JPR Advance Access originally published online on January 19, 2007
Journal of Plankton Research 2007 29(2):107-113; doi:10.1093/plankt/fbm009
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
HORIZONS |
Vitamins, phytoplankton and bacteria: symbiosis or scavenging?
Scottish Association For Marine Science, Dunstaffnage Marine Research Laboratory, Oban, Argyll, PA37 1QA, Scotland
* Corresponding Author: m.droop{at}ukonline.co.uk
Received on July 19, 2006; accepted on December 14, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
The conclusion that over 25% of global primary production depends on direct algal/bacterial symbiosis involving vitamin B12 [Croft et al., (2005)
Algae acquire vitamin B12 through a symbiotic relationship with bacteria. Nature, 438, 90–93] is patently false, for it is based on a misconception of the probable level of the vitamin B12 requirement in marine pelagic algae. A review of the various published attempts at measuring this requirement suggests that it is likely to be so low that oceanic and coastal concentrations of the vitamin would usually be sufficient to sustain the populations that occur without the assistance of direct algal/bacterial symbiosis. The levels measured are discussed in relation to method (batch or continuous culture) and protocols used. Requirement values considered by the author to be acceptable range from 0.1 to 0.3 pM for the vitamin growth saturation constant (KS) and from 30 to 100 µL algal biomass pmol–1 vitamin for the yield. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
The recent renewal of interest in the marine ecology of vitamin B12 has included a paper (Croft et al., 2005
), the hyperbole and implication of whose title warrants careful scrutiny. The authors describe a neat experiment demonstrating that algae and certain bacteria can form a loose symbiotic association involving exchange of photosynthetically produced carbon for vitamin B12. However, I question the global inference drawn from the results, suggesting that this is quite unwarranted. The idea that the amount of free vitamin in the sea and fresh waters is not sufficient to support the phytoplankton without the involvement of symbiosis rests on a misconception of the level of requirement in plankton algae. Croft et al. appear to have so misinterpreted the literature on the topic that one may even question whether they were aware of the full extent of it. I refer to work on pelagic genus Thalassiosira (Carlucci and Silbernagel, 1969; Swift and Taylor, 1974) and to some of mine (Droop, 1968, 1970) on Skeletonema costatum and the haptophyte Pavlova (Monochrysis) lutheri, all of which, incidentally, bears a modicum of responsibility for the general lack of interest in vitamin B12 on the part of marine ecologists. The thrust of my previous two communications on this topic (Droop, 1968, 1970) was centred on the kinetics of nutrient uptake and growth per se, using vitamin B12 as a model system. While the relation between the level of requirement of the vitamin and its availability in the sea was not emphasized, the published data contained the relevant information. It now appears that a pointed interpretation of the data from that continuous culture work, together with a summary of work leading up to it, would be useful, and could also be usefully discussed in relation to other work in the area. |
REVIEW |
|---|
|
TOPABSTRACTINTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
The yield relation
The yield in batch culture of P. lutheri (cell vol. 200 µm3) worked out at 
1/3 µm3 algal biomass per vitamin molecule (201 µL pmol–1, since 1 µL is equivalent to 109 µm3 and 1 mole to 6.02 x 1023 molecules) over a range of vitamin concentrations from 0.07 to 14 pM (Droop, 1957). This, however later proved to be an overestimate that only applied to cells well into the stationary phase of growth. A ‘subsistence quota’, the reciprocal of yield, of 2.25 fmol per million cells obtained in continuous culture by extrapolation to zero specific growth rate indicated a yield of 89 µL pmol–1, a more realistic estimate than the previous one, since it referred to cells with their metabolic machinery still largely intact (Droop, 1968). The previous calculations should therefore be decreased by a factor 89/201. It may be argued that P. lutheri, an inhabitant of supralittoral rock pools, may not be a good model for truly marine species, but the subsistence quota of the pelagic diatom S. costatum (cell vol. 250 µm3), similarly culled from chemostat data, was 65 µL pmol–1, a result similar to the Pavlova figure (Droop, 1970). This may be a pure chance, but similar reasoning points to requirements of the same order in Euglena gracilis (cell vol. 2000 µm3) (Hutner and Provasoli, 1951), Stichococcus (cell vol. 12 µm3) (Lewin, 1954) and Thalassiosira pseudonana (cell vol. 200 µm3) (Swift, 1984), one might be tempted to hazard that this figure may represent the yield requirement in algae generally, namely that each pmol of vitamin can yield 30–100 µL of algal biomass. (Assuming a relationship of 200 gC L–1 cell volume, this equates to 1 pmol B12 supporting 6–20 mgC biomass.) If so, judging by the size of oceanic populations quoted by Bainbridge (1957), it seems unlikely, even given the variability in C-biomass per unit of cell volume are taken into account, that more vitamin than 0.1 pM would be required to support them.
Aside from the above, there is some justification in supposing that the relation between specific growth rate and substrate concentration is of greater ecological and competitive significance than is the yield relation. Accordingly, again with batch cultures of P. lutheri (Droop, 1961), the rate-limiting concentrations proved to be below 0.07 pM, a figure I found hard to believe in view of Ford's rate data for Ochromonas malhamensis (Ford, 1958) but I could not fault it. O. malhamensis is an obligate chemotroph and it is possible that chemotrophy in the dark and the high temperature require more vitamin B12 than phototrophy. I concluded, having regard for the levels of vitamin being recorded in the sea (see the legend to Figs 1 and 2), that it was unlikely that the vitamin was of ecological importance, particularly in view of the general dearth of nitrogen. Batch cultures however do not leave sufficient room to manoeuvre when the interesting concentrations are vanishingly small. Continuous culture offers a way out of this impasse, but it has problems of its own. Paramount among these is the vitamin binding phenomenon, which is very relevant to the interpretation of the steady-state data that follows.
|
%µ/µmax) against concentration of free vitamin, as log pM. The vertical lines, the means of some assays of marine waters. A: North Atlantic and north North Sea, summer. B: north North Sea, autumn. C: Butt of Lewis, Scotland, spring (Cowey, 1956
|
The binding protein
Proteins released by cultured algal cells that combined reversibly with vitamin B12 and rendered it unavailable, were first reported by Kristensen (1956) in cultures of E. gracilis and by Ford (1958) with O. malhamensis. These authors observed that the phenomenon, if general, could interfere with attempts to establish the levels of requirement. So it has proved, for in steady-state situations the kinetics are such that interference is so obvious that the classical equations of chemostat operation (Monod, 1950; Herbert et al., 1956) cannot be applied without modification (Droop, 1968). [Symptoms of the interference are first, that the curve of growth rate on apparent substrate concentration is C-shaped, rather than a rectangular hyperbola passing through the origin and second, that the saturation constant (KS, the substrate concentration producing half maximal growth rate) appears to increase with increasing biomass and decreasing growth rate].
The modification I adopted, which is entirely successful in regard to chemostat operation, is based on the single assumption of constant rate of protein excretion per cell. This leads directly from the mass-action equation (Dixon and Webb, 1958) to a simple expression for the ratio of total to free (i.e. available) substrate in the chemostat, namely
|
|
/ 
), a composite constant being the ratio of the rate of excretion of inhibitor to the dissociation constant of the complex]. The ratio ST/SA defines the ratio of the spurious to the true saturation constant, i.e. (in the chemostat), (KS)x=i/(KS)x=0. The value of (/) is obtained from the chemostat data and is given by the regression
|
|
). Failure to take into account the binding protein, if present in continuous cultures with vitamin B12, will show total concentration which, at lower dilution rates, can reach a 100-fold that of the available vitamin.
In the application to batch cultures
|
|
|
|
In the open sea, on the other hand, the importance of the binding phenomenon is likely to be negligible. One can obtain an idea of the scale of sequestration from the S. costatum data (Droop, 1970). The value of (/) in the experiment upon which Fig. 2 is based is 16 mL medium µL–1 biomass day–1 and the ratio of total to free vitamin in the environs of an exponential population will be 1 + e(16x/µ), (x in µL mL–1, µ in day–1). S. costatum, a dominant member of the spring bloom of phytoplankton in British coastal waters, can reach 104 cells mL–1 in a good year. Assuming a cell volume of 250 µm3, this population amounts to 0.0025 µL biomass mL–1. It gives the ST/SA ratio as 1 + 0.04e (=1.108), showing that protein binding is of no consequence here; but matters could be different when populations reach red tide proportions.
The Pavlova and Skeletonema chemostat data
Figures 1 and 2 are based, respectively, on P. lutheri (Droop, 1968) and S. costatum (Droop, 1970), and use available B12 concentration, computed as detailed above, rather than total as entered in the published tables. The abscissa scale is logarithmic for the sake of clarity; on a linear scale, the points follow a typical Monod hyperbola. Any conclusions drawn from the figures strictly apply to organisms of dimension similar to the subjects and one would expect, for any particular substrate concentration, uptake to vary as the cell surface area and the growth rate to vary inversely as the cell volume, the latter therefore as the reciprocal of the linear dimension. However, since the restraint applies equally to the maximum as to the lower rates, specific growth rate expressed as a percentage of the maximum should be independent of cell size.
I remarked above that subsistence quotas of the two organisms, P. lutheri and S. costatum, indicate that each pmol of vitamin was sufficient to yield 70–90 µL of biomass. Figures 1 and 2 show that both species grow at over 80% of their maximum rates in a concentration of 1.0 pM, while in a concentration of 0.1 pM (i.e. a concentration that might be expected in the open sea) the equivalent rates are 60 and 20%, respectively. It appears that in the open sea an organism like S. costatum, if not limited by another nutrient, would reach a population of 26 million cells (=6.5 µL) L–1, while growing at 20% of its maximum rate, whereas in coastal waters, with vitamin titres 1.0 pM or above there is enough vitamin to produce populations growing at their maximum rate to a size more than an order of magnitude greater than that normally recorded. Thus, the results from continuous culture go some way in confirmation of the previous batch data, both as to yield and rate. The diatom apparently has a growth saturation constant (KS, from the curve fit of µ on SA with standard error in parentheses) about double that of the haptophyte [0.19 (25%), 0.086 (12%) pM, respectively].
Therefore, I suggest that, at any rate on the evidence of these two species, the supposed general need for symbiosis rests on a misconception of the level of requirement. In the sea, there is no mystery about the provenance of the vitamin in outline; it follows the route taken by other nutrients, namely decay and sinking of spent plankton, bacterial regeneration and upwelling.
|
COMMENT |
|---|
|
TOPABSTRACTINTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
The level of requirement
These conclusions from work, now some 35 years old, are at such variance with those reported by Croft et al. (Croft et al., 2005) that some comment is in order. The work was sufficiently focused at least to warrant some discussion; accordingly, I invite them to reconsider the global inference they make on a very doubtful premise. They may wish to consider the relative relevance of simple enrichment experiments of the type they report, vis-à-vis detailed quantitative work, and also consider the influence of the physiological state of the inoculum on their results as well as the methodological implications of the binding protein and its carryover with the inoculum. Incidentally, Croft et al. (Croft et al., 2005) include in their argument the figure 10 µg L–1 as the amount of vitamin that most cultures contain. This is hardly relevant; moreover that figure is a 100-fold the highest quantity I ever used when requiring a medium with ample excess vitamin.
To give an example on the effect of the binding protein, a 3 week old source culture of P. lutheri with, say, 8 million cells mL–1 would contain some 1600 pM of binding protein (in vitamin equivalents), which after a 100-fold dilution on transfer, the ratio of total to free and available vitamin would be something of the order of 6.3.
|
|
The two constants of the mass-action relation, and , were measured by Davies and Leftley (Davies and Leftley, 1985) for P. lutheri, and later by Sahni et al. (Sahni et al., 2001) for T. pseudonana with similar result. The example above is rather extreme, but equation (5) underlines the need to keep biomass (x) as small, and the dilution factor (d), as large as possible.
I may be accused of generalizing from results gathered from a mere two species, and Provasoli and Carlucci (Provasoli and Carlucci, 1974) are correct to caution that algae have been found to differ from one another in the level of requirement for vitamin B12. They suggest that P. lutheri may not be typical, but the two examples they cite are unfortunate. The first, dark growth of O. malhamensis, an obligate chemotroph whose normal mode of nutrition is phagotrophy (Ford, 1958), is not, as I mentioned, a good example of phototrophic phytoplankton. The second, an unpublished chemostat study (Wood, 1962) of S. costatum, did not take sequestration into account and would show substrate concentrations in error of up to two orders of magnitude at lower chemostat dilution rates. In my hands, the growth saturation constant (of free substrate) for S. costatum was less than double the mean of my P. lutheri entries in Table I and compares with those of T. pseudonana (Swift and Taylor, 1974) and T. oceanica (Carlucci and Silbernagel, 1969), which ranged from 0.13 to 0.26 pM.
|
Some measurements of the six species listed by Swift (1984)
are shown in Table I. The 20-fold discrepancy in the P. lutheri batch entries can be attributed to protocol: an initial count of 100 exponential phase cells mL–1 used in 1961 versus one of 20 000 possibly stationary phase cells mL–1 in 1974. It would appear that inoculum size affects the result. This is most likely due to interference by the yield relation when the inoculum is heavy. For example, with an inoculum of 20 000 cells mL–1 of P. lutheri, with a KS even as small as 0.1 pM vitamin, growth would have virtually ceased in 4 days in any culture having initially less than 1.4 pM vitamin. As I remarked earlier, a batch system leaves little room for manoeuvre. Also, possible hysteresis in the cell metabolism at the sudden change in substrate concentration and the need to allow time for its avoidance cannot be ruled out, while on the other hand, the effect of binding protein in the 1974 experiment would not be great, with ST/SA at the very most 2.0. The same considerations apply to the Isochrysis galbana entry. P. lutheri chemostat data indicate that the lower figure for KS is the more likely. Indeed, the figure 0.086 pM indicated by the curve fit of the chemostat µ on SA data is confirmed by the independent batch uptake data (Droop, 1968). This is possible because the Monod and Cell Quota equations are equivalent in the steady state, and they reveal the following identities when the Michaelis equation for uptake is combined with them (Droop, 1973).
|
|
|
|
|
|
Swift and Taylor (1974) point out that the chemostat-derived KS has not the same interpretation as that derived directly from the Monod equation. Very true, the one is based on SA, the other on ST. However, S in the Monod equation is necessarily always SA, and I note that the validity of my derivation of SA was not questioned. Swift and Taylor (1974) do not appear to have thought the matter through, for if there is protein binding of the vitamin then SA < ST, so that
|
|
The inocula of the Thalassiosira experiments were probably sufficiently small (1800 cells mL–1) for the conclusions concerning them to be reliable. In which case four of the entries in Table I, three of which are pelagic marine diatoms, are very similar in their vitamin B12 requirement, both on a yield and a rate basis, and it is very low; they would not generally be limited by the amounts of vitamin occurring in the sea. Other pelagic organisms may have greater requirements, but where is the evidence?
Of course more is likely to be involved in the control of phytoplankton than simply the level of requirement as measured in the laboratory. It is known that clones cultivated for long periods may adapt to the conditions in which they are kept, and with a plentiful nutrient supply, they might be expected eventually to loose an ability to grow at high rates with low nutrient concentrations and hence come to need higher levels than they did in the wild. I have not considered the provenance and requirement for the various variants of the vitamin molecule, nor of temperature effects. Even the meaning of ‘requirement’ becomes blurred: Swift and Guillard (1978), for instance, found significant stimulation of a number of non-auxotrophic diatoms by the addition of vitamin B12. Another example that also clouds the issue is a recent paper by Sañudo-Wilhelmy et al. (2006) demonstrating that a natural population, brought into the laboratory and enriched with 500 pM vitamin, can result in a 10-fold increase in biomass, but only in the >5 µm fraction, a result that is difficult to interpret in view of the enormous quantity of vitamin employed.
Regarding the source of vitamin B12, that it is eventually bacterial is not in question (Hall et al., 1950; Haines and Guillard, 1974), nor is the fact that aquatic bacteria are normally associated with solid matter, nor that some live in close association with algae and can be regarded as symbiotic (Cole, 1982). This, however is a long step from most, especially pelagic, algae being necessarily in symbiotic association as suggested by Croft et al. (Croft et al., 2005) on the evidence of one experimental example (of a facultative terrestrial species). Furthermore, regarding the measured marine concentrations of vitamin B12, Andersen (Andersen, 2005) points out, quoting Menzel and Spaeth (Menzel and Spaeth, 1962), that ‘there is strong circumstantial evidence that this vitamin is produced on the ocean floor at depths where darkness makes it unlikely that an algal-bacterial symbiosis can exist’.
In the 50s and 60s of the last century interest, engendered by the ideas of Lucas (Lucas, 1949) and Lefèvre et al. (Lefèvre et al., 1952) in the possible importance of micronutrients and so-called ‘ectocrines’ in marine and fresh waters, was strong and the question of close association and even symbiosis with bacteria was very much in our minds. In one attempt at an answer I and my colleague Keith Elson (Droop and Elson, 1966) looked for and tested for bacteria associated with several hundred cells of the large centric diatom Coscinodiscus concinnus from tow-nettings over a period of several months in the winter of 1963. We saw no bacteria on the cells and only 12 individuals out of 533 tested positive for bacteria. We also took chains of S. costatum, also directly from tow-nettings, in the following spring and examined them for attached bacteria under phase contrast. We found that during the exponential phase of the spring bloom (14 days) the bacterial count remained constant at from 3 to 5 per 100 diatom cells and remained so for the first week of the stationary phase, then quite suddenly the count began a rapid increase and was accompanied by signs of diatom lysis. These observations indicate, quite conclusively, that for these two species at least symbiosis was not involved during the growth phase. One cannot of course generalize from two, albeit typical and common, species, but the finding is not in contradiction of the general impression that healthy pelagic diatoms are remarkably clean (Waksman et al., 1937; Cole, 1982).
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
Having regard for the protocols used for measuring the level of vitamin B12 requirement in the few marine phytoplankton for which there are data, it would appear that the requirement, on the present evidence, and in the light of marine vitamin available, is sufficiently low for there not to be a general need for direct bacterial symbiosis involving the vitamin to account for the populations occurring.
‘Algae acquire vitamin B12 through a symbiotic relationship with bacteria’ (Croft et al., 2005). Symbiosis certainly, but in the sense that mankind is in symbiotic association with wheat.
|
ACKNOWLEDGEMENTS |
|---|
I am indebted to Dr John Leftley for discussions and especially for his knowledge of the literature, and to Dr Fiona Hannah for facilitating our access to it.
|
Notes |
|---|
Communicating editor: K.J. Flynn
Written responses to this article should be submitted to Kevin Flynn at k.j.flynn{at}swansea.ac.uk within two months of publication. For further information, please see the Editorial ‘Horizons’ in Journal of Plankton Research, Volume 26, Number 3, Page 257.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONREVIEWCOMMENTCONCLUSIONREFERENCES |
|---|
Andersen R. A. (2005) Algae and the vitamin mosaic. Nature 438 33–35.[CrossRef][Medline]
Bainbridge R. (1957) The size, shape and density of marine phytoplankton concentrations. Biol. Rev. 32 91–115.
Carlucci A. F. (1970) The ecology of the plankton off La Jolla, California in the period April through September 1967, II. Vitamin B12, thiamine and biotin. Bull. Sci. Inst. Ocean. 17 23–30.
Carlucci A. F. and Silbernagel S. B. (1969) Effect of vitamin concentrations on growth and development of vitamin-requiring algae. J. Phycol. 5 64–67.[Medline]
Cole J. J. (1982) Interactions between bacteria and algae in aquatic ecosystems. Ann. Rev. Ecol. Syst. 13 291–314.[CrossRef][ISI]
Cowey C. B. (1956) A preliminary investigation of the variation of vitamin B12 in oceanic and coastal waters. J. Mar. Biol. Ass. UK 35 609–620.
Croft M. T., Lawrence A. D, Raux-Deery E, et al. (2005) Algae acquire vitamin B12 through a symbiotic relationship with bacteria. Nature 438 90–93.[CrossRef][Medline]
Davies A. G. and Leftley J. W. (1985) Vitamin B12 binding by microalgal ectocrines: dissociation constant of the vitamin binder complex determined using an ultrafiltration technique. Mar. Ecol. Prog. Ser. 21 267–273.
Dixon M. and Webb E. C. (1958) EnzymesLondon Longmans pp. 782.
Droop M. R. (1955) A suggested method for the assay of vitamin B12 in sea water. J. Mar. Biol. Ass. UK 34 435–440.
Droop M. R. (1957) Vitamin B12 in marine ecology. Nature 180 1041–1042.
Droop M. R. (1961) Vitamin B12 and marine ecology: the response of Monochrysis lutheri. J. Mar. Biol. Ass. UK 41 69–76.
Droop M. R. (1968) Vitamin B12 and marine ecology, IV. The kinetics of uptake, growth and inhibition in Monochrysis lutheri. J. Mar. Biol. Ass. UK 48 689–733.
Droop M. R. (1970) Vitamin B12 and marine ecology, V. Continuous culture as an approach to nutritional kinetics. Helgoland Wiss. Meeresunters. 20 629–636.[CrossRef]
Droop M. R. (1973) Some thoughts on nutrient limitation in algae. J. Phycol. 9 264–272.[CrossRef][ISI]
Droop M. R. (1974) The nutrient status of algal cells in continuous culture. J. Mar. Biol. Ass. UK 54 825–855.
Droop M. R. and Elson K. G. R. (1966) Are pelagic diatoms free from bacteria? Nature 211 1096–1097.[CrossRef]
Droop M. R., Mickelson M. J., Scott J. M., et al. (1982) Light and nutrient status of algal cells. J. Mar. Biol. Ass. UK 62 403–434.
Ford J. E. (1958) B12-vitamin and growth of the flagellate Ochromonas malhamensis. J. Gen. Microbiol. 19 161–172.[Medline]
Haines K. C. and Guillard R. R. L. (1974) Growth of vitamin B12 requiring marine diatoms with vitamin B12 producing marine bacteria. J. Phycol. 10 245–252.[CrossRef][ISI]
Hall H. H., Benjamin J. C., Bricker H. M., et al. (1950) A survey for B12 producing microorganisms. Proc. Soc. Amer. Bact. 50 21.
Herbert D., Elsworth R., Telling R. C. (1956) The continuous culture of bacteria; a theoretical and experimental study. J. Gen. Microbiol. 14 601–622.[Medline]
Hutner S. H. and Provasoli L. (1951) The phytoflagellates. In Lwoff A. (Ed.). Biochemistry and Physiology of ProtozoaNew York Academic Press pp. 29–128.
Kristensen H. P. O. (1956) A vitamin B12-binding factor formed in cultures of Euglena gracilis, var. bacillaris. Acta Physiol. Scand. 36 8–13.
Lefèvre M., Jacob H., Nisbet N. (1952) Auto- et hétéroantagonisme chez les algues d'eau douce. Stn. Cent. Hydrobiol. Appl. 4 197.
Lewin R. A. (1954) A marine Stichococcus sp. which requires vitamin B12 (cobalamin). J. Gen. Microbiol. 10 93–96.[Medline]
Lucas C. E. (1949) External metabolites and ecological adaptations. Symp. Soc. Exp. Biol. 3 336–356.[Medline]
Menzel D. W. and Spaeth J. P. (1962) Occurrence of vitamin B12 in the Sargasso Sea. Limnol. Oceanogr. 7 151–154.
Monod J. (1950) La technique de culture continue; th
orie et applications. Ann. Inst. Pasteur, Paris 79 390–410.
Provasoli L. and Carlucci A. F. (1974) Vitamins and growth regulators. In Stewart W. D. P. (Ed.). Algal Physiology and Biochemistry.Oxford Blackwell pp. 741–787.
Sahni M. K., Spanos S., Wahrman M. Z., et al. (2001) Marine corrinoid-binding proteins for the direct determination of vitamin B12 by radioassay. Anal. Biochem. 289 68–76.[CrossRef][ISI][Medline]
Sañudo-Wilhelmy S. A., Gobler C. J., Okbamichael M., et al. (2006) Regulation of phytoplankton dynamics by vitamin B12. Geophys. Res. Lett. 33.
Swift D. G. (1984) Algal assays for vitamins. In Shubert L. E. (Ed.). Algae as Ecological IndicatorsNew York Academic Press pp. 281–313.
Swift D. G. and Guillard R. R. L. (1978) Unexpected response to vitamin B12 of dominant centric diatoms from the spring bloom in the Gulf of Maine. J. Phycol. 14.
Swift D. G. and Taylor W. R. (1974) Growth of vitamin B12-limited cultures: Thalassiosira pseudonana, Monochrysis lutheri, and Isochrysis galbana.. J. Phycol. 10 385–391.[CrossRef][ISI]
Waksman S. A., Stokes J. L., Butler M. R. (1937) Relation of bacteria to diatoms in sea water. J. Mar. Biol. Ass. UK 22 359–373.
Wood E. A. (1962) An evaluation of the role of vitamin B12 in the marine environment. PhD Thesis,New Haven, USA Yale University.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||