Ecoinformatics Revisited September 22, 2005

 

Past Topics Addressed During Ecoinformatics Meetings:

Environment and species composition correlations (Peet)

Neutral model design (Jobe)

Phylogenetics (Ott)

Plot methods and Stohlgren evaluation (Wentworth, White, Fridley, Peet)

Specialist vs generalist species (Vandermast, Kuppinger, Fridley, Peet)

Species-area relationships (Fridley, Wentworth, White, Gramling, Peet)

Species-richness at the continental level (Fridley, Weakley, White, Peet)

Species pools (Gramling, Wentworth, Fridley, Peet)

Richness, rarity, uniqueness, and endemism (Peet, Weakley, Wentworth, etc.)

Manuscripts from CVS Ecoinformatics Work:

Fridley, J.D., D.B. Vandermast, D. Kuppinger, R.K. Peet and M. Manthey  (In preparation) How wide the niche? Scale-dependence assessment of habitat generalists and specialists using survey co-occurrence of Southeastern U.S. woody flora.

Wentworth, T.R. et al. (In limbo?) Nested vs. non-nested characterization of vegetation composition and species richness at multiple spatial scales.

Fridley, J.D., R.K. Peet, T.R. Wentworth and P.S. White. 2005. Connecting fine- and broad-scale patterns of species diversity: species-area relationships of Southeastern U.S. flora. Ecology 86:1172-1177.

Peet, R.K., J.D. Fridley and Joel M. Gramling. 2003. Variation in species richness and speices pool size across a pH gradient in forests of the southern Blue Ridge Mountains. Folio Geobotanica 38:391-401.

Assigned tasks:

 


A plant community-based "ecoinformatics" shopping list

Originally compiled by Jason Fridley and Bob Peet, the list has and should continue to evolve so that it reflects the collective insight of the group rather than that of any one person.


Questions below are based on combinations of three data types:

a) composition data (plot x species matrix)
- implied that plot spatial distribution known
- many sampling scales per plot rare

b) environmental data (plot x env descriptor matrix)
- GIS-based data can be derived
- site-sampled data (eg, soils) rare

c) trait data (species x trait matrix)
- taxonomic data usually available (eg, Kartesz)
- life-history data rare
- geographic distribution data


Questions, group A: Composition data only

1. Are species distributions more clustered than that predicted by a "neutral model" (sensu Hubbell, Bell)?

2. What is the distribution of relative abundance; what is the relationship between local and global abundance? What is the correlation between local abundance and geographic range? How does abundance vary with location in range (test of the Whittaker bellcurves).

3. How do species accumulate with area (z values)?

4. Co-occurrence analysis:

4a. Species pools: Can "species pools" be identified in a quantitative way? At what scale? What is the relationship between regional (species pool) richness and local richness? Do species pools "exist" outside a plot-centered context ("hard" vs. "soft" species pool concept)? [Joel Gramling and I have been tinkering with these questions, as has Tom]

4b. Generalist vs. specialist species: given equal sampling effort, what is the distribution of total co-occurrences per species? [Dave Vandermast, Dane Kuppinger and I have been working on this, also on question 7, below]

4c. Nestedness: to what extent are some communities nested subsets of other communities, and where does nestedness tend to occur?

4d. Explore co-occurrences at different spatial scales. [One woman from the Smithsonian is doing a fascinating study of phylogenetic co-occurrences of oaks in Florida, and finding that closely related species co-occur less often than expected. But she only sampled oaks at one scale--we could use different scales to ask whether closely related species co-occur even less often as scale decreases. In other words, we'd be using scale as a proxy for mechanism, similar to how Becky used it to examine the native-exotic richness question.]

5. Spatial autocorrelation. Look at the similarity of 10x10s in a CVS PLOT and how this drops off with proximity to each other. Could vary with topography, size of trees, etc. Would be easier if the R data were split by module. We are constrained to looking at the 10x10s agains the average for the Rs and the other 10x10s.


Questions, groups A and B: Composition and environment data

5. Species/area and environment:

5a. Large scales: To what extent do environmental factors (and which) constrain species accumulation as compared to random (area only) species accumulation? ("proximity" analysis) [I've been playing with this]

5b. Small scales: How do environmental factors (and habitat types) influence small-scale (within-plot) rates of species accumulation (z values)?

6. What is the relationship between species richness and environmental variables?

7. (Related to question 4b): Do "specialist" species occur in certain types of habitats?

8. (Related to questions 4a and 4c): Does environment influence the regional-local richness relationship or nestedness?

9. (Related to question 2): Does environment influence relative abundance distributions?

10. Niche breadths (based on environment):

10a. Do niche breadths consistently differ between adults and juveniles (trees)?

10b. How is a generalist-specialist analysis based on co-occurrences related to a niche breadth analysis based on environment?


Questions, groups A and C: Composition and trait data

11. Distribution of life-history traits: are species of different life histories more likely to co-occur than a null model? (eg, M. Cody's work in SW deserts)

12. Distribution of taxonomic traits: are species of different taxa more likely to co-occur than a null model? (eg, oak sub-genera, congener focus, family focus, etc) (B. Enquist Nature paper in review on this)


Questions, groups A, B, and C: Relating traits to environment through composition data "4th-corner" problems: new analytical tools provided by Legendre et al.

13. These questions are nearly infinite: what environments favor adaptations of evergreenness, lobed leaves, biennial-ness, shade tolerance, small seeds, and on and on.


Other questions

14. Overyielding in nature: do species (eg, trees) perform better next to different species?

15. Optimal sampling: what is the best way to sample communities that best describes community structure and richness? (Whittaker vs. Stohlgren vs. CVS)