Resource gradients affect competitive interactions of tree seedlings under a heterogeneous canopy

Background/Question/Methods

Forest management techniques that emulate natural disturbance regimes have become increasingly popular as a tool for increasing forest complexity and diversity, yet still remain largely untested.  We lack information about how the spatial pattern of canopy trees will alter resource heterogeneity in the understory and therefore competitive interactions and productivity of regeneration.   Species with different light and nutrient demands should respond very differently to the variation in competitive environments created with different patterns of tree retention after thinning.  In this large scale and long-term experiment conducted in red pine forests in Minnesota, USA, we used a likelihood approach to model the growth of jack, red and white pine seedlings as a function of resource availability (light and nutrients) and competition from residual trees and understory shrubs in stands with different spatial patterns of canopy trees.  We implemented a randomized-block design, replicated four times.  Each block consisted of four ~16 ha treatment stands: three overstory treatments (a thinning with residual trees dispersed evenly throughout the stand and patch cutting that created either 0.1 or 0.3 ha gaps in a forest matrix) and an unharvested control.  We measured light and soil nutrients throughout our treatment stands and mapped trees in 16.1 m diameter plots around each sample point.  Planted seedlings were measured repeatedly over 5 years. 

Results/Conclusions

After preliminary analyses, final models explained 40-75% of the variation in individual seedling growth, depending on the species, but all models included significant effects of light, tree and shrub competition, and soil resources.  Models revealed non-linear declines in diameter growth with decreased light and increased competition from neighboring trees.  Soil resources modified the effects of competition on tree growth. Competitive effects on red pine and white pine seedlings intensified on high nitrogen and high phosphorus soils, respectively.  Interestingly, competitive effects on jack pine growth were greater on high nitrogen but low phosphorus sites.  The nonlinear and species-specific responses of seedling growth to multiple abiotic and biotic factors supports the idea that overstory pattern can be manipulated to increase complexity and species diversity in forests.  Neighborhood models provide an effective mechanistic approach to understanding complex ecological processes in heterogeneous ecosystems.