← Back to results
Page 51 · DO #4926 · 231_Rogers_Bx9FF8_r
- Collection
-
David James Rogers (1918–2007) papers
- Item/Folder
-
Manihot - Correspondence; Publication, 1972–1973
- Digital Object
-
DO #4926,
page 51
- Collection-level dates
- 1948–1977
Page transcription
connectors between the groups (more than one pair with the same similarity value in different groups).
As noted before, we modified or restricted the selection of nodal cultivars to those cultivars that were classified at the early levels (higher c-values or similarity ratios). Logically, any of the cultivars that satisfied the criterion for being a node, regardless of the level or amount of similarity, could have been selected. The hierarchical arrangement could have been continued and groups based on nodal cultivars with lower c-values could have been designated as subgroups. However, since the understanding of the biology of the species and genus is very limited at the present time, in spite of its importance as a world crop, it has not been considered either wise or productive of a better understanding of the complex to divide the groups of cultivars any further.
There would also have been good taxonomic justification for establishing a dozen or so single membered groups of cultivars having low similarity values because of the unusual attributes or attribute combinations present in these cultivars. We have placed them in the group with which they show the greatest affinity [S(a,b) value]. Additional relationships for these distinctive cultivars can easily be found in Tables 33, 34 or in the respective group figures.
At the same time, the groups established satisfy the taxonomist's need to have some method of indicating relationships in groups which have mosaic or reticulate interrelationships. The groups will disappoint him to the extent that seldom can one expect an invariant set of exclusive descriptions that will encompass all included cultivars. Previous attempts that have been made to resolve reticulate relationships have been (1) select those characters that are the most easily determined and divide the population on a ranking of these characters, and (2) select characters on the basis of their phylogenetic or ancestral importance. The first method almost invariably results in placing
