[R-sig-eco] Independence of vegetation samples

[Krzysztof Sakrejda]

I agree with Sarah Goslee's answer---you are looking at two spatial scales,
one of which is much smaller.  The samples obtained at the smaller scale
are going to be similar because of where they are regardless of which
environmental variables you choose to use to describe locations at a larger
spatial scale.

Having multiple samples does improve your measurement of the local
community structure.  Community structure is hard enough to measure that
this is valuable even if you don't treat them as independent samples.

Krzysztof





Krzysztof Sakrejda

Department of Biostatistics and Epidemiology
University of Massachusetts, Amherst
319 Morrill Science Center South
611 N. Pleasant Street
Amherst, MA 01003

work #: 413-325-6555
email: sakrejda at umass.edu

2015-09-03 11:19 GMT-04:00 Michael Marsh <swamp at blarg.net>:

> Reading David Warton's reply to Rajendra made me realize that my question
> (below) is related to his, and that I should specify the objectives of my
> investigation. I'm interested to see if there are separate, distinguishable
> plant communities/associations related to environmental variables, and if
> these relationships differ among the different vegetation life forms (e.g.,
> shrubs, graminoids, herbaceous annuals or perennials). I've used mrpp() to
> test for closeness of relationships among clusters distinguished by
> hclust() and mvpart() clustering, and used manova() as a further test of
> the responses of life forms to different environmental variables. ,
>
> Original question:
> Is there a method in R for testing for independence of vegetation samples,
> for example because of relative proximity of different samples? I would
> like to treat the 3 radially arranged transects of Jornada Line Point Index
> plots as different sample units.
> Mike Marsh
> Washington Native Plant Society
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>

	[[alternative HTML version deleted]]