Soil matters: soil variables remarkably increase explained variation - - PowerPoint PPT Presentation

Soil matters: soil variables remarkably increase explained variation of tree beta diversity in forest dynamics plot

David Zelený1, Li-Wan Chang2,3, Ching-Feng Li1, Shau-Ting Chiu3,4 and Chang-Fu Hsieh2

1Department of Botany and Zoology, Faculty of Sciences, Masaryk University, Czech Republic 2Institute of Ecology and Evolutionary Biology, National Taiwan University, Taiwan 3Technical Service Division, Taiwan Forestry Research Institute, Taiwan 4Department of Biology, National Museum of Natural Science, Taiwan

niche-based processes dispersal-based processes

Community assembly

variation explained by environment species composition environmental data (e.g. pH) RDA, CCA

Variance partitioning of species composition among environmental and spatial variables

variation explained by space species composition spatial variables (e.g. PCNM) RDA, CCA

Variance partitioning of species composition among environmental and spatial variables

Explanation of community spatial distribution

may be everything is just dispersal...

Explanation of community spatial distribution

• r it could be perfectly explained by

environment?

Explanation of spatial distribution

• r may be both environment and

dispersal?

Explanation of spatial distribution

+

• r may be we just haven’t measured

everything?

Explanation of spatial distribution

+

variation explained by space variation explained by environment species composition spatial variables (e.g. PCNM) environmental data (e.g. pH) RDA, CCA RDA, CCA

Variance partitioning of species composition among environmental and spatial variables

species composition spatial variables (e.g. PCNM) environmental data (e.g. pH) RDA, CCA RDA, CCA

Variance partitioning of species composition among environmental and spatial variables

variation explained by environment variation explained by space

[a] [b] [c] niche-based processes ([a]+[b]) dispersal-based processes ([c])

PCNM variables generated from grid data

http://zmrzlina-misa.cz/retro

PCNM as a kaleidoscope...

http://wikimedia.org

Context:

• variance partitioning among environment and space
• applied on vegetation data from large forest permanent plots
• interpretation in terms of niche- vs dispersal-based processes

Questions:

• How will interpretation of variation partitioning change if we

measure better/more environmental variables?

• How to know if we measured everything important and

what’s left is really caused by dispersal?

Gutianshan forest dynamics plot (24 ha, China)

Gutianshan forest dynamics plot (24 ha, China)

• Fig. 3b from Legendre et al. (2009)

Ten forest dynamics plots around the World

Chang et al. (2012): Linehuachih Subtropical Evergreen Broadleaf Forest Dynamics Plot: Tree Species Characteristics and Distribution Patterns. Taiwan Forestry Research Institute.

Lienhuachih Forest Dynamics Plot (25 ha, Taiwan)

[a] = 0.011 [b] = 0.234 [c] = 0.408 Unexplained (residual) = [d] = 0.346

Gutianshan (China) Lienhuachih (Taiwan)

• nly

topography

[a] = 0.011 [b] = 0.234 [c] = 0.408 Unexplained (residual) = [d] = 0.346 [a] = 0.017 [b] = 0.473 [c] = 0.170 Unexplained (residual) = [d] = 0.340

Gutianshan (China) Lienhuachih (Taiwan)

• nly

topography topography and soil (16 variables, including C, N, C/N, pH, K, Ca, Mg, Fe, Mn, Cu, Zn and P, water content and texture)

broad scale – environmental variables fine scale – dispersal processes

broad scale – environmental variables fine scale – dispersal processes

Chang, Zelený, Li, Chiu & Hsieh (accepted): Better environmental data may reverse conclusions about niche- and dispersal-based processes in community

• assembly. Ecology

Thank you for your attention!

The study was financially supported by: Taiwan Forestry Research Institute (97 AS-7.1.1.F1-G1) Taiwan Forestry Bureau (tfbm-960226) Czech Science Foundation (P505/12/1022)