search

rosemckeon / ploidy

How does disturbance on a landscape affect the establishment of new polyploid plant species?
3 stars 0 forks source link

Should there be more pollen than ova? #35

Closed rosemckeon closed 5 years ago

rosemckeon commented 5 years ago

At the moment N_gametes is used to create numbers of pollen and ova equally. Then pollen is lost and we decide how many ova are landed on.

@bradduthie, should plants create more pollen than ova to simulate the biology more realistically, or is this unimportant? I've restructured the data so that it's in a format that will handle uneven numbers of either gamete.

bradduthie commented 5 years ago

@rozeykex I think it's good if the model can take uneven numbers, but I don't see it as a problem to start by assuming equal numbers. What really matters is the probability that pollen and ova interact, and competition among pollen. For example, if there are 1000 pollen grains for each ova, but the probability that an ova is pollinated, and that any given male successfully sires the resulting offspring, does not change, then this is really just adding to computation time unnecessarily.

rosemckeon commented 5 years ago

@bradduthie OK, I think that's covered. Though I'm not sure the probability changes... There's a set probability which removes pollen, just like survival, to leave those that find ova. Then ova IDs from the same coordinates are assigned to pollen (with replacement so that some pollen can fight for the same ova). Only winners become seeds. Lots of ova lose out too because the pollen is limiting due to the initial loss. But all this is basically chance. The only thing that really changes is whether selfing or outcrossing are the result, depending on which plants are present in a cell.

Should there be variation in the amount of gametes made by different individuals? So that the probability of any given plant being more successful than another changes? Maybe bigger plants could make more gametes so have a better chance of siring the successful offspring? This would be another layer in selection for faster growth rate.

bradduthie commented 5 years ago

@rozeykex Sorry, I'm not sure if I explained that well (I agree that I think you're covered). What I mean to say is that, as I understand it, a lot of pollen is made, and the pollen that do not find ova are removed? This means that each pollen grain finds an ova with some probability (which will presumably be the same for all pollen grains from a given location?). My only thought was that if you can get to the exact same end result (random subset of pollen retained on a ova to compete -- or even just the one random winner) without having to make and then destroy large numbers of pollen in the code, it might give a bit more speed (this entirely depends, of course, on how much pollen is actually being made). I think it's good that selfing versus outcrossing is what really changes; this nicely isolates the mechanism of interest.

I guess it would make sense if bigger plants made more gametes, but I'm not sure if this is really necessary, or how it helps understand the focal question. Maybe leave an option open for this to be possible, then you can always give it a try to see how selection on gamete production changes things?

rosemckeon commented 5 years ago

@bradduthie I get you now! The making of gametes and choosing winners takes fractions of seconds now I've postponed giving them genomes until seed creation. I guess it probably could be done in an even more streamlined way, but I'm not sure it's worth it at the moment. Perhaps if we need to make the gamete numbers huge and find those parts slow down too.

Almost the entire generation time is taken up by giving genomes to seeds. I expect this will get a little faster when I add in genome duplication, as there will be some gametes that just get whole parent genomes copied instead of being sampled.

I've been starting to think about the genome duplication #12 and changing gamete number might be important there. Triploids and other polyploids with uneven copies are usually infertile in nature. I was wondering if this should be included by reducing gamete numbers?

So far I've been juggling this with revision, but my last exam is getting close now so I'm going to try and put this down until after the 19th (I really don't want to!). I think it might be coming to a point where we should chat again, and include Mario?