COMMENTS TO THE AUTHOR:

Summary

In this manuscript, the authors provide a comprehensive overview of the practical use and highlight possible negative effects of inert ingredients in crop protection products such as pesticide formulations and agricultural spray adjuvants. These substances are declared as inert by their manufacturers, but there is evidence that they may be harmful to insect pollinators. Currently, there are few studies on this topic, and most studies have been conducted on honey bees. Thus, there is a clear need to focus research on these substances as well and relate them to potential hazards. In addition, legislators need to adapt regulatory procedures for these products to be marketed with appropriate testing.

In this systematic review, all relevant literature on the topic has been considered and its context has also been adequately discussed. The synthesis will reach a wide audience such as researchers, ecotoxicologists, non-governmental organizations, regulators and industry. However, some details were presented incorrectly or inaccurately - see my comments below. Since the authors submitted the article, German legislation has changed considerably due to recent findings from a laboratory study on spray adjuvants. I therefore advocate that this should also be implemented here due to its high importance. Overall, this review is of high quality and importance and should therefore find its place in this journal.

That being said, I suggest a major revision to the editor. Please find my detailed comments below.

GENERAL
The manuscript currently exceeds the 10-page limit. However, the text is not yet formatted in the final typesetting and layout of the journal. Without providing a more precise quantitative measure such as word count by the journal, I find it difficult to judge whether or not the authors are within the limits. In my opinion, the text is comprehensive overall and contains important information throughout that I would not cut. However, if it is necessary to limit the length of the article, the authors need to identify appropriate passages of text that can be removed without leaving the reader without a common thread.
You are using the term honey bee inconsistently. Please make sure to either use honey bee or honeybee throughout your text.
INTRODUCTION
L40: I suggest emphasizing that these adjuvants are specifically "spray adjuvants" - this is also the term commonly used in agricultural practice. Furthermore, a more detailed description would be appropriate here, such as "spray adjuvants are separate products added to a pesticide spray solution by the operator." This will be supplied later, but I think it should be mentioned at this point.
L52: Over 360 spray adjuvants are registered in Germany as well (https://www.bvl.bund.de/EN/Tasks/04_Plant_protection_products/01_ppp_tasks/05_ppp_Adjuvants/ppp_Adjuvants_node.html)
L62: German legislators recently changed the registration for spray adjuvants that may be added to insecticide formulations. Since October 26, 2021, it says briefly:
For the registration of adjuvants intended for mixing with insecticides according to the application for authorization, directions for use or product label, it is necessary that the applicant submits additional studies with honey bees. It cannot be excluded that the adjuvant may influence the effect of the insecticidal mixing partners, for example by surface-active components. The purpose of these bee studies is to investigate the efficacy enhancing potential of the adjuvant when mixed with insecticides. Excluded from this are adjuvants applied for seed dressing.
Testing should be done as follows:
The adjuvant will be tested alone and with three insecticides of different classes and the respective insecticide-adjuvant combination.
(see more details https://www.bvl.bund.de/EN/Tasks/04_Plant_protection_products/03_Applicants/12_ApprovalAdjuvants/ppp_ApprovalProcedure_adjuvants_node.html and especially the PDF here https://www.bvl.bund.de/SharedDocs/Downloads/04_Pflanzenschutzmittel/Zusatzstoffe_Studienanforderungen-Honigbienen.html?nn=11031052 which is currently only available in German, but can be translated with online tools).
Germany is thus the first European country to rely on scientific findings resulting not least from a recent (own) study that investigated the effects of different spray adjuvants in combination with different classes of insecticides (https://link.springer.com/article/10.1007/s41348-021-00541-z). Since this study is of great importance, it should be included in this review at the appropriate point as well as the fact that German legislation has changed its rules on adjuvants.
L77-78: See comment above.
L105: You use the units ppm and ppb throughout the manuscript. Since these are not SI units, I suggest replacing them with mg/kg or mg/L, which is more scientifically and ecotoxically appropriate. It should also be noted that this particular example may be conflicting. Pyrethroids per se have a repellent effect on honey bee flower visitation (see http://www.bulletinofinsectology.org/pdfarticles/vol56-2003-131-134thompson.pdf ).
L129 ff: The ball was started rolling when more than 12,000 honey bee colonies were affected by the abrasion of seed treatment compounds containing the neonicotinoid clothianidin in the Rhine Valley in Germany 2008 which contaminated floral resources bees were foraging on.
(see https://www.cabi.org/isc/FullTextPDF/2011/20113401182.pdf ).
After this incident, research began to look at these types of insecticides and came to three main conclusions: 1) there are major differences in susceptibility between honey bees and other (mostly solitary) bees, 2) there are differences in susceptibility when comparing effects on individual honey bees with those on colonies, and 3) there are differences in susceptibility when comparing laboratory versus field studies.
I don't think researchers would have put so much effort into studying neonicotinoids and honey bees without this incident. It was assumed that these substances were safe because seed coatings don't usually come into contact with bee pollinators, dust fall was not on the radar, and only honey bees were used as model organisms. Therefore, I would not say that regulation failed. It simply lacked that input at the state of knowledge at the time, which was then provided by science in the aftermath.
Recent evidence shows that when applied according to manufacturers' recommendations, even harmful pesticides can be "safe" for the model organisms with which they were tested (https://doi.org/10.1016/j.scitotenv.2021.146084 ).
I think it should be made clear that honey bees may not be the appropriate proxy to measure broader effects on pollinators that are not cared for by a beekeeper. This is because there has been a documented increase in the number of bee colonies worldwide since 2008. As you correctly mention, real life exposure scenarios are key to judge the hazard of these adjuvants and other model organisms are needed to dive into the full detail of risks and hazards.

METHODS
The methods are adequately described, appropriate, and additional details and raw data are provided in the supplemental information. The result of only 16 studies does not lend itself to an adequate meta-analysis, which could have revealed some more details.
RESULTS AND DISCUSSION
Even if it means investing additional time, I believe that the inclusion of the Wernecke et al. 2021 study mentioned earlier (https://link.springer.com/article/10.1007/s41348-021-00541-z) would benefit the outcome and discussion.
L235-236: If possible, also highlight these differences in the graph using colors.
L258: What does American surfactand adjuvants mean? Are they only available on the U.S. market or are they manufactured by a U.S. company? Please make that more clear.
L260: Just a side note: these compounds appear to be lipophilic and therefore accumulate in matrices such as beeswax, where they are persistent over time. This could later lead to undesirable co-exposures with active ingredients (miticides, pesticides) and not least to contamination of bee products marketed to consumers.
L282: I fully agree! Residue data are so very important and nation wide, ultimately pan-European residue monitorings are urgently needed.
L299: I have lost you here; how does 1% relate to 10,000 ppm (mg/kg)? Where is the reference to this percentage? To my knowledge, 1% refers to the final spray volume used in the tank of the field sprayer. As this tank is limited in size, it comes usually by a range of 300 – 1000 L volume. That means 1% would relate to the volume in the tank (i.e. 3-10 L of adjuvants). I do not see how you calculate the concentration of 10,000 ppm out of this.
L307: I suggest toning down this statement to “surfactants may have strong” as not all of them do.
L342-343: I cannot fully agree with this statement. While the scenario is more representative than the droplet on the thorax, the conditions are not representative of a spray in the field. This is because the subject and the way it is held (artificial conditions, laboratory conditions, age structure, etc.) alone can influence or bias the outcome of the study. Not to mention the weather and other environmental factors. I propose to rephrase the sentence.
L351: You state two LD50s here without reference: A dose always stands for weight/volume of substance per weight/volume of the test subject or similar (whole individual, e.g. bee). Please provide the correct doses here, i.e. 357 µg/bee.
L355: What is the range required to detect lethal effects and is it field relevant? The reference Donovan and Elliot is quite weak and contains many flaws. For example, they incorrectly state LD50’s as the endpoints, but state LC50’s because they only provide the concentration of the spray solution, not the weight or volume of the adjuvants received per individual bee. It is also not clear how the LC50 was calculated and how many concentrations were tested to achieve this endpoint. Therefore, confidence intervals ar missing for the correct interpretation. In addition, many details are missing to adequately reproduce the study. However, the concentrations of the adjuvants given as a percentage of the spray solution seem reasonable (most likely a recommendation from the manufacturer), which is why I asked the initial question: Why do you think a higher excipient concentration should have an impact when they have already tested the highest practically relevant concentration?
A source that is more reliable in my opinion, is the paper by Wernecke et al., which provides similar data with a more robust setup. They confirm, that spray adjuvants alone have no effect (mortality) on caged honey bees. This tends to confirm the conclusion that the tested substances alone are not "acutely" toxic, since only 72 hours post-exposure were evaluated.
L366: 0.1% of what? If you state a concentration with absolute measure of 1,000 ppm or 1,000 mg/kg, the relative measure does only make sense with a reference. A reference could be 0.1% of the total spray volume of 400 L or similar, so you must be precise. With a spray volume of 500 L, 0.1% may not be 1,000 mg/kg anymore. I suggest deleting this measure as it is confusing.
L370: Again, you state a concentration here and I find it hard to relate to the correct exposure of the bees. Usually, a dose should be provided to enable a clear interpretation. This can be considered a flaw even in the Fine et al. 2017 study as they state “This concentration was selected following an oral communication with an industry representative suggesting a thousand fold dilution from the tank mix concentration of 1% as a field realistic dose of OSS.”… a concentration is not a dose! A concentration is the amount of a.i. present in the (feeding or spray) solution. The dose is what the individual actually intakes and is more or less independent of the concentration. Thus, one can take in a lot from a low concentration or a little from a high concentration and receive the same dose. In Chen et al. 2018, the authors also do not specify a dose, which they could have easily done by weighing the feeders before and after each day and relating consumption to the number of bees per cage. These simple ecotoxicological standards seem to be missing from many studies looking at pesticide exposure.
L371: Since we don't know the actual dose a bee was exposed to, I find it difficult to draw firm conclusions. Especially since the Mullin group has done most of the recent research on this subject. We simply cannot judge whether or not the dose that caused mortality was realistic field exposure. Wernecke et al. at least could not confirm any of these results, especially for the OSS adjuvants. This is just my point of view, so no action is required on the part of the authors.
L384 ff: dose: 20 µg adjuvant per bee?/xx? Please add correct parameter.
L393: As this applies to other bees rather than honey bees, I agree. However, managed hives are heavily influenced by the beekeeper and beekeeping practices such as artificial queen rearing, and I do not believe that this factor plays a decisive role in colony health, as queens can be easily replaced. The authors should try to draw a clearer line between managed and wild bees where this is obvious throughout the manuscript.
L397: Yes, I fully agree, but the correct name is Smodis-Skerl.
L409, 458, 461: dose or concentration? Please check for yourself.
L440: Please replace the bracket with a comma.
L489: 17 including Wernecke et al.
L492, 518: I fully agree!

This manuscript has been revised in many parts and all my remarks and objections have been thoroughly and adequately considered. The authors' explanations of my comments have been made sufficiently clear. I therefore recommend that this manuscript be accepted with the current changes, but suggest that my minor comments below be considered.

L382: Amistar is mentioned here with a trademark symbol, but all other commercial formulations are not. I suggest to remove it.
L420: Change doses to concentrations
L424: Change dose to concentration
L547: Canadian regulation does require testing of adjuvants, too:
https://www.canada.ca/en/health-canada/services/consumer-product-safety/reports-publications/pesticides-pest-management/policies-guidelines/regulatory-directive/1993/registration-requirements-adjuvant-products-dir93-15.html