Recovery of understory assemblage along 50 years after shelterwood cut harvesting in Nothofagus pumilio Southern Patagonian forests.
Harvesting modifies the understory due to changes in the overstory structure, microclimate and soil properties. The impact is directly related to the harvesting intensity, although studies are usually developed in logged areas (low impact) avoiding stockpiled areas (high impact). Understory resilience is quite variable, and recovery time remains unknown for different management practices and different temperate forests. The objectives of this work were: (i) to evaluate recovery of understory vascular plants and forest structure in Nothofagus pumilio forests of South Patagonia (Argentina) along a chronosequence of years-after-harvesting (YAH) (1-5, 6-10, 20-30, >50) in shelterwood cuts compared to unharvested forests (UF), and (ii) to analyze differences in the understory recovery in areas with different impact intensity due to the harvesting: low impact harvesting (LIH) in logged areas vs. high impact harvesting (HIH) in stockpiled areas. We selected 12 harvested areas with different YAH, and in each area we sampled three forest treatments (LIH, HIH, UF) with 5 replicas (N=12×3×5=180). In each sampling unit we analyzed the forest structure (primary or remnant overstory and secondary forests), environmental (soil moisture) and understory variables (richness, cover and species assemblage patterns) through univariate and multivariate statistical analyses. Remnant forest structures presented significant decay in the dominant height along the chronosequence, and an increasing in the secondary forest variables, as well as in the crown cover (remnant + secondary forests). Lower values of forest structure variables were observed in HIH than LIH, but without significant differences between them. Most of treatment levels presented significant differences with UF, and values become similar with time. Understory presented significant differences in richness and cover along the chronosequence, with values become more similar to UF on time. Harvesting impact levels also presented differences, e.g. 104% and 144% more exotic species cover in LIH and HIH compared with 20% in UF. However, greater significant differences were found among treatments for their species assemblage patterns, mainly in the first stages of the chronosequence (e.g., between UF and LIH at 6-10 YAH); meanwhile in the last stage (>50 YAH), species assemblage patterns become similar with UF, both for LIH and HIH. We conclude that understory vascular plant assemblage of N. pumilio forests recovers patterns after 50 YAH, being more similar to original ones as the crown cover reaches similar values to unharvested forests, showing N. pumilio forests as highly resilient to shelterwood cut harvesting, even in high impacted areas.