with remarkable sedimentary facies that allow the different
profiles to be correlated. Indicators of a small sliver of water
or of a drought located in the profile (mollusk shells, coarse
sand beds, or paleosols) as well as former perched shorelines
provide direct data on lake level fluctuations. The mineralogical composition of the sediments, their grain size, the
geochemical properties of the various fractions (mineral or
organic, detrital or authigenic), and the assemblages of fossilized organisms are indirect indicators of past environmental conditions.
In the outcrops of the Bolivian Altiplano, we focused our
investigations on the diatoms. These are microscopic unicellular algae (3–100 lm), which are very sensitive to
variations in environmental parameters, such as the depth or
salinity of water, and each cell produces an easily fossilized
bivalve siliceous outer skeleton. This external skeleton,
called a frustule, consists of an inner layer of silica tetrahedra
(Si 2 O, nH 2 O) bonded to each other, and of a hydrated outer
layer made of organic material, mainly proteins and
polysaccharides. Diatom valves are extracted from sediments by physicochemical treatment. In optical microscopy,
the species are identified by the structure of their valves:
their absolute (number of valves/g) and relative (percentage)
abundance can then be estimated. Knowing the self-ecology
of each species identified makes it possible to characterize
the aquatic environment in which they developed.
The presence of oxygen in the inner layer of the frustule
allows the isotopic composition in oxygen 18 to be measured. Preliminary dehydration is necessary because of the
hydrated nature of the silica so that only the isotopes in the
oxygen of the internal structure are measured. After extraction of oxygen and purification in a preparation line, the O 2
molecule is analyzed by mass spectrometry (Crespin et al.
2008, Crespin et al. 2010).
Quantification of the Oxygen Isotopic
Composition of Lakes
In lacustrine environments, the isotopic oxygen composition
of diatoms (d
18 O diatoms ) varies with the temperature and
isotopic composition of the lake water (d
18 O lake water ).
Diatomaceous silica precipitates in isotopic equilibrium with
the lake water, and calibration studies have established the
thermo-dependent relationship [([d
18 O diatoms − d
18 O lake
water ] (‰ vs. VSMOW) = a*T lake water (°C) + b)] to
express the variation in the isotopic composition of diatoms
as a function of temperature (Brandriss et al. 1998; Moschen
et al. 2005; Crespin et al. 2010; Dodd and Sharp 2010; Dodd
et al. 2012; Alexandre et al. 2012). These relationships
between the oxygen isotopic composition of the diatoms and
the water during formation have been experimentally
established for lacustrine diatoms under contemporary
environmental conditions, taking the seasonal to
multi-annual variations into account (Fig. 19.2).
These relationships all show a low dependence on temperature, of around −0.20‰/°C. On the other hand, the
fractionation factors are very different, which may be due
either to sampling problems or to different methodological
approaches when carrying out these calibrations (Crespin
et al. 2010; Alexandre et al. 2012). Nevertheless, applied in
different contexts, this method proved relevant to the
reconstruction of the isotopic composition of lakes in the
past (Leng and Barker 2006). In addition, the application of
these different calibrations enables sensitivity tests to be
conducted depending on the potential temperature range
applicable during the period considered.
So that the thermo-dependence of the isotopic fractionation between the diatoms and the formation water is taken
into account, the temperature of the lake water in which they
develop must also be known. Diatoms are photosynthetic
organisms which grow in the epilimnion and reflect the
surface temperatures of the lake. For current periods, field
observations are essential; for past periods, assumptions
based on simulated atmospheric temperature data from climate models and data obtained from other proxies in the
same region are used.
Hydro-Isotopic Modeling and Paleoclimatic
Interpretation
A quantified estimate of the hydrological balance of lakes or
of the P-E is obtained by hydrological modeling. The more
or less sophisticated models used all rely on the balance
equation for lake water (Eq. 1). The first step is to derive the
contemporary water balance from the available instrumental
data and the relationships between H L , V L and S L constructed from a digital terrain model (DTM). Estimating
evaporation, a function of temperature, solar radiation,
humidity, wind and vegetation cover, is always complex. It
involves the use of different hydrological and climatic concepts and methods. It is better constrained by connecting the
water balance to the salt balance (Vallet-Coulomb et al.
2001) with a watershed runoff model (Legesse et al. 2004) or
with an energy balance model (Kutzbach 1980). Application
of the model to the past is usually done for a time interval
where the lake is considered to have been in equilibrium
(DV L = 0). We know S B , H L (from which we calculate V L
and S L ), the solar radiation provided by astronomical calculations, paleotemperatures and paleo- vegetation on the
watershed can be deduced from data e.g. from palynology.
19 Air-water Interface: Tropical Lake Diatoms and Isotope …
215
profiles to be correlated. Indicators of a small sliver of water
or of a drought located in the profile (mollusk shells, coarse
sand beds, or paleosols) as well as former perched shorelines
provide direct data on lake level fluctuations. The mineralogical composition of the sediments, their grain size, the
geochemical properties of the various fractions (mineral or
organic, detrital or authigenic), and the assemblages of fossilized organisms are indirect indicators of past environmental conditions.
In the outcrops of the Bolivian Altiplano, we focused our
investigations on the diatoms. These are microscopic unicellular algae (3–100 lm), which are very sensitive to
variations in environmental parameters, such as the depth or
salinity of water, and each cell produces an easily fossilized
bivalve siliceous outer skeleton. This external skeleton,
called a frustule, consists of an inner layer of silica tetrahedra
(Si 2 O, nH 2 O) bonded to each other, and of a hydrated outer
layer made of organic material, mainly proteins and
polysaccharides. Diatom valves are extracted from sediments by physicochemical treatment. In optical microscopy,
the species are identified by the structure of their valves:
their absolute (number of valves/g) and relative (percentage)
abundance can then be estimated. Knowing the self-ecology
of each species identified makes it possible to characterize
the aquatic environment in which they developed.
The presence of oxygen in the inner layer of the frustule
allows the isotopic composition in oxygen 18 to be measured. Preliminary dehydration is necessary because of the
hydrated nature of the silica so that only the isotopes in the
oxygen of the internal structure are measured. After extraction of oxygen and purification in a preparation line, the O 2
molecule is analyzed by mass spectrometry (Crespin et al.
2008, Crespin et al. 2010).
Quantification of the Oxygen Isotopic
Composition of Lakes
In lacustrine environments, the isotopic oxygen composition
of diatoms (d
18 O diatoms ) varies with the temperature and
isotopic composition of the lake water (d
18 O lake water ).
Diatomaceous silica precipitates in isotopic equilibrium with
the lake water, and calibration studies have established the
thermo-dependent relationship [([d
18 O diatoms − d
18 O lake
water ] (‰ vs. VSMOW) = a*T lake water (°C) + b)] to
express the variation in the isotopic composition of diatoms
as a function of temperature (Brandriss et al. 1998; Moschen
et al. 2005; Crespin et al. 2010; Dodd and Sharp 2010; Dodd
et al. 2012; Alexandre et al. 2012). These relationships
between the oxygen isotopic composition of the diatoms and
the water during formation have been experimentally
established for lacustrine diatoms under contemporary
environmental conditions, taking the seasonal to
multi-annual variations into account (Fig. 19.2).
These relationships all show a low dependence on temperature, of around −0.20‰/°C. On the other hand, the
fractionation factors are very different, which may be due
either to sampling problems or to different methodological
approaches when carrying out these calibrations (Crespin
et al. 2010; Alexandre et al. 2012). Nevertheless, applied in
different contexts, this method proved relevant to the
reconstruction of the isotopic composition of lakes in the
past (Leng and Barker 2006). In addition, the application of
these different calibrations enables sensitivity tests to be
conducted depending on the potential temperature range
applicable during the period considered.
So that the thermo-dependence of the isotopic fractionation between the diatoms and the formation water is taken
into account, the temperature of the lake water in which they
develop must also be known. Diatoms are photosynthetic
organisms which grow in the epilimnion and reflect the
surface temperatures of the lake. For current periods, field
observations are essential; for past periods, assumptions
based on simulated atmospheric temperature data from climate models and data obtained from other proxies in the
same region are used.
Hydro-Isotopic Modeling and Paleoclimatic
Interpretation
A quantified estimate of the hydrological balance of lakes or
of the P-E is obtained by hydrological modeling. The more
or less sophisticated models used all rely on the balance
equation for lake water (Eq. 1). The first step is to derive the
contemporary water balance from the available instrumental
data and the relationships between H L , V L and S L constructed from a digital terrain model (DTM). Estimating
evaporation, a function of temperature, solar radiation,
humidity, wind and vegetation cover, is always complex. It
involves the use of different hydrological and climatic concepts and methods. It is better constrained by connecting the
water balance to the salt balance (Vallet-Coulomb et al.
2001) with a watershed runoff model (Legesse et al. 2004) or
with an energy balance model (Kutzbach 1980). Application
of the model to the past is usually done for a time interval
where the lake is considered to have been in equilibrium
(DV L = 0). We know S B , H L (from which we calculate V L
and S L ), the solar radiation provided by astronomical calculations, paleotemperatures and paleo- vegetation on the
watershed can be deduced from data e.g. from palynology.
19 Air-water Interface: Tropical Lake Diatoms and Isotope …
215