We sampled plants in 5 plots x different habitats (mature, degraded and secondary forests). In every habitat we measure functional traits for the present species. Our traits are a mix of continuous and categorical variables .
There are differences in the continuous traits between the same species in the different habitats.
We want to compare if there are differences on functional diversity between habitats. As in the attached paper.
I was trying with the function "treedive" of the "vegan" package of R. But unfortunately, I can just supply one database of traits that were measure one time to all the species. In our case, we have several repetitions of traits (continuous) of the same species but at different habitats resulting in different values.
I was thinking in divide the dataset in static traits (categorical) and compare with the treedive function. And with changing traits (continuous) and compare them with multivariate analysis. But I am not sure if this will affect the nature of my study (comparing functional diversity).
Another option that I was thinking was just compare the distance matrix (all traits) of each habitat.
What do you think or suggest?
Thanks in advance.
http://onlinelibrary.wiley.com/doi/10.1111/j.1461-0248.2008.01255.x/pdf
Hello Jorge.
The indices measuring functional diversity are more about the differences between species (inter specific variability in trait values) than about the variability inside a given species (intra specific variability in trait values). Basically, you would detect a shift in functional diversity along an environmental gradient if you have species replacement or change in their relative abundance for species having contrasting trait values.
As you said in your question, to compute an index of functional diversity, you need a trait database with one trait value per trait and per species. Usually, for continuous traits, authors take the mean value for all the species their captured for trait measurement.
However, after a quick research on google scholar, I found this article dealing with intraspecific variability in trait values and functional diversity measures: http://onlinelibrary.wiley.com/doi/10.1890/07-1864.1/epdf
In the trait database, they seem to replace the rows containing species by rows containing individuals.
However, you should test, in the first place, some hypothesis before computing your FD index. If the interest of you study relies in the impact of intraspecific vs interspecific variability on trait values at the community level, you have to check for that before. If you find that trait values at the community level (SLA for example) are more driven by changes in community structure than in intraspecific variability, you may chose to compute a classical FD index. If you find that trait values are more driven by changes in intraspecific variability (due to phenotypic plasticity along environmental gradients), you should account for that in the FD index.
Here is a nice article that could help you dealing with the measurement of intraspecific trait variability: https://www.researchgate.net/publication/229149953_When_and_how_should_intraspecific_trait_variability_be_considered_in_plant_ecology
I hope it helped!
Romain
Article When and how should intraspecific trait variability be consi...
Dear Romain,
Thank you very much for your answer. You help a lot in order my ideas. I will start the test about intra vs inter functional diversity.
Best regards
Dear Jorge,
If your main interest is in the trait differences between habitats, then you have two additional options:
(1) Treat all traits as a matrix of response variables (multivariate normally distributed) and enter them in a multivariate linear model e.g. using lm() followed by Anova() in John Fox´s "car" library:
model1=cbind(trait1,trait2,trait3...)~habitat
Anova(model1)
You can set up contrast matrices for the response matrix to test individual differences and hypotheses on sets of traits. You may also introduce random effects if desired.
See an example provided in the link below.
(2) Alternatively, you could just look at individual traits or groups of traits in structural equation models, e.g. using the lavaan or piecewiseSEM packages.
I hope these suggestions are helpful in some way for you - don´t know if they work but maybe worth a try.
Best wishes,
Christoph
http://www.christoph-scherber.de/content/PDF%20Files/Scherber%20et%20al%20Nature%202010%20Bottom-up%20effects%20of%20plant%20diversity%20on%20multitrophic%20interactions%20S1.pdf
I know its probably not helpful for Jorge anymore, but for everybody else thinking about the same issue: there's a nice new package Article mFD: an R package to compute and illustrate the multiple fac...
Here you can calculate various indices. the default requests steady traits across "baskets" or plots. if you measure traits, like we did, the hack around it is to name species*treatment in the "Species x traits data frame". For me it worked.
Additionally we have figured out to calculate trait spaces with the "chull" and "Polygon" functions...
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology