The Seasonal Variation in Biological Composition of Certain Plankton Samples from the North Sea in Relation to Their Content of Vitamin A, Carotenoids, Chlorophyll, and Total Fatty Matter

As a preliminary step to the further investigation of these problems it was decided to study the seasonal variation of the carotenoid, chlorophyll and vitamin A contents of North Sea plankton. In addition it was planned to examine the biological composition parallel with the chemical examination, with the idea of seeking a correlation, if such exists. The present paper describes these results and also the result of an examination of a single bulkier sample consisting almost entirely of the diatoms Rhizosolenia styliformis and Biddulphia sinensis.

The material consisted of plankton collected each month, from January to November 1936, at six stations between Flamborough Head and the Dogger Bank. The first position was latitude 54° 8’ N., longitude 0° 0’ E., thereafter the stations fell at $1212$ mile intervals on a course east by north magnetic. This line of stations has been regularly used for some years and is discussed elsewhere in greater detail (Wimpenny, 1933, 1937).

Temperature and salinity observations were recorded at 10 m. depth intervals at each station by the use of a Nansen-Petterson water-bottle, and on two cruises samples for phosphate determination were collected by the same means. These latter were examined by Mr M. Graham using the Lowestoft Photometer (Graham, 1936), and we are indebted to him for allowing us to use these results. Three vertical hauls were also made with a Hensen net having a silk cone of 60 meshes to the inch. The catch from the first haul was preserved in weak neutral formalin (circa$212$%) for biological examination. The second haul was passed through filter paper and dried in the engine room until a biscuit-coloured solid was obtained. This was then sealed up in a dry jar, labelled and left for subsequent fat assessment. The third sample was transferred to a in. tube containing 20 c.c. acetone, except when the catch was too bulky, when either a half or only one-thirtieth part was so treated. The latter dilution was achieved by making the plankton up to 150 c.c. in a whirling flask and taking 5 c.c. with a stempel pipette. On return to port the acetone-preserved samples were immediately dispatched to Manchester for chlorophyll, carotenoid, vitamin A and total solid determinations, whilst the samples for the estimation of ether-soluble matter and biological composition were dealt with at Lowestoft. An estimate of the numbers and kinds of organisms present was carried out on the same lines as similar earlier work (Wimpenny, 1937).

The ether-soluble matter in the dried plankton was determined by extracting the samples in glass tubes (⁠$14$ in. bore) having small holes in their bases plugged with glass-wool. These tubes were first weighed and the “biscuit “of dried plankton, after powdering, was poured into them and dried in an oven at 80° C. for 2 days prior to a second weighing to determine the weight of the powder. The tubes and powder were then extracted together in a Soxhlet apparatus for 9 hr. with ether (British Drug Houses “Analar”). After this they were again dried at 80° C. for 2 days and then weighed. This last weighing subtracted from the second weight gave the weight of ether-extracted matter, and applied to the total weight of the powder gave a percentage of ether-soluble matter to the dry weight at 80° C. This method has been in regular use by one of us (R. S. W.) since 1932, and has been found to give satisfactory results with the small amounts of material that so often have to be dealt with. The values obtained in this investigation ranged from 1.4 to 32.3%.

Chlorophyll was detected in the samples by its green colour, its characteristic red fluorescence and its equally characteristic absorption spectrum.

Fucoxanthin was tested for (in representative samples) by treating an ether extract of the material (before saponification) with 25 % aqueous hydrochloric acid, when a blue colour would indicate the presence of the typical algal carotenoid, fucoxanthin, although the less common flavoxanthin and violaxanthin also give the same reaction. Positive tests were given in the case of the August stations nos. 2, 3 and 4 only (cf. Table I), but it is probable that the negative tests indicated fucoxanthin below the amount necessary to give the test rather than complete absence.

Carotene was detected as follows : Five to ten extra plankton samples were mixed together and extracted with acetone. The solutes were transferred to light petroleum and the extract washed with 90 % methyl alcohol, when the main bulk of pigment remained in the petrol phase. The two fractions were adsorbed separately on alumina from light petroleum, the epiphasic fraction, containing the major part of the pigment, yielding three coloured zones on the chromatogram. The colours of the zones and the absorption maxima of the eluted pigments were as follows: (a) top, green, 670 and 660 mµ (chlorophylls), (b) reddish yellow, 484, 453 mµ (carotene), (c) buff-yellow, 479 and 450 mµ (values for light petroleum B.P. 6o-8o°).The hypophasic fraction on adsorption yielded an upper green zone of chlorophyll and a small red-brown zone, presumably of a xanthophyll (maxima 487 and 455 mµ in light petroleum).

In a separate examination of a mixed catch of almost pure Rhizosolenia styli-formis and Biddulphia sinensis which were available in larger quantities the carotene : xanthophyll ratio was found to be 1 : 1.82, which is comparable with the 1 : 2 ratio commonly found for land plants.

Biological tests (as mentioned above) have shown that gross plankton samples in general exhibit vitamin A activity when fed to rats on a diet deficient in this factor. When, however, the question is raised whether the biological activity is due to vitamin A precursors, such as carotene, or to vitamin A itself, the evidence is not so clear. It might have been expected from the analogy of land plants and animals that phytoplankton would contain carotenoids and none of the true vitamin A, whereas zooplankton would contain vitamin A itself with or without carotene. Previous work on the subject is somewhat contradictory; thus, phytoplankton has been recorded having high biological activity and giving a strong blue colour with antimony trichloride, but whether both these reactions were due solely to carotene is not clear. On the other hand, several specimens of zooplankton have been reported as possessing no biological activity and giving no blue colour with antimony trichloride (Drummond & Gunther, 1930).

Our examination of the gross plankton shows clearly that it contains both carotene and vitamin A. The evidence for the presence of the former has been given, whilst for the latter we have the following data :

• Routine samples examined throughout the year normally exhibit an inflexion at 320-330 mµ (alcohol) in the absorption spectra curves of their rm-saponifiable fractions.

• Certain samples, notably nos. 4, 5 and 6 of the March series, exhibited well-defined absorption bands of high persistence typical of that shown by vitamin A.

• The ordinary samples usually give transient blue colours with antimony trichloride which rapidly fade. The residual yellow or red of the solutions makes spectroscopic examination of the blue component very difficult (cf. also Drummond & Gunther, 1930). However, by collecting a series of samples together and saponifying them, a stronger solution in chloroform was obtained which gave a fairly stable blue colour with antimony trichloride exhibiting clear bands at 617 and 570 mµ. That the 617 mµ band was due to vitamin A is supported by the fact that a trace of 7-methyl-indole reduced the intensity of the band whilst a little more suppressed it completely. The ultra-violet absorption spectrum of this bulked plankton sample exhibited, as usual, only an inflexion instead of a band between 320-330 mµ.

We have made a careful examination of the only sample of phytoplankton that we have been able to obtain in reasonably pure form and in sufficient quantity. This consisted almost entirely of Rhizosolenia styliformis and Biddulphia sinensis and was obtained in the southern North Sea by means of tow-nets at various depths on 18 October 1937. It was examined especially for the presence or absence of vitamin A, as follows : 100 c.c. of the thick green suspension in water were saponified with 20 % alcoholic KOH and subsequently extracted with ether. The suspension contained 3.1 % of dry matter, and the total carotenoids were found to be o-io% on dry weight. The solutes in ether were transferred to light petroleum and extracted seven times with 90 % methyl alcohol (volume for volume). This should extract all the xanthophyll and all the vitamin A, if any were present (Gillam & Senior, 1936). The methyl alcohol extracts were bulked and diluted with water and the solutes driven into ether. After evaporation of the ether the whole of the residue was treated with antimony trichloride when only a weak olive brown colour was obtained instead of what would have been quite a strong blue had any appreciable vitamin A been present. It is to be concluded, therefore, that these particular phytoplanktonic species contain no detectable vitamin A (at least at this time of year).

After preliminary tests the following scheme of quantitative analysis was adopted. The gross planktonic organisms were filtered from the acetone extract and washed successively with small portions of acetone and ether until colourless. The filtrates were made up to known volume, whilst the insoluble residue was dried and weighed, the result being the weight of fat-free solid in the sample.

The intensity of absorption at 665 mµ of an alcoholic dilution of the etheracetone extract was first determined (Hilger-Nutting visual spectrophotometer). From the value so obtained the chlorophyll in the sample was calculated using a value of (alcohol) for “chlorophyll”. This value was obtained experimentally on a mixture of pure chlorophyll A and pure chlorophyll B in the ratio of 2.5 : 1, and agrees with published spectrophotometric data on the chlorophylls (Heierle, 1935).

Having determined chlorophyll on a small portion of the solution and returned it to the main bulk, the solvents were removed and the product saponified together with the solid residue previously separated. This was carried out under reflux with 10 ml. of 20 % alcoholic KOH for 30 min. The resulting soaps were diluted with twice their volume of water and extracted with ether (4 × 25 ml.). The washed ether was dried (Na₂SO₄) and removed in a stream of nitrogen, the residue being dissolved in alcohol and made up to known volume. The absorption spectra were then determined photographically in both the visible and the ultra-violet regions of the spectrum, using a Hilger E₃ quartz spectrograph and Spekker Photometer.

The total carotenoids (as carotene) were determined from the absorption at 452 mµ taking at this wave-length for pure carotene as 2500.

Vitamin A was determined from the intensity of absorption at 328 mµ (after correction for the absorption due to carotene), the absorption of pure vitamin A being taken provisionally as (alcohol) = 1600 (Carr & Jewell, 1933).

In a few of the samples the ultra-violet absorption spectrum exhibited a well-defined band of high persistence, but in the majority of specimens the band was masked by the presence of considerable non-selective absorption. The presence of light absorbing material other than vitamin A is supported by the fact that in a comparison of the results of quantitative blue tests with those obtained by the ultraviolet evaluation of the vitamin A in a representative bulked sample, the blue test gave very much lower results. Part of this discrepancy, however, might be caused by inhibition of the blue colour—a very common phenomenon. In any case the ultra-violet values alone have been recorded and must be regarded as maximal.

The detailed results of the spectroscopic determinations of pigments and vitamin A are shown in Table I, the seasonal variations being best seen from Fig. 1.

The four most important diatoms recorded were Rhizosolenia styliformis, R. alata, Coscinodiscus concinnus and Chaetoceras boreale. Rhizosolenia styliformis occurred in two periods of the year, i.e. from April to June and from September to November. On both occasions the flowering lay mainly on the Dogger Bank side of the line, and the maximum was found at the Dogger Bank station in November. The earlier flowering was much the lighter. The other large diatom, Coscinodiscus concinnus, was also found at two different periods, from February to June and from September to November. The maximum for the year was in June with an autumn maximum in October. It was the smaller diatoms, however, that constituted the majority of the great numbers found in May and August. Their numbers were so much greater than that of the larger species that even if we give them the reasonable allowance of 1/100th the size of the larger species, they would still represent the greater part of the diatom crop by volume. Chaetoceras boreale, which occurred from January to June and again in November, was almost entirely responsible for the numerical superiority of the May outburst. It will be suggested below that the diatom flowering of August had a special local origin in the movement of a thermocline, and that in this it differed in nature from the spring and autumn outbursts usual in temperate seas. In this connexion it is interesting to note that the August diatom maximum was characterized by the enormous numbers of the small species Rhizosolenia alata, large diatoms being absent, though two other small species of Rhizosolenia, R. shrubsolei and R. stolterfothii, were present in numbers. Considering the disposition of R. alata along the stations of the line in the month of its maximum, it should be noted that it occurred most abundantly at the three stations where the bottom water was relatively cold and probably more representative of northern water than that occurring a month earlier (Fig. 2). The species occurred during all months from July to November. In November it exhibited a small secondary maximum, and in August it was observed to be forming auxospores.

The peridinians were composed chiefly of the genus Ceratium, of which C. macroceros was by far the most characteristic species and rose to its absolute maximum in August. Considered in relation to the other peridinians, however, C. macroceros had its greatest relative abundance in October.

The general numerical distribution of the zooplankton throughout the year has been most affected by that of the small copepods, of which Pseudocalanus elongatus was by far the most important component. This species occurred throughout the year but was most abundant from May to August, rising to its peak in the latter month (Fig. 3). There was a minor peak at the time of the first diatom flowering. Females with eggs were observed in April, June and August. Another small calanid which occurred in all months was Paracalanus parvus. There were very small numbers of this species until August, whence and until November it was more frequent. Females with spermatophores were found in September and November, when, and also in October, the only males were found. There are signs, therefore, that a spawning had taken place towards the end of the year. The largest copepod, Calanus finmarchicus, differed from the other important zooplankton organisms in preferring the earlier half of the year. The species was present at most stations in all months, but its main occurrences took place between April and July, with a minor period of abundance in September and October. The distribution of the ova indicates two spawnings, one between March and May, and another and smaller one between August and October, again culminating in the last month.

It will be apparent that both were more abundant in the later half of the year and that 5. setosa was much more frequent than S. elegans. Compared with 1934 (Wimpenny, 1937), however, 5. setosa vfus, much reduced in numbers, whilst S. elegans remained nearly the same.

The residual current of the North Sea moves down the east coast of Britain towards the Dutch coast, whence, reinforced by a current entering the English Channel, it moves north-east and finally northwards up the Norwegian coast–giving rise to several swirls in its course. The stations from which our samples were taken lie across a submarine channel between Flamborough Head and the Dogger Bank, and down this channel there passes a south-going current of water of a higher salinity than that which it is about to enter. At this juncture it is also associated with an anti-clockwise swirl around the area south-west of the Dogger Bank, and there is no doubt that there are other minor circulations in the neighbourhood in which the water masses we are considering may become involved from time to time. These, however, will not affect the main drift to the shallower water of the south.

The chief hydrological features are shown in Fig. 2. It will be seen that the invasion of the area by salter water was mainly in the early part of the year and lasted until June. This spring and early summer access of water of oceanic origin bears with it the nutrient salts, which, when the conditions of illumination are adequate must contribute to the success of the first diatom outburst. The surface temperature rose steeply from May to July, and reached its maximum in September. The warming up of the surface water each summer causes its sharp separation from cooler underlying water in the north and central North Sea. The result is that the warm water in the upper illuminated zone continues to produce an algal vegetation until it has exhausted its supply of nutrient salts. These latter are present in the dark, cool bottom layers and become available at the surface around the edge of the area of thermal stratification, where there is a vertical interchange of water. The edge of such a thermocline often produces an abundant diatom flowering. This would be particularly likely when the water movement was from the thermocline area and out across the edge. In the area we are discussing the thermocline approached in May and August and doubtless was the sole cause of the diatom outburst in August. The temperature sections on the right of Fig. 2 show that the thermocline had entered the channel in September, being very well marked at the two middle stations.

The phosphate determinations show that in August these values were low in water less than 20 m. deep, and indicate their utilization in the diatom maximum of that month. In September, on the other hand, the diatom outburst had subsided and there was a considerable regeneration of phosphate, particularly in the deeper water.

In October each year the thermocline breaks down in this area (Savage & Wimpenny, 1936), and it is likely that it is this breakdown, bringing nutrient salts into the upper illuminated layer, which causes the last and smallest flowering of diatoms.

The seasonal distribution of the various chemical and biological constituents, with respect to which the plankton has been examined, are shown as mean values per cubic metre for the whole six stations for each month in Figs. 1, 3 and 4 and in Table I.

The biological train of events may now be shortly summarized (Fig. 4). The year begins with a scanty plankton, the animals being in a numerical majority. The numbers rise until May, when the diatoms reach their greatest numerical abundance for the year. The zooplankton also rise in numbers to a peak in May, but both plankton communities show a decrease in June and July. At this time the zooplankton, though numerically less than the phytoplankton, is sufficiently numerous to make up practically the whole mass of the catch. In August the diatoms rise to a large secondary maximum and the zooplankton reaches its highest numbers for the year. Thereafter the abundance of all organisms falls away until November, although there is a slight rise in the number of diatoms and a check in the decline of zooplankton organisms in October. The occurrence of peridinians throughout the season forms a rather close parallel with that of the zooplankton.

The monthly course taken by the fat-free solids, chlorophyll and the carotenoids, resembled that of the plankton. Ether-soluble matter had its minimum in February and its maximum in July (Fig. 4). There was a secondary maximum in March-April, and it is to be noted that each maximum precedes by one month the main diatom outbursts and the times when the occurrence of ova show that zooplankton breeding has taken place (Table II and Fig. 6). On the whole it appears that the plankton was more fatty in the early part of the year when the temperature was lower.

The monthly distribution of vitamin A is plotted on the same part of Figs. 1 and 4 as the total solids. It shows a vigorous spring maximum in May and June which corresponds to the biological and chemical values just discussed. For the rest of the year there is an inverse relation to the course taken by fat-free solids–a type of behaviour we might expect from an autocatalyst of growth. The main development of vitamin A, therefore, coincided with the first diatom maximum and the zooplankton breeding period, being possibly used up as the main mass of the plankton was developed later in the year.

It is notable that Drummond & Gunther (1934), examining plankton samples from the Irish Sea in 1928, found no evidence of vitamin A (as distinct from carotene) in either phyto- or zooplankton. On the other hand, a seasonal variation in the vitamin A content of halibut-liver oil has been demonstrated by Lovern and others (1933) for the year 1932. They found a maximum in May with a minor rise in the autumn and point to a striking correspondence between these values and the mean seasonal distribution of diatoms for the period 1907-20. Our present work agrees in showing a clear correspondence with the vernal diatom outburst. It has been stated above, however, that an examination of a considerable sample made up almost entirely of diatoms showed no trace of vitamin A. In these circumstances it seems likely that the vitamin A found in our monthly samples had its origin in the zooplankton.

During the biological assessment of the plankton the proportion feeding was estimated for a few of the principal components by a method already described (Wimpenny, 1937). The results for Pseudocalanus elongatus, Paracalanus parvus and Sagitta setosa are given below :

These values and the corresponding strengths of the specific populations for the different months have been plotted on the same graph as the appropriate carotenoid values in Fig. 3. A consideration of this figure shows two things. First, that the proportion feeding tends to fall away as the population increases and the number of mouths to feed increases in relation to the food supply, and secondly, that the anomalous peak for carotenoids occurring in July coincides with a smart drop in the proportion feeding for all the species (Fig. 3). July appears to be a biological interlude between the spring outburst of diatoms and that of late summer. It also precedes the breeding periods of several of the important species and it is suggested that carotenoids, which may earlier have been used by the zooplankton as a source of the abundant vitamin A of May and June, are not being made use of in the same way as they have been before or after this time. There is no evidence of the abundant production of vitamin A after this month, but there is not to be excluded the possibility that the carotenoids may be utilized in the production of other chemical substances.

The number of organisms present at each station has been plotted for the different months in Fig. 5, whilst in Figs. 5-7 the same method has been used in respect of percentage ether-soluble matter (“fat”), total solids, diatoms, non-larval zooplankton, ova, chlorophyll, carotenoids and vitamin A.

It will be noticed that the number of organisms appears to have spread in from the right-hand side of the diagram (Fig. 6), which represents the Dogger Bank side of the line, to the left, which is towards the Yorkshire coast. This spread inwards is also shown by total solids, the non-larval zooplankton and diatoms, and may be taken as meaning that organic production begins earliest in the Dogger Bank neighbourhood spreading therefrom later. The maximum for the number of organisms lay in mid-channel in May, whilst there was a secondary one near the Yorkshire coast in August. For total solids the maximum was near the coast in August. This difference between numbers and weight indicates that the larger organisms (zooplankton) were more abundant in August. Indeed, if we compare the distribution of total solids with the larger organisms represented by the non-larval zooplankton, we get a broad resemblance in the diagrams, and when we compare the distribution of total numbers per cubic metre with that of the millions of tiny diatoms, we get almost identical pictures (Figs. 5 and 6).

The distribution of fat-content (Figs. 4 and 5) does not present an orderly picture. Nevertheless, it gave signs that its peaks precede the two great diatom maxima.

Chlorophyll, carotenoids, and vitamin A, all showed maxima corresponding to the great spring diatom maximum. Thereafter, however, only chlorophyll followed the distribution of diatoms, carotenoids having a second maximum in July at stations 12 and 24 miles from Flamborough Head, and vitamin A occurring more often than not in inverse relation to the other constituents. Even for chlorophyll the correspondence with the number of diatoms was only a broad one, and in these circumstances it seems unlikely that any arbitrary colorimetric measurement based on the amount of chlorophyll present in an acetone or ether suspension will give more than a rough estimate of the diatoms present in any sample. In the case of vitamin A it should be noted that the area of very high values in May, June, and July suggested a movement towards the Dogger Bank. This is the opposite of the extension from the Dogger Bank towards the shore apparent in the distribution of the plankton organisms and that of fat-free solids. It recalls the tendency to seasonal inversion in these same relations.

The zoological and hydrological sequence of events recorded above for the year 1936 is approximately the same qualitatively each year, although variations on the quantitative side are naturally to be expected, in addition to effects produced by the imposition of cycles of longer duration than the annual one considered here. This is also probably true of the chemical data, but we are continuing our investigations into this point.

A combined chemical and biological study of the plankton of the southern North Sea has been made. Three vertical hauls, with a Hensen net (silk cone, 60 meshes to the inch) were taken at six positions between Flamborough Head and the South-West Patch of the Dogger Bank, monthly from January to November 1936. One haul was used for biological examination, the second for determinations of percentage of ether-soluble matter and the third to determine carotenoids, chlorophyll, vitamin A and total solids.

The biological results showed that the total plankton depended on three diatom outbursts, the largest in May, another little inferior in numbers in August, and a small one in October. Peridinians and the zooplankton occurred successively in relatively greater numbers with each outburst, but reached their maximum in August when the mass of the plankton was at its greatest.

Carotene, chlorophyll and vitamin A were definitely detected in the gross plankton extracts, whilst positive tests for fucoxanthin were only obtained on a few isolated occasions. An examination of a large phytoplankton sample (obtained by tow net) containing Rhizosolenia styliformis and Biddulphia sinensis only, showed that no vitamin A as such was present. Carotene and xanthophyll, however, were present, in the ratio of 1 : 1.82, which is comparable with the ratio typical of land plants. Total carotenoids equalled 0.1 % calculated on dry weight.

Of the chemical constituents the seasonal variation of chlorophyll most nearly coincided with the total mass of the plankton. The carotenoids reached their peak slightly before the maximal biological development, whilst the vitamin A content reached its maximum in the month after the spring diatom outburst and much preceded the maximum for the plankton crop as sampled by the Hensen net.

We are indebted to Prof. I. M. Heilbron, F.R.S., for his stimulating interest and for helpful discussions during the course of this investigation.

Note added in proof, 4 November 1938. Lederer [Bull. Soc. Chim. biol., Paris, 20, 567(1938)] has recently obtained crystalline astacine in relatively large quantities from particular bulked specimens of the organism Calanus finmarchicus, thus confirming an earlier indication of its presence in this species [Euler, Hellstrom & Klussmann, Hoppe-Seyl. Z. 228, 77 (1934)]. The values for “total carotenoids” recorded in the present investigation would not include astacine which is held back as a sodium salt after saponification. Further experiments on the carotenoids of plankton which are now in progress have been modified to include astacine in the determination.