Abstract

This study evaluated early mortality patterns of planted Nothofagus dombeyi (Mirb.) Blume (evergreen, shade intolerant) and Nothofagus alpina (P. et E.) Oerst. (deciduous, mid-tolerant to shade) seedlings established in scarified gaps in a Nothofagus old-growth forest in the Chilean Andes that was subjected to a shelterwood cut. Soils were scarified with heavy machinery that removed up to 40 cm of the upper horizons (e.g. it completely removed the litter layer and organic horizon, and in some cases partially the A horizon) to eliminate bamboo (Chusquea culeou E. Desv) shoots and rhizomes from the scarified areas. Two years after planting, seedlings were growing free of understory competition and mortality patterns were assessed using logistic regression models with the following predictor variables: (1) basal diameter growth in the prior year, and (2) soil strength measured as resistance to penetration. The probability of mortality of both species was highly predictable based on recent basal diameter growth. N. dombeyi seedlings appeared to be less affected by mortality, as reflected in lower sensitivity of mortality at lower growth rates compared to N. alpina. High values of resistance to penetration (> 2,000 kPa) induced a higher probability of mortality in N. alpina (up to 40 % at 3,500 kPa). In contrast, N. dombeyi mortality was not significantly influenced by soil strength, which illustrates the more stress-resistant and pioneer character of this species. This study showed that recent growth rates can be used as a predictor of future mortality, but these relationships need to be calibrated for each species. In addition, our results suggest that species characteristics that relate to resource limitations, such as shade tolerance, are not adequate to understand effects of other stressors, such as soil disturbance, on mortality patterns.