J Plant Ecol ›› 2017, Vol. 10 ›› Issue (2): 406-413 .DOI: 10.1093/jpe/rtw035

• Research Articles • Previous Articles    

Do different measures of maternal fitness affect estimation of natural selection on floral traits? A lesson from Linum Pubescens (Linaceae)

Laura Bigio1,2, Merav Lebel1,3 and Yuval Sapir1,*   

  1. 1 The Botanical Garden, Department of Molecular Biology and Ecology of Plants, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel; 2 Earth Science Institute, The Hebrew University of Jerusalem, Givat Ram, Jerusalem 91904, Israel; 3 Nature-Parks Authority, Science Division, Am-Veolamo 2, Jerusalem 95463, Israel
  • Received:2016-01-24 Accepted:2016-04-11 Published:2017-03-25
  • Contact: Sapir, Yuval

Do different measures of maternal fitness affect estimation of natural selection on floral traits? A lesson from Linum Pubescens (Linaceae)

Abstract: Aims Floral traits in animal-pollinated plants are proposed to be selected by the behavior of pollinators, which create differential fitness. However, fitness estimation is not trivial and there is no agreed or universal measure of fitness. In plants, fitness can be measured in various ways, including the number of fruits or seeds, probability of survival, or growth rate.
Methods Because each fitness measure can be associated with different pollinator behavior, estimating pollinator-mediated selection on floral traits can differ among various fitness measures. We used the annual Mediterranean plant, Linum pubescens, as a model to assess the relative role of various fitness measures in estimating pollinator-mediated selection on floral traits. We compared four maternal fitness measures and estimated their relative advantage in assessing natural selection on floral traits of L. pubescens. To identify whether each trait is under pollinator-mediated selection, we used mediational analysis.
Important findings We found that each floral trait was under natural selection assessed by a different fitness measure. The color combination of floral tube throat and stamen color was under selection when using seed mass as a fitness measure, and floral diameter was under positive directional selection when the number of seeds or number of fruits was used as a fitness measure. Selection on floral color was not mediated by pollinators' visitation rates, while flower diameter showed partial pollinator-mediated selection. We conclude that using several fitness measures gives an insight into the mechanisms underlying phenotypic selection on a floral trait, and facilitates the understanding of pollinator-mediated selection.

Key words: fitness, flower size, fruit-set, Linum pubescens, natural selection, pollinator-mediated selection, pollinator visitation rate, seed mass

摘要:
Aims Floral traits in animal-pollinated plants are proposed to be selected by the behavior of pollinators, which create differential fitness. However, fitness estimation is not trivial and there is no agreed or universal measure of fitness. In plants, fitness can be measured in various ways, including the number of fruits or seeds, probability of survival, or growth rate.
Methods Because each fitness measure can be associated with different pollinator behavior, estimating pollinator-mediated selection on floral traits can differ among various fitness measures. We used the annual Mediterranean plant, Linum pubescens, as a model to assess the relative role of various fitness measures in estimating pollinator-mediated selection on floral traits. We compared four maternal fitness measures and estimated their relative advantage in assessing natural selection on floral traits of L. pubescens. To identify whether each trait is under pollinator-mediated selection, we used mediational analysis.
Important findings We found that each floral trait was under natural selection assessed by a different fitness measure. The color combination of floral tube throat and stamen color was under selection when using seed mass as a fitness measure, and floral diameter was under positive directional selection when the number of seeds or number of fruits was used as a fitness measure. Selection on floral color was not mediated by pollinators' visitation rates, while flower diameter showed partial pollinator-mediated selection. We conclude that using several fitness measures gives an insight into the mechanisms underlying phenotypic selection on a floral trait, and facilitates the understanding of pollinator-mediated selection.