Introgression/Admixture?

We wanted to look at whether there has been any introgression/admixture from the D Barb/Oak/Mud species into our pulex ponds. From the pooled 2017 data I identified SNPs that were fixed alt in D Barb and D Oak. I then looked at the distribution of these SNPs in the Dorset ponds.

First looking at D8 across the four time points. In each case the graphs are the following from left to right. 1) Distribution of prop alt across a random set of SNPs. 2) Distribution of prop alt across the SNPs fixed alt in D Barb and D Oak. 3) The same as 2, but only looking at prop alt above 0.05.

D8 2012 (actual pool)

D8 2016 (artificial pool)

D8 2017 April (artificial pool)

D8 2017 May (actual pool)

So far, the distribution of prop alt for SNPs fixed in D Barb and D Oak does not look that different from the distribution of prop alt for all SNPs.

Now looking at some of the other Dorset ponds. DBunk and D10 look similar to D8.

DBunk 2017 April (artificial pool)

 DBunk 2017 May (actual pool)

D10 2012 (actual pool)

D Oily looks different. It shows a signature that looks more like introgression to me.

DOil 2017 May (actual pool)

And DMud looks like we would expect given the mix of species.

 DMud 2017 May (actual pool)

I was struck by the fact that most of the Dorset ponds show similar distributions of prop alt for all SNPs and those fixed as alt in D Oak and D Barb. I wondered if the SNPs that look like D Oak and D Barb in the Dorset ponds were always the same or not. Maybe they are just shared polymorphism between the two species or something? So, I tried something a bit different. I looked at SNPs that were fixed alt in D Oak and D Barb, but had a prop alt of less than 0.1 in W6, a pond that I thought had D. pulex like our Dorset ponds, but far enough away from D Oak and D Barb and the Dorset ponds that it should not have the same kind of introgression. Here are the distributions of prop alt for this set of SNPs. But only looking at ponds/samples where I have true pools.

D8 2012

D8 2017 May

DBunk 2017 May

D10 2012

D Oily 2017 May

D Mud 2017 May

From this, I think that it looks like the only pond (for which we have pooled data) in which there is current/recent introgression, is D Oily. D Oily is downstream (intermittently) of DBunk, but also down hill and close by to D Mud. D Oily is also more ephemeral than D8 or D Bunk.

I wanted to look at this a bit more. Here are some graphs of prop alt for these SNPs plotting D Mud versus D Oil, DBunk, and D8, and then each of those three ponds against each other.

D8 2017 May versus D Mud 2017 May

DBunk 2017 May versus D Mud 2017 May

Oil 2017 May versus D Mud 2017 May

Notice how the cloud of points around propaltDMud 0.75 is pulled out to the left when graphing against D Oily, but not D8 or D Bunk.

D8 2017 May versus D Bunk 2017 May

 D8 2017 May versus D Oily 2017 May

 D Bunk 2017 May versus D Oily 2017 May

Notice how when D8 is graphed versus D Bunk the points all cluster tightly around 0. But when graphing D8 or D Bunk against D Oily the points are pulled upwards a bit for propaltDOil.