The manuscript does address relevant scientific questions within the scope of BG. It also presents partly novel data.

Are there substantial conclusions reached or not is questionable as the manuscript intends to show results from feeding experiments that where obviously to short or failed. N. labradorica is usually a deep infaunal (limited by the availability of oxygen) dwelling species that covers its test with a sedimentary cyst. In these experiments the sedimentary cyst was removed, which should have led to significant stress for the animals as thereby also a significant amount of ectoplasm is removed. Being placed in petri dishes with artificial seawater and after what is called a starvation period of 2-4 hours (why this range?), that from my experience is too short for foraminifera specimen to recover, either cultured M. sedimenti was added or not. The maximum experimental incubation time was 20 h. Of the 17 specimens selected only 2 specimens showed very a total of 3 supposed methanotrophic bacteria close to the reticulopodial network that rests in the final chamber of rotaliids. Obviously, none of the identified bacterial remains could be definitively related to the provided bacterium. Furthermore, obviously all specimens showed clay particles in the final chamber, that was not provided during the experiment. The fact that clay was observed in the last chamber of all specimens and only 2 specimens showed 3 bacteria that eventually from which two were not even provided during the experiment, indicate that clay and bacteria were eventually in place before the start of the experiment. Although I consider it logical that N. labradorica may facultatively nourish on bacteria, the experiments are no proof. Rotaliids extend and retreat there pseudopodial network into the final chamber and whatever attaches to these filaments will be found there. If we look at the colorful Fig. 2a, it becomes obvious that we would expect any nutritious material predominantly in the older test parts (see also Wollenburg et al, 2018 Fig. 8b). The final chamber is a place of activity like a garage for pseudopodia or for short storage, but if you don’t find traces of the provided bacteria in older chambers, there presence in the last chamber could just be accidental. Support for this suggestion comes from the very few observations of bacteria and the absence of M. sedimenti. So, I consider the experiment as too short or rather failed and would focus the manuscript on the great TEM figures and place some speculative assumptions on experimental pictures (by the way I like the pictures), or to redo the experiments for a significantly longer period of time and with labelled bacteria. This would also proof that the specimens even survived the treatment, a 24-h experiment on specimens that immediately died during the treatment would show no different TEM pictures, and as no pseudopodial activity was reported, the survival rate is unknown.

The scientific methods are sufficiently described and can be reproduced by scientists.

I don’t consider the experimental results sufficient to support the interpretations and conclusions, but I would like to see a revised version focusing on the TEM pictures especially from the TEM pictures with the paired clay and bacteria vacuoles. I don’t understand why the authors elaborate on geochemistry methods, when neither d13C or other geochemistry aspects have been applied to the investigated specimens. I would recommend to either get rid of this method chapter, except for the basic oceanographic feature or to put them in context to the foraminiferal data. A basic information on the sedimentary composition at the coring site is needed to put the clay particles in context. Futhermore, is there a preferred grain size spectrum in the food vacuoles? If yes, why? What does this tell us.

The description of experiments is sufficiently complete and precise to allow their reproduction by fellow scientists.

I think that the authors should leave their inner circle of referencing. Especially regarding carbon isotopes there is a misinterpretation of published work that essentially relies on the assumption that foraminiferal shells reflect the bottom or pore water signature they are dwelling in not their nutrition. And regarding Wollenburg et al (2007) the statement they place is wrong. This study confirmed that foraminiferal shells reflect the emanation of methane and its signature.

The authors usually give proper credit to related work and clearly indicate their own new/original contribution.

The title clearly reflects the contents of the paper, but should be addressed to N. labradorica only

The abstract has to be shortened to be concise.

The overall presentation is not well structured and should be streamlined.

The language is fluent.

The whole text contains a lot of repetitions and should be rewritten in a way that one aspect is addressed predominantly at only one place in the introduction and discussion. There are also a lot of typos and missing blancs.

The supplementary material could be improved by showing successful TEM pictures of all investigated specimens.

In this study, living individuals of the calcareous benthic foraminifera Nonionella labradorica were collected from active Arctic methane-emission site sediments and used for a one-day feeding experiment to investigate the uptake of the prokaryot methanotroph Methyloprofundus sedimenti. Transmission Electron microscopy pictures of the foraminiferal cytoplasm were analyzed to test the hypothesis that Nonionella labradorica is a deposit feeder and is able to ingest and feed on Methyloprofundus sedimenti. The presented TEM micrographs are of high quality, and show the deposit feeding character of Nonionella labradorica. But the feeding experiment itself demonstrates only week results.

The original scientific question of the manuscript fits into the field of biogeosciences, but the results presented do not fulfil this claim and the concept and data can only partly be described as novel. Therefore, no significant conclusions can be drawn with respect to the initial questions.

The scientific methods and assumptions are valid and clearly outlined, although I want to comment the following points:

• It is not clear for me why the authors put a lot of effort to the geochemical analyses and molecular genetics but did not show these data in detail and do not really link them to the experiment. Why they determined sulfate decline and the SMTZ if they collect their foraminifera from the 0-1 cm surface and feed them with methanotrophs from the water column? It would have been better to concentrate on the feeding experiment.
• If I understand correctly, the methanotroph food shows a natural labelling (line 63: methanotrophs produce the biomarker diplopterol, which has an extremely light δ13C signature (− 60 ‰)). It would have been of big advantage to use this kind of natural label or any other labelling to track a possible uptake by the foraminifera.
• The incubation time seems to have been planned too short, especially at these cold temperatures, because no uptake was observed.
• The authors mentioned a starvation phase before the experiment of 2-4h. Do they mean 2-4 days? Otherwise, 2-4 h are no starvation time, I would say. Foraminifera probably do not feed 24/7.
• The authors mention that for each experimental time point, 4 out of 5 foraminifera were examined. Was the selection for these 4 purely random?

The results are not sufficient to support the interpretations and conclusions, and I want to comment the following points:

• The TEM analyses showed that all investigated specimens (collected in the field and in the experiment) contain degradation vacuoles containing clay/inorganics, which is a strong indicator for sediment uptake and deposit feeding in Nonionella labradorica, indeed. But this does not have something to do with the presented experiment, because sediment uptake took part before the individuals were placed in the petri dishes and the experiment started.
• The observed bacteria inside vacuoles or at the aperture area of the foraminifera are still few and not convincing, there seems no significant correlation. Living foraminiferal tests are surrounded by a sticky cytoplasm cover. Fig 4 shows sediment particles around the aperture. This means that specimens were not cleaned properly before placed in the petri dishes (contradict to the text in line 189). This kind of sediment may also be very sticky, as well as pseudopodia remnants around the aperture. Added methanotrophs may stick here by accident after some time. This seems to be more an artefact because of the experimental set up and should not be interpreted as an “association of foraminifera and methanotrophs” as noted in line 354 of the discussion.

The description of the experiment is very clear and allows reproduction. The authors give proper credit to related work and clearly indicate their own new contribution. The number and quality of the references is appropriate. The language is fine, although there are some minor mistakes in the text. The title reflects the contents of the paper, and the abstract provides a good summary. The introduction is very informative, but in parts drive off from the topic and raises high expectations in the experiment, e.g. that a direct link can be drawn from methanotrophic bacteria to the extremely light 13C-signals in foraminiferal tests. Concerning the supplementary material, I miss the data of the geochemical measurements.

The overall presentation is well structured and clear described. The TEM micrographs are of high quality, and show the deposit feeding character of Nonionella labradorica. This should be published! But I recommend publishing the data in a more foraminiferal-focused journal and changing the focus of the manuscript.

The manuscript „ Deposit feeding of a foraminifera from an Arctic methane seep…” by Christiane Schmidt and co-authors describes a feeding experiment with Nonionellina l. from a seep site with cultured methanotrophs (Methyloprofundus s.).

The methods are described clearly and great care has been taken to ensure the viability of the foraminifers. Impressive photos of the foraminifera are presented.

The experimental set-up seems to me (as a non-expert for foraminifera) a bit weak: 

-    There were 5 specimens in each set up, but results for only 4 are reported

-    The incubation time and/or incubation temperature was too short or too low, as hardly any feeding (bacteria in vacuoles or near the aperture) was observed. Unfortunately, an extended incubation or slightly warmer temperatures, with an extension or repetition of the experiment is not possible....

-    The presence of storage granulas and of gram-negative cell walls in the observed bacteria is not specific for methanotrophs, only the ICMs are characteristic for methanotrophs

In the discussion, the relation of the study to porewater chemistry is a bit superficial and not necessary for the experiment. 

Also, the discussion on the SMTZ and anaerobic methane oxidation is miss-leading, as the foraminifera have been sampled from the sediment surface, and also Methyloprofundus is an aerobic methane oxidizer, presumable from the sediment surface. 

As so few bacteria have been found in or in front of the foraminifera the conclusion that they can feed on them is not justified.