Jump to Main Content
Tree growth and competition in an old‐growth Picea abies forest of boreal Sweden: influence of tree spatial patterning
- Fraver, Shawn, D'Amato, Anthony W., Bradford, John B., Jonsson, Bengt Gunnar, Jönsson, Mari, Esseen, Per‐Anders, Turnbull, Lindsay
- Journal of vegetation science 2014 v.25 no.2 pp. 374-385
- tree growth, kriging, inventories, linear models, boreal forests, regression analysis, Picea abies, trees, dendrochronology, stand basal area, tree mortality, mortality, prediction, Sweden
- QUESTION: What factors best characterize tree competitive environments in this structurally diverse old‐growth forest, and do these factors vary spatially within and among stands? LOCATION: Old‐growth Picea abies forest of boreal Sweden. METHODS: Using long‐term, mapped permanent plot data augmented with dendrochronological analyses, we evaluated the effect of neighbourhood competition on focal tree growth by means of standard competition indices, each modified to include various metrics of trees size, neighbour mortality weighting (for neighbours that died during the inventory period), and within‐neighbourhood tree clustering. Candidate models were evaluated using mixed‐model linear regression analyses, with mean basal area increment as the response variable. We then analysed stand‐level spatial patterns of competition indices and growth rates (via kriging) to determine if the relationship between these patterns could further elucidate factors influencing tree growth. RESULTS: Inter‐tree competition clearly affected growth rates, with crown volume being the size metric most strongly influencing the neighbourhood competitive environment. Including neighbour tree mortality weightings in models only slightly improved descriptions of competitive interactions. Although the within‐neighbourhood clustering index did not improve model predictions, competition intensity was influenced by the underlying stand‐level tree spatial arrangement: stand‐level clustering locally intensified competition and reduced tree growth, whereas in the absence of such clustering, inter‐tree competition played a lesser role in constraining tree growth. CONCLUSIONS: Our findings demonstrate that competition continues to influence forest processes and structures in an old‐growth system that has not experienced major disturbances for at least two centuries. The finding that the underlying tree spatial pattern influenced the competitive environment suggests caution in interpreting traditional tree competition studies, in which tree spatial patterning is typically not taken into account. Our findings highlight the importance of forest structure – particularly the spatial arrangement of trees – in regulating inter‐tree competition and growth in structurally diverse forests, and they provide insight into the causes and consequences of heterogeneity in this old‐growth system.