Cypsela diversity in some members of Compositae through LM study and its taxonomic significance.

Abstract

Compositae (Asteraceae), also known as the sunflower family has evolved to become the most diverse and advanced angiosperm family. However, with high diversity, taxonomic complexities also arise at several levels. A reliable tool in resolving such taxonomic conundrums in this family is the characteristic fruit of Compositae known as "cypsela." While most of the previous cypsela-based studies have focused on specific tribes and species of Compositae, in this study, we shift our focus to the genus level and seek a broader understanding of the taxonomic relationships that prevail in this family. In order to achieve this, we sampled 25 genera of Compositae and subjected them to qualitative and quantitative analysis to resolve them into groups based on the degree of similarity. This study was conducted by employing light microscopy techniques for gathering the qualitative and quantitative morphological data of the cypselae. The qualitative data were collected on six key morphological characteristics-shape, color, base, surface ornamentation, ribs, and pappus types for all the members. These data were used to yield an artificial dichotomous key to genera for the 25 members. The quantitative data on dimensions (length, breadth) of the entire cypsela as well as the pappus and ribs were subjected to principal component analysis, to understand the covariance among these variables. Hierarchical clustering followed by dendrogram construction was used to visualize the patterns of correlation within the samples and make taxonomic conclusions. The resulting dendrogram showed four closely related clades that consisted of 23 out of the 25 members whereas 2 members emerged as outgroups. The results of this study commensurate with DNA-based resolution of the concerned Compositae members, hence highlighting the reliability and significance of both light microscopy and morphology-based studies that seem to be fading away otherwise.