Reproductive compensation and embryo screening drive the evolution of polyembryony
Abstract: Simple polyembryony -- where one gametophyte produces multiple embryos with different sires but the same maternal haplotype -- is common in conifers, ferns, horsetails and other vascular plants. By providing a backup for inviable embryos, polyembryony may evolve as a mechanism of reproductive compensation. Alternatively, polyembryony may provide an opportunity for embryo screening and the preferential exclusion of low fitness embryos from the seed, perhaps acting as a mechanism of Self-Incompatibility (SI). To date, the evolution of polyembryony has not been modeled, and these verbal hypotheses have not been evaluated. We develop an infinite-site, forward population genetics model to test how these factors can favor the evolution of polyembryony, and how these underlying benefits of polyembryony shape the genetic load under a range of biological parameters. We find that reproductive compensation strongly favors the evolution of polyembryony, and embryo screening less strongly favors polyembryony. When embryo screening favors the evolution of polyembryony it increases embryo competitiveness, but does not act as an SI mechanism, as it does not trade low-fitness selfed offspring for high fitness outcrossed offspring. Remarkably we find nearly identical results in cases in which mutations impact either embryo or post-embryonic fitness (no pleiotropy), and in cases in which mutations have identical fitness effects embryo or post-embryonic fitness (extreme pleiotropy). In sum, the consequences of polyembryony depends on its function -- decreasing the mean embryonic fitness when acting as a mechanism of embryo compensation and increasing mean embryonic fitness when acting as a mechanism of screening.
Summary (3 min read)
- To better understand how and when these factors favor the evolution of polyembryony, the authors vary the distribution of dominance and fitness effects and the probability of selfing.
- When polyembryony does evolve, the authors ask how its evolution shapes the genetic load and its architecture.
- The life cycle begins with N = 1000 diploid seeds, each of which has one or two embryos, depending on whether mothers are mono-or polyembryonic.
- Following embryo selection, surviving seed parents for the next generation are chosen with replacement with a probability reflecting their post-embryonic fitness.
Parameters and model details
- Genome structure and mutation rate: Every generation, each haploid genome expects a Poisson distributed number (mean U ) of de novo deleterious mutations to arise, each at any one of an infinite number of unlinked sites (i.e. an infinite sites model).
- The authors investigate cases with U = 0.5 mutations per haploid chromosome per generation.
- The authors focus on the case in which half of de novo deleterious mutations impact embryonic fitness and the other half impact post-embryonic fitness.
- So the authors do not investigate this pleiotropic model here.
- Both embryos of polyembryonic mothers are fertilized independently, and pollen parents of non-selfed seed are sampled with from the population with replacement in proportion to each genotypes post-embryonic fitness.
The distribution of fitness and dominance effects of new mutations:
- To investigate the impact of mutational architecture on the evolution of polyembryony, the authors compare models with a different value of fitness (s), and dominance (h) effects of new mutations.
- Thus, mutation effects span the range from quite deleterious to lethal, but will not reach fixation by random genetic drift.
- In all simulations, the authors assumed that the distribution of fitness and dominance effects did not differ for mutations impacting the embryo and adult.
- With a probability equal to the p self (which the authors systematically varied from zero to one in increments of 0.2) the seed parent was also chosen to be the pollen parent, also known as Selfing.
- The authors note that this random mating does not preclude selfing.
- For all parameter combinations, the authors forward simulated ten replicates process for 2000 generations, ensuring that populations achieved mutationselection-drift balance by visually examining the variability in the number of deleterious mutations over time and among replicates .
- For most parameter values, equilibrium was reached within this time frame .
- For recessive mutations in predominantly outcrossing populations (with selfing rates of 0, 0.2, or 0.4) this was not enough time to reach equilibrium.
- For these slowly equilibrating cases, the authors increased the burn-in period until 3000 generations, at which point equilibrium was largely achieved.
- Finally, with complete selfing and a non-recessive load with s = 0.1, the number of deleterious mutations seems somewhat unstable .
Invasion of polyembryony:
- For each burn in replicate, the authors ran many introductions of a dominant acting polyembryony allele, introduced at a frequency of 1/2N , and kept track of the fate of this allele (loss or fixation) for each introduction.
- Due to computational considerations, the authors varied the number of introductions from 500 to 1000 for each model of polyembryony for each burn-in replicate.
- That is, when polyembryony was strongly favored, a given simulation took longer to complete (because fixation from 1/2N takes more time than loss from 1/2N ).
- The R (R Core Team 2020) code for these forward simulations is available on github https:.
- In simple polyembryony with two embryos, the seed has three possible combinations of viable and inviable embryos.
Models of polyembryony
- The authors aim to dissect the contribution of reproductive compensation and embryo competition to the evolution of polyembryony.
- Thus, this model includes both potential benefits of polyembryony.
- Each ovule in the seed of a polyembryonic genotype survives independently with a probability determined by its embryonic fitness.
- As such, polyembryony provides the benefit of increasing the probability that a seed survives, but does not provide the added benefit of embryo competition.
- The authors discuss the results from their burn-in simulations, as they set the scene for the evolution of polyembryony.
- Genomes saved at the end of the burn-in are available for download here.
- Intriguingly, with an intermediate selfing rate of 0.4, the population appears to reach an equilibrium, relatively modest number of recessive mutations, until this rapidly and dramatically increases, presumably reflecting a transition from effective purging to interference among deleterious mutations (Lande et al.
- The authors observe a strong positive correlation between embryo and post-embryo fitness for recessive gene action and intermediate selfing rates, but no relationship otherwise (Fig. 3D , and Fig. S2D ).
Invasion of polyembryony
- The authors compare the fixation probability of a new mutant that confers polyembryony, across all models described above.
- The dashed pink line displays the expectation under neutrality.
- In cases with additive gene action, the fixation probability of a polyembryony allele decreases with the selfing rate, again reflecting the lack of within-seed variance in fitness.
- The benefit of embryo competition also favors the evolution of polyembryony.
- Fixation probabilities are approximately five-to ten-fold lower for this model than for the reproductive compensation model.
Evolutionary consequences of polyembryony
- The authors compare how different models of the evolution of polyembryony shape key features of a population, including the proportion of surviving seeds, the realized selfing rate and the architecture of genetic load.
- The benefit of compensation (in both the compensation and all benefits model) resulted in a strong increase in seed survival.
- Higher fitness embryos, the benefit of competition alone subtly increased seed survival for all models of dominance investigated so long as the selfing rate was not too large (Fig. 5 ).
- Figure 6 shows that the allele frequency spectrum is comparable in the no benefit and competition model, arguing against the idea that competition favored selfsacrifice in the form of an excess of rare recessive lethals.
- By contrast, there is a slight increase in the count of deleterious recessive mutations across all frequency classes in the compensation and all benefits models, reflecting the relaxation of embryo selection in these cases.
- The authors present four models to test the plausibility of the compensation and competition theories for the evolution of polyembryony.
- By contrast, the benefit of embryo competition more weakly favored the evolution of polyembryony, resulting in between a zero-fold increase with high selfing rates, and a twenty-fold increase, with intermediate to low selfing rates and a recessive load, relative to neutral expectations.
- As such, embryo competition does not offer a more refined view into postembryo fitness than is automatically accounted for by "hard selection" on seed viability imposed in their model.
- That is, the authors must consider biological processes outside of their model as they interpret their model results.
- Competition, compensation and conflict in a pine nutshell: Gymnosperm seed with a maternal haploid megagametophyte, multiple geneti-cally distinct embryos, genetically identical embryos, and strong inbreeding depression is a stage of evolutionary drama that deserves more attention, and the authors hope that the provided model will be used to broaden the investigations on the evolutionary dynamics outside the angiosperm sphere.
Did you find this useful? Give us your feedback
Related Papers (5)
Frequently Asked Questions (1)
Q1. What are the contributions mentioned in the paper "Reproductive compensation and embryo competition drive the evolution of polyembryony" ?
4. 0 International license available under a ( which was not certified by peer review ) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.