Respective importance of ecological conditions and stand composition on Abies alba Mill. dominant height growth

In mixed stands, abiotic effects and interspecific interactions on tree height growth remain largely unknown. We investigated the effects of climate, nutrition and stand composition on silver fir (Abies alba Mill.) height growth in the Vosges Mountains (northeastern France). Data were collected from 143 plots established on seven stand types of silver fir (Abies alba Mill.): pure silver fir stands and six mixed stands of silver fir with Fagus sylvatica, Picea abies, Pinus sylvestris, Quercus petraea and Acer pseudoplatanus, along both elevation and nutritional gradients. Silver fir dominant height was characterised by the site index, which was related by multiple-regression analyses with measured nutritional variables, climatic variables and stand composition descriptors. The results showed that silver fir site index was significantly correlated with climate, nutrition and stand composition. Among climatic variables, elevation was responsible for a large portion of the site index variance (R2 = 0.13). Slope aspect showed an important positive effect of eastern oriented plots on silver fir site index, as did mild winter temperature. Variables related to water availability showed a positive effect on site index only at elevations below 600 m, where lower precipitations and higher temperatures lead to a water-limiting factor. Nutritional resources were the second factor determining silver fir site index by both nitrogen nutrition and protons and aluminium acidity or toxicity. An environmental model, including climatic and nutritional variables, explained 64% of site index variance with a standard error of 2.7 m. For equivalent sites, basal area of the dominant cohort, as well as basal area of conifer species, showed a positive effect on site index, while broadleaved basal area did not show any significant effect in addition to that of conifers. A final model including abiotic variables and conifer basal area explains 71% of site index variance with a standard error of 2.5 m in the study area. The search for ecological drivers of tree height that control access to light is of great importance in the context of environmental changes in predicting tree species relationships in mixed stands.