the Canadian Cordillera

7.3.1.1 Extratropical Pacific Ocean

Box 7.3 Climate signals in Patagonian upper-elevation

a north-south direction provides the opportunity to examine the relationships

Box Fig. 7.5 Nothofagus pumilio chronologies in the Patagonian Andes

Atlantic Oceans clearly show that the temperature reconstructions contain

7.3.1.2 Tropical Pacific Ocean

7.3.1.3 High-Latitude Oscillations

7.3.2 Precipitation-Sensitive Records

Box 7.4 Spatial patterns of drought and wetness regimes

The network of moisture-sensitive tree-ring chronologies now available for

Box 7.5 Western United States droughts in medieval times

Several multicentury shifts in precipitation over the central Great Basin

picture would be consistent with the Pacific-wide pattern proposed here.

In 1999, a gridded network of 154 Palmer Drought Severity Index reconstruc-

Box 7.6 A network of tree-ring chronologies for northern

Box Fig. 7.11 Tree-ring chronologies in Mexico

Ahuehuete or Montezuma bald cypress (Taxodium mucronatum), a widely

Box 7.7 The Polylepis tarapacana chronologies: The highest

Polylepis stands are almost exclusively restricted to volcanic slopes, with a

1400 1500 1600 1700 1800 1900 2000

Traditionally, the wood of Polylepis tarapacana has been used by local

—Jaime Argollo, Claudia Soliz, Jorge Moya, Janette Pacajes, Mariano S.

7.3.2.1 Subtropical Precipitation and ENSO

Box 7.8 Tree-ring chronologies from Austrocedrus chilensis

Several tree-ring collections of Austrocedrus chilensis were taken in cen-

It is important to note that based on instrumental records, the strongest telecon-

7.3.2.2 Dominant Oscillations in Precipitation Variations

These spectral analyses basically reaffirm the existence of common oscillatory

7.4 Future Research

rings. Interannual variability up to 8‰ in δ18 O cellulose is associated with year-to

Box 7.9 Monitoring of tree growth dynamics to improve

Fine-scale monitoring of tree growth dynamics is essential to improve our

In Nevado de Colima, a 3900 m elevation site in central western México,

7.5 Discussion and Conclusions

resemble those of the tree-ring reconstructions of Arctic mean annual temperatures