A pH Method for Determining the Carbon Dioxide Exchanges of Marine, Brackish - Water, and Freshwater Organisms

Various methods, depending upon indicatometric measurements of the hydrogen-ion exponent, have been devised and employed for determining the carbon dioxide exchanges of small animals and plants, e.g. Haas (1916) and Osterhout (1918). Some of these involve the addition of a non-toxic indicator to the water in which the organisms are suspended, noting the colour-changes of the indicator after various intervals of time, and comparing them with the changes produced by known increments or decrements of carbon dioxide, under similar conditions. The principle has been extended to the study of subaerial organisms by equilibrating a known volume of water containing indicator with the air surrounding the plants or animals under investigation, subsequently making allowance for the distribution of carbon dioxide between the solution and vapour phases at the temperature of the experiment (Jacobs, 1920).

In the case of plants, where the rate of carbon-assimilation is likely to be influenced by the colour of the incident light, the presence of an indicator in the reacting medium is undesirable. This difficulty is overcome in the simple and effective apparatus described by Irwin (1920), in which air is circulated in a closed system consisting of two tubes, one containing the organism in water, and the other the indicator solution, which rapidly attains CO₂-equilibrium. A U-tube attachment, or by-pass, containing caustic soda, may be temporarily brought into the circuit when, in respiratory experiments, the concentration of carbon dioxide has reached the permissible physiological limit, or the sensitive limit of the indicator. It is, of course, essential that the volume, alkalinity and temperature of the solutions in the two tubes should be either exactly equal, or else accurately known.

The indicatometric method, in the case of freshwater animals, has been standardised by Saunders (1.923, 2), who caused the animals to respire, out of contact with the air, in very dilute (0.0025 N) solutions of sodium carbonate, previously equilibrated with the carbon dioxide present in normal outdoor air and containing the appropriate indicator. The gradual decrease from the initial pH, 8.6o, at 18° C., referred to a table, affords a measure of the CO₂-output The table is based upon a formula deduced by Prideaux (1915), on theoretical grounds, connecting the total CO₂-content of dilute solutions of carbonates and bicarbonates with the hydrogenion concentration, and its accuracy has been independently tested by Saunders at several points.

In the case of marine organisms, the buffer-action of the sea-water, due to the presence of bicarbonates of magnesium and calcium, greatly minimises the change of pH consequent upon the gain or loss of a given amount of carbon dioxide, and the extent of the pH change is also dependent upon the total concentration of bases combined with the buffer acids—the so-called “excess base,” or less appropriately, “alkalinity.” Despite the steepening of the curve due to the buffer-effect, a difference of 0.05 pH unit—readily detectable by the eye when the brilliant sulphone phthalein indicators are used—corresponds to a gain or loss of about 0.001 mgm. of CO₂ when 10 c.c. of sea-water are used in the determination. M’Clendon and his co-workers (1917) have investigated very fully the relations existing between the hydrogen-ion concentration, CO,-tension and CO₂-content of normal (oceanic) sea-water, containing various proportions of excess base, and give conversion tables for determining all of these factors when any two are known. Thus, in the case of normal sea-water of excess base 24 (vide p. 60) and pH 8.0, the carbon dioxide content is 45.7 c.c. per litre, as measured at N.T.P., and the corresponding tension 0.00043 atmospheres.

Data are available, therefore, for the application of the pH method to the measurement of the gaseous exchanges of animals and plants over two limited ranges—fresh waters, with very low carbonate and bicarbonate content, and seawater, with excess base ranging from 23 to 26. Between these extremes, there lies a wide range of natural waters, littoral and estuarine, supporting a luxuriant and varied fauna and flora, for which no physical data in a form immediately useful to the experimental biologist have been hitherto available.

The investigations of Buch (1917) on the chemical hydrography of the low-salinity waters of the Baltic Sea afford data from which the necessary values may be computed and interpolated. It is important to note that variations in salinity, as such, have an almost negligible effect upon the carbon dioxide tension and content of sea-water. This is because a very small proportion only, always less than 1 per cent., of the total carbon dioxide that can be extracted from acidified sea-water exists naturally in true solution, obeying Henry’s Law. The remainder, over 99 per cent, of the total, is present as bicarbonate and carbonate, and is not affected, except indirectly and to a very slight extent, by the presence of neutral salts. Variations in excess base, however, are of great importance. When normal sea-water is diluted with distilled water, the salinity and excess base decrease pari passu, but under natural conditions none but the roughest general proportionality exists between the two values. It is necessary, therefore, first to determine the excess base in the water used, irrespective of its source or salinity. A final determination is necessary also, in the case of calcareous organisms, whether plant or animal, since they may have altered the total basic content of the water, either by abstraction of calcium carbonate for the building of their skeletons, or by solution of their limy reserves to restore the acid-base equilibrium disturbed by accumulation of carbon dioxide (Collip, 1920). Since, however, the main purpose of the pH method is to secure results in the shortest possible time, it is improbable that either of these factors will have an appreciable influence.

The determination of excess base is best carried out by 59 M’Clendon’s method (1917). To 100 c.c. of the water, contained in a conical resistance-glass flask, a few drops of bromo-cresol purple solution are added, and then N/100 HC1 is run in from a burette until the original purple colour of the indicator has become yellow. The water is then boiled, care being taken, by the use of a sufficiently large vitreosil plate, to prevent access of flame-gases to the flask. The purple colour is soon restored by loss of CO₂ and is again just discharged by addition of a further quantity of N/100 HC1. This is continued, alternately boiling and adding more acid, until finally just enough acid has been added to prevent the return of the purple colour after boiling for a further five minutes. Any diminution in volume, due to evaporation, is made up, from time to time, by addition of boiling freshly-distilled water. Normal sea-water will require about 24 or 25 c.c. of N/100 HC1 per 100 c.c., and this number is used, in the accompanying curves, to denote the “excess base.”

The determination of pH is carried out in the usual way, preferably with the use of some form of comparator in which the tubes are brought side by side, and not separated by a dark or light interspace, as in some forms at present on the market. For details of manipulation necessary to secure readings of high relative accuracy, reference must be made to Atkins (1923) and Saunders (1923, 1). As to the indicators to be used, phenol red is serviceable over the range pH 6.8 to 8.4, and cresol red from/H 7.2 to 8.8. The higher alkalinities attained in algal cultures are covered by thymol blue, which is effective up to pH 9.6. The salt-error of the indicator, to be corrected for when the kation-content of the solution tested is not equal to that of the buffer solution used for comparison, has been accurately determined, in the case of cresol red, by Wells (1920), and of the other indicators named, by Saunders (1923, 1). For details of the correction, reference must be made to the original papers. With the Clark and Lubs buffer-mixtures commonly used, and cresol red as indicator, a sample of sea-water, of full normal salinity, in which the indicator-tint matches that given by the buffer at pH 8.5 is actually at pH 8.3. The correction varies in different parts of the pH range. M’Clendon and co-workers (1917) have prepared buffer Determining Carbon Dioxide Exchanges solutions of salt content equal to that of normal sea-water, thus obviating the correction for salt-error, but for the purpose immediately in view, i.e. the application of the pH method to solutions of widely varying salinity, the raison dêtre of the M’Clendon mixtures disappears.

In this extended application of the pH method, no new experimental technique is suggested. The indicatometric method is equally applicable whether the organisms are suspended in water or the water merely equilibrated with the gases produced by them, and any of the forms of apparatus previously described may be used, with the added advantage that freshwater, brackish-water and marine forms are now equally amenable to the method, and that serial experiments, designed to test the effect of varying salinity upon the metabolic activity of a given organism, may be undertaken simply and effectively. The writer, in the course of investigations upon the gaseous exchanges of dinoflagellates, has used the method in the simplest form—standard suspensions of the organisms being kept in a series of paraffin-stoppered tubes, one or more of which were removed from time to time for pH determination.

The pH method, though simple, rapid and convenient, has an important limitation. It can afford, at best, a measure of the carbon dioxide exchanges, and although it may be assumed, for the roughest general determinations, that the concentration of dissolved oxygen varies in an inverse proportion, yet for accurate work, and whenever the respiratory quotient may differ appreciably from unity, parallel determinations of oxygen content, by Winkler’s or other suitable method, should be made. Finally, the results obtained are expressed in cubic centimetres of CO₂ per litre of solution. This arrangement is dictated by graphical convenience, and by the fact that it readily permits of checking the determinations by gasometric measurements, and of contrasting them, when necessary, with oxygen values. So far as the organism is concerned, however, the tension of carbon dioxide, and not the total amount, is the prime consideration. The gaseous exchanges across respiratory and photosynthetic surfaces are determined by the partial pressures of the reacting gases in the immediately adjacent spaces. Such pressures depend upon concentration, excess base and temperature, and in view of the intended purpose of the curves it was thought undesirable to complicate them by introducing the tension factor.

The curves, as previously mentioned, have been computed from the formulae and observations of three or four workers, using different methods, and it was only to be expected that some discrepancy should exist between their several results. By plotting smoothing curves, however, and by careful interpolation, it has been possible closely to reconcile the existing data, and to produce a series of curves which, while not of a high order of absolute accuracy, are consistent among themselves, and should afford relative results of useful significance. The temperatures at which the determinations were originally made lay between 15° and 20° C. Within these limits, the error introduced by neglect of temperature lies within the range inseparable from determinations on the living organism. To obtain comparable results, however, in an experimental series, it is important to maintain a steady temperature during the period of observation. The disturbing influence of temperature is less, in general, in the more saline waters in virtue of the dissociation of their buffer-reserves.

The curves from excess base 1 to 5 follow Prideaux’s formula (1915) exactly; from 5 to 10 they are slightly influenced by the results of Buch (1917), to whom the values from 10 to 20 are exclusively due; from 20 to 26 M’Clendon’s (1917) data have been followed, with some interpolations, but above pH 8.0 the values deviate somewhat from his straight line approximations. At the higher values of excess base data are available only over a very limited range of pH variation. Extrapolation is undesirable in so critical a region, and it would, in any case, be unwise to extend the curves until it were shown that the higher and lower pH values did not lead to abnormal reactions in organisms accustomed to so limited a range of pH variation as that occurring in the sea.