SOIL NETROGEN AND CARBON FRACTIONS AS RELATED TO QRGANIC AMENDMENTS AND RESPONSE OF OATS IN THE GREENHOUSE Thesis for flu Degree of Dh. D. MICHEGAN STATE {ENEVERSITY George W‘iiliam Wright 1964 IHESIS This is to certify that the thesis entitled Soil Nitrogen and Carbon Fractions as Related to Organic Amendments and Response of Oats in the Greenhouse presented by George William Wright has been accepted towards fulfillment of the requirements for Ph.D. degree in Soil Science MWA M74 Major professor Date Febm 251 1961! L I B R A R Y Michigan State University -‘ ‘ 1-1: .I . ABSTRACT SOIL NITROGEN AND CARBON FRACTIONS AS RELATED TO ORGANIC AMENDMENTS AND RESPONSE OF OATS IN THE GREENHOUSE by George William wright Multiple regression analysis was used to relate short term and long term responses of oats in the greenhouse to experimentally imposed variations in several chemical fractions of carbon and nitrogen in soil. Lignin was isolated from hardwood sawdust by acid treatment to re- move carbohydrates. Oshtemo sand was amended with lignin or sawdust at rates calculated to supply 2% tons per acre of lignin in each material (9 3/4 tons sawdust, containing 25.6 percent lignin). Urea was added to amended and control soil at rates equal to O, 200 and #00 pounds N per acre. Nitrogen additions were calculated to give C:N ratios fer sawdust of 40:1 and 2021. Mineral salts were added to assure adequate supplies of major nutrients other than nitrogen. After 5 weekS' incubation in the greenhouse, growth and nitrogen uptake by cats were determined in triplicate short term (11 days) and long term (54 days) experiments. Similar determinations were made ' again after 15 weeks, using soil from previously cropped and previously uncropped series of pots. Soil samples were taken just before planting each crop. Soil nitrogen and carbon were fractionated into water-soluble, exchangeable, acid;hydrolyzable and non4hydrolyzable forms. Nitrate, other water-soluble nitrogen (presumably ammonium), and exchangeable nitrogen were extensively immobilized in the presence of George William wright sawdust. The immobilized nitrogen appeared primarily in the hydrolyz- able fraction. Where urea was added with the sawdust, peak immobiliza- tion had occurred by the fifth week. Where no nitrogen was added with sawdust, and in the presence of root residues in previously cropped pots, additional increases in hydrolyzable nitrogen were observed at 15 weeks. Essentially no immobilization of nitrogen occurred with lignin. although small and statistically significant increases in non-hydrolyz— able nitrogen, increasing with level of urea addition, occurred with both lignin and sawdust. Multiple regression analysis indicated that nitrate, other water- soluble and exchangeable ferms of nitrogen were utilized by the first long term crop of cats and that increasing levels of hydrolyzable carbon seriously reduced the availability of all three forms. In the second long term crop, variations in growth and nitrogen uptake were determined largely by the level of nitrate present at planting time, although a significant contribution from the hydrolyzable nitrogen fraction was also indicated. In the short term growth experiments, nitrate and water-soluble materials were removed by leaching just before placing the soil in contact with the root mat of 15-day old, nitrogen deficient oat plants. In the first experiment, 5 weeks after soil amendment, the production of dry matter was stimulated independently of nitrogen uptake by some factor or factors present in both sawdust and lignin. Multiple regres- sion analysis associated the stimulating activity primarily with the non-hydrolyzable carbon fraction. Nitrogen uptake reflected this George William Wright stimulus but was additionally influenced by variations in eXChangeable forms of nitrogen. In the second short term experiment, 15 weeks after soil amendment, similar stimulating effects of lignin and sawdust were observed. There also appeared to be a specific inhibitory effect of root residues in previously cropped soil. In this second crop, neither the stimulatory nor the inhibitory effects were clearly associated with any single fraction of carbon or nitrogen by relationships defined in the regres- sion function. SOIL NITROGEN AND CARBON FRACTIONS AS RELATED TO ORGANIC AMENDMENTS AND RESPONSE OF OATS IN THE GREENHOUSE By GEORGE WILLIAM WRIGHT A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Soil Science 1964 ACKNOWLEDGEMENTS The author wishes to express his sincere thanks to Professor A. R. Wblcott fbr encouragement and untiring help during this study. His sincere appreciation goes also to Mrs. Arlene King of the M. S. U. Computer Center. Additional acknowledgement is due Dr. Cook who made it possible for this work to be carried out. He is also grateful to his family for their patience and coopera- tion during the course of this work. ii I. II. III. V. VI. TABLE OF CONTENTS m TRO DU CTION I O O 0 O 0 O O O O O O O O O O O O O O O O 0 mm OF LITEMTUM O O O O O O O O O O O O O O O O O O 0 WT“ mom 0 O O O O O O O O O 0 O O O O O O O O A. B. C. D. Design of Greenhouse Experiment . . . . . . . . . . . Short Term Oat Response Study . . . . . . . . . . . . Fractionation of 8011 Carbon and_Nitrogen . . . . . . StatiStical Treatment 0 o o o o o o o o o o o o o o 0 WT“ “SULTS . 0 O O O O O C O O O O O O 0 C O O O A. B. C. D. Effects of Treatment on Soil Variables . . . . . . . . Long Term Responses of’Oats . . . . . . . . . . . . . Short Term Responses of Oats . . . . . . . . . . . . . Functional Relationships among Crop Variables . . . . 1. Short term relationships . . . . . . . . . . . . . 2. Long term relationShips . . . . . . . . . . . . . Functional Relationships among Crop and Soil Variables 1. Intercorrelations among soil variables . . . . . . 2. Short term crop and soil variables . . . . . . . . 3. 'Long term crop and soil.variables, . . . . . . . . SWY AND CONCLUSIONSa O O I O O 0 0 O O O O O O O O O O BIEIOGWY O O O O O O 0 O O O O O O O O O O O O O O O 0 iii 17 17 21 22 24 26 26 53 53 58 ‘ 64 73 78 10 11 12 13 14 15 16 LIST OF TABLES Description of residue and nitrogen treatments 0 . . . . . . Effects of residue and nitrogen treatments on soil variables measured prior to the first cropping period. . . . ’ Nbin effects of residue treatments on measured SOII variables. 0 e o e e e o e e e e e o o e e e e o e e e Nbin effects of nitrogen treatments on measured soil variables 0 e e e e o e o e o o e e e e e e e e o o e e e 0 Effects of previous cropping, residues, and nitrogen treatments on soil variables measured prior to the second cropping periOd e e e e e e e o e o e e e o o e e e 0 Effects of residues and nitrogen treatments on long term crop variables for first cropping period . . . . o . . . . . Effects of residue treatments on measured long term crop variabICSe O O O O O O O O O O O O O 0 O O 6 O O 0 O O O O O Whin effects of nitrogen treatments on measured long term crop variables 0 O O O O O O O 0 C O O 0 0 O 0 O O O O O O 0 Effects of previous cropping, residues, and nitrogen treat- ments on long term crop variables measured in the second cropping period . . . . . . . . . . . . . . . . . . . . . . Effects of residue and nitrogen treatments on short term crop variables for first cropping period . . . . . . . . . . Effects of residue treatments on measured short term cr0p variables. 0 O O O O O O O O O O O C O O O O O O O C O O O 0 Effects of nitrogen treatments on measured short term crop variables. 0 O O O O O O O O» O O O O O O O O I O O O O O O 0 Effects of previous cropping, residues, and nitrogen treat- ments on short term crop variables measured in the second cropping period... 0 e e e e e e e e e e o e e e e e e e e 0 Linear, semi-log, and log correlations among soil variables measured prior to the first crepping period. . . . . . . . . Linear, semi-log, and log correlations among soil variables measured prior to second cropping period . . . . . . . . . . —2 Coefficients of multiple determination (R ) for short term crop variables as functions of soil variables in different COlnbinations O O O O O O 0 O O O I O O O O O O O 0 O O O O O 8.. 30 31 32 35 37 37 38 4O 42 42 43 54 56 59 ‘ —2 Coefficients of multiple determination (R ) for long term crop variables as functions of soil variables in different COUlbinationS O O O O O O O O 0 O O O 9 O O o 0 9 65 UIG LIST OF FIGURES Functional relationships among short term crop variableSo e e o e o o e e o o e o o e e o 0 Functional relationships among long term crop variables 0 e o e e o e o e e e o e o e e o e e 0 Functional relationships among first short term crop variables and soil variables 0 . o . . . . . Functional relationships among long term crop variables and soil variables. . . . o . . . . . . Dry weight of the first and second long term oat crops vs. nitrate nitrogen of the soil at planting time . . . . . . . . . . . . . . . o . . vi 5O 61 67 71 INTRODUCTION Agronomists have been attempting to find ways for estimating the availability of soil nitrogen from the time it was learned that most of the nitrogen taken up by plants is obtained from the soil. The mineralization process by which organic nitrogen is converted into available mineral forms has been rather well defined since about 1890. but as yet, no satisfactory method has been devised for determining the nitrogen supplying power of a soil under field conditions. This fact gives mute evidence to the complexity of nitrogen compounds and nitrogen transformation processes occurring in the soil. There are several reasons why the nitrogen supplying power of a soil is so difficult to assess. If one attempts to isolate the organic from the inorganic fraction of the soil, rather vigorous methods must be employed. These may cause unexplained and often unappreciated changes in the labile organic fractions. Even if the organic fraction could be separated without drastic changes, not enough is known about pits chemistry to translate quantitative measurements into expected rates of release of nitrogen under varying soil and climatic condi- tions. Another factor that leads to difficulties in estimating the availability of soil nitrogen is the immobilizing effects of root residues which remain in the soil or of surface residues which may be incorporated into the soil. Other effects of added residues may be wholly unrelated to nitrogen supply. such as effects on moisture or oxygen supply, soil structural effects, and growth regulatory effects which may be either stimulatory or inhibitory. Biological assays are at present the most userI methods for estimating the nitrogen supplying power of soils. Among these biologi— cal assays are various greenhouse procedures for measuring short term growth or uptake of nitrogen by indicator plants. The objective of this investigation was to attempt to relate, empirically by multiple regression analyses, short term and long term responses of oats in the greenhouse to experimentally imposed variations in several chemical fractions of carbon and nitrogen in the soil. It was hoped that, by measuring both carbon and nitrogen in different fractions. it might be possible to estimate both the potential nitrogen supplying power and the immobilizing potential of carbonaceous constituents. REVIEW OF LITERATURE Norman (31) leads off his chapter on lignin with the following statement. "Despite a great volume of research many aspects of the Chemistry of lignin remain obscure and controversial". Part of the difficulties encountered in the study of lignin lies in the fact that it has been difficult to separate lignin from asso- ciated plant constituents without seriously altering the lignin itself. Thus the bulk of the work done on lignin.has not been carried out on an unaltered pure substance. Norman goes on to criticise the assumption that "lignin found in nature is more or less unifbrm and homogeneous". He is led to this conclusion since much literature suggests that lignin derived from different plant species has a different character. Even lignin derived from the same species of plant at different physiologi- cal ages may be separated into fractions that differ in their properties. The view has been expressed that there is no compound of lignin as such, but that what has been called lignin is the product of'methylated carbo- hydrates that have been altered by the action of organic acids found in the soil. This view has not received wide acceptance. Lignin is generally considered to be a complex mixture of compounds with similar structure but differing in side chains or substituting groups or in chain length or degree of polymerization. Lignin is referred to by Fraser (21) as a polymer formed mainly from units of substituted derivatives of phenylpropane. It has been deduced that lignin derived from plant materials as they break down forms the major constituent offlhumus. This cannot be accepted as an 'establiShed fact until more is known about the structure of lignin from various sources. Ensminger and Pearson (19) make some reference to the composition of organic residues. They suggest that oneuthird to oneehalf of the organic matter found in the soil is microbially derived, if plant de- composition takes place under aerobic conditions. It is stated that lignin or lignin-derived materials make up a.major portion of the organic fraction of soils. Gottleib and Hendricks (22) suggest that soil lignin is very similar to "alkali lignin" and that under natural conditions lignin in the soil undergoes transformations similar to lignin treated with alkali. Brauns (6) has quoted Hagglund as follows: ”In no other field of chemistry does there exist such a discrepancy in opinions as to the constitution of a substance as there is on lignin." Concerning the discrepancies that are found in lignin literature, Brauns goes on to say “by far the principal cause is the fact that lignin is one of the most complicated and elusive natural products known to chemists.“ In an attempt to clarify some of the confusion, Brauns has defined the terms ”lignin building stones" and ”lignin building units". The use of these terms is based on the assumption that the lignin molecule is made up of units in a manner somewhat similar to the units which make up starch and related compounds. The ”building stones" of lignin are made up of a phenylpropane carbon structure. Four or five of these building stones are linked together to ferm a lignin ”building unit". A.number of lignin building units are combined to make up the lignin molecule. Investigators who have attempted to determine the molecular weight of lignin have reported values all the way from 250 to 11.000. These values vary with the type of lignin, and also with the same lignin preparation using different methods for estimation. At the present time, the most widely accepted molecular weight for lignin is 840. This corresponds to 5 building stones as a lignin building unit. It should be pointed out that this is in no means a final accepted figure, but it does give a basis for fUrther work. Broadbent (11+) has pointed out that as plant residues are added to the soil. the cellulose and.hemicelluloses are readily broken down by the soil microorganisms whereas the lignin is quite resistant. For this reason, the relative proportion in the residue increases. This supports the theory that a large part of the soil organic matter is either lignin or lignin-derived. This is further supported by the simi- larities between a soil fraction and lignin as they are dissolved in alkali and precipitated by the addition of excess acid. Gottlieb and Hendricks (22) were unable to isolate definitely characterizable products from soil organic matter as they had been able to from wood lignin, using the same techniques. This would indicate that changes had taken place in plant lignin during decomposition proc- esses in the soil. These two workers hypothesized the preduction of fused ring structures through a condensation of demethylated molecules. Another factor that points to the dissimilarity between wood lignins and soil organic matter is that wood lignin usually contains more than 60% carbon while the lignaceous fraction of surface soils rarely contains more than 52%. The carbon content in the deeper layers is even less than this. Broadbent (14) suggested that more emphasis should be placed on.microbial processes involved in altering soil organic matter. How- ever, he feels that the methods that have been used to gain more understanding of wood lignin should continue to be used as an aid to understanding the lignaceous fraction in soil organic matter.‘ It has been pointed out by Rodrigues (36) and Bremner (10) that some of the nitrogen in the soil that had been thought to be in organic combinations. was actually present in the form of ammonia trapped in the clay lattice. Waksman and Iyer (43, 1+4, 1+5, 46) postulated the existance of resistant nitrogen compounds which they termed "ligno- proteins”. Allison (2) stated that the nitrogen found in humus is very heterogeneus and probably consists of proteins, aminoasugars, nucleic acid, chitin, heterocyclic compounds, and complexed lignouprotein come pounds. Mattson and KoutlerGAndersson (29) studied the fixation of'ammonia by lignin and feund that the greatest fixation occurs under aerobic conditions with simultaneous oxidation and that the fixed ammonia is very resistant to acid and fairly resistant to alkaline hydrolysis. They observed this type of fixation of ammonia liberated by bacteria during the breakdown of nitrogenous organic matter. Parker gt_§l, (33) applied several materials on the surface or incorporated them into the soil with and without nitrogen fertilizer. Incorporation without fertilizer decreased nitrogen uptake while surface application had little effect. From their data using corn as an in- dicator crop, it seemed that adequate nitrogen fertilizer incorporated with the residue gave good results, however, no better than the surface application. The use of sawdust on crops was studied by White gt’gl, (49). Sawdust was used both as a mulch and incorporated into the soil. They feund that, as a.mulch, it improved crop growth due to moisture con- servation. However, it did not show beneficial effects whenincorpo- rated. Both types of application caused a decrease in soil nitrate, and, in all cases where sawdust was used, extra additions of nitrogen fertilizer were necessary. 7 Based on the fact that as much as 30% of the total soil nitrogen is resistant to acid or alkali hydrolysis, Bremner (7) suggested that much of the nitrogen present in the soil is nonmprotein in nature. He 'was of the opinion that part of the organic nitrogen was present as heterocyclic compounds. In some humic acid preparations, as much as 60% of the nitrogen was not released by hydrolysis with strong acids (9). ‘Wbrk conducted by Dyck and McKibbin (18) indicated that not all organic nitrogen was determined by the Kjeldahl method. Bremner (7) suggested that this discrepancy could be due to heterocyclic nitrogen compounds containing direct nitrogen to nitrogen linkages. Later Bremner (11) studied the Kjeldahl method of nitrogen determination in great detail. He compared a number of the different modifications and found that the method as outlined by the Association of Official Agricultural Chemists. gave results that were from 20 to 37% low as compared with other methods. He further stated that if the digestion time were increased, satisfac- tory results could be obtained. In this same paper, Bremner suggested that there are few or no highly refractory nitrogen compounds containing N-N or N-O linkages in soils. This refutes some of his earlier state- ments (7). Bremner and Shaw (12) reported that in some soil humic acid prepa- rations, nitrogen was released at a rate intermediate between that of lignin-ammonia and that of’lignin-protein complexes. They proposed that neither the lignin-ammonia theory of'Mattson and Koutler-Andersson (29) nor the lignin-protein theory as advanced by waksman and Iyer (43, 44) was adequate and that the best explanation of the observed properties of the humic acid complex fractions of the soil could be explained better by combining both of these theories into a single concept. Stevenson (40) disagreed with Bremner (8) in that he did not believe that amino acids isolated from the soil were necessarily struc- tural constituents of proteins. He went on to suggest that the amino acids can be protected from microbial decomposition by adsorption on clay surfaces or by association with soil organic colloids. His data were in agreement with the concept that humic acids may fix amino com- pounds in the soil. Mortland and'Wolcott (30) have recently made an extensive review of the literature dealing with the fixation of inorganic nitrogen by the soil constituents. A number of'methods have been used in the investiga- tion of soil nitrogen compounds, including eray diffraction, kinetic and thermodynamic adsorption studies, and infrasred adsorption. From these studies, a number of concepts have been fermulated. The ammonium ion is formed when ammonia is adsorbed by either hydrogen or base-saturated clay minerals. Coordinate covalent amine compounds may be formed if good electron acceptors are present on the exchange complex. Ammoniate combinations, analogous to hydrates. may also be fermed. Another way in which ammonia or ammonium may be held in the soil is by'hydrogen bonding. Simple physical bonding may be active in the adsorption of ammonia, but this is of a transitory nature and a stronger type bonding is soon formed. The stability of these nitrogen compounds ranges from water solu- ble ferms to complex heterocyclic compounds from which nitrogen is not released by hydrolysis with strong acids at high temperatures. A large number of the soil organic matter reactions, including the fixation of nitrogen, are responsive to the oxidation-reduction status of the soil. Relationship between redox potentials and decom- posing soil organic matter are very difficult to fellow due to the complexity of decomposition processes and the many different factors 9 that cause the redox potential to vary. Mortland and'W01cott went on to suggest that it was probable that the incorporation of nitrogen into complex heterocyclic compounds was accidental in that, when nitrogen was present, nitrogen bonds might form between adjacent molecules during condensation and polymerization where, in the absence of nitrogen, oxygen bonds would be expected. Under alkaline conditions, ammonia fixation takes place rapidly and extensively. The rate and amount of fixation is markedly reduced by lowering the pH of the system. At a low pH the slow fixation that continues is probably due to the continuing exposure of active sites by enzymatic or mineral catalysis of oxidative reactions. Gaseous nitrogen compounds at an intermediate stage of oxidation- reduction, such as N20 and N have been observed in systems involving 2. the breakdown of organic matter. It has been suggested that a reaction similar to the Van Slyke reaction may take place in the soil. Such an overall reaction might proceed through stepewise reactions, involving intermediate complexes with clays or organic matter, so that the re- action normally observed under drastic laboratory conditions could take place in a field soil. A factor which may promote such stepcwise reactions is the polarizing and catalytic effect that clay mineral surfaces exert on organic and inorganic reactants in soil. Such catalytic reactions have been demonstrated under controlled conditions using silica and hydrated iron oxides. In the breakdown of organic matter, as described by waksman and Iyer (#3), the carbon is used as a source of energy by soil organisms, with part of the carbon and a corresponding proportion of nitrogen being combined in the synthesis of microbial tissue. Part of the nitrogen is thus held in the soil in the form of proteins. As the 1O organisms die and are subject to decomposition, the protein nitrogen is released as ammonia, which is oxidized to nitrites and nitrates. Harmsen and Van Schreven (23) pointed out that the liberation in mineral form of organically combined nitrogen by microbial Processes in the soil has been recognized since 1893, but these processes have undergone continuing study in an effort to better understand their impact on plant nutriticn. The overall process by which organically combined nitrogen in the soil becomes available to plants has been termed "mineralization". It has been considered to take place in two steps. The first step, termed "ammonification", involves the degradation of protein to give rise to ammonia. The second step, called "nitrification”, is the bacterial oxidation of the ammonia, first to nitrites and then to nitrates. Pinck‘gt‘gl. (35) studied nitrogen availability in soils treated with from O to 4 tons of straw per acre. They feund that residue ad— dition reduced nitrogen availability. Applied nitrogen fertilizer com- pletely counteracted the effect of the residue. An incubation period following the addition of the residue increased the availability of the soil nitrogen. Similar effects have been reported by others (4, 13. 33). The recovery of fertilizer nitrogen from soils to which corn and alfalfa residues had been added was studied by Bartholomew and Hiltbold (5). Both materials were ground to pass a 10 mm. seive and #% g. of plant material were added to pots containing 1800 g. of soil. Part of the pots were incubated and plants were grown on others. After the incubation or growth period, the entire 1800 g. of soil were extracted with normal sodium chloride at pH of 1.0.. Aliquots of the extract were analyzed fer mineral nitrogen. From their study, 11 they concluded that plant yields and nitrogen uptake were reduced by the addition of the corn residue. The total recovery of added ferti~ lizer nitrogen ranged from 27 to 54%. Bartholomew (4) presented information showing that decomposition of organic matter and the release of available nitrogen was dependent on such things as crop, cropping sequence, soil texture, structure, and moisture conditions during the actual growing season of the crop. It has been reported by WOodruff (52) that the nitrogen uptake from a particular soil was influenced by the type of crop grown. He reported twice as much nitrogen delivered to corn as to a small grain. Che might consider here that these two crops grow at different seasons, and that corn growing during the warm season would be expected to have a greater uptake than a cool season crop such as the small grain due to seasonal differences in rates of mineralization and plant metabolism. Smith (39) feund that soil texture was an important modifier of nitrogen release from the soil. He reported that a clay or clay loam would release 1% to 2%% of its total nitrogen in one season, a silt loam 1% to 3%, and a sand or sandy loam 4 to 5%. He also stressed the important influence of moisture and temperature. Harmsen and Van Schreven (23) reviewed the work of Norman and Werkman (1943), Broadbent and Norman (1947), Broadbent (1948), Broadbent and Bartholomew (1949). and Thornton (1946) in their dis- cussion concerning the claims that the stable portions of the soil organic matter were subjected to fresh attack by microorganisms upon the addition of a fresh organic residue such as a green manure crop. Lbhnis had reported such findings as early as 1926. The conclusion drawn by Harmson and Van Schreven was that the addition of fresh energy material to the soil caused such an increase in numbers and activity of 12 the soil microbial population that some of the "stable" organic matter was broken down along with the fresh material. They suggested that the ”stable” humus was much less stable than.had been supposed, and that its rate of decomposition was due in part to insufficient microbial activity. The addition of fresh material was spoken of as a "forced draft on the smoldering bacterial fire". These conclusions support the opinion of Broadbent (13) that a large part of the nitrogen taken up by a crop following a green manure crop has its origin in the ”stable" soil organic fraction rather than the freshly added plant residue. Harmson and Van Schreven (23) cited the work of Barnes (1953) on the mineralization rate of plant residues added to the soil. During the incubation in soil of'mustard, vetch, and grain straw, only straw, with its wide carbonetomnitrogen ratio, inhibited the mineralization of nitrogen fer the entire 89 weeks of the incubation experiment. waksman and Tenney (47) criticised the use of only the carbonutou nitrogen ratio when predicting the mineralization rate of applied organic amendments. They carried out a study comparing lignin, cellua lose, and sugars as organic amendments. The lignin had only a slight effect on mineralization rates whereas both sugar and cellulose had an inhibitory influence. ‘Waksman and Tenney (48) also studied mineraliza- tion rates as affected by the physiological age of plants making up organic residues. Older plants retarded the mineralization rate more than younger plants. Two reasons fer this were given: the older plants had a higher lignin content and, also, a wider carbon-to- nitrogen ratio. They concluded that if a plant residue had a nitrogen content of 1.7%, it was just sufficient to cover the nitrogen re» quirements of microorganisms during the period of decomposition. If the nitrogen content were below 1.7%, nitrogen would have to be taken 13 from other sources and would cause a serious shortage of available nitrogen in the immediate area. Harmsen and Van Schreven (23) have considered the nitrogen content required in plant residues added to the soil if tiemup of soil nitrogen is to be averted. They have suggested that the organic material added Should contain from 1.5 to 2.5% nitrogen. They further stated that if organic matter added to the soil has a carbonmtomnitrogen ratio wider than 20 to 1. then the release of mineral nitrogen will not occur until the carbon-tounitrogen ratio reaches approximately 20 to 1. This would imply that the nitrogen content of added residue should be about 2.5% if’mineralized nitrogen is to be released during the early stages of decomposition. Seasonal variations in the mineral nitrogen content of a heavy loam soil in Holland were reported by Harmsen and Van Schrevan. In a fallow soil, the mineral nitrogen content rose in the spring, stayed at a high level during the summer, and decreased in the fall. The decrease in the fall was explained by slower production of mineral nitrogen during the cooler season, and by leashing caused by fall pre- cipitation. In the case of a soil under grass vegetation, the mineral nitrogen content remained at a very low level throughout the entire year. This was caused by the continued crop use of the nitrogen as it was made available. The correlation of nitrogen uptake by plants with various soil tests for available nitrogen was studied by Peterson §§,§l, (34). They feund that, under greenhouse conditions, initial nitrate nitrogen content accounted for 94% of the variation in nitrogen uptake in the first cropping period. This reliability did not hold fer a second crop. 14 An inverse relationship between the ratio of hydrolyzable carbon to hydrolyzable nitrogen and the ratio of total nitrogen to hydrolyzable nitrogen was reported by Kamerman and Klintworth (2?). Lyon §t_gl. (28) reviewed reports of various workers and observed that as much as #5 pounds of nitrogen per acre per year could not be accounted for by crOp removal. leaching, or erosion. They suggested that at least a part of this was lost through the formation of volatile nitrogen compounds. Nitrogen losses have been studied by Hiltbold and Adams (24). They worked on soil acidity changes caused by applied nitrogen ferti- lizer. Among other things, they fOund that when urea was applied to soils having a pH above 7.0. a large loss of nitrogen occurred during a three-month incubation period. In this investigation, glucose was added as an energy source for soil microorganisms. Clark 2; £45.06) also found that soils with a pH range from 7.0 to 7.5 were subject to losses of nitrogen added as urea. They reported losses as high as 40% of the applied nitrogen. Harmson and Van Schreven (23) considered that the loss of nitrogen through the volatilization of ammonia might be quite high in some in- stances. They pointed out that the loss was greatest in waterlogged soils with a high pH. It might be noted here that these conditions often occur in greenhouse culture, and that losses probably occur. They also reported that the risk of nitrogen loss increased with tem- perature, pH, calcium carbonate content, and a decrease in soil mois- ture. Mineral or organic colloids, if present in substantial amounts, could prevent or limit these losses due to their ability to adsorb ammonia. Mann and Barnes (1951). as reviewed by Harmson and Van Schreven. 15 feund that no more than 40 to 51% of added nitrogen could be accounted for at the end of an experiment in which two successive crops of barley 'were grown on well aerated soils under greenhouse conditions. Their experiments covered a total time of 18 months. Allison (1) has given a comprehensive review of a large number of investigations in which attempts have been made to account for all nitrogen in a system during crop production. In some of these experi- ments, only about half of the nitrogen could be accounted fer. Allison considered that at least a part of this nitrogen was lost by volatiliza- tion.f Several mechanisms for volatile nitrogen loss were discussed. Ammonia losses from soils with a pH of 6 to 7 were barely detectable, however. losses at a higher pH were marked. wet soils lost little ammonia but losses were very high as the soil dried if the pH was above 7. Temperature increases were found to increase volatilization rates. Soils with a low exchange capacity lost more ammonia than a soil with a high exchange capacity under similar physical conditions. Ammonia losses tended to be high, even in acid soils, where nitrogenous organic matter was undergoing rapid decomposition. Chemical reduction of nitrogenous compounds and the loss of nitro~ gen as nitric oxide or nitrogen gas was also considered by Allison. He concluded that this type of loss was of minor importance. The recent review by'Mortland and'Wblcott (30) suggests that some elemental nitro- gen may be lost from the soil. Further work is necessary to evaluate this pathway of nitrogen loss from the soil. Bacterial denitrification and nitrogen loss was mentioned by Allison. The presence of free oxygen greatly reduces bacterial deni- trification and it has been thought that this type of nitrogen loss was completely inhibited in moderately well aerated soils. Recent work has 16 indicated that bacterial denitrification may continue in the presence of molecular oxygen and that this type of nitrogen loss may be of some importance. Proper evaluation of this type of nitrogen loss will re- quire teChniques with smaller sampling errors than those currently in use. Schwartzbeck gt.a;, (37) measured the loss of gaseous nitrogen from soil in the forms of‘N2 and N29 using infrared and mass spectro- scopy. They feund that there were substantial losses from saturated soils, and smaller losses from soils at field capacity. Nitrogen fertilizer was added in both nitrate and ammonia ferms, and it was feund that losses were markedly influenced by the type of fertilizer added. N29 losses were highest when a 50~5O mixture of nitrate and ammonia nitrogen were added. Different soil types lost different amounts of nitrogen under similar conditions. EXPERIMENTAL METHODS Design of Greenhouse Experiment The objective of the experiment was to relate short term and long term response of cats to various chemical fractions of nitrogen and carbon in the soil. Johnston (26) had shown that large variations in soluble, acidphydrolyzable and nonehydrolyzable forms of nitrogen in soil resulted from the addition of sawdust and lignin isolated from sawdust, together with varying proportions of urea nitrogen. Large differences associated with treatment were still apparent after 40 week' incubation in the greenhouse. Additional variations resulted when wheat was grown during the incubation period. Similar amendments and a cropping variable were used in the present study to impose a suitable range of variation in measured carbon and nitrogen fractions fbr multiple correlation with various response measurements on short term and long term crops of oats. It was hoped that, by measuring fractional forms of carbon as well as nitrogen, it would be possible to identify forms of nitrogen which contribute uniquely to the avail- able supply and ferms of carbon which contribute uniquely to immobiliza- tion and reduced availability of nitrogen. Mixed.hardwood sawdust was obtained from the Michigan State University experimental sawmill. It was dried at 50° C. and ground in a Wiley mill to pass a 20-mesh screen. Lignin was prepared from an aliquot of this sawdust by acid ex- traction of cellulose and other carbohydrates, as described by Brauns (6). Two-hundred-fifty grams of ground sawdust were added to 1 liter of 72 percent sulfuric acid at 15° C. The mixture was stirred to avoid the fbrmation of lumps. The mass first turned green, then 17 18 black. After a reaction time of 15 minutes, the mass became fluid. The digestion mixture was then diluted with 2 liters of distilled water and filtered onto hardened filter paper in a Buchner funnel. The residue was washed with 500 ml. of 3 percent sulfuric acid and then taken up in 1 liter of 3 percent sulfuric acid and refluxed on a hot plate for four hours. During this period, the color changed from nearly black to a brown color. The residue in the reflux chamber was then filtered, waShed free of acid, and resuspended in 1 liter of 0.5 percent hydrochloric acid. After heating for 12 hours on a water bath, it was again filtered, washed with distilled water and dried at 1050 C. The resulting cake of lignin was ground in a Wiley mill. using a 20 mesh screen, and stored for later use. A soil low in organic matter was used so that changes in nitrogen and carbon due to treatment would be more readily detected through use of larger soil aliquots than would be possible with a soil already high in carbon and nitrogen. The soil chosen was an OShtemo sand with an exchange capacity of 3.5 me. per 100 g., and containing 0.4 percent carbon and 0.023 percent nitrogen. Soil was obtained in the fall of the year, dried, put through a 1/h-inch screen, thoroughly mixed and stored fbr fUture use. Soil tests showed the following nutrient status: pH, by glass electrode in a 121 water suspension, was 5.8; available P using Bray's weak acid extractant (.025 §.HC1, .03fl_NHuF) was 2.5 ppm.; K at 48 ppm., Ca at 350 ppm., and Mg at 8 ppm. - all exchangeable to fl.NHuQAc at pH 7.0. At potting time, mineral amendments were made to insure adequate supplies of major nutrients other than nitrogen. To 12,000 g. of soil in each 3~gal. pot, the following were added: Ca (0H)2 6.0 g. Ca HPOL, . ZHZO 3.0 g. Mao 1.2 g. KCl 3.0 g. At the same time, organic amendments and urea were added according to the schedule in table 1. Immediately after potting, the soil mois- ture level was brought to about 80 percent of field capacity with dis- tilled water. This moisture level was maintained by bringing the pots to the constant predetermined weight with distilled water every other day, or more frequently as needed. The treatments in table 1 were established in two series of trips licate pots for each treatment. After incubating for 37 days, one series of pots was planted to oats. Sixtyeeight day later (105 days after treatment) oats were planted again, this time in both series of pots. Populations were adjusted shortly after emergence to 15 plants per pot. Both crops were grown for 54 days, at which time the oats were fully'headed. The aboveeground portions of the plants were har- vested. Fresh weights were taken, and dry weights after drying at 60° C. The dry material was ground and its nitrogen content determined by'macro-Kjeldahl procedure, using salicylic acid to include nitrate. Following the removal of the first crop, the soil in the 27 pots of the first series was removed. The roots were separated, dried, crumbled by hand and remixed with the soil. The soil was returned to the pots and allowed to incubate two weeks before the second crop was planted. Just prior to planting each crop, a 500~g. sample of soil was taken from each pot for the short term oat response studies and for 20 Table 1. - Description of residue and nitrogen treatments. Treatment Residue Nitrogen Added Rate of Carbon In As C/N No. Code Material addition added residue urea Total Ratio ppm. ppm. ppm. ppm. PPm- 1 R1N1 Control a w a 0 0 - 2 R1N2 Control a m u 106 106 — 3. R1N3 Control m - m 213 213 - 4 32111 Lignin" 2 500 1 500 6 0 6 242 5 R2N2 Lignin 2500 1500 6 100 106 14 6 RZNB Lignin 2 500 1500 6 207 213 7 7 R3N1 Sawdust 9750 4202 16 0 16 269 8 R3N2 Sawdust 9750 #202 16 90 106 40 9 R3N3 Sawdust 9750 #202 16 197 213 20 *'Lignin added in quantities equivalent to that added in sawdust, which yielded 25.6 percent lignin by sulfuric acid digestion. 21 the fractionation of soil carbon and nitrogen, which are described in succeeding sections. Soil tests for pH and nutrient status were also made on these samples. Soil tests after the first crop averaged 55 ppm. P, 60 ppm. K, 325 ppm. Ca and 150 ppm. Mg. Soil tests at the same time in the uncropped series averaged 60 ppm. P, 110 ppm. K, 440 ppm. Ca and 150 ppm. Mg. Differential declines in P and K with different treatments due to differences in crop removal or other treatment ef_ fects were adjusted fer by appropriate additions of these nutrients to the first series of pots prior to planting the second crop. Initial addition of bases exceeded exchange capacity and gave rise to a soil pH of 7.5 in unamended soil. Release of’NH3 from urea re- sulted in additional pH increases up to 8.0. Soil pH declined rapidly as ammonium was nitrified. However, these declines were less rapid in sawdust-treated soil and in the absence of cropping. After the first crop, pH in control and ligninmtreated soils ranged from 6.6 to 7.0; with sawdust pH ranged from 7.0 to 7.4. The lower values in eaCh case were associated with previously cropped pots and those receiving the higher additions of urea. Pots were randomly located in the greenhouse and rotated periodi- cally to minimize position effects and allow for statistical analyses in accordance with a completely random design. Short Term Oat Response Study The short term responses of oats were observed, using freSh soil samples taken just prior to the planting of each long term crop. The availability of nitrogen in non-water-soluble fractions was of primary concern in these short term studies. For this reason, ZOO-g. aliquots of soil were thoroughly leached (soils leachete = 1:6) 22 before being exposed to the root mat of 15mday~old, nitrogen-deficient oat plants, in accordance with the procedure described by De Ment 22 .él~ (17). The above-ground portions of the plants were harvested after 11 days' contact with the soil. Green weight was determined, and dry weight after drying at 500 C. Kjeldahl nitrogen was determined, using salicylic acid to include nitrate. Fractionation of Soil Carbon and Nitrogen A flow diagram for the soil fractionation procedure will be found on the following page. Total carbon, total nitrogen, nitrite and nitrate were determined in a 126 water leachete of the fresh soil. The leachate was brought to a boil, cooled quickly and stored at 40 C. until the analyses could be completed. Aliquots of the leached soil were used for moisture determination and fer the short term response study. The balance was divided into duplicate samples and quick frozen until the remainder of the frac- tionation could be undertaken according to procedures described by Sinsh (38)- "Extractable” nitrogen and carbon represent compounds exchangeable to y K2804 at pH 1. 5 to 2.0 (g/ 10 H2804). Two-hundred ml. of this ex- tracting solution were added to each of duplicate 110 g. aliquots of the frozen soil and shaken at room temperature for 30 minutes. The extract was recovered by filtration and combined with three distilled water waShes to give a final volume of 500 ml. Total carbon and nitro- gen were determined. The extracted soil was dried at 70° C. for 10 hours and its total carbon and nitrogen determined. Twenty-five g. of the dried soil, in duplicate, was digested in 80 ml. of 80% sulfUric acid at room 23 FLOW DIAGRAM FOR SOIL FRACTIONATION 00 . Moist Soil from Greenhouse Pot """"1 Moisture sample Leeched with 3000 ml. distilled water Leachate f F I I Total C Total N Nitrite Nitrate Leached Soil l 1 Short term Moisture crop response sample Extracted with N K2804 in NZ10 H280“, Extract 1 ' l Total C Total N Extracted Soil r’ l Total C Total N Hydrolyzed with 80% H2504 Hydrolyzate Residue 1 [7 ' F' I Total C Total N Total C Total N (By difference) 24 temperature for 2 hours, with frequent shaking. The volume was then made up to 350 ml. with distilled water and autoclaved at 15 pounds pressure for four hours in a flask fitted with a Bunsen valve. After cooling, the residue was collected on filter paper in a Buchner funnel. "Non-hydrolyzable" nitrogen and carbon were determined in the residue. ”Hydrolyzable" carbon and nitrogen were calculated by difference. Total carbon in the whole soil and in the various fractions was determined by wet combustion and gravimetric measurement of C02 as described by Allison (3). Total nitrogen was determined by'macroe Kjeldahl methods described by Jackson (25), using CuSOn and HgO as catalysts. Thiosulfate was added with the alkali during distillation to reduce mercuric compounds. Methyl purple was used as indicator.. Analyses for nitrate in the water leachete were made, using the phenoldisulfonic acid procedure as described by Jackson (25). Nitrite analyses fellowed the naphthylamineasulfanilic acid method described by Fraps and Sterges (20). Duplicate samples of soil from each pot were carried through the entire fractionation and analyzed separately. The mean of duplicate determinations was used as the unit observation for each pot. Statistical Treatment All experimental values were subjected to analyses of variance in accordance with a completely random design with three replications. The significance of mean differences was tested by multiple range and multiple F tests according to Duncan.1 Multiple correlation and 1 Duncan, D. B. Multiple range and multiple F tests. Biometrics. Vol. II. pp. 1-42. 1955. 25 regression analyses made use of an exponentialupower function described by Halter gt‘gl. (15) and proposed for use in fertilizer input-output studies by Sundquist and Robertson (41). This function is of the following fOrm: Log Y = a + b1logX1 + c1X1 + bglogXZ + c2X2 + .... + bilogxi + cixi The use of this function has some advantages over the use of a polynomial. It is, in reality, a crossmproduct functions b1 X1 b2 X2 bi Xi As a result, a degree of interaction between independent variables is expressed without the loss of degrees of freedom which occurs when separate cross-product terms are introduced in polynomial functions. The function also allows for polymodal changes in direction of the response curve. For these reasons it was anticipated that more useful multiple correlations and more highly significant regression coef- ficients might be obtained among a relatively large number of variables using relatively few observations. EXPERIMENTAL RESULTS Effects of Treatment on Soil Variables The first soil samples were takentfive weeks after treatment, just befbre planting the first crop of oats. Soil fractionation data for this sampling are presented in Table 2. Increasing additions of urea nitrogen were reflected by increases in nitrate, other waterasoluble and saltuextractable forms of nitrogen in control soil and soil amended with lignin. In the presence of sawdust, these forms of nitrogen were sharply reduced. This was likely due to microbial immobilization,.since cor- responding increaSes‘occurred in the hydrolyzable fraction. Changes in the non-hydrolyzable nitrogen fraction were small, but significantly higher levels were obtained with lignin and sawdust than in the control. This suggests that chemical complexing of ammonia by lignaceous cons stituents may have occurred to a limited extent. Carbon added as lignin appeared primarily in the nonmhydrolyzable fraction, whereas the addition of sawdust was reflected in both the hydrolyzable and non-hydrolyzable carbon fractions. Changes in CzN ratio reflected both residue and nitrogen treatments. The nature of water-soluble nitrogen other than nitrate in table 2 was not determined. Only traces of nitrite (less than 2 ppm.) were fbund. No carbon was found in the water leaChate, so it is assumed that this fraction was principally ammonium. It is unlikely that any urea would have remained after five weeks. The saltaextractable nitro- gen was largely exchangeable ammonium, although same'organic compounds were included, since a small amount of carbon appeared in the potassium sulfate extract. Nitrogen in these two fractions had been extensively 26 27 Table 2. - Effects of residue and nitrogen treatments on soil variables N0 «N Water 501. K2804 Ext. Hydrolyzable .__Ireatment* N** . N c N 9 ppm. PPmo ppm. ppm. ppm. PPmo Control N1 19 u 10 9 198 1l+26 abc c de ab b f N2 17 25 14 6 218 1469 abcd bc bcd c b ef N3 26 80 32 7 211 1321 a a a be b fg Lignin N1 12 3 7 10 209 1826 bcd c e a b d N2 25 21 16 7 212 1213 a bc be be b g N3 22 76 30 6 207 1598 ab a a c b e Sawdust N1 1 2 9 8 200 2336 e c de bc b b N2 7 7 11 7 258 2818 de c cde bc a a N 11 46 19 8 257 2137 3 cd b b abc a c * N1 = no nitrogen. N2 and N3 = nitrogen added to give CzN ratios for ** Other than nitrate or nitrite. a, b, c, --- g. Ranges of equivalence. For a given soil variable, different at 5 percent. 28 measured prior to the first cropping period. Non Hydrolyzable Total Non water Soluble Treatment L c N 0 cm ppm. PPmo ppm. ppm. Control N1 45 1248 250 2584 10.6 e f e d e N2 51 1510 282 2986 10.6 d e bcd d e N3 51 1472 294 2801 9 . 5 d e b d e Lignin N 58 2704 274 4541 16.6 1 be cd cd e be N2 59 2917 288 4136 14.4 bc be be c d N 60 2586 296 4190, 14.2 3 abc d b c d Sawdust N 56 3240 266 5585 21.0 1 c a de ab a N2 62 3156 332 5981 18.0 ab ab a a b N 65 3160 341 5305 15.6 3 a ab a b cd sawdust of 40:1 and 2021 respectively. numerical values with a common literal subscript are not significantly 29 converted to nitrate 68 days later when the second sampling was made at the beginning of the second cropping period (tables 3 and 4). Some additional immobilization of nitrogen in the hydrolyzable fraction also occurred with sawdust during this period (table 3). As a result there was little change in total nonowater soluble nitrogen with sawdust, whereas some loss of insoluble forms occurred in the controls and in lignin-amended soil. There was a marked change in the nature of carbonaceous soil mate- rials during this time interval as shown by net decreases in the non- hydrolyzable fraction and net increases in the less resistant hydrow lyzable ferms. Extensive losses of carbon with the sawdust treatment occurred primarily from the nonehydrolyzable fraction (table 3), and these losses were enhanced by increasing levels of applied nitrogen (table 4). In the case of the control and lignin treatments, net increases in total carbon reflected additions in the form of root residues from oats grown, and these additions appeared primarily in the hydrolyzable fraction. Removal of nitrogen by the first crop of oats was reflected in levels of nitrate remaining for the second crop (table 5). A portion of'the nitrogen removed by oats remained in the soil in the form of root residues and appeared primarily in the hydrolyzable fraction. The apparent contribution of both carbon and nitrogen to this fraction from oat root residues was greater in soil amended with lignin or saw- dust and increased with increasing level of applied nitrogen. The greatest residual accumulations of nitrogen in organic forms at this time were found in the sawdust treated soil. The principal mineral form at this time was nitrate which was depressed in the pres- cence of sawdust as compared with the control and lignin treatments. npflz common mcaddoao m magmas mosam> Hmofiuosdd .manmawm> HHom co>wm m pom .pcoonod m we pdmpommfiv hapcmofimacwfim pom ohm pdfipompsm amuopfia mossoo m .oomoam>flsdm mo mowcmm 00 on a“ .oaaueaa no oneness amen tempo . n m m n m a m m m m n o.ea mane arm seem an seem emm m e o mm amsnssm m m n m m n n o m w m e.oa own: mom omom an ema_ mom o o_ o no eaemaq o n o o p p Q n m w m m.NF omom mam some m: can? mm? m P? 0 so aowpdoo possum wmflddono vcooom m m o m m m m m n p n N.ma emom mam nw_m so omen mmm m m_ me o pmeessm Q n n n m D n m m m m P .m F mam: 0mm mm mm on man P mom 0 we mm om madman o o o o n o n m m m m N.oa seam new case as men? men a e_ on em Honeeoo uoanom mqaddouo pmuam .sdd .sad .sdd .sdd .smd .emd .de .sdd .sdm .smd awo o 1m o m. Ir 2 pmospmmne . o. z o z . maseaon noes: eoz asses .mapesaaotsam eoz nammmaammmam .mmmwuaummm .Hom momma. a-noz .moanmahm> Haom vousmmos co mucospmohp cavemen mo mpoommm admz.u .m canes 31 .pcoopod m we pcohmmmfiv havqmoamammfim 902 one pdfiuomnsm amnopda cosEoo m suds moaned meaddomo m menus: mosam> Hmowposdm .oanmfium> Haom co>fim w you .oocoam>asdo mo mqumm .o .n .m .opfihpdc ho opmupwc gasp honeo in .mam>apomdmmp .Faom use vac: no pmsbzmm you capes zuo o>aw op bosom momoupfi: n ma use N2 .cmmoupfic oquaz.* D m m 9 mm m m m m w m m T? mama lam mm? em $9 omm m S a N9 2 D m m m m m m m m w p of. moo: 0mm 38 mm 39 mmm N. 2 o 3 NZ m w n m D m n w m m o P s5 85 Sm omom Nm Q8 9.: a o. o m z powwow mmfldmouo bmooom o m m D m D m n m m m m if. $3 2m worm an 32 nmm N. am so om z D m m m m a m D Q 9 pm a .a no? as new a Rs. an a a 2 2 as m w n n p m n m o o n For RN: 8m 39. mm 89 mom a m m : Fz soaked McAddowo pmnflm .eee .eae .ene .53 and .58 .eee and and and 1% .ollllnmllllm o m1 o m o m1 :z .paoesmone Season teem: eoz H38. oanmuhHOHmmMImoz manmnhHOAUhm .uexm mom; .aom hopmz. zumoz .moanmfium> Hfiom vonsmuos no mpcospmouw demons“: mo mpoommo ads: I .3 canes 32 Table 5. - Effects of previous cropping, residues, and nitrogen treat- . N03eN Hydrolyzable N . Treatment Uncroppgg, Cropped Uncropped Cropped ppm. ppm ppm. PPm- Control N1 23 6 196 177 ef ef cde e N2 86 20 194 181 be ef de de N3 163 66 196 188 ' a cd cde de Lignin N1 16 8 193 180 ef ef de de N2 111 17 213 218 b ef cde cd N3 181 67 208 197 a cd cde cde Sawdust N1 1 2 23 215 f f be cde N2 45 13 259 284 e ef ab a N3 103 31 260 270 be ef ab ab Means for Cropping 81 24 217 212 a b a a 7 N1: no nitrogen. N2 and N3 = nitrogen added to give CsN ratios fer a, b, c, --- e. Ranges of equivalence. Within a pair of columns for a significantly different at the 5 percent level. 33 ments on soil variables measured prior to the second cropping period. Hydrolyzable C Non Water Soluble N Treatments Uncropped Cropped Uncrcpped Cropped ppm. ppm. ppm. ppm. Control N1 1710 1891 254 241 bed bcd de 6 N2 1713 1783 257 2‘10 bcd bcd cde 6 N3 1455 1710 252 248 d bcd de de Lignin N1 1628 1567 260 243 bed cd cde e N2 1342 2247 284 285 d abcd cd cd N3 1646 1917 272 260 bed bcd cde cde Sawdust N1 3095 2216 294 274 a abcd bc cde N2 2481 2117 326 347 abc bcd ab a N3 2473 2586 333 337 abc ab a a Means for Cropping 1949 2004 281 275 a a a a sawdust of 40:1 and 20:1 respectively. given soil variable, numerical values with a common subscript are not 34 Nevertheless, the peak immobilization effect of sawdust had been expressed earlier where nitrogen had been applied. Nitrate was now accumulating where the initial CsN ratio had been 4021 (N2), as well as at CzN = 2031 (N3). Long Term Responses of Cats Two successive crops of oats were planted on one series of pots. A second series was uncropped during the first cropping period. The first crop was planted five weeks after addition of residue and nitroa gen amendments. The second crop was planted 68 days later. Both crops were narvested 54 days after planting When the cats were fully headed out. Determinations made on abovesground portions of the first crop are presented in table 6. In sawdust treated soil, there was very little growth of cats after germination where no nitrogen was applied. Nitrogen added with sawdust to give a CzN ratio cf 403? produced yields equal to those in the control without nitrogen. Yields were practically doubled when additional nitrogen was used to give a CsN ratio of 2031. Nevertheless, yields were significantly less with saw- dust at all levels of nitrogen than in controls or with lignin. Lignin had no effect on yields, but it did reduce percent nitrogen and nitrogen uptake at the higher level of nitrogen addition. In- creasing nitrogen additions increased succulence (decreased percent dry matter) and increased percent nitrogen. Both sawdust and lignin reduced succulence when added without nitrogen. Sawdust markedly reduced percent nitrogen and nitrogen uptake at all levels of applied nitrogen. The depressing effect of sawdust on yields, nitrogen content and 35 Table 6. - Effects of residues and nitrogen treatments on long term crop variables for first cropping period. * Green Dry Dry Nitrogen Treatment ‘Weight weight Matter Nitrogen Uptake g. mg’ % % mgo Control N1 95 19.1 20.1 1.09 .209 d c be d f N2 212 39.4 18.5 2.09 .823 abc a be b c N3 226 39.8 17.7 2.60 1.035 a a c a a Lignin N1 87 19.1 22.1 1.08 .207 d c ab d f N2 209 39.6 18.9 2.13 .842 be a ho b c 2N3 223 41.8 i8.8 2.29 .957 ' ab a c b b Sawdust N1 3 0.8 24.5 0.75 .006 e d a e g N2 96 19.9 20.6 1.35 .270 d 0 abc c e N3 200 35.2 17.6 2.19 .768 c b c b d t N = no nitrogen. N2 and N = nitrogen added to give a CzN ratio f0r sawdust of 40:1 and 2021 respectively. a, b, c, --- e. Ranges of equivalence. For a given crop variable, numerical values with a common literal subscript are not significantly different at 5 percent. 36 nitrogen uptake were still apparent in the second crcp (table 7). Residual responses to the initially applied nitrogen were also ex- pressed on the second crop (table 8). The general level of yields in the second crop was about oneehalf that in the first crop. This appears to have been due to the fact that the second crop was grown during the long, hot days of summer. These are conditions unfavorable for a cool season crop such as oats. Nitrogen does not appear to have been limiting at the higher levels of nitrogen addition. The nitrogen responses shown in table 8 represent overall averages. The response of the second crop to residual nitrogen from the initial applications was strongly influenced both by residue treatment and by previous cropping. Dry matter yields and nitrogen uptake for previousm ly cropped and uncropped pots are presented in table 9. Soil tests showed that phosphorus and potassium had been differs entially removed by differential growth of oats in the first crop. These variations were evened out by application of additional phosw phorus and potassium befbre planting the second crop. As will be seen in a later section, variations in yield and nitrogen uptake in table 9 were essentially a function of nitrogen availability. Removal of nitrogen in the tops and immobilization in the presence of roots from the first crop sharply reduced yields and nitrogen uptake with most treatments. Of primary interest in table 9 is the fact that the immobilizing effect of sawdust had completely disappeared at both levels of nitrogen addition (N2 and N3) , although growth was still completely inhibited where no nitrogen had been applied (N1). In both cropped and uncropped soils, yields of dry matter leveled off at the N2 level of addition with all residue treatments. However, Table 7. - Effects of residue treatments on measured long term crop variables. Green Dry Dry Nitrogen _Treatment weight -'Weight Matter Nitrogen Uptake g. mg. i % mg, First Cropping Period Control 178 32.8 18.8 1.93 .689 a a a a a Lignin 173 33. 5 19.9 1. 83 . 669 a a a a a Sawdust 100 18.6 20.9 1.20 .348 b b ’a b b Second Cropping Period Control 75 14.4 19.9 2.26 .346 a a a a a Lignin 80 14.2 18.9 2.27 .341 a a ab a a Sawdust 57 10.8 18.2 2.18 .258 b b b a b Table 8. - Main effects of nitrogen treatments on measured long term crop variables. * Green Dry Dry Nitrogen Treatment weight ‘Weight Matter Nitrogen Uptake g. mg. % % mg. First Cropping Period N1 62 13.0 22.2 0.97 .141 c c a c c N2 172 33.0 19.3 1.86 .645 b b b b b N3 216 3809 11800 2003 0920 a a b a a Second Cropping Period N1 40 8.2 19.9 1.52 .116 c b a c c N2 81 1506 1908 2920 0358 b a a b b N3 92 15.7 17.3 2.99 .472 a a b a a * N = no nitrogen. N2 and N = nitrogen added to give a CzN ratio fer sawdust of 40:1 and 2031 respectively. 38 Table 9. - Effects of previous cropping, residues, and nitrogen treatments on long term crop variables measured in the second cropping period. Nitrogen Uptake Dry‘WEight _Ireatment* Uncropped Cropped Uncropped Crppped mg. ‘ mg. mg. mg. Control N1 16.4 8.4 237 99 abcd h de f N2 17.8 12.8 576 209 ab g a e N3 16.9 14.3 536 418 abc defg ab c Lignin N 13.9 8.6 219 104 1 efg h e f N2 17.1 14.4 526 199 ab cdefg ab e N3 15.5 15.8 535 462 bcdef abode ab bc Sawdust N, 0.8 1.0 15 18 i i g g N2 18.0 13.3 423 213 . ab fg c e N3 18.1 13.6 575 305 a efg a d Means for Cropping 14.9 11.4 405 225 a b a b * N = no nitrogen. N2 and N nitrogen added to give CzN ratios for sawdust of 40:1 and 20:1 respectively. a, b, C, ‘3'"- go Ranges of equivalence. Within a pair of columns for a given crop variable, numerical values with a common subscript are not significantly different at 5 percent. 39 in previously uncropped soil, nitrogen uptake was additionally in» creased following the N3 level of nitrogen addition. With sawdust, there was a similar increase in nitrogen uptake at the N3 over the N2 level in previously uncropped soil. Thus, residual effects of ini- tially applied nitrogen were reflected in nitrogen uptake beyond the point where yields were influenced. Nitrogen was apparently taken up in excess of plant needs at high levels of available nitrogen in the soil. Short Term Response of Cats It was recognized that long term growth of oats would be influu enced primarily by soluble forms of nitrogen. To differentiate between responses to soluble and insoluble forms of nitrogen, 200 gram aliquots of soil which had been leached with 1200 ml. of distilled water were placed in containers. Oats previously germinated in nitrogenofree sand cultures were grown for 1% days, according to the procedure described by DeMent‘gt‘gl. (17). This was done at the beginning of each cropping period and is referred to here as the "short term crop”. Data for the first short term crop are presented in table 10. In contrast to the long term response where yields were sharply depressed by sawdust (tables 6 and 7), both green and dry weight yields were significantly increased by both sawdust and lignin. This was true. even where no nitrogen was added. The top growth with lignin and sawdust was less succulent (higher in dry matter content). Al- though nitrogen content was lower than in the controls, nitrogen uptake was greater. It would appear that both lignin and sawdust provided some factor or factors which stimulated the synthesis of dry matter by oats during this early stage of growth. The principal stimulating 1+0 Table 10. - Effects of residue and nitrogen treatments on short term crop variables for first cropping period. * Green Dry Dry Nitrogen Treatment weight weight Matter Nitrogen Uptake g. mgo 7% % mg. Control N1 4.050 444 11.0 1.64 7.3 c d cd 0 e N2 4.076 375 9.2 1.87 7.0 c e e b e N3 4.969 518 €0.4 2.07 10.7 ab 0 d a b Lignin N1 5.030 654 13.0 1.40 9.1 ab ab a d cd N2 5.037 640 12.7 1.52 9.7 ab a ab cd bc N3 5.320 631 1i.9 1.90 1..9 a ab bc ab Sawdust N 5.110 638 12.5 1.38 8.8 1 ab ab ab d cd N2 4.667 600 12.9 1.39 8.3 b b ab d d N3 5.387 689 12.8 1.41 9.7 a a ab d be * N = no nitrogen. N and N = nitrogen added to give CzN ratio for s wdust of 40:1 and :1 respectively. a, b,.c, --.e. Ranges of equivalence. For a given crop variable, numerical values with a common literal subscript are not significantly different at 5 percent. 41 factor does not appear to have been nitrogen in any form, since there were no consistent relationships between dry matter yields and the amount of nitrogen which had been applied two weeks before leaching. The level of nitrogen applied was reflected rather consistently in nitrogen content and nitrogen uptake, so that nitrogen supply, as a secondary factor, did influence response to some extent. The stimulating effect of sawdust and lignin on yields and nitro- gen uptake were still apparent in the second short term planting made 68 days later (table 11). The nitrogen effect, however, had completely disappeared (table 12) . Previous cropping of the soil reduced growth and nitrogen uptake of the second short term crop, except where sawdust had been incorpow rated without nitrogen (table 13). The nature of this cropping effect is not clear. It may reflect a toxic effect of root residues from the preceding crop of cats. There would have been a negligible amount of such residues in the soil where sawdust was applied without nitrogen, since there had been essentially no growth during the first cropping period (table 6). Functional Relationships Among Crop Variables §hort term relationships It appeared from the short term response data that dry matter production may have been affected by factors other than nitrogen uptake. Dry weight and nitrogen uptake were positively correlated (r = .593 in the first crop, r = .840 in the second). Dry weight was negatively correlated with percent nitrogen and positively correlated with percent dry matter in the first crop, however, in the second crop the reverse 42 Table 11. - Effects of residue treatments on measured short term crop variables. Green Dry Dry Nitrogen Treatment 'Weight 'Weight Matter 4Nitrogen Uptake s- mg. % mg. First Cropping Period Control 4.37 446 10.2 1.86 8.3 b b b a c Lignin 5.13 642 12.5 1.61 10.2 a a a b a Sawdust 5.06 629 12.7 1.39 8.9 a a a b Second Cropping Period Control 4.60 630 13.9 1.43 9.0 c a ab a b Lignin 4.91 688 14.2 1.36 9.4 b a a a ab Sawdust 5.23 704 13.5 1.43 10.1 a a b a a Table 12. - Effects of nitrogen treatments on measured short term crop variables. * Green Dry Dry Nitrogen Treatment Weight, weight Matter Nitrogen thake g. mg. % mg. First Cropping Period N1 4.73 579 12.2 1.47 8.4 b a a c b N2 4.60 525 11.6 1.59 8.3 b b b b b a a b a Second Cropping Period a a a a a a a a a a a a a a a * N = no nitrogen. N2 and N = nitrogen added to give CzN ratios for sawdust of 40:1 and 20:1 respectively. 43 Table 13. - Effects of previous cropping, residues. and nitrogen treat— ments on short term crop variables measured in the second cropping period. Wei ht Nitrogen Uptake Treatment* Uncropped Cropped Uncropped CrOpped mg. mg. mg. mg, Control N1 702 588 10.2 8.5 a a abc de N2 696 563 10.1 7.6 a a abc 3 N3 681 551 908 706 a a abc e Lignin N1 768 573 11.2 7.6 a a ab e N2 747 640 11.1 7.8 a a ab e N3 775 623 11.4 7.3 a a a e Sawdust N1 695 750 10.5 11.0 a a abc ab N2 709 657 10.7 8.8 a a abc cde a a abc bode Means for Cropping 723 625 10.6 8.4 a b a b * N = no nitrogen. sawdust of 40:1 and a. b, C, '"l'- f. N50 Ranges of equivalence. = nitrogen added to give CxN ratios fbr ectively. Within a pair of columns for a given crop variable, numerical values with a common literal subscript are not significantly different at 5 percent. 44 was true. There was no significant correlation between nitrogen uptake and either nitrogen or dry matter content in the first crop. whereas in.the second crop nitrogen uptake was positively correlated with percent nitrogen and negatively correlated with percent dry matter. For these reasons, it seemed desirable to investigate functional relationships among the measured crop variables, using the exponential Carter-Halter function. These are shown for the two short term crops in figure 1. The plotted functions were fitted to unit observations. Treatment means fer three_replicate observations are also shown to give visual evidence of the extent-to which the functions accounted for variability in the dependent variable plotted on the vertical axis. In figure 1-A it is apparent that increasing dry weight was es- sentially a function of increasing percent dry matter, although a significant positive relationship with percent nitrogen was also ex- pressed. Both responses were curvilinear. Dry weight increased more rapidly with increasing nitrogen and dry matter contents above their mean values. The R2 = .865 indicates that 86.5 percent of the total variation in dry weight was accounted f0r by these functional relation- ships with nitrogen and dry matter contents. It is significant that the simple negative correlation observed between dry weights and nitro- gen percentages by themselves was reversed when percent dry matter was also considered. The equation for figure 1-B accounted fer only 35 percent of the variation in dry weight of the second crop. Nevertheless. regression coefficients for the log and semi-log terms fer dry matter content were statistically significant. The plotted function shows an increasingly rapid decline in dry weight with increasing percent dry matter. If figures 1-A and 1-B are considered together, it would appear 45 Figure 1. - Functional relationships among short term crop variables. Figure 1—A. - Dry weight (DWt.) of first crop of oats vs. dry mat- ter (DM) and nitrogen (N) in tissue: Log DWt. = - 7.0013 - 4.16’ log N + 117? N + .339 16g DM + 5.94 DM ' NZ = .865 Figure 1-B. - Dry weight (DWt.) of second crop of cats vs. dry mat- ter (DM) and nitrogen (N) in tissue: Log DWt. = + 14.629* - .426 log N + 16.8 N + 9.69* log DM - 32.5* DM NZ = .346 Figure 1—C. - Nitrogen uptake (NUpt.) of first crop of oats vs. nitrogen (N) and dry matter (UN) in tissue: Log NUpt. = - 7.00 — 3.16 log N + 117.‘ N + .339 10g DM + 5.94 UM R2 = .809 Figure 1-D. - Nitrogen uptake (NUpt.) of second crop of oats vs. nitrogen (N) and dry matter (DM) in tissue: Log NUpt. = + 14.625* + .574 log N + 16.8 N + 9.69‘ log DM - 32.5* DM R2 = .720 * Regression coefficient significant at 5 percent. ** Regression coefficient significant at 1 percent. ‘. . 46 BOO - I- A - l- 8 FIRST CROP SECOND CROP 700’— N'Z'O ® 900‘ 1. /, 5 (E35411 '600 ‘; ,._ [13 mass >- .- 3500 % 0 © Ava: II. NILI'Ieffi $400 0) \r < 1: 6 Q ‘2 4300 d 2 1 1 1 400__1—1__1_1 2009 .40 u .2 .3 I2 13 14 15 I6 PERCENT DRY MATTER PERCENT 0'" "”75“ 12 |-c [E/AVEDM '3‘!sz I-D 5‘" [FIRST CROP "2% g SECOND AVE-0N ,5 . CROP - '4 1‘ 3" no / a H b KEY 011- 7' ‘3) 5 014-1311. 8 — f, o 8 0 No Residue gm / 310— 6 1:1 Liqnln E [21 1: / O Seeduet _ 2 z 9 I Bold faced symbols ”,9” / '3’ _ -Previouely uncropped 3 /o~1.13-/. : / (9 1.2.3 -Appnod '28— / 9. 8— N level 9 ,, j [E] j - RED ON-Ie v. i o/ 3 El / 7 1 1 © J 7 1 1 1 1.0 L3 L6 1.9 2.2 1.0 1.3 I-6 I-9 22 PERCENT NITROGEN PERCENT NITROGEN Piqure l 47 that an optimum for dry matter synthesis by oats at this stage of growth might occur at 13 to 14 percent dry matter. In the first crop, some factor or factors in lignin or sawdust promoted dry matter syn- thesis and decreasing succulence over the positive leg of the response curve. In the second crop, decreasing synthesis of dry matter was associated with some factor in previously cropped soil which reduced succulence excessively. This depression in previously cropped soil was essentially un- related to nitrogen content of tissue or to nitrogen treatment. In previously uncropped soil, dry weight increased with both nitrogen and dry matter contents, as in the first crop. The Carterohalter function did not give expression to this obviously different relationship in previously uncropped soil as compared with that in previously cropped soil. In figure 1-C, the scatter of mean values for nitrogen uptake was rather well covered by functional values calculated over the observed range of mean values fer dry matter and nitrogen content. Nitrogen uptake increased with both nitrogen and dry matter contents, as had dry matter in figure 1-A. In the second crop (figure 1-D), the varia- tion in nitrogen uptake in previously cropped soil was fairly well accounted for by a positive relationship with nitrogen content and a negative relationship with dry matter. In previously uncropped soil. however, the response to unknown factors in sawdust and lignin exceeded the observed limits of'the function. Functional relationships in figure 1 support the earlier inference that dry matter synthesis was specifically stimulated by non-nitrogenous constituents in sawdust and lignin. Thus, oats in residue amended pots in the first crop were segregated by the function into a separate 48 population from oats in control pots on the basis of differences in dry matter percentage (or succulence). Nevertheless, in both popula- tions there was a similar increase in dry weight and nitrogen uptake with increasing nitrogen content (figures 1mA and 1_C). In the second crop, there appeared to be an inhibitory effect of root residues from the previous crop of cats. This inhibitory effect on dry matter production exceeded the functional limits imposed by the range of mean nitrogen and dry matter contents actually observed (fig- ure 1-B). Thus, cats on previously cropped and previously uncropped soils appear to belong to two populations differentiated by some factor or factors other than nitrogen content or dry matter content. Dry matter contents appeared to be excessively high in the previously cropped population which suggests that water uptake or normal tissue hydration was interfered with. Nitrogen uptake associated with these excessively high dry matter contents was reduced, and the reduction was fully accounted fOr within the functional limits actually observed (figure 1-D) . In figure 1-D nitrogen uptake in excess of the functional limits was associated with sawdust and lignin treatments which had stimulated dry matter production in a direction opposite to the general trend ex- pressed by the function in figure 1-B. It is apparent that the inc clusion of data for both cropped and uncropped pots in one function was not fully justified. Also, the functional ferm used allows for an inflexible, additive type of interaction in which inflexions or peaks associated with one independent variable are not free to vary with variations in another. Thus, a maximum dry weight in figure 1~B is expressed at 13 percent dry matter, regardless of Uhe level of tissue nitrogen. A quadratic function with interaction terms would allow for 49 flexibility in expression of this maximum at different levels of ni- trogen. In spite of these shortcomings, the CarteroHalter function led to useful inferences and is employed in succeeding sections. Lopg term relationships Functional relationships among long term crop variables in figure 2 were fitted to unit observations. Experimental means of triplicate observations are also shown. The functional relationShips in the first crop between dry weight (figure 24A) and nitrogen uptake (figure 240), on the one hand, and tissue nitrogen and dry matter content, on the other, were very similar. The curves suggest that optimum production of dry matter occurred at a nitrogen content of 2.0 percent. Optimum nitrogen uptake occurred at a somewhat higher nitrogen content (about 2.2 percent). Expected dry weight and nitrogen uptake, as defined by the function, declined rapidly at nitrogen contents above this optimum. These declines, how- ever, were opposed by increasing dry matter content. The regression coefficients for tissue nitrogen content in both equations were highly significant and 83 and 90 percent, respectively, of the total variation in dry weight and nitrogen uptake was accounted fbr. In the second crop (figures 2-B and 2-D), no optimum level of tissue nitrogen was expressed fer either dependent variable. Only 36 to 54 percent of the total variation was accounted for and only the semi-log term for tissue nitrogen acquired a statistically signifi- cant regression coefficient in either equation. Nevertheless, the functional relationships are revealing. In figure 2-B, increasing percent nitrogen was accompanied by 50 Figure 2. - Functional relationships among long term crop variables. Figure 2—A. - Dry weight (DWt.) of first crop of oats vs. nitrogen (N) and dry matter (BM) in tissue: Log DWt. + 24.441-+10.7‘*1og N - 2.32**N + .508 10g DM + 2.23 DM 82 = .827 Figure 2~B. - Dry weight (DWt.) of second crop of oats vs. nitro~ gen (N) and dry matter (DM) in tissue: +2.439 - 3.87 log N + 112? N + 10.2 16g DM - 16.0 DM 82 = .358 Log DWt. Figure 2-C. - Nitrogen uptake (NUpt.) of first crop of cats vs. nitrogen (N) and dry matter (DM) in tissue: Log NUpt. = +24.439 + 11.7**1og N - 232T*'N + .507 10g DM + 2.25 DM 2? = .898 Figure ZnD. - Nitrogen uptake (NUpt.) of second crop of cats vs. nitrogen (N) and dry matter (UN) in tissues Log NUpt. = + 2.439 - 2.87 log N + 112?‘N + 10.2 16g nu - 16.0 DM Re = .542 * Regression coefficient significant at 5 percent. we Regression coefficient significant at 1 percent. 60. N 0‘ Q U 0 O O O l T I T GRAMS DRY WEIGHT 5 I IZOO [ IOOO I O O O T O O O l 400 ~ 200 — MILLIGRA M5 NITROGEN UPTAKE 1 I 1.0 2.0 3.0 PERCENT NITROGEN 2-C FIRST l I I-0 2-0 3-0 PERCENT NITROGEN 607 2 - B KEY 25% PND O No Residue D Lignin 50s 0 Sawdust Bold teced symbols . Previously uncropped ,_ 40 - l,2,3 - Applled ‘3’: N level - AVE.DM Isl 3 30+- ' l9 % —\ ). a: o 20 (I) I < 5 IO 0 0."). L l-O 2-0 3-0 PERCENT NITROGEN 6001' 2-0 SECOND @ CROP 00/ Isl 500 ~ I x < f; / 3 400— D" ' 23 70 z u 8 a: 300» t z AVE.DM m 200— . I9 96 I < e: 3100—6 MDMPIGTB :..-l 3 og—“4lggl H__.__.. .1 L0 2.0 3 0 Figure 2 PERCENT NITROGEN 52 decreasing percent dry matter. The expected increases in yield of dry matter failed to materialize. It appears likely that this was due to respiratory losses of photosynthate promoted by unfavorably high temper» atures and long days during the summer months when this crop was grown. In the case of nitrogen uptake (figure 24D), the essential func- tional relationships were similar to those fer dry weight, but large increases in nitrogen uptake were actually obtained over the whole range of observed values. Maximum uptake of nitrogen was associated with maximum succulence (minimum dry matter percentage). Comparing figures 2_B and 2-D, it isapparent that nitrogen taken up in excess of 300 mg. was in excess of the cropt1potential for inn creased growth. Such excessive accumulations of nitrogen occurred at nitrogen percentages in excess of about 2.0. This approximates the Optimum nitrogen content expressed for the first crop by the functions in figures 2.A and 2_C. It was inferred from patterns of response fer the short term crops in figure 1 that dry matter synthesis and nitrogen uptake can be influenced independently by soil or environmental factors. This infers once is supported by response patterns fer the long term crops in fig- ure 2. It is apparent that neither dry matter yields nor nitrogen uptake, individually, are adequate to characterize crop responses to ‘soil treatments or environmental variables. In spite of the limitations of the mathematical model used, it was useful in bringing out meaningful relationships among the various crop measurements. Useful information was obtained even where regres- sion coefficients fell short of statistical significance or where 8'2 values indicated that only a portion of the variation in the dependent variable was described by the function. 53 Functional Relationships among Crop and Soil Variables Intercorrelations among soil variables The interpretation of'multiple regression analyses becomes dif- ficult where independent variables are highly intercorrelated. Simple correlation coefficients among soil variables measured prior to the first cropping period are given in table 14. Four correlation co- efficients are given for each pair of factors, since each factor appears in linear and log form in the functions used. Coefficients not significant at the 5 percent level of probability are not listed. Nitrate and other water soluble forms of nitrogen were positive- ly correlated only with extractable nitrogen. Nitrate was negatively correlated with hydrolyzable and nonmhydrolyzable carbon. Other water- soluble and extractable nitrogen were negatively correlated with extractable and.hydrolyzable carbon. Hydrolyzable and nonAhydrolyzable nitrogen and carbon were all positively correlated among themselves. Simple correlation coefficients for soil variable measured prior to the second cropping period are given in table 15. Much of the intercorrelation observed in the earlier sampling had disappeared. Significant correlations involving watercsoluble nitrogen and extrac- table carbon are of little consequence because these were present in very small quantities (tables 3 and 4). Positive correlations between hydrolyzable nitrogen and carbon, and between nonehydrolyzable nitrogen and carbon in both samplings tend to be prejudicial to some of the inferences to be made in the fellowing sections. V However, all pairs fall Short of perfect correlation, and most have a large degree of independence. Thus, correlations of the order 54 a mz m2 m2 mz 1mmmumcg mz mz mz mz Hmocaanmog mz mz mz mz mcHIHeecaa mz mz m2 m2 Amanda .0 .apxm mz 11mz .Nos.1 82 ..nsw.1 ecaammm mz mz .emm.1 mz ..Nmm.1 cmecaaamcg mz mz .sma.n m2 .eams.1 mcHIHeecag m2 mz .emm.1 mz .em3.1 access 2 .ccxm mz mzl, .aam.1 mz ..esn.1 .11. .ewce.+ mmmummm» mz mz .me.- mz eesom.1 armem.+ seecaalmca mz mz mz mz .mms.1 eemoe.+ mcaaemecaa m2 mz mz mz .mm:.1 cease.+ seemed ,2.Hoeeomm :26... as :68 .. a 1% .931. Race... 82-48! .eenn.1 mz .emme.1 mz mz rams.+ pmmm.+ ceecaaumcq eecam.a mz eemos.1 mz mz caec.+ steam.+ moaueeecaq sewsm.1 mz eesmc.1 mz mz rmns.+ mz emcee; zumoz o z o .m11 o 2 mm1 coacecsa eageace> N U sumo oHnmNmHouohm odpmpomuexm .aownommvl, . aflom pond“ one op Aofind scammmos moanmaam> 1‘. .ooanom mcaddono Hfiom moose mcofipmaoncoo moa one .moHIHEom .nmoqaq I .:F canes 55 .cceccec F as ccecacacmam es .pcoccoc 0 pm pcwoamwcmwm * .opdhpwc no women“: swap compo F essew.+ moaumoq as:©w.+ amocHHIwoq semow.+ moauhmocaq 1.53 .+ .86qu z .Ennncoz ..mew.+ .emow.+ 11ecwummm. eemmm.+ eeeom.+ seecHHIMoa ssrmm.+ *st:.+ moasamocfiq .enmm.+ .eems.+ scenes 0 .esm fags... {*mmd...‘ {imom 6+ 3 saw:.+ sswrn.+ ssomm.+ nmmcaaumoq .mo:.+ . ..sms.+ ..smn.+ ecH1ceecaq ewes.+ .ewom.+ essam.+ access .2 .esm o 1m o z o 2 F2 cowpocsm mammanm> mammmwammmmmwmmm. AddmemwammmHmIl .1 caneceeccme. .HoesommI Haom .Accccacccov 1 .2? canes 56 2.8m... an an .83. 848% .eoms.1 m2 mz cesm.+ seesaalmcn 1m R .1 mz mz comm .+ mcaaeeefiq ..scs.1 mz mz .ecem.+ seecaq c .ccxa wz ma mz 1mz Ilmz ecHIecm. m2 mz mz mz mz eeecaaamoq mz m2 m2 mz mz mcaaseccaa mz m2 mz mz mz scenes 2 .ccxm mz 111m2 m2 m2 m2 m2 ,unmmnswm mz mz mz mz m2 m2 seesaalmoq mz mz mz mz mz mz moanseecaq . mz mz m2 mz mz m2 access ,2.H6810Nm .Ilm1z 2 feed... |mIz Lmzl mz isms..+ 8% mz mz .ewmm.1 mz mz mz cram.+ seecaalwcq mz mz .mem.a mz mz mz .eoas.+ mcH18eecaq m2 m2, .mem.1 m2 m2 mz sewem.+ ceecan zlmoz o w as o 1m hr, 11m, . .2 N assesses easease> Imammmwammmsmhmmm .mammmuamwmwml. capecceccxm Hoeso m Hwom .ooapmu moan . Idopo scooom op pofipd ansmmos moanmfinm> Haom mcosm moofipwaoapoo moH one .moHstmm .hmocau I .m— canoe 57 .pcoouod ? pm undefimacmaw cs .pcoouom n as pqmoamacmam s .ecaccac cc escapes secs sense a ssmnm.+ moaumoq :98 .+ 82:1ch scumm.+ moaunmocda ssn:m.+ seamen z .ohnlcoz IIIuZI mm. mmwummJWI mz mz umocAHImoq mz mz moasuwocaq m2 mz .883 .o .esm aluz emmm .+ tin 0.. icomummq) mz smmm.+ ssmom.+ Hmocaaumoq mz mz *smrm.+ moanuwocfiq mz mz sswom.+ nwocaq z .omm o z o z o 11a cm 8.983 6% «3 cans snags? dang 4810mm. . . flow Aceaacscov .. .3 cases 58 of r =-.7 between nitrateenitrogen and hydrolyzable carbon (table 14) correspond to coefficients of determination of the order r2 = .49. In other words, only about 50 percent of the variation in nitratemnitrogen was associated with inverse variation in hydrolyzable carbon. The higher correlations between water-soluble nitrogen and extractable nitrogen and between nonmhydrolyzable nitrogen and carbon in the first sampling (table 14) are more damaging. With such a.high degree of interdependence, a functional relationship shown for one independent variable may actually reflect a more fUndamental relationship involv- ing the other. §hort term crop and soil variables For each short term crop variable, five separate functions employed. In the first three listed across the top of table 16, nitro- gen and carbon were used as log and semimlog terms in separate functions fer the extractable, hydrolyzable and nonehydrolyzable fractions. In the fourth function, all these ferms of nitrogen and carbon were used as separate log and semiolog terms in a single function. In the fifth, the sums of nitrogen and carbon in these three nonmwater—soluble frac— 'tions were used. Nitrate and other waterasoluble forms of nitrogen ‘were not included, since these were removed by leaching prior to plac- ing the germinated oats in contact with the experimental soil aliquots. In the first cropping period (table 16), #5 to 69 percent of the 'variation in green weight, dry weight and percent dry matter was ac- counted for by variations in nonéhydrolyzable nitrogen and carbon. The amount of infbrmation provided by this fraction alone was greater than fer the total non-water-soluble nitrogen and carbon and almost as great as in the feurth function involving all three fractions as separate 59 .sodpocda - mamcam m ca magma opmuwmom mm copnwo cam ammopvwc manwshaouvhnvcoc cam manmuhaopcmn .oanmpomupxm e mmo.+ :mo.+ Fmo.l Fmo.+ mmo.+ hmpems hue pcmopmm mmo.n mmo.n 300.: mro.c Nmo.u comoupfic camoumm moo.+ mmo.+ mmo.+ mma.+ moo.a mxmpms cmmoupfiz on". .+ 09.... moi... 3:. .+ 30.... 2303 b5 ome.+ mdP.+ m:o.+ va.+ omo.+ pnmflms comma coaumm mcammouo ccoomm . m:m.+ Nmm.+ moo.+ eoo.+ Nmo.l pmppws mac unmopmm 33.... mmm.+ 0%.... mmm.+ 49.... comofimfla 2.8.8.“ mmm.+ at}. new; 5.... $3.... 33% 8332 wom.+ mmm.+ :mw.+ wmo.n umr.u pnMflmz aha owl}. 59+ m9}. Nwo... So... Emma; compo uodpmm mcammouo pmuam o e5 .21 0 can D was a o a .Homapopmzococ coapocsm moanmanm> mono Hmpoe a confinsoo mHnmNmHOHUhnwcoz manmahaonehm canmpomupxm shop ppocm r .moapmanmi» dwom mo 3930ch now Umpcsooom 3 M .mGoapm:HDEOO econommac ca modpmdnm> Haom mo mcoapocdm mm moanmamm> mono shop puosm mom Ammv soapw:H3pmpmc oaawpada.wo mpcmaowmucoo a .mfi mHnme 60 terms. The extractable and hydrolyzable fractions were completely without effect on green or dry weight or percent dry matter. Nitrogen uptake and percent nitrogen, on the other hand, were influenced by variations in nitrogen and carbon in both the extractable and nonuhydrolyzable fractions. Percent nitrogen also reflected some influence of the hydrolyzable fractions. For these two plant variables, the fourth function provided a substantial increase in information over that supplied by the individual fractions or by the sum of nonmwatera soluble nitrogen and carbon. In the second crop, functional relationships with soil variables were much reduced in their contribution to variation in short term crop variables (table 16). Functional relationships calculated for unit observations in the first crop are depicted graphically in figure 3. Experimentally ob» served means for triplicate observations are also shown. The function for figure BmA accounted fer 73 percent of the varia- tion in dry matter percentage and provided statistically significant regression coefficients fer nonahydrolyzable carbon and hydrolyzable nitrogen. The major response indicated was to carbonaceous constituents supplied by both lignin and sawdust and appearing in the non-hydrolyzable fraction. The function, therefbre, substantiates the inference made from inspection of the data in tables 10 and 11. The basic response was modified by nitrogenous constituents in the hydrolyzable fraction, 81- though the observed range of this effect was inadequate to cover the range of observed.means. Additional increases above the mean function would.have been associated with extractable nitrogen, but these would have been countered by a depressing effect associated with extractable carbon. These moderating effects were not statistically significant but 61 Figure 3. a Functional relationships among first short term crop variables and soil variables. Figure 3-A. - Dry matter content (DM) of cats vs. nonmhydrolyzable carbon (NHC) and hydrolyzable nitrogen (HN) in soil: Log DM = + 1h.1399 - .145 log EXN + .0033 EXN + .916 log EXC - .0368 EXC - 7.47* log HN + .0145‘ HN _ .166 16g HC - .00001 HO - 3.46 log NHN + 00259 NHN + 1.31* log NHC - .0002 NH0 .2 R = .732 Figure 3-B. a Dry weight (DWt.) of oats vs. nonmhydrolyzable carbon (NHC) in soils Log DWt. = + 17.780 .. .243 1.0g EXN + .0088 EXN + 1.10 log Exc - .0434 EXC - 7.78 log HN + .0150 HN + .0385 log HC - .00004 HO - 4.69 log NHN + .0300 NHN + 1.80 log NHC — .0002 NHC -2 R = .725 Figure 3mC. - Nitrogen content (N) of cats vs. extractable nitrogen (EXN) in soils ~ 13.067* - .132 EXN + .0070** EXN - .533 log EXC + .0174 sxc + 4.80 log HN - .0092 EN - .0293 10g BC + .000003 HC .+ 2.27 log NHN - .0184 NHN - .165 10g NHC - .00002 NHC .2 R 2: 0923 Figure 3-D. - Nitrogen uptake (NUpt.) of cats vs. extractable nitrogen (EXN) and nonahydrolyzable carbon (NHC) in soils Log NUpt. =.,.4,714,_,375 lpg EXN + .0157** EXN + .570 10g EXC ; - .0260 EXC + 2.97 log HN + .006 HN + .067 10% H0 - .00001 HO - 2.42 log NHN + .0156 NHN + 1.68 log NHC .. .0003 NHC i2 = .779 * Regression coefficient significant at 5 percent. ** Regression coefficient significant at 1 percent. ( OATS I NITROGEN PERCENT (OATS) PERCENT ORY MATTER «T: N 2.2 2-O I-Z I-O 3-A 311137 cnop HN - 250 or 200 J 1 1 1 ISOO 2000 2500 3000 PPM NON-HYDROLYZABLE CARBON (eeIl) _ 3'0 FIRST CROP ® ® [3] _ Q) Li] 4369 ® IO 20 30 PPM EXTRACTABLE NITROGEN (SOIL) MILLIGRAMS ORY WEIGHT (OATS) IOOO GOO GOO 4OO ZOO MILLIGRAMS NITROGEN UPTAKE (OATS) O 3-8 _ FIRST CROP _ 1:151“?! 30 (3 KEY E) O No reeidue P (D C] LIgnIn O Sawdust I.2.3 - Applied " N leveI I l L 1 IOOO 2000 2500 3000 PPM NON-HYDROLYZABLE CARBON (eoIII L 3-0 FIRST CROP — AVE. NHC I 2444 r- m® V Pm® . \/ sac-1600 G) ® _ 1 1 1 IO 20 30 PPM EXTRACTABLE NITROGEN (SOIL) Fiqure 3 63 are theoretically realistic. As was observed in figure 1wA, dry weight was essentially a function of percent dry matter, whereas the effect of nitrogen content was minor. It is not surprising, therefore, that dry weight was also related primarily to nonmhydrolyzable carbon by the function in figure 3GB. Other soil variables contributed very little to explaining the deviation of experimental points away from this line. The functional relationships in figure 3mC suggest that percent nitrogen was influenced primarily by extractable nitrogen. Other soil variables, in their combined effect, accounted fer much of the deviau tion of experimental points away from this line. However, their in- dividual effects were small, and no attempt was made to show them in the figure. Nitrogen uptake in figure 3~D was influenced by soil factors which affected both dry matter production and nitrogen percentage. Thus, a major part of the variation in observed experimental means was en- compassed within the functional limits established by observed values fer extractable nitrogen and nonahydrolyzable carbon. This figure is strikingly similar to figure 1-C where relationships between nitrogen uptake and tissue dry matter and nitrogen contents are shown. The relationships in figure 3 clearly suggest that a primary stimulus to early growth of cats was expressed by some constituent in the nonéhydrolyzable carbon fraction. The stimulating factor was present in both sawdust and lignin obtained from the sawdust by acid extraction of cellulose and other carbohydrates. This stimulus was of the order of a hormonal response and was expressed specifically on processes of dry matter synthesis. In addition to this regulatory effect on dry matter production, 64 nitrogen uptake also reflected a greater availability of nitrogen in the extractable fraction. Nitrogen in the hydrolyzable fraction may have influenced the early response of cats, also, as indicated by the function for percent dry matter in figure 3mA. These relationships in the first short term crop of oats were asso- ciated with early stages of decomposition of the added residues. Similar relationships in the second crop were much less marked and were less ef- fective in accounting for variation in the measured crop variables (table 16). It appeared that effects associated with initial residue treat- ments were confounded with effects of root residues from the previous oat crop. As noted in discussion of figure 1aB, these root residues appeared to have an inhibitory influence on early growth of the second crop. Long term crop and soil variables Coefficients of multiple determination (fie) for functions applied to data for the long term cat crops are present in table 17. As seen in the first column, most of the variation in growth and nitrogen up- take in both crops was accounted fer by variations in nitrate and other water soluble forms of nitrogen. There was little change in R? fer these three crop measurements when waterasoluble nitrogen was replaced in successive functions involving extractable. hydrolyzable or non- hydrolyzable N and C. The fit was somewhat improved when all soil variables were included in a single function. In the first cropping period the use of total non-water-soluble nitrogen and carbon in a function with nitrate tended to detract slightly from the information obtained with nitrate and water-soluble nitrogen. Variations in percent nitrogen were well accounted for in the first crop by nitrate and water-soluble nitrogen. Some effect of the 65 .eoapocpm mamnam a ma manna mummwdom mm conpmo 6cm comoppfic manwuhaomemnwcoc new .oanmumaouehn .odpmpowupxo .cemoppa: mansaom poems .moowupaz e 3mm. med. mam. 0mm. BFN. 3mm. mopeds hue vemohom woo. woo. new. one. mmo. 0:0. comatose pcoopom Re. is. mom. was. 8m. Km. 23% memos; 6%. 2m. 6%. Em. Km. 3m. 23.... .86 :a. as. me. is. Na. ma. Ema... $36 coated mcwddomo becomm owe. own. man. Nwm. 3mm. awn. genome mac pcoommm mmm. mmm. new. 5mm. mum. 9mm. coMOMpHe peoonmm New. new. Sm. 3m. Rm. Nam. $.on 5632 :8. mam. . . 3mm. :9. «so. mam. 232. in 3.... as 9a.. 6:... am. as. saw... 8.3 beamed meaddono pmpfim o 6% a cdnzaom 0 one 2 0 one 2 0 one 2 numpmSecoc deuce :oapocdm mammahaouehnwcoc manmumaonehn manwpowupXo z .Howuhopms moanwfiam> mono .mopmapfiz pocaneoo .mmpmnpwz .mmpwupaz .mopmupaz .opwupflz sump mcoq .moanmfiaw> Haom mo mcoapocsm an new eopcsooom w ca hpdaanmfinm> e.um . H .mcodpmcaQEoo pcmnommde ca moanwdhm> Haom mo mcoflpocdm mm moanmamm> mono shop mcoa pom Ammv soapmcasuopoc camapase mo mucoaoammmoo 1 .mw capes 66 extractable fraction was apparent, but the hydrolyzable and nonm hydrolyzable fractions contributed little information beyond that which would have been supplied by nitrate alone. In the second crop, percent nitrogen was essentially a function of nitrate, although the hydrolyzable and nonmhydrolyzable fractions tended to contribute ad. ditional information. Percent dry matter was influenced by soil variables to a much lesser extent than the other crop variables. particularly in the second crop. Nevertheless, there was evidence that watercsoluble and extractable fractions contributed to variability in the first crop. In the second crop, there was a marked increase in F2 for dry matter when all fractions were combined in a single function (column 5, table 17) . Functional relationships between soil variables and dry weight and nitrogen uptake of both long term crops are presented in figure 4. Again the functions were fitted to unit observations. Experimental means for triplicate observations are shown for reference. Essentially all of the variation in dry weight was accounted for by the function for figure 4=A. Statistically significant regression coefficients in the function suggest that the most influential soil factors were nitrate and hydrolyzable carbon. Examination of the curves for low and high levels of hydrolyzable carbon would suggest that deviations from the average curve for nitrate were due to increas- ing immobilization of nitrogen in the presence of increasing carbonau ceous energy supplies represented by hydrolyzable forms of carbon. This would not, however, explain the two low points for the control and lignin treatment without nitrogen, since both of these were rela- tively low in hydrolyzable carbon (table 2). 67 Figure 4. - Functional relationships among long term crop variables and soil variables. Figure 4-A. - Dry weight (DWt.) of first crop of cats vs. nitrate (N03). hydrolyzable carbon (HC), soluble nitrogen (SN) and extrac- table nitrogen (EXN) in soil: Log DWt. = + 6.534 + 1.33“ log N03 - .0257“ N03 + .206 16g SN .0043 SN - .95? log EXN + .0274 EXN m .440 log EXC .0257 EXC ~ 5.35 log HN + .0139 HN + 1.18” log HC .0004* HC - 2.91 log NHN + .0443 NHN + 1.48 log NRC .0004 NRC -‘7 R” .".'.' , 99 3 Figure 4-B. - Dry weight (DWt.) of second crop of cats vs. nitrate (N03) and hydrolyzable nitrogen (HN) in soil: Log DWt. = + 4. 32.789** + .854** 10g N0 - .0046** NO + .0443 16g SN .0189 SN + .464 log EXN a .0054 EXN + 2.56 log EXC .193 Exc - 9.17** log HN + .0187** HN + 1.23 log HC .0002 HC - 9.56 log NHN - .0728 NHN - 2.64 log NHC .0006 NRC Figure 44C. - Nitrogen uptake (NUpt.) of first crop of cats vs. nitrate (Egg , hydrolyzable carbon (HC) and extractable nitrogen ( ) Log NUpt. = + in soil: 13.187 + 1.21** log NO - .0178** N0 + .416* 16g SN .0067 SN - 1.38 log EX + .0402 EXN a 1.37 log EXC .0043 3x0 - 9.86 log HN + .0251 HN + 1.73 log HC .0007** HC - 6.25 log NHN + .0845 NHN + .248 NRC .0007* NRC a? = .995 Figure 4-D. - Nitrogen uptake (NUpt.) of second crop of cats vs. nitrate (N03), and hydrolyzable nitrogen (HN) in soil: Log NUpt. = + +1! 16.891** + .932** 10g N03 - .0032** N03 + .0430 10g SN .0070 SN + .112 log EXN + .0020 EXN + .592 log EXC .0610 Exc - 4.16** log HN + .0092** HN + .642 10g HC .0001 HC - 7.53 log NHN + .0584 NHN _ 1.11 log NRC .0002 NRC R2 = .974 * Regression coefficients significant at 5 percent. ** Regression coefficients significant at 1 percent MILLIGRAMS NITROGEN UPTAKE (OATS) GRAMS DRY WEIGHT (OATS) (I O J 0 0| 0 N O 3 I200 IO 00 GOO GOO 4OO ZOO O 4-A Inner asap l-lc-IZOO _Ex11-2s /l§]% <3) wane-1794 mess-29 r )AVE.EXN-IS su-z _ ® |@ Hc-zeoo 1. ' l I l 1 l J 5 10 1s 20 25 so PPM 1103-11 (3011.) 1' 4'6 FIRST CROP © l-lc-IZOO E] 51111-25 *— ~ Hc-IZOO 9 v exu-Ie AVE.HC IITSQ /" AVE.EXN8 I6 )6; 5 IO I5 PPM NOf-N ZO ZS 30 (SOIL) 20 _ 4 - B . SECOND CROP :2 0 g l6~ ‘9 1- [Z] : ® 2 12+ II.) 3 / AVE. MN I Z|4 z:- 3 1111-150 or 250 O (D 2 44 C o k 0 1 1 1 J 50 I00 ISO 200 PPM NOS-N (SOIL) 50°F HN- 240 sscoua “Oi/\u CROP 500 ’4EJ'IN'ZI5 4OOL KEY O Na resldue 300 C] LIgnln O Sawdust ZOO IOO Bald faced symbols -Prevlausly uncropped l.2.3 - Applled N level MILLIGRAMS NITROGEN UPTAKE (OATS) 0 Figure 4 W®\g& I l I ISO ZOO (SOIL) 1 I00 N03- N 0_ 0 PPM 69 The effect of varying other soil variables one at a time in the function fOr figure 4aA revealed that deviations below the average function were associated primarily with variations in watermsoluble ni- trogen other than nitrate. Deviations above the average function were associated with variations in extractable nitrogen. As seen in table 14, these forms of nitrogen, as well as nitrate, were negatively cor- related with hydrolyzable carbon. Because of these and other inter- correlations, the depressing effect of’hydrolyzable carbon was exag- gerated by transfer to hydrolyzable carbon terms in the function of effects associated with other soil variables. It is likely that a function which would permit more flexibility in expression of interaction effects would be more useful in segreu gating out specific effects of individual variatles. Nevertheless. relationships brought out by the function are meaningful and reasonable. Distortions can be rationalized by reference to observed experimental values. Functional relationships for nitrogen uptake by the same crop (figure 4-C) were essentially similar. The major response to nitrate was influenced by a depressing effect of hydrolyzable carbon. A portion of this depressing effect, however, was transferred to non- hydrolyzable carbon by reason of the high positive correlation between these two forms of carbon and their.negative correlation with nitrate. Deviations below the average curve were again associated with variations in soluble nitrogen, deviations above the curve with extractable nitrom gen. In the second crop (figures 4mB and 4mD) the depressing effect of hydrolyzable carbon was greatly minimized by the function. This was due in part to a much reduced intercorrelation between nitrate and 70 hydrolyzable carbon (cf. tables 14 and 15). More significantly. however, nitrogen which had accumulated residually in the hydrolyzable fraction was released during the second cropping period and added appreciably to the supply of available nitrogen represented by nitrate present at planting time. Some experimental means fall outside the functional limits set by observed mean values for nitrate and by... drolyzable nitrogen, indicating that minor effects were also assom ciated with other soil or environmental variables. In the first crop, an optimum for dry matter production was exm pressed at a soil nitratemnitrogen level of about 22 ppm. (figure 4nA), whereas the optimum for nitrogen uptake was somewhat higher9 about 28 ppm. (figure 4mC). In the second crop, the functional optimum for dry weight was about 85 ppm. (figure 44B), and the optimum expressed for nitrogen uptake was about 130 ppm. nitrateenitrogen (figure 4aD). These differences reflect the greater efficiency of dry matter ace cumulation during the short days and cool greenhouse temperatures of late winter and spring when the first crop was grown, as compared with greater respiratory dissipation of products of photosynthesis under the high summertime temperatures to whiCh the second crop was exposed. The larger quantities of nitrogen available to the second crop tended to compensate for the unfavorable environmental conditions. However, as may be seen in figure 5, the major increases in dry weight in both crops occurred at soil nitrate levels less than 30 ppm. nitrate- nitrogen. On this basis, nitrogen taken up by the second crop (figure 4-D) in excess of about 300 mg. exceeded the capacity of the crop to efficiently assimilate it. As noted in the dis¢Ussion of figure 2—D, such excessive accumulations of nitrogen in tissue occurred at nitrogen contents in excess of 2.2 percent. In the first crop a distinct optimum 71 .meflp ochCmHo om Hfiom one Cw cmoomowc memmec .m> macho umo Emmy ecoH pcoomm pom pmpwm age we “Lewes >HQ 1 m masowm Anzomv mhdmtz San. ON 00 On 0? on ON 0. O _ q H 1 _ — a _ 00.1mm ozanmo 0200mm e61 I®11 .05. 2 02:34 a m.m._ Seesaw 0 see: D 323. oz 0 >mx _ 16 ® 6% any—mun. wzimomu ...mm..... av O IHOIBM A30 SWVHO O N 0 [O 0 ¢ (SIVO) 72 for dry matter production was expressed at 2.0 percent nitrogen (figure 2.11). Thus it would appear that 20 to 30 ppm. nitrateanitrogen in soil and 2.0 to 2.2 percent nitrogen in tissue may represent critically optimum levels for growth of cats under widely varying environmental conditions. Functional analysis as employed in this study provides a tool for identifying such critical optima with greater precision than is possible by examination of means by conventional analysis of variance. SUMMARY AND CONCLUSIONS Large changes in the chemical nature of carbon and nitrogen com- pounds in soil were observed during the first three month's decomposi- tion of added sawdust. ‘watermsoluble and saltmextractable forms of nitrogen were initially immobilized, appearing primarily in the acid- hydrolyzable fraction after five weeks. This initial immobilization was due almost entirely to hydrolyzable forms of carbon and did not occur with lignin isolated from the sawdust by acid hydrolysis and removal of cellulose and other carbohydrates. Small, but statistically significant. increases in nonmhydrolyzable nitrogen were observed after 5 weeks with both sawdust and lignin. These increases were greater with increasing levels of added urea nitro- gen. It appeared likely that immobilization of nitrogen in nonAhydroe lyzable forms was due to the formation of chemical complexes between lignin degradation products and ammonia. Immobilization in.hydrolyza able forms was probably due mainly to incorporation of mineral nitrogen into microbial proteins. Where urea nitrogen was added to give CsN ratios for sawdust of 40:1 and 2031. peak immobilization of nitrogen occurred prior to the sampling made 5 weeks after amendment. During the next 10 weeks, ad- ditional immobilization of nitrogen in this fraction occurred where no nitrogen was used with sawdust, and in the presence of roots of the first long term crop of cats. In sawdust amended soil, this continued immobilization was associated with extensive conversion of nonmhydro- lyzable forms of carbon to hydrolyzable forms. Growth and uptake of nitrogen in 54 days by the first long term 73 74 crop of cats was influenced primarily by levels of nitrate and hydrolyze able carbon found at planting time, 5 weeks after amendment. Multiple regression analyses revealed that, in addition to nitrate, other forms of nitrogen (primarily ammonium), which were soluble in water or exo tractable with N.K2504 at pH 1.5 to 2.0, contributed extensively to nitrogen availability. The effect of increasing levels of hydrolyzable carbon was to depress the availability of these three forms of nitrogen. Growth and uptake of nitrogen in a similar period by a second oat crop planted 105 days after amendment was essentially a function of nitrate present at planting time, although a significant contribution of nitrogen from the hydrolyzable fraction was indicated. In both long term crops. 91 to 99 percent of the observed varia- tion in dry weight and nitrogen uptake was accounted for by functional relationships defined by an exponentialmpower function in which all measured forms of nitrogen and carbon were included as separate terms. In both crops, near optimum production of dry matter was associated with soil tests of 20 to 30 ppm. nitratemnitrogen and tissue nitrogen contents of about 2.0 percent. Optimum levels of soil nitrate and tissue nitrogen for nitrogen uptake were higher but represented nitrom gen taken up in excess of the crop's ability to assimilate the nitrogen. An unexpected result was encountered 5 weeks after amendment when nitrate and other soluble materials were removed by leaching with water before placing the roots of nitrogenudeficient oat plants in contact with the soil. Dry matter synthesis in the first 11 days of growth was stimulated independently of nitrogen uptake by some factor or factors present in both lignin and sawdust. Multiple regression analysis asm sociated the stimulating activity primarily with the nonmhydrolyzable carbon fraction. Reduced percent nitrogen in harvested tissue and 75 increases in nitrogen taken up in the same period reflected increased dry matter production and increased vigour, but were additionally in- fluenced by the supply of nitrogen present in saltmextractable forms in the soil. When nitrogen uptake was plotted against percent nitrogen in tissue, the distribution of experimental points and the expressed func- tional relationShips with percent dry matter were analogous to those obtained when nitrogen uptake was plotted against extractable nitrogen and varying levels of nonmhydrolyzable carbon in the soil. In both cases, about 80 percent of the total variation in nitrogen uptake was accounted for by functional relationships defined in the mathematical model. Similar stimulating effects of sawdust and lignin on short term growth and nitrogen uptake were observed in the second short term oat planting made 105 days after amendment. However, this stimulating action was accompanied by what appeared to be an inhibitory effect of root residues from the first long term crop of cats in previously cropped pots. Neither the stimulatory activity of sawdust and lignin nor the inhibitory activity in root residues was clearly associated with any measured fraction of carbon or nitrogen by functional rela» tionships defined in the mathematical model. In spite of shortcomings in the exponentialapower function which was employed as a mathematical model, its use did lead to meaningful inferences regarding the contribution to plant response of various fractional forms of carbon and nitrogen in the soil. The fractions measured were particularly meaningful at early stages of decomposition of sawdust incorporated into the soil. After decomposition had pro- ceeded for three months, the contribution of carbon and nitrogen frac- tions, other than nitrate, was only weakly expressed by the function. 76 This is not surprising, since the hydrolytic fractions studied rem present gross fractions in which forms of carbon or nitrogen with high specific activity would have been masked by mixture with forms of little or no significance for crop response. Nevertheless, the degree of functional association which was encountered indicated that multiple correlation and regression analysis is a useful tool for isolating activity in chemically characterizable fractions of carbon and nitrogen in soil. In future work, more precisely characterizable fractions or groups of organic compounds should be considered. Not all of the responses of crops to additions of carbonaceous residues are to be understood in terms of their effects on nutrient availability. The data reported here indicate that specific growth regulatory effects are associated with early stages of decomposition of crop residues remaining or added to the soil. These effects may be inhibitory as well as stimulatory and may involve carbon compounds which contain little or no nitrogen. These hormonal effects of carbon compounds may be as important as the immobilizing effects of care bonaceous energy materials in influencing the effective nitrogen supplying power of soils. Short term growth or nitrogen uptake of plants may be very misleading when used as a bicassay for the availa- bility of insoluble forms of nitrogen in soil recently enriched with crOp residues. Of direct practical importance was the observation that inhibitory effects on long term growth of cats, due to nitrogen immobilization. had completely disappeared 3 months after addition of sawdust at the rate of 10 tons per acre when 200 pounds per acre of urea nitrogen was also applied. This quantity of nitrogen was calculated to give a CzN ratio for the sawdust of 4031. No reduction in yield occurred after 77 5 weeks when nitrogen was applied at the rate of’400 pounds per acre to give a CsN ratio of 20:1. 10. 11. 12. 13. 14. 15. BIBLIOGRAPHY Allison, F. E. The enigma of soil nitrogen balance sheets. In, A. G. Norman, ed. Advances in Agronomy Vol. VII. Acad. Press, Inc., New York. pp. 2133237. 1955. Allison, F. E. Estimating the ability of soils to supply.nitrogen. Agr. Chem. 11 (No. 10:46:48. 1956. Allison. L. E. 'Wetacombustion apparatus and procedure for organic and inorganic carbon in soil. Soil Sci. Soc. Amer. Proc. 24:36m “O. 1960. BartholomeW5'W. V. Availability of organic nitrogen and phosphorus from plant residues, manures and soil organic matter. Papers of the Soil Microbiology Conference. Perdue University. 195a. , and Hiltbcld, A. E. Recovery of fertilizer nitro_ gen by cats in the greenhouse. Soil Sci. 733193~7 i. 1952. Brauns. F. E. The Chemistry of Lignin. Acad. Press Inc., New York. p. 157. 1952. Bremner. J. M. A review on recent work on soil organic matter: I. J. Soil Sci. 2:67m82. 195i. . The nature of soil nitrogen complexes. J. Sci. Food Agr. 3:497a500. 1952. ' . Studies on humic acids. J. Agr. Sci. b83352“ 360. 1956:— . Determination of fixed ammonium in soil. J. Agr. 5°10 523147.3160. 19590 . Determination of nitrogen in soil by the Kjeldahl method. J. Agr. Sci. 55:11a33. 1960. , and Shaw, K. The mineralization of some nitrogenous materials in soils. J. Sci. Food Agr. 8:3h1m347. 1957. Broadbent, F. E. Nitrogen release and carbon loss from soil organic matter during decomposition of added plant residues. Soil Sci. Soc. Amer. Proc. 12:246-249. 1947. . The soil organic fraction. Ln; A. G. Norman, ed. Advances in Agronomy Vol. V. Acad. Press Inc., New York. pp. 153- 181. 1953. tranmendental production fimction. Jour. of Farm Econ. 393966.. 971‘". 1957' 78 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 79 Clark. F. E.. Beard9 W. F., and Smith, D. H. Dissimilar nitrifying capacities of soil in relation to losses of applied nitrogen. Soil Sci. Soc. Amer. Proc. 24:50-5u. 1960. DeMent, J. D., Stanfbrd, 0.. and Hunt, C. M. A method for measur- ing shortaterm nutrient absorption by plants: III. Nitrogen. Soil Sci. Soc. Amer. Proc. 233371w374. 1959. Dyck, A. W. J., and.McKibbin, R. R. The nonmprotein nature of a fraction of soil organic nitrogen. Can. J. Res. 133264e268. 1935. Ensminger,. L. E.. and Pearson. R. W. Soil nitrogen. IQ,A. G. Norman, ed. Advances in Agronomy Vol. II. Acad. Press, Inc., New York. pp. 81‘1090 19500 Fraps, G. S., and Sterges, A. J. Estimation of nitric and nitrous nitrogen in soils. Texas Agr. Exp. Sta. Bul. h39..pp..19~21.. 1931. Fraser, 0. K. Soil organic matter. In F. E. Bear, ed. Chemistry of the Soil. Reinhold Pub. Corp.. New York. pp. 199m176. 1955. Gottlieb, S., and Hendricks, S. B. Soil organic matter as related to newer concepts of lignin chemistry. Soil Sci. Soc. Amer. Proc. 103117-125, 1945. Harmsen, G. W., and Van Schreven, D. A. Mineralization of organic nitrogen in soil. gn’A. G. Norman, ed. Advances in Agronomy Vol. VII. Acad. Press, Inc., New York. pp. 300m383. i935. Hiltbold, A. E., and Adams, F. Effects of nitrogen volatilization on soil acidity changes due to applied nitrogen. Soil Sci. Soc. Amer. Proc. 24345:.47. 1960. Jackson. M. L. Soil Chemical Analysis. PrenticemHall, Inc.. Englewood Cliffs, New Jersey. pp. 183m201. 1958. Johnston, H. H. Studies on the chemical, physical and biological properties of soil organic matter. Ph. D. Thesis, Department of Soil Science, Michigan State University. 1958. Kamerman. P.. and Klintworth, H. Influence of fertilizer on the nitrogen and carbon cycles in soils. Union of South Africa, Department of Agr. Sci. Bul..137:1¢26. .1934. . Lyon. T. L.. Buckman, 0. H.. and Brady, N. C. The Nature and Properties of Soils. The Macmillan Company, New York. pp. #55. Mattson. 5., and Koutler-Andersson, E. The acidmbase condition in vegetable litter and humus: VI. Ammonia fixation and humus nitro- gen. Annals Agr. College Sweden 11:107a13u. 1943. 31. 32. 36. 37. 38. 39. 41. 42. 43. 80 Mortland, M. M., and.wolcott, A. R. Sorption of inorganic nitrogen compounds by soil materials. (MSS. submitted for pub- lication ig,Soil Nitrogen. Monograph series. American Society of Agronomy.) 1963. Norman, A. G. The Biochemistry of Cellulose, the Polyuronides, and Lignin. Oxford, at the Clarendon Press. 1937. Norman, A. G. Problems in the chemistry of soil organic matter. Soil Sci. Soc. Amer. Proc. 727m15. 1942. Parker, D. T., Parson, W. E., and Bartholomew, W. V. Studies on nitrogen tienup as influenced by location of plant residues in soils. Soil Sci. Soc. Amer. Proc. 21:608m611. 1957. Peterson, L. A., Attoe, O. J., and Ogden, w. B. Correlation of nitrogen soil tests with nitrogen uptake by the tobacco plant. Soil Sci. Soc. Amer. Proc. 243205m209. 1960. Pinck, L. A., Allison, F. E.. and Caddy, V. L. The nitrogen requirements in the utilization of carbonaceous residues on soils. Agron. Jour. 382410.420. 1946. Rodrigues, G. Fixed ammonia in tropical 50115. J. Soil Sci. 53264c274. 1954. Schwartzbeck, R. A., MacGregor, J. M., and Schmidt, E. L. Gaseous nitrogen losses from nitrogen fertilized soils measured with infrared and mass spectroscopy. Soil Sci. Soc..Amer..Proc. 25:186-189. 1961. Singh, B. N. Mineral and organic forms of nitrogen in some Michigan soils and an agromeconomic evaluafigq of their poten_ tial userIness for advisory purposes. Pb. D. Thesis, Department of Soil Science, Michigan State University. 1960. Smith, G. E. Soil fertility and corn production. Missouri Agr. Exp. Sta. B111. 583. 1952. Stevenson, F. J. Effects of some longutime rotations on the amino acid composition of the soil. Soil Sci. Soc. Amer. Proc. 20: 204-208. 1956. ‘ Sundquist, W. B., and Robertson, L. 8., Jr. An economic analysis of some controlled fertilizer inputmoutput experiments in Michigan. Michigan State Univ. Tech. Bul. 269. 1959. waksman, S. A. Soil Microbiology. John Wiley & Sons, Inc., New York. 1952. . and Iyer, K. R. N. Contributions to our knowledge of the chemical nature and origin of humus: I. On the synthesis of the ”humus nucleus". Soil Sci. 34:43~69. 1932. 45. 46. 47. 48. 50. 51. 520 81 , . Contributions to our knowledge of the chemical nature and origin of humusz II. The influence of ”synthesized” humus compounds and of "natural humus" upon soil microbiological processes. Soil Sci. 34371m79. 1932. . Contributions to our knowledge of the chemical nature and origin of humuss III. The base exchange capacity of ”synthesized humus" (ligno protein and "natural humus" complexes). Soil Sci. 36257s67. 1933. . Contributions to our knowledge of the chemical nature and origin of humus 3 IV. Fixation of proteins by lignins and formation of complexes resistant to micron bial decomposition. Soil Sci. 36369m82. 1933. waksman. S. A., and Tenney, F. G. On the origin and nature of the soil organic matter or soil "humus": IV. The decomposition of the various ingredients of straw and of alfalfa meal by mixed and pure cultures of microorganisms. Soil SCi. 223395m406. 1926. . Composition of natural organic materials and their decompos1tion in the soil. II. Influence of age of plant upon the rapidity and nature of its decompositionmm rye plants. .Soil Sci. 243317~333o 1927. White, A.'W..Jr., Giddens, J. E., and Morris. H. D. The effect of sawdust on crop growth and physical and biological properties of Cecil soil. Soil Sci. Soc. Amer. Proc. 233365-368. 1959. White, W. C., and Pesek, J. Nature of residual nitrogen in Iowa soils. Soil Sci. Soc. Amer. Proc. 23339m42. 1959. welcott, A. R.. and Johnston, H. H. Immobilization and mineraliza- tion of nitrogen in soils following organic amendments. Agron. Abstr. p. 17. 1958. WOodruff, C. M. Estimating the nitrogen delivery of soil from the organic matter determination as reflected by the Sanborn field. Soil Sci. Soc. Amer. Proc. 14:208w212. 1949. G 9 an" ‘9 3":-