The resulting object contains the following tables:
• relfrq ¼ relative frequency of the species in each group ¼ number of sites
where species is present/number of sites in the group
• relabu ¼ relative abundance of the species across groups ¼ total abundance in
group/grand total abundance
• indval ¼ indicator value (IndVal) of each species
• maxcls ¼ cluster where the species has highest IndVal
• indcls ¼ highest IndVal of the species
• pval ¼ permutational p-value of IndVal
Correct the p-values for multiple testing:
pval.adj <- p.adjust(iva$pval)
The following lines of code extract the significant indicator species with their
group with highest IndVal, the corresponding IndVal, the species’ indicator p-value
and their total frequency in the data set.
# Table of the significant indicator species
gr <- iva$maxcls[pval.adj <= 0.05]
iv <- iva$indcls[pval.adj <= 0.05]
pv <- iva$pval[pval.adj <= 0.05]
fr <- apply(spe > 0, 2, sum)[pval.adj <= 0.05]
fidg <- data.frame(
group = gr,
indval = iv,
pvalue = pv,
freq = fr
)
fidg <- fidg[order(fidg$group, -fidg$indval), ]
fidg
# Export the result to a CSV file (to be opened in a spreadsheet)
write.csv(fidg, "IndVal-dfs.csv")
On this basis, what do you think of the results? How do they relate to the four
ecological zones presented in Chap.1?
Note that the indicator species identified here may differ from the members of the
species assemblages identified in Sect. 4.10. The indicator species are linked to
predefined groups, whereas the species assemblages are identified without any
prior classification or reference to environmental conditions.
In Sect. 4.12 you will find another application in relationship with the MRT
method. Some detailed results will be presented.
122
4 Cluster Analysis
• relfrq ¼ relative frequency of the species in each group ¼ number of sites
where species is present/number of sites in the group
• relabu ¼ relative abundance of the species across groups ¼ total abundance in
group/grand total abundance
• indval ¼ indicator value (IndVal) of each species
• maxcls ¼ cluster where the species has highest IndVal
• indcls ¼ highest IndVal of the species
• pval ¼ permutational p-value of IndVal
Correct the p-values for multiple testing:
pval.adj <- p.adjust(iva$pval)
The following lines of code extract the significant indicator species with their
group with highest IndVal, the corresponding IndVal, the species’ indicator p-value
and their total frequency in the data set.
# Table of the significant indicator species
gr <- iva$maxcls[pval.adj <= 0.05]
iv <- iva$indcls[pval.adj <= 0.05]
pv <- iva$pval[pval.adj <= 0.05]
fr <- apply(spe > 0, 2, sum)[pval.adj <= 0.05]
fidg <- data.frame(
group = gr,
indval = iv,
pvalue = pv,
freq = fr
)
fidg <- fidg[order(fidg$group, -fidg$indval), ]
fidg
# Export the result to a CSV file (to be opened in a spreadsheet)
write.csv(fidg, "IndVal-dfs.csv")
On this basis, what do you think of the results? How do they relate to the four
ecological zones presented in Chap.1?
Note that the indicator species identified here may differ from the members of the
species assemblages identified in Sect. 4.10. The indicator species are linked to
predefined groups, whereas the species assemblages are identified without any
prior classification or reference to environmental conditions.
In Sect. 4.12 you will find another application in relationship with the MRT
method. Some detailed results will be presented.
122
4 Cluster Analysis