I thank the reviewer for having checked my latest changes and for having
spotted several new problems that I have now corrected. My changes are
again marked in blue.

> 1.      As shown in Fig.3, the spectra of ions and neutrals diverge at
> large wavenumbers, which indicates the coupling is not perfect. The author
> should more clearly explain whether the single-fluid approximation used at
> these wavenumbers is valid and whether this would affect the conclusions.

This is now explained in the last paragraph of section 3.1.2, where I
clarify why the strong coupling approximation fails below the viscous
scale, and why this does not affect the magnetic field and therefore
also not the basic conclusions of this paper.

> 2.      After Eq. (12), the AD wavenumber is defined, "where
> the turbulent and AD timescale are comparable". Here it should be
> further clarified that the "turbulent timescale" here is not eddy
> turnover time as u_rms0 is used, and what is the relation of the "AD
> wavenumber" to the dissipation scale of magnetic fluctuations due to AD.

I have now added such a comment just a the end of the paragraph after
Eq.(12). I have then returned to this question on page 5 at the end of
the paragraph after eq.(21), where restate more clearly that the magnetic
dissipation wavenumber is in fact independent of AD and that this
confirms the results of Brandenburg & Subramanian (2000).

> 3.      In Fig. 2, different colors represent different values of
> tau_AD^prime, but in the caption of Fig. 3, it is written that different
> line styles are used to represent different values of tau_AD^prime. I
> feel this is very confusing to the reader. Please examine it.

The caption to Fig.3 was indeed wrong and has now been corrected.

> 4.      Please correct the typos through the paper, e.g., after
> Eq. (12), n_i \approx .... n_i^1/2, the second n_i should be n_n, and in
> the first sentence of Section 4, "import" should be "important".

I hope I have now fixed this and other obvious bugs.