Phenotypic selection mosaic for flower length influenced by geographically varying hawkmoth pollinator proboscis length and abiotic environment

New Phytol. 2020 Jan;225(2):985-998. doi: 10.1111/nph.16192. Epub 2019 Oct 30.

Abstract

Biotic and abiotic context may affect the intensity of interspecific interactions and subsequently drive locally particular phenotypic selection patterns on interacting traits. We evaluated the geographical variation of matching traits of the brush-type flowers of Caesalpinia gilliesii and of the proboscis length of its guild of hawkmoth pollinators, as well as their relationship with environmental variables. We assessed the geographical variation of interacting traits (style and filament vs mean proboscis length of the guild of hawkmoths) across seven populations and estimated phenotypic selection on the plant side. Interacting traits showed similar relationships with environmental variables. Phenotypic selection on the plant side was influenced by proboscis length and by environmental conditions. Mean proboscis length of the guild was shorter than previously recorded for the same study area, thus probably shifting the selective optima of flower length. We observed two presumptive coevolutionary cold spots where one-sided negative directional selection is acting on style length. The lack of selection on the pollinator side should be further confirmed. We provided joint evidence, mostly lacking, about the geographical variation of selective pressures on the plant side associated with both proboscis length and abiotic conditions. We suggest that recent environmental change may be shifting floral length optima.

Keywords: abiotic environment; coadaptation; geographic range; nocturnal plant-pollinator interactions; phenotypic selection mosaic; trait matching.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Biological Evolution
  • Caesalpinia / anatomy & histology
  • Environment*
  • Flowers / anatomy & histology*
  • Geography*
  • Moths / anatomy & histology*
  • Multivariate Analysis
  • Phenotype
  • Pollination / physiology*