Hi Katharine Are you intending to do a cluster analysis on the original variables to group so called 'like' estuaries using euclidean distance measures and/or factor analytic methods, and then compare the results of the cluster analysis to results obtained from the distance matrix of functional dissimilarity between estuaries?
Ling Ling Huang Sacramento City College --- On Fri, 9/28/12, Neahga Leonard <naturalistkni...@gmail.com> wrote: From: Neahga Leonard <naturalistkni...@gmail.com> Subject: Re: [ECOLOG-L] Visualizing functional diversity To: ECOLOG-L@LISTSERV.UMD.EDU Date: Friday, September 28, 2012, 3:45 PM Hello Katharine, Interesting question (once it comes down to plain language). One thing I'd add to the conversation is as follows. In addition to processing energy and material ecosystems develop complex, non-linear internal relationships. An ecosystem is defined by these relationships, the material and energy flow are a way to facilitate these relationships, but describing only these does a poor job of describing the ecosystem as a whole. We often focus on these aspects because they are relatively easy to quantify, but when relationships come into play simplicity is not synonymous with accuracy. Many things process materials and energy, but are not ecosystems. Part of the difficulty with a question like this is that you are treading into the, "What is life," type pf question. Every answer we give is an oversimplification. Cheers, Neahga Leonard On Fri, Sep 28, 2012 at 5:00 PM, Nicolas PERU <nicolas.p...@univ-lyon1.fr>wrote: > Martin, > > You're perfectly right, this is was an oversight from my part and David > gave a more general definition of my point of view (including flow of > material). > > Now, if I go back on Katharine question, this is important to understand > that traits are just a kind of proxy to evaluate ecosystems functions and > more particularly ecosystems functions realized by estuarine fishes. > Consequently, in traits studies we are quite far from a pure evaluation of > functional diversity. My PhD thesis was on freshwater invertebrates traits > and some people said to me that I wasn't measuring a "functional diversity" > because there was no direct link with nutrient or energy. This is true of > course. But this is the reason why I wrote in my first message that > functional diversity measures "the different way to manage energy" (or > material). As traits highlight organisms strategies to maximize their > success in ecosystems, we can consider them as a good way to quantify one > part of the possible ways to transport energy and material. > > Mathematically speaking, I'm not sure that beta diversity derived from > Rao's index can be considered as a real distance (ie with all properties of > a distance) and so included in a cluster analysis. This is a quite > complicated index giving a disproportionate weight to abundant species. > Consequently, by using Rao's index, you consider that species with high > counts are more important for ecosystems functions. Quite simply, a prey > (generally abundant) is more abundant than a predator to ensure flows of > energy and material. In addition, Rao's index express the mean (functional, > biological...) distance of two individual taken at random in the community. > As a mean, this index tend to stabilize as the number of species increase > (at a rate depending on the distance matrix used and so on the chosen > traits) leading to the conclusion that the more species you have the more > functionally redundant they are. > > I could say many things on Rao's index but the most important is that we > must be very careful about our biological hypotheses on functional > diversity to check if indices are able to really illustrate them. > > HTH again > > Nicolas > > > Le Fri, 28 Sep 2012 18:41:02 +0200, Martin Meiss <mme...@gmail.com> a > écrit: > > > Nicolas, >> Why would you restrict your interest to the flow of energy, and >> not >> include the flow of material, such as a nutrient like fixed nitrogen, or >> potassium? >> >> Martin M. Meiss >> >> 2012/9/27 Katharine Miller <kmill...@alaska.edu> >> >> Hi, >>> >>> I wanted to thank everyone for their responses and recommendations. Some >>> of >>> them were quite helpful and have got me thinking in new ways. >>> >>> With respect to the use of the Rao index, I didn't express my question >>> very >>> well. What I was really trying to discern was whether it was appropriate >>> to >>> use the Rao index values as a distance matrix of functional dissimilarity >>> between estuaries that could then be evaluated using standard >>> multivariate >>> methods (i.e. clustering). I have not seen Shannon entropy used this way >>> either, but it is understood that pairwise beta diversity calculated by >>> either of these approaches is a measure of dissimilarity between sites. >>> So, >>> on that basis, it doesn't seem too much of a stretch. Also, the index >>> values >>> are used as dissimilarities in Mantel tests or other matrix calculations. >>> >>> I am not sure whether the reason these indices have not been used this >>> way >>> is because it would be inappropriate statistically or mathematically, or >>> whether there is some ecological reason for not doing it. >>> >>> Thanks again. >>> >>> - Katharine >>> >>> >>> >>> > > -- >
