Geoscience Reference
In-Depth Information
problem involved with obtaining quantitative estimates of changes in precipitation
and temperature from analysis of the
18
O composition of ice cores, alluded to in
Section 21.4
when discussing carbonates in fossil soils in the Andes, involves the
choice of
18
O precipitation lapse rate. This is because elevation is only one of a
group of factors controlling the lapse rate, some of the others being distance travelled
by the air mass and the intensity of convectional downpours during the passage of
moist air across Amazonia. These difficulties apply equally to the paleoclimatic ana-
lysis of speleothems collected from limestone caves in South America (Cruz et al.,
2009
).
Thompson et al. (
2013
) have recently obtained an annual record of climatic fluctu-
ations for the last 1,800 years from the Quelccaya ice cap. The
18
O oxygen isotopic
values showed an increase in snow accumulation rates and a decrease in temperature
during the latter half of the Little Ice Age, from 1681 to 1880 AD. High concentrations
of nitrate and ammonium, and
18
O depleted isotopes in the ice cores, coincided with
decreased percentages of Ti in the Cariaco Basin marine record, and vice versa. This
indicates that when run-off and precipitation along the north-east coast of tropical
South America were reduced, conditions were wetter further south over the Amazon
Basin, and conversely.
21.5.4 Lake fluctuations
A number of lakes located at high elevations in the central Andes and semi-arid
Bolivian Altiplano provide evidence of late Quaternary fluctuations in temperature
and precipitation (Sylvestre,
2009
). Many of these lakes are now shallow and saline
saltpans known as
salars
. Lake Titicaca was deep and fresh between 25 and 15 ka
(Baker et al.,
2001
). In the southern Bolivian Altiplano, the
salar
of Coipasa was a
shallow salt lake between 24.6 and 20.9 ka, when conditions were more humid than
today (Sylvestre et al.,
1998
; Sylvestre,
2002
). The
salar
of Uyuni was also less arid
between 26.1 and 14.9 ka. The lake chronology for the Uyuni-Coipasa Basinwas based
on both
14
Cand
230
Th/
234
U ages (Sylvestre et al.,
1999
). One difficulty that Sylvestre
and her colleagues noted during this work was the varying reliability of radiocarbon
ages. For example, the lakes began to rise a little before 16,000 radiocarbon years
ago (
14
C yr BP) and reached maximum levels between 13,000 and 12,000
14
CyrBP.
Following a dry spell, the lake rose again to a lower level between about 9,500 and
8,500
14
C yr BP. There was good agreement between the
14
Cand
230
Th/
234
Uages
for the first and highest lake phase but a lack of accord for the second lake phase
(Sylvestre et al.,
1999
). They concluded that this discrepancy was probably caused
by a delayed response of the groundwater-table during the early Holocene dry phase
and used a correction of about 2,000
14
C years for this reservoir effect.
Other workers in the southern Andes and Atacama have faced the same problem
of very large reservoir effects. Geyh et al. (
1999
) investigated the timing of the late
glacial/early Holocene humid phase along a high-altitude transect between 18
°
Sand
