We discussed this paper on Feb. 19th, led by Doris Bachtrog. After the enthusiastic discussion, a few questions were raised. So I emailed the lead authors (Steven Van Belleghem and Markus Moest).
Q1, why don’t you keep using Tajima’s D as opposed to CLR? A hard sweep and an adaptive introgression would have more similar footprints for the latter than the former.
Steven Van Belleghem responded:
“In short, much of the signal we found in Tajima’s D was very noisy. While in theory a nice volcano pattern may occur, we think that (1) recurrent introgression events, (2) multiple loci under selection and (3) the fact that we sample one ‘natural simulation event’ (compared to an averaged pattern of 100 Slim simulations) generates noisy signals that are hard to differentiate as resulting from introgression or de novo sweeps. I believe also the amount of divergence between the donor and recipient populations is in a gray zone to really differentiate these patterns. Look for instance at this pattern in H. m. plesseni from Ecuador:”
Q2, we wondered what is the main driving force for the spatial heterogeneity in the strength of selection. Is local mimetic species richness positively associated with the selection strength on the color genes?
Steven Van Belleghem answered:
“That’s a hard question and maybe your hypothesis is correct. However, the strength of selection is calculated from the SF results and we also show that the strength of this signal rapidly reduces over time. The selection strength we present is thus somewhat a composite of selection strength and age of the sweep. If selection strength turns out to be all the same among geographic populations, maybe we are just looking at the age of the adaptations in those populations… It thus subject to some discussion…”
Markus Hartmann Moest added as well:
“as estimated from SF2 is reflecting strength of selection and time since sweep has completed. The calculation also depends on a more or less accurate estimate of local recombination rate and effective population size. This makes the interpretation a bit tricky but our strongest sweeps reflect s values that indicate strong selection even if we use very conservative assumptions. Variation of s across space partly reflects differences in timing of the sweeps but most likely also variation in selection regimes. Some sort of variation of selection is most likely required to explain the variation of colour patterns across Central and South America and can explain why we see such recent signatures of selection. Fluctuations of predator density or of the availability of edible/more tasteful prey for the predators may result in periods of relaxed selection allowing colour patterns to drift and subsequent increases of selection pressure will fix the most common pattern at that time point. Quasi-Batesian mimicry (i.e. variation in the toxicity of the members in the mimicry ring) or other selection pressures on the colour patterns (e.g. temperature regulation, visibility in different light environments) may likewise underly this variation and dynamic. All exciting questions that are still unanswered. Disentangling timing of sweeps and selection strength is also one of our aims for the future. You’re explanation is also very interesting – I would argue that in general the number of individuals is more important than species richness but if we, for example, invoke some extent of Quasi-Batesian mimicry or multiple selection pressures acting on the colour patterns (also unrelated to mimicry) increased species richness should make these systems more dynamic I think – food for thoughts…Thanks!”
Q3, we wanted to parse out this process more and wonder where is the evidence that introgression supplied color genetic variation for divergent selection to work on? Or where is the evidence of adaptive introgression facilitating speciation/divergence?
Steven Van Belleghem's answer:
“I think the evidence for this is still mostly indirect. We have strong arguments to show that introgression has allowed mimicry to happen easily (Dasmahapatra et al. 2012; Wallbank et al. 2016; Van Belleghem et al. 2017; Edelman et al. 2019), and when tis happens, divergence from the ancestral population, at least, is an obvious consequence. Linking this to speciation is a hard one. H. heurippa has been an important example here (which seems to be mostly a H. timareta which has the optix locus introgressed) and has evolved reproductive barriers as well (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2607310/).”
Markus Hartmann Moest added:
"Agree with what Steven said. I think the evidence is still circumstantial (but I am sure many Helcionius researchers would disagree…). In H. heurippa, the colour pattern causes some reproductive isolation but the fate of this population is still unclear. The fact that pheromones also play an important role for mate choice - colour patterns are important for the attraction from the distance but pheromones then kind of take over when it comes to mating – might complicate this (at least in todays populations). Most of these cases of introgression btw. melpomene and timareta are not very old and speciation may need more time or opportunity (e.g. additional barriers/factors. I could well imagine that if a population of a species that has white colour patterns and preferences for white introgresses red colour patterns AND the preference for red, this population has higher chances to eventually evolve into a distinct species – but I also think that colour patterns are only one (albeit a very apparent and attractive) aspect of speciation in Heliconius and pheromones, habitat use, host plants etc… are also important."
Reference:
Moest, M., Van Belleghem, S., James, J. E., et al. 2020. Selective sweeps on novel and introgressed variation shape mimicry loci in a butterfly adaptive radiation. Plos Biol. https://doi.org/10.1371/journal.pbio.3000597
Comments