Technical Bulletin No. 96 September 15, 1942 Hitter? of COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION THE MICROBIOLOGICAL OXIDATION OF AMMONIA IN DESERT SOILS, I. THRESHOLD PH VALUE FOR NITRATIFICATION BY A. B. CASTER, W. P. MARTIN, AND T. F. BUEHRER PUBUSHED BY of Arizona TUCSON, ARIZONA OKGANIZATION BOARB OF KEGENTS Sidney P. Osborn (ex officio), . —.—................Governor of Arizona E. D. King, BA. (ex officio).. State Superintendent of Public Instruction Albert M, Crawford, B.S., President.,, Prescott William H. Westover, U-.B......... Yuma Martin Gentry, LL.B.... Wilicox Cleon T. Knapp, LLJB., Treasurer Tucson Jack B. Martin, Secretary..... Tucson M. O. Best Phoenix Clarence E. Houston, 1X.B., B.A Tucson Mrs. Joseph Madison Greer, B.A , Phoenix Alfred Atkinson, D.Sc President of the University EXPEBIMENT STATION STAFF Paul S. Burgess, PhD Ralph S. Hawkins, PhD. .Director Vice-Director DEPARTMENT OF AGBICULTTOAL CHEMISTRY AND SOILS William T. McGeorge, M.S , Agricultural Chemist James F, Breazeale, B.S ,...., Biochemist Theophil F. Buehrer, Ph,D Physical Chemist Howard V, Smith, M.S......................... .........Associate Agricultural Chemist William P. Martin, PhD Assistant Soil Microbiologist George E, Draper, M.S ...........Assistant Agricultural Chemist (Phoenix) Alfred B. Caster, PhD, „ Assistant Agricultural Chemist TABLE OF CONTENTS PAGE 475 INTRODUCTION LITERATURE REVIEW Ammonia as Compared with Nitrate Nitrogen in Plant Nutrition.... Influence of Soil Alkalinity on the Nitrification Process Toxic Effect of Ammonia on Nitrifying Bacteria PURPOSE AND PLAN OF THIS INVESTIGATION 476 476 477 479 479 EXPERIMENTAL PROCEDURE AND RESULTS 480 Description of Soils 480 Rate of Nitrification of Ammonia and Other Nitrogen Compounds in Desert Soils 483 FURTHER EVIDENCE SUBSTANTIATING THE EXISTENCE OF A THRESHOLD pH VALUE OF 7.7±0.1 FOR THE NITRATIFICATION PROCESS 487 Influence of Initial pH Value on the Rate of Nitrite Accumulation and the Formation of Nitrates 489 Series A—Control; No Treatment 490 Series B—Treated with NH t OH-fCa(OH) 2 490 Series C—NELOH Alone 492 Series D—Treatment with NHiOH+0.5tf H»SOt 493 Series E—Treatment with NHiOH+ltf H2SO, 493 Relationship between pH Value and Nitrate Formation 494 Correlation of Initial pH Value with Nitrite and Nitrate Formation 495 Effect of Maintaining the Soil at a pH Value above the Threshold of 7.7±0.1 496 DISCUSSION 499 Energy Relations Involved in the Hydroxylamine-Hyponitrous Acid Mechanism in the Nitrification of Ammonia 502 SUMMARY 505 LITERATURE CITED., 507 ILLUSTRATIONS PAGE FIGURE 1.—RELATIVE RATES OF NITRIFICATION IN LAVEEN LOAM UNDER DIFFERENT TREATMENTS FIGURE 2.—RELATIVE CHANGES IN PH VALUE- AND NITRATE CONTENT OF VIRGIN GILA SANDY LOAM TREATED WITH 300 P.P.M. OF NITROGEN AS AMMONIA FIGURE 3.—RELATIVE CHANGES IN pH VALUE AND NITRATE CONTENT OF CULTIVATED GILA SANDY LOAM TREATED WITH 300 P.P.M. OF NITROGEN AS AMMONIA , FIGURE 4.—CORRELATION BETWEEN pH VALUE AND NITRATE FORMATION IN GILA SANDY LOAM UNDER DIFFERENT TREATMENTS FIGURE 5.—THRESHOLD pH RANGE IN THE NITRIFICATION OF AMMONIA IN GILA SANDY LOAM FIGURE 6.—RATE OF NITRIFICATION OF AMMONIA IN GILA SANDY LOAM AS INFLUENCED BY CONTROLLED ALKALINITY 485 487 489 494 495 498 THE MICROBIOLOGICAL OXIDATION OF AMMONIA IN DESERT SOILS, I. THRESHOLD PH VALUE FOR NITRATIFICATION* BY A. B. CASTER,! W. P. MARTIN, AND T. F. BUEHRER INTRODUCTION In recent years ammonia gas dissolved in irrigation water has become agriculturally important as a source of nitrogen in the fertilization of soils, particularly those of irrigated regions. Gaseous ammonia, as distinguished from the common nitrogen fertilizers, has several noteworthy advantages: (1) an unusually high percentage of nitrogen—the highest, in fact, of all nitrogen fertilizers; (2) the fact that by being wholly utilized by the plant it leaves no salt residue to contribute to the salinity of the soil; (3) the ease with which it can be applied to the soil even after the crop is too high to permit the application of granular fertilizers by the conventional methods; and (4) the economy with which the amount applied and its distribution can be controlled. In addition, the loss of ammonia by volatilization when so applied is remarkably slight on account of its prompt fixation by base exchange. Superficially it may seem anomalous to apply a highly alkaline ammonia solution as a fertilizer to an already alkaline soil, and the established practice has generally been to use a nitrogen compound having an acid or neutral reaction. Nevertheless, on many soils where ammonia has been employed, it has produced increases in yield of magnitudes such that the anticipated undesirable effects of high alkalinity were entirely overcome by the rapidity with which the ammonia was oxidized. Since the rate of this oxidation process is without doubt a factor affecting the crop response which may be expected as a result of ammonia fertilization, it was deemed of interest to investigate the interrelationships of the physical, chemical, and microbiological factors which influence it. The authors are well aware that the mechanism of ammonia oxidation in the soil has for many years engaged the attention of investigators, and that its quantitative relationships are fairly well known. In the present instance, however, it is evident that there may be conditions that are unfavorable to the process, either (1) by slowing down certain steps, (2) by allowing intermediate products toxic to the nitrifying bacteria to accumulate, or (3) by inhibiting nitrification entirely. The extent to which *The authors desire to express their appreciation to the Shell Chemical Company of San Francisco, California, for sponsoring the fellowship under which this investigation was carried out and for financing in part its publication. tShell Fellow, 1939-41. 475 476 TECHNICAL BULLETIN NO. 96 these effects manifest themselves in any given soil is evidently a function of various factors, notably its alkalinity. This bulletin sets forth the results of experiments to determine to what extent the alkalinity affects the process, and whether or not there is a threshold pH value above which the conversion of ammonia to nitrite, or of nitrite to nitrate, will not occur regardless of how favorable the other conditions may be, LITERATURE REVIEW Inasmuch as the use of free ammonia as a fertilizer is a comparatively recent development in agricultural practice, the amount of the published research on its properties and behavior is necessarily limited in extent. Furthermore it is not possible to deduce with any degree of certainty the probable behavior of ammonia when so used, from the results of investigations on the various ammonium salts or of the organic nitrogen compounds commonly applied as fertilizers. On the other hand, the background of existing research on the varied factors affecting the microbiological oxidation of nitrogen in both the organic and inorganic forms and the utilization of those forms by the plant has yielded valuable suggestions as to an advantageous approach to the present problem. AMMONIA AS COMPARED WITH NITRATE NITROGEN IN PLANT NUTRITION The ability of plants throughout their entire growth period to utilize ammonia per se has not yet been conclusively proved. Physiological differences in plants with respect to the intake of other nutrient elements, the development of their root systems, and fruiting characteristics have been observed repeatedly where plants were grown in nutrient solutions or sand cultures in which either the nitrate or ammonium ion was present as the sole source of nitrogen. Some investigators, notably Prianishnikov (46), claim to have established the fact that plants under certain conditions absorb ammonia in preference to nitrate. Studies on ammonium-ion absorption have, however, generally involved the use of neutral ammonium salts in nutrient solutions, in sand cultures or in waterlogged soils—all of which represent conditions which do not obtain in normal, arable soils. Shive and his co-workers (2, 14, 15, 20, 48, 49, 50), Jones and Skinner (29), Stewart et al. (52), McGeorge (34), and Naftel (41) have made substantial contributions to our knowledge of the relative absorption and assimilation of ammonium ion and nitrate by plants. Considered collectively, their results indicate that the ion absorbed and assimilated is a function of the age and species of the plant and/or the reaction of the external medium. The preponderance of evidence, however, indicates that nitrates are essential to the plant at some stage of its development and growth. As noted recently by Arnon (1), the possibility of nitrification of ammonium ion in any study which aims to compare its relative absorption with that of nitrate is a complicating THRESHOLD pH VALUE FOR NITRATIFICATION 477 factor which renders the use of, or even the presence of, soil undesirable. On the other hand, since the conditions are not comparable, it is poor logic to assume that the results of nutrient solution studies will apply directly to the soil. From an experimental point of view it would seem preferable to make a study of the state or condition of the nitrogen in the soil at various intervals of time after its application—that is, to determine its nitrification rate—when subjected to conditions such as those which actually prevail in the soil. Although numerous compounds have been studied as a source of nitrogen in fertilizer programs, practically no work bearing either on field or laboratory studies with gaseous ammonia has appeared in the literature. Ammonium sulfate has been extensively used as a standard for the comparison of different nitrogen compounds in nitrification studies, but its residually acid nature makes it difficult to apply the results so obtained in the prognosis of the probable behavior of ammonia when applied to a soil. The majority of the published papers concern themselves with nitrification studies on humid soils which are dominantly acid and may possess a microflora in general different from that of the dominantly alkaline soils of the desert. INFLUENCE OF SOIL ALKALINITY ON THE NITRIFICATION PROCESS Because of the alkalinity produced when ammonia is dissolved in water, the pH factor so introduced may influence not only the rate at which the ammonia is nitrified but also certain physical properties of the soil. When ammonia is added in an amount equivalent to 300 p.p.m. of nitrogen, it is sufficient not only to neutralize the carbon dioxide and bicarbonate ion which the irrigation water may have contained but also to leave a certain amount of free alkalinity. The effect of the added base on the pH value of the soil will of course be determined by the extent to which the latter is buffered against base, which in alkaline soils is usually slight. In this manner an amount of free alkalinity is added to the soil which, though temporary in duration, may nevertheless have important effects on the microbial population of the soil, particularly the nitrifying organisms. Hence their ability to remain active over the range of pH values which may be produced under such conditions may be in considerable measure reduced. These are some of the fundamental aspects of the problem which received particular consideration in the course of the present study. That nitrification proceeds more efficiently under neutral or slightly alkaline conditions has been noted by various investigators in this field, but no study appears to have been made on the optimum or limiting pH values for nitrification in alkaline calcareous soils. Fraps and Sterges (21) found that soils which failed to nitrify ammonium sulfate could be made to do so by the addition of cultures from actively nitrifying soils and/or of 478 TECHNICAL BULLETIN NO. 96 calcium carbonate. These authors (22), furthermore, found that in acid soils calcium carbonate favored rapid nitrification more so than did dicalcium phosphate, rock phosphate, magnesium carbonate, or dolomite. Similar studies by Tandon and Dhar (53) showed that nitrification is favored more by calcium carbonate than by magnesium carbonate. Waksman (54) in his nitrification studies stressed the necessity of neutralizing the acid formed in the oxidation of ammonium sulfate and indicated that an acid condition was harmful to the process. He observed, for example, that the accumulation of nitrates stopped when a pH value in the range of 4.4 to 4.8 was reached and concluded that nitrate accumulation in any soil depends upon its initial pH value, its buffer capacity, and the final pH value of the soil, "more so than on the bacteriological activity." The minimum pH value reported by Waksman is essentially in agreement with that found by Humf'eld and Erdman (28) and by Gaarder and Hagem (23); the latter, having worked with acid solutions, placed the minimum pH range below which the nitrifiers no longer function between 3.9 and 4.5. Naftel (42) investigated the rate of nitrification in five widely different soils as a function of pH value and base saturation and found, in general, that the rate increased with base and calcium saturation. His results indicated that the extent of nitrification in different soils may vary widely even though their pH values may be the same. This may be due in part to the degree of saturation with bases, since only those soils which had the highest percentage of their base exchange complex saturated with calcium nitrified ammonium sulfate to any considerable extent. That the nitrifying bacteria are capable of adjusting themselves to a fairly wide variation in pH value was found by Olsen (44) working with strongly acid humus soils in which the pH value had been adjusted and maintained at the desired levels with lime. He concluded that nitrification can take place between pH 3.7 and 8.8, the optimum being 8.3, so long as ammonia is not the limiting factor. Meyerhof (37, 38, 39), Gowda (25), and Meek and Lipman (36) have reported results which pertain more specifically to nitrification in the alkaline range. In a detailed study of both types of bacteria, Meyerhof found that ammonia was oxidized most quickly between pH 8.5 and 8.8. The optimum pH for nitrite oxidation was between 8.3 and 9.3. These values are in substantial agreement with those obtained by Gowda. Meek and Lipman, on the other hand, found that both nitrite- and nitrate-forming organisms from garden soil were alive and functioning at pH 13.0, but the nitrifiers from acid peat were unable to produce nitrates above a pH value of 9.5. They concluded that the nitrate-forming bacteria are somewhat more resistant to alkalinity than are the nitrite formers. It is to be noted, however, that the above investigators worked exclusively with nutrient solutions and not with soil. THRESHOLD pH VALUE FOR NITRATIFICATION 479 Ayers and Jenny (4) and Waynick (56) studied the nitrification of ammonia and ammonium sulf ate in relation to the physical and chemical changes resulting from the application of these fertilizers. The former used one acid and one slightly alkaline soil in their studies. Waynick worked with a series of soils ranging from 7.1 to 8.1 in their initial pH values, but since his determinations were made on the soil suspension, his pH values are obviously higher than those obtained on the same soil at field moisture contents (35). Nelson (43) found that the toxicity of manganese was reduced by additions of lime to the soil and concluded that the shift in reaction so produced rendered the conditions more favorable for the action of the nitrifying organisms. TOXIC EFFECT OF AMMONIA ON NITRIFYING BACTERIA There is some evidence to indicate that free ammonia may have a toxic or inhibiting effect on the nitrifying organisms. Willis and Piland (57) found that the free ammonia formed from the hydrolysis of diammonium phosphate decreased the rate of nitrification when this salt was used in pot culture experiments. Such toxicity, strangely enough, was not observed in the case of ammonium sulf ate, chloride, or nitrate; nor did the alkalinity of the diammonium phosphate appear to be responsible for the injurious effect. They found further that calcium salts were able to counteract the toxicity of the ammonia. It is therefore reasonable to expect that such toxicity would not be likely to occur in calcareous soils, notwithstanding the alkalinity produced when free ammonia is added as the source of nitrogen. Similarly Waksman (54) reported that sufficient free ammonia is evolved in the rapid decomposition of dried blood in alkaline and poorly buffered soils to have an injurious effect upon the activity of the nitrifying bacteria in the soil. He did not, however, mention the antagonistic effect of calcium as reported by Willis and Piland, and since his report referred specifically to alkaline soils, it is felt by the present writers that further studies should be conducted on this phase of the problem to clarify the apparent anomaly. PURPOSE AND PLAN OF THIS INVESTIGATION In view of the increasing interest in the use of gaseous ammonia as a fertilizer for citrus, truck crops, small grains, and hay which are extensively grown under irrigation on the alkaline calcareous soils of the desert, it seemed desirable to study some of the fundamental aspects of its behavior when applied to such soils. It is recognized that the rate and mechanism of ammonia nitrification in such soils may be a function of the alkalinity, salinity, soil type, and dominant microflora, and perhaps other factors which influence it to a greater or lesser degree. The results to be presented in this bulletin concern themselves primarily with the pH relationships which are involved in the nitrification mechanism. 480 TECHNICAL BULLETIN NO. 96 The investigation was planned as a laboratory incubation study under controlled conditions so as to afford a comparison between the rate of microbial oxidation of ammonia with that of other types of nitrogen fertilizers commonly used—namely, an inorganic ammonium salt, ammonium sulfate, and an organic nitrogen compound, urea. Six typical but widely separated desert soils, having textures varying from sands to clays, were chosen. Three of these were in the virgin state and three were from areas under cultivation. All of them except one were calcareous and decidedly alkaline. EXPERIMENTAL PROCEDURE AND RESULTS DESCRIPTION OF SOILS The soils used in this study were as follows: 1. Superstition sand, from near the University Farm on the Yuma mesa; virgin, calcareous. 2. Gila sandy loam, from a ranch adjacent to the Cortaro Farms at Marana, Arizona; virgin, calcareous. 3. Gila sandy loam, same general locality as (2) but from a field in cotton on Cortaro Farms at Marana, Arizona; under cultivation, calcareous. 4. Pima clay loam, from a farm in the Gila River bottom near Safford, Arizona; under cultivation, mildly calcareous. 5. Laveen loam, from the University of Arizona Farm at Mesa, Arizona; under cultivation, calcareous. 6. Palos Verdes sandy loam, from north of the Rancho Palos Verdes near Tucson, Arizona; virgin, noncalcareous. The samples were taken within a period of about a month during the summer of 1939. After gentle rolling to break up the aggregates, and removal of foreign matter, the soils were screened to 10-mesh, thoroughly mixed, and stored. Prior to the nitrification studies, it was deemed of value to make determinations of certain chemical constituents and physical properties of these soils which might have some bearing on the nitrification process. The following were accordingly determined by the accepted methods: pH values by the Beckman pH meter; soluble salts by conductivity of the 1:5 aqueous extract; total carbonate by the official gasometric method of the A.O.A.C. (3); nitrate by phenoldisulfonic acid; organic carbon by wet oxidation with chromic acid; and total and specific buffer capacities on the 1:5 suspension according to the method of Pierre (45). The data are assembled in Table 1. The pH values were determined both on the 1:5 suspension and on the soil at a moisture content of 70 per cent of its water-holding capacity. The data given in the table illustrate the frequently observed fact that the pH value of a soil at field moisture content is considerably lower than that of the same soil in the 1:5 suspension, and in most of the above soils the difference amounted to as much as 1.1 pH unit. None of these soils at field moisture con- 482 TECHNICAL BULLETIN NO. 96 tent can be said to be excessively alkaline, being, in fact, fairly close to neutrality, and one of them, the Palos Verdes sandy loam, actually has a value on the acid side of neutrality. With the exception of the Pima and Laveen loams, the soluble salt contents of these soils are relatively low, of a magnitude which is not great enough to affect seriously the activity of the nitrifying organisms. The high salt content of the two soils referred to accounts in part for their relatively lower pH values. The data for total organic carbon are consistent with the typically low percentage of organic matter in desert soils. All of the soils except the Palos Verdes sandy loam contained appreciable amounts of carbonate, particularly the Laveen loam, to which the latter soil owes its unusually high buffer capacity. The specific buffer capacity data afford a significant indication of the stability of these soils toward a change in pH. When the specific buffer capacity is plotted against the total carbonate content, the points fall on a straight line within limits of experimental error, thus indicating that calcium carbonate is the principal buffering compound present. For this reason and also because the soils on the alkaline side of neutrality are largely saturated with bases, they are buffered primarily toward acid. Being so slightly buffered toward base, as the buffer determinations have shown, it is not surprising that a considerable rise in pH is realized when solutions of ammonia or of other free bases are added to the soil. In the case of the Palos Verdes sandy loam, these conditions are reversed. Since buffer capacity determinations are usually made at a constant dilution of 1:5, it is conceivable that a soil may exhibit a considerably higher buffer capacity at field moisture content. This was found to be the case in later experiments when equal volumes of normal and half-normal sulfuric acid solution were added to samples of Gila sandy loam at field moisture content, with no significant difference in pH value. These buffer considerations have an important bearing upon how a given soil will react when treated with ammoniated irrigation water and the extent to which such a soil will resist any considerable decrease in pH value upon its subsequent nitrification. Some of the soils contained appreciable amounts of nitrate and organic matter, particularly those which have been under cultivation for some time. This is illustrated in the case of the Pima clay loam, one of the most fertile agricultural soils of Arizona, whose fertility can be accounted for by the fact that the farm from which the sample came had been systematically fertilized with manure at the rate of 12 tons per acre for the past 10 years. The foregoing determinations are presented to emphasize the fact that the soils chosen for the nitrification studies to be set forth in this bulletin represent a fair cross section of the agricultural areas where ammonia is already being used, or may be used, as a nitrogen fertilizer—a circumstance which must be borne in THRESHOLD pH VALUE FOR NITRATIFICATION 483 mind in making generalizations concerning the microbiological transformations it may undergo under field conditions. RATE OF NITRIFICATION OF AMMONIA AND OTHER NITROGEN COMPOUNDS IN DESERT SOILS The first series of experiments in this investigation was designed to ascertain not only the general nature of the nitrification curve for ammonia, ammonium sulfate, and urea over an extended period of incubation but also the nature of the process insofar as it may be affected by the character of the soil. Replicate 100-gram samples of soil were weighed into glass tumblers with close-fitting covers. The samples were divided into four equal groups and treated with the following fertilizer solutions: Group 1: control, untreated; Group 2: 30 mg. of nitrogen as ammonia; Group 3: 30 mg. of nitrogen as ammonium sulfate; Group 4: 30 mg. of nitrogen as urea. On this basis the nitrogen was present in each sample to an extent of 300 p.p.m. of air-dry soil. Since ammonia is fixed very promptly by the soil, there is no perceptible loss when so applied. By keeping the ammonia solution in a closed delivery system, its concentration could be maintained constant. Following application of the above solutions distilled water was added to bring the moisture content of each sample to 70 per cent of the water-holding capacity of the particular soil. The tumblers were then weighed and placed in an incubator maintained at 30 degrees C. The moisture lost during the incubation was periodically restored by bringing the samples back to their original weights by the addition of distilled water. After the desired incubation intervals, duplicate samples from each series were removed from the incubator and the pH values determined on the moist soil. The entire 100-gram sample of soil was then transferred to a large wide-mouthed bottle, 500 ml. of distilled water added, and the resulting 1:5 suspension shaken for 20 minutes to bring the nitrites and nitrates, formed during the incubation, into solution. Carbon dioxide was then bubbled through the suspension for a few minutes to coagulate the colloids and facilitate the filtration. The clear filtrates were then analyzed for nitrates by the standard methods, the final colorimetric determinations being made with a Cenco Photelometer. The data are reported in terms of parts per million of nitrogen as nitrate on the basis of the air-dry soil. Throughout this investigation the analyses were made on duplicate samples of each treatment of each soil. This procedure eliminates possible systematic errors resulting from the use of aliquot portions of only one soil sample and, in addition, the error arising from variation in the soil from sample to sample. Thus each analytical result, as reported in the tables and graphs to follow, is an average of determinations on two independent soil samples and hence proportionately more significant and trustworthy. By the same token the curves, although continuous, do not represent a continuous change in any 484 TECHNICAL BULLETIN NO. 96 single sample but rather the changes in similarly treated, replicate samples. The results presented in Table 2 give the successive amounts of nitrate observed in the three respective treatments on the six soils over a period of about 100 days. The results for Laveen loam are shown in Figure 1. Because of limitations of space and similarity in the general character of the nitrification curves, it was deemed desirable to limit the graphs to one of the soils in which the trends of the nitrification process were fairly well exemplified. It will be noted that the curves representing the rate of nitrification are typically exponential and hence are similar to the growth rate curves for the nitrifying bacteria as found by numerous investigators (12). In the cultivated Gila sandy loam, Laveen loam, and Pima clay loam, the rates of nitrification of ammonia, ammonium sulfate, and urea were practically identical with the exception that in the cultivated Gila sandy loam the ammonia was oxidized somewhat more slowly than either of the other two nitrogen compounds. It should be noted incidentally that the corresponding curve for the control sample exhibits only a very slight increase in nitrate during the first several days of incubation and thereafter remains horizontal over the entire period of incubation. This slight increase in nitrate probably resulted from the nitrification of the organic nitrogen originally present in the soil. The remaining three soils—namely, the Superstition sand, virgin Gila sandy loam, and Palos Verdes sandy loam—differed in their rates of nitrification from those discussed above. Referring to Table 2, it may be noted that ammonium sulfate was nitrified less rapidly than the urea or ammonia in the Palos Verdes sandy loam, which is diametrically opposite to the behavior observed in the other soils. This behavior is very probably due to the fact that during the oxidation of the ammonium sulfate the pH value of the soil is lowered sufficiently to inhibit nitrification. At the end of 40 days' incubation and with only one third of the ammonia nitrogen oxidized, the pH value had already dropped to 4.65, which incidentally is in close agreement with that of Waksman (54) and Humfeld and Erdman (28)—namely, 4.4—as the "limiting reaction for nitrification on the acid side of neutrality. In the Superstition sand (Table 2), nitrates did not begin to accumulate for nearly 30 days, regardless of the nature of the nitrogen fertilizer added, and even then such accumulation occurred only in the case of ammonium sulfate and urea. Not even a trace of the ammonia was found to have been oxidized during the entire 107 days of incubation. That the rate of nitrification was rather slow in the case of Superstition sand is not surprising considering its inherently low native fertility. It is most surprising, however, that none of the ammonia had been oxidized, whereas the urea and ammonium sulfate were readily nitrified, The data in Table 2 show that the ammonia solution had raised THRESHOLD pH VALUE FOR NITRATIFICATION 485 360 320 LATEEN IQAU 280 — Kd Treatment 240 300 p.p.m.(N) as NH4OH — 300 p.p.m.(N) as (NHA)PSOA 300 p.p.m.(N) as Urea 200 120, 80 40 days 30 40 INCUBATION TIME Figure 1.—Relative rates of nitrification in Laveen loam under different treatments. 10 20 the initial pH value of this soil to 9.5, and that the alkalinity had, during the incubation period, decreased to a value equivalent to pH 8.35 corresponding to the urea but higher than that of the ammonium sulfate—namely, 8.21—which existed in the soil prior to the beginning of the incubation. This observation, in addition to the anomalous behavior of ammonia in virgin Gila sandy loam, 486 TECHNICAL BULLETIN NO. 96 next cited, was the first definite indication that high alkalinity may limit, or even prevent, the oxidation of ammonia to nitrate in calcareous desert soils. In the ammonia-treated samples of virgin Gila sandy loam, no nitrates were observed until the forty-ninth day and then only in one of the duplicate samples. Since duplicates had in all of the previous instances been in good agreement, it was at first thought that this relatively enormous discrepancy was due to an error which had crept into the technique employed in the nitrate determination on one of the samples. Particular care was therefore taken with the corresponding samples taken off on the fifty-third day, but again a considerable difference in nitrate content between the duplicate samples was noted. Throughout the remainder of the incubation, the behavior of the samples was unusual; on the fifty-ninth day nitrates were entirely absent in both samples; on the seventy-second they were present in one but not the other; on the eighty-sixth nitrates were present in considerable amount in both samples; and on the one hundred and seventh the amount of nitrate present was again of the same order of magnitude as in the control. This peculiar behavior was difficult to account for, except insofar as it might be related to the pH values of the samples in question. The data of Table 2, as well as the curves in Figure 2, show, in fact, that there is a close correlation between the pH value and the presence or absence of nitrates. It will be noted from Figure 3 that on the forty-ninth day, When nitrates first appeared in one of the samples, the pH value of this sample was 7.9. In all other instances the failure of the ammonia to be nitrified correlated closely with a pH value of 7.7 it 0.1. Similarly, all of the samples which contained nitrates exhibited a pH value below this figure. As a consequence of this striking coincidence, the data for all of the soils tested were re-examined, with the result that in only one instance out of 133 in which nitrification was observed to have taken place did nitrates make their appearance at a pH value above 7.7 ±. 0.1. In view of this finding and of the supporting evidence which is to follow, the authors feel justified in concluding that: There appears to exist in alkaline desert soils a threshold pH value of 7.7 zb 0.13 above which the complete oxidation of ammonia to nitrate will not occur, and to which the pH value of such soils must first be reduced before nitratification will take place. The foregoing conclusion is of unusual interest and significance, particularly in view of the results of Meek and Lipman (36) who placed the optimum value at pH 8.5 to 8.8; Olsen (44) at 8.3, and Meyerhof (37, 38, 39) between 8.3 and 9.2. The significance of such a value in connection with the microbiological oxidation of ammonia lies in the fact that in desert soils their inherent alkalinity must in some manner be reduced, and that the organisms are virtually ineffective, so far as the primary nitrification process is THRESHOLD PH VALUE FOR NITRATIFICATION 487 300 9.0 240 8.0 180 7.0 GILA, SAKDT LOAM (VIRGIN) 60 5.0 'NITRATE I 20 40 60 80 100 days INCUBATION TIMS Figure 2.—Relative changes in pH value and nitrate content of virgin Gila sandy loam treated with 300 p.p.m. of nitrogen as ammonia. concerned, until the pH value of the soil has been reduced to the threshold value. The foregoing generalization regarding the existence of a threshold pH value for nitratification suggested certain confirmatory experiments, the results of which will now be presented. FURTHER EVIDENCE SUBSTANTIATING THE EXISTENCE OF A THRESHOLD pH VALUE OF 7.7+0.1 FOR THE NITRATIFICATION PROCESS In the previous sections of this bulletin the process of nitrification has been referred to in its broad sense, including all of the intermediate stages through which nitrogen may pass in the course of its oxidation from ammonia to nitrate. In discussing 488 TECHNICAL BULLETIN NO. 96 the subsequent phases of this investigation, it will be advantageous to divide the process into two main steps: the first, which will hereafter be referred to as nitritification, includes all of the possible intermediate steps in the oxidation of nitrogen from any of the lower forms in which it may occur to nitrite; and the second, to be termed nitratification, will signify the last stage of the process in which the nitrite is oxidized to nitrate. From the drop in pH value of the soil prior to nitrification, as is evident from the data in Table 2, it follows that the nitrifying bacteria native to such soils do not function effectively until a threshold pH value of 7.7 has been reached. This fact is unique in view of the prevailing opinion that a decrease in pH occurs concurrently with the oxidation of the nitrogen to nitric acid. In the present study the pH drop in the incubated samples was found to occur prior to the detection of nitrates. It is self-evident that the formation of sulfuric and nitric acid must, or should, result in reducing to some extent the alkalinity of the soil. If, therefore, a pronounced decrease in pH value occurred before nitrates began to be formed, it seems reasonable to assume (in the case of the urea, and ammonia-treated samples in which no sulfate radical is present) that the reduction in pH must have come about through the formation of compounds other than nitric or sulfuric acids. The over-all oxidation of ammonia to nitrite is represented by the following equation: 2 NH3 + 3 O2 + 2 OH" = 2 N(V + 4 H2O in which the reduction in pH value results from the fact that hydroxyl ions are used up in the reaction. It is possible, of course, for some intermediate nitrogen compound other than nitrite to be formed. Assuming that the foregoing reaction actually accounted for the observed changes in pH, it seemed likely that the nitrite ion might be detectable in the soil in appreciable amount during the course of the oxidation, especially in those samples in which a considerable difference existed between the initial pH value of the soil and the threshold value. Under normal conditions and in arable soils, however, nitrites are considered to be too transitory to accumulate in significant amounts. It was considered of great interest, therefore, to determine whether or not nitrite may, if formed under the alkaline conditions of desert soils, actually accumulate in measurable amounts, and if so, whether the reduction in pH value observed could be explained in terms of such accumulation. Experiments were accordingly planned embodying the following phases of the problem: (1) the influence of the initial pH value of the soil upon the rate of nitrite accumulation and the formation of nitrates; and (2) the effect of maintaining the pH value at a level above the threshold value by the addition of a base, such as calcium hydroxide. In order to permit a greater variation of other factors, this phase of the study was confined to the cultivated Gila THRESHOLD pH VALUE FOR NIT RAT IF 1C AT ION 489 sandy loam, a soil of relatively low salt content, medium buffer capacity, and especially one which in previous experiments had been found to permit the rapid oxidation of ammonia, as shown in Table 2 and Figure 3. I 30<P T NITRATE 9.0 240 '.0 180 7.0 6.Ot GIIA SMDY LOAM (CULTIVATKD) 60 INCUBATION TIME 5.0 100 days Figure 3.—Relative changes in pH value and nitrate content of cultivated Gila sandy loam treated with 300 p.p.m. of nitrogen as ammonia. INFLUENCE OF INITIAL PH VALUE ON THE RATE OF NITRITE ACCUMULATION AND THE FORMATION OF NITRATES As in previous experiments, 100-gram samples of soil were weighed into glass tumblers with close-fitting covers. These were then divided into five groups and subjected to the following treatments: (A) control—no treatment; (B) 10 ml. of ammonia solution containing 30 mg, of nitrogen plus 10 ml. of saturated calcium hydroxide; (C) 10 ml. of the same ammonia solution plus 10 ml. distilled water; (D) 10 ml. of the same ammonia solution plus 10 490 TECHNICAL BULLETIN NO, 96 ml. 0.5N sulfuric acid; and (E) 10 ml. of the same ammonia solution plus 10 ml. of 12V sulfuric acid. These treatments served to fix the initial pH values at different levels with respect to the threshold value. The volumes of solution, totaling 20 ml. in each case were so chosen that the resulting moisture content of the samples was 70 per cent of the water-holding capacity of this soil. In order to insure a rapid and approximately uniform distribution of the added liquids throughout the soil mass, each sample of soil was transferred from the tumbler to a clean sheet of paper and divided into approximate fourths. Five milliliters of the solution to be used were then put into the empty tumbler and one fourth portion of the soil sample added to it. A second 5 ml. portion of solution was added, then a second quarter of the soil, and so on, until all of the solution and soil for each tumbler had been added. In this manner the initial pH value of the soil was quickly established, and, more important still, an even distribution of the ammonia was achieved throughout the soil mass. The samples were again incubated at 30 degrees C., and the moisture lost by evaporation was restored at intervals. Periodically, duplicate samples from each treatment were removed and analyzed for nitrite and nitrate, and the pH values on the moist soil were determined. The results obtained in this study are presented in Table 3 as well as in Figures 4 and 5, in which p.p.m. of nitrate formed, as well as the pH values, are plotted as ordinates against incubation (9.5 (A) (B) NHiOH PLUS Ca(QH)g (C) NH4OH ALOUE i£THKESH01I> pH 7.5 7.0 6.5 10 15 INCUBATION TIKE 25 days Figure 4.—Correlation between pH value and nitrate formation in Gila sandy loam under different treatments. 492 TECHNICAL BULLETIN NO. 96 time as abscissae. The purpose of this study was considered achieved as soon as the pH value had dropped to a point sufficiently low so that nitrification could proceed rapidly. For this reason the data presented in the afore-mentioned figures were made to extend through the twenty-fifth day only. A* final analysis was made, however, on the fifty-eighth day to determine the maximum amount of nitrate that had formed. This determination was made primarily to compare the results of the C treatment with the data obtained in the previous experiments on the rate of nitrification of ammonia, which had been carried out on the same soil and under identical conditions of incubation. The behavior of the various series of treatments will now be discussed. Series A—Control; no treatment Eeferring to Figure 4, it will be noted that the pH values and nitrate contents of the samples of the A series remained practically unchanged over a period of 25 days. The initial pH value was at the threshold level of 7.7, and in the subsequent samples it increased to as high as 7.9. Nitrates were present to the extent of only 8 p.p.m. Nitrites were found to be present only as mere traces throughout the incubation period. Series B—Treated with NH,OH+Ca(OH)2 This series of samples, as a result of the addition of calcium hydroxide, exhibited the highest initial pH value—namely 9.21. It will be noted from Table 3 that the pH value gradually decreased during the first 8 days of incubation to about 8.4, remaining practically constant at that value up to the seventeenth day. Thereafter a sudden drop in pH occurred, accompanied by the appearance of nitrite, and decreased gradually to 7.60 at the end of the experiment. The concentration of nitrite in the soil samples continued to increase, however, reaching a maximum of 71 p.p.m. on the twenty-first day just as the pH dropped to the threshold value. Not until the twenty-fifth day, while the pH value was still at the threshold level, did nitrates make their first appearance and nitrites begin to disappear. Series C—NH^OH alone The data of this series are of particular interest since they confirm the results obtained under the same experimental conditions as in the nitrification rate study already reported in Table 2 and Figure 3. In the present instance the samples had a relatively high initial pH value—namely, 9.0—and the decrease during the first 11 days of incubation was approximately 0.6 pH unit. Up to the fourteenth day, as shown by the data in Table 3, no additional nitrate had been formed, but a definite quantity of nitrite (22 p.p.m.) had made its appearance. As the build-up of nitrites continued, the pH value dropped somewhat, and eventually fell to a point below the threshold value by the time the maximum amount of nitrite— THRESHOLD PH VALUE FOR NITRATIFICATION 493 namely, 94 p.p.m.—had accumulated. This relatively high maximum for nitrite was reached on the twenty-first day of incubation, but thereafter the amount decreased sharply, so that by the twenty-fifth day only a trace of nitrite was detectable in the samples. Almost immediately after the pH value had dropped below the threshold value (which occurred on the seventeenth day), nitrates began to form and thereafter accumulate. From the above data it is evident that nitrites must have been present in the soil for nearly 2 weeks, a fact which is rather surprising, in view of the generally accepted fact that nitrites do not exist as such to any appreciable extent in well-aerated soils, at normal moisture contents, being oxidized to nitrate almost as soon as they are formed. It is of interest to compare the nitrate data for the C series of samples with those obtained in the initial nitrification experiment illustrated in Table 2 and Figure 3. It will be recalled that the same Gila sandy loam (cultivated) had been fertilized with ammonia to exactly the same extent—namely, 30 rng. of nitrogen— and incubated under identical conditions. Referring to Table 3, it will be noted that the first appreciable increase in nitrate content occurred on the nineteenth day of incubation, and on the twentyfifth day nitrates were found to be present to an extent of 190 p.p.m., which represented a nitrogen recovery of about 63 per cent. In the previous experiment shown in Figure 3, the first evidence of nitratification was noted on the eighteenth day of incubation, and on the twenty-fifth day a maximum of 193 p.p.m. of nitrates was observed. Here the recovery was about 64 per cent of the total nitrogen applied as ammonia. Inasmuch as 97 per cent of the nitrogen applied in the first series had been recovered at the end of 7 weeks, it was decided to allow a final set of the samples to incubate for approximately the same period—namely, 58 days—to observe how closely the final values recorded in Table 3 would be confirmed. It was found, in fact, that in the C series 99 per cent of the nitrogen initially added as ammonia had been oxidized to nitrate by the end of this period. This phenomenal agreement between two entirely independent incubation studies indicates how reproducible the results of such studies on nitrogen transformations can be and constitutes a confirmation of the threshold pH value in nitratification processes as had been repeatedly observed in the course of the present investigation. Series D—Treatment with NH4OH-f 0.5N HoSO* and Series E—Treatment with NHiOH+lN H2SO* In this experiment an attempt was made to adjust the initial pH value to a point below the threshold, by adding equal volumes of 0.5JV and IN sulfuric acid to the 100-gram samples of soil after treatment with ammonia. From the difference in acid concentration it was anticipated that the pH values of the samples would be 494 TECHNICAL BULLETIN NO. 96 reduced to different levels below the threshold, as indicated by the buffer curves obtained on the 1:5 suspension. The pH values actually obtained—namely, 7.47 with the 0.5IV solution and 7.50 with the normal acid solution—are identical within limits of precision of the pH determination. Obviously the soil at 70 per cent of its water-holding capacity is buffered considerably more than in its 1:5 suspension. This experiment is, nevertheless, of interest in that it shows how reproducible the extent of microbiological nitrogen transformations actually is when the pH value is held constant. The resulting curves for these two series of incubations (D and E), shown in Figure 4, are found to be so nearly identical that they are practically superimposable. The numerical data for this experiment, given in Table 3, indicate that the accumulation of nitrite begins considerably sooner than in the B and C series. Similarly, the period of time during which nitrites were produced in appreciable amounts is considerably shorter in the acidified than in the more alkaline samples. Nitrates were first observed at a much earlier date in the acid-treated samples; hence by the twenty-fifth day a considerably greater amount of nitrate had accumulated in the D and E series than in the B and C series, where the formation of nitrate was retarded by the high initial alkalinity. RELATIONSHIP BETWEEN PH VALUE AND NITRATE FORMATION The dependence of nitrate formation upon the attainment of the threshold pH value of 7,7 is strikingly shown by the curves in Figure 4. The control series A, in which the amounts of nitrogen originally present in the samples were rather minute, could not be expected to show a correlation between pH value and nitrate formation. It is quite significant, however, that nitrates were entirely absent from the samples of the B and C series until the pH had dropped to the threshold value, after which time nitrates began to accumulate. The importance of this threshold value is further illustrated by the data in Table 3, which show that nitrate formation began relatively promptly—i.e., on the sixth day after the beginning of the incubation—in those samples whose initial pH values were below the threshold. Therefore the conclusion may be drawn that, if the initial pH value lies above the threshold, nitratification begins promptly when the threshold value is reached, and hence represents a response of the microorganisms to an environment favorable to their activity. In samples whose initial pH value lies below the threshold, nitratification appears to depend simply upon the beginning of an appreciable microbiological activity following the normal lag phase. Since, in the D and E series the initial pH values were close to the threshold and during most of the incubation period did not drop to any considerable extent below it, the foregoing deduction is consistent THRESHOLD PH VALUE FOR NITRATIFICATION 495 with the normal growth or activity curve, the character of which is indicated graphically in Figure 1. CORRELATION OF INITIAL PH VALUE WITH NITRITE AND NITRATE FORMATION The influence of the various treatments upon the initial and final pH values of the soil samples and their relation to nitrite and nitrate formation are shown in the form of a bar graph (Fig. 5). In this figure are also given the pH values at the time nitrites were first observed, when the amount of nitrite formed reached a maximum, and when nitrates made their first appearance. At the top of each bar is given the day of incubation on which the observation was made. lst - DAY , pH 25th. DAY pH 1st. NOjj OBSERVED pH HAX. NOJ OBSERVED pH 1st. NOJ OBSERVED NH4OH - 0.5 N H2S04 NH4OH - I N H204 Figure 5. — Threshold pH range in the nitrification of ammonia in Gila sandy loam. (Numbers above bars indicate time in days of incubation after which measurement was made.) It will be noted in each instance that the pH value was equal to, or less than, the threshold value before nitrates began to appear. The time interval between the beginning of the incubation and the first appearance of nitrates appears to vary directly with the initial pH value of the sample. Thus in the most alkaline series (B), 25 days had elapsed before nitrates appeared, and in the less alkaline ammonia-treated series (C), 19 days elapsed prior to the formation of nitrates. In the two acidified series (D and E), the corresponding time interval was only 8 days. It is obvious, therefore, that the rate of nitrification of ammonia in desert soils among other things is a function of the pH value, and, in general, 496 TECHNICAL BULLETIN NO. 96 the higher it is above the threshold value of 7.7 =t 0.1 the longer will be the time which elapses before nitrates begin to make their appearance. Although it was not the primary object of this investigation to study the influence of pH upon the formation of nitrites in the micr'obial oxidation of ammonia, sufficient evidence has been accumulated to justify the formulation of a provisional rule as follows: The more alkaline the soil, the longer will be the time which elapses during the incubation before nitrites appear in amounts detectable by analysis. Figure 5 also indicates that a large decrease in alkalinity occurred in the B and C series prior to the formation of nitrites. This decrease amounted to 1.4 pH units in the B and 0.9 pH unit in the C series. The fact that complete equilibrium may not have been attained within the soil samples can account for only a small portion of this decrease. Hence it appears obvious that the ammonia must have been oxidized to some intermediate product (or products) before passing over into the nitrite form, in order to account for the observed drop in pH value. Further reference will be made to this phenomenon in the discussion of results which is to follow. Another point of interest shown in Figure 5 is that the build-up of nitrites appears to be conditioned by the length of time which elapses, prior to the attainment of the threshold value, before the oxidation of nitrite to nitrate begins. For example, in the B series, nitrites were first observed on the seventeenth day of incubation; nitrates on the twenty-fifth day. Thus there was a period of 8 days during which the accumulation of nitrites in the soil samples could take place. By way of contrast, the corresponding time interval in the C series was 5 days, and in the D-E series, only 2 days. While nitrites tend to accumulate in greatest amounts under alkaline conditions, it will be seen from Table 3 that even in the acid-treated samples, fairly large quantities of nitrite may accumulate, of the order of 70 to 90 p.p.m. Considering the process quantitatively one may regard the total amount of nitrite formed in the oxidation of ammonia as determined by two factors: the rate of its build-up and the time during which such an accumulation can occur. This time interval is, as shown above, the length of time required to bring the pH value of the soil down to the threshold value, at which point nitrites would begin to disappear by their oxidation to nitrate. It appears that high alkalinity reduces the activity of the organisms, which is equivalent to reducing the rate of nitritification. It is quite possible that ammonia is oxidized to nitrite more rapidly at the lower pH values than under the more alkaline conditions and at a more rapid rate than it can be oxidized to nitrates by the nitratifiers. EFFECT OF MAINTAINING THE SOIL AT A PH VALUE ABOVE THE THRESHOLD OF 7.7±0.1 In the preceding experiments the incubations were set up and begun under a particular set of conditions, and the pH value of THRESHOLD pH VALUE FOR NITRATIFICATION 497 the soil in each case changed during the period of incubation. It seemed of interest therefore to carry out an incubation experiment in which the pH value of the soil, initially at a level considerably above the threshold due to the addition'of ammonia, was maintained at a relatively high level by the periodic addition of base in the form of calcium hydroxide solution. As in previous series, 100-gram samples of the cultivated Gila sandy loam were treated with ammonia solution, added in 5-ml. portions to successive quarter portions of soil, so that the total nitrogen added amounted to 30 mg. The samples after weighing were incubated without the close-fitting covers, so that evaporation would be hastened and restoration of moisture lost could be made with saturated calcium hydroxide solution. Determinations of pH, nitrate, and nitrite were made on duplicate samples as before. The data are shown graphically in Figure 6. The first samples were analyzed on the third day of incubation, at which time the pH value was 8.72. During the following 18 days, the pH value gradually dropped in spite of the added base, but the nitrate accumulations were relatively slight, amounting to only 12 to 15 p.p.m. On the twenty-second day, however, notwithstanding the pH value was still 7.90, the nitrate content had increased to nearly 35 p.p.m. This amount is small, to be sure, in comparison with the amount formed in previous experiments, but nevertheless significant, in view of the fact that the pH value was still slightly above the threshold value. This result was interpreted to mean that the pH-nitratification curve did not jail immediately to zero once the threshold pH value had been passed, "but that nitratification continued to proceed slowly and nitrates to accumulate in measurable amounts after the lapse of a sufficient length of time. It might be taken to indicate that the pH value within a part of the soil mass was less than the threshold value, which would enable part of the nitrifying bacteria, at least, to function normally. Since the nitrites appeared to have reached a stationary value, at least did not appear to increase or decrease, after the seventeenth day of incubation, and nitrates appeared to be forming slowly within at least a portion of the soil mass, it was decided to withhold additional base until the threshold value had been reached, and then to observe whether or not such treatment had favored the formation of nitrates. Figure 6 shows the result: On the twenty-sixth day the pH value had decreased to a point within the threshold range—namely, 7.68—and the nitrate content had risen to 75 p.p.m. On the thirtieth day, when the pH value was still within the threshold range, the amount of nitrate formed had more than doubled in amount, totaling 180 p.p.m. Evidently, after the threshold pH range has once been attained, the nitrate content of the soil increases normally as a result of ameliorated pH conditions, and this increase in nitrate content appears to be independent of the time interval which elapses prior to the attainment of the threshold value. 498 TECHNICAL BULLETIN NO. 96 Figure 6.—Rate of nitrification of ammonia in Gila sandy loam as influenced by controlled alkalinity. If high alkalinity is the factor which prevents nitratification, it might be expected that the saturated calcium hydroxide solution added to the surface of the soil samples to restore moisture lost by evaporation would retard or prevent the activity of the nitrifying organisms with which it came in contact. It is probable, however, that this treatment did not affect all of the organisms throughout the sample equally, since there was no means available to bring about a uniform distribution of the base. The appearance of nitrate under these conditions can therefore be accounted for only on the assumption that the organisms were not all affected in the same manner or to the same extent by the calcium hydroxide solution added. The experiment here described was, however, continued by restoring the pH after the thirtieth incubation day to a value above the threshold to determine whether, in line with the foregoing reasoning, the rate of accumulation of nitrate should decrease. Figure 6 shows, in fact, that the slope of the nitrate curve dropped off quite promptly indicating that nitrates form much less rapidly when the pH value of the soil is raised above the threshold value. This experiment was terminated after 52 days of incubation, at which time the nitrate content of the soil was found to have reached a value of 225 p.p.m. This result is rather striking in view of the fact that the same soil fertilized with ammonia in the absence of calcium hydroxide had by the forty-ninth day of in- THRESHOLD pH VALUE FOR NITRATIFICATION 499 cubation (as shown in Table 2) attained a nitrate concentration of approximately 300 p.p.m. The nitrite curve in the same figure illustrates the general result, observed in all of the experiments, that the nitrite content of the soil falls rapidly to a mere trace as soon as the threshold pH value is reached and nitrates have begun to form at a fairly rapid rate. DISCUSSION In the foregoing studies several new and highly significant facts have been brought to light: first, ammonia exhibits a nitrification rate similar in magnitude to that of ammonium sulfate and urea under similar experimental conditions; second, there exists a threshold pH value for the nitratification of these compounds— namely, 7.7±0.1—above which nitrites are not oxidized to nitrates to any appreciable extent; third, nitrites may form in considerable amount even in a well-aerated desert soil maintained at the optimum moisture content. The fact of outstanding interest is the existence of a threshold pH value for nitratification, particularly in view of the work of Gerretsen (24), Gowda (25), Meyerhof (37, 38, 39), Olsen (44), and others, who reported pH values or ranges considerably higher than 7.7 as the optimum for the nitrifying bacteria. Meyerhof placed his optimum between 8.3 and 9.3, basing his deductions on results obtained with pure culture media. Under such conditions the mineral nutrients required by the organisms are usually present in an easily available form, so that the bacteria should be able to work most efficiently. In the soil, however, such may not be the case. If the needed mineral nutrients are present in the soil in a less available form due to high alkalinity, the bacteria cannot function efficiently, if at all, in the nitrifying process until the pH value has been reduced to a point where the required nutrients are more readily available. Thus the threshold value may prove to be a factor which affects both the biological behavior of the nitrifying bacteria and their nutrition as well. The threshold value of 7.7±0.1 here found is in striking agreement with the value of 7.6 established by Breazeale and McGeorge (11), beyond which plants are unable to absorb phosphate and/or nitrate from the soil solution. Basing his deductions upon thermodynamic principles, Buehrer (13) showed that since plants are unable to utilize phosphate from solution at pH values above 7.6, the form of phosphate ion used by them must be the H2PO4~ ion. Similarly Greene (26) found that Azotobacter grew best and fixed nitrogen most abundantly in soils at reactions where phosphorus is present chiefly as the H2PCV ion, and concluded that certain bacteria behave like plants with respect to their absorption of phosphate. It is therefore very probable that in alkaline calcareous soils the nitratifiers are not able to perform their normal functions without phosphorus, for at reactions above the threshold value the amount of phosphorus present in the form of 500 TECHNICAL BULLETIN NO. 96 H2PO4~ ion is relatively small. In solution cultures, on the other hand, it is quite possible for nitratification to occur at pH values considerably higher than 7.7, in view of the fact that even at higher alkalinity the ionic relations of the phosphate equilibrium (see Table 3 and Ref. 13) permit the existence of an amount of H2P04~ sufficient for bacterial metabolism. It is evident that the percentage of the gross phosphate concentration represented by the H2P04~ ion decreases as the pH rises. In solution cultures where the total concentration of soluble phosphate is usually high, the concentration of H2PO4" ions notwithstanding the high pH value may still be sufficiently high to enable the nitrifying organisms to function. In the soil, however, where both the total concentration of soluble phosphate and the percentage of H2PO4" present at higher pH values are small in magnitude, the actual concentration of H2PO4~ ion may be too small to satisfy the bacterial requirements. It is a fairly well-established fact that the nitrifying organisms can adapt themselves to free alkalinity if exposed to such conditions over sufficiently long periods of time. Olsen (44) working with strongly acid humus soil adjusted the pH with lime and found nitrification to occur over an unusually wide range of pH values. It is quite probable that under these conditions he did not succeed in neutralizing the acidity within the soil aggregates even though the pH value, as determined on the entire soil mass, appeared to have been on the alkaline side of neutrality. Among the nitrification studies on ammonia, ammonium sulfate, and urea, reported in the foregoing sections of this bulletin, three observations are of particular interest: first, the failure of the ammonia to nitrify in Superstition sand, as shown in Table 2; second, the apparent irregularity in the nitratification occurring in the virgin Gila sandy loam samples that had been fertilized with ammonia, as shown in Table 2; and third, the slower rate of nitratification of all three of the fertilizers in Palos Verdes sandy loam, as illustrated by the data of Table 3. The application of ammonia to Superstition sand, which exhibited a low buffer capacity toward base, being highly calcareous, raised the hydroxyl-ion concentration to such an inordinately high value—550 times the alkalinity existing at neutrality, for example—that nitrate formation did not take place. In the Palos Verdes soil, however, the buffer conditions are reversed, the soil having a low specific buffer capacity toward acid. This buffer capacity is characteristic of acid soils of the type with which Waksman (54) worked. He recommended the addition of a base to neutralize the acid which is continuously formed during the nitrification. The Palos Verdes soil had an initial pH value of 6.5 and exhibited a characteristically slower rate of nitrate formation than is shown by soils of a more alkaline reaction. During the nitrification of ammonium sulfate in this soil, the rate of nitrate formation decreased simultaneously with the decrease in pH value. This retardation of nitratification rate at the lower pH values is THRESHOLD pH VALUE FOR NITRATIFICATION 501 consistent with the findings of Waksman (54), Naftel (42), Gaarder and Hagem (23), and Humfeld and Erdman (28). In the present investigation, the series of samples treated with ammonium sulfate attained a minimal pH value of 4.65, which is not low enough to verify the pH range of 3.9 to 4.4 adopted by the foregoing investigators as the extreme lower limit of pH for nitrate formation. The challenging aspect of the present series of experiments was to establish definitely the actual existence of a threshold pH value for nitratification, and particularly to determine whether the decrease in pH from the high initial values was accompanied by, or resulted from, the formation of nitrites. The data for Series B and C, as well as their graphical representation in Figure 5, showed that a large drop in pH occurred prior to the beginning of nitrite formation. Several factors may have contributed to this effect: one is the possibility of a lag in the establishment of equilibrium in the soil samples following application of the solutions; some time is manifestly required to establish moisture- and base exchange-equilibrium in the soil. The effect manifests itself as an almost immediate initial drop in the pH curve. It is a wellknown fact that the pH value of any soil tends to fall for some time after the addition of water while the various equilibria are being established, but the lowering of the pH value from this cause is small in comparison with the large decreases in pH which were observed prior to the beginning of nitrite formation. Another factor which may have a bearing upon this decrease in pH prior to nitritification is the fact that the first observations of nitrite, as shown in Series B and C of Table 3, were not obtained immediately after the formation of nitrites had begun. The first appearance of nitrites in the B series was observed on the seventeenth day, at which time about 30 p.p.m. of nitrite had accumulated, or nearly half of the maximum amount which finally formed. Similarly about 22 per cent of the maximum value for nitrites in the C series had made its appearance by the fourteenth day, when nitrites were first detected. If this reasoning is correct, the remaining 50 per cent of the total amount of nitrite accumulated in Series B and 78 per cent in Series C was evidently much less effective in reducing the pH value than the amounts mentioned above. Each p.p.m. of nitrite formed would, if one assumes the equation 2 NH8 + 3 O2 + 2 OH- = 2 NO>~ + 4 H2O to represent the process, remove an equivalent amount of hydroxyl ion; hence there is no justification for the assumption that there may be a change in ratio of nitrite formed to hydroxyl removed. It is apparent that there must be some intermediate reaction which is responsible for the lowering in pH prior to the formation of nitrites. Various investigators have assumed the presence of one or more intermediate compounds in the formation of nitrite from am- 502 TECHNICAL BULLETIN NO. 96 monia. Beesley (7), Mumford (40), Kluyver and Donker (30), and Corbet (16, 17, 18, 19) have produced direct or indirect evidence of such intermediate products in their oxidation studies. Beesley found that as much as 44 per cent of the applied ammonia disappeared prior to the formation of nitrite, and concluded that the nitrogen must have passed through some intermediate stage, "which must be regarded as more or less hydroxylated." Mumford advanced the theory that the oxidation of ammonia involves the successive hydroxylation of the hydrogen atoms with its attendant removal of water, and reported the presence of hydroxylamine and salts of hyponitrous acid. Kluyver and Donker (30) similarly incline to the hydroxylamine-hyponitrous acid mechanism for the oxidation of ammonia, assuming that these intermediate products result from alternate hydration and dehydrogenation characteristic of microbiological reactions. Corbet (17) definitely reports the presence of these two compounds in systems involving ammonia oxidation. He writes as follows: It appears that the reaction proceeds through the formation of hydroxylamine and hyponitrous acid as intermediate compounds, but while the first-named can never have more than an ephemeral existence under the experimental conditions, hyponitrous acid present in the form of the calcium salt may account for as much as 40% of the total nitrogen present. The actual presence of hydroxylamine and hyponitrous acid in such a process is difficult to prove analytically even under the most favorable conditions. Rao et al. (47) attempted to demonstrate the formation of these compounds by using culture technique, but without success. They are of the opinion that Corbet's analytical procedure is not specific for hyponitrites. The presence of hydroxylamine, on the other hand, has been suggested (8, 9, 10, 31, 33) as an intermediate product not only in the oxidation of ammonia to nitrite but also in denitrification and nitrogen fixation—both of which are microbiological processes. ENERGY RELATIONS INVOLVED IN THE HYDROXYLAMINEHYPONITROUS ACID MECHANISM IN THE NITRIFICATION OF AMMONIA The pH changes associated with the nitrogen transformations in the microbiological oxidation of ammonia, set forth in the preceding sections, can be reconciled in part with the hydroxylaminehyponitrous acid mechanism as proposed by Kluyver and Donker. To be tenable, such a mechanism must involve not only the removal of hydroxyl ions, which is responsible for the drop in pH value, but it must also be consistent with the free energy changes accompanying the respective steps in the process. The free energy change, by virtue of its magnitude and sign, offers a criterion of whether or not the postulated intermediate reactions are likely to take place, or can take place if conditions are favorable. In other words, it is a measure of the tendency with which the reaction tends to take place, but it also represents the maximum amount THRESHOLD pH VALUE FOR NITRATIFICATION 503 of available energy liberated or absorbed when one mol of the reacting constituent is converted reversibly to another form. Baas-Becking and Parks (5) employed such a free energy approach in a theoretical study of the energy efficiency of various types of autotrophic bacteria, including the nitrifiers. They calculated the free energy change for the overall transformation of ammonium ion to nitrite, and of nitrite to nitrate, basing their calculations on certain data of Meyerhof (37, 38, 39) for the optimum concentrations of the ions involved in the respective transformations. The data in question were obtained in experiments with nutrient solutions, and from the analytically determined ion activities it was possible to calculate the free energy change attending the process under those conditions. In applying such free energy methods to transformations occurring in the soil, however, it is difficult to convert the standard free energy changes to the actual activities of the constituents as they exist in the soil solution, because neither hydroxylamine nor hyponitrous acid can be satisfactorily determined by existing analytical methods. It was decided, therefore, as a first approximation, to calculate the free energy changes for the constituents considered as being present in their standard states—namely, at unit activity. The appropriate data for these nitrogen transformations have been assembled by Latimer (32) from the most reliable thermodynamic studies available. In making such free energy calculations it is obviously possible to base them either (1) on the half reaction involving the nitrogen compounds or ions, which would take place at the negative electrode of a reversible galvanic cell, or (2) on the entire oxidation reaction involved in each step. In the latter procedure one must introduce the oxidizing agent which is generally written as molecular oxygen but which probably involves the microorganism in some manner. The free energy data so calculated are shown in Table 4, in which the values are given both for the half reaction and the entire oxidation reaction. For comparison, the data have been calculated for the respective steps under both alkaline and acid conditions. It will be noted that when the nitrification of ammonia occurs under alkaline conditions, as in the soils used in this study, each of the steps (as shown by the equation for the half reaction) uses up hydroxyl ions. As a result, there should be a decrease in pH value resulting from each of these steps. Furthermore, there is a much greater consumption of hydroxyl ion—namely, seven mols—when one mol of ammonium ion goes to nitrite, than when the nitrite is transformed to nitrate, which requires only two mols. The fact that the hydroxylamine and hyponitrous acid steps alone involve a total of five mols of hydroxyl accounts in part for the phenomenal drop in pH which occurs prior to both nitrite and nitrate formation. THRESHOLD PH VALUE FOR NITRATIFICATION 505 The free energy changes attending these steps give a hint as to a justification for the proposed mechanism. It will be noted that the first step in which the hydroxylamine is formed involves a large positive free energy change, "indicating that the process is not spontaneous. If the reaction occurs at all, energy must be supplied from some source, possibly by the microorganisms. Moreover, the high positive value of the free energy suggests that the hydroxylamine must be unstable, which in turn may account fof the difficulty of identifying it in the soil. When it changes to hyponitrous acid, the process is evidently spontaneous, since it involves a considerable decrease in free energy (33,680 calories per mol for the half reaction and 51,870 calories per mol for the entire reaction). The fact that energy must be put into the system in Step 1 may also account in part for the lag period which is always observed in the initial stages of ammonia oxidation. The free energy data for the transformation of nitrogen under acid conditions are of interest in the respect that in every step a high positive free energy value is involved. The reaction is quite evidently not favored by acid conditions since hydrogen ion is formed in each step, which would tend to reverse the process. In the case of the nitrifying organisms it has in fact been found (6) that microbiological oxidation processes cease when the pH value drops to 5.5, below which the oxidation is believed to be primarily chemical and catalyzed perhaps by hydrogen ion. Below a pH value of 4.5, the oxidation processes represented in Table 4 cease entirely. It is of interest therefore to note that the oxidation of ammonia, which is of primary importance when used as a fertilizer, proceeds favorably on the alkaline side of neutrality, where the free energy values for the reactions are dominantly negative. The fact that it requires the catalytic influence of the microorganisms to proceed emphasizes the repeatedly observed fact that the oxidation will not proceed with maximum efficiency if the environment of the soil is not favorable to the metabolism of the bacteria and their multiplication. Therefore, although an excessively alkaline condition in the soil would theoretically favor the nitrogen transformations in question, it is evident that unless man can otherwise ameliorate such an adverse condition, it must be done by the bacteria themselves, by reducing the pH value to the point where their own nutrition processes become normal and nitratification can take place efficiently. SUMMARY 1. A comparative study has been made of the rates of nitrification of ammonia, ammonium sulfate, and urea in six typical Arizona soils. 2. A threshold pH value of 7.7±0.1 has been found for the nitratification of the ammonia type of fertilizers in desert soils above which the complete oxidation of ammonia will not occur, and to which the pH value of such soils must first be reduced before nitrification can proceed to completion. 506 TECHNICAL BULLETIN NO. 96 3. In the microbiological oxidation of nitrogen applied in the three above-mentioned forms, there is considerable nitrite formation, even in well-aerated soils under favorable conditions of temperature and moisture, so long as the pH value of the soil is considerably above the threshold value. It was greatest in those cases where the soil had been rendered strongly alkaline by addition of calcium hydroxide. 4. A pronounced decrease in pH value occurs in the soil prior to both nitrite and nitrate formation. 5. Nitrite accumulation appears to be inhibited by a high concentration of calcium ions and/or by high alkalinity. 6. In all instances in which a significant accumulation of nitrites was noted, the amount of nitrite decreased almost simultaneously with the formation of nitrates. 7. Ammonia is not toxic to the nitrifying organisms even at a concentration as high as 300 p.p.m. The failure of the ammonia to nitrify in some instances is attributed to the high alkalinity of the soil. 8. The fact that practically all of the nitrogen added as ammonia can be analytically accounted for indicates that losses by volatilization from the soil or by the spontaneous decomposition of ammonium nitrite are negligible. 9. Under constant (uniform) experimental conditions it is found that equal amounts of nitrate are formed in soil samples treated with the different nitrogen fertilizers after a given period of time. 10. In alkaline calcareous soils ammonia nitrifies as rapidly as ammonium sulfate and urea, provided the soil is sufficiently well buffered toward base to withstand the initial change in pH. By virtue of the microbial oxidation, the pH value of a soil treated with ammonia will be reduced to substantially the same limiting value as when an equivalent amount of ammonium sulf ate or urea has been added to supply the nitrogen. THRESHOLD PH VALUE FOR NITRATIFICATION 507 LITERATURE CITED 1. Arnon, D. I. 1937. Ammonium and nitrate nitrogen nutrition of barley at different seasons in relation to hydrogen-ion concentration, manganese, copper and oxygen supply. Soil Sci. 44:91-113. 2. Arlington, L. B., and Shive, J. W. 1935. Rates of absorption of ammonium and nitrate nitrogen from culture solutions by ten day-old tomato seedlings at two pH levels. Soil Sci. 39:431-35. 3. Association of Official Agricultural Chemists. 1935. Official and Tentative Methods of Analysis. Fourth Edition. A.O.A.C., Washington, B.C. 4. Ayers, A. D., and Jenny, H. 1939. Private communication to Shell Chemical Company, San Francisco, California. 5. Baas-Becking, L. G. M., and Parks, G. S. 1927. Energy relations in the metabolism of autotrophic bacteria. Physiol. Rev. 7:85-106. 6. Barritt, N. W. 1933. The nitrification process in soils and biological filters. Ann. Appl. Biol. 20:165-76. 7. BeesleyrR. M. 1914. Experiments on the rate of nitrification. Trans. Chem. Soc. 105:1,014-24. 8. Blom, J. 1928. Determination of hydroxylamine. Biochem. Zeit. 194:385-91. 9. Blom, J. 1928. Formation of hydroxylamine in the reduction of nitrates by microorganisms. A contribution to the problem of amino acid formation by microorganisms. Biochem. Zeit. 194:392-409. 10. Blom, J. 1931. An attempt to clarify the chemical processes involved in the assimilation of molecular nitrogen by microorganisms. Centralbl. Bakt. 84:60-86. 11. Breazeale, J. F., and McGeorge, W. T. 1932. Nutritional disorders in alkaline soils as caused by deficiency of carbon dioxide. Ariz. Agr. Exp. Sta. Tech. Bull. 41. 12. Buchanan, R. E., and Fulmer, E. I. 1928. Physiology and Biochemistry of Bacteria. Williams and Wilkins Co., Baltimore. Vol. I, p. 16. 13. Buehrer, T. F. 1932. The physico-chemical relationships of soil phosphates. Ariz. Agr. Exp. Sta. Tech. Bull. 42. 14. Clark, H. E., and Shive, J. W. 1934. The influence of the pH of a culture solution on the rates of absorption of ammonium and nitrate nitrogen by the tomato plant. Soil Sci. 37:203-25. 15. Clark, H. E., and Shive, J. W. 1934. The influence of the pH of a culture solution on the assimilation of ammonium and nitrate nitrogen by the tomato plant. Soil Sci. 37:459-76. 508 TECHNICAL BULLETIN NO. 96 16. Corbet, A. S. 1934. The formation of hyponitrous acid as an intermediate compound in the biological or photochemical oxidation of ammonia to nitrous acid. I. Chemical reactions. Biochem. Jour. 28:1,575-82. 17. Corbet, A. S. 1935. Biological Processes in Tropical Soils with Special Reference to Malaysia. W. Heffer and Sons, Ltd., Cambridge. 18. Corbet, A. S. 1935. The formation of hyponitrous acid as an intermediate compound in the biological or photochemical oxidation of ammonia to nitrous acid. II. Microbiological oxidation. Biochem, Jour. 29:1,086-96. 19. Corbet, A. S. 1936. The biological and chemical oxidation of ammonia to nitric acid. Proc. Third Int. Cong. Soil Sci. 1:133-34. 20. Davidson, O. W,, and Shive, J. W. 1934. The influence of the hydrogen-ion concentration of the culture solution upon the absorption and assimilation of nitrate and ammonium nitrogen by peach trees grown in sand cultures. Soil Sci. 37:357-85. 21. Fraps, G. S., and Sterges, A. J. 1933. Causes of low nitrification capacity of certain soils. Soil Sci. 34:35363. 22. Fraps, G. S., and Sterges, A. J. 1937. Basicity of some phosphates as related to nitrification. Jour. Amer. Soc. Agron. 29:613-21. 23. Gaarder, T., and Hagem, O. 1922-23. Nitrification in acid solutions. Bergens Mus. Aarbok.; Naturv. Raekke No. 1, 26 pp. Original not seen; Chem. Abs. 19:2,508 (1925). 24. Gerretsen, F. C. 1925. Over den invloed van de waterstof-ionen-concentratie op bacteriologische processen. VersZagen van Landbouwkundige Onderzoek. der Rijkslandbouwproefstat. 30:1-44. (Summary in English.) 25. Gowda, R. N. 1924. Oxidation of ammonia and nitrites by microorganisms under different conditions. Soil Sci. 17:57-64, 26. Greene, R. A. 1933. Some factors limiting the applicability of biological methods for determining the availability of plant food elements in calcareous soils. Soil Sci. 36:261-66. 27. Halversen, W. V. 1928. The value of nitrification tests on soils representing extreme contrast in physical and chemical properties. Soil Sci. 26:221-31. 28. Humfeld, H., and Erdman. L. W. 1927. The significance of the hydrogen-ion concentration in soil nitrification studies. Iowa Acad. Sci. Proc. 34:63-67. 29. Jones, C. D., and Skinner, C. E. 1926. Absorption of nitrogen from culture solutions by plants. N. J. Agr. Exp. Sta. Ann. Rept pp. 360-65. 30. Kluyver, A. J., and Donker, H. J. L. 1926. Die Einheit in der Biochemie. Chem, d, Zelle u. Gewebe 13:134-90. Original not seen; cited by Stephenson (51). 31. Kostychev, S., and Tsvetkova, E. 1920. The utilization of nitrates by molds for the production of nitrogenous compounds. Zeit. physioL Chem. 111:171-200. THRESHOLD pH VALUE FOR NITRATIFICATION 509 32. Latimer, W. M. 1938. The Oxidation States of the Elements and Their Potentials in Aqueous Solutions. Prentice-Hall, Inc., New York. 33. Lindsay, G. A., and Rhines, C. M. 1932. The production of hydroxylamine by the reduction of nitrates and nitrites by various pure cultures of bacteria. Jour. Bact. 241:489-92. 34. McGeorge, W. T. 1923. The assimilation of nitrogen by sugar cane. Nitrates vs. ammonia salts. Planters' Rec. 27:347-52. 35. McGeorge, W. T., and Martin, W. P. 1940. pH determination of alkali soils. Jour, Assoc. Off. Agr Chem 23:205-19. 36. Meek, C. S., and Lipman, C. B. 1922. The relation of the reaction and of salt content of the medium on nitrifying bacteria. Jour. Gen. Physiol. 5:195-204. 37. Meyerhof, O. 1916. Untersuchungen iiber den Atmungsvorgang nitrifizierender Bakterien. Pflug. Arch, ges. Physiol. 164:353. Original not seen; cited by Stephenson (51). 38. Meyerhof, O. 1916. Untersuchungen uber den Atmungsvorgang nitrifizierender Bakterien. II. Beeinflussung der Atmung des Nitratbildners durch chemische Substanzen. Pflilg. Arch. ges. Physiol. 165:229. Original not seen; cited by Stephenson (51). 39. Meyerhof, O. 1917. Untersuchungen liber den Atmungsvorgang nitrifizierender Bakterien. IV. Die Atmung des Nitritbildners und ihre Beeinflussung durch chemische Substanzen. Pflug. Arch. ges. Physiol. 166:240. Original not seen; cited by Stephenson (51). 40. Mumford, E. M. 1914. The mechanism of nitrification. Jour. Chem. Soc. 30:36. 41. Naftel, J. A. 1931. The absorption of ammonia- and nitrate-nitrogen by various plants at different stages of growth. Jour. Amer. Soc. Agron. 23:142-58. 42. Naftel, J. A. 1931. The nitrification of ammonium sulfate as influenced by soil reaction and degree of base saturation. Jour. Amer. Soc. Agron. 23:175-85. 43. Nelson, D. H. 1929. Some effects of manganese sulfate and manganese chloride on nitrification. Jour. Amer. Soc. Agron. 21:547-60. 44. Olsen, C. 1928. On the significance of hydrogen-ion concentration for the cycle of nitrogen transformation in the soil. Compt. rend. Lab. Carlsl>erg 17: (8): 21. Abstracted in Nature (London) 123:44 (1929). 45. Pierre, W. H. 1927. Buffer capacity of soils and its relation to the development of soil acidity from the use of ammonium sulfate. Jour Amer. Soc. Agron. 19:332-51. 46. Prianishnikov, D. N. 1929. Ammonia in fertilizers and its relation to the life of plants. Trans. Sci. Inst. Pert. (Moscow) 61:99-103. Abstracted in Chem. Abs. 23:5,264 (1929). 510 TECHNICAL BULLETIN NO. 96 47. Rao, W. V. S,, Krishnamurti, P. V., and Rao, G. G. 1938. Mechanism of the microbiological oxidation of ammonia. I. Formation of intermediate products. Jour. Indian Chem. Soc. 15:599-603. 48. Sessions, A. C,, and Shive, J. W. 1933. The effect of culture solutions on growth and nitrogen fractions of oat plants at different stages of their development. Soil Sci 35:355-74. 49. Stahl, A. L., and Shive, J. W. 1933. Studies on nitrogen absorption from culture solutions. I. Oats Soil Sci. 35:375-99. 50. Stahl, A. L., and Shive, J. W. 1933. Further studies on nitrogen absorption from culture solutions Buckwheat. Soil Sci. 35:469-83. II 51. Stephenson, Marjory. 1939. Bacterial Metabolism. Longmans, Green and Co., New York pp 257-72. 52. Stewart, G. R., Thomas, E. C., and Horner, J. 1925. The comparative growth of pineapple plants with ammonia and nitrate nitrogen. Soil Sci. 20:227-42. 53. Tandon, S. P., and Dhar, N. R. 1934. Influence of temperature on bacterial nitrification in tropical countries. Soil Sci. 38:183-89. 54. Waksman, S. A. 1923. Microbiological analysis of soils as an index to soil fertility. V. Methods for the study of nitrification. Soil Sci. 15:241-60. 55. Waksman, S. A. 1932. Principles of Soil Microbiology. 2nd Ed. Williams and Wilkins Co., Baltimore, Md. 56. Waynick, D. D. 1934. Anhydrous ammonia as a fertilizer. Calif. Citrograph 19:295. 57. Willis, L. G., and Piland, J. R. 1931. Ammonium-calcium balance: A concentrated fertilizer problem Soil Sci. 31:5-17.
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project