A Surfeit of Teeth

Last week the work I came here to do began in earnest with me giving a three hour, non-stop, seminar of the principles underpinning the CLAMP method of obtaining ancient climate information from fossil leaves. CLAMP stands for Climate Leaf Analysis Multivariate Program and was initially devised by my late post-doctoral mentor Jack A. Wolfe.

As a botany student at Imperial College in the early 1970s I had heard about Jack’s interest in the way that leaf form, particularly leaf margin architecture and size are related to the mean annual temperature and water availability in the place where they grow. Anxious to be involved in this research I elected to study under Jack when I obtained a LindemannPostdoctoral Fellowship. It also helped that Jack was based in California just south of San Francisco and, after all, this was in the wake of the “flower power” era.

My own research into the way that fossil leaf assemblages may, or may not, reliably reflect the vegetation from which they are derived had involved the use of, for that time, advanced multivariate statistics and computing. Jack, on the other hand was, well, not that way inclined. However over dinner we discussed the possibility of exploring whether or not more leaf features might carry more climate information.

In 1993, long after I returned to the UK, Jack published the first detailed account of CLAMP. In the intervening years he had doggedly set about sampling leaves from forests in different climate regimes and devising ways of mathematically describing the leaves in ways that captured climatic information. Subsequently we continued to develop and apply the technique until Jack’s death in 2005.

It is now clear that CLAMP is potentially a very powerful climate proxy and has been cross-calibrated with non-biological techniques such as oxygen isotope methods. However CLAMP in its current form is limited by the climate and type of modern vegetation used to calibrate it, being mostly based on observations in North America and Japan. More data are needed from sub-tropical and tropical forests and may even require a separate calibration for these types of climate altogether.

As we rattled along a dusty road to the Kukrail Forest site near Lucknow to take our first Indian CLAMP sample my colleague, Dr Rakesh Mehrotra, casually remarked that this was also a “crocodile reserve”. Suddenly my light-weight walking books that I had especially selected with ankle support to limit the damage done by insect or, at worst, snake bites, seemed woefully inadequate. After all I had not reckoned with invading the territory of large vertebrates with big mouths full of an indecent number of teeth.

Dr Rakesh Mehrotra and PhD student
Gaurav Srivastava collecting a
CLAMP sample in Kukrail Forest

The forest itself was tinder dry and at this point in the dry season many trees begin to shed their leaves. This is an ideal time to take a CLAMP sample as the leaves are mature and in a state where they would naturally be shed and perhaps enter the fossil record. I reassured myself that no respecting crocodile would appreciate such dryness.

 As we moved through the forest collecting all the observable different shapes and sizes of leaves from species after specie,s we eventually came to a small river where water buffalo were wallowing.  Surely no self-respecting buffalo would be doing such a thing if there were the slightest chance of being regarded as lunch…

As it happened we did see crocodiles, or more accurately Ghariyals (Ghavialis gangeticus) . The long-snouted crocodilians were once common in the Ganges and Brahmaputra Rivers, but there they were safely behind a wire fence in an area set aside for the breeding programme. Throughout the day we collected 29 species of tree, shrub and vine leaves and are now on our way to building up an Indian CLAMP calibration.

Preliminary analysis indicates that, with the existing calibration, CLAMP underestimates the temperature of the Lucknow area, just as we suspected it might. This is because many of the species in the Kukrail Forest, like the crocodiles, have teeth and toothed leaves are most often found in cooler climates. The difference between the CLAMP estimate 15.4±3.4°C and the measured mean annual temperature of 24°C is significant enough to justify continued sampling of forests in India to improve the technique.