Ecological divergence and evolutionary transition of resprouting types in Banksia attenuata

Abstract

Resprouting is a key functional trait that allows plants to survive diverse disturbances. The fitness benefits associated with resprouting include a rapid return to adult growth, early flowering, and setting seed. The resprouting responses observed following fire are varied, as are the ecological outcomes. Understanding the ecological divergence and evolutionary pathways of different resprouting types and how the environment and genetics interact to drive such morphological evolution represents an important, but under-studied, topic. In the present study, microsatellite markers and microevolutionary approaches were used to better understand: (1) whether genetic differentiation is related to morphological divergence among resprouting types and if so, whether there are any specific genetic variations associated with morphological divergence and (2) the evolutionary pathway of the transitions between two resprouting types in Banksia attenuata (epicormic resprouting from aerial stems or branch; resprouting from a underground lignotuber). The results revealed an association between population genetic differentiation and the morphological divergence of postfire resprouting types in B. attenuata. A microsatellite allele has been shown to be associated with epicormic populations. Approximate Bayesian Computation analysis revealed a likely evolutionary transition from epicormic to lignotuberous resprouting in B. attenuata. It is concluded that the postfire resprouting type in B. attenuata is likely determined by the fire's characteristics. The differentiated expression of postfire resprouting types in different environments is likely a consequence of local genetic adaptation. The capacity to shift the postfire resprouting type to adapt to diverse fire regimes is most likely the key factor explaining why B. attenuata is the most widespread member of the Banksia genus.

Keywords: Epicormic; evolutionary transition; genetic differentiation; lignotuber; morphological divergence; resprouting.

Figure 1

Map of the studied populations relative to the distribution of Banksia attenuata and concordance with Bayesian clustering analysis. The area west of the green dotted line corresponds to the approximate range of the species; the red dotted line is the suggested boundary between the growth forms (Cowling and Lamont ; Taylor and Hopper 1988). Gray markers represent population surveyed, colored markers represent population genotyped. Gray curve indicates rough 300-mm isohyet which defines the east boundary of South Western Australia Floristic Region.

Figure 2

The distinctive morphologies of the two postfire resprouting types of Banksia attenuata. (A) Tree form (8.6 m in the figure); (B) epicormic resprouting (1 year after fire); (C) shrub form (1.8 m in the figure); (D) lignotuberous resprouting (1 year after fire).

Figure 3

Genetic differentiation measured as Jost's D of the 11 microsatellite loci. Blue represents with differentiation within lignotuberous populations, green for within epicormic populations, red for between lignotuberous and epicormic populations. The line represents the lower and upper values generated from 1000 bootstrapping.

Figure 4

F_ST values for 11 microsatellite loci examined in B. attenuata populations plotted against the heterozygosity. The lines represent the median and 99% quantiles (dotted lines) of the expected F_ST values from a neutral model. (A) Analysis including all populations; (B) epicormic populations only; (C) lignotuberous populations only.

Figure 5

Comparison of the hypotheses for the evolutionary transition of postfire resprouting types in Banksia attenuata. (A) The three tested scenarios (Pop1 and N1: lignotuberous population and its effective population size; Pop2 and N2: epicormic population and its effective population size; Nb: new population with derived resprouting type; t₁: time before present when new derived resprouting type first appeared; db: the time period of population bottleneck); (B) direct comparison of the posterior probabilities of the three scenarios; (C) logistic regression of the posterior probabilities of the three scenarios.