I'm trying to find differences between habitats based on the traits of the species living there in order to prove whether alien species are present in some habitats for having some specific traits. In you opinion, which statistical method is more suitable for this purpose?
These papers could be helpful:
Ecological Monographs, 74(4), 2004, pp. 635–662: MULTIPLE TRAIT ASSOCIATIONS IN RELATION TO HABITAT DIFFERENTIATION AMONG 17 FLORIDIAN OAK SPECIES
Plant, Cell & Environment Volume 34, Issue 9, pages 1536–1547, September 2011: Morphological and physiological differentiation of seedlings between dry and wet habitats in a tropical dry forest
Annals of Botany 110: 1429 – 1438, 2012: Seedling traits, plasticity and local differentiation as strategies of invasive species of Impatiens in central Europe
I would start with ordinations to see if there are differences in species composition between habitats. Then I'd search for the subset of species, which is best at separating the habitats. From these species I'd then take the traits and run a PCA.
These papers could be helpful:
Ecological Monographs, 74(4), 2004, pp. 635–662: MULTIPLE TRAIT ASSOCIATIONS IN RELATION TO HABITAT DIFFERENTIATION AMONG 17 FLORIDIAN OAK SPECIES
Plant, Cell & Environment Volume 34, Issue 9, pages 1536–1547, September 2011: Morphological and physiological differentiation of seedlings between dry and wet habitats in a tropical dry forest
Annals of Botany 110: 1429 – 1438, 2012: Seedling traits, plasticity and local differentiation as strategies of invasive species of Impatiens in central Europe
I would start with ordinations to see if there are differences in species composition between habitats. Then I'd search for the subset of species, which is best at separating the habitats. From these species I'd then take the traits and run a PCA.
Zonal statistics in GIS also could be helpful. For this, you need provide species distribution maps along with all rasters related to habitat traits. Then zonal statistics will help you to find those traits which are more important. Data mining
Also could be helpful. Back stepwise regression method and PCA also may be fruitful.
If you are interested in the relationship between habitat characteristics and species traits, RLQ and/or fourth-corner analyses may help you.
For RLQ you can look at:
http://link.springer.com/article/10.1007%2FBF02427859
http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2745.2010.01743.x/abstract (adding phylogenetic and spatial signal)
For fourth-corner you can look at:
http://www.esajournals.org/doi/abs/10.1890/08-0349.1
There must be more up-to-date papers, but these ones provide the background.
Very informative replies and links. It is good to learn about design in special case
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Environmental Science