Interactive comment on “ Nutritive and photosynthetic ecology of subsurface chlorophyll maxima in the Canadian Arctic waters ” by

3) Comment: The authors state that SCM contribute largely to “total water column” production; the depth for which they have estimated production is only given for one station (and it is 71 m. The depth of the water column at the sampling stations is not indicated in the paper. Response: Actually, 71 m is not the depth of the water column for station 303 but the depth of the euphotic zone. Nevertheless, your comment made us realize that the bathymetry is missing in Fig.1. We added it to the figure. Introduction 4) Comment: Temperature is mentioned in the abstract, results and discussion but does not appear in the introduction or as a question for the work. Response: Indeed. We apologize for the oversight and added a presentation of temperature effects to the introduction. C3169 BGD 9, C3167–C3177, 2012

samples (on deck or in situ) as previously done by others in the Arctic (e.g. Simpson et al. in prep;Brugel et al. 2009;Fouilland et al. 2007;Smith et Harrison 1991;Harrison et al. 1982;Platt et al. 1982) and elsewhere.We recognize that our approach was not fully explained in the introduction.Firstly, it should be noted that the 24-hr incubations advocated by the referee are not a "gold standard".These 24-hr incubations are subject to artefacts (e.g.unrealistic photoinhibition of surface samples, nutrient depletion) and bottle effects.Secondly, sampling at 6-7 depths provides better vertical coverage but it still is a discrete approach that poorly resolves the structure of SCM layers.Thirdly, light-gradient incubations are the only means to separate the light-dependent and light-independent components of nitrogen uptake (see reply 11).Finally, the results of 24-hr incubations provide "biogeochemical" estimates of primary production at a given station, but these cannot be extrapolated beyond the day of sampling since incident irradiance in the Arctic fluctuates greatly at short time scales.
Using a dynamic approach yielding parameters that assess physiological state, acclimation and responses to irradiance and nutrient additions seemed to be the most promising means to understand the ecology SCM communities, assess their productivity and generate parameters for remote-sensing approaches and numerical models.With these parameters it is possible to "reconstruct" primary production by combining continuous calibrated fluorescence profiles (thus resolving the SCM without the aliasing caused by discrete sampling) with a continuous record of incident irradiance and light attenuation in the water column.
For the reasons exposed above we believe that a direct comparison of our productivity estimates with discrete 24-hr incubations is of limited value.We attempted to do it nevertheless, but realized that most of the papers mentioned above reported vertically-integrated rates only but not the vertical profiles.The only profiles shown by Harrison et al. 1982 for open waters of the Canadian Arctic indeed show a pronounced C2847 subsurface maximum of primary production and nutrient uptake (see their Fig. 2, stn.93).We also found the following statement in Tremblay et al. 2006 (LO: 51), who used 24-hr incubations at 7 depth for Baffin Bay: "Moderate RPINO3 , VNO3 and f-ratios in the nitracline were associated with a persistent layer of the centric diatom Chaetoceros socialis, which drove most of the new production after the early (surface) bloom".
3) Comment: Furthermore, SCM has been well studied in the ocean, and light, nutrients, and temperature are well recognized to be important for the SCM communities (Lalli and Parsons, 1993).I receive the impression that their finding is lacking novelty and remains local interest.
Response: Indeed we recognize that light, nutrients and temperature are important for the phytoplankton and were studied numerous times in the world ocean.However we disagree that SCM have been "well" studied from a global perspective.Lalli and Parsons 1993 did not consider the Arctic at all.Most of the available knowledge is derived from temperate or warm oceans and the paper of Martin et al. (2010) and the present one actually demonstrate that Arctic SCM are unique and do not function like tropical or subtropical ones.That's novelty!We also believe that investigating the ecology of a 2500 x 3000 km portion of the Arctic Ocean, the most understudied ocean on Earth and yet the most perturbed by climate change, is not a matter of "local" interest.We have slightly modified the conclusion (or discussion) to better place the results in a larger context.

4) Comment:
Introduction One of the objectives in this study is to establish contemporary parameters for use in ecosystem models and remote-sensing algorithms (P6448 L9-11).However, by only reading this introduction, I cannot understand the necessary for obtaining the parameters.Therefore it is difficult to understand why the authors needed to conduct experiments for photosynthesis or nitrogen uptake-irradiance curve (P6451 L7-P6452 L5).

Response:
The pressure (< 250 mmHg) is now provided in the text.

7) Comment: P6452 L6:
Recent study demonstrated that f-ratio has been overestimated by nitrification near surface (Yool et al., 2007).Influence of nitrification on nitrate uptake and f-ratio should be addressed.

Response:
We agree that nitrification may introduce a bias in the estimation of the f-ratio (as stated by Raimbault et al. 1999 andYool et al. 2007) and have added a caveat to the text.Since bacterial nitrification is inhibited by light (Horrigan et al. 1981), Martin et al. (2010) surmised that nitrification should play a lesser role for Arctic SCM located relatively high in the euphotic zone (which receives light 24-hr per day).These SCM also track the nitracline, where ambient nitrate concentrations are relatively high and must strongly dilute small inputs by nitrifyers.C2849 Results

8) Comment:
This comment has been merged with comment #2.9) Comment: Table 1: Incubations of surface waters were only conducted in late summer-fall.The authors demonstrate that plankton community properties did not only vary between at the surface and the SCM (Fig. 2) but also with the seasons (Fig. 5 and 6).This sampling bias might influence on interpretation of obtained data.

Response:
We agree that our representation of vertical phytoplankton structure is partial due to the lack data for surface waters in the spring.On the other hand, as mentioned P6467 L1-6, Palmer et al. ( 2011) observed a continuous and rapid acclimation (within 4 to 10 days) of the phytoplankton during the initiation of the growth season.While acknowledging the limitation that this lack of data can provide, we assume nevertheless that this observation may be appropriate for a large part of the production season.
10) Comment: Table B1,B2: It is difficult to understand the difference between Table B1 and B2.What is the difference of correlation coefficient for nitrate uptake between Table B1 and B2?
Response: Table B1 and B2 represent statistical analysis on two different data sets.
As mentioned in Materiel and Methods (P6449 L2-5) and in the result section (P6454 L13 and P6455 L20), NO3/NH4 experiment at SCM depth represent one data set (Table B1) and surface/SCM experiments another one (Table B2).