UNIVERSITY OF CALIFORNIA PUBLICATIONS IN AGRICULTURAL SCIENCES Vol. 4, No. 12, pp. 397-405, 3 figures in text December 22, 1922 THE EFFECT OF REACTION ON THE FIXATION OF NITROGEN BY AZOTOBACTER* BY HARLAN W. JOHNSON and CHARLES B. LIPMAN INTRODUCTORY Azotobacter has been given considerable attention in the literature of soil bacteriology and it is thought that it plays an important part in keeping up the supply of nitrogen in many soils. It has been recognized, almost since its first discovery, that Azoto- bacter is especially sensitive to the reaction of the medium in Avhich it grows. The effect of acidity on the organism is so well established that its absence from a soil has frequently been taken as an indica- tion that the soil needed lime. In fact, Christensen has worked out a method wherein he uses these organisms to determine the lime require- ment of soils. The point has been repeatedly stressed that the acidity of soils must be neutralized if Azotobacter is to fix nitrogen effectively. Since the influence of hydrogen-ion concentration on bacteria has been recognized, the following investigators have reported on its effect on these organisms. Fred and Davenport1 reported that Azotobacter is very sensitive, the limits for its growth being between PH 6.6 and PH 8.4 to 8.8. Gainey2 in a preliminary report states that in 90 soils studied all but 3 of the 37 in which no Azotobacter was found had a PH value of 5.9 or less and all but 3 of those containing the organisms had PH values of 6.0 or greater. In later papers3 he reported that Azotobacter rapidly disappears when inoculated into soils whose PH value is below 6.0 and that, in a study of 382 soils, using the hydrogen electrode for PH measurements, 158 samples with PH values below 6.0 and 20 above 6.0 * This study was undertaken at the suggestion of Dr. C. B. Lipman, in whose laboratory the work was carried out. 398 University of California Publications in Agricultural Sciences [Vol. i contained no Azotobacter, while 165 samples whose PH values were above 6.0 and 39 samples below 6.0 showed Azotobacter. The average nitrogen fixed in 186 samples containing Azotobacter was 7.9 mgs., while the average fixed in 181 samples lacking the organism was 4.6 mgs. Waksman4 in a study of cranberry soils found no Azotobacter in an unlimed soil with PH values of 5.4 to 5.6, but in an adjacent limed soil with a PH value of 6.2-6.4 found it to be present. These investigators agree that slight acidity inhibits the growth of Azotobacter. None of them, however, reports on the effect of the reaction on the nitrogen fixing efficiency of the organism. Gainey reports the nitrogen fixed in the soils containing Azotobacter as com- pared with soils lacking them, but does not state the effect of the reaction in the soils where they were present. The work here presented was undertaken to determine the effect of various hydrogen-ion concentrations on the ability of Azotobacfi r chroococcum to fix nitrogen. EXPEEIMENTAL The organism used was 'a strain of Azotobacter chroococcum isolated from a California soil which in previous work had been found to be very efficient in nitrogen fixation and to produce abundant pigment. The medium was made up as follows: Mannite, 15.0 gms. MgS04 • 7H20, .2 gms. NaCl, .2 gms. CaSO, • 2H.O, .1 gms. H3P04, 1.5 c.c. Distilled water, 1000 c.c. This solution was titrated with N/2 KOH to give definite PH values, using the hydrogen electrode in the titration. The titration curve of the medium is shown in figure 1. The inoculations were made into 100 c.c. portions of the medium in 800 c.c. Erlenmeyer flasks. To secure uniform inoculation, 50 c.c. portions of the medium in a small Erlenmeyer flask were inoculated with Azotobacter, and after a char- acteristic membrane had formed the flask was shaken vigorously, and (he suspension was used as the inoculum. ^» Cn Ph Values <© »o Kt <4 6 S 5 c \ o ft) \ "^1 L- . 0 0 <5> v. Co * *> m o < I 400 University of California Publications in Agricultural Sciences [Vol.4 The First Series In this series the medium was titrated to give PH values of approxi- mately 3, 4, 5, 6, 7, 8, 9, 10, and 11. These values changed during sterilization and incubation. Four flasks of each reaction were inoculated and incubated three weeks at 28° C. Three of the flasks of each reaction were left undisturbed. The others were used for PH measurements, 5 c.c. being withdrawn at intervals from each for electrometric determinations. The PH values of the solutions in these flasks at the time of inoculation and at each of the succeeding periods are shown in figure 2. After incubating twenty days the nitrogen in each of the un- disturbed flasks was determined and the amounts fixed are given in table 1. TABLE 1 Nitrogen Fixed in Solutions op Various Ph Values Series 1 N Fixed Mgs. Ph at time Ph at time Flask Flask Flask of titration of inoculation No. 1 No. 2 No. 3 Averag 3.14 3.47 .07 .0 .0 .02 4.10 4.80 .0 .07 .0 .02 5.06 5.31 .0 .0 .0 .00 5.98 6.07 .70 1.12 .0 .61 7.00 7.08 1.82 2.66 2.52 2.33 8.15 7.69 2.59 1.75 2.45 2.26 9.01 7.98 2.42 2.28 2.42 2.37 10.04 8.08 2.40 3.38 3.52 3.10 11.03 8.86 4.95 4.11 3.41 4.16 It will be noticed that the PH values of the media were not the same when inoculated, as at the time of titration. This may be due to incomplete reaction at the time of titration, although the titration required two days for completion. It might be due to the absorption of carbon dioxide by the more alkaline media, since these media showed the greatest changes and several days intervened between sterilization and inoculation. Probably both factors had their effect. It will be noted that the solutions with PH values of 6.0 and 7.0, whose reactions remained remarkably constant, lie in the region which is shown by 1922] Johnson-Lipman : The Fixation of Nitrogen by Azotobacter 401 the curve in figure 1 to be most highly buffered, while those with PH values between 8.0 and 11.0, which are in a poorly buffered region, show the greatest changes. The growth of Azotobacter in the more alkaline solutions also affects the reaction. The more acid solutions, although in a poorly buffered region, show little change in reaction. Fig. 2. Changes in Ph Values. First Series II -t.16 6- .61 S~ 00 4 — O.Z -3 0.2. i/„ J a. J * S 6 7 fff Jilrsfiui 8 9 ID II 13. 13 if /S 16 17 is /9 zo Days after Inoculation The end point for nitrogen fixation on the acid side of the neutral point evidently lies very close to PH 6.0. At that reaction only a very small amount of nitrogen was fixed and in the solutions with lower PH values no fixation occurred. The alkaline reaction inhibiting fixation was not reached in this series; in fact the largest amount of N was fixed in the most alkaline medium used. 402 University of California Publications in Agricultural Sciences [Vol.4 The Second Series The second series was planned to find more accurately the lower critical PH value and an alkaline reaction inhibiting fixation. Conse- quently solutions were made up with PH values of approximately 5, 6. 6.2, 6.4, 6.6, 6.8, 7, 8, 9, 10, 11, and 12. Four flasks of each PH value were again inoculated and incubated, one of each PH value being used as before for PH determinations. Only three PH determinations were made, viz., at inoculation, ten days, and seventeen days later. It is regretted that no determination was made at the end of incubation. The reactions at the time of inoculation and later are shown in figure 3. Again those in the buffered region between PH 6.0 and 7.0 showed little change, while those above in the less buffered region were markedly changed. The change in the highly buffered solution of PH 12 is remarkable, but must most likely be due to absorption of carbon dioxide since sterilization did not materially change the reaction. It should be noted that an extra flask, uninoculated, changed exactly the same as the inoculated one. These cultures were incubated thirty days, since the amounts of nitrogen fixed in the period of twenty days in the first series were rather small. The amounts of nitrogen fixed in this series are given in table 2. TABLE 2 Nitrogen Fixed in Solutions of Various Ph Values Series 2 N Fixed Mgs. Ph at time Pn at time Flask of titration of inoculation No. 1 5.00 5.34 .0 5.99 5.97 .28 6.20 6.29 5.88 (5.41 6.49 3.36 6.63 6.70 3.64 6.83 6.83 3.78 6.97 6.91 7.50 8.00 7.93 5.46 8.99 8.18 5.46 10.02 8.52 5.60 11.02 9.26 .14 12.05 11.87 .0 Flask Flask No. 2 No. 3 Averaga .0 .28 .09 .84 1.96 1.02 4.70 4.20 4.95 4.62 5.18 4.38 4.90 1.34 4.28 4.90 4.34 4.34 8.12 2.66 7. SI 5.18 4.48 r..04 4.90 5.34 5.23 5.88 .42 T 3.97 8.26* .0 .07 .0 .14 .05 ■iitly some factor hai l affected tlio * Not included in averages since very < results. The amounts of nitrogen fixed in this series were somewhat larger than those in the first series and show a somewhat different effect of reaction. In this series the neutral solution shows the greatest fixation. 1922] Johnson-Lip man: The Fixation of Nitrogen by Asotobaeter 403 It is evident again that PH 6.0 is near the critical limit for fixation. The very great increase in N fixed in the solution about .2 PH higher is remarkable. The lower limit for fixation is apparently very definite. The upper limit for nitrogen fixation was reached in the solution whose PH value at inoculation was 9.26. Fig. 3. Changes in Ph Values. Second Series s a s fff irtoculaTtan flf- hfrafton Days after Inoculation 404 University of California Publications in Agricultural Sciences [Vol. 4 GENERAL DISCUSSION These experiments -were planned to determine the effects of hydrogen-ion concentration on nitrogen fixation by Azotobacter. The results obtained, however, can only be used to show that nitrogen fixation is not seriously affected until the critical limits of reaction are closely approached, when an abrupt decrease occurs in the amount of the nitrogen fixed. The variations in amounts of nitrogen fixed in solutions between these limits are not sufficient, considering the number of cultures used and the variations between cultures of the same reaction, to be of definite significance. The averages given in the tables are simply for convenience, and it is realized that a large number of flasks of each reaction would be necessary in order to secure conclusive evidence of the effects of various reactions. The results show that there is an abrupt decrease in the amount of nitrogen fixed between PH 6.2 and PH 6.0, in other words, that the limiting hydrogen-ion concentration for good nitrogen fixation is a definite value between those two points. This corroborates the results of the previous investigators who found Azotobacter in soils whose PH was 6.0 or above and none in soils with lower PH values. It shows that we can expect no nitrogen fixation by Azotobacter in many of our soils, since it has repeatedly been shown that PH values below 6.0 are frequently encountered. These organisms apparently react much more sensitively than do most of our other soil organisms, since ammoni- fication, nitrification, and other forms of bacterial activity are active in soils whose acidity is higher than PH 6. The alkaline limit for nitrogen fixation is apparently near PH 9.0, since in the first series the solution whose PH value was 8.86 at inocu- Lation showed good fixation, while in the second series the solution inoculated at PH 9-26 showed no nitrogen fixed. It is doubtful whether many soils ever attain such a reaction. Sharp and Hoagland report two soils whose alkalinity is greater, but these were exceptional alkali soils. From this study, Fred's1 limits of PH 6.6 and S.4 to 8.8 would seem to be too narrow. Different strains of Azotobacter, however, may show variations in the effect of reactions on growth, and since this strain was an especially vigorous one its limits might be expected to be wider. 1922] Johnson-Lipman : The Fixation of Nitrogen by Azotobacter 40f SUMMARY A vigorous strain of Azotobacter chroococcum was grown in solutions whose reactions were definitely determined by the hydrogen electrode. The nitrogen fixed in the solutions of each reaction was determined and the changes in reaction during incubation were measured. It was found that the reaction of the solutions below PH 8.0 changed very little, those below PH 6.0 because no growth occurred, and those between 6.0 and 8.0 because the solution in this region was highly buffered. Above PH 8.0 the reaction changed greatly, possibly due to incom- plete reaction of the alkali at the time of titration, but more probably due to absorption of carbon dioxide by the strong alkali. The amount of nitrogen fixed was not greatly affected by reactions between PH values of 6.2 and 8.8 although reactions around PH 7.0 and 8.0 seemed to be most favorable. Slight changes outside of these values caused an abrupt decrease in fixation. LITERATURE CITED i Fred, E. B., and Davenport, A. 1918. Influence of reaction on nitrogen assimilating bacteria. Jour. Agr. Res., vol. 14, pp. 317-336. 2 Gainet, P. L. 1918. Soil reaction and the growth of Azotobacter. Jour. Agr. Res., vol. 14, pp. 265-271. a Gainey, P. L. 1922a. A study of some factors controlling the presence of Azotobacter. 1922b. Correlation between the presence of Azotobacter in a soil and the hydrogen-ion concentration of the soil. Scientific Proceedings of the Society of American Bacteriologists, abstracts 32 and 33, in Abst. of Bact., vol. 6, pp. 14-15. * Waksman, S. A. 1918. The occurrence of Azotobacter in cranberry soils. Science, vol. 48, p. 653.