Congratulations, the second round review has recommended publication after attention to minor remaining points. The issue of removing remaining implications that strong flux attenuation was observed is essential and should bring the paper into accord with your revised part 1 companion paper. The addition of the second point regarding spore production at depth is optional, but I suggest would add an interesting nuance to your paper. Thank you for your excellent revisions of both papers. Please return your revised paper with your changes clearly marked.

Thank you for a thorough revision. Please make the following technical corrections.

Technical corrections (part 2)
Line Number Correction
30 suggest change from “annual carbon flux was low (34%)” to “annual carbon flux was lower (34%)”, because 34% is a reasonably large fraction of the total.
44, 45, 55, 882 replace Ocean by ocean
62 the phrase “light-mixing regime” is not very clear, suggest replacement by “light levels as modulated by insolation and surface layer mixing” or something similar.
94 add full stop, ., after “surface feature”
100 change “bloom is’ to “bloom was”
325 update Laurenceau et al., 2014 to Laurenceau-Cornec, 2015
329 correct feacel to faecal
337 citation of Legendre and Legendre, 1998 needs better accord with reference to Legendre – fix one or the other.
354 change “are already reported” to “are also reported”, as there is no timing aspect, only juxtaposition.
423 change “Small centrics” to “small centrics”
613 change “a strict dominance into” to “strict dominance of”
614 change “A resting spore formation at some depth (possibly implying a light control) would” to “Resting spore formation as some depth below the summer mixed layer (possibly implying a light control) could”
644 change pellet type to pellet types
764-5 consider adding citation to another special volume paper on BSi:POC ratios at A3, by changing: “is significantly higher than the BSi:POC ratio of 0.32±0.6 in the mixed layer of the same station during spring (Lasbleiz et al., 2014).” to “is significantly higher than the BSi:POC ratios of 0.3 to 0.7 in the mixed layer of the same station during spring (Lasbleiz et al., 2014; Trull et al., 2015).”
[see the data in Table 2 of
Trull, T. W., Davies, D. M., Dehairs, F., Cavagna, A. J., Lasbleiz, M., Laurenceau, E. C., d'Ovidio, F., Planchon, F., Leblanc, K., Quéguiner, B., and Blain, S.: Chemometric perspectives on plankton community responses to natural iron fertilization over and downstream of the Kerguelen Plateau in the Southern Ocean, Biogeosciences 12, 1029–1056, 10.5194/bg-12-1029-2015, 2015.

797 change to “downstream of the Crozet islands”

799 change to “as a major carbon sinker”

800 change to “as a strong silica sinker”

822 change to “resistant to grazing’

853 change to “primarily on suspended particles”

863 change to “the lower contribution of”

References
Correct formatting of Legendre book reference and update to Rembauville et al., 2015 and Laurenceau-Cornec et al., 20015

Table 1.
Update citation to Rembauville et al., 2015. (also in Table 7)
Consider shortening title, e.g. to “Table 1. Sediment bulk flux and composition results”, by moving the details to notes after the table.
Add an asterisk to the annual export row and a note stating that the "values assume no flux during the unsampled portion of the year".
To this row, add values for all numerical columns, to give the total days, mass flux, % opal, POC/PON and BSi:POC values.

Table 2, 3, 6, 7. consider shortening title by moving details to notes below the table
Table 3. Consider adding info on Tabular shapes in this row, such as “highly variable shapes, see text”

Tables 4 and 5. add notes at bottom explaining how the % column was calculated, e.g. is cells/total cells? or full cells/total full cells, or what?

Figure caption 1., line 1418-19, change to “Error bars are standard deviations”

Figure 5a, consider using log scale on y-axis to allow the low values to be seen.