Trypsin inhibitor activities as useful taxonomic traits of sweetpotato varieties

Summary

Trypsin inhibitor activities (TIAs) of roots of 53 sweetpotato varieties were used as traits for intraspecific grouping. Three numerical taxonomic techniques were evaluated on the basis of breeding records of 42 varieties. When UPGMA was used as the clustering method, phenograms from average Euclidean distance and Manhattan distance performed better than those from correlation, and 8-character matrix gave better phenograms than 4- or 11-character matrix in average, when both node counts and detailed assignments of correct pairs were compared. The best phenogram simultaneously assigned 10 pairs exactly with a probability by chance of 2.84×10⁻¹⁰. For cladistic analysis, mid-point rooting Wagner in the Wagner78 program package was applied to construct 3 cladograms using 4-, 8-, or 11-character matrix as input data. The cladograms from 8- and 11-character matrices were better than that from 4-character matrix, when total relative path length required to search for group members of 11 varieties and assignments of correct pairs were compared. In all, TIAs are useful traits for intraspecific grouping of sweetpotato varieties used. Eleven varieties with unknown breeding records were assigned to appropriate groups accordingly.