Assessing dendroecological methods to reconstruct defoliator outbreaks on Nothofagus pumilio in northwestern Patagonia, Argentina

PARITSIS, J, TT VEBLEN

Boulder, CO, EE.UU.

Conferencia; 59th Annual Western Forest Insect Work Conference; 2008

The history of insect outbreaks is crucial in understanding and managing forested landscapes. However, there is little quantitative information about past insect outbreaks for the temperate forests of the southern hemisphere. In the Patagonian Andes in Argentina, the southern beech Nothofagus pumilio experiences extensive defoliations caused by Ormiscodes (Saturniidae) species. Dendrochronology offers a method to reconstruct the history of past insect defoliations over broad time scales in forested landscapes. We evaluated the effectiveness of a native conifer, Austrocedrus chilensis, as a non-host species in reconstructing past Ormiscodes amphinome outbreaks on N. pumilio and we assessed the effectiveness of alternative and supplemental procedures to reconstruct these defoliations. We selected five sites for dendrochronological sampling (paired N. pumilio host stand and A. chilensis non-host stand). Correlation functions between monthly climate data and tree-ring chronologies of N. pumilio and A. chilensis were used to test for similar responses in tree ring growth to climatic variability. We used four methods to reconstruct the outbreaks: 1. Host vs. non-host correction (NH): A. chilensis used as non-host to correct for climate variation, 2. Host vs. host regional correction (HR): N. pumilio regional chronology used to correct for climate variation, 3. Detection of extreme tree-ring growth reductions using a kernel (K): Identification of narrow rings, and 4. Morphological variations in annual ring formation (S): Identification of a typical morphological anomaly caused by the defoliation. In addition to the 1986 defoliation, on two sites, 1878 appears to be an outbreak year, as three methods indicate defoliations on that year. A. chilensis appears to be more effective as a non-host species towards the more xeric sites compared to the more mesic sites (i.e. 1500-2500 mm) due to the similarities in its growth response to climatic variability compared to N. pumilio. To reconstruct insect defoliations on N. pumilio, we recommend application of all four of the reconstruction methods tested here. An outbreak event can be inferred when three or more methods indicate the same defoliation year. This study illustrates the importance of explicitly testing the assumption of a similar growth response between host and non-host species in the area in which the outbreaks will be reconstructed.