Fig 1.
Comparison of the continuum index (left) and shade tolerance index (right) in southern Wisconsin.
Similarly to the original study [24], the species were split in two groups: major (top) and lesser (bottom) species. Bars indicate the standard error of the mean.
Fig 2.
Conceptual model of shade tolerance index dynamics after a major disturbance.
The initial proportion of seedling of shade tolerant and intolerant species can be very different due to the nature of the disturbance and biological legacies. Shade intolerant species outgrown shade tolerant species during the stem exclusion stage of the forest stand development, leading to the index decrease. Shade tolerant species gain an edge over shade intolerant species via canopy gap dynamics, resulting in the index increase during the understory reinitiation stage, and its further stabilization as the forest approaches the old growth stage.
Fig 3.
Computer simulation of White Pine—Eastern Hemlock forest stand.
Shade tolerance index as a function of the implied stand age (left) and cumulative basal area of White Pine and Eastern Hemlock (right) as a function of the implied stand age. The simulation is conducted using the crown plastic SORTIE model following [62] for one hectare and 1000 years, the parameter values are similar to the original paper except the mortality of both species was reduced to 5%.
Fig 4.
Approximated chronosequence of White Pine—Eastern Hemlock forest stands in the northeastern US.
Analyzed data consist of 1375 USDA FIA plots, where these species account for more than 75% of the total basal area (S3 Appendix). Shade tolerance index (left) and cumulative basal area of White Pine and Eastern Hemlock (right) as a function of the implied stand age (S1 Appendix). The shaded areas represent the standard error of the mean. The last bin “120+” includes 3% of the plots.
Fig 5.
The stand-level characteristics of plots for all years demonstrate very heterogeneous forest types in the US, with no obvious common distribution pattern between indicators.
Maps of subsets of plots from different decades are similar and can be found in S1 Appendix.
Fig 6.
Representations for each ecoregion with n > 150 re-sampled plots of: -the shade tolerance index transition matrices (tables); -the current distribution of states (bar graphs).
In each transition matrix, the value located in row i and column j is the probability of transition from state i to state j after 3 years (expressed in percents). The current distribution of the shade tolerance index is important to interpret these matrices, as transitions affecting the most numerous shade tolerance states have the most influence on future distributions. We indicate these distributions with a bar graph to the right of each matrix, and report it as colored rows in the matrices for easier reading (from white to red, for sparsely to largely represented shade tolerance index states). Placements have been chosen to approximately reflect geography, with northeastern ecoregions located in the top-right corner of this figure.
Fig 7.
Classification of ecoregions based on shade tolerance distribution width and spectral gap.
Top: the disparity of shade tolerance is measured as the length of interval encompassing 95% of the distribution. Middle: the spectral gap of the transition matrices quantifies the convergence rate after a disturbance, where low values mean that one plot is quick to reach equilibrium. Bottom: classification of ecoregions. Provinces where forest succession is primarily driven by shade tolerance exhibit a wide distribution of shade tolerance index as well as a fast convergence toward equilibrium.