search

rosemckeon / honours-project

When rare gametes meet: an individual-based model of polyploid evolution in angiosperms.
https://github.com/rosemckeon/honours-project/blob/master/thesis/_thesis_2417024.pdf
0 stars 0 forks source link

Test the effects of increasing ploidy_rate > 0.01 #33

Closed rosemckeon closed 4 years ago

rosemckeon commented 4 years ago

In a quick test while debugging I found increasing the rate lead to chance of extinction. so, looks like this will give some interesting results.

rosemckeon commented 4 years ago

@bradduthie The data I've gathered for increasing the rate looks pretty interesting. Once the duplication rate reaches just under 0.2, it causes lots of extinctions. Both the polyploids and diploids quickly crash. Do you think it would be useful to run a big batch that focus in around that rate?

frequency~rate-diploids-vs-polyploids

rosemckeon commented 4 years ago

Increasing the polyploids in the system disrupts the seed counts due to mismatched pairings during fertilisation attempts.

Edit: that righthand facet actually has ploidy_rate all the way up to 0.5

life-stages

bradduthie commented 4 years ago

Huh, interesting @rosemckeon -- definitely something you can write about (sort of a reproductive interference between diploids and polyploids). Should I run another 25 from the same script, or did you want to tweak things first?

rosemckeon commented 4 years ago

@bradduthie I wondered if you thought I should just focus in and do more simulations between 0.15 and 0.2?

Also... I had a thought this morning. It could be difficult for the whole plant population to grow when the rate of duplication is too great because of the small starting population size. I start in generation 0 with 60 individuals. This might be influencing the extinctions. what do you think?

rosemckeon commented 4 years ago

@bradduthie I think I'm regretting running them with small starting numbers. It means that the population is quite susceptible to drift while it gets started. I didn't give that part enough thought. It was a bit arbitrary; just 20 of each life stage.

I think it might make more sense to run them with starting populations at full size, so 100000 seedlings and 75000 seeds... Or, to start just with a number of seeds, as if simulating a pioneer species in a new empty space. Is it worth re-running everything with this updated as well as the ploidy_rate range?

bradduthie commented 4 years ago

Okay @rosemckeon -- no worries. I think either option works, depending on what kind of a system you want to discuss and model (i.e., it's fine to make the assumption that you have either an established population, or an empty space that is newly colonised). You just need to make sure that you're interpreting the model accurately in light of the assumptions.

Also, the small population size isn't necessarily that bad. You've learned that a small (maybe initially establishing?) population has a higher extinction risk when the rate of duplication is high. If you crank up the starting population size and see the extinction go away, then this doesn't necessarily mean that your previous results were wrong or flawed -- just that you can put them in the context of relatively small versus large population sizes.

Happy to run more, in any case! Just modify the script, and I'll pull it and run it 24+ times.

rosemckeon commented 4 years ago

@bradduthie Thanks! I'm tweaking it now with some adjustments to get a range of starting population sizes. May as well test that out seeing as the results come through so quickly.

bradduthie commented 4 years ago

Yeah @rosemckeon -- feel free to simulate anything you think will help or be interesting!