MULTILEVEL ANALYSIS OF MORPHOMETRIC DATA FROM NATURAL PLANT POPULATIONS: INSIGHTS INTO ONTOGENETIC, GENETIC, AND SELECTIVE CORRELATIONS IN DALECHAMPIA SCANDENS

Publisher Website
Google Scholar
Abstract
I examined patterns of covariation of three morphometric blossom characters [gland area (GA), gland–stigma distance (GSD), and bract length (BL)] within genets, among genets, and among populations of the tropical vine, Dalechampia scandens (Euphorbiaceae). Covariance between BL and GA was evenly distributed among the three levels. This observation, coupled with developmental information, indicates that the two characters change size similarly during development, that there is probably genetic covariance between them (apparently caused by pleiotropy), and that the genetic covariance may have constrained (at least proximally) the course of population differentiation with respect to these characters. Most covariance between GSD and GA occurred at the among-population level. This observation, coupled with developmental information, indicates that there is negligible ontogenetic covariance and that within populations there is probably little or no genetic covariance between the two characters. Among-population covariance has probably been caused by natural selection operating in a correlated fashion on characters that functionally interact in pollination.
Funding Information
  • National Science Foundation (BSR‐8509031)