Abiotic and biotic determinants of height growth of Picea abies regeneration in small forest gaps in the Swiss Alps

In managed mountain forests, height growth of Norway spruce (Picea abies (L.) Karst) regeneration is a decisive factor for the gap-filling process, especially when the silvicultural goal is to provide continuous protection against natural hazards such as snow avalanches or rockfall. For the planning of management interventions, robust predictions on height growth of regeneration at the scale of forest gaps are thus needed. However, there is a lack of such data for trees of intermediate sizes, and existing studies fail to cover large environmental gradients. The goal of our study is to identify the key factors that influence height growth of Norway spruce regeneration in small gaps of spruce-dominated forests in the Swiss Alps. Furthermore, we assess whether there are site-specific differences of height growth or whether it follows a similar pattern along a large gradient of temperature (i.e., along elevation) and of water and nutrient availability (i.e., among different phytosociological site types) within the upper montane and subalpine vegetation belts. On 124 plots, >2′000 observations of annual height increments of Norway spruce regeneration (10 cm tree height to 12 cm stem diameter) in gaps were collected. Using linear mixed effects models and cross-validation for model selection, we identified the best variable combinations to predict annual height growth. Consistently across the entire gradient, the most important factors were 1) the positive effect of tree size, 2) the negative effect of competition by the surrounding stand, and 3) local topography. We found site-specific differences in height growth patterns such as gap size and therefore direct radiation being the most important competition measure in subalpine sites, as opposed to diffuse radiation in high montane sites. However, the pooled model for the entire environmental gradient allowed for predictions of regeneration height growth with similar explanatory power as the more specific models while containing comparable effect sizes. Furthermore, competition can be equally well expressed by metrics based on basal area measurements as by metrics derived from hemispherical photography. Based on these relatively simple models, accurate and robust predictions of the development of Norway spruce regeneration in gaps of managed mountain forests are possible.