This manuscript corresponds to the first "PCR-free" DNA metabarcoding experiment for analyzing bulk animal samples. I fully agree with the authors that PCR has many drawbacks during classical DNA metabarcoding experiments. Clearly, the shotgun approach presented here represents the future of DNA-based biodiversity assessment. I strongly support the publication of this manuscript that unambiguously demonstrates that the shotgun approach is reliable.

Major Compulsory Revisions

(i) I have no major compulsory revisions to propose.

Minor Essential Revisions

(ii) The journal Molecular Ecology recently published a special issue on environmental DNA. In the editorial of this issue (Taberlet et al. 2012a), in order to facilitate bibliographic surveys, we tried to propose a standardization of the terms used within the context of DNA-based species identification. This manuscript corresponds to "DNA metabarcoding" as many species are identified at once from a bulk sample. I strongly suggest to include the terms "DNA metabarcoding" either in the abstract or in the keywords (and not only in the running title). Accordingly, the terms "environmental barcoding" and "metagenetics" could be removed from the keyword list.

(iii) The limitations of the PCR approach in DNA metabarcoding studies has already been thoroughly discussed (Taberlet et al. 2012b). This paper also proposed the shotgun approach. It must be cited in the introduction.

Discretionary Revisions

(iv) Abstract line 30 and text lines 362-371: I am not sure that the terms "false positive" are appropriate here. Can this sequence be an artifact of the assembling process? If yes, the terms "false positive" can be used. If not, the terms "false positive" are not appropriate, as this Lepidoptera was probably in the initial sample, but not well preserved (for example, microlepidoptera are known to be very fragile, and can easily be missed by visual inspection of the bulk sample). This means that the shotgun approach can reveal some species that were not recorded by the morphological approach.

(v) lines 31 and 374: I suggest to replace "nucleotide number of sequence data" or "number of nucleotides" by "sequence coverage".

(vi) lines 38-40: I fully agree that further improvements in mitochondria enrichment are needed. However, in my opinion, there is also room for further improvements at the bioinformatic level. This can be mentioned both in the abstract and in the discussion.

(vii) line 70: I would avoid the terms "metagenetics" or "metagenomics" here (see comment (ii) above).

(viii) lines 89-90: The sentence "..., the performance of the new primers designed in silico is difficult to predict in the real world when the investigated fauna is largely unknown" is questionable. By using the ecoPCR program (Bellemain et al. 2010; Ficetola et al. 2010), we are quite confident that a thorough in silico analysis can provide a very good estimate of the performance of any new primer pair in the real world.

(ix) lines 108-114: Concerning the correspondence between the number of NGS reads and the taxonomic abundance, two other papers showing this correspondence can be cited. Fist Soininen et al. (2009) showed a good correspondence in stomach content estimated with the number of sequence reads and with a micro-histological approach. Second, the results of Kowalczyk et al. (2011) on bison winter diet clearly showed that the DNA metabarcoding data can be interpreted quantitatively: the proportion of tree in diet decreases with the intensity of the feeding with hay (see Fig. 1).

(x) line 456: How the individuals were homogenized? Such information is crucial to reproduce the experiments.

(xi) lines 495-500: The program SOAPdenovo was used for the assembly. According to Namiki et al. (2012) and to Yang et al. (2012) both MetaVelvet and VICUNA assemblers seems to work better than SOAPdenovo, especially for metagenomic/metabarcoding data. It would be interesting to try MetaVelvet and VICUNA on these two samples.

Cited references

Bellemain E, Carlsen T, Brochmann C, Coissac E, Taberlet P, Kauserud H (2010) ITS as an environmental DNA barcode for fungi: an in silico approach reveals potential PCR biases. BMC Microbiology, 10, 189.

Ficetola GF, Coissac E, Zundel S, Riaz T, Shehzad W, Bessière J, Taberlet P, Pompanon F (2010) An in silico approach for the evaluation of DNA barcodes. BMC Genomics, 11, 434.

Kowalczyk R, Taberlet P, Coissac E, Valentini A, Miquel C, Kamiński T, Wójcik JM (2011) Influence of management practices on large herbivore diet - case of European bison in Białowieża Primeval Forest (Poland). Forest Ecology and Management, 261, 821-828.

Namiki T, Hachiya T, Tanaka H, Sakakibara Y (2012) MetaVelvet: an extension of Velvet assembler to de novo metagenome assembly from short sequence reads. Nucleic Acids Research, 40, e155.

Soininen EM, Valentini A, Coissac E, Miquel C, Gielly L, Brochmann C, Brysting AK, Sønstebø JH, Ims RA, Yoccoz NG, Taberlet P (2009) Analysing diet of small herbivores: the efficiency of DNA barcoding coupled with high-throughput pyrosequencing for deciphering the composition of complex plant mixtures. Frontiers in Zoology, 6, 16.

Taberlet P, Coissac E, Hajibabaei M, Rieseberg LH (2012a) Environmental DNA. Molecular Ecology, 21, 1789-1793.

Taberlet P, Coissac E, Pompanon F, Brochmann C, Willerslev E (2012b) Towards next-generation biodiversity assessment using DNA metabarcoding. Molecular Ecology, 21, 2045–2050.

Yang X, Charlebois P, Gnerre S, Coole MG, Lennon NJ, Levin JZ, Qu J, Ryan EM, Zody MC, Henn MR (2012) De novo assembly of highly diverse viral populations. BMC Genomics, 13.

Level of interest: An article of outstanding merit and interest in its field

Quality of written English: Acceptable

Statistical review: No, the manuscript does not need to be seen by a statistician.

Declaration of competing interests: I declare that I have no competing interests