Short-term effects of variable-density thinning on regeneration in hardwood-dominated temperate rainforests

Abstract

Second-growth temperate forests usually have simpler composition and structure than comparable old-growth stands. We evaluated the application of variable density thinning (VDT) as a way to increase old-growth attributes, specifically tree species composition and density in two types of second-growth forests: a mixed evergreen-dominated forest (SE, stem exclusion phase) and a Nothofagus dombeyi-dominated forest (UR, understory reinitiation stage) in southern Chile (40 degrees S Lat.). Six one-ha plots were established in each forest. We applied VDT in three plots in each forest, which included the following sub-treatments: a thinned matrix (M), large and small patches (LP and SP), and large and small reserves (LR and SR). We also established three one-ha plots in an old-growth forest (OG), which served as a reference for overall stand structural variables and the tree regeneration layer. OG had significantly more total regeneration and shade-tolerant regeneration, than the untreated SE and UR. After VDT, canopy openness was significantly greater in SE than in UR within the treated portions of the plots (M, LP and SP). The similarities in stand characteristics between untreated and pre-harvest treated plots in each second-growth forest suggest that harvesting was responsible for the lower densities observed in thinned plots. Regeneration patterns two growing seasons after harvests were reflective of the interaction between forest type and sub-treatments. While both second-growth forests showed increases in regeneration densities, none of these increases was significant, although proportionally and numerically they were greater in the SE forest. The sub-treatments showed differences only for shade-intolerant and shade-tolerant species in both forests. These differences in general illustrated a preference of shade-intolerant species for patches (only in the smaller height class in both forests), and of shade-tolerant species for the matrix or the reserves in all height classes from > 50 cm through saplings (only in the SE forest). Overall, tree regeneration has been more responsive to VDT in the forest currently in an earlier successional stage, suggesting that second-growth forests in more advanced stages may need to be treated more intensively (larger patches, lower residual densities). Correspondingly, VDT protocols must be developed to account for the interaction between harvest disturbances, developmental stage, and silvics of the component species.

Second-growth temperate forests usually have simpler composition and structure than comparable old-growth stands. We evaluated the application of variable density thinning (VDT) as a way to increase old-growth attributes, specifically tree species composition and density in two types of second-growth forests: a mixed evergreen-dominated forest (SE, stem exclusion phase) and a Nothofagus dombeyi-dominated forest (UR, understory reinitiation stage) in southern Chile (40 degrees S Lat.). Six one-ha plots were established in each forest. We applied VDT in three plots in each forest, which included the following sub-treatments: a thinned matrix (M), large and small patches (LP and SP), and large and small reserves (LR and SR). We also established three one-ha plots in an old-growth forest (OG), which served as a reference for overall stand structural variables and the tree regeneration layer. OG had significantly more total regeneration and shade-tolerant regeneration, than the untreated SE and UR. After VDT, canopy openness was significantly greater in SE than in UR within the treated portions of the plots (M, LP and SP). The similarities in stand characteristics between untreated and pre-harvest treated plots in each second-growth forest suggest that harvesting was responsible for the lower densities observed in thinned plots. Regeneration patterns two growing seasons after harvests were reflective of the interaction between forest type and sub-treatments. While both second-growth forests showed increases in regeneration densities, none of these increases was significant, although proportionally and numerically they were greater in the SE forest. The sub-treatments showed differences only for shade-intolerant and shade-tolerant species in both forests. These differences in general illustrated a preference of shade-intolerant species for patches (only in the smaller height class in both forests), and of shade-tolerant species for the matrix or the reserves in all height classes from > 50 cm through saplings (only in the SE forest). Overall, tree regeneration has been more responsive to VDT in the forest currently in an earlier successional stage, suggesting that second-growth forests in more advanced stages may need to be treated more intensively (larger patches, lower residual densities). Correspondingly, VDT protocols must be developed to account for the interaction between harvest disturbances, developmental stage, and silvics of the component species.