A two hour drive and a 40 minute boat ride along the Ganges north of Calcutta is Green Island. As the name suggests this is an island where the original vegetation that once covered the Ganges Delta remains relatively undisturbed. I say relatively because while human disturbance is limited, every year when monsoon storms and high tides swell the river level Green Island is flooded, sometimes to a depth a little over a meter.
Studying this type of vegetation is crucial to my work as it represents exactly the kind of plant community that has the potential to be preserved in the fossil record. As the floodwaters scour the leaf litter from the soil surface, and the island margins are eaten away by erosion, leaves are washed into the river where they may be deposited in muds and silts and eventually fossilised. We see many such deposits in the rock record from which we deduce information on past vegetation and climate.
I was in the company of Professor Subir Bera from Calcutta University and his wife who had organised the day. We had special police permission to visit and collect on the island and even had a police escort. On the boat with us were several people from the nearby village who had (as we discovered later) prepared a wonderful lunch of fish and meat curries, rice and fruit, all served on banana leaves.
As the open wooden boat neared the island we could see whole trees, still with green leaves, that had recently fallen in to the water as the riverbanks were eroded. Caught in the branches of one such tree was the body of a goat. Now for most this might seem gruesome but for me it was fascinating because it was another example of taphonomic processes – taphonomy is the study of fossilisation.
As the Ganges undercuts the edges of Green Island whole trees fall into the river - perhaps on their way to being fossilised.
Green Island is a little over a kilometre in length and a few hundred metres wide. Here we collected the populations of leaves from 56 different species of trees shrubs and vines. After pressing, drying and mathematically scoring them CLAMP analysis positioned the Green Island vegetation near the Kerala sites we had previously analysed, but in an area of the three-dimensional plot that indicated they were from a slightly cooler site.
Because Green Island is on a flat delta plain I could use meteorological data in the form of a grid in which observations from individual meteorological stations can be interpolated (mathematically extrapolated) for sites such as Green Island that does not have its own measurements.
CLAMP showing the positions of modern forests determined by their leaf architecture
The plot above shows the positions of the Indian forests that I have analysed so far. In this plot each ball represents a forest. Green Island is Labelled “Green” and the other labelled balls are forests in Kerala. The positions of balls are determined by the numerical score that describes leaf architecture for at least twenty species of woody trees, shrubs and vines from each forest.
Balls that plot close together indicate forests with similar leaf architectures, while that that plot far apart are very different. The balls are colour coded such that blue represents cool climates and red ones hot climates. Orange, green and light blue indicate forests growing in intermediate climates. It is easy to see that the Indian forests (coded maroon) all lie in the warm end of the plot.
In this plot the Indian forests have been treated as if they were fossils. They have found their own position with respect to the other sites for which the climate is known. Despite the fact that the Indian sites all plot close to other warm sites they form a group beyond the limits of the existing cloud of sites, and using this calibration all the Indian sites yield a climate prediction that is several degrees colder than that which is observed.
The next stage is to include the Indian observed climate information so that the shape of the plot will change and the ability of the method to give accurate results for fossils that represent ancient forests growing in warm climates does not suffer from the same error of underestimating temperatures.