Lieubeau Vincent, Genthon Pierre, Stievenard M., Nasi R., Masson-Delmotte V. (2007). Tree-rings and the climate of New Caledonia (SW pacific) preliminary results from Araucariacae. Palaeogeography Palaeoclimatology Palaeoecology, 253 (3-4), p. 477-489. ISSN 0031-0182.
Titre du document
Tree-rings and the climate of New Caledonia (SW pacific) preliminary results from Araucariacae
Lieubeau Vincent, Genthon Pierre, Stievenard M., Nasi R., Masson-Delmotte V.
Source
Palaeogeography Palaeoclimatology Palaeoecology, 2007,
253 (3-4), p. 477-489 ISSN 0031-0182
The dendroclimatologic potential of some Araucariacae of New Caledonia (including Agathis, or kauris, and Araucaria) is assessed using ring thickness and delta O-18 measurements. New Caledonia is a group of islands in the SW Pacific that are currently under influence of ENSO events. Endemic to New Caledonia, the long-living species of Agathis lanceolata and A. ovata, growing on poor ultramafic-detived soils may provide valuable proxies for the local climate and for ENSO. These trees present visible growth bands of changing thickness along their circumference. However, several bands are locally absent, and the growth axis is generally offset with respect to the geometrical axis of the tree. This led us to compute so-called composite ring thickness profiles, accounting for the geometry of growth bands on the whole surface of a tree disk. Our computational method involves 10 optical density profiles measured along 10 equally spaced radii drawn from the bark toward the growth axis, and 10 to 20 master rings, that can be easily identified on the whole disk. Growth bands visible on less than 5 radii were discarded. Our method is similar to the cross-dating method used by dendrochronologists, except that it is applied here to a single tree disk. Our samples consist of three disks of A. lanceolata, one disk of A. ovata, and one disk of Araucaria columnaris. Multiple regressions have been computed between composite profiles and climatic variables i.e. monthly and yearly temperatures and rainfall amounts. The best correlation is found between the width of the ring growing between July (n-1) and June (n) with the rainfalls of June (n), June (n-1) and June (n-2). Monthly rainfalls allow to explain between 20% to 50% of the ring thickness variance, a result similar to that obtained with other studies on Agathis of New Zealand. No temperature parameter appears in the most stable regressions. 30 measurements of tree ring cellulose delta O-18 have been conducted on one single disk selected for the strong climate-ring width correlation. While earlier studies have used 6180 measurements to identify seasonal cycles in tropical woods and date the rings, our data suggest that the direct use of delta O-18 is misleading due to false rings that do not correspond to a complete growth year. When these false rings are identified from the disk analysis and discarded, a fair visual correlation with the total rainfall during the growth season is obtained. This requires information that cannot be found in single growth band thickness profiles, for example as obtained by coring. Thus, Araucariacae of New Caledonia may present a valuable potential for dendroclimatology. However, reconstructing a chronology of this region will require more extensive sampling and possibly an account of additional species.
