This document is a supplement to the article “Ordination analysis in sedimentology, geochemistry and paleoenvironment - background, current trends and recommendations” https://doi.org/10.31223/X5N31Q

We use the transformed dataset and also standardize the variables to make them comparable.

reduced3t_std <- decostand(reduced3t, method="range", MARGIN =2)
pairs(reduced3t_std,
      lower.panel = NULL, 
      col = brewer.pal(7, name="RdBu")[sample.prop$`Lithological descripton`])
par(xpd = TRUE)
legend("bottomleft", fill = brewer.pal(7, name="RdBu"), legend = c( levels(sample.prop$`Lithological descripton`)), cex=0.7)

Teh distributions should remain the same, but standardized to the same range (0 to 1). Now we can perform DCA and extract sample and variable scores.

reduced3_DCA <- decorana(reduced3t_std, ira=0)
variablescores <- as.data.frame(scores(reduced3_DCA, display = "species"))
variablescores$species <- rownames(variablescores)
samplescores <- as.data.frame(scores(reduced3_DCA, display="sites"))
samplescores$sample <- rownames(samplescores)
samplescores$Lithology <- sample.prop$`Lithological descripton`

We use extracted sample and variable scores to plot, because they are easier to handle as separate objects, compared to the clunky DCA output object.

ggplot() + 
  geom_text(data=variablescores,aes(x=DCA1,y=DCA2,label=species),size=3) +  
  geom_point(data=samplescores,aes(x=DCA1,y=DCA2,shape=Lithology,colour=Lithology),size=5) + 
  geom_text(data=samplescores,aes(x=DCA1,y=DCA2,label=sample),size=2.5,vjust=0, alpha=0.5) +
  scale_colour_brewer(palette="RdBu") +
  coord_equal()+
    theme_bw()+
  theme(legend.position = "none")

As with NMDS, the plot will be more legible if we focus on samples, but the above graphic allows us to interpret the “gradient” reconstructed by DCA.

Adding convex hulls:

grp.Limestone <- samplescores[samplescores$Lithology == "Limestone",][chull(samplescores[samplescores$Lithology == "Limestone", c("DCA1", "DCA2")]), ]

grp.Dolomite <- samplescores[samplescores$Lithology == "Dolomite", ][chull(samplescores[samplescores$Lithology == "Dolomite", c("DCA1", "DCA2")]), ]  

grp.Marly_dolomite <- samplescores[samplescores$Lithology == "Marly dolomite", ][chull(samplescores[samplescores$Lithology == "Marly dolomite", c("DCA1", "DCA2")]), ]  


hull.data <- rbind(grp.Limestone, grp.Dolomite, grp.Marly_dolomite)  

ggplot() + 
    geom_polygon(data=hull.data,aes(x=DCA1,y=DCA2,fill=Lithology,group=Lithology),alpha=0.30)+
  scale_fill_manual(values = c("#FDDBC7","#67A9CF","#D1E5F0"))+
  geom_point(data=samplescores,aes(x=DCA1,y=DCA2,shape=Lithology,colour=Lithology),size=3) + 
  geom_text(data=samplescores,aes(x=DCA1,y=DCA2,label=sample),size=2.5,vjust=0, alpha=0.5) +
  scale_colour_brewer(palette="RdBu") +
  coord_equal()+
    theme_bw()+
  theme(legend.position = "none")