Accessory costs of seed production and the evolution of angiosperms

Janice M. Lord*, Mark Westoby

*Corresponding author for this work

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Accessory costs of reproduction frequently equal or exceed direct investment in offspring, and can limit the evolution of small offspring sizes. Early angiosperms had minimum seed sizes, an order of magnitude smaller than their contemporaries. It has been proposed that changes to reproductive features at the base of the angiosperm clade reduced accessory costs thus removing the fitness disadvantage of small seeds. We measured accessory costs of reproduction in 25 extant gymnosperms and angiosperms, to test whether angiosperms can produce small seeds more economically than gymnosperms. Total accessory costs scaled isometrically to seed mass for angiosperms but less than isometrically for gymnosperms, so that smaller seeds were proportionally more expensive for gymnosperms to produce. In particular, costs of abortions and packaging structures were significantly higher in gymnosperms. Also, the relationship between seed:ovule ratio and seed size was negative in angiosperms but positive in gymnosperms. We argue that the carpel was a key evolutionary innovation reducing accessory costs in angiosperms by allowing sporophytic control of pre- and postzygotic mate selection and timing of resource allocation. The resulting reduction in costs of aborting unfertilized ovules or genetically inferior embryos would have lowered total reproductive costs enabling early angiosperms to evolve small seed sizes and short generation times.

Original languageEnglish
Pages (from-to)200-210
Number of pages11
JournalEvolution
Volume66
Issue number1
DOIs
Publication statusPublished - Jan 2012

Keywords

  • accessory costs
  • angiosperms
  • gymnosperms
  • reproduction
  • seed mass

Fingerprint Dive into the research topics of 'Accessory costs of seed production and the evolution of angiosperms'. Together they form a unique fingerprint.

Cite this