Appendix A. Description of the variables included in the structural equation model and figures showing the transformed relationships between species richness, evenness, and root, shoot, and litter biomass, and the initial structural equation model.
TABLE A1. Description of the variables included in the structural equation model.
Variable |
Description |
Species richness |
Number of vascular plant species in 0.5 m by 0.5 m permanent quadrats in July 2005. Square-root transformed. |
Initial species richness |
Number of vascular plant species in 0.5 m by 0.5 m permanent quadrats in May 2003. Square-root transformed. |
Evenness |
Evenness calculated from the cover data collected in 0.5 m by 0.5 m permanent quadrats in July 2005.using the "odds measure of evenness" (Simpson’s D`-1)/(Species richness -1) (Kvålseth 1991). Not transformed. |
Initial evenness |
Evenness calculated from the cover data collected in 0.5 m by 0.5 m permanent quadrats in May 2003 using the "odds measure of evenness" (Simpson’s D`-1)/(Species richness -1) (Kvålseth 1991). Not transformed. |
Shoot biomass |
Aboveground live biomass (g/m^{2}) clipped from a 0.1 by 1 m quadrat. Ln transformed. |
Root biomass |
Belowground biomass (g/m^{2}) washed from a 5.3 cm diameter by 12 cm deep core. Ln transformed. |
Litter biomass |
Aboveground dead biomass (g/m^{2}) collected from the 0.1 by 1 m quadrat used for shoot biomass. Ln transformed. |
Initial ground cover |
Proportion of the ground surface in May 2003 covered by litter plus bare soil. The amount of bare soil without cover by lichens or bryophytes in this system is very small relative to the cover or litter, so this value is a reasonable proxy for pre-treatment litter levels. Arcsine transformed. |
Light interception |
Light interception is the percentage of the total photosynthetically active radiation intercepted by the shade covers, aboveground biomass, and litter. Arcsine transformed. |
FIG. A1. The initial structural equation models. |
FIG. A2. The transformed relationships between species richness and evenness and shoot, root, and litter biomass. While the transformations did not eliminate the nonlinear relationships between species richness and root and shoot biomass, or between evenness and litter, the dominant pattern in all three cases is for a linear change over the range of biomass sampled. |
LITERATURE CITED
Kvålseth, T. O. 1991. Note on biological diversity, evenness, and homogeneity measures. Oikos 62:123–127.