- , , ,
Evolution of flowering time in a selfing annual plant: Roles of adaptation and genetic drift
Separating adaptation from drift: A cautionary tale from a self-fertilizing plant
In recent years many studies have documented shifts in phenology in response to climate change, be it in arrival times in migrating birds, budset in trees, adult emergence in butterflies, or flowering time in annual plants (Coen et al. 2018; Piao et al. 2019). While these changes are, in part, explained by phenotypic plasticity, more and more studies find that they involve also genetic changes, that is, they involve evolutionary change (e.g., Metz et al. 2020). Yet, evolutionary change may occur through genetic drift as well as selection. Therefore, in order to demonstrate adaptive evolutionary change in response to climate change, drift has to be excluded as an alternative explanation (Hansen et al. 2012). A new study by Gay et al. (2021) shows just how difficult this can be.
The authors investigated a recent evolutionary shift in flowering time by in a population an annual plant that reproduces predominantly by self-fertilization. The population has recently been subjected to increased temperatures and reduced rainfalls both of which are believed to select for earlier flowering times. They used a “resurrection” approach (Orsini et al. 2013; Weider et al. 2018): Genotypes from the past (resurrected from seeds) were compared alongside more recent genotypes (from more recently collected seeds) under identical conditions in the greenhouse. Using an experimental design that replicated genotypes, eliminated maternal effects, and controlled for microenvironmental variation, they found said genetic change in flowering times: Genotypes obtained from recently collected seeds flowered significantly (about 2 days) earlier than those obtained 22 generations before. However, neutral markers (microsatellites) also showed strong changes in allele frequencies across the 22 generations, suggesting that effective population size, Ne, was low (i.e., genetic drift was strong), which is typical for highly self-fertilizing populations. In addition, several multilocus genotypes were present at high frequencies and persisted over the 22 generations, almost as in clonal populations (e.g., Schaffner et al. 2019). The challenge was thus to evaluate whether the observed evolutionary change was the result of an adaptive response to selection or may be explained by drift alone.
Here, Gay et al. (2021) took a particularly careful and thorough approach. First, they carried out a selection gradient analysis, finding that earlier-flowering plants produced more seeds than later-flowering plants. This suggests that, under greenhouse conditions, there was indeed selection for earlier flowering times. Second, investigating other populations from the same region (all populations are located on the Mediterranean island of Corsica, France), they found that a concurrent shift to earlier flowering times occurred also in these populations. Under the hypothesis that the populations can be regarded as independent replicates of the evolutionary process, the observation of concurrent shifts rules out genetic drift (under drift, the direction of change is expected to be random).
The study may well have stopped here, concluding that there is good evidence for an adaptive response to selection for earlier flowering times in these self-fertilizing plants, at least under the hypothesis that selection gradients estimated in the greenhouse are relevant to field conditions. However, the authors went one step further. They used the change in the frequencies of the multilocus genotypes across the 22 generations as an estimate of realized fitness in the field and compared them to the phenotypic assays from the greenhouse. The results showed a tendency for high-fitness genotypes (positive frequency changes) to flower earlier and to produce more seeds than low-fitness genotypes. However, a simulation model showed that the observed correlations could be explained by drift alone, as long as Ne is lower than ca. 150 individuals. The findings were thus consistent with an adaptive evolutionary change in response to selection, but drift could only be excluded as the sole explanation if the effective population size was large enough.
The study did provide two estimates of Ne (19 and 136 individuals, based on individual microsatellite loci or multilocus genotypes, respectively), but both are problematic. First, frequency changes over time may be influenced by the presence of a seed bank or by immigration from a genetically dissimilar population, which may lead to an underestimation of Ne (Wang and Whitlock 2003). Indeed, the low effective size inferred from the allele frequency changes at microsatellite loci appears to be inconsistent with levels of genetic diversity found in the population. Moreover, high self-fertilization reduces effective recombination and therefore leads to non-independence among loci. This lowers the precision of the Ne estimates (due to a higher sampling variance) and may also violate the assumption of neutrality due to the possibility of selection (e.g., due to inbreeding depression) at linked loci, which may be anywhere in the genome in case of high degrees of self-fertilization.
There is thus no definite answer to the question of whether or not the observed changes in flowering time in this population were driven by selection. The study sets high standards for other, similar ones, in terms of thoroughness of the analyses and care in interpreting the findings. It also serves as a very instructive reminder to carefully check the assumptions when estimating neutral expectations, especially when working on species with complicated demographies or non-standard life cycles. Indeed the issues encountered here, in particular the difficulty of establishing neutral expectations in species with low effective recombination, may apply to many other species, including partially or fully asexual ones (Hartfield 2016). Furthermore, they may not be limited to estimating Ne but may also apply, for instance, to the establishment of neutral baselines for outlier analyses in genome scans (see e.g, Orsini et al. 2012).
Cohen JM, Lajeunesse MJ, Rohr JR (2018) A global synthesis of animal phenological responses to climate change. Nature Climate Change, 8, 224–228. https://doi.org/10.1038/s41558-018-0067-3
Gay L, Dhinaut J, Jullien M, Vitalis R, Navascués M, Ranwez V, Ronfort J (2021) Evolution of flowering time in a selfing annual plant: Roles of adaptation and genetic drift. bioRxiv, 2020.08.21.261230, ver. 4 recommended and peer-reviewed by Peer Community in Evolutionary Biology. https://doi.org/10.1101/2020.08.21.261230
Hansen MM, Olivieri I, Waller DM, Nielsen EE (2012) Monitoring adaptive genetic responses to environmental change. Molecular Ecology, 21, 1311–1329. https://doi.org/10.1111/j.1365-294X.2011.05463.x
Hartfield M (2016) Evolutionary genetic consequences of facultative sex and outcrossing. Journal of Evolutionary Biology, 29, 5–22. https://doi.org/10.1111/jeb.12770
Metz J, Lampei C, Bäumler L, Bocherens H, Dittberner H, Henneberg L, Meaux J de, Tielbörger K (2020) Rapid adaptive evolution to drought in a subset of plant traits in a large-scale climate change experiment. Ecology Letters, 23, 1643–1653. https://doi.org/10.1111/ele.13596
Orsini L, Schwenk K, De Meester L, Colbourne JK, Pfrender ME, Weider LJ (2013) The evolutionary time machine: using dormant propagules to forecast how populations can adapt to changing environments. Trends in Ecology & Evolution, 28, 274–282. https://doi.org/10.1016/j.tree.2013.01.009
Orsini L, Spanier KI, Meester LD (2012) Genomic signature of natural and anthropogenic stress in wild populations of the waterflea Daphnia magna: validation in space, time and experimental evolution. Molecular Ecology, 21, 2160–2175. https://doi.org/10.1111/j.1365-294X.2011.05429.x
Piao S, Liu Q, Chen A, Janssens IA, Fu Y, Dai J, Liu L, Lian X, Shen M, Zhu X (2019) Plant phenology and global climate change: Current progresses and challenges. Global Change Biology, 25, 1922–1940. https://doi.org/10.1111/gcb.14619
Schaffner LR, Govaert L, De Meester L, Ellner SP, Fairchild E, Miner BE, Rudstam LG, Spaak P, Hairston NG (2019) Consumer-resource dynamics is an eco-evolutionary process in a natural plankton community. Nature Ecology & Evolution, 3, 1351–1358. https://doi.org/10.1038/s41559-019-0960-9
Wang J, Whitlock MC (2003) Estimating Effective Population Size and Migration Rates From Genetic Samples Over Space and Time. Genetics, 163, 429–446. PMID: 12586728
Weider LJ, Jeyasingh PD, Frisch D (2018) Evolutionary aspects of resurrection ecology: Progress, scope, and applications—An overview. Evolutionary Applications, 11, 3–10. https://doi.org/10.1111/eva.12563