Referee 1's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

The paper provides an example of non-transitive aggressive interactions in Drosophila. The authors suggest that no single aggressive strategy exists that can dominate all others. They claim that virgin females do not choose males based on male ability to hold a territory, but rather that female choice depends upon the interaction between male and female genotypes. The authors suggest that these behaviours could provide a mechanism for the maintenance of genetic variation in levels of aggression, and they highlight the importance of studying interactions between individuals of specific genotypes, rather than studying population averages.

I'm not sure whether the title is accurate. To me this suggests that some genotypes are better at getting copulations and others are better at territory holding.

I think the introduction needs a more detailed account of the rock-paper-scissors scenario in side-blotched lizards. They do not explain 'active mechanisms' that may explain the lek paradox. They do not provide a rationale behind using Drosophila to study territorial behaviour, except to say that it may be another interesting organism. I found that in general the introduction did not give enough background as to why these questions are important.

The aims of the paper are poorly set out. They claim to test several hypotheses related to maintenance of genetic variation. These hypotheses should have been clearly discussed in the intro, perhaps with predicted outcomes for each hypothesis. The 1st time the rock-paper-scissors concept is introduced is in the aims.

I find confusion about whether the authors are claiming that the non-transitive aggressive interactions maintain genetic variation or whether the genotype-specific mate choice maintains variation. I don't see why the non-transitive effects are so important if females don't choose territorial males. If, from previous work, it is reasonable to assume that females do choose territorial males then the use of virgins in this study needs to be justified.

The authors cite previous work that showed that mated females prefer territorial males. Since this is key to the study of territoriality, I would find a discussion of the use of virgin vs mated females useful. Why did this study use virgins? Do females mate multiple times or just once and so is the behaviour of virgins or mated females most relevant for understanding selection on male aggression? Why might virgins and mated females show different mate choice patterns?

The main non-transitive result was between the selected Agg and Neut lines and Winters line 89. This rock-paper-scissors pattern was not found between the Winters lines. What does this mean and are these patterns likely to be found in natural populations?

Since this study concerns mating success of territorial/aggressive males it is important to distinguish between female preference and female coercion. Is there evidence for female control over mating? Did they measure courtship behaviour? I feel that claiming that mating success is determined by female choice (although plausible since females did not always choose territorial males) needs justification.

Did the study control for body size? This could determine the outcome of aggressive interactions and also female preference. Does body size differ between the lines or is it a potentially confounding factor?

Results
In figure 1 the graph shows continuous data but the x axis is the isogenic line, which is surely categorical.
I think there is too much discussion of results that do not hold under Bonferroni correction and I find this detracts from the significant findings.
The legends for the tables are not detailed enough to allow easy understanding of the tables. Could the tables be presented as graphs?

I think it would be useful to understand the mechanisms behind success over territories and genotype-specific mate choice. What are the behaviours underlying why super aggressive males sometimes lose territories? Why does only one of the female lines show preference for male lines?

Minor points:
Abstract line 3 - 'sexual display traits should be depleted, but a great of variation...' - word missing
Page 2 of introduction - 'Theoretical work suggest that even in homogenous environments, variation in IGEs may not be lost' - this needs a reference
Page 3 of introduction - 'First, we were curious to see whether territorial success in males is a transitive trait, where, as has been suggested' - something missing
Reference 8 - needs a journal cited