Understorey vegetation comprises the majority of species diversity and contributes greatly to ecosystem functioning in boreal forests. Although patterns of understorey abundance, species diversity and composition associated with forest stand development are well researched, mechanisms driving these patterns remain largely speculative. We sampled fire‐origin stands of varying stand ages and overstorey compositions on mesic sites of the boreal forest of Canada and used structural equation modelling (SEM) to link time since fire (stand age), light availability and heterogeneity, substrate heterogeneity and soil nitrogen to understorey vegetation cover and species diversity. The most parsimonious model for total understorey cover showed a positive direct effect of stand age (r = .43) and an indirect effect via mean light level (0.18) and shrub cover (−0.11), with a positive total effect (0.50); the per cent broadleaf canopy had a direct negative effect (−0.22) and an indirect effect via shrub cover (−0.11). The model for total understorey species richness showed an indirect effect of stand age via mean light (0.24), light heterogeneity (0.10) and substrate heterogeneity (0.07), with a positive total effect (0.52); per cent broadleaf canopy had an indirect effect via light heterogeneity (0.09), and substrate heterogeneity (−0.10). Soil nitrogen did not significantly influence either understorey cover or species richness. The models for vascular plants followed similar trends to those for total understorey cover and species richness; however, there was an opposite indirect effect of light heterogeneity for both cover and species richness of non‐vascular plants. Shrub cover had positive direct and negative direct and indirect effects on both vascular and non‐vascular cover and species richness. Synthesis. Our findings indicate that understorey cover and species diversity are driven by time since disturbance, light availability as influenced by overstorey and shrub layers, but with important additional effects mediated by light and substrate heterogeneity. Non‐vascular understorey vegetation is more strongly determined by time since disturbance than vascular vegetation, and negatively affected by broadleaf tree abundance. The overall results highlight the importance of colonization, light availability and heterogeneity, substrate specialization and growth dynamics in determining successional patterns of boreal forest understorey vegetation.

No abstract

No abstract

No abstract

Question Although the importance of coarse woody debris (CWD) for understorey species diversity has been recognized, the relative effects of coarse woody debris decay class and substrate species on understorey species composition have received little attention. We examined how the species composition of understorey vegetation change with CWD decay class and substrate species. Location Boreal mixed‐wood forests, Ontario, Canada. Methods To cover a wide range of CWD decay classes and substrate species, we sampled fire‐origin boreal forest stands that varied in stand age and canopy tree species composition. Vegetation on CWD was sampled by visually estimating percentage cover of each species within a 0.1 m × 0.5 m quadrat, randomly laid lengthwise on top of each sampled CWD log. We also recorded the forest floor vegetation by establishing an adjacent plot of the same size at a distance of 1.0 m in a random direction from the CWD vegetation sample. Results Multivariate analysis showed that understorey species composition differed among decay classes and substrate species. A NMDS ordination of understorey species composition revealed a clear separation of decay classes 1 and 2 from higher decay classes, and that decay classes 4 and 5 shared several species with the forest floor. The species composition on the forest floor was completely different from the species composition on CWD decay classes 1, 2 and 3. Two distinct groupings of substrates according to CWD species composition were found: conifer species (Pinus banksiana and Picea spp.) and broad‐leaf species (Betula papyrifera and Populus spp.), with Abies balsamea taking an intermediate position. Indicator species analysis showed distinct understorey species affiliations to substrate species at advanced decay classes. Understorey species composition on the CWD of P. banksiana showed particularly pronounced changes from the dominance of lichens on decay classes 2 and 3 to dominance by mosses and vascular species on decay classes 4 and 5. Conclusions Understorey species composition on CWD not only differed with decay class, but also with CWD substrate species. Conservation strategies should aim at retaining diversity of CWD in terms of both decay classes and species composition in boreal forests.