Date: Sat, 26 Sep 2015 14:07:07 +0200
From: Eduard Sz?cs <szoecs at uni-landau.de>
To: r-sig-ecology at r-project.org
Subject: Re: [R-sig-eco] get species within sites ordihull polys
Message-ID: <56068A6B.4040000 at uni-landau.de>
Content-Type: text/plain; charset=windows-1252
Dear Tim,
Sound's like your looking for indicator species ?
Maybe the indicspecies packages suites your needs.
Dufr?ne, M and P. Legendre. 1997. Species assemblages and indicator
species: the need for a flexible assymetrical approach. Ecological
Monographs 67: 345 - 366.
De C?ceres, M., P. Legendre, and M. Moretti, 2010. Improving indicator
species analysis by combining groups of sites. Oikos 119: 1674 -1684.
Cheers,
Eduard
On 25/09/15 20:58, Howard, Tim G (DEC) wrote:
This isn't the best example because species points don't fall in more than
one of the black polygons. But, my question: Can I easily identify which blue
points (species) fall within the polygon? Or can I easily identify which species
are 'most important' (most abundant?) for defining each of the three
groups?
Thanks for any pointers
Tim Howard
--
Eduard Sz?cs
Quantitative Landscape Ecology
Institute for Environmental Sciences
University Koblenz-Landau
Tel. +49 6341 280 31552
http://www.uni-koblenz-landau.de/en/campus-
landau/faculty7/environmental-sciences/landscape-
ecology/Staff/eduardszoecs
------------------------------
Date: Sat, 26 Sep 2015 21:18:24 -0700
From: Michael Marsh <swamp at blarg.net>
To: r-sig-ecology at r-project.org
Subject: Re: [R-sig-eco] get species within sites ordihull polys
Message-ID: <56076E10.4050301 at blarg.net>
Content-Type: text/plain; charset=windows-1252; format=flowed
Tim, I've used mvpart to cluster, and then rpart.pca the resulting
regression clustering.
mvpart requires a corresponding environmental data set.
The pca plot has what you require, polygons like ordihull based on and
showing plots (rows in your data), and vectors to named species.
I have assumed that the distance from the centroid to each species
corresponds to its importance in configuring the output, but I'm a
novice andwould like more information on that.
Mike Marsh
On 9/26/2015 3:00 AM, r-sig-ecology-request at r-project.org wrote:
get species within sites ordihull polys
Date: Fri, 25 Sep 2015 18:58:19 +0000
From: "Howard, Tim G (DEC)"<tim.howard at dec.ny.gov>
To:"r-sig-ecology at r-project.org" <r-sig-ecology at r-project.org>
Subject: [R-sig-eco] get species within sites ordihull polys
Message-ID:
<CY1PR09MB0266DBF8A724047334E58D6CA8420 at CY1PR09MB0266.n
amprd09.prod.outlook.com>
Content-Type: text/plain; charset="us-ascii"
All -
Consider clusters of points in an NMDS with those clusters determined in
some way (I'll use hclust below).
Then consider plotting the species on that ordination. I'd like to automatically
find which species are 'most associated' with each cluster. Perhaps that
translates to finding those species that fall within an ordihull of each group.
Before I stumble down into the world of the sp package and spatial overlaps
perhaps this is already a part of vegan or another package.
###Example:
library(vegan)
data(dune)
ord <- metaMDS(dune)
# get some groups based on hclust
dis <- vegdist(dune)
clus <- hclust(dis, "average")
plot(clus)
rect.hclust(clus, 3)
grp <- cutree(clus, 3)
#plot the mds with the groups
mdsPlot <- plot(ord, type="n", display = "sites")
points(ord, display = "sites", col="red", pch=19)
ordihull(ord, grp)
#plot the species
points(ord, display = "species", col = "blue", pch=19)
###End example
This isn't the best example because species points don't fall in more than one
of the black polygons. But, my question: Can I easily identify which blue
points (species) fall within the polygon? Or can I easily identify which species
are 'most important' (most abundant?) for defining each of the three
groups?
Thanks for any pointers
Tim Howard