A multi-millennial temperature-sensitive tree-ring record derived from living and subfossil trees in the northern Patagonian Andes

BONINSEGNA, J.; ROIG, F.; BOTTERO, R; VILLALBA, R.

Mendoza

Conferencia; Third American Dendrochronology Conference. Ameridendro 2016; 2016

Old growth or late successional forests has long recognized as a valuable source of long tree-ring series useful for paleoclimatic studies. However, scarcely diversity of aged trees, a relatively short geographical distribution and the frequency and intensity of natural disturbances and harvesting history reduce the virgin tracts to areas left by lumbermen because of inaccessibility. In South America, a reduced suit of tree species has the attributes to be considered as long-lived trees. A relictual old-growth forest community of the long-lived tree Fitzroya cupressoides, distributed in a basin of nearly 200 km2 on the climatically cool-temperate and rainy north Patagonian Andes, is the source of woods in this study. Well-preserved dead logs were essential to substantially push the living records back several centennial to millennial age beyond the critical lifespan of the tree species. The dating of 188617 rings from 441 trees was carefully checked by cross-matched techniques giving a series inter-correlation of 0.513 (p < 0.05). As a result, a well-replicated tree-ring record, enough to provide past perspectives of fine-resolution growth, was obtained for the past 2,133 years (-130 BC-2002 AD). The living-subfossil composite chronology was confronted against other 9 of the same species and macro region yielding correlations between 0.39 and 0.91 (n=1087) indicating a good fitness of tree growth on a larger spatial scale. The MultiTaper spectral analysis estimates spectral peaks at 2.2-3.7, 23.8, 35.6, 47.6, 73.1, and 215.6 yrs. The composite chronology was subjected to a wavelet analysis to examine multidecadal patterns of growth variation. Oscillatory modes in the 30-70 yr domain persisted almost repeatedly along the last two millennia, while high power is evident during the period 0-1000 AD around the 256-yr band, but lack for the last millennia. Austral summer maximum temperature, for both prior and current growth period, was identified as the prevalent feature affecting tree growth for the calibration interval 1901-2002. To sort any drawback in a time-dependent interpretation of this relationship, an applied evolutionary response function indicates the stability of this climate-tree growth relationship. The significance of this proxy-climate record for paleoclimate inferences will be discussed.