Schefferville - transitions; K-means clustering and comparisons of these clusters
Tammy L. Elliott
Date: March 21, 2015
R version 3.1.0
# set global chunk options:
library(knitr)
opts_chunk$set(cache=FALSE, fig.align='center')Plot PCoA with 3 groupings based on k-means clustering for environmental distances
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.env.3<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.3[,1]
y<-vasc.env.3[,2]
plot(vasc.env.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.3, vasc.env.clust.3$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.3, vasc.env.clust.3$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.env.3), cex=0.4)
vasc.env.3<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.3[,1]
y<-vasc.env.3[,2]
plot(vasc.env.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.3, vasc.env.clust.3$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.3, vasc.env.clust.3$cluster, col="grey70")
box(lwd=2)
### Plot PCoA with 4 groupings based on k-means clustering for environmental distances
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.env.4<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.4[,1]
y<-vasc.env.4[,2]
plot(vasc.env.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.4, vasc.env.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.4, vasc.env.clust.4$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.env.4), cex=0.4)
vasc.env.4<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.4[,1]
y<-vasc.env.4[,2]
plot(vasc.env.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.4, vasc.env.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.4, vasc.env.clust.4$cluster, col="grey70")
box(lwd=2)
Plot PCoA with 5 groupings based on k-means clustering for environmental distances
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.env.5<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.5[,1]
y<-vasc.env.5[,2]
plot(vasc.env.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.5, vasc.env.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.5, vasc.env.clust.5$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.env.5), cex=0.4)
vasc.env.5<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.5[,1]
y<-vasc.env.5[,2]
plot(vasc.env.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black")
ordispider(vasc.env.5, vasc.env.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.5, vasc.env.clust.5$cluster, col="grey70")
box(lwd=2)
Non-abundance weighted Chao’s beta diversity
sp.dist.abd <- dist(vasc.pairs.abd0.mat)
# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((sp.dist.abd),2,15)
plot(ccas.dist,sortq=TRUE)
veg.abd.k.clust<-kmeans(sp.dist.abd, 3)
veg.abd.k.clust.4<-kmeans(sp.dist.abd,4)
veg.abd.k.clust.5<-kmeans(sp.dist.abd,5)PCoA on Chao’s beta diversity with three clusters based on K-means clustering
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
veg.abd<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-veg.abd[,1]
y<-veg.abd[,2]
plot(veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with three clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd, veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd, veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(veg.abd), cex=0.4)
veg.abd<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with three clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd, veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd, veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
PCoA on Chao’s beta diversity with four clusters based on K-means clustering
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
veg.abd.4<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-veg.abd.4[,1]
y<-veg.abd.4[,2]
plot(veg.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with four clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd.4, veg.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd.4, veg.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(veg.abd.4), cex=0.4)
veg.abd.4<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(veg.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with four clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd.4, veg.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd.4, veg.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
PCoA on Chao’s beta diversity with five clusters based on K-means clustering
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
veg.abd.5<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-veg.abd.5[,1]
y<-veg.abd.5[,2]
plot(veg.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with five clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd.5, veg.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd.5, veg.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(veg.abd.5), cex=0.4)
veg.abd.5<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(veg.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with five clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd.5, veg.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd.5, veg.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
Abundance weighted Chao’s beta diversity
# Abundance weighted Chao's beta diversity
abd.sp.dist.abd <- dist(vasc.pairs.abd1.mat)
# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((sp.dist.abd),2,15)
plot(ccas.dist,sortq=TRUE)
abd.veg.abd.k.clust<-kmeans(abd.sp.dist.abd, 3)
abd.veg.abd.k.clust.4<-kmeans(abd.sp.dist.abd,4)
abd.veg.abd.k.clust.5<-kmeans(abd.sp.dist.abd,5)Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted beta diversity with three clusters based on K-means clustering (vasculars)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.veg.abd<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.veg.abd[,1]
y<-abd.veg.abd[,2]
plot(abd.veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with three clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.veg.abd), cex=0.4)
abd.veg.abd<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with three clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted beta diversity with four clusters based on K-means clustering (vasculars)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.veg.abd.4<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.veg.abd.4[,1]
y<-abd.veg.abd.4[,2]
plot(abd.veg.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with four clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd.4, abd.veg.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd.4, abd.veg.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.veg.abd.4), cex=0.4)
abd.veg.abd.4<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.veg.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with four clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd.4, abd.veg.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd.4, abd.veg.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted diversity with five clusters based on K-means clustering (vasculars)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.veg.abd.5<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.veg.abd.5[,1]
y<-abd.veg.abd.5[,2]
plot(abd.veg.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with five clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd.5, abd.veg.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd.5, abd.veg.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.veg.abd.5), cex=0.4)
abd.veg.abd.5<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.veg.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with five clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd.5, abd.veg.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd.5, abd.veg.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
Pairwise phylogenetic distances (not weighted) between pairs of plots
#Pairwise phylogenetic distances (not weighted) between pairs of plots
trans.dist<-dist(vasc.ph.pairs.abd0.mat)
# Make euclidean for further analyzes
head(sqrt(trans.dist))## [1] 1.5915041 1.7298445 0.9721087 1.4180747 1.5518220 1.1105326is.euclid(trans.dist)## Warning in is.euclid(trans.dist): Zero distance(s)## [1] TRUE# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((trans.dist),2,15)
plot(ccas.dist,sortq=TRUE)
vasc.dist.clust.two<-kmeans(trans.dist,2)
vasc.dist.clust.three<-kmeans(trans.dist,3)
vasc.dist.clust.four<-kmeans(trans.dist,4)
vasc.dist.clust.five<-kmeans(trans.dist,5)Principal Coordinate Analysis (PCoA) with three clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (vasculars)
#Plot PCoA with 3 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.abd.k.3<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.abd.k.3[,1]
y<-vasc.abd.k.3[,2]
plot(vasc.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.3, vasc.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.3, vasc.dist.clust.three$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.abd.k.3), cex=0.4)
vasc.abd.k.3<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(vasc.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.3, vasc.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.3, vasc.dist.clust.three$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with four clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (vasculars)
#Plot PCoA with 4 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.abd.k.4<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.abd.k.4[,1]
y<-vasc.abd.k.4[,2]
plot(vasc.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.4, vasc.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.4, vasc.dist.clust.four$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.abd.k.4), cex=0.4)
vasc.abd.k.4<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(vasc.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.4, vasc.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.4, vasc.dist.clust.four$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with five clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (vasculars)
#Plot PCoA with 5 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
vasc.abd.k.5<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.abd.k.5[,1]
y<-vasc.abd.k.5[,2]
plot(vasc.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.5, vasc.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.5, vasc.dist.clust.five$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(vasc.abd.k.5), cex=0.5)
vasc.abd.k.5<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(vasc.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.5, vasc.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.5, vasc.dist.clust.five$cluster, col="grey70")
box(lwd=2)
Pairwise phylogenetic distances (not weighted) between pairs of plots
#Pairwise phylogenetic distances (not weighted) between pairs of plots
abd.trans.dist<-dist(vasc.ph.pairs.abd1.mat)
# Make euclidean for further analyzes
head(sqrt(abd.trans.dist))## [1] 1.430152 1.860524 1.735937 1.877979 1.578909 1.782176is.euclid(abd.trans.dist)## [1] TRUE# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((abd.trans.dist),2,15)
plot(ccas.dist,sortq=TRUE)
abd.vasc.dist.clust.two<-kmeans(abd.trans.dist,2)
abd.vasc.dist.clust.three<-kmeans(abd.trans.dist,3)
abd.vasc.dist.clust.four<-kmeans(abd.trans.dist,4)
abd.vasc.dist.clust.five<-kmeans(abd.trans.dist,5)Principal Coordinate Analysis (PCoA) with three clusters based on K-means clustering on Chao’s abundance-weighted phylogenetic beta diversity pairwise distances (vasculars)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.vasc.abd.k.3<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.vasc.abd.k.3[,1]
y<-abd.vasc.abd.k.3[,2]
plot(abd.vasc.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.3, abd.vasc.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.3, abd.vasc.dist.clust.three$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.vasc.abd.k.3), cex=0.4)
abd.vasc.abd.k.3<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.vasc.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.3, abd.vasc.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.3, abd.vasc.dist.clust.three$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with four clusters based on K-means clustering on Chao’s abundance-weighted phylogenetic beta diversity pairwise distances (vasculars)
#Plot PCoA with 4 groupings based on k-means clustering for abundance-weighted Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.vasc.abd.k.4<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.vasc.abd.k.4[,1]
y<-abd.vasc.abd.k.4[,2]
plot(abd.vasc.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.4, abd.vasc.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.4, abd.vasc.dist.clust.four$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.vasc.abd.k.4), cex=0.4)
abd.vasc.abd.k.4<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.vasc.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.4, abd.vasc.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.4, abd.vasc.dist.clust.four$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with five clusters based on K-means clustering on Chao’s abundance-weighted phylogenetic beta diversity pairwise distances (vasculars)
#Plot PCoA with 5 groupings based on k-means clustering for abundance-weighted Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.vasc.abd.k.5<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.vasc.abd.k.5[,1]
y<-abd.vasc.abd.k.5[,2]
plot(abd.vasc.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.5, abd.vasc.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.5, abd.vasc.dist.clust.five$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.vasc.abd.k.5), cex=0.5)
abd.vasc.abd.k.5<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.vasc.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.5, abd.vasc.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.5, abd.vasc.dist.clust.five$cluster, col="grey70")
box(lwd=2)
Angiosperms non-abundance weighted Chao’s beta diversity
angio.dist.bd <- dist(angiopairs.abd0.mat)
# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((angio.dist.bd),2,15)
plot(ccas.dist,sortq=TRUE)
angio.dist.bd.k.clust<-kmeans(angio.dist.bd, 3)
angio.dist.bd.k.clust.4<-kmeans(angio.dist.bd,4)
angio.dist.bd.k.clust.5<-kmeans(angio.dist.bd,5)Principal Coordinate Analysis (PCoA) on Chao’s beta diversity with three clusters based on K-means clustering (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
pcoa.angio.dist.bd<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-pcoa.angio.dist.bd[,1]
y<-pcoa.angio.dist.bd[,2]
plot(pcoa.angio.dist.bd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with three clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd, angio.dist.bd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd, angio.dist.bd.k.clust$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(pcoa.angio.dist.bd), cex=0.4)
pcoa.angio.dist.bd<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(pcoa.angio.dist.bd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with three clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd, angio.dist.bd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd, angio.dist.bd.k.clust$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s beta diversity with four clusters based on K-means clustering (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
pcoa.angio.dist.bd.4<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-pcoa.angio.dist.bd.4[,1]
y<-pcoa.angio.dist.bd.4[,2]
plot(pcoa.angio.dist.bd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with four clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd.4, angio.dist.bd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd.4, angio.dist.bd.k.clust.4$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(pcoa.angio.dist.bd.4), cex=0.4)
pcoa.angio.dist.bd.4<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(pcoa.angio.dist.bd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with four clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd.4, angio.dist.bd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd.4, angio.dist.bd.k.clust.4$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s beta diversity with five clusters based on K-means clustering (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
pcoa.angio.dist.bd.5<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-pcoa.angio.dist.bd.5[,1]
y<-pcoa.angio.dist.bd.5[,2]
plot(pcoa.angio.dist.bd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with five clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd.5, angio.dist.bd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd.5, angio.dist.bd.k.clust.5$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(pcoa.angio.dist.bd.5), cex=0.4)
pcoa.angio.dist.bd.5<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(pcoa.angio.dist.bd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with five clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd.5, angio.dist.bd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd.5, angio.dist.bd.k.clust.5$cluster, col="grey70")
box(lwd=2)
Abundance weighted Chao’s beta diversity for angiosperms only
angio.dist.abd <- dist(angiopairs.abd1.mat)
# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((angio.dist.abd),2,15)
plot(ccas.dist,sortq=TRUE)
abd.angio.abd.k.clust<-kmeans(angio.dist.abd, 3)
abd.angio.abd.k.clust.4<-kmeans(angio.dist.abd,4)
abd.angio.abd.k.clust.5<-kmeans(angio.dist.abd,5)Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted beta diversity with three clusters based on K-means clustering (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.angio.abd<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.angio.abd[,1]
y<-abd.angio.abd[,2]
plot(abd.angio.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with three clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd, abd.angio.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd, abd.angio.abd.k.clust$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.angio.abd), cex=0.4)
abd.angio.abd<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.angio.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with three clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd, abd.angio.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd, abd.angio.abd.k.clust$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted beta diversity with four clusters based on K-means clustering (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.angio.abd.4<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.angio.abd.4[,1]
y<-abd.angio.abd.4[,2]
plot(abd.angio.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with four clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd.4, abd.angio.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd.4, abd.angio.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.angio.abd.4), cex=0.4)
abd.angio.abd.4<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.angio.abd.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with four clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd.4, abd.angio.abd.k.clust.4$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd.4, abd.angio.abd.k.clust.4$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) on Chao’s abundance-weighted beta diversity with five clusters based on K-means clustering (angiosperms)``
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
abd.angio.abd.5<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-abd.angio.abd.5[,1]
y<-abd.angio.abd.5[,2]
plot(abd.angio.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with five clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(abd.angio.abd.5), cex=0.4)
abd.angio.abd.5<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.angio.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with five clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
Pairwise phylogenetic distances (not weighted) between pairs of plots - angiosperms
angio.trans.dist<-dist(angioph.pairs.abd0.mat)
# Make euclidean for further analyzes
head(sqrt(angio.trans.dist))## [1] 1.298212 1.607143 1.321965 1.326673 1.795988 1.035656is.euclid(angio.trans.dist)## Warning in is.euclid(angio.trans.dist): Zero distance(s)## [1] TRUE# Suggestions for optimal K-means clusters
ccas.dist<-cascadeKM((angio.trans.dist),2,15)
plot(ccas.dist,sortq=TRUE)
angio.dist.clust.three<-kmeans(angio.trans.dist,3)
angio.dist.clust.four<-kmeans(angio.trans.dist,4)
angio.dist.clust.five<-kmeans(angio.trans.dist,5)Principal Coordinate Analysis (PCoA) with three clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 3 groupings based on k-means clustering for not abundance-weighted Phylogenetic beta diversity values (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.distk.3<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.distk.3[,1]
y<-angio.distk.3[,2]
plot(angio.distk.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.3, angio.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.3, angio.dist.clust.three$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.distk.3), cex=0.4)
angio.distk.3<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.distk.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.3, angio.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.3, angio.dist.clust.three$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with four clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 4 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.distk.4<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.distk.4[,1]
y<-angio.distk.4[,2]
plot(angio.distk.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.4, angio.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.4, angio.dist.clust.four$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.distk.4), cex=0.4)
angio.distk.4<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.distk.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.4, angio.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.4, angio.dist.clust.four$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with five clusters based on K-means clustering on Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 5 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.distk.5<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.distk.5[,1]
y<-angio.distk.5[,2]
plot(angio.distk.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.5, angio.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.5, angio.dist.clust.five$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.distk.5), cex=0.5)
angio.distk.5<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.distk.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.5, angio.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.5, angio.dist.clust.five$cluster, col="grey70")
box(lwd=2)
Pairwise phylogenetic distances (weighted) between pairs of plots
abd.angio.dist<-dist(angioph.pairs.abd1.mat)
# Make euclidean for further analyzes
head(sqrt(abd.angio.dist))## [1] 1.052225 1.161007 1.407660 1.220391 1.890975 1.659321# Suggestions for optimal K-means clusters
ccas.dist.angio<-cascadeKM((abd.angio.dist),2,15)
plot(ccas.dist.angio,sortq=TRUE)
abd.angio.dist.clust.two<-kmeans(abd.angio.dist,2)
abd.angio.dist.clust.three<-kmeans(abd.angio.dist,3)
abd.angio.dist.clust.four<-kmeans(abd.angio.dist,4)
abd.angio.dist.clust.five<-kmeans(abd.angio.dist,5)Principal Coordinate Analysis (PCoA) with three clusters based on K-means clustering on abundance-weighted Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 3 groupings based on k-means clustering for abundance-weighted Phylogenetic beta diversity values (angiosperms)
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.abd.k.3<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.abd.k.3[,1]
y<-angio.abd.k.3[,2]
plot(angio.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.3, abd.angio.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.3, abd.angio.dist.clust.three$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.abd.k.3), cex=0.4)
angio.abd.k.3<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.abd.k.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.3, abd.angio.dist.clust.three$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.3, abd.angio.dist.clust.three$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with four clusters based on K-means clustering on abundance-weighted Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 4 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.abd.k.4<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.abd.k.4[,1]
y<-angio.abd.k.4[,2]
plot(angio.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.4, abd.angio.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.4, abd.angio.dist.clust.four$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.abd.k.4), cex=0.4)
angio.abd.k.4<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.abd.k.4, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with four\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.4, abd.angio.dist.clust.four$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.4, abd.angio.dist.clust.four$cluster, col="grey70")
box(lwd=2)
Principal Coordinate Analysis (PCoA) with five clusters based on K-means clustering on abundance-weighted Chao’s phylogenetic beta diversity pairwise distances (angiosperms)
#Plot PCoA with 5 groupings based on k-means clustering for Phylogenetic beta diversity values
#dev.new(width=11.8, height=8)
par(mfrow=c(1,2))
angio.abd.k.5<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-angio.abd.k.5[,1]
y<-angio.abd.k.5[,2]
plot(angio.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with five\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.5, abd.angio.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.5, abd.angio.dist.clust.five$cluster, col="grey70")
box(lwd=2)
text(x, y, rownames(angio.abd.k.5), cex=0.5)
angio.abd.k.5<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.abd.k.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.5, abd.angio.dist.clust.five$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.5, abd.angio.dist.clust.five$cluster, col="grey70")
box(lwd=2)
My objective classifications of habitat type
#try polygons per habitat type
library(ggplot2)## Warning: package 'ggplot2' was built under R version 3.1.1#dev.new(width=8, height=6)
hab<-site[,c(3:5), drop=FALSE]
hab.types<-hab[,1, drop=FALSE]
fen<-hab[which(hab$Habitat.Description=="Fen"),]
smf<-hab[which(hab$Habitat.Description=="Spruce moss"),]
smf.fen<-hab[which(hab$Habitat.Description=="Spruce moss - Fen"),]
tun<-hab[which(hab$Habitat.Description=="Tundra"),]
shr<-hab[which(hab$Habitat.Description=="Shrubs"),]
tun.shr<-hab[which(hab$Habitat.Description=="Tundra - Shrubs"),]
#use convex hull function to plot habitats
find_hull <- function(df) {
res.ch <- df[chull(hab$Latitude..WGS.84.datum., hab$Longitude..WGS.84.datum), ]
res <- bezier(res.ch)
res <- data.frame(x=res$Latitude..WGS.84.datum.,y=res$hab$Longitude..WGS.84.datum)
res$z <- res$z
res
}
hab.ggplot<-ggplot(hab, aes(hab$Latitude..WGS.84.datum., hab$Longitude..WGS.84.datum., colour = hab$Habitat.Description)) + geom_point(size=4) +
theme_bw() + theme(panel.border = element_rect(colour="black"), panel.grid.major = element_blank(),
panel.grid.minor = element_blank(), axis.line = element_line(colour = "black"))+ xlab("Latitude") + ylab("Longitude") +
ggtitle("Habitat Types for Plots in Large \nSampling Grid (Tammy's Classifications)")
hab.ggplot+theme(axis.title.y = element_text(size=16, vjust=1),axis.title.x = element_text(size=16), axis.text.x = element_text(size=12), axis.text.y= element_text(size=12,angle=90,hjust=0.5))+
theme(plot.title = element_text(vjust = 2, size=15, face="bold")) + theme(legend.title = element_text(size=12))+
scale_linetype_discrete("Habitat Type") +
scale_shape_discrete("Habitat Type") +
scale_colour_discrete("Habitat Type")
Locations on grid of the three groupings
#make dataframe for large transects; with plot number, Habitat.description, latitude, longitude
site.trunc<-site[,c(3:5)]
#dev.new(width=15, height=8.5)
par(mfrow=c(3,3))
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.env.clust.3$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on \nK-means clustering of environmental distances", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(veg.abd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means \nclustering of Chao's beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.veg.abd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.dist.clust.three$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.vasc.dist.clust.three$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.bd.k.clust$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.abd.k.clust$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.clust.three$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.dist.clust.three$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3"), pch=21,
pt.bg=c("black","grey50", "white"),col="black", cex=1.1, bty="n")
Locations on grid of the four groupings
#Show on grid where four clusters correspond to:
#dev.new(width=15, height=8.5)
par(mfrow=c(3,3))
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.env.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on \nK-means clustering of environmental distances", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(veg.abd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means \nclustering of Chao's beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.veg.abd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.dist.clust.four$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.vasc.dist.clust.four$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.bd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.abd.k.clust.4$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.clust.four$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.dist.clust.four$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("topleft", c("Group 1","Group 2","Group 3", "Group 4"), pch=21,
pt.bg=c("black", "gray30", "gray80","white"),col="black", cex=0.95, bty="n")
Locations on grid of the five groupings
# Show on grid locations of five clusters for nine different methods
#dev.new(width=15, height=8.5)
par(mfrow=c(3,3))
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.env.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on \nK-means clustering of environmental distances", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(veg.abd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means \nclustering of Chao's beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.veg.abd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.vasc.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's weighted phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.bd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.abd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(abd.angio.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's weighted phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
# Compare numbers among clusters
size.3## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## Group 1 42 45 55 72 100
## Group 2 86 57 64 61 7
## Group 3 48 74 57 43 69
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Group 1 38 64 67 58
## Group 2 60 71 72 56
## Group 3 78 41 37 62size.4## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## [1,] 34 34 38 31 6
## [2,] 45 56 53 34 30
## [3,] 55 55 32 71 89
## [4,] 42 31 53 40 51
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## [1,] 32 49 101 54
## [2,] 35 32 29 44
## [3,] 56 53 25 22
## [4,] 53 42 21 56size.5## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## Group 1 32 30 32 72 44
## Group 2 10 43 53 16 51
## Group 3 23 11 25 40 46
## Group 4 37 56 20 16 29
## Group 5 74 36 46 32 6
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Group 1 22 52 17 56
## Group 2 28 34 50 41
## Group 3 45 15 30 25
## Group 4 49 28 26 32
## Group 5 32 47 53 22# Show within group sum of squares
withinss.3## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## Group 1 35.97047 3306.602 5366.880 4102.819 3034.183
## Group 2 49.24220 2240.539 7128.447 1734.179 2276.252
## Group 3 34.96651 4170.653 3671.677 1566.483 3235.393
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Group 1 2572.728 9226.259 3625.735 3508.737
## Group 2 2868.798 6542.412 2146.382 7202.903
## Group 3 5039.601 3356.442 1405.717 1445.043withinss.4## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## Group 1 9.450085 2095.118 2652.304 979.5204 2036.833
## Group 2 30.605065 2737.750 4909.335 878.4684 1109.265
## Group 3 19.079068 2128.677 1727.124 4030.3091 2504.894
## Group 4 35.970475 1236.884 3115.649 581.3797 1263.229
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Group 1 1330.889 3202.646 3545.8550 6921.3836
## Group 2 2201.212 2125.532 922.9113 1748.6033
## Group 3 2776.249 5520.618 825.8882 794.3296
## Group 4 2168.654 3535.861 380.6387 792.5493withinss.5## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc Phy.Abd.Bdiv.vasc
## Group 1 20.41801 1837.63838 1727.124 4102.81900 484.4514
## Group 2 1.30683 1382.63917 4909.335 69.77186 1198.3972
## Group 3 12.17893 98.57614 1257.154 581.37972 987.9169
## Group 4 18.02882 2891.87654 1213.403 197.82068 1067.8001
## Group 5 34.96041 1333.60517 2274.467 757.22465 2036.8328
## Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Group 1 1051.949 5419.677 636.1840 792.5493
## Group 2 1759.896 2395.241 1413.9549 1486.5580
## Group 3 1434.520 1116.326 1019.9035 764.2236
## Group 4 1976.840 1537.045 477.7594 3401.8341
## Group 5 1330.889 2838.457 1278.5311 794.3296# Compare sum-of-squares within and between clusters
totss.all<-rbind(totss.3, totss.4, totss.5)
rownames(totss.all)<-c("Total SS 3", "Total SS 4", "Total SS 5")
totss.all## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc
## Total SS 3 586.5044 21517.85 31209.84 18917.66
## Total SS 4 586.5044 21517.85 31209.84 18917.66
## Total SS 5 586.5044 21517.85 31209.84 18917.66
## Phy.Abd.Bdiv.vasc Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio
## Total SS 3 21098.37 23217.48 35774.2 20389.45
## Total SS 4 21098.37 23217.48 35774.2 20389.45
## Total SS 5 21098.37 23217.48 35774.2 20389.45
## Phy.Abd.Bdiv.angio
## Total SS 3 35238.74
## Total SS 4 35238.74
## Total SS 5 35238.74tot.withinss.all<-rbind(tot.withinss.3, tot.withinss.4, tot.withinss.5)
rownames(tot.withinss.all)<-c("Tot. within SS 3", "Tot. within SS 4", "Tot. within SS 5")
tot.withinss.all## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc
## Tot. within SS 3 120.17918 9717.793 16167.00 7403.481
## Tot. within SS 4 95.10469 8198.430 12404.41 6469.678
## Tot. within SS 5 86.89300 7544.335 11381.48 5709.016
## Phy.Abd.Bdiv.vasc Bdiv.angio Abd.Bdiv.angio
## Tot. within SS 3 8545.829 10481.128 19125.11
## Tot. within SS 4 6914.221 8477.004 14384.66
## Tot. within SS 5 5775.398 7554.095 13306.75
## Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Tot. within SS 3 7177.834 12156.683
## Tot. within SS 4 5675.293 10256.866
## Tot. within SS 5 4826.333 7239.495betweenss<-rbind(betweenss.3, betweenss.4, betweenss.5)
rownames(betweenss)<-c("Between SS 3", "Between SS 4", "Between SS 5")
betweenss## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc
## Between SS 3 466.3252 11800.06 15042.84 11514.18
## Between SS 4 491.3997 13319.42 18805.43 12447.99
## Between SS 5 499.6114 13973.51 19828.36 13208.65
## Phy.Abd.Bdiv.vasc Bdiv.angio Abd.Bdiv.angio Phy.Bdiv.angio
## Between SS 3 12552.54 12736.35 16649.08 13211.61
## Between SS 4 14184.15 14740.48 21389.54 14714.15
## Between SS 5 15322.98 15663.39 22467.45 15563.11
## Phy.Abd.Bdiv.angio
## Between SS 3 23082.06
## Between SS 4 24981.87
## Between SS 5 27999.25# Compare within/between
with.betw<-rbind(tot.withinss.3/betweenss.3, tot.withinss.4/betweenss.4, tot.withinss.5/betweenss.5)
rownames(with.betw)<-c("Within/Between SS 3", "Within/Between SS 4", "Within/Between SS 5")
with.betw## Env.dist Bdiv.vasc Abd.Bdiv.vasc Phy.Bdiv.vasc
## Within/Between SS 3 0.2577154 0.8235378 1.0747311 0.6429880
## Within/Between SS 4 0.1935384 0.6155245 0.6596188 0.5197369
## Within/Between SS 5 0.1739212 0.5399025 0.5740003 0.4322180
## Phy.Abd.Bdiv.vasc Bdiv.angio Abd.Bdiv.angio
## Within/Between SS 3 0.6808045 0.8229301 1.1487186
## Within/Between SS 4 0.4874610 0.5750834 0.6725090
## Within/Between SS 5 0.3769110 0.4822773 0.5922677
## Phy.Bdiv.angio Phy.Abd.Bdiv.angio
## Within/Between SS 3 0.5432974 0.5266724
## Within/Between SS 4 0.3857030 0.4105723
## Within/Between SS 5 0.3101136 0.2585603Mantel plot comparisons
mantel.statistic## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.000 0.470 0.398 0.412 0.131
## Bdiv.vasc 0.470 1.000 0.629 0.972 0.582
## Phy.Bdiv.vasc 0.398 0.629 1.000 0.529 0.577
## Bdiv.angio 0.412 0.972 0.529 1.000 0.608
## Phy.Bdiv.angio 0.131 0.582 0.577 0.608 1.000mantel.statistic.abd## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.000 0.377 0.150 0.304
## Abd.Bdiv.vasc 0.377 1.000 0.384 0.923
## Phy.Abd.Bdiv.vasc 0.150 0.384 1.000 0.277
## Abd.Bdiv.angio 0.304 0.923 0.277 1.000
## Phy.Abd.Bdiv.angio 0.117 0.519 0.320 0.585
## Phy.Abd.Bdiv.angio
## Env.dist 0.117
## Abd.Bdiv.vasc 0.519
## Phy.Abd.Bdiv.vasc 0.320
## Abd.Bdiv.angio 0.585
## Phy.Abd.Bdiv.angio 1.000mantel.statistic.signif## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.000 0.001 0.001 0.001 0.001
## Bdiv.vasc 0.001 1.000 0.001 0.001 0.001
## Phy.Bdiv.vasc 0.001 0.001 1.000 0.001 0.001
## Bdiv.angio 0.001 0.001 0.001 1.000 0.001
## Phy.Bdiv.angio 0.001 0.001 0.001 0.001 1.000mantel.statistic.abd.signif## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.000 0.001 0.001 0.001
## Abd.Bdiv.vasc 0.001 1.000 0.001 0.001
## Phy.Abd.Bdiv.vasc 0.001 0.001 1.000 0.001
## Abd.Bdiv.angio 0.001 0.001 0.001 1.000
## Phy.Abd.Bdiv.angio 0.001 0.001 0.001 0.001
## Phy.Abd.Bdiv.angio
## Env.dist 0.001
## Abd.Bdiv.vasc 0.001
## Phy.Abd.Bdiv.vasc 0.001
## Abd.Bdiv.angio 0.001
## Phy.Abd.Bdiv.angio 1.000Partition comparisons
Use Rand index Values between 0 and 1, with 0 indicating that the two data clusters do not agree on Any pair of points and 1 indicating that the data clusters are exactly the same.
Partition comparison between three cluster groupings
compPart## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.00 0.66 0.70 0.63 0.59
## Bdiv.vasc 0.66 1.00 0.71 0.92 0.70
## Phy.Bdiv.vasc 0.70 0.71 1.00 0.67 0.62
## Bdiv.angio 0.63 0.92 0.67 1.00 0.71
## Phy.Bdiv.angio 0.59 0.70 0.62 0.71 1.00Partition comparison between three cluster groupings for abundance-weighted groupings
compPart.abd## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.00 0.66 0.59 0.63
## Abd.Bdiv.vasc 0.66 1.00 0.63 0.76
## Phy.Abd.Bdiv.vasc 0.59 0.63 1.00 0.57
## Abd.Bdiv.angio 0.63 0.76 0.57 1.00
## Phy.Abd.Bdiv.angio 0.55 0.64 0.51 0.61
## Phy.Abd.Bdiv.angio
## Env.dist 0.55
## Abd.Bdiv.vasc 0.64
## Phy.Abd.Bdiv.vasc 0.51
## Abd.Bdiv.angio 0.61
## Phy.Abd.Bdiv.angio 1.00Partition comparison between four cluster groupings
compPart.4## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.00 0.70 0.70 0.70 0.59
## Bdiv.vasc 0.70 1.00 0.69 0.97 0.65
## Phy.Bdiv.vasc 0.70 0.69 1.00 0.69 0.60
## Bdiv.angio 0.70 0.97 0.69 1.00 0.64
## Phy.Bdiv.angio 0.59 0.65 0.60 0.64 1.00Partition comparison between four cluster groupings for abundance-weighted groupings
compPart.4.abd## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.00 0.67 0.64 0.67
## Abd.Bdiv.vasc 0.67 1.00 0.63 0.91
## Phy.Abd.Bdiv.vasc 0.64 0.63 1.00 0.61
## Abd.Bdiv.angio 0.67 0.91 0.61 1.00
## Phy.Abd.Bdiv.angio 0.63 0.69 0.57 0.69
## Phy.Abd.Bdiv.angio
## Env.dist 0.63
## Abd.Bdiv.vasc 0.69
## Phy.Abd.Bdiv.vasc 0.57
## Abd.Bdiv.angio 0.69
## Phy.Abd.Bdiv.angio 1.00Partition comparison between five cluster groupings
compPart.5## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.00 0.73 0.69 0.71 0.69
## Bdiv.vasc 0.73 1.00 0.71 0.77 0.70
## Phy.Bdiv.vasc 0.69 0.71 1.00 0.69 0.68
## Bdiv.angio 0.71 0.77 0.69 1.00 0.74
## Phy.Bdiv.angio 0.69 0.70 0.68 0.74 1.00Partition comparison between five cluster groupings for abundance-weighted groupings
compPart.5.abd## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.00 0.68 0.67 0.68
## Abd.Bdiv.vasc 0.68 1.00 0.71 0.90
## Phy.Abd.Bdiv.vasc 0.67 0.71 1.00 0.69
## Abd.Bdiv.angio 0.68 0.90 0.69 1.00
## Phy.Abd.Bdiv.angio 0.67 0.73 0.72 0.74
## Phy.Abd.Bdiv.angio
## Env.dist 0.67
## Abd.Bdiv.vasc 0.73
## Phy.Abd.Bdiv.vasc 0.72
## Abd.Bdiv.angio 0.74
## Phy.Abd.Bdiv.angio 1.00Optimal number of grouping according to Calinski criteria
#vasc.env.clust.3
#veg.abd.k.clust
#abd.veg.abd.k.clust
vasc.dist.clust.eight<-kmeans(trans.dist,8)
abd.vasc.dist.clust.fourteen<-kmeans(abd.trans.dist,14)
angio.dist.bd.k.clust.2<-kmeans(angio.dist.bd, 2)
#abd.angio.abd.k.clust.5
angio.dist.clust.two<-kmeans(angio.trans.dist, 2)
abd.angio.dist.clust.seven<-kmeans(angio.dist.abd,7)Show PCoA’s with clusters for optimal numbers of groupings
# Not - abundance weighted
# Plot PCoA with 3 groupings based on k-means clustering for environmental distances
#dev.new(width=11.8, height=8)
par(mfrow=c(2,3))
vasc.env.3<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
x<-vasc.env.3[,1]
y<-vasc.env.3[,2]
plot(vasc.env.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.env.3, vasc.env.clust.3$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.3, vasc.env.clust.3$cluster, col="grey70")
box(lwd=2)
veg.abd<-cmdscale(sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with three clusters based on K-means clustering", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(veg.abd, veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(veg.abd, veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
vasc.abd.k.8<-cmdscale(trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(vasc.abd.k.8, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with eight\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.abd.k.8, vasc.dist.clust.eight$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.abd.k.8, vasc.dist.clust.eight$cluster, col="grey70")
box(lwd=2)
pcoa.angio.dist.bd.2<-cmdscale(angio.dist.bd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(pcoa.angio.dist.bd.2, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's beta \ndiversity with two clusters based on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(pcoa.angio.dist.bd.2, angio.dist.bd.k.clust.2$cluster, col="grey40", label=TRUE)
ordiellipse(pcoa.angio.dist.bd.2, angio.dist.bd.k.clust.2$cluster, col="grey70")
box(lwd=2)
angio.distk.2<-cmdscale(angio.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.distk.2, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with two\nclusters based on K-means clustering on Chao's phylogenetic \nbeta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.distk.2, angio.dist.clust.two$cluster, col="grey40", label=TRUE)
ordiellipse(angio.distk.2, angio.dist.clust.two$cluster, col="grey70")
box(lwd=2)
Show PCoA’s with clusters for optimal numbers of groupings (abundance-weighted)
#abundance-weighted
#Plot PCoA with 3 groupings based on k-means clustering for environmental distances
#dev.new(width=11.8, height=8)
par(mfrow=c(2,3))
vasc.env.3<-cmdscale(site.env.dist, k = 54, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(vasc.env.3, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with three \nclusters based on K-means clustering on \npairwise environmental distances", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(vasc.env.3, vasc.env.clust.3$cluster, col="grey40", label=TRUE)
ordiellipse(vasc.env.3, vasc.env.clust.3$cluster, col="grey70")
box(lwd=2)
abd.veg.abd<-cmdscale(abd.sp.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.veg.abd, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with three clusters based on K-means clustering (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey40", label=TRUE)
ordiellipse(abd.veg.abd, abd.veg.abd.k.clust$cluster, col="grey70")
box(lwd=2)
abd.vasc.abd.k.14<-cmdscale(abd.trans.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.vasc.abd.k.14, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with fourteen\nclusters based on K-means clustering on Chao's abundance-weighted \nphylogenetic beta diversity pairwise distances (vasculars)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.vasc.abd.k.14, abd.vasc.dist.clust.fourteen$cluster, col="grey40", label=TRUE)
ordiellipse(abd.vasc.abd.k.14, abd.vasc.dist.clust.fourteen$cluster, col="grey70")
box(lwd=2)
abd.angio.abd.5<-cmdscale(angio.dist.abd, k = 100, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(abd.angio.abd.5, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) on Chao's abundance-weighted \nbeta diversity with five clusters \nbased on K-means clustering (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey40", label=TRUE)
ordiellipse(abd.angio.abd.5, abd.angio.abd.k.clust.5$cluster, col="grey70")
box(lwd=2)
angio.abd.k.7<-cmdscale(abd.angio.dist, k = 75, eig = FALSE, add = FALSE, x.ret = FALSE)
plot(angio.abd.k.7, type="n", xlab="", ylab="", main="Principal Coordinate Analysis (PCoA) with seven\nclusters based on K-means clustering on abundance-weighted \nChao's phylogenetic beta diversity pairwise distances (angiosperms)", cex=1, pch=16, col="black", cex.main=0.9)
ordispider(angio.abd.k.7, abd.angio.dist.clust.seven$cluster, col="grey40", label=TRUE)
ordiellipse(angio.abd.k.7, abd.angio.dist.clust.seven$cluster, col="grey70")
box(lwd=2)
Partition comparison for optimal number of groups
compPart.optim## Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio Phy.Bdiv.angio
## Env.dist 1.00 0.66 0.68 0.50 0.57
## Bdiv.vasc 0.66 1.00 0.73 0.66 0.66
## Phy.Bdiv.vasc 0.68 0.73 1.00 0.46 0.59
## Bdiv.angio 0.50 0.66 0.46 1.00 0.51
## Phy.Bdiv.angio 0.57 0.66 0.59 0.51 1.00Partition comparison for optimal number of groups for abundance-weighted groupings
compPart.optim.abd## Env.dist Abd.Bdiv.vasc Phy.Abd.Bdiv.vasc Abd.Bdiv.angio
## Env.dist 1.00 0.66 0.63 0.65
## Abd.Bdiv.vasc 0.66 1.00 0.72 0.75
## Phy.Abd.Bdiv.vasc 0.63 0.72 1.00 0.78
## Abd.Bdiv.angio 0.65 0.75 0.78 1.00
## Phy.Abd.Bdiv.angio 0.64 0.75 0.83 0.89
## Phy.Abd.Bdiv.angio
## Env.dist 0.64
## Abd.Bdiv.vasc 0.75
## Phy.Abd.Bdiv.vasc 0.83
## Abd.Bdiv.angio 0.89
## Phy.Abd.Bdiv.angio 1.00Tammy’s classification per habitat type reduced to five different groupings
library(ggplot2)
#dev.new(width=8, height=6)
hab<-site[,c(3:5), drop=FALSE]
hab.types<-hab[,1, drop=FALSE]
# Show number of times that each habitat type represented
table(hab.types[,1])##
## Fen Shrubs
## 17 31
## Spruce lichen Spruce lichen - Shrubs
## 2 2
## Spruce moss Spruce moss - Fen
## 67 16
## Spruce moss - Shrubs Spruce moss - Spruce lichen
## 4 6
## Tundra Tundra - Shrubs
## 23 8#Change values so that less levels - split categories go into the smaller of the two factors
# put Spruce lichen - shrubs into Spruce lichen
# put Spruce moss - Spruce lichen into Spruce lichen
# put Spruce moss - Shrubs into Spruce moss
# put Spruce moss - Fen into Fen
# put Tundra - shrubs into Tundra
library(plyr)## Warning: package 'plyr' was built under R version 3.1.1hab.types.smaller<-revalue(hab.types[,1], c("Spruce lichen - Shrubs"="Spruce lichen" , "Spruce moss - Spruce lichen"="Spruce lichen",
"Spruce moss - Shrubs"="Spruce moss", "Spruce moss - Fen" = "Fen", "Tundra - Shrubs"="Tundra"))
table(hab.types.smaller)## hab.types.smaller
## Fen Shrubs Spruce lichen Spruce moss Tundra
## 33 31 10 71 31hab.new<-cbind(hab.types.smaller,hab[,c(2:3)])
hab.new.ggplot<-ggplot(hab.new, aes(hab.new$Latitude..WGS.84.datum., hab.new$Longitude..WGS.84.datum., colour = hab.new$hab.types.smaller)) + geom_point(size=4) +
theme_bw() + theme(panel.border = element_rect(colour="black"), panel.grid.major = element_blank(),
panel.grid.minor = element_blank(), axis.line = element_line(colour = "black"))+ xlab("Latitude") + ylab("Longitude") +
ggtitle("Habitat Types for Plots in Large Sampling Grid \n(Tammy's Classifications with Regroupings)")
hab.new.ggplot+theme(axis.title.y = element_text(size=16, vjust=1),axis.title.x = element_text(size=16), axis.text.x = element_text(size=12), axis.text.y= element_text(size=12,angle=90,hjust=0.5))+
theme(plot.title = element_text(vjust = 2, size=15, face="bold")) + theme(legend.title = element_text(size=12))+
scale_linetype_discrete("Habitat Type") +
scale_shape_discrete("Habitat Type") +
scale_colour_discrete("Habitat Type")
Plot Tammy’s classifications with other 5 cluster groupings (without abundance-weighted groupings)
# Plot this with other non-abundance weighted grid representations
#dev.new(width=15, height=8.5)
par(mfrow=c(2,3))
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(hab.new$hab.types.smaller))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on \nTammy's clasifications (5 groups)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.env.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on \nK-means clustering of environmental distances", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(veg.abd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means \nclustering of Chao's beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(vasc.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups based on K-means clustering \nof Chao's phylogenetic beta diversity (vasculars)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.bd.k.clust.5$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
myColorRamp <- function(colors, values) {
v <- (values - min(values))/diff(range(values))
x <- colorRamp(colors)(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
cols <- myColorRamp(c("black", "gray25", "gray50", "gray85", "white"), as.numeric(angio.dist.clust.five$cluster))
colfunc <- colorRampPalette(c("black", "gray25", "gray50", "gray85", "white"))
par(mar=c(5.1,4.1,4.1,2.1))
plot(site.trunc$Latitude..WGS.84.datum., site.trunc$Longitude..WGS.84.datum., xlab="Latitude", ylab="Longitude",
bg=cols, pch=21, col="black",cex=1.75, cex.lab=1.5, cex.axis=1.25, main="Locations of different groups from K-means clustering \nof Chao's phylogenetic beta diversity (angiosperms)", cex.main=1.25)
xl <- 1
yb <- -2.3
xr <- 1.5
yt <- 2.7
legend("bottomright", c("Group 1","Group 2","Group 3", "Group 4", "Group 5"), pch=21,
pt.bg=c("black", "gray25", "gray50", "gray85", "white"),col="black", cex=0.85, bty="n")
Comparing partitions of five groupings with Tammy five group classifications (for non-abundance weighted data)
compPart.5.tam## Tammy Class Env.dist Bdiv.vasc Phy.Bdiv.vasc Bdiv.angio
## Tammy Class 1.00 0.76 0.70 0.67 0.71
## Env.dist 0.76 1.00 0.73 0.69 0.71
## Bdiv.vasc 0.70 0.73 1.00 0.71 0.77
## Phy.Bdiv.vasc 0.67 0.69 0.71 1.00 0.69
## Bdiv.angio 0.71 0.71 0.77 0.69 1.00
## Phy.Bdiv.angio 0.68 0.69 0.70 0.68 0.74
## Phy.Bdiv.angio
## Tammy Class 0.68
## Env.dist 0.69
## Bdiv.vasc 0.70
## Phy.Bdiv.vasc 0.68
## Bdiv.angio 0.74
## Phy.Bdiv.angio 1.00