Biodiv Sci ›› 2019, Vol. 27 ›› Issue (7): 772-777.DOI: 10.17520/biods.2019067

• Original Papers • Previous Articles     Next Articles

Phenotypic selection analysis of flower traits in Delphinium kamaonense var. glabrescens (Ranunculaceae)

Lu Ningna1,2,*(),Liu Zhenheng3,Ma Yan2,Lu Guangmei2,Meng Xiuxiang1,*()   

  1. 1 School of Environment & Natural Resources, Renmin University of China, Beijing 100872
    2 School of Life Sciences, Northwest Normal University, Lanzhou 730000
    3 Grassland Station, Maqu, Gannan Prefecture, Gansu 747300
  • Received:2019-03-08 Accepted:2019-06-20 Online:2019-07-20 Published:2019-08-21
  • Contact: Lu Ningna,Meng Xiuxiang

Abstract:

It has been widely accepted that pollinator-mediated selection is responsible for the evolution of flower traits. To date, two distinct methods of phenotypic manipulation and phenotypic selection analyses have been used to demonstrate flower function and potential selection. However, empirical studies combining both methods are rare in wild plants. Through manipulating sepal size of Delphinium kamaonense var. glabrescens (Ranunculaceae) and performing phenotypic selection analysis, this study examined functions of two flower traits and the potential selection on them. Sepals manipulated to be smaller significantly decreased visitation rate of pollinators, but did not affect seed number per flower or seed set. Sepal size mainly affected pollen export mediated by pollinator attraction rather seed production. Phenotypic selection estimated through female fitness showed no significant gradient with varying sepal width and length. However, significant linear selection differential and selection gradient on the spur length were observed, showing higher seed production in individuals with longer floral spurs. This study suggests different function and potential fitness consequences of floral traits that were subject to natural selection.

Key words: Delphinium kamaonense var. glabrescens, phenotypic selection, phenotype evolution, fitness, flower size, trait manipulation