Evolution of flowering time in a selfing annual plant: Roles of adaptation and genetic drift

Abstract

When facing changing environments, organisms can persist by one of three strategies: fleeing (migration), coping (plasticity), or adapting. If migration and plasticity can lead to rapid and reversible changes in the average phenotype of a population, adaptation proceeds through genetic changes and toward phenotypes with the highest fitness in a given environment. The literature describing adaptation in natural populations is vast (Bay et al., 2017;Côté & Reynolds, 2012;Kremer et al., 2012;Olson-Manning et al., 2012), and the recent rise of next generation sequencing has enabled tremendous progress in our knowledge about the genetic architecture of adaptation at the species level (

Genome scan of landrace populations of the self-fertilizing crop species rice, collected across time, revealed climate changes’ selective footprints in the genes network regulating flowering time

“…For instance, the flowering time advancement was of seven days on average, for 43 common plant species of North America for a temperature rise of 2.5 °C, over some 150 years (Miller-Rushing, 2008). It was of two days in a 22-generation resurrection experiment of the model legume Medicago truncatula (Gay et al, 2021), and of 5 to 10 days °C -1 , according to ecotypes, in Arabidopsis thaliana , under experimental conditions (Footitt et al, 2021). In comparison, the advance in DTHD of the O. sativa subpopulation Osj-2 relative to Osj-1 (3 days on average) was on the lowest side of the reported flowering time advancement.…”

“…The temporal clean approach (Lande 1977; Goldringer & Bataillon, 2004) we have implemented offers the advantage of using plant material in which the populations have actually adapted in contemporary times to changes in climate. However, the majority of the statistical methods detecting selection footprint were developed for outcrossing populations, and assumes random mating, making challenging the detection of loci under selection in predominantly selfing populations (Hartfield et al, 2017; Navascués et al, 2021). Indeed, given the extent of LD, selective sweeps in selfing populations can involve large genomic segments and hamper distinction of the genetic features between neutral and adaptive loci.…”

“…The Navascués et al (2021) method (hereafter referred to as “drift-based” method), relies on principles first proposed by Goldringer & Bataillon (2004), that is using the estimated magnitude of drift between two time samples as a null model to test for homogeneity of differentiation across loci. In the case of complete neutrality, all sampled loci should provide estimates of genetic differentiation drawn from the same distribution.…”

“…Loci under directional selection or linked to such loci are expected to exhibit larger differentiation values than expected under the neutral hypothesis. In order to account for the mainly selfing reproduction system, Navascués et al (2021) developed an implementation procedure that relies not on the observed allele frequency, but on genotype ( AA, Aa and aa ) frequency. Indeed, in a mainly selfing species, allele copies within an individual are not independent samples from the population, while genotypes are independent samples.…”

“…Selfing rate is assumed constant across generations. The method was implemented using DriftTest script (Navascués & Vitalis, 2020).…”