Continental-Scale Patterns in Modern Wood Cellulose δ18O: Implications for Interpreting Paleo-Wood δ18O
Abstract
Ancient wood from Siberia and Arctic Canada is readily available and spans > 40 my of changing climate. Its isotopic composition might provide useful climate information, but there are several uncertainties that must be considered when interpreting the analyzed values of this potential resource. Aside from possible problems associated with diagenetic changes, oxygen isotopes in wood cellulose (δ18Ocel) are not likely to be good paleothermometers for the distant past when global patterns of precipitation δ18O (δ18Oppt) and temperatures were different, nor might they be good predictors of δ18Oppt because paleo-relative humidity, a contributor to δ18Ocel, is difficult to quantify. We sampled naturally growing, predominantly native trees in forested regions of North American and the Caribbean to determine how applicable the established relationships among δ18Ocel, temperature, relative humidity and δ18Oppt were at the continental scale, and how much relative humidity modifies the δ18Oppt signal. Different species at the same site may or may not have access to groundwater; they may have different phenological and physiological constraints that could cause them to have different δ18Ocel. Accordingly, we assessed species composition and depth to water table as potential sources of variability in δ18Ocel. We found up to 4‰ differences among species growing at the same site with conifer cellulose more enriched than angiosperm cellulose by 1.5‰ (P<.00001). Differences in landscape position, reflecting differing access to the water table, produced small (1‰) differences in δ18Ocel. While δ18Ocel was correlated with MAT (R2=.83-.91, P<.001), average summer minimum relative humidity (RH) combined with δ18Oppt explained most of the variability (R=.96) in δ18Ocel across North American and Caribbean forests. The isotopic enrichment that occurs during the processes leading to the incorporation of precipitation oxygen in cellulose is not constant, but varies from 35-44‰ and is directly correlated with the isotopic composition of the sourcewater. The relationships among δ18Oppt, RH and δ18Ocel established for N. America and the Caribbean applied reasonably well for most species on other continents, reflecting primarily the large spatial scale over which the relationship between temperature and δ18Oppt is similar, and that the tree species in this sample appear to fractionate sourcewater similarly. Accurate prediction of MAT and δ18Oppt from δ18Ocel requires RH. Predictions of MAT and δ18Oppt made from δ18Ocel alone (without RH) produced errors of up to 16°C and 8‰, respectively.
- Publication:
-
AGU Spring Meeting Abstracts
- Pub Date:
- May 2006
- Bibcode:
- 2006AGUSMPP51A..06R
- Keywords:
-
- 4900 PALEOCEANOGRAPHY (0473;
- 3344);
- 3344 Paleoclimatology (0473;
- 4900)