Referee 2's review:

As it is generally the rule with this research group, this is a piece of very original and significant research. Degenerate primers based on conserved sequences are an important tool for (RT-) PCR detection of RNA viruses as well as of other micro-organisms, and the performance of the primers depends on the conservation of the targeted genomic region. To my knowledge this is the first analysis of how conserved regions in genome "decay" with new information available. By exploring this process for the Potyvirus genus of plant viruses, the authors describe a process of general biological and technical interest. Hence, this paper is of a general relevance and fits well with the scope of PLoS ONE.

However, before acceptance, the authors should pay attention to the rather careless presentation of the paper, particularly in the results section:

1) References should be indicated by numbers. They are not always indicated in this way, but in the reference section they are ordered according to appearance in the text. As a result, it is hardly possible to identify them.
2) Fig. 1 could be presented as a text figure or, better, the gene where a conserved site is, could be indicated in Table 1.
3) The section "Sequence and species" could be reduced. Moreover, I do not agree that the cumulative data in Figs. 1 and 2 show asymptotic trends. For partial sequences (Fig. 1) a decrease in the number reported occurs only in 2005, and a later increase cannot be discarded. This is even more so for complete sequences, particularly taking into account the advances in sequencing and the decrease in price. Anyhow, since this is not a central issue in the paper, there is no need to stress these possible trends.
4) Figs 1 and 2 could better be merged into a single one.
5) Fig. 3 is not central, and could be presented as supplementary material.
6) Table 3 is unnecessary and its two columns could be presented as additional rows in Table 2. Moreover, if the standard deviation is presented, the minimal and maximal values are no longer needed.
7) Figs. 4 and 5 could also be merged.
8) The evolution of the N score values with time could be analysed by adjusting the data to a model by regression analyses, and compared by means of the regression parameters. This would considerably improve the descriptive comparisons presented in page 5.
9) The Discussion is occasionally repetitive of the results (e.g., page 7, third paragraph)

**********
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.