Re: Nature magazine article. review

[John Kineman <***>]

If I understand Jamie's parsimonious comment, I think this is correct.
Niches are defined by a combination of ranges of various environmental
variables. One approach that has been popular to predicting ecoregions has
been cluster analysis of environmental variables, which then gives a map of
environmental uniqueness. One may then assume that these areas would be
biologically unique because of adaptation. So, given that genetic
difference reflects environmental difference, one should get the same map
of general categories from environment or niche models. However, one should
not get the same distribution of species except within its adaptive range,
and as written the article seems to imply that all species are equivalent,
which is clearly not true. The example given, species abundances within a
particular forest type, makes sense because one can assume they are all
operating within their niche limits. In that case the niches are
essentially equal and so the neutral model will work fine. But I doubt that
they claim that mangroves will invade the Yukon delta any time soon. It
sounds more like grandstanding than anything to me.

An analogy may be appropriate. We know that mountains are created by very
intricate processes deep in the earth -- volcanism and tectonics
determining mineral contents and uplift. And yet we can produce very
realistic looking landscapes using a simple fractal equation. This is not
the same thing as predicting what a given mountain will be like. It is
modeling a gross generality.

JJK

But the article seems to say that

At 09:09 AM 3/27/04 -0800, you wrote:
> Dear Judith,
>
> Not sure why it's such a mystery.  The formulae that
> co-mingle several 'random distribution' vectors are
> de facto generators of niches (or, ala standard complexity
> notions, 'attractor basins').
>
> If the concern is that the math is 'neutral' to what
> any specific qualia/characteristics are for the possible
> niches, then the only thing missing is identification
> of this or that bundle of vectors as being associable
> with one or another niche qualia, where they can be shown
> to interactively track/map one another.
>
> Might be an interesting empirical way to be able
> to conceptually move between tiers of qualia in
> 'Roseneanly complex' systems.
>
> James
>
>
>
>
>
> > Judith Rosen wrote:
> >
> > Hi Folks,
> >
> > There is an article that was published last year in Nature magazine
> about some
> > ecology theories and models developed by two men; Hubbell and Bell.
> Neutrality
> > Theory, it's being called. I read it because my oldest child had a paper to
> > write and thought some of "Pop's stuff" (her name for her grandfather)
> might be
> > applicable. She was right.
> >
> > I'm curious, before I infect the group with my own point of view, what
> some of
> > the ecologists think. The link is here:
> > http://www.nature.com/nsu/020527/020527-13.html
> >
> > I would also like to get contact info for Hubbell, if anyone knows how to
> > achieve that?
> >
> > Judith