Worked example 2

Worked example 2 illustrates the use of Diversity.R functions to perform multivariate analyses that are consistent with diversity measurements, the CA-richness and NSCA-Simpson strategies from Pélissier et al. (2003). It is based on analyses of PSE data set, from a 10-ha rainforest plot at Piste de St-Elie station (PSE) in French Guiana (Sabatier et al. 1997).

1 - PSE data set

Once the library diversity has been installed and loaded:

data(PSE)
str(PSE)

#List of 3
# $ Tfac: Factor w/ 113 levels "Alexa wache..",..: 1 1 2 2 3 3 3 4 5 6 ...
# $ Sfac: Factor w/ 9 levels "Alt","DVD","DhS",..: 7 5 2 1 2 1 2 3 1 3 ...
# $ Qfac: Factor w/ 80 levels "q1","q10","q11",..: 10 13 46 64 45 20 48 52 31 15 ... 

PSE$Tfac is a factor of 113 species names assigned to 381 trees ;
PSE$Sfac is a factor of 9 soil classes
assigned to 381 trees ;
PSE$Qfac is a factor of 80 quadrat codes assigned to 381 trees.


2 - CA-richness and NSCA-Simpson strategies with respect to the soil classes

Use the following commands to perform a CA-richness strategy with respect to the soil classes:

rich.sol<-ca.richness(PSE$Tfac~PSE$Sfac)
#Select the number of axes:
2
summary(rich.sol)

#class:  div between dudi
#metric: Richness
#call: ca.richness.formula(formula = PSE$Tfac ~ PSE$Sfac)
#total diversity: 112
#explained diversity: 2.33
#ratio of explained diversity: 0.0208
#Pr(>ratio): 0.551 based on 999 replicates

plot(rich.sol)


Use the following commands to perform a NSCA-Simpson strategy with respect to the soil classes:

simp.sol<-nsca.simpson(PSE$Tfac~PSE$Sfac)
#Select the number of axes:
2
summary(simp.sol)
#class:  div between dudi
#metric: Simpson
#call: nsca.simpson.formula(formula = PSE$Tfac ~ PSE$Sfac)
#total diversity: 0.933
#explained diversity: 0.0375
#ratio of explained diversity: 0.0402
#Pr(>ratio): 0.001 based on 999 replicates

plot(simp.sol)


These results conform to Table 2 and Figure 2 in Pélissier et al. (2003).

3 - CA-richness and NSCA-Simpson strategies with respect to the partition into quadrats

Use the following commands to perform a CA-richness strategy with respect to quadrats:

rich.quad<-ca.richness(PSE$Tfac~PSE$Qfac)
#Select the number of axes:

2
summary(rich.quad)
#class:  div between dudi
#metric: Richness
#call: ca.richness.formula(formula = PSE$Tfac ~ PSE$Qfac)
#total diversity: 112
#explained diversity: 23.6
#ratio of explained diversity: 0.211
#Pr(>ratio): 0.301 based on 999 replicates

plot(rich.quad)


Projection of the soil classes at the weigthed mean position of their occurrences (as in Fig.4a in Pélissier et al. 2003) can be obtained from:

s.class(rich.quad$ls,PSE$Sfac,cell=0,cstar=0,cpoint=0,xlim=c(-1.8,0.3),ylim=c(-0.9,0.6))




Use the following commands to perform a NSCA-Simpson strategy with respect to quadrats:

simp.quad<-nsca.simpson(PSE$Tfac~PSE$Qfac)
#Select the number of axes:

2

summary(simp.quad)

#class:  div between dudi
#metric: Simpson
#call: nsca.simpson.formula(formula = PSE$Tfac ~ PSE$Qfac)
#total diversity: 0.933
#explained diversity: 0.214
#ratio of explained diversity: 0.229
#Pr(>ratio): 0.001 based on 999 replicates

plot(simp.quad)

Projection of the soil classes at the weigthed mean position of their occurrences (as in Fig.4b in Pélissier et al. 2003) can be obtained from:

> s.class(simp.quad$ls,PSE$Sfac,cell=0,cstar=0,cpoint=0,xlim=c(-0.4,0.2),ylim=c(-0.2,0.2))


 

These results conform to Table 3 and Figure 4 in Pélissier et al. (2003).


Literature cited

Sabatier, D., Grimaldi, M., Prévost, M.-F. , Guillaume, J., Godron, M., Dosso, M. and Curmi, P. 1997. The influence of soil cover organization on the floristic and structural heterogeneity of a Guianan rain forest. Plant Ecology, 131:81-108

Pélissier, R., Couteron, P., Dray, S. and Sabatier, D. 2003. Consistency between ordination techniques and diversity measurements: two strategies for species occurrence data. Ecology, 84:242-251.