Interpretation of negative values in AMOVA

[GrimPerez]

Please place an "x" in all the boxes that apply

• I have the most recent version of poppr and R
• I have found a bug
  • I have searched for solutions on this site and the poppr forum
  • I have a reproducible example
• I want to request a new feature

---

Please include a brief description of the problem with a code example:

# insert reprex here

With all due respect I ask the following.

I carried out AMOVA based on STRUCTURE clustering and got the following results for % variance

Among populations 13%
Among individuals within populations -11%
Within individuals 98%
Total 100%

•What does the negative value in 'Among individuals within populations' (-11%) say?
•If the variation among populations (STRUCTURE clusters) is 13 percent, then how to calculate the variation within populations? -11% + 98% =87% Or do I need to carry out some other calculation?

Sincerely,
Grim

---

[zkamvar]

This is answered in the poppr forum: https://groups.google.com/g/poppr/c/NSag-55d6bs/m/PasHn3C7AQAJ

The "Among individuals within populations" is the variation within populations, you don't need to do a separate calculation.

[GrimPerez]

Thank you very much for the reply.

Regarding the variation within populations, I saw some research articles report it like this
Article 1:
"Analysis of molecular variance (AMOVA) identified 4% variance among and 96% variance within subpopulations"

Article 2:
"The AMOVA on the landrace population vs. cultivar population showed a much greater variation within a population (67.36% + 17.08% = 84.44%) than among the populations (15.56%, p < 0.001)"

It seems like they add the 'Among individuals within populations' and 'Within individuals' %. Can you please provide your input as to why they report it like this?

Sincerely,
Grim

[zkamvar]

I would suggest for you to post to R-SIG-Genetics about this inquiry because there are people on that list with much more expertise in AMOVA than I (poppr only provides an interface for two different amova functions in the ade4 and pegas packages). I'm afraid I cannot interpret single quotes and table images from random articles without citations or context.