This is the second revision of the manuscript titled Microbial strong organic ligand production is tightly coupled to iron in hydrothermal plumes" by Hoffman and coauthors. I was rev1 for the first revision.
I am undecided about how to proceed. On one hand, the authors replied to all my questions and introduced changes in the text, but on the other hand, most of those changes were merely cosmetic. None of the (IMO) main shortcomings of the manuscript have been corrected. I think that the data base must be published, specially the chromatographic part, but the rejection to recognize the limitations of the technique and the refusal to comment or amend the oversimplification of data and processes makes the revision process unnecessarily long and frustrating. This attitude brings a clear problem on the long run, some misconceptions are perpetuated just because they have been published. I suggest another round of revision.
1 The first matter is the “organicity” of ligands. Some lines have been added mid-way the manuscript about the possible presence of inorganic phases and erroneously these are restricted to oxyhydroxides. Why inorganic phases are necessarily “weak” ligands? I insist, CLE-AdCSV does not provide any insight about the nature of the fraction that is not labile after equilibration with a competing commercial ligand. And this is not arguable. The manuscript should clearly state that the analytical technique cannot discriminate between organic, inorganic or mixed phases and modify the abstract to reflect that the contribution of inorganic and mixed phases to the ligand concentration is likely. Such phases are described in most of the bibliography. It must also be noted that calculations of the “ligand” concentration are based in a 1:1 Fe:ligand ratio model and that in the presence of mixed phases or inorganic phases this is just a parametrization and possibly does not respond to the actual speciation.
After my comment in the initial review, no biological process, apart of siderophore release, which is clearly insufficient, has been presented as a possible source of organic ligand concentrations of tens of nM. There is no bibliography showing such ligand concentrations, even in the absence of phytoplankton. I contacted one of the first authors involved in the electrochemical speciation in hydrothermal plumes and this researcher confirmed me that despite the label of organic ligands, no test was carried out to study their “organicity”. And the whole bibliographic body about the presence of persistent inorganic phases has been either ignored or just cited ignoring the part referred to stabilized inorganic phases. The study of hydrothermal vents is complex and requires to take into account that many precipitation processes are of very slow kinetics, especially at low temperatures and that metastable (as in non thermodynamically stable) phases may be present for very long periods (I recommend a reading of Stumm and Morgan’s book). Furthermore, colloidal phases are included in the operational definition of dissolved iron from filtration by 0.2 um. The authors argue throughout the manuscript in terms of equilibrium and this is not correct for colloids. (Fitzsimmons et al., 2017) (cited when denying the possible inorganic character of the ligands) described dFe in hydrothermal plumes as a mix of two components, one of them would be inorganic oxyhydroxides that would persist even when the plume is so dilute that most of iron is background ocean dFe. Many articles argue about the presence and persistence of nanopytites. There is irony behind the use of an article titled “Hydrothermal vents as a kinetically stable source of iron-sulphide-bearing nanoparticles to the ocean” (Yücel et al., 2011) to argument the complete precipitation of inorganic iron. Nanopyrites and the onion model are also discussed in (Hawkes et al., 2013) (cited in this manuscript) when referring to components of the “inert” fraction of hydrothermal dFe. Pyrites are known for their high insolubility; aged oxyhydroxides become more and more refractories even to acidification protocols (Raiswell et al., 2010).
This discussion pertains not only to ligand concentrations but also to the so-called L1. If iron inorganic phases are organized in more or less refractory clusters, it is highly probable that SA, especially when used at such a low alpha coefficient, cannot "peel" the colloid off during the overnight equilibration. Iron not available for ligand exchange would manifest as a complex with a very high K but limited to an upper threshold. This upper limit is caused by the limitation of the analytical technique known as the analytical window that forces K (or alpha if the authors prefer) into the analytical window. The possible presence of inorganic phases and their possible inert character should be added to the abstract and throughout the manuscript and the role of siderophores toned down.
2 The use of L1 as a proxy for siderophore concentration is also at least arguable and there is not consensus in the field as the text seems to point. Recent works by Slagter, Gerringa, Laglera and Sukekava have shown that the so-called L1 is present in Arctic waters where humics are overwhelmingly the majority of ligands, and that competing desferrioxamine B only extracts a fraction of iron bound to reference humic material. The possible ascription of non siderophore ligands to the so-called L1 pool has also been recently discussed by Mahieu and co authors. It is very likely that nanopyrites and nano aged Fe oxyhydroxides may contribute to electrochemical L1 in hydrothermal plumes (this is considered in (Kleint et al., 2016) and for the case of copper complexes in (Sander et al., 2007)).
3 The weight of the role of siderophores. In their response the authors state We have edited some sections of the manuscript to make it clear that we do not think that siderophores are a major part of the L1 pool, but rather that their presence in these systems is intriguing and suggests that the other strong ligands we observed in the neutrally-buoyant plumes might also be microbially-produced (whether or not they are siderophores). This line of argument is impeccable but not compatible with this line in the introduction: “The tight coupling between strong ligands and dissolved iron within neutrally buoyant plumes across distinct hydrothermal environments, and the presence of dissolved siderophores with siderophore-producing microbial genera, suggests that biological production of ligands exerts a key control on hydrothermal dissolved iron concentrations”. Here again, it is conferred a controlling role to a group of substances found at a maximum of 4% of the ligand concentration and the possible contribution of inorganic colloids and nanoparticles, which presence is commonly accepted, completely ignored. The ecological importance of siderophores is huge but at the concentrations shown here, and in the absence of other organic candidate to contribute to organic ligands, the data base is far from even suggesting the biological control of dFe by microbial exudation or lysis products.
4 The trend to self citation has not been corrected at all. In the response, the authors argue in favour of the relevance of their bibliography and the need to keep the number of citations low. Case by case this makes sense but when I go to the bibliography and see that authors of the weight in the iron speciation community such as Gerringa and van den Berg or in the study of hydrothermal vents such as Sander are cited a single time this does not look right. A reader not specialized in the field would think that the authors here are almost the only ones worth to be cited. The result is the absence of works that do not align with their vision of the topic. More specifically, the bibliography about the presence of inorganic phases in hydrothermal plumes has been ignored or badly cited (omitting such parts).
An illustrative example was my comment about the two self citations used to support the description of iron transition in estuaries. From the early works of Boyle and Sholkovitz, many authors have used electrochemical techniques to find ligands in estuaries and made more contributions than the authors here: van den Berg, Laglera, Whitby, Yang, Muller, Waeles/Pernet Coudrier, etc, . Furthermore, many of these studies actually indicate that iron is exported from estuaries as humic complexes, therefore, implying that, if only the iron bound to strong ligands survives the estuarine transition, as the authors suggest, a significant part of humics must belong to L1.This is also discussed in (Slagter et al., 2019; Sukekava et al., 2024). I guess that since this presence of refractory DOM reduces the theoretical contribution of siderophores to the so-called L1 concentrations, citations of these articles are not in the interest of the authors.
The authors have the logical aim for this work to transcend, but I do not think that ignoring previous effort by other authors in the field and not discussing but ignoring bodies of bibliography that do not concur with their vision of the topic, is the best approach.
5 From the revised version: The strong coupling between dFe and ligands was only observed at sites where L1 ligands were detected. Some sampling locations, such as in the buoyant plume or closer to the vent orifice, contained high concentrations of weaker ligands (log KFe’,FeLcond < 12, Table S2) with no correlation to dFe. This is consistent with these environments likely being dominated by inorganic forms of Fe as hydrothermal fluids initially mix with oxygenated seawater
This is based in assuming that inorganic iron can only be ascribed to a weak ligand group. The refractory character of iron oxyhydroxides (for instance to competition by an added ligand) is obviously a matter of aging as the aggregates move from pure hydroxides (fresh ferrihydrite) to oxyhydroxides. In this process, not only iron aggregates (in the sense of increasing size) but also becomes more refractory, which also relates to the onion model by Mackey and Zirino. There are Raiswell references (Raiswell et al., 2010) where different oxyhydroxides are subjected to the same extraction protocols including acidification with a enormous range of efficiencies, almost of them low. Nanopyrites are insoluble and should also be refractory to ligand competition.
6 from the response by the authors: and all ligand parameters reported were confirmed to be within the analytical window of the SA measurements,
This is no control of quality. Please revisit classic works from Apte and van den Berg for definitions of analytical window and its extent. The alpha coefficient of FeSA defines the CENTER of the analytical window. The WIDTH of the analytical window is given by the analytical error. All ligands can only be resolved in the absence of data error. The analytical error, even if small, prevents the accurate determination of tiny concentrations of FeL (for weak ligands) or FeLad (for strong ligands) and such ligands are not properly characterized. As a result, alpha values and therefore log K values of strong ligands are forced to fall into a window (centred around alpha FeLad) and do not correspond to the real stability of the complex. This is a limitation of the technique independently of the quality of the analysis. For the case of copper titrations this is explained in (Laglera et al., 2013; Pižeta et al., 2015) and references there in. This means that ligands of alpha coefficient stronger than the upper limit of the analytical window are detected (these ligands have not exchanged iron with the added ligand) but K is brought down so alpha falls into the analytical window. Ligands of weaker alpha are simply not measured. So, that ligand solutions fall close to the alpha of the artificial ligand is not surprising. However, in this manuscript there is something else, log alpha values of ligands fall approximately from 1.3 to 4.9 (more than 4 orders of magnitude, table in supplement) and the centre of the window should be around 1.7 from 10 uM SA and K and B values from (Abualhaija and van den Berg, 2014). The upper limit of the analytical window should be around 3.7, which implies that some of the values cannot be “confirmed”.
7 Forward and reverse titrations should appear in different tables. Since the definition of the analytical window is different, the distribution of analytical error throughout the titration and the statistical weight of data during the data treatment is very different, forward and reverse titrations cannot return the same solution for the same sample if there is more than one ligand present. This is evident from works where both methods have been used in natural samples for copper (Nuester and van den Berg, 2005; Santos-Echeandía et al., 2008; Wiwit et al., 2021).
8 The bibliography of the tracked changes version has not been updated since at least the Hassler paper is missing
9 from the response to reviewers: Operationally defining classes of organic ligands by their conditional stability constant is an accepted and preferred strategy by the field in the application of CLE-AdCSV methods (Gledhill and Buck 2012).
This is simply not true. The definition and adscription of K values to classes (as in L1 siderophores, L2 HS and L3 EPS or similar) is the preferred strategy by a part of the field, mostly the authors in here. Other authors do not use it or directly challenge it (Mathieu, Laglera, Sukekava, Gerringa, Slagter recent manuscripts cited above)
10 from the response to reviewers: We are not sure which inorganic ligands the reviewer is specifically referring to, but if they are referring to the formation of inorganic iron oxy(hydro)oxides, then the CLE-AdCSV method would not artificially detect these compounds as “organic ligands” because the side reaction coefficient of these is 𝛼 = 10 or 11 which is outside of the analytical window used in this work.
The discussion about the presence of metastable phases, nanopyrite and aging of oxyhydroxides I wrote above is of relevance here. This inorganic alpha is not exactly the strength of the inorganic complexation, it is the ratio Fe3+:Fe’ . This alpha is summed to log K if the constant is referred to Fe3+ instead of Fe’. It does not make full chemical sense because includes the insoluble Fe(OH)3 as part of the inorganic complexes and therefore it cannot be used to compare organic/inorganic complexation.
11 Section amended: “Over the last few decades, observations and modelling efforts have increased our understanding about the critical role organic ligands play in the cycling, transport, and utilization of trace metals (Tagliabue et al., 2017; Buck et al., 2018; Bundy et al., 2018; Moore et al., 2021). Organic ligands in seawater have a wide range of sources, which are only just beginning to be understood, but observations suggest that microbial production of siderophores, humic-like substances and exopolysaccharides are some of the major contributors (Hassler et al. 2017), linking microbial activity to impacts on Fe cycling. For example, microbial communities can influence Fe cycling in environments ranging from hydrothermal plumes (Cowen and Bruland, 1985; Cowen et al., 1990) to the open ocean (Lauderdale et al., 2020).”
I would add the expression “in deep waters”. In upper waters there are sources of organic ligands more important than microbial production of ligands. Phytoplankton cultures and phytoplankton blooming waters have shown substantial ligand increments (Boye et al., 2005; Boye and van den Berg, 2000; Kondo et al., 2008; Laglera et al., 2020), riverine humics are also an important source of ligands in areas of the open ocean (van den Berg, Whitby, Laglera/Sukekava, Yang, Slagter/Gerringa, Muller, Dulaquais manuscripts).
With the rest of the manuscript I agree with the changes performed.
Abualhaija, M.M. and van den Berg, C.M.G., 2014. Chemical speciation of iron in seawater using catalytic cathodic stripping voltammetry with ligand competition against salicylaldoxime. Marine Chemistry, 164: 60-74.
Boye, M. et al., 2005. Major deviations of iron complexation during 22 days of a mesoscale iron enrichment in the open Southern Ocean. Marine Chemistry, 96(3-4): 257-271.
Boye, M. and van den Berg, C.M.G., 2000. Iron availability and the release of iron-complexing ligands by Emiliania huxleyi. Marine Chemistry, 70(4): 277-287.
Fitzsimmons, J.N. et al., 2017. Iron persistence in a distal hydrothermal plume supported by dissolved–particulate exchange. Nature Geoscience, 10(3): 195-201.
Hawkes, J.A., Connelly, D., Gledhill, M. and Achterberg, E.P., 2013. The stabilisation and transportation of dissolved iron from high temperature hydrothermal vent systems. Earth and Planetary Science Letters, 375: 280-290.
Kleint, C., Hawkes, J.A., Sander, S.G. and Koschinsky, A., 2016. Voltammetric investigation of hydrothermal iron speciation. Frontiers in Marine Science, 3: 75.
Kondo, Y. et al., 2008. Organic iron (III) complexing ligands during an iron enrichment experiment in the western subarctic North Pacific. Geophysical Research Letters, 35(12).
Laglera, L.M., Downes, J. and Santos-Echeandía, J., 2013. Comparison and combined use of linear and non-linear fitting for the estimation of complexing parameters from metal titrations of estuarine samples by CLE/AdCSV. Marine Chemistry, 155(0): 102-112.
Laglera, L.M. et al., 2020. Iron organic speciation during the LOHAFEX experiment: Iron ligands release under biomass control by copepod grazing. Journal of Marine Systems, 207: 103151.
Nuester, J. and van den Berg, C.M.G., 2005. Determination of metal speciation by reverse titrations. Analytical Chemistry, 77(1): 11-19.
Pižeta, I. et al., 2015. Interpretation of complexometric titration data: An intercomparison of methods for estimating models of trace metal complexation by natural organic ligands. Marine Chemistry, 173(0): 3-24.
Raiswell, R., Vu, H.P., Brinza, L. and Benning, L.G., 2010. The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology, 278(1–2): 70-79.
Sander, S.G., Koschinsky, A., Massoth, G., Stott, M. and Hunter, K.A., 2007. Organic complexation of copper in deep-sea hydrothermal vent systems. Environmental Chemistry, 4(2): 81-89.
Santos-Echeandía, J., Laglera, L.M., Prego, R. and van den Berg, C.M.G., 2008. Copper speciation in estuarine waters by forward and reverse titrations. Marine Chemistry, 108(3-4): 148-158.
Slagter, H.A., Laglera, L.M., Sukekava, C. and Gerringa, L.J.A., 2019. Fe-binding Organic Ligands in the Humic-Rich TransPolar Drift in the Surface Arctic Ocean using Multiple Voltammetric Methods. Journal of Geophysical Research: Oceans, 124: 1491-1508.
Sukekava, C.F., Downes, J., Filella, M., Vilanova, B. and Laglera, L.M., 2024. Ligand exchange provides new insight into the role of humic substances in the marine iron cycle. Geochimica et Cosmochimica Acta, 366: 17-30.
Wiwit et al., 2021. Wide-range detection of Cu-binding organic ligands in seawater using reverse titration. Marine Chemistry, 230: 103927.
Yücel, M., Gartman, A., Chan, C.S. and Luther, G.W., 2011. Hydrothermal vents as a kinetically stable source of iron-sulphide-bearing nanoparticles to the ocean. Nature Geoscience, 4(6): 367-371. |
