Is there a model for detection of few reads for low abundance targets?

Hello,

Is there a model (or a reason) why genome with low abundance only show a few portions of the reference genome?

I am doing a WGS on virus sequences; the mean coverage was 25, and for endogenous retroviruses I have such kind of coverage:

But for other, such as HHV7, I only have sparse reads with low coverage:

The low coverage I can understand: there is just a little bit of viral DNA, thus there is a low probability of sequencing it.

What I am less comfortable with is the fact the only a portion of the genome is covered.

Is this normal? Is there a model or some paper describing this phenomenon? Are there some factors that increase the sequencing of a certain portion of a low abundance genome rather than others?
Thanks