I am working with some data from botanical records. But I only measure the presence-absence of the species. I am interested in compare three habitats (natural, disturbed and secondary) with herbaceous angiosperms along an altitudinal gradient.
Dear Jorge,
citing C. Ricotta (Ecology, 2010, 91(7): 1981-1983): "Different beta metrics are measuring different quantities (i.e. average number of species not observed
for additive beta, or ‘‘effective number of communities’’ for multiplicative beta). Therefore, the key question we should ask of a beta metric is: does it measure the thing we are biologically interested in? If the metric has statistical properties that make patterns in beta easy to analyze and interpret, so much the better. But if not, this
is not necessarily a good reason to abandon it in favor of something statistically well behaved that is not actually the quantity we are most interested in."
That is, it is very difficult to help in absence of information on what you are interested in, based on your presence-absence data. In addtion to good food for thoughts, the cited forum paper includes the most relevant references to reports on descrption and comparative analysis of different beta-diversity metrics, since the first formulation by Whittaker. Hope it helps.
Thanks a lot Guido. I am interested in compare three habitats (natural, disturbed and secondary) with herbaceous angiosperms along an altitudinal gradient.
Jorge, again I suggest to read a paper, (Koleff et al. 2003, J of Animal Ecology 72: 367–382) in which the authors not only review 24 measures of beta diversity for presence/absence data, but, most useful for your application, express many of them for the first time in common terms, and compare some of their basic properties, as related to specific information provided by each measure. In particular, 22 measures are grouped as either 'measures of continuity' or 'measures of gain and loss' of species richness along environmental gradients.
Which of the two issues are you more interested in?
Then, the decision on what is the most suitable for your specific purpose, is of course up to you.
A general suggestion, after selecting the most relevant group of measures, would be to test different measures on your dataset , and compare the results to assess the measure-dependency of the outcomes, as well as the outcome consistency across different measures.
Hi,
you can very easily use this R package: betapart (A. Baselga, D. Orme & S. Villeger 2013). It gives the turnover and nestedness components of beta diversity. 0/1 data are fine.
Baselga A. (2010). Partitioning the turnover and nestedness components of beta diversity. Global Ecology and Biogeography, 19, 134-143.
Baselga A., Orme D. & Villeger S. (2013). Betapart : Partitioning beta diversity into turnover and nestedness components R.
For a nice graphical representation, see for example Albouy C., Guilhaumon F., Araujo M.B., Mouillot D. & Leprieur F. (2012). Combining projected changes in species richness and composition reveals climate change impacts on coastal Mediterranean fish assemblages. Global Change Biology, 18, 2995-3003.
Hope it helps!
cheers
Nina
As Nina advice you, for your absence and presence data it would be better for you to use R package.
Best
Jean Bruno
Hi Jorge,
I too would suggest Koleff et al. 2003 to start with (attached).
Best wishes,
Siddharth.
Thanks a lot for everyone for your interesting answers. I will start to run several analysis to have a better view of this phenomena.
hi,
I found Anderson et al. also helpful. In attachments...
cheers
Nina
I used a rarefaction analysis to compared species richness in different sized samples. iNext in R program can provide the results.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India