¹²C/¹³C in cellulose and lignin as palaeothermometers

Abstract

SEVERAL workers^1–3 have attempted to retrieve palaeoclimatic information from the ¹²C/¹³C ratios in tree rings. Their use of whole wood, however, has been unsatisfactory, in that the extractives are isotopically much lighter than cellulose⁴ and vary considerably in content between heartwood and sapwood. Further, lignin is usually ∼ 3‰ lighter than cellulose and it is well known from the extensive literature on the chemistry of wood that the lignin-to-cellulose ratio varies not only across one ring, that is, in late wood as compared to early wood, but up and down and radially across the tree⁵. The science of dendroclimatology is based on the observation that narrow rings represent a cold (or dry) year. In most conifers the amount of late wood remains relatively constant and it is the thickness of early wood that varies. A narrow ring would be expected to have a smaller proportion of early wood than a wide ring. The different proportions of lignin in the two rings would then lead to a different isotopic ratio for the whole wood. If whole wood is used the δ¹³C measurements will reflect the relative proportions of lignin and cellulose. We therefore separated the lignin and cellulose from the wood before making δ¹³C determinations, hoping that the temperature coefficients of their production reactions might be large enough to enable the temperatures at which the carbon was fixed from the atmosphere to be measured. We also wished to investigate the possibility that the difference in δ¹³C between the cellulose and lignin might reflect temperature.