TREES TELL OF PAST CLIMATES - BUT ARE THEY SPEAKING LESS CLEARLY TODAY

The annual growth of trees, as represented by a variety of ring-width, densitometric, or chemical parameters, represents a combined record o f different environmental forcings, one of which is climate. Along wit h climate, relatively large-scale positive growth influences such as h ypothesized 'fertilization' due to increased levels of atmospheric car bon dioxide or various nitrogenous compounds, or possibly deleterious effects of 'acid rain' or increased ultra-violet radiation, might all be expected to exert some influence on recent tree growth rates. Infer ring the details of past climate variability from tree-ring data remai ns a largely empirical exercise, but one that goes hand-in-hand with t he development of techniques that seek to identify and isolate the con founding influence of local and larger-scale non-climatic factors. By judicious sampling, and the use of rigorous statistical procedures, de ndroclimatology has provided unique insight into the nature of past cl imate variability, but most significantly at interannual, decadal, and centennial time-scales. Here, examples are shown that illustrate the reconstruction of annually resolved patterns of past summer temperatur e around the Northern Hemisphere, as well as some more localized recon structions, but ones which span 1000 years or more. These data provide the means of exploring the possible role of different climate forcing s; for example, they provide evidence of the large-scale effects of ex plosive volcanic eruptions on regional and hemispheric temperatures du ring the last 400 years. However, a dramatic change in the sensitivity of hemispheric tree-growth to temperature forcing has become apparent during recent decades, and there is additional evidence of major tree -growth (and hence, probably, ecosystem biomass) increases in the nort hern boreal forests, most clearly over the last century. These possibl y anthropogenically related changes in the ecology of tree growth have important implications for modelling future atmospheric CO2 concentra tions. Also, where dendroclimatology is concerned to reconstruct longe r (increasingly above centennial) temperature histories, such alterati ons of 'normal' (pre-industrial) tree-growth rates and climate-growth relationships must be accounted for in our attempts to translate the e vidence of past tree growth changes.