Reconstruction of Climate
These interpretations are essentially qualitative. Transfer
functions and the analog method (see Chap. 10, Sect. 10.1.3)
provide quantitative information on climate. Here, we use
the analog method, not on the percentages of taxa, but on the
PFT scores (Fig. 12.2.), which has many advantages: it
reduces the number of variables and groups together the taxa
with similar behavior, making the approach more robust.
Twelve analogs for each fossil spectrum were retained.
Figure 12.2 shows these reconstructions with the shaded
area representing the range of variability between the analogs. The quality of the reconstructions is controlled by
applying the same method to modern data: for each modern
spectrum, the best analogs are determined (obviously
excluding the spectrum itself) and the present climate is
reconstructed so that it can be compared to direct observations. In this case, we find a coefficient of determination (r
2 )
of 0.64 for the annual precipitation (Pann), 0.89 for temperatures in January (Tjan) and 0.93 for annual average
temperatures (Tann). The estimates obtained for precipitation will therefore contain larger errors than the other variables. This is confirmed by Fig. 12.2. In terms of climate, we
see that the Younger Dryas was very cold (14 ± 7 °C colder
than now) and dry (400 ± 400 mm/year less rainfall than
now over the annual average), but that the uncertainties are
large (in this method, this is the variability between analogs
rather than an actual error bar). The temperature maximum
(4 ± 2 °C more than now) occurred at around 10,000 years
BP, when oak dominated, and the rainfall maximum
(100 ± 250 mm more) occurred around 8200 years BP,
with the arrival of the mixed forest. However, the 8200 BP
event does not appear to have been short-lived since this
forest continued for several millennia longer.
Several other methods have been proposed to reconstruct
climate from pollen data (see Brewer et al. 2007; Birks 2011;
Juggins 2013). They all have their strengths and weaknesses.
It is not always easy to find the optimal method. It is recommended to try several and compare the results. The
convergence of estimates is an indication of the robustness
of the reconstruction, and their divergence is often a sign that
the initial assumptions used for the reconstruction were not
entirely valid. In particular, when the climate changes
rapidly, vegetation adapts with a certain delay, which makes
it difficult to find current analogs. Another problem is estimating the impact of non-climate constraints. For example,
atmospheric CO 2 concentration during the Quaternary
glaciations was lower than the current concentration (about
200 ppm instead of more than 280 ppm). We know that this
level is an activator of photosynthesis and so it is unlikely
that the principle of uniformity applies: vegetation will not
respond to climate change in the same way when atmospheric CO 2 levels change. This problem was solved by the
use of mechanistic models of vegetation (Guiot et al. 2000).
Another avenue takes pollen dispersal and associated biases
into account when reconstructing the landscape (Sugita
2007; Trondman et al. 2015). The relative abundance of
pollen of a taxon depends on the height and distance of the
productive plant, its pollen productivity, the weight and
shape of the pollen grain, the dimensions of the lake into
which it falls.
To improve climate reconstructions without resorting to
overly complex models, the ‘multiproxy’ approach is recommended. This involves taking several simultaneous climate indicators into account: pollen, macro remains of plants
or animals, carbon isotopes, lake levels, sedimentological
parameters etc. as was done by Cheddadi et al. (1997). This
multiproxy approach is also supported by mechanistic
models (Rousseau et al. 2006. Guiot et al. 2009). These
avenues of research are being explored both to reconstruct
the climate of the past and to understand how it has influenced the vegetation.
References
Birks, H. J. B. (2011). Strengths and weaknesses of quantitative climate
reconstructions based on late-quaternary biological proxies. The
Open Ecology Journal, 3, 68–110. https://doi.org/10.2174/
1874213001003020068.
Brewer, S., Guiot, J., Barboni, D. (2007). Pollen Data as Climate
Proxies. In Encyclopedia of Quaternary Sciences, Elsevier.
Cheddadi, R., Yu, G., Guiot, J., Harrison, S. P., & Prentice, I. C.
(1997). The Climate of Europe 6000 years ago. Climate Dynamics,
13, 1–9.
Guiot, J., Torre, F., Jolly, D., Peyron, O., Boreux, J. J., & Cheddadi, R.
(2000). Inverse vegetation modeling by monte carlo sampling to
reconstruct palaeoclimates under changed precipitation seasonality
and CO 2 conditions: Application to glacial climate in mediterranean
region. Ecological Modelling, 127, 119–140.
Guiot, J., Wu, H., Garreta, V., Hatti, L., & Magny, M. (2009). A few
prospective ideas on climate reconstruction from a statistical single
proxy approach towards a multi-proxy and dynamical approach.
Climate of the Past, 5, 571–583.
Jolly, D. et al. (1998). Biome reconstruction from pollen and Plant
Macrofossil data for Africa and the Arabian Peninsula at 0 and 6 ka.
Journal of Biogeography, 25, 1 007–1 028.
Juggins, S. (2013). Quantitative reconstructions in palaeolimnology:
New paradigm or sick science? Quaternary Science Review, 64.
https://doi.org/10.1016/j.quascirev.2012.12.014.
Lotter, A. F., & Zbinden, H. (1989). ‘Late-glacial pollen analysis,
Oxygen-Isotope and Radiocarbon stratigraphy from Rotsee
(Lucerne), Central Swiss Plateau. Ecologae Gologicae Helvetiae,
82, 191–202.
Moore, P. D., Webb, J. A., & Collinson, M. E. (1991). Pollen Analysis.
Oxford: Blackwell Scientific Publications.
Prentice, I. C., Guiot, J., Huntley, B., Jolly, D., & Cheddadi, R. (1996).
Reconstructing biomes from Palaeoecological data: A general
154
J. Guiot
These interpretations are essentially qualitative. Transfer
functions and the analog method (see Chap. 10, Sect. 10.1.3)
provide quantitative information on climate. Here, we use
the analog method, not on the percentages of taxa, but on the
PFT scores (Fig. 12.2.), which has many advantages: it
reduces the number of variables and groups together the taxa
with similar behavior, making the approach more robust.
Twelve analogs for each fossil spectrum were retained.
Figure 12.2 shows these reconstructions with the shaded
area representing the range of variability between the analogs. The quality of the reconstructions is controlled by
applying the same method to modern data: for each modern
spectrum, the best analogs are determined (obviously
excluding the spectrum itself) and the present climate is
reconstructed so that it can be compared to direct observations. In this case, we find a coefficient of determination (r
2 )
of 0.64 for the annual precipitation (Pann), 0.89 for temperatures in January (Tjan) and 0.93 for annual average
temperatures (Tann). The estimates obtained for precipitation will therefore contain larger errors than the other variables. This is confirmed by Fig. 12.2. In terms of climate, we
see that the Younger Dryas was very cold (14 ± 7 °C colder
than now) and dry (400 ± 400 mm/year less rainfall than
now over the annual average), but that the uncertainties are
large (in this method, this is the variability between analogs
rather than an actual error bar). The temperature maximum
(4 ± 2 °C more than now) occurred at around 10,000 years
BP, when oak dominated, and the rainfall maximum
(100 ± 250 mm more) occurred around 8200 years BP,
with the arrival of the mixed forest. However, the 8200 BP
event does not appear to have been short-lived since this
forest continued for several millennia longer.
Several other methods have been proposed to reconstruct
climate from pollen data (see Brewer et al. 2007; Birks 2011;
Juggins 2013). They all have their strengths and weaknesses.
It is not always easy to find the optimal method. It is recommended to try several and compare the results. The
convergence of estimates is an indication of the robustness
of the reconstruction, and their divergence is often a sign that
the initial assumptions used for the reconstruction were not
entirely valid. In particular, when the climate changes
rapidly, vegetation adapts with a certain delay, which makes
it difficult to find current analogs. Another problem is estimating the impact of non-climate constraints. For example,
atmospheric CO 2 concentration during the Quaternary
glaciations was lower than the current concentration (about
200 ppm instead of more than 280 ppm). We know that this
level is an activator of photosynthesis and so it is unlikely
that the principle of uniformity applies: vegetation will not
respond to climate change in the same way when atmospheric CO 2 levels change. This problem was solved by the
use of mechanistic models of vegetation (Guiot et al. 2000).
Another avenue takes pollen dispersal and associated biases
into account when reconstructing the landscape (Sugita
2007; Trondman et al. 2015). The relative abundance of
pollen of a taxon depends on the height and distance of the
productive plant, its pollen productivity, the weight and
shape of the pollen grain, the dimensions of the lake into
which it falls.
To improve climate reconstructions without resorting to
overly complex models, the ‘multiproxy’ approach is recommended. This involves taking several simultaneous climate indicators into account: pollen, macro remains of plants
or animals, carbon isotopes, lake levels, sedimentological
parameters etc. as was done by Cheddadi et al. (1997). This
multiproxy approach is also supported by mechanistic
models (Rousseau et al. 2006. Guiot et al. 2009). These
avenues of research are being explored both to reconstruct
the climate of the past and to understand how it has influenced the vegetation.
References
Birks, H. J. B. (2011). Strengths and weaknesses of quantitative climate
reconstructions based on late-quaternary biological proxies. The
Open Ecology Journal, 3, 68–110. https://doi.org/10.2174/
1874213001003020068.
Brewer, S., Guiot, J., Barboni, D. (2007). Pollen Data as Climate
Proxies. In Encyclopedia of Quaternary Sciences, Elsevier.
Cheddadi, R., Yu, G., Guiot, J., Harrison, S. P., & Prentice, I. C.
(1997). The Climate of Europe 6000 years ago. Climate Dynamics,
13, 1–9.
Guiot, J., Torre, F., Jolly, D., Peyron, O., Boreux, J. J., & Cheddadi, R.
(2000). Inverse vegetation modeling by monte carlo sampling to
reconstruct palaeoclimates under changed precipitation seasonality
and CO 2 conditions: Application to glacial climate in mediterranean
region. Ecological Modelling, 127, 119–140.
Guiot, J., Wu, H., Garreta, V., Hatti, L., & Magny, M. (2009). A few
prospective ideas on climate reconstruction from a statistical single
proxy approach towards a multi-proxy and dynamical approach.
Climate of the Past, 5, 571–583.
Jolly, D. et al. (1998). Biome reconstruction from pollen and Plant
Macrofossil data for Africa and the Arabian Peninsula at 0 and 6 ka.
Journal of Biogeography, 25, 1 007–1 028.
Juggins, S. (2013). Quantitative reconstructions in palaeolimnology:
New paradigm or sick science? Quaternary Science Review, 64.
https://doi.org/10.1016/j.quascirev.2012.12.014.
Lotter, A. F., & Zbinden, H. (1989). ‘Late-glacial pollen analysis,
Oxygen-Isotope and Radiocarbon stratigraphy from Rotsee
(Lucerne), Central Swiss Plateau. Ecologae Gologicae Helvetiae,
82, 191–202.
Moore, P. D., Webb, J. A., & Collinson, M. E. (1991). Pollen Analysis.
Oxford: Blackwell Scientific Publications.
Prentice, I. C., Guiot, J., Huntley, B., Jolly, D., & Cheddadi, R. (1996).
Reconstructing biomes from Palaeoecological data: A general
154
J. Guiot