f 1 SYMBIOTIC NITROGEN FIXATION AS INFLUENCED BY THE NITROGEN IN THE SOIL BY WILLIAM ALBERT ALBRECHT A . B. University of Illinois, 1911 B. S: University of Illinois, 1914 M. S. University of Illinois, 1915 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Agronomy in the Graduate School of the University of Illinois 1919 SYMBIOTIC NITROGEN FIXATION AS INFLUENCED BY THE NITROGEN IN THE SOIL BY WILLIAM ALBERT ALBRECHT A. B. University of Illinois, 1911 B. S. University of Illinois, 1914 M. S. University of Illinois, 1915 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Agronomy in the Graduate School of the University of Illinois 1919 Sz.C^ :^ C^U \ 6^ n„ ©X &-3> CONTENTS V Introduction 275 II. Historical 276 III. Experimental 283 Plan of the experiment 284 Analytical methods 287 Discussion of crop series and results 288 Series 1 (soybeans) 288 Nitrates in soil 290 Nodule production 291 Nitrogen balance 292 Series 2 (cowpeas) 296 Nodule production 298 Nitrates in soil 298 Nitrogen balance 299 Series 3 (Cowpeas) 302 Nodule production 303 Nitrates in soil 304 Nitrogen balance 305 Nitrogen changes in soils during growth of legumes 309 Distribution of nitrogen in tops and roots of cowpeas 311 IV. Summary 314 V. Conclusions 316 Digitized by the Internet Archive in 2010 with funding from The Library of Congress http://www.archive.org/details/symbioticnitrogeOOalbr Reprinted from Son. Science, Vol. IX, No. 5, May, 1920 SYMBIOTIC NITROGEN FIXATION AS INFLUENCED BY THE NITROGEN IN THE SOIL^ WILLIAM ALBERT ALBRECHT Formerly Fellow in Agronomy, University of Illinois^ Received for publication February 10, 1920 I. INTRODUCTION Of the important elements necessary for growing plants, nitrogen is the one that presents our most serious problem and is most apt to be deficient. Early studies on the different forms in which nitrogen may serve for plant growth seemed to have it definitely settled that only combined nitrogen could be used by plants (6, 30). Atwater (2), however, showed that legumes, quite contrary to earher beliefs, were able indirectly to utilize the elementary nitrogen. This prompted many researches and opened many discussions, which soon estab- lished the fact that certain plants, belonging to the Leguminosae, were able to obtain the gaseous nitrogen of the air through the action of bacteria living in the nodules on their roots. When the fact became known that legumes are able to use the nitrogen of the air through a mutually beneficial relationship with bacteria, numerous studies of these plants were undertaken to determine the manner in which they take nitrogen from the air and incorporate or fix it in their tissues. This process of "symbiotic nitrogen fixation," as it has been named, has taken on considerable significance in the attempt to maintain the nitrogen supply for plant growth. The fact that it offers a means of utiHzing the unlimited supply of nitrogen of the air in place of the costly nitrogenous fertilizers, has served as an incentive to study this process and the factors which influence its highest development. Any information giving a clearer understanding of the process of symbiotic nitrogen fixation may be justified as contributing to the large agricultural problem of maintaining the supply of nitrogen in the soil in sufficient amounts to insure maximum crop produc- tion. The following research is a contribution to the process of nitrogen fixation by legumes as influenced by the amount of nitrogen present in the soil in both the organic and inorganic forms. 1 A thesis submitted to the Faculty of the Graduate School of the University of Illinois in partial fulfillment of the requirements for the Degree of Doctor of Philosophy, June, 1919. 2 Present Associate Professor of Soils, University of Missouri. 275 SOIL SCIENCE, VOL. IX. NO. 5 276 WILLIAM ALBERT ALBRECHT II. HISTORICAL The literature on the subject of symbiotic nitrogen fixation is rather exten- sive and has been well collected in bibliographies by Jacobitz (27), by Burrill (7), and others, so that no extensive review on the subject is necessary. Only those papers deaUng particularly with this process as influenced ^by the nitro- gen content of the soil will be considered. " ' - " ' General statements are common in saying that the legume fulfills its needs for nitrogen from the soil and later resorts to the supply in the air. Conn (8) makes the statement that "legumes appear to prefer taking their nitrogenous material directly from the nitrogenous foods in the soil when these are present in abundance. But if the soil does not furnish the proper nitrogen, then recourse is had to atmospheric nitrogen, through the agency of tubercle organisms." Van Slyke (46) ventures a similar opinion in which he says, " when supplied with available nitrogen compounds, the bacteria fail to make use of atmospheric nitrogen," Hopkins (25, p. 217) agrees in substance with this. "Clover and other legumes," he says, "take available nitrogen from the soil in preference to the fixation of free nitrogen from the air, the latter being drawn upon only to supplement the soil's supply and thus balance the plant-food ration." Early works show that nodules are present when the bacteria become estab- lished and when the plant uses atmospheric nitrogen. In much of the litera- ture the nodules represent nitrogen fixation and unless this modification of the root is present, no use of gaseous nitrogen is believed to be taking place. The importance of nodules in nitrogen fixation was established early by Hellriegel and Wilfarth (19) as one of the fundamental facts when they say, "the nodules of the roots must not be considered as simply reservoirs of albuminoid substances; their relation to the assimilation of free nitrogen is that of cause to effect." In most works cited in the following discussion, the effects of the nitrogen in the soil as a cultural medium, are reported as favorable or unfavorable to the nodule production, and hence to the nitrogen fixation. Moore (34) apparently does not agree with this general conception of the importance of nodules. He believes that it is possible for the bacteria to enter the roots and be of benefit without evincing their presence by such external evidence as nodules. He fails to beheve that even if the absence of nodules might permit some nitrogen fixation, this would not prohibit the same performance in their presence. Rautenberg and Kuhn (43) ventured perhaps the earliest statement con- cerning the relation of nitrogen fixation to the nitrogen in the medium for plant growth. In their work Viciafaba, growing in a nitrogen-free solution, developed numerous nodules, while in the presence of nitrates no nodules appeared. De Vries (10) obtained similar results while studying nodules as a storage for nitrogen. In the absence of nitrogen from the cultural solution, many nodules of normal structure were produced, but otherwise scarcely any developed. SYMBIOTIC NITROGEN FIXATION 277 Schindler (44) working with clovers, vetch, serradella, kidney vetch, and beans in water cultures concludes that, in general, solutions rich in nitrogen are less disposed toward giving nodules than those lacking in this element. These same plants grown in soil fertihzed with compost or manure as com- pared with soils low in nitrogen gave larger and more numerous nodules in every case with the latter soil. Tschirch (45) in studying nodules, as a means of nitrogen storage, says that it is established that nodules grow more profusely in soils poor in nitrogen than in those rich in humus. Vines (47) treated soils with 1 per cent potassium nitrate and found nodular development decreased, and also an indication that as the amount of nitrate diminished, the development of nodules became more marked. Similar depressive effects on nodule production by nitrates are reported also by Baszler (3) and by Laurent (28). These studies on the influence of the nitrogen content of the soil on the nodule development, take this development as a measure of nitrogen fixation, but fail, however, to substantiate their contentions with careful chemical analyses to show that the total nitrogen in the plant and soil has truly increased. Frank (13) studied conditions of the plant as influenced by inoculation and other factors. In using soils rich in humus as compared with those very poor in this respect, he was led to believe that when humus is present in sufficient amounts the bacteria are dispensable and serve with no benefit to the plant. Where humus is lacking the bacteria are active. This explains why legumes can be grown on sand when all minerals are supplied even when no humus is present. Atwater (2) in growing peas in sand supplied with nutrient solutions con- taining varying amounts of potassium and calcium nitrates, found nitrogen fixation taking place. He measured it by the increase in the total nitrogen present at the close of the experiment as compared with that at the start. Table 1 is taken from his data and shows fixation when a large nitrogen ration is supplied. Woods (50) (of Connecticut) grew scarlet clover in sand with a nutritive solution and accounted for all the nitrogen in the experiment, as did Atwater, by analysis of material at the outset and at the close. With no nitrogen in the solution, 18 plants fixed an average of 37 mgm. per plant. When 40 mgm. of nitrogen as calcium nitrate and potassium nitrate were added at the start, the fixation was reduced to 30 mgm. per plant. Vetch responded differently, fixing 20 mgm. in the former case and 47 in the latter. Cowpeas treated with nitrate fixed amounts varying from 87 to 129 mgm. of nitrogen. His data show a larger part of the nitrogen in the roots when nitrate was omitted than when it was added. He says, "all the plants grown without added nitrogen gained in nitrogen. Some of the plants supplied with nitrate showed a loss. The gain has occurred where root nodules are developed and without them there was no gain of any account." Prazmowski (42) worked on an experiment similar to that of Woods to find out if the nitrates in the soil hindered or aided the bacteria in entering the 278 WILLIAM ALBERT ALBRECHT plant and fixing nitrogen. Peas in sterile sand cultures containing 300 mgm, of nitrogen fixed 50 mgm. With no nitrogen in the sand and only 12 mgm. in the seed, 70 mgm. were taken from the air. In working with water cul- tures, nitrates held down nitrogen fixation and nodule production. In the absence of nitrates the fixation varied from 17 to 83 mgm. per plant. He says that the nitrogen content of the soil influences the time at which the nodule empties itself. The presence of nitrogen in the soil brings this change at the time of seed formation, but its absence from the soil permits the pro- cess of nodule growth and nitrogen fixation to go on slowly during the life of the plant with an increase in it at the time of seed formation. TABLE 1 Nitrogen fixed by peas grown in nutrient solutions containing nitrates (Taken from Atwater) NUMBER OF NITROGEN SUPPLIED NITROGEN AT CLOSE GAIN OR EXPERI- MENT In seeds In solution Total Vines, roots, etc. Residual solution Total NITROGEN mgm. mgm. mgrn. mgm. mgm. mgm. mgm. 1 36.7 59.4 96.1 116.4 1.4 117.8 -21.7 3 72.6 59.4 132.0 158.9 3.8 162.7 -30.7 Group I. Small ni- 5 34.2 59.4 93.6 156.1 0.0 156.1 -62.5 trogen ration 7 71.5 59.4 130.9 158.1 0.0 158.1 -27.2 9 35.3 59.4 94.7 186.5 1.4 187.9 -93.2 11 72.5 59.4 131.9 210.9 2.7 213.6 -81.7 2 34.4 136.9 171.3 178.9 2.0 180.9 -9.6 4 75.2 136.9 212.1 200.6 12.8 213.4 -1.3 Group II. Large ni- 6 34.8 136.9 171.7 149.6 1.2 150.8 -20.9 trogen ration 8 70.3 136.9 207.2 197.5 12.7 210.2 -3.0 10 34.6 136.9 171.5 277.8 35.7 313.5 -142.0 12 68.8 136.9 205.7 260.2 45.7 305.9 -100.2 This work cited last, gives a different degree of effect for nitrates in a sand or open medium than it does for nitrates in water cultures. This fact may be of significance in explaining the injurious effects on nodule growth and nitrogen fixation which are attributed in many cases to the nitrates. The use of solutions may be inadequate for an experiment of this nature. To test the effects of different forms of nitrogen on the nodule growth of legumes, Frank (14) used calcium nitrate, ammonium sulfate and urea. He analyzed seeds at the start and total plants at the close, and found that for the lupine the greatest growth and nitrogen increase in the plant, as well as the most profuse nodule production, took place in the absence of all nitro- genous compounds. The pea behaved similarly. He measured the nitrogen fixed in sand and soil by lupines, peas and red clover, obtaining the results given in table 2. SYMBIOTIC NITROGEN FIXATION 279 Frank's data show that the yellow lupine fixes less nitrogen on a humus soil than on a sand soil, or that soils with a higher nitrogen content fix less nitrogen with this legume. It may be possible that physical differences or other factors were responsible, for he gives no detailed description of relations other than nitrogen in the soil. For the peas and clover the case is different. Both gave a decided increase in the soil rich in nitrogen. For the lupine he beheves that its nitrogen- fixing power is less in a nitrogen-rich soil than in one very poor in this respect. Nevertheless, a soil already rich in nitrogen may be enriched in this element TABLE 2 Nitrogen fixed in sand and soil by lupines, peas and red clover (Taken from Frank) KIND or son. DRY WEIGHT HAR- VESTED In seed and in- oculum In harvest IN- CREASE NITROGEN IN SOIL Outset Close INFECTION Lupine Sand Humus soil Humus soil Sand Sterilized sand not in oculated Humus soil gm. gm. gm. per cent per cent 5 14.760.035 0.3609 10.3 0.0096 0.0157 4 23.32 0.0364 0.2816 7.7 0.1076 0.1208 1 large nodule 1 to 7 nodules Peas 37.98 0.0282 0.7467 26.5 0.1076 0.1253 Of bean size Red clover 44.33 0.0457 0.7087 15.5 0.0073 0.0105 7.18 0.0457 0.0687 1.5 0.0073 0.0079 222.02 0.0457 4.6406 105.5 0.1076 0.1184 Rich in nodules Few nodules Few nodules by means of legumes. Peas and clover, he believes, reach their maximum fixing capacity only when using nitrogenous substances, especially nitrates, to supplement the bacteria on their roots, even though the direct opposite is true for the yellow lupine. The nitrogen enriching effect of legumes takes place not only in soils poor in nitrogen, but also in the better soils, rich in humus. The above conclusion is quite the opposite to that of Maercker (31), who used the yellow lupine in sterilized and inoculated sand with varying amounts of potassium nitrate added. He found in this experiment that nitrates did not hinder or lessen the abihty of the lupine as a nitrogen fixer. Nobbe and Hiltner (35) reported the diameter of Robinia nodules as 8 mm. in nitrogen-free soil, and 0.5 mm. in soil treated with nitrates. 280 WILLIAM ALBERT ALBRECHT Perhaps the most careful work in the early study of nitrogen fixation is that of Aeby (1) in his attempt to see if non-legumes would give nitrogen fixation. He used two soils, one a clay soil with 0.0783 per cent nitrogen and the other a "humus-rich" soil with 0.4050 per cent nitrogen. They were used in growing peas without nitrogen treatment, and with nitrogen added at four intervals to make a total of 2 gm. per pot of 4 kgm. of soil. Analyses were made for total nitrogen in all materials at the beginning and at the close. Any increase present at the close over that at the beginning represented fixa- tion of nitrogen. This method of analysis gave a loss of nitrogen in fallow pots but a decided increase for those on which peas were grown. Peas grow- ing in a rich soil fixed 1.976 gm. nitrogen per pot, while in the same soil treated with nitrogen, the fixation was but 1.621 gm. In the clay soil the corre- sponding figures are 2.759 gm. and 1.987 gm., respectively; a decrease of 0.355 and 0.772 gm. due to the nitrogen added. This indicates that in the soil rich, as well as one poor in nitrogen the addition of nitrogen depressed the fixation of atmospheric nitrogen. Furthermore, in the soil which was low in nitrogen the amount taken from the air by the peas was greater than in the rich soil, both when untreated and treated with nitrogen. Accordingly there was less fixation with increased amounts of nitrogen in the soils, which agrees with some of the preceding works cited. In the soil left fallow, he failed to recover as much nitrogen at the close as was present at the beginning. In the distribution of nitrogen in the roots and tops of plants, his results agree with those of Woods. Nobbe and Hiltner (36) go farther in their statements than many others and conclude from a study of cross inoculation, that nodules have no influence on plant growth when plenty of soil nitrogen is available. Salts containing nitrogen were used by Marchall (32) and found to inhibit nodule production in the following concentrations: alkaline nitrates 1 part in 10,000 and ammonium salts 1 part in 2000. By using soybeans and meas- uring the nitrogen increase in t^rms of the crop, when sterile or inoculated, or treated with nitrate nitrogen, Nobbe and Richter (38) found that with increased amounts of soluble nitrogen or humus substance added, the total nitrogen content of the crop decreased. They beHeved that small amounts of soluble nitrogen are beneficial to the young plants — at least until bacteroids are formed. According to them, inoculation was best in the absence of nitrates and decreased with the increase of the latter. To judge the value of legumes as nitrogen fixers, Wohltmann and Bergene (51) used a variety of soils ranging in nitrogen from 0.046 per cent to 0.205 per cent and a peat soil with 1.650 per cent. These soils were treated with either ammonium nitrate or ammonium sulfate and planted to a number of legumes. They judged the amount of nitrogen taken from the air in terms of the number of nodules produced, and give the following results and conclusions in regard to the influence of the nitrogen in the soil on the amount of nitrogen fixed. On all soils to which ammoniimi nitrate was added the nodules failed to SYMBIOTIC MTROGElv nXATION 281 develop and the plants grew weU in tke absence of tliem, Anamonimn sulfate suppressed nodtde production completely in nine cases and almost completely in two cases. These men belie\^e that legumes do not need the help of baaeria and atmos- pheric nitrogen when there is present in the soil an abimdant supply of avail- able nitrogen, and that the>^ use soil nitrogen ahnost exclusively. According to these conations, legum^ used as green manure would not add nitrogen to that soil whose a\^ailable supply of this essential element is high. The}' point out that it is the "a^^ailable" nitrogen rather than the "total" nitrogen that has a detrimental influence on nitrogen fixation. Similar resnlts were obtained about 1904 by Kobbe and Richter (39) who undertook to determine the efiect of soluble nitrogen in soil on the amomit of nitrogen taken from the air by vetch. The)^ measured fixation by differ- ence between total nitrogen in plants that were inoculated and those that were not so treated. The increase in the nitrogen content caused by bacteria was TABLES mtrogen in »stc*, inocuLaied wni urdnocidaUd, IreateA iviih nUraie niirogen 'Trom Nob>>; arid Eichtsr) ys irrnfjszsi ' 500 maL. \ 1000 jsgil ADKEI/ ^STTEOGZy ADHED 3S3IE.OGEJJ tJJOZSi T-,— vv^r^ V-. 1.533 No: 'viArjoi^v&i (gm.) X/ir erence ' gm. ) Increase f^per cait of t- liamana uz; lot thai hrauQii cm nodule piodscdaon witi peas, beans and vetch. He found tiaat potasaom nitrate pMisHated nodule formation when used m amounts as low as 1 part in 10,000; sodium nitrate required >^ut 1 part m 282 WILLIAM ALBERT ALBRECHT 2000, or the equivalent of 1000 pounds per acre. Urea, oxamide, and potassium cyanide were very prejudicial to the production of nodules even when used in very dilute solution. Ammonium salts, either as nitrate or sul- fate, prohibited nodules on vetch. Table 4 gives the proportions in which these compounds were fatal to nodule production. This work was done in water cultures and as in previous works cited may have given more significant effects than would be true of soils. The contention that assimilable nitrogen hinders nodule production is further supported by A, Hercke (20) who concludes that, "when the soil contains sufficient assimilable nitrogen, the presence of nodules on the roots has no influence on the nitrogen content of lupines. When soil is poor in nitrogen the presence of nodules increases the absolute as well as the per- centage of nitrogen content of the plant." This is contradicted by S. Hercke (21) who found that nitrogen compounds as ammoniimi sulfate, potassium nitrate and asparagin favored the growth of nodule bacteria. TABLE 4 Nitrogen concentrations fatal to nodule production (From Flamand) Potassium nitrate . , Sodium nitrate Calcium nitrate. . . . Ammonium nitrate Ammonium sulfate Pisum sativum 1:10,000 1:10,000 1: 2,000 1:10,000 1:10,000 Vicia narbonensis 1:10,000 1: 2,000 1:10,000 1 : 20,000 1:20,000 Faba equina 1:10,000 1: 2,000 1:20,000 1: 2,000 1:10,000 Fred and Graul (17) tested the effect of soluble nitrogenous compounds on nitrogen fixation in different kinds of soil and with different kinds of plants. In general, they found that increasing amounts of soluble nitrogen depressed nodule production, but for this effect a larger amount of soluble nitrogen was necessary than would probably ever occur under field conditions. They are supported in this by the field experiments of Ewart (11). When it was found that bacteria are responsible for nitrogen fixation in conjunction with the plant, the question arose concerning their behavior in this respect independent of a host. Several researches have been carried out to see how nitrogen in various forms in the media influences the nitrogen- assimilating capacity of this organism when separated from the plant. In 1891 Beijerinck (4) found nitrogen fixation taking place by these bacteria (since named Ps. radicicola) independent of a plant, when the media contained ammonium, sodium, and potassium nitrates. Other early workers who reported fixation independent of the plant in the presence of nitrogenous compounds were Prazmowski (42), Berthelot (5), and Frank (14), but they said it was too small to be significant. Larger nitrogen fixation was reported SYMBIOTIC NITROGEN FIXATION 283 for the organism living independently by Maze (33) and by Lewis and Nichol- son (29), the former finding as much as 23 mgm. of nitrogen fixed in 100 cc. of medium in 16 days. The presence of nitrogen is considered injurious to the bacteria themselves according to Moore (34). He beUeves that, "the cultivation of bacteria upon media containing appreciable quantities of nitrogen for any length of time is sufficient to cause them to lose both the power of infection and that of fixing atmospheric nitrogen." This is strongly refuted, however, by Burrill and Hansen (7) who grew Ps. radicicola on nitrogen media for 30 months without a loss of its special adaptations or its capacity for producing nodules. Very recently Fred (15) found that Ps. radicicola fixed more nitrogen when a trace of this element was present in the medium, but larger amounts of it retarded the process. He is supported by still more recent work of Hills (23) who found that fixation independent of the plant was increased by nitrate nitrogen. His results vary somewhat, and the fixation is large enough to be significant, which may not be wholly true of Fred's results. This review of literature indicates that on the fixation of nitrogen by legumes, one may expect a significant influence from the nitrogen content of the soil. According to several researches soluble or assimilable nitrogen has a depressing influence, especially in water cultures and in soils when used in larger amounts. Nobbe and Hiltner (37) showed that the water-culture method itself was unfavorable for the development of nodules, and conse- quently nitrogen additions under such conditions may have been wrongly interpreted. The use of soil seems to be the best procedure and duplicates most nearly the field relations. Little has been done, however, in using carefully analyzed soil to measure the influence of soil nitrogen on the process of nitrogen fixation under those conditions generally prevailing in the soil. The opinions on this question are varied, though the more recent ones em- phasize the assimilable or soluble nitrogen of the soil as being depressive to the efficiency of nitrogen-fixing bacteria. In most experiments cited the soluble nitrogen was applied as ammonia or nitrate salts and no account was taken of the nitrogen in organic matter. Such works fail to settle the signifi- cance of the organic matter in the soil concerning symbiotic nitrogen fixation. ni. EXPERIMENTAL The following experimental work was undertaken in order to study further the relation of symbiotic nitrogen fixation to the nitrogen content of soils, especially the total nitrogen as obtained by usual soil analysis. In advising the use of legumes in a rotation cropping system, especially for the soils of the Corn Belt, which are moderately high in total nitrogen, the following questions often arise: Will the legumes fix any atmospheric nitrogen in a soil already con- taining large amounts of this element? If they do, how efiicient will they be, and what significance does the nitrogen content of the soil have on the process? The following experimental work was undertaken with the hope of contributing a possible answer to these questions. 284 WILLIAM ALBERT ALBRECHT Plan of the experiment In experimental work necessitating the measurement of nitrogen fixation, two general methods have been widely accepted. The first one consists in growing one set of legume plants without bacteria on the roots, and another set under Uke conditions but with the organism applied. The difference in the nitrogen content of the two sets of plants represents the increase due to bacteria. The second method consists of careful analyses of all materials to determine the total nitrogen present both at the outset and at the close of the experiment. Under carefully controlled conditions, with no nitrogen added, the increase of this element at the close over that at the beginning represents nitrogen drawn from the atmosphere. Objections have been made to the first method on the ground that the soil or medium in which the plant grows is not taken into consideration. An in- crease in the nitrogen content of the plant because of the bacteria may mean that the plant was able to take more nitrogen from the soil and does not prove that all such increase in nitrogen came from the air. The second method is more detailed and painstaking and gives difficulty because of lack of refined methods for determining total nitrogen. However, it measures the total nitrogen in all materials concerned at the beginning, and again at the close, so that the increase must come from some other source than the soil. In carefully controlled conditions this source must be the at- mosphere. The latter method was used in this study. The soil used for the pot cultures was a yellow silt loam from the unglaci- ated area of southern Illinois and contained 625 pounds of nitrogen per 2,000,- 000 pounds of surface soil. It was extremely low in organic matter, hence in such a poor physical condition that at first sight it would appear much like a clay or clay loam. It gave an acid reaction, and in order to neutralize this it was treated with calcium carbonate at the rate of 2 tons per acre. Plant nutrient elements other than nitrogen were added in soluble form during the time the plants were growing to assure good fertility, except for nitrogen which is the main deficiency in the soil (26). This soil was chosen because it was so low in nitrogen, and would offer a low nitrogen basis, which had become stable because of years of weathering, and which could be increased by several large increments without surpassing the nitrogen content of the more fertile soils. Two kinds of legumes were grown, soybeans and cowpeas. The first series included soybeans but failed to do as well as expected in greenhouse work. They were replaced by cowpeas for the later series which were grown out-of- doors during most of the time. Insects molested the plants in all series and were combatted with chemicals containing no nitrogen. Red spiders were especially prevalent and were sprayed with cold water or potassium sulfide solution. During the first series a fungus gnat (Sciaria mycetophilidae) (18) infested the soils which were high in organic matter. Later troubles from this insect were not experienced. Insect infestations were prevalent on all plants, SYMBIOTIC NITROGEN FIXATION 285 SO that even though they may have hindered growth to some extent, their disturbances would not prevent comparable results. In the first series which contained soybeans, four duplicate series of 1 -gallon pots were used with the soil treatments given in table 5. TABLE s Soil treatments in crop series 1 (Soybeans) NITROGEN ADDED TO SOU (POUNDS PER 2,000,000 POUNDS OF SOU,) Pot series 1 1-2 3-4 5-6 7-8 9-10 None (not inoculated) None 10 pounds as nitrate 50 pounds as nitrate 150 pounds as nitrate Pot series 2 11-12 13-14 15-16 17-18 19-20 1000 pounds as clover tops 2000 pounds as clover tops 3000 pounds as clover tops 4000 pounds as clover tops 5000 pounds as clover tops Pot series 3 21-22 23-24 25-26 27-28 29-30 1000 pounds as clover tops and 10 pounds as nitrate 2000 pounds as clover tops and 10 pounds as nitrate 3000 pounds as clover tops and 10 pounds as nitrate 4000 pounds as clover tops and 10 pounds as nitrate 5000 pounds as clover tops and 10 pounds as nitrate Pot series 4 31-32 33-34 35-36 37-38 39-40 1000 pounds as clover tops and 50 pounds as nitrate 2000 pounds as clover tops and 50 pounds as nitrate 3000 pounds as clover tops and 50 pounds as nitrate 4000 pounds as clover tops and 50 pounds as nitrate 5000 pounds as clover tops and 50 pounds as nitrate The addition of nitrate nitrogen and clover tops was made on the basis of one acre, or 2,000,000 pounds of soil. In all cases a constant amount of soil was used, and the increasing amounts of clover tops added gave increasing amounts of substrate for the plants in the series. This treatment also brought about decided changes in the soil's physical condition. Four pots were used as checks. All pots in the series were inoculated except two of the check pots, no. 3 and 4, which remained uncontaminated and showed no nodules at the close of the experiment. 286 WILLIAM ALBERT ALBRECHT TABLE 6 Soil treatments in crop series 2 (Cowpeas) NITROGEN ADDED TO SOIL (POUNDS PER 2,000,000 POUNDS OF SOIL) Pot series 1 la-2a 3a-4a 5a-6a 7a-8a 9a-10a None 10 pounds as nitrate 25 pounds as nitrate 50 pounds as nitrate 100 pounds as nitrate Pot series 2 1-2 3-4 11-12 13-14 15-16 17-18 19-20 None (not inoculated) None 1000 pounds as clover tops 2000 pounds as clover tops 3000 pounds as clover tops 4000 pounds as clover tops 5000 pounds as clover tops TABLE 7 Soil treatments in crop series 3 (Cowpeas) NITROGEN ADDED TO SOIL (POUNDS PER 2,000,000 POUNDS OF SOIL) Pot series 1 lb-2b None 3b-4b 50 pounds as nitrates 5b-6b 100 pounds as nitrates 7b-8b 150 pounds as nitrates 9b-10b 200 pounds as nitrates llb-12b 250 pounds as nitrates 13b-14b 50 pounds as nitrates (added at intervals) Pot series 2 1-2 None 3-4 None 11-12 1000 pounds as clover tops 13-14 2000 pounds as clover tops 15-16 3000 pounds as clover tops 17-18 4000 pounds as clover tops 19-20 5000 pounds as clover tops SYMBIOTIC NITROGEN FIXATION 287 The second series was seeded mth cowpeas, and contained two pot series, one treated with nitrates, and one with organic matter. Original soil, the same as used previously, was treated with nitrate nitrogen, while the series treated with organic matter consisted of a part of those soils used in the first crop series The four check pots and the ten pots that had been treated with nitrogen as clover tops only, were taken to complete this series. These soils were in better physical condition and the stage of most rapid decay seemed to have been passed. Constant amounts of soil (3300 gm. water-free soil) were weighed into each pot. All pots were inoculated except the two check pots, but e\ddently these were already inoculated, or became contaminated, for nodules were present at the close of the series. This crop series with its treat- ment is summarized in table 6. The third series was similar in soil treatment to that of the second. Some of the original soil was treated ^iih nitrates for one pot series, while those soils which had been used in both pre\dous series were used for the other pot series. Cowpeas were grown as the crop. The treatments are given in table 7. The last two series in which cowpeas were grown were kept out-of-doors near the greenhouse for much of the time. This seemed to favor better growth, and lessened the insect troubles. Care was taken to move the plants inside before rain and only once did rain fall on each series. It was only a slight shower and the effects in adding nitrogen were negligible. Analytical methods The usual analytical methods were employed. To measure the total nitro- gen in organic matter such as seeds and plants, the materials were digested with mercury and sulfuric acid or with acid and sodiimi sulfate, neutralized with sodium hydroxide and potassium sulfide, distilled into standard acid and titrated with standard alkali. For the total nitrogen determination on soils, 10-gm. samples were first dried for 8 hours at a temperature of 107°-108°C., the loss in moisture determined, and then transferred for digestion. Sulfuric acid containing salicylic acid was added and allowed to stand for some time, usually over night. Sodiimi thiosulfate w^as introduced and heated slowly untn frothing ceased. Mercury was then added and digested to clearness, potassiimi permanganate being used to assure complete oxidation. This was then neutralized with alkali and distilled into standard acid, according to the usual procedure. For the determination of nitrate nitrogen, samples of soil were dried at 107°- 108°C. for 8 hours and then extracted with 0.0625 .V hydrochloric acid. After being made alkaline, the ammonia was boiled off, and the original volume re- placed by nitrogen-free water. Devarda's metal was added, the reduced nitrogen distilled into 0.0357 T standard sulfuric acid and titrated. 288 WILLIAM ALBERT ALBRECHT In all the determinations the greatest care was exercised. The organic matter was dried, weighed, thoroughly ground, and kept in air-tight containers to prevent decided variations in moisture content before an entire series could be analyzed. Purest chemicals were used and all reagents were made up in large amounts to insure greater uniformity in results. Standard acid for ni- trogen in soil was of 1/14 normaUty and with sodium alizarin sulfonate as an indicator, titrations checked within 0.2 cc, a variation equivalent to 0.2 mgm. nitrogen in 10-gm. samples, or 40 pounds of nitrogen in 2,000,000 pounds of soil. For nitrate determinations, a sulfuric acid of 1/28 normality was used and 0.2 cc. allowed as variation in duplicates. Determinations in moisture loss for 10-gm. samples were usually less than 6 mgm. All calculations and determinations for soil were on the water-free basis to insure uniformity. Samples were run in triplicate and quadruplicate to offset errors. Only doubly distilled nitrogen-free water was used, both in analytical work and in the pot cultures. Discussion of crop series Series 1 (Soybeans) The soil for this series was thoroughly mixed in a quantity large enough for all the pots. To determine the nitrogen in it, a large sample was taken and ground to pass a 100-mesh sieve and then analyzed for total nitrogen. Ac- cording to the analysis, 30-gm. quantities of air-dry soil, which were equivalent to 29.1970 gm. water-free soil, contained 9.1568 mgm., or 0.03136 per cent nitrogen on the latter basis, as an average of five determinations varying less than 0.2 mgm. Each pot received 3500 gm. of soil with a moisture content of 14.01 per cent, or the equivalent of 3421 gm. of water-free soil, and a nit ogen content according to the above analysis of 1072 mgm. For one pot series the soil was treated with nitrate nitrogen only, at the rate of 17.2, 86.0 and 258 mgm. of nitrogen per pot, or the equivalent of 10, 50 and 150 pounds per 2,000,000 pounds of soil. For three additional pot series the nitrogen in the soil was increased at increments equivalent to 1,000 pounds of nitrogen per 2,000,000 pounds of soil by means of finely ground clover tops. The clover tops contained 2.709 per cent of nitrogen, as an average of 12 analy- ses. With 3500 gm. of soil per pot an increment of 1000 pounds per 2,000,000 — one part in 2000 — required 1.75 gm. of nitrogen, or 64.6 gm. of clover tops. Accordingly, for the second pot series enough clover tops were added to give a series of the original soil, whose nitrogen was increased at the rate of 1000, 2000, 3000, 4000 and 5000 pounds per acre. The third series was similar to the second except that in addition to the organic matter sodium nitrate also was added to all the pots at the rate of 10 pounds of nitrogen per 2,000,000 pounds of soil, while the fourth series differed from the third only by the addi- tion of sodium nitrate at the rate of 50 pounds of nitrogen instead of 10. This gave one pot series whose nitrogen was increased as nitrate, one with increas- SYMBIOTIC NITROGEN FIXATION 289 ing nitrogen in organic matter as clover tops, another increasing with clover tops and 10 pounds of nitrate nitrogen and the fourth whose nitrogen increased as clover tops and 50 pounds of nitrate nitrogen. The addition of varying amounts of clover tops to 3500 gm. of soil increased the amount of soil in the pots, so that the weight of soil in the different pots was not a constant figure. According to Hilgard (22) and others, the best moisture content of soils for proper plant development is one-haK of the moisture-holding capacity. On this basis, determinations of moisture-holding capacity were made on these soil series as modified by additions of clover tops, and the water applied to the pots in quantities to give the equivalent of 50 per cent of the moisture-holding capacity. An attempt was made to maintain such conditions throughout the experiment by weighing the pots at intervals, but the labor in- volved was large and the growth of the crops prohibited absolute accuracy. It was therefore necessary to apply the water according to one's best judg- ment. The presence of large amounts of organic matter caused trouble in keeping the moisture at optimum amounts. Soybean seeds were selected so that each seed weighed 140 mgm. Six of these were planted in each pot, in order that five vigorous plants might be assured in spite of low vitality and poor germination. The sixth plant or seed was removed after sufficient time had elapsed for the plants to get well started. According to the analyses of the beans, each seed contained an average of 9 mgm. of nitrogen. The growth of this series was fair. Numerous replantings were necessary because the large amounts of fresh organic matter interfered with proper germi- nation of the seeds, a difiiculty experienced also by Fred (16) at Wisconsin and by others at Illinois with oats, cowpeas and clover after turning under alfalfa. The injurious effects were more marked with the increase in organic matter so that on the pots receiving clover tops at the rate of 5000 pounds of nitrogen per 2,000,000 pounds of soil, only small plants were produced. These effects were not offset by the nitrate nitrogen added at the rate of 10 pounds and 50 pounds per 2,000,000 along with the organic matter. Some plants "damped- off" at the age of four weeks, while others died later. The growth was much influenced by the organic matter in the soil, being the best of the entire series where 1000 and 2000 pounds of nitrogen had been added as clover tops, but decidedly poorer in soil with larger amounts of this material added. The plants showed a less deep color, smaller leaves, less branching and more trans- location, with the increase in organic matter above the equivalent of 2000 pounds of nitrogen per acre. The effect of the nitrates was marked in the soil when these were added alone, but gave no appreciable effect when coupled with the organic matter. In the soil alone, the increasing amounts of nitrates gave im- proved plant growth for the early part of the plant season, but these differences were obliterated when the crop matured, as was shown by the weights and nitrogen content of the crop. The differences in the four pot series just before their close are shown in plate 1. 290 WILLIAM ALBERT ALBRECHT Nitrates in soil growing soybeans. Before removing the plants at harvest, samples of soil were taken and nitrates determined by extraction and reduction with Devarda's alloy. Forty-gram samples of water-free soil were used. De- terminations on those pots with organic matter added were somewhat erratic and failed to check well in duplicate. Since large amounts of organic matter are known to interfere with reduction methods (9, 48) and give high results, full credence cannot be placed on the results from these soils with organic matter. Table 8 gives the results of the determinations of nitrates. TABLE 8 Nitrate nitrogen in soil growing soybeans POT NITROGEN ADDED (POUNDS PER 2,000,000 POUNDS OF SOIL) MILLIGRAMS IN 40 GRAM SOIL MILLIGRAMS PER POT POUNDS PER 2,000,000 POUNDS or SOIL 1 None 0.337 28 16.8 2 None 0.366 31 18.3 3 None 0.378 32 18.9 4 None 0.337 28 16.8 5 10 pounds as nitrate 0.397 33 19.8 6 10 pounds as nitrate 0.378 32 18.9 7 50 pounds as nitrate 0.337 28 16.8 8 50 pounds as nitrate 0.397 33 19.8 9 150 pounds as nitrate 1.353 115 67.6 10 150 pounds as nitrate 1.533 131 76.6 11 1000 pounds as clover tops 2.031 176 100.0 12 1000 ix)unds as clover tops 0.816 71 40.0 13 2000 pounds as clover tops 2.442 216 122.0 14 2000 pounds as clover tops 5.835 516 291.0 15 3000 pounds as clover tops 15.852 1425 792.0 16 3000 pounds as clover tops 12.585 1132 629.0 17 4000 pounds as clover tops 9.459 864 472.0 18 4000 pounds as clover tops 9.757 892 487.0 19 5000 pounds as clover tops 8.403 780 420.0 20 5000 pounds as clover tops 8.722 810 436.0 From the above table it is evident that the pots receiving the equivalent of 10 and 50 pounds of nitrate nitrogen per acre contained no more nitrogen in this form at the close of the series than the pots left untreated. Evidently the plants consumed all that was applied, since soil conditions were unfavorable to denitrification, the other possible chance for nitrate removal. With the pots receiving the equivalent of 150 pounds per acre the conditions were dif- ferent and the nitrate content was much higher. The application of such a large amount, far above that needed for the plant growth, left some in the soil, since no leaching took place and the open soil prohibited denitrification. This would suggest that such a large apphcation suppHes more nitrogen than a single crop of soybeans can remove. SYMBIOTIC NITROGEN FIXATION 291 In the soils with organic matter the figures indicate excessive amounts of nitrates, reaching the maximum in the pots receiving 3000 pounds of nitrogen as clover tops per 2,000,000 pounds of soil, and decreasing somewhat with the heavier applications. The high results may have been caused partly by the interference of the organic matter. Nodule production by soybeans. When the plants were harvested, the roots were removed from the soil as completely as possible and studied for nodule production. Pots 1 and 2, untreated, which had not been inoculated at the outset were still sterile in respect to the soybean bacteria, containing no nod- ules and showing no contamination from the inoculated pots all about them. This indicates that there is no great danger of inoculation by contamination in an experiment of short duration of this kind with soybeans when no special precautions are taken. In the pots in this series, treated with sodium nitrate, many good-sized nodules were present giving no visible effect from the nitrate treatment either on the size or on the number of the nodules, except perhaps in the two pots receiving the equivalent of 150 pounds of nitrogen per acre. In these the nodules were smaller, but this difference was offset by increased numbers. In the pot series treated with clover tops, nodule production decreased with added amounts of organic matter. However, in the soils with the highest ap- plication there was such a poor development of the plants that very few nodules could be expected. The low vitality and poor plant growth rather than any disturbing factors in the soil may have been the cause of insignificant nodule formation. There was no evidence to indicate that with healthy plants nodule production would have been prohibited by these large amounts of or- ganic matter. In the soil treated with such large amounts of clover tops as would give 5000 pounds of nitrogen in 2,000,000 pounds of soil, the original composition and physical make-up were radically altered. Such modifications of soil condi- gions invited the infestation by a winged fungus gnat {Sciara mycetophilidae) which may have consumed and removed much of the nitrogen. The open soil structure and a plentiful food supply in the form of decaying organic matter furnished an excellent habitat for these small insects. The infestation origin- ated in the pots with the heaviest applications of organic matter, but spread to all pots so treated. Its duration was Hmited to a certain stage or period in the decomposition of the organic matter, and the insects were present only long enough to develop the larvae and pupae of one generation. They may have been a disturbing factor of which no account could be taken, and may have been partly responsible for the losses of nitrogen in these series. In tables 9, 10, 11 and 12 is given the nitrogen balance for the four pot series with soybeans. Many of the analytical data are omitted and only the sum- mation figures are given to show the balance between the total nitrogen in the soil and seed at the beginning, and that of the soil and crop at the close. Fig- ures given are the results of quadruplicate determinations in the soil, while in SOIL SCIENCE, VOL. IX, NO. 5 292 WILLIAM ALBERT ALBRECHT the crop the total material was analyzed. Four pots given in the first series are used as the checks for the remaining pot series. Following the tables is a graphical representation of the increases and losses in total nitrogen shown in the tables. The ordinates represent milligrams of nitrogen gained or lost while the abscissas represent nitrate nitrogen added to the soil in case of one curve, and the total nitrogen present in the soil at the outset for the remaining curves. Instead of showing an increase in nitrogen, most pots show a decrease. In- creases in nitrogen, when they occur, are insignificant as compared with the TABLE 9 Nitrogen balance — Soybeans Pot series 1, soil treated with nitrates POT nitrate nitro- gen ADDED SAMPLE WATER- FREE SOIL NITROGEN IN SAMPLE NITRO- GEN IN POT AT CLOSE CROP WEIGHT NITRO- GEN IN CROP NITROGEN AT CLOSE (SOIL-f CROP) NITRO- GEN AT OUTSET (SOIL -t- seed) INCREASE OR FIXATION AVER- AGE mgm. gm. mgm. mgm. gm. mgm. mgm. mgm. mgm. mgm. 1 None (uninoc- ulated) 9.8394" 2.8713^ 9981= 2.9 36" 10.34 1117" -83 2 None (uninoc- ulated) 9.8215 2.9714 1032 4.0 58 1090 1117 -55 f -36 3 None 9.8260 2.9S48 1040 3.5 98 1138 1117 21 4 None 9.8288 2.8864 1003 3.85 89 1092 1117 -25 J 5 17.2 9.7864 2.8783 1005 3.85 95 1100 1134 -341 -16 6 17.2 9.8025 2.9714 1037 4.75 98 11.35 11.34 -1/ 7 8 86 86 9.8029 9.8183 2.9140 2.8498 1016 992 5.12 4.25 113 98 1129 1090 1203 1203 -74\ -113/ -93 9 258 9.8087 3.2243 1124 4.25 124 1248 1375 -1271 -122 10 258 9.8190 3.1790 1106 3.70 151 1257 1375 -118/ * Average of four determinations, or duplicates made at two different times. ^ Based on 3421 gm. water-free soil in each pot. ° Found by analyzing entire crop, roots and tops. ^ Based on the analyses of the soil at the outset giving 0.03136 per cent of nitrogen, or 1072 mgm. of nitrogen in each pot with 3421 gm. of soil, and 5 seeds containing 45 mgm. of nitrogen. Additions of nitrate were equivalent to 17.2, 86 and 258 mgm. of nitrogen. losses. When the soil was treated with nitrates only, one of the duphcate pots showed an increase in nitrogen at the close over that present at the outset, but this increase was within the limit of difference between duplicate determi- nations which, when calculated per pot, was 60 mgm. of nitrogen. The limits of error in nitrogen per pot were less, however, than this figure, as a result of quadruphcate analysis. The largest loss of nitrogen was 127 mgm. As more nitrate nitrogen was added at the beginning, the loss at the close increased, corresponding closely to the amount added. This indicates that perhaps the SYMBIOTIC NITROGEN FIXATION 293 analytical procedure failed to detect the nitrate nitrogen added. Had the nitrate nitrogen been incorporated into the plant tissue as protein, it would certainly have been found; but it might be possible, in spite of all care in this respect, that such small amounts of nitrate were not detected by the analysis for total nitrogen at the close. Later crop series receiving far larger amounts of nitrogen as nitrate, failed to show any such discrepancies. TABLE 10 Nitrogen balance — Soybeans Pot series 2, soil treated with clover tops POT NITRO- GEN ADDED AS CLOVER TOPS SAMPLE WATER - FREE SOIL NITROGEN IN SAMPLE SOIL IN POT NITRO- GEN IN POT AT CLOSE CROP WEIGHT NITRO- GEN IN CROP NITRO- GEN AT CLOSE (SOIL + CROP) NITRO- GEN AT OUTSET (soil -h seed) in- crease, OR fixation aver- age mgm. em. mgm. gm. mgm. gm. mgm. mgm. mgm. mgm. mgm. 11 1750" 9. 8361 b 7.7461b 3480 2740° 2.85 122 2862 2867 -5\ 79 12 1750 9.8385 8.0541 3480 2848 4.70 182 3030 2867 163/ 13 3500 9.8279 12.3984 3539 4464 4.15 189 4653 4617 36\ 90 14 3500 9.8216 12.7212 3539 4583 4.05 179 4762 4617 145 J 15 16 5250 5250 9.7951 9.7905 15.4721 16.3141 3598 3598 5683 5995 3.55 2.70 199 154 5882 6149 6367 6367 -485 \ -218J -351 17 18 7000 7000 9.7815 9.8248 f 19.9884 19.3026 3657 3657 7473 7184 1.85 1.70 131 99 7604 7283 8117 8117 -513\ -834j -673 19 20 8750 8750 9.7879 9.8146 23.2323 23.3436 3716 3716 8820 8838 1.25 1.20 91 82 8911 8920 9867 9867 -956\ -947 J -951 » Clover tops added were figured on the basis of 3500 gm. of soil per pot instead of 3421, hence are not truly equivalent to the number of pounds per 2,000,000 pounds of soil as the increments were intended. ^ Average of four determinations, or duplicates made at two different times. " Based on water-free soil in pot as given in fifth column. "^ Found by analyzing entire crop, roots and tops. » Based on the analyses of the soil at the outset giving 0.03136 per cent of nitrogen, or 1072 mgm. of nitrogen in each pot with 3421 gm. of soil, and 5 seeds containing 45 mgm. of nitrogen. ' Two determinations only. In the three series receiving organic matter alone and organic matter in conjunction with 10 and 50 pounds of nitrogen as nitrate, there was an increase in nitrogen for all pots receiving clover tops equal to 1000 pounds of nitrogen per acre, and for two pots treated with organic matter equivalent to 2000 pounds of nitrogen. All other pots in these three series showed decided losses. In the pots with the equivalent of 5000 pounds of nitrogen added as clover tops this loss ran as high as 1150 mgm. of nitrogen in some few cases. :.:^K 294 WILLIAM ALBERT ALBRECHT These large losses must have resulted from the very rapid decomposition of the fresh organic matter and doubtless the nitrogen of the clover tops escaped as gaseous ammonia. The excessive application encouraged rapid decomposi- tion, and with 320 gm. of clover tops incorporated in 3421 gm. of soil, the pro- portion of soil may have been too small to absorb all the ammonia. The losses increased fairly regularly with larger applications of clover tops, though in no definite mathematical relation. TABLE 11 Nitrogen balance — Soybeans Pot series 3, soil treated with clover tops and 10 pounds of nitrate nitrogen 21 22 23 24 25 26 27 28 29 30 NITROGEN ADDED mgm. fl750^ as clover < tops and [17.2 as nitrate 3500 as clover tops and 17.2 as nitrate (5250 as clover tops and 17.2 as nitrate [7000 as clover \ tops and 17.2 as nitrate 8750 as clover tops and 17 . 2 as nitrate w H 5 14 nd H O (4 SAMPLE WATER- NITROGEN H O H O < o Z— bT W M O Www w2 FREE IN SAMPLE Z ^ O P< o a Bi Sjs w <% SOIL 2 f-' ns o B. O O ex! a! o o u OS Hi 1 Bi P , H + gS H U oE LO Z O Z Z z gm. mgm. gm. mgm. gm. mgm. mgm. mgm. mgm. 9.7732b 7.9007b 3480 2813 « 3.85 144 d 2957 2884" 73] 9.7804 8.1348 3480 2894 4.1 167 3061 2884 177J 9.7684 11.8428 3539 4290 4.2 167 4457 4634 -177] 9.7537 12.1518 3539 4409 3.9 169 4578 4634 -56] 9.7672 14.8770 3598 5480 2.2 123 5603 6384 -781] 9.7629 15.7567 3598 5806 2.0 126 5932 6384 -452 J 9.7648 18.6387 3657 6980 1.9 105 7085 8134 -949] 9.7642 19.2363 3657 7204 1.85 126 7330 8134 -8O4J 9.7626 22.6532 3716 8628 0.05 73 8701 9884 -1183] 9.7556 22.4843 3716 8653 1.80 110 8763 9884 -II21J 125 -116 -616 -876 -1152 * Clover tops and nitrate added were figured ou the basis of 3500 gm. of soil per pot in- stead of 3421, hence are not truly equivalent to the number of pounds per 2,000,000 pounds of soil as the increments were intended. b Average of four determinations, or duplicates made at two different times. " Based on weights of soil given in fifth cohimn. ^ Found by analyzing entire crop, roots and tops. * Nitrogen in soil 1072 mgm., nitrogen in seed 45 mgm., nitrogen m 64.59 gm. clover tops equivalent to 1000 pounds per acre 1750 mgm., and nitrogen as nitrate 17.2 mgm. From the data of this series, one cannot say with any great certainty that the soybeans so grown used atmospheric nitrogen. The treatment of nitrate apparently prevented an increase in nitrogen while the treatment of clover tops equal to 1000 pounds of nitrogen gave an increase of nitrogen. This must SYMBIOTIC NITROGEN FIXATION 295 have come from the air in the form of nitrogen fixation. The fact that there is a loss of total nitrogen in the system does not deny the possibiHty of the plants fixing nitrogen, for the legume may have used atmospheric nitrogen at the same time this escape from the soil was taking place. The crop growth on all these series was small, and the total nitrogen in the crop of any pot never exceeded 200 mgm. With such small plant growth no TABLE 12 Nitrogen balance — Soybeans Pot series 4, soil treated w ith clover tops and 50 pounds of nitrate nitrogen POT NITROGEN ADDED SAMPLE WATER- FREE SOU NITROGEN IN S.\MPLE H O g iJ )-i O tn a < !5 H m a p< o O !5 OS as H o < O z-^7: W H O o S « O o " H + 1^ H 3 « » i_ °Z Z 1 < mgm. gw. mgm. gm. mgm. gm. mgm. mgm. mgm. mgm. mgm. 31 [1750* as clover \ tops and 9.7603b 8.4575b 3480 3015° 3.6 154d 3169 2953" 216 114 32 86 as nitrate 9.7656 7.9314 3480 2826 3.3 140 2966 2953 I3J 33 3500 as clover tops and 9.7694 12.0295 3539 4357 3.25 144 4501 4703 -202] -187 34 86 as nitrate 9.7561 12.0698 3539 4378 3.05 152 4530 4703 -I73J 35 5250 as clover tops and 9.7481 15.0377 3598 5550 1.5 93 5643 6463 -820] -490 36 86 as nitrate 9.7474 16.7461 3598 6181 1.9 122 6303 6463 -I60J 37 [7000 as clover tops and 9.7386 19.6224 3657 7368 1.5 101 7469 8203 -734] -950 38 86 as nitrate 9.7546 18.4956 3657 6934 1.65 103 7037 8203 -1166J 39 ('8750 as clover 9.7601 23.3127 3716 8875 0.9 47 8922 9953 -10311 tops and I -1148 40 86 as nitrate 9.7570 22.6049 3716 8609 1.05 79 8688 9953 -1265 J * Clover tops and nitrate added were figured on the basis of 3500 gm. of soil per pot in- stead of 3421, hence are not truly equivalent to the number of pounds per 2,000,000 pounds of soil as the increments were intended. b Average of four determinations or duplicates made at two different times. " Based on weights of soil given in fifth column. ^ Found by analyzing entire crop, roots and tops. • Nitrogen in soil 1072 mgm., nitrogen in seed 45 mgm. , nitrogen in 64.59 gm. clover tops equivalent to 1000 pounds per acre 1750 mgm., and nitrogen as nitrate 86 mgm. great fixation was possible for the fixation of atmospheric nitrogen consists in the utiHzation of this form of nitrogen by the plant for tissue building, and unless plant growth is significant, no marked use of atmospheric nitrogen can be expected. The conditions of this part of the experiment gave too narrow a margin between the possible nitrogen fixed and the limits of variation, and 296 WILLIAM ALBERT ALBRECHT even though increases were shown in some cases, they failed to give assurance of any significant fixation. Unless a reasonable amount of fixation was taking place, the influence of the amount of nitrogen in the soil upon it could not be measured. leo Fig. 1. Graphs Showing Nitrogen Fixation by Soybeans on Soils Treated WITH Nitrates, Clover Tops, and with Nitrates and Clover Tops Series 2 (Cowpeas) For the second crop series cowpeas were grown in the hope that they would be less subject to the infestation by insects, more apt to do well under green- house conditions and more able as nitrogen fixers. Only two soil treatments were employed, one in which increasing amounts of nitrate nitrogen were SYMBIOTIC NITROGEN FIXATION 297 added and the other one in which the nitrogen had been increased by organic matter. Some of the original soil left over when the previous crop series was made up and stored in the dry condition was used for the treatment with nitrate ni- trogen. The equivalent of 3300 gm. of water-free soil was weighed into each pot and enough sodium nitrate added in a solution to give increases in nitrogen equivalent to 10, 25, 50, and 100 pounds of nitrogen in 2,000,000 pounds of soil. For the series whose nitrogen increase was in the form of organic matter, the soils used in the previous series in the pots 1 to 4 and 11 to 20, inclusive, were used again. Pots 1 to 4 had received no treatment with nitrogen and again served as checks. Pots 11 to 20 were also used without modification, save that the amount of soil in each pot was limited to 3300 gm. on the water- free basis. The nitrogen content in these soils no longer showed increments of 1000 pounds per 2,000,000 pounds of soil as a result of losses while growing soybeans, but were used because their differences in nitrogen content were very marked and they were similar in all respects except this one. Plant-foods other than nitrogen were supplied in a soluble form. Calcium carbonate was added at the rate of 2 tons per 2,000,000 pounds. Six cowpeas of uniform weight and known nitrogen content were planted in each pot and later when the plants were well started were reduced to five. The nitrogen added by the five cowpea seeds was equivalent to 43 mgm. During the early spring the pots were kept in the greenhouse but later were kept outdoors in a screened area. No difficulty in germination was experi- enced, and the growth was decidedly better than that of the soybeans. Differences in these two series were soon noticeable. For the series receiving nitrate the plants were taller, the leaves larger and the color deeper where greater amounts of nitrate were added. These differences later disappeared so that by the time of harvest, there were no significant variations within the entire series, either in crop weight or total nitrogen. The soil series treated with nothing but organic matter and which had been previously used for soybeans, gave a decidedly better growth of cowpeas than the series treated with nitrates. The differences are shown in plates 2 and 3. In this series the early growth was best in the pot receiving 1000 pounds of nitrogen per acre as clover tops. There was a decrease in growth with the increase of added organic matter, but even all these were equal to the check, receiving no organic matter. In the latter part of the growing season these differences were reversed, and the largest crop yields and the largest total amounts of nitrogen in the crop were produced on the pots receiving most organic matter. The crops were grown for about 135 days and then harvested. Some plants had produced blossoms and a few had set pods. The fact that the pots were much shaded by some large trees lengthened the vegetative period and delayed seed development. 298 WILLIAM ALBERT ALBRECHT When the crops were harvested the roots were removed as completely as possible and carefully examined for nodules. It is important to note that the nitrate series produced many nodules, even with 100 pounds of nitrogen in this form. Where organic matter was added, the nodules were most numerous in the lesser appHcations but were larger in size as the applications increased. There were no indications in any pots that insufficient nodules were present for nitrogen fixation. If the presence of the nodule is certain evidence that atmospheric nitrogen is used by the plant, then both these soil series permitted fixation to take place in all pots regardless of treatments. Determinations of nitrate nitrogen in the soils in all the pots were made just before the plants were harvested. In the soil treated with sodimn nitrate, no nitrogen in this form was found. The determinations of nitrates gave re- TABLE 13 Nitrogen balance — Cowpeas I Pot series 1, soil treated with nitrates o H O Z 5 S o < O H Bi O H 0< WEIGHT OF CROP NITROGEN IN THE CROP 5S Z-S-r- <^_ Z H O H M H O m W O H "2 « Pi H OT Z °Z U h S m H O o O H o o OS < mgm. gm. mgm. mgm. gm. gm. mgm. mgm. mgm. mgm. mgm. mgm. la None 9. 7673 "^ 3.27S5« 1106^ 15.75 8.855 415 152 1673 1077° 596\ 688 2a None 9.6763 3.7410 1275 14.75 9.435 418 164 1857 1077 780/ 3a 16.5 9.7201 3.3414 1134 14.80 7.820 306 139 1579 1093 486\ 533 4a 16.5 9.7034 3.1594 1074 16.40 7.800 440 159 1673 1093 580/ 5a 41.0 9.6933 3.1163 1061 14.45 7.145 292 147 1500 1118 382 \ 512 6a 41.0 9.7568 3.3418 1130 17.65 7.781 445 185 1760 1118 642/ 7a 82.0 9.7691 2.9373 992 16.15 8.408 404 162 1558 1159 399 \ 453 8a 82.0 9.7215 3.1031 1053 16 65 7.675 445 168 1666 1159 507/ 9a 157.0 9.7441 3.0102 1019 15.45 6.798 347 124 1490 1234 256\ 264 10a 157.0 9.7641 2.9108 983 14.80 5.610 385 138 1506 1234 272/ * Average of three determuiations. ^ Calculated on the basis of 3300 gm. of water-free soil. " Based on 3300 gm. of water-free soil with 0.03136 per cent of nitrogen, or 1077 mgm. of nitrogen, and five cowpea seeds with 43 mgm. of nitrogen. suits no larger than those of the blank determinations. For the series with the organic matter, this treatment interfered with the analytical procedure for nitrates so that no definite statement of the amounts of nitrate present can be made. Results were erratic, but indications pointed to the presence of comparatively large amounts of nitrate nitrogen. Evidently nitrification was going on and nitrates were present in the soil but this had not prohibited the production of many large-sized nodules on the roots of the cowpeas. The harvested crops were thoroughly dried, weighed and later ground to permit uniform sampling. The crop growth was too large to be analyzed in total and only triphcate 2-gm. samples were used. The soil was air-dried, SYMBIOTIC NITROGEN FIXATION 299 ground and sampled. Analyses were made by the same methods used in the previous series, but instead of using four samples, only triplicate determina- tions were made. On the basis of these analyses, the total nitrogen in the soil and the crop at the close of the series was calculated. Tables 13 and 14 give the nitrogen balance for the second crop series with cowpeas. Only the summations of analytical data are given. Since the soils in pots 1 to 4 and 11 to 20 were analyzed at the close of the soybean series no analyses were made on these soils at the beginning of this series, but the amount TABLE 14 Nitrogen balance — Cowpeas I Pot series 2, soil treated with clover tops z w a o o H g ■is s O H 1' WEIGHT OF CROP NITROGEN IN THE CROP 5^ W M O O S M O o tJ Www g§ + s w-2 5 1 a. o H o o PS a. o o Pi Ihs. gm. mgm. mgm. gm. gm. mgm. mgm. mgm. mgm. mgm. 1 None 9.7140" 3.5308'' 1199>' 21.70 6.976 604° 135 d 1938 1006" 9321 2 None 9.7324 3.4744 1178 19.65 6.040 559 126 1863 1041 822 j 3 None 9.7423 3.3086 1120 18.75 7.170 504 157 1779 1045 734\ 4 None 9.7026 3.2257 1097 21.90 10.182 502 178 1777 1012 765/ 11 995' 9.7421 6.8328 2314 34.75 11.705 1066 201 3581 2641 941 \ 12 995 9.7419 7.0649 2393 35.30 9.525 1032 205 3630 2744 886J 13 1960 9.7324 9.5746 3246 33.00 10.550 977 192 4415 4206 2091 14 1960 9.7412 9.5912 3249 39.75 12.711 1125 209 4576 4317 259/ 15 2890 9.7462 11.5307 3904 39.20 11.959 un 204 5325 5255 70l 16 2890 9.7514 11.6965 3958 36.45 12.270 1124 226 5308 5541 -223] 17 3790 9.7549 13.6027 4601 43.65 9.975 1256 153 6010 6786 -776\ 18 3790 9.7468 13.4631 4558 37.25 8.347 1290 189 6037 6526 -489/ 19 4660 9.7461 15.6151 5287 45.70 10.255 1594 125 7006 7875 -8961 20 4660 9.7516 15.9400 5394 41.45 8.215 1386 153 6933 7891 -958J mgm. 877 749 913 234 -81 -632 -927 " Average of three determinations. ^ Calculated on the basis of 3300 gm. of water-free soil. " Found by analyzing three 2-gm. samples and calculating from the weight of tops. ^ Found by analyzing the entire root system. ® Based on analysis made at the close of the soybean series (data in table 10). ' The additions of nitrogen as clover tops made on these soils previous to the growth of soybeans are given as pounds of nitrogen added per 2,000,000 pounds of the mixture of soil and clover tops. of nitrogen in 3300 gm. of soil was calculated from the preceding analyses. The soil which had been stored and used in this series was analyzed and found to contain the same amount of nitrogen as at the beginning of the soybean series, or 0.03136 per cent of nitrogen on the water-free basis, which gave 1034 mgm. of nitrogen per pot. The crop weights and the amounts of nitrogen in the crop are given separately as tops and roots. 300 WILLIAM ALBERT ALBRECHT Following the tables is a graphical representation of the nitrogen balance in the series (fig. 2). In the series treated with sodium nitrate the amounts of total nitrogen present at the close were larger in every pot than those present at the outset. 1000 800 600 400 N 200 200 400 600 800 TT 10(10 2r aoDo 60 51 )90 Po soi IL 40 30 ids ni trea trog ted 1 50,30 en per with c ^acre li Lover tcps. IDO \ Potmds nitrogen adjded per X acre as nitrate Fig. 2. Graphs Showing Nitrogen Fixation by Cowpeas on Soils Treated WITH Nitrates, and with Clover Tops This indicates decidedly that nitrogen was fixed from the atmosphere. The amounts so obtained were very significant, and in every case far above the limits of variation in analytical determinations. The average nitrogen in- -.n>^ SYMBIOTIC NITROGEN FIXATION 301 crease, or nitrogen fixation, from duplicate pots shows a gradual decrease with increase in nitrate nitrogen added to the soil. The largest fixation appeared in the pots whose soil received no nitrogen additions and was reduced about the same in the two soils treated at the rate of 10 and 20 pounds of nitrate nitrogen per 2,000,000 pounds of soil. On the soil treated with the equivalent of 100 pounds of nitrate nitrogen the fixation was about 38 per cent of that where no treatment was applied. The general decrease in fixation with in- creased treatment would lead one to believe that one or two more increments beyond the appHcation of 100 pounds would have prohibited fixation com- pletely. The table indicates that increasing amounts of nitrate nitrogen in the soil reduce nitrogen fixation by cowpeas, and may perhaps prohibit it, but the amounts required for significant reduction of this process are far greater than ever occur in a soil or are ever applied. The series treated with organic matter and previously used for growing soy- beans had some pots which failed to show any increase in total nitrogen at the close over that at the outset. Significant increases were obtained for the check pots and those originally treated with the equivalent of 1000 and 2000 pounds of nitrogen per acre, but for the rest of the treatments a negative fixation, or loss, took place. This was greater with larger applications of ni- trogen. It is highly probable that the organic matter put on the soil in such heavy applications one season previously was still undergoing decomposition rapidly enough to lose nitrogen as ammonia or as gaseous nitrogen. Decom- position had not yet gone far enough to change all the organic matter of such heavy applications into a more stable form, from which no losses could take place. Such losses from the total nitrogen present in soil and crop at the close, as compared with that present in soil and seed at the beginning, do not prove that the plant failed to draw on the atmosphere for some of its nitrogen supply. Nodules were plentiful in the soils receiving the heaviest applications of organic matter and some use may well have been made of gaseous nitrogen. Any fixation that could have taken place was offset by the losses from the soil, and could not be detected in this method of determination. Considering the averages for duplicate pots, the amounts fixed in the check pots 1 and 2, which had previously grown the uninoculated soybeans, were larger than that fixed by pots 3 and 4, on which the soybeans were inocu- lated. This difiference between the two sets of check pots resulted from a larger amount of nitrogen being present in both the crop and the soil in the first two pots. The reverse was true, however, of the total crop weights, largely because of greater root development in pots 3 and 4. In pots 11 and 12 the average fixation was the largest in the series. This would indicate that the treatment with organic nitrogen corresponding to 1000 pounds per acre was beneficial to nitrogen fixation by cowpeas in a soil so low in this element as this soil was. This statement cannot be made, however, for all the higher applications. A decided increase was shown by the pots receiving 2000 pounds of nitrogen per acre, but those receiving more showed a loss rather than a gain. 302 WILLIAM ALBERT ALBRECHT Evidently the lower applications do not prohibit nitrogen fixation, but whether the larger ones prohibit the process cannot be said from the preceding data. The losses coming doubtlessly from the decaying organic matter in the soil more than balance any nitrogen from the air added to the plant. In terms of total nitrogen in the soil rather than nitrogen added, those pots showed nitro- gen fixation whose soils contained the equivalent of approximately 625, 1625 and 2600 pounds of nitrogen per 2,000,000 pounds. From this it may be safely said that a soil with these amounts of nitrogen does not prohibit the symbiotic nitrogen fixation even when 2000 pounds of nitrogen were in a readily decomposable form. The soils with higher nitrogen content failed to show similar results because the large amount of nitrogen applied to the soil was partly lost in a volatile form. Series 3 (Cowpeas) The third crop series was very similar to the second, with a few modifica- tions. It included one series of pots with soils treated with nitrates and an- other series with soils treated some time previously with organic matter in the form of clover tops. For treatment with nitrates some soil saved by careful storage from the first crop series of soybeans was used. Sodium nitrate was added in solution to increase the nitrogen content at increments equivalent to 50, 100, 150, 200 and 250 pounds of nitrogen per 2,000,000 pounds of soil. Two additional pots receiving the equivalent of 50 pounds of nitrate nitrogen were included, but this was added in three applications at intervals extending over the greater part of the growing period. Three kilograms of soil were used per pot. For soils whose nitrogen was increased by organic matter, those pots used in both preceding crop series were used again without change, save that the amounts of soil were less, and only 2800 gm. of water-free soil were weighed into each pot. The nitrogen content of these soils did not increase in constant amounts for the series, but corresponded very closely to amounts equivalent to 625, 1625, 2600, 3250, 4025 and 4775 pounds of nitrogen per 2,000,000 pounds of soil. These soils had been made up originally with enough clover tops to give a series whose nitrogen content increased by units of 1000 pounds, but the losses during the two preceding crop series reduced it to the above figures. It was thought advisable to use these soils again since their nitrogen content was accurately known, and the growth of two crops allowed sufficient time to permit decomposition of the organic matter to have gone far enough to prohibit further losses in volatile nitrogen. The soils were treated with calcium carbonate at the rate of 2 tons per 2,000- 000 pounds of soil. Phosphorus, potassium and other mineral plant-food elements were applied in solution when the dry soil was first moistened and then at intervals during the crop growth. All soils were treated similarly in this respect. Attempts were made to keep the moisture content of the soil SYMBIOTIC NITROGEN FIXATION 303 at optimum by weighing the pots, but this method was inaccurate when the plants became larger, so that water was applied at intervals to keep the mois- ture content near what seemed optimum. Cowpeas were grown as the crop. Six seeds weighing in total an average of 1.109 gm. and containing 38 mgm. of nitrogen were planted in each pot. As the seedhngs were well started, one was removed to leave five vigorous plants. The growth of the crop was good. No difi&culty was experienced in germi- nation, and though the early growth was somewhat slow, the remainder of the growing season found the plants doing well. Differences due to treatment were manifested early. With increased amounts of nitrate applied the plants were taller, of a deeper color and the leaves more fleshy, but these marked differences disappeared later. The plants on soils treated with organic matter were larger and of deeper color as the nitrogen content of the soil increased. These variations became less prominent and were completely lost before the close of the experiment. The total plant growth was similar to that in the preceding crop series, but with less variations. Plate 4 illustrates the crop growth of the two series just before harvest, which was about 22 weeks after planting. Shortly before the crops were harvested, several blossoms were formed but were removed to keep these few plants from setting seed. All blossoms and fallen leaves were collected and saved for analysis. The harvested plants were dried in the greenhouse, and later weighed and thoroughly ground for analysis. The roots were carefully removed, examined for nodules and thoroughly dried. No great differences in nodule production were evident as correlated with the treatment. There were no significant differences shown by the nod- ules in the soils treated with nitrates. Nodules were numerous in all pots and especially so in those receiving the equivalent of 250 pounds of nitrogen per 2,000,000 pounds of soil. There was no evidence in either of the duphcate pots that the appHcation of such large amounts of nitrate interfered with nodule production. This application was equivalent to 12.5 mgm. of nitrogen per 100 gm. of soil and its failure to suppress nodule growth agrees well with the results of Fred and Graul (17) who found that 10 mgm. of nitrate nitrogen per 100 gm. of Miami silt loam were required before nodules of alfalfa and crimson clover were decreased by such treatment. Evidently the application in the yellow silt loam used in this case was still less than was required for serious effect on nodule production. Nodules were present in all pots treated with organic matter and gave indi- cations of being larger and better distributed within those soils receiving the heaviest apphcations This may have been due to the better physical condi- tion and greater aeration caused by such treatment. These results are not in accord with those found by Frank (14) for red clover, on which the nodules were fewer and smaller in the soil rich in organic matter. 304 WILLIAM ALBERT ALBRECHT At the close of this series, determinations of nitrate nitrogen were made on all the soils. Those treated with nitrate failed to show any nitrogen of this form present in the soil. This result is quite different from that found for soybeans, in which case an application of 258 mgm. per pot left much nitrate in the soil. The total nitrogen in the crop of soybeans was only 150 mgm. With the profuse growth of cowpeas in this series the application of 390 mgm. of nitrogen as nitrate was far below the nitrogen in the crop and might well be expected to be removed on account of its solubility. The removal was so complete that no more than a trace of nitrate was found in any of the soils. TABLE 15 Nitrogen balance — Cowpeas II Pot series 1, soil treated with nitrates A g ^ H 55 O WEIGHT OF NITROGEN IN H^ H O B) < V. HO N __^_-N<,N;]^«b ItmlM 321 PLATE 2 •CowPEAS, AT 50 Days, on Soil Treated with Nitrates (Above) and Cowpeas AS THE Second Crop on Soil Treated with Clover Tops (Below) 322 SYMBIOTIC XTTROGEX FTX.\TIOX WILIIAII ALBERT AISSXCHT PLATE 2 '%S^ ^f<^ "^^^ **T>^ ^ son. scEzxci:, tol. ix, xo. o PLATE 3 CowPEAS, AT 75 Days, on Soil Treated with Nitrates (x^bove) and Cowpeas AS the Second Crop on Soil Treated with Clover Tops (Below) 324 SYMBIOTIC NITROGEN FIXATION WILLIAM ALBERT ALBRECHT PLATE 3, "* Tlfli^' 325 PLATE 4 CowPEAS ON Soil Treated with Nitrates (Above) and Cowpeas as the Third Crop on Soil Treated with Clox^er Tops (Below) 326 SYMBIOTIC NITROGEN FIXATION WILLIAM ALBERT ALBRECHT PLATE 4 327 BIOGRAPHICAL SKETCH William Albert Albrecht was born on a farm near Flanagan, Livingston County, Illinois, September 12, 1888. He secured his common-school educa- tion in the district school, and his preparatory work in the academy of Bluff ton College, Bluffton, Ohio. In the fall of 1907 he entered the University of Illinois, College of Liberal Arts and Sciences, from which he received the degree of Bachelor of Arts in 1911. The following year was spent as teacher of Latin in Bluffton College, Bluffton, Ohio, after which he returned to the University of Illinois to enter the College of Agriculture, where he received the degree of Bachelor of Science in 1914. During that year he was granted a scholarship and received the degree of Master of Science. In 1915 he was a fellow in agronomy in the University of Illinois, enabling him to pursue graduate work during that time. In the fall of 1916 he went to the University of Missouri, College of Agriculture, where he is now associate professor in soils. He is author of the following papers: "Changes in the Nitrogen Content of Stored Soils," published in the Journal of the American Society of Agronomy, Vol. 10, February, 1918, and "Soil Inoculation for Legumes," accepted for publication as Missouri Agricultural Experiment Station Circular 86. He is a member of the following societies: Gamma Alpha, Alpha Zeta, Sigma Xi, American Society of Agronomy, and Society of American Bacteriolo- gists. 002 756 447 1