Q. Describe Cluster analysis?

After making a similarity table, it is then rearranged so that OTU's whose members have the highest mutual similarity are brought together. This can be done by several methods and related taxas or groups are recognised. These clusters are called phenons and can be arranged hierarchically in a tree diagram or dendrogram.

The groups or clusters thus recognised may be treated as equivalent to the categories having ranks in classical taxonomy, such as the genus, family, order, etc. A problem faced by many taxonomists is whether there is any equivalence between the ranks in different taxonomic groups of organisms. Is a family of flowering plants, for example equivalent in any sense to one of algae, or other organisms? To overcome this problem, numerical taxonomists have advocated a new terminology. Here, the tern "phenon" is introduced and the particular phenons are designated by numerical prefixes (e.g. 80 similarity) showing the level of. resemblance by which they are defined. The delimitation of the phenons is done by drawing horizontal lines across the dendrogram at a chosen similarity value. Such a dendrogram will have reference to a given study only, and cannot be generalised. Thus, phenons will be arbitrary and relative to groups within the limits of only one analysis.