MINERAL AND ORGANIC FORMS OF NITROGEN IN SOME MICHIGAN SOILS AND AN AGED-ECONOMIC EVALUATION OF THEIR POTENTIAL USEFULNESS FOR ADVISORY PURPOSES TI‘Iesls Ior ”10 Degree OI le. D. MICHIGAN STATE UNIVERSITY Bhubneshwar Narain Singh 1960 This is to certify that the thesis entitled Mineral and Organic Forms of Nitrogen in Some Michigan Soils and an Agro-Economic Evaluation of Their Potential Usefulness for Advisory Purposes presented by Bhubneshwar Narain Singh has been accepted towards fulfillment of the requirements for Doctorate degree in Soil Science r"; -/ . f) ”Hi A?” L!" r {1:1' L Major professor Date September 12, 1960 0-169 LIBRARlT-LI Michigan State University I MINERAL AND ORGANIC FORMS OF NITROGEN IN SOME MICHIGAN SOILS AND AN AGRO-ECONOMIC EVALUATION OF THEIR POTENTIAL USEFULNESS FOR ADVISORY PURPOSES BY Bhubne shwar Narain Singh AN ABSTRACT Submitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Soil Science Field of Agronomic Extension 1960 l I ’L/ ) ,311/ ,/ -_;_ Approved [for f/Mw/Jz’ v I Bhubne shwar Narain Singh ABSTRACT Air-dry soil samples from five established field experiments were analyzed for exchangeable ammonium, nitrate, and two organic nitrogen These two organic fractions included the portion hydrolyzed fractions. by digestion with strong sulfuric acid and the portion resistant to acid hydrolysis. Attempts were made to correlate these measured forms of nitrogen with crop yields. Ammonium levels in air-dry soils were several times higher than would be expected in field fresh soils, indicating release by breakdown of soil organic materials during storage. The quantities found were higher in soils high in total organic nitrogen than in soils low in organic nitrogen. There was no relationship to crop yields or to residual yield variance not explained by current fertilizer treatments. Nitratevlevels in soils sampled in the fall of the year reflected rotational differences and levels of previous nitrogen application. In soil samples taken in the spring, nitrate was low and unrelated to prior treatment. No correlation with crop yields or yield residuals was observed. The two organic fractions and their total showed a tendency to increase with increasing level of nitrogen applied one year previously. Where However, these increases were not statistically significant. supplemental nitrogen had been applied on corn, beans and barley in a rotation including two years of alfalfa-brome, significant increases in each fraction and in their total were observed at the end of the first five-year rotational cycle. The increases in total organic nitrogen ranged from 352 to 648 pounds per acre, exceeding by a factor of 3 to 5 ii Bhubne shwar Narain Singh the 120 pounds total supplemental nitrogen which had been applied on the three crops preceding alfalfa. In a second experiment on the same soil type (Sims clay loam) where supplemental nitrogen had been applied on row crops and cereal grains over four cycles of two 5v~year rotations, no significant increases in soil organic nitrogen were found. Residual organic nitrogen was significantly higher, by 400 pounds, in the livestock rotation which included manure and two years of alfalfa than in the cash crop rotation. The ratio of nonhydrolyzable to hydrolyzable nitrogen varied under different systems of management. The prOportion of nonhydrolyzable nitrogen was higher where alfalfa was included in the rotation, or where supplemental nitrogen was used. This effect of supplemental nitrogen was enhanced when combined with a high rate of application of other fertilizer nutrients. In one experiment a maximum of 61 percent of yield variance was found to be associated with regression in a five-variable polynomial equation involving either total organic nitrogen or hydrolyzable nitrogen. However, most of the variance was associated with rotation or supple- mental nitrogen treatments. A maximum of 26 percent of yield variance was associated linearly with total soil organic nitrogen when rotation, fertility level and supplemental nitrogen treatments were ignored. Only 15 percent and 5 percent of yield variance were similarly associated with hydrolyzable and nonhydrolyzable nitrogen, reSpectively. iii MINERAL AND ORGANIC FORMS OF NITROGEN IN SOME MICHIGAN SOILS AND AN AGRO-ECONOMIC EVALUATION OF THEIR POTENTIAL USEFULNESS FOR ADVISORY PURPOSES BY Bhubne shwar Na rain Singh A THESIS Submitted to‘the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOC TOR OF PHILOSOPHY Department of Soil Science Field of Agronomic Extension 1960 ACKNOWLEDGMENT The author wishes to express his sincere appreciation to Dr. A. R. Wolcott for patient counsel and assistance throughout the course of this study. He is particularly indebted to Dr. R. L. Cook who made arrangements for financial assistance and who was instru— mental in arranging the author's return for a second period of graduate study at Michigan State University. Special thanks are due Dr. 0. Ulrey, US-TCM Expert, located at Ranchi Agricultural College, Bihar, India, for his interest and aid in securing a travel fellowship from the Council on Economic and Cultural Affairs, Inc. The financial assistance from the council is gratefully acknowledged. He is also grateful to Dr. Glenn L. Johnson for his valuable suggestions and criticisms during preparation of the manuscript, and to Mr. B. Hoffnar for his help in the statistical computations. ************** TABLE OF CONTENTS Page INTRODUCTION.............. ........... . 1 LITERATURE REVIEW ...................... 3 .............. 3 4 Nature of Soil Organic Nitrogen Availability of Organic Nitrogen .......... . . . . Procedures for Determining Nitrogen Availability ..... 11 MATERIALS AND METHODS ................... 16 Field Treatments and Cropping Histories ......... 16 Fertilizer experiments ................ 16 Residue experiment . ............... . l7 Rotation experiment ............... . . .19 Laboratory Determinations ................. 20 Exchangeable ammonium ............... 20 Nitrate ..... y. ................... ZO Hydrolyzable and nonhydrolyzable nitrogen . . . . . 21 EXPERIMENTAL RESULTS ................... 23 Fertilizer Experiments ................... 23 Residual effects of fertilizer on ammonium and nitrate levels ...... .l . . . . . . . . . . . . 23 Residual effects on organic forms of nitrogen 30 3O Relation of treatment and nitrogen fractions to yield. 31 Organic Residue Experiment ...... . . . ....... 32 0000000000000 Inorganic forms of nitrogen Organic forms of nitrogen . .......... 32 Corn yields . . . . . . . . ............ . 34 Ferden Farm Rotation Experiment . . . ..... . . . . . 35 Inorganic forms of nitrogen . .......... . . 35 Organic forms of nitrogen . . . . . . . . ...... 35 39 Corn yields . . Production Function Analysis of Corn Yields for Rotations . . . . . 43 1and6 . . . ............. 44 Rotations 1 and 6 vi TABLE OF CONTENTS .- Continued Page 44 O O 0 Total nitrogen as soil nitrogen variable Hydrolyzable nitrogen as soil nitrogen vari- ° 0 0 46 able .................. Nonhydrolyzable nitrogen as soil nitrogen variable ................. . . 46 Rotation 1 . .................... . . 47 . . . 47 Total nitrogen as soil nitrogen variable . Hydrolyzable nitrogen as soil nitrogen vari» able ..... . ........... . . . . 48 49 Rotation 6 ....................... Total nitrogen as soil nitrogen variable 49 Hydrolyzable nitrogen as soil nitrogen vari- o o o o 49 able ...... . . . . . . DISCUSSION .......................... 51 Significance of Mineral Forms of Nitrogen ........ 51 Nitrate ......................... 51 Ammonium ................... . . . 52 ........ 53 Significance of Organic Forms of Nitrogen Factors affecting the proportion of nonhydrolyzable 54 00000000 to total nitrogen . ...... . . Significance of nonhydrolyzable nitrogen to nitrogen availability . . ............. . ..... 58 Significance of hydrolyzable nitrogen to nitrogen availability . . . . . . . . . . ........... 59 Evaluation of Experimental Designs . . ..... . . . . . 60 'Evaluation of Functional Analyses . ........... 62 Agro-Economic Considerations . ........... . 66 SUMMARY AND CONCLUSION ................. 69 BIBLIOGRAPHY . . . . . . ....... . . . ..... . . . . . 71 APPENDIX I--Individual Observations ....... . . . . . . . 78 APPENDIX II--Computational Formulae Used in Functional o o o o o o o ..... 99 Analyses . . . . . . APPENDIX III--Locations of Field Experiments and Soil Type Descriptions. . . ...... . . 103 vii LIST OF TABLES Page TABLE 1. Effects of one annual application of N, P and K on residual forms of soil nitrogen and yields of oats follow- ing a second annual application of the same fertilizer Fick farm. Kalamazoo sandy loam. 1955- treatments. 56. . ..... . . . Effects of N, P and K on residual forms of soil nitrogen, observed yields of oats, and yield residuals not function- ally associated with current nutrient inputs. Fick farm. 1955—56. . . . ......... Kalamazoo sandy loam. Effects of one annual application of N, P and K on resid- ual forms of soil nitrogen, and yield of wheat following a second annual application of the same fertilizer nutrients Campbell farm. Kalamazoo sandy loam. 1955-56. . . . Effects of N, P and K on residual forms of soil nitrogen, observed yields of wheat, and wheat yield residuals not functionally associated with current nutrient inputs. Campbell farm. Kalamazoo sandy loam. 1955-56. . . . Effects of one annual application of N, P and K on resid-- ual forms of soil nitrogen, and yields of beans following a second annual application of the same fertilizer nutrients. Thompson farm.‘ Sims loam. 1955-56. . . . Effects of N, ‘P and K on residual forms of soil nitrogen, observed yields of beans, and bean yield residuals not functionally associated with current nutrient inputs. Thompson farm. 1955-56 ...... . . . . ...... Residual effect of residue treatment and supplemental nitrogen treatment on mineral and organic forms of Ferden farm. Sims clay nitrogen and yields of corn. loam. 1959 ....... . ........ Residual effect of rotation, fertility 1eVel and supple- mental nitrogen treatments on mineral and organic forms of nitrogen. Ferden farm. 1958 ........... . . viii 24 25 27 28 29 33 36 LIST OF TABLES - Continued TABLE 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Correlation and regression of hydrolyzable and non- hydrolyzable forms of nitrogen on total organic nitrogen in Sims clay loam under three systems of management . Percentage nonhydrolyzable nitrogen as related to supplemental nitrogen treatment and levels of applied 0 O O O O O O O nutrients. Sims loam. . . . . . . . . Percentage nonhydrolyzable nitrogen as related to supplemental nitrogen and residue treatment. Sims loam ..... . Percentage nonhydrolyzable nitrogen as related to level of applied nutrients. Kalamazoo sandy loam. Fick farm....... Residual ammonia and nitrate nitrogen one year after fertilizer treatment and yield following a repeated appli- Kalamazoo sandy cation of N, P and K. Fick farm. loam. 1955-56........... Residual hydrolyzable and nonhydrolyzable nitrogen one Fick farm. Kalamazoo year after fertilizer treatment. sandy loam. 1955’56 o o o o o o o o o a o o o o o o o o 0 Residual inorganic and organic nitrogen one year after Fick farm. Kalamazoo sandy fertilizer treatment. loam. 1955-56 . . . . . Residual ammonia and nitrate nitrogen one year after fertilizer treatment and yield following a repeated application of N, P and K. Campbell farm. Kalamazoo sandy loam. 1955-56 . . . . . . Residual hydrolyzable and nonhydrolyzable nitrogen one year after fertilizer treatment. Campbell farm. Kalamazoo sandy loam. 1955-56. . . . ......... Residual total inorganic and organic notrogen one year Campbell farm. Kalamazoo after fertilizer treatment. sandy loam. 1955—56 . . . . . . . . Residual ammonia and nitrate nitrogen one year after fertilizer treatment and yield following a repeated appli— cation of N, P and K. Thompson farm. Sims loam. 1955-56. ..... . . . . . . . ix Page 56 56 57 79 80 81 82 83 84 85 LIST OF TABLES - Continued TABLE Page 20. Residual hydrolyzable and nonhydrolyable nitrogen one year after fertilizer treatment. Thompson farm. Sims loam. 1955-56. . . . . ....... . . . . . . . 86 21. Residual inorganic and organic nitrogen one year after fertilizer treatment. Thompson farm. Sims loam. 1955-56............. ..... 87 22. Ammonia nitrogen in Sims clay loam as affected by or- ganic amendmentsand fertilizer nitrogen.= . . . . . . . 88 23. Nitrate nitrogen in Sims clay loam as affected by organ- ic amendments and fertilizer nitrogen ...... . . . . 89 24. Inorganic nitrogen in Sims clay loam as affected by organic amendments and fertilizer nitrogen ....... 9O 25. Hydrolyzable nitrogen in Sims clay loam as affected by organic amendments and fertilizer nitrogen ..... '. . 91 26. Nonhydrolyzable nitrogen in Sims clay loam as affected by organic amendments and fertilizer nitrogen ..... 92 27. Organic nitrogen in Sims clay loam as affected by or- ganic amendments and fertilizer nitrogen. . . ..... 93 28. Corn yields in Sims clay loam as influenced by organic amendments, fertilizer nitrogen ........ . . . . . 94 29. Ammonia and nitrate nitrogen in Sims clay loam as influenced by cropping sequence, fertility level, and supplemental nitrogen treatment ........ . . . . . 95 30. Hydrolyzable and nonhydrolyzable nitrogen in Sims clay loam as influenced by cropping sequence, fertility level, and supplemental nitrogen treatment ....... 96 31. Mineral and organic nitrogen in Sims clay loam as influenced by cropping sequence, fertility level, and supplemental nitrogen treatment ............. 97 32. Corn yields in Sims clay loam as influenced by crop- ping sequence, fertility level, and supplemental nitro- gen treatment .......... . . ....... . . . . 98 X LIST OF FIGURES FIGURE Page 1. Correlation and regression of the ratio of nonhydro— lyzable nitrogen to hydrolyzable nitrogen on total organic nitrogen in Sims clay loam under two 5~year rotations containing alfalfa-brome. . . . . . . . . . . . . ..... 40 2. Basic response of corn to rotation treatments as related to supplemental nitrogen treatment, fertility level, and total organic nitrogen (A), hydrolyzable nitrogen (B), and nonhydrolyzable nitrogen (C) in Sims clay loam as reflected in treatment means. . . . . . . . ........ 42 xi INTRODUCTION Decisions that farmers make regarding the use of fertilizers are based primarily upon the economic returns they hepe to realize from the investment. Estimates of fertilizer costs and expected returns must be weighed against similar estimates for alternative production factors. Decisions resulting from such weighting of alternatives may be quali- tative, resulting in total rejection or adoption of a given practice. Or they may be quantitative and expressed in terms of how much of one factor should be substituted for how much of another. Agronomic research and farmers' experience have built up com- pelling evidence for the qualitative affirmative decision to use fertilizer. They have also established broad quantitative limits to the range of pounds-per-acre inputs over which the decision is valid. These limits have been established with some degree of refinement for different crops, soil management groups and systems of management. Soil tests for P, K, and pH have contributed additional refinement by providing a basis for assessing the probability that fertilizer applied to a given field will result in increased yields of a given crop. However, the opportunities for quantitative interpretation of avail- able agronomic data fall short of the requirements for quantitative decision making. Over what range of nutrient combinations can a cheaper nutrient be substituted for a dearer? What part of the cost of fertilizer nutrients may be credited to residual benefits to succeeding crops ? At what level of fertilizer input do expected net returns approach marginal equivalence to expected returns from equivalent expenditures for alternative pro- duction factors ? Deductive economic principles for maximizing returns from combinations of production factors involve manipulation of clearly delineated functional relationships. Currently available agronomic data does not lend itself to effective functional analysis. In recent years, cooperative activities of agronomists and agri- cultural economists have been directed towards the design of experiments for obtaining fertility and yield data appropriate for functional analysis. These have been concerned primarily with defining fertilizer input and cr0p yield output relationships. However, the degree of unexplained yield variation encountered has led to the attempted use of soil tests and other measured soil or climatic variables as independent variables in various formulated functions. Several difficulties have appeared as regards the use of soil test data for additional control over variance unexplained by input—output Among these are the lack of agreement among agronomists as functions . Reservations are to the scientific validity of available soil tests. particularly strong as regards tests for availability of soil nitrogen, although it is agreed that tests for P and K are far from being as informa- tive as might be desired. A further difficulty is the lack of precise theoretical concepts which might be used to deduce appropriate mathe- matical formulations for specifying functional relationships between cr0p yields, applied nutrients, and residual soil nutrients as estimated by soil tests. A principal objective of the present'study was to evaluate several chemically derived fractions of soil nitrogen in terms of their sensitivity as measures of residual nitrogen from previous treatment. A secondary objective was the application of functional analysis as a statistical tool for evaluating the significance of measured nitrogen fractions to crOp performance. These investigations must be considered preliminary in scope. However, they were motivated by the ultimate objective of provid- ing appropriate agronomic information‘for economic Optima studies leading towards efficient» fertilizer use. LITERATURE REVIEW Nature of Soil Organic Nitrogen Present knowledge about the nature of soil organic nitrogen is based on the studies of nitrogen compounds released by extraction or hydrolysis of soil by chemical agents, usually strong acids or bases. Practically all the nitrogen present in surface soil is in combination with organic com- pounds. Gortner and Morrow (29) fractionated the nitrogen present in mineral and organic soils and showed that a large part of organic nitrogen was in the form of protein and proteinaceous compounds. Hobson and Page (34) performed numerous studies on soil organic matter and concluded that the humic materials contain a complex of non— nitrogenous humic acids and protein. A smaller portion of the total nitrogen extracted from soils with cold soda was found to be in the amino form than in protein of animal or vegetable origin. From this they con— eluded that. the protein was of a different source than plant or animal protein. At about the same time Waksman and Iyer (80, 81, 82) postulated that protein existed in the soil in the form of a resistant ligno-protein complex, and this accounted for its apparent low availability to micro- organisms and plants. Kojima (40) and Bremner (10) showed that 30-40 percent of total nitrogen was in the form of amino acid in soil organic matter hydrolysates. Bremner, (it a_._l. (13), and Sowden (69), estimated that 5-10 percent was in the form of hexosamines. Numerous amino acids and hexosamines have been identified by Bremner (ll, 14), but their modes of linkage in soil have not been established. Adams (:3: a}. (1), and Anderson (5), indicated that of the order of 2 percent of the total nitrogen in surface soil occurs in the forms of adenine, cystein, thymine. The proportion in which these nitrogen bases occurred indicated that they came previously from microbial nucleic acids. Bremner (12) has estimated that not more than 10 percent c ” total soil nitrogen is present as nucleic acid Allison (2) pointed out that nitrogen in humus is very hetrogeneous nucleic acid, in nature and is believed to consist of protein, amino sugars chitin heterocyclic compounds and ligno-protein complexes Rodrigues (6 3) and Bremner (15) have shown that some of the nitrogen 1n 5011 pr ev1oisl" considered to be organically combined is in the form of ammonium trapped in the lattice of clay minerals. Allison (2) pointed out that these nitrogen compounds are so intimately bound with clays that they are largely free from biological attack. However, present evidence indicates that more than 95 percent of the total nitrogen in the surface 5011 15 organlcallv combined. It must be mineralized (converted to inorganic form) before it is available for plant. absorption. Availability of Organic Nitrogen An abundant literature exists on the availability of organlc nitrogen There is much that is ambiguous or contradictory in 131115 to plants. it is a universally accepted concept that organic literature . However, The nitrogen must be mineralized before it is available for plant uptake. process by Whlch this is carried out is called "mineralizatmn" and is the result of microbial activity. The stages in mineralization of organic nitrogen have been outlined as follows: Amm oni\fication Nitrification f-‘V/ ' MHW \O a N Ammo 1a Nitrite Nitrat rg nic —-———} _ n —) VD Available N '2‘..-» Nim N In a normal soil, nitrate is the end product of nitrogen rninerali— zation. Marchall (45) pointed out that ammonia was the first form of mineral nitrogen to appear in the break-down of nitrogenous organic materials. This is oxidized to nitrate via nitrite. Nitrate formed by ammonification and nitrification during some finite period, such as a growing season in the field or the period of an incubation in the laboratory, is referred to as "nitrifiable" nitrogen. Organisms involved in the decomposition process use organic substances as food. Part of the materials entering the metabolic processes of decomposition is used in the synthesis of microbial cell tissue and part is converted to inorganic (mineral) form. The nitrogen mineralized in excess of the needs of the microorganisms represents nitrogen "available" for higher plants. Harmsen and Van Schreven (32) have summarized the areas of major agreement in the literature dealing with mineralization of organic nitrogen in soil as follows: (a) In ordinary soil, the rate of oxidation to nitrate is greater than the formation of nitrite. Nitrite formation, in turn, is faster than the rate of ammonification. As a result, ammonium and nitrite do not accumulate in soils except under abnormal conditions . (b) In fallow soils, the mineral nitrogen content is lowest in winter, rises in Spring, is highest in summer, and goes down in autumn. (c) In cropped soils, the minimum mineral nitrogen content is accompanied by maximum plant growth, and maximum mineral nitrogen content occurs after harvest. (d) The winter minimum is due to heavy leaching in humid climates. The rise in Spring is probably due to a "partial sterilization effect" as a result of frost, giving rise to an enhanced activity of the surviving population as the soil warms up in the spring. (e) The mineral nitrogen content of soil under perennial crops remains very low at all times. . (f) The nitrogen content of organic materials below which no mineralization occurs correSponds to a C:N ratio of 20 to 25:1. Nitrate is considered the form primarily used by most crop plants. Mehlich (48) has stated that N03 ion is rapidly reduced in plants to NH3, probably in the presence of a molybdenum-containing enzyme. The N113, in combination with organic acids, forms amino acids, which are the building stones of proteins. Lyon, Buckman and Brady (44) have pointed out that young plants of almost all kinds are capable of using nitrogen in the form of ammonia, although they seem to grow better if some nitrate nitrogen is also available. They pointed out that plants such as lowland rice even prefer ammoniacal nitrogen instead of nitrate. Wallace and Mueller (78) found an average ratio of ammonia to nitrate absorption of 1. 84', ammonia absorption increased with rising pH, but nitrate absorp- tion decreased with increasing pH. Burris (20), discussing the relative effectiveness of nitrate and ammonia in plant nutrition, observed that his own experimental data indicate that plants usually assimilate ammonia more readily than nitrate. Chapman and Liebig (21) have concluded that whenever two con- ditions occur, namely, fairly high ammonium concentrations and neutral to alkaline soil conditions, more or less accumulation of nitrite in plants may be expected. Nightingale (54) has reported that nitrate is reduced rapidly to the ammonia form after absorption by plants and is then con- verted into organic nitrogen compounds in the same manner as ammonia nitrogen. Mevius and Dikussar (49) reported that sweet corn can utilize nitrogen of nitrites in either neutral or alkaline solution. The optimum concentration is approximately 50 mg of nitrite nitrogen per litre, but at a. pH of 7. O as much as 200 mg per litre is not injurious. An increase in the amount of nitrite in a medium is followed by a rapid increase in assimilation and in the total amount of nitrogen inplants. Thus the toxicity of nitrite and ammonia nitrogen is probably dependent upon the relative rates of absorption and utilization or detoxifi- cation within the plant. Grogan and Zink (30) and Tiedjens and Robbins (75) have concluded that if assimilation or detoxification of ammonia or nitrite nitrogen within the plant keeps pace with absorption so that no accumulation occurs, injury to the plant is prevented. Bremner (l9), Salter and Green (66), and Woodruff (89) have esti- mated that nitrogen is released from soil organic matter to the extent of 1 to 3 percent during each growing season. Allison (2) listed factors which affect the rate of release of nitrogen from soil organic matter. These are: (a) nature of the soil organic matter itself, (b) temperature, (c) moisture, (d) aeration, (e) reaction, (f) supply of inorganic nutrients, and (g) nature of the soil microflora. Millar. 3t a_._l. (51) in studying the effect of decomposing plant residues on accumulation of nitrate in soil, found a significant corre- lation between accumulation of nitrates in the soil and the carbon and nitrogen contents of the materials added. Materials with a wide C:N ratio depressed the accumulation of nitrates in soil to a'greater extent than materials having a narrow C:N ratio. Waksman and Hutchings (79) concluded that the source and chemical nature of organic materials added to the soil strongly influence the retention of nitrogen in organic forms in the soil. Waksman and Tenney (84) suggested that 1. 7 percent nitrogen in decomposing rye was sufficient for microbial need, and that a nitrogen content in excess of this value was rapidly mineralized. Waksman and Tenney (83), Pin'ck e} a} (59), and Norman (55) have reported that, at a critical nitrogen content of l. 2 to l. 7 percent, neither tie-up (immobilization) or release (mineralization) of nitrogen takes place. Broadbent (16), Jansson and Clark (36), and Broadbent and Bartholomew (17) have pointed out that different kinds of organic matter incorporated in the soil either alone or together with an inorganic nitro- gen source, will influence net mineralization differently. Millar _e_t _a_1. (51) reported that materials high in nitrogen tend to decompose faster at first than materials low in nitrogen. After prolonged decomposition, Turk (77) found that plant materials high in nitrogen gave a greater retention of carbon than did materials lower in nitrogen. The rate of residue addition is also a factor affecting the avail- ability of nitrogen in soil. Wright (90), Miller. 91; a_._1. (52), Lohnis St 341. (42), Allison and Sterling (3), Patrick (57), Dunn and Wheeting (22), and Pinck it a_._l. (59) have noted that the rate of tie-up or release of nitrogen from organic materials tended to decrease with increasing rate of addition or concentration of the plant residues or manures in soils. Bartholomew (7) has pointed out that where the addition rates have been either high (5 to 20 tons per acre) or low (0. 5 to 2 tons per acre) or where the range in the rate has not been wide, the influence of concen- tration of residues on nitrogen tie-up has not always been evident. However, the influence of concentration of residues on rate of decompo- 'sition and nitrogen tie-up had been most evident where low rates of nitrogen were compared to high rates. Periodic addition of fresh plant materials in the form of cr0p residues and green manures has long been accepted as fundamental to good soil management. Their influence upon the soil nitrogen supply within a very few weeks after application is often marked. Very little carryover or delayed influence has been noted in crops succeeding the initial one . The application of inorganic nitrogen to low nitrogen residues has been reported both to hasten and to retard decomposition of these materials. ' Starkey (70), Tenney and Waksman (74), McCalla (47), Jansson and Clark (36) have found that the addition of inorganic nitrogen both in and apart from soil tended to Speed up decomposition. However, Chapman and Liebig (21) found that additional nitrogen did not influence the level of carbonaceous material retained in the soil after prolonged decomposition. Since mineralization is mainly biochemical in nature, environ- mental factors which affect the number and activity of microorganisms, also'influence the processes of ammonification and nitrification. The microorganisms connected with ammonification include aerobic and anaerobic forms, but the bacteria involved in nitrification are strictly aerobic. Thus the relative amounts of soil nitrogen in the form of ammonia nitrogen and nitrate nitrogen are affected by the amount of oxygen available in soil air. Plummer (60) observed under strictly anaerobic condition that there was somewhat less ammonia produced than when oxygen was present at the beginning. Oxygen was found to be the limiting atmospheric constituent for nitrification. Fathi and Bartholomew (24) indicated that the minimum oxygen concentration for nitrification was below 0.4 but greater than 0. 2 percent, and the optimum concen- tration was about that contained in ordinary air. Soil moisture is another factor affecting the number and activities of the organisms connected with mineralization. Gainey (27) found that nitrate accumulation was directly proportional to moisture content of soils. Bollen (9) found that 60 percent moisture saturation capacity was optimum for ammonification and nitrification. However, 75 per cent of saturation capacity was found to be the optimum content for carbon dioxide evolution in prolonged respiration experiments. Fitts it 31' (25) found that 100 cm of water tension was optimum for the production 10 of nitrate under laboratory conditions. Depending on the texture of the soil, this tension resulted in 25 to 35 percent moisture. Mineralization is also greatly affected by temperature. Ensminger and Pearson (23) have summarized the effects of temperature as follows: "It is a generally accepted fact that the greatest accumulation of nitrate takes place during the summer months and the least during winter months. " Panganiban (56) found that ammonification took place between 15 and 60 degrees C, the rate increasing with rise in temperature. Nitrification tookplace at 15 and 40 degrees C. The Optimum temperature for nitrification was 35 degrees C or slightly higher. Rothwell and Frederick (64) have found that nitrification proceeds at a low but significant rate at temperatures as low as 5 degrees C. Sabey it a_._1. (65) reported that nitrification rate decreased with diminution in soil temperature. However, the relationship was not linear over the entire temperature range and was dissimilar in different soils. In dealing with environmental conditions the role of soil reaction cannot be overlooked. Allison and Sterling (3) found that lime produced a greater increase in nitrification in a low nitrogen soil than in a high nitrogen soil. In all cases, lime had a stimulating effect on mineralization, continuing for a long time. ' Halvorson andCaldwell (31) reported that the presence of large amounts of calcium carbonate inhibited nitrification. Stojanovic and Alexander (71) observed the inhibition of Nitrobactor by free NH, at alkaline pH. They found that concentrations of ammonia greater than 250 mg/ml reduced nitrate formation proportionally to the concentration of NH‘. Oxidation of ammonium to nitrite was unaffected. Harmsen and Van Schreven (32) stated that in most normal soils nitrite seldom accumulates to a measurable level, but the formation of nitrite is enhanced by a high pH level to a greater degree than the transformation of nitrite to nitrate. 11 Anadequate supply of nutrients, notably phosphorus, is also needed for rapid decomposition. Kaila (38) noted that O. 2 percent phosphorus was an average level below which decomposition was retarded and immobilization of mineral phosphates occurred. Ensminger and Pearson (23) have concluded that soil treatment such as the addition of limestone, phosphorus, and potassium affects the production of nitrates in the field, as do tillage operations such as plowing, cultivation, fallow- ing, and mulching. McCalla (47) found more nitrate nitrogen liberated from sweetclover residues which were plowed under than from the same residues left on the surface of the soil as mulch. The presence in the soil of roots of growing crops and the sequence of crops in a rotation have been observed to affect mineralization of nitrogen. Goring and Clark (28) concluded that less net mineralization of nitrogen occurred in cropped soils than in fallow soils. They believed that nitrogen unaccounted for in cropped soil was immobilized in the soil rather than lost to the air. Brown (19) observed that the rotation of craps resulted in greater numbers of soil organisms as well as greater ammonifying and nitrifying powers in soil than continuous cropping to corn or clover. Brown suggested that carbonaceous matter exuded from roots favours development of nitrate consuming organisms in soil with a consequent transformation of nitrates to insoluble organic forms. Lyons gt a_._l. (43) presented data in support of the postulation that certain plants differ in their ability to take up nitrogen from the soil because of characteristic differences in the amount and composition of the organic compounds liberated by their roots. Procedures for Determining Nitrogen Availability Several types of procedures for determining nitrogen availability in soils have been advocated. Allison (2) has classified them. into four general categories: 12 l. Vegetative tests 2. Nitrification tests 3. Soil nitrogen released by chemical reagents 4. Determination of total nitrogen either directly or by measuring total organic matter. Vegetative tests: Field plot experiments or greenhouse experiments are conducted and nitrogen uptake by crops or test plants are determined. Different rates of nitrogen addition are needed to provide a basis for the interpretation of values obtained. Allison (2) points out that because of time and expense involved the vegetative test can only be used to a limited extent. Lyon (it al. (43) and Bartholomew (7) have observed that this test does not give a measure of total available nitrogen as it is normally influenced by the previous crOp residues. Nitrification tests: Release of nitrate during incubation has been widely used as a measure of availability of nitrogen in soils. It has been based on incubation techniques of which the Iowa test is representative. Fitts El: 11. (25) demonstrated that nitrate produced during incubation under standardized conditions provided a basis for predicting the nitrogen requirement of corn under Iowa conditions. They obtained a highly significant negative correlation between nitrifiable nitrogen and yield response of corn to nitrogen fertilization. Harmsen and Van Schreven (32), in reviewing methods for estimating the nitrogen mineralization capacity of soils, pointed out that the determination of the momentary amount of mineral nitrogen (ammonium or nitrate) in the soil has a very dubious value. They expressed the concensus of numerous investigators that the results from incubated samples are in no way comparable to the mineralization process under field conditions. Under field condition variable factors related to crop, management practices, and climate are involved, whereas incubation experiments provide information on 13 the short term potential of the soil for mineralizing nitrogen. Saunder e_t a_.l. (67) found that the nitrogen mineralized in laboratory incubated soil, sampled towards the end of the dry season, gave a good index of the nitrogen likely to be available for crops grown under field conditions during the subsequent growing season. It was necessary to determine the time when nitrification began for reliable estimates of minerali- zation or nitrification rates. ' Allison (2), discussing the incubation tests, has suggested three factors to be considered: (1) A standardized system of treatment of soil samples following removal from the field, (2) optimum incubation conditions with respect to moisture and aeration, and (3) the increase of nitrification rate by addition of lime. Chemical extraction of nitrogen: A third type of procedure involves the chemical extraction of a fraction of the soil nitrogen. Truog (76) proposed a procedure involving partial oxidation of soil organic materials with alkaline permanganate. It was assumed that the permanganate attacked the readily oxidizable portion of the soil organic matter. Nitrogen is released as ammonia and measured together with exchange- able ammonia. Kresge and Merkle (41) investigated the alkaline per- manganate distillation procedure in laboratory and field studies. Their studies showed that a good correlation existed between this determination and incubation nitrification rates. The amounts of nitrogen released by nitrification or by alkaline permanganate oxidation were useful in evalu- ating soils and soil management practices. Peterson <_-2_t a}. (58) found highly significant correlation coefficients between total soil nitrogen and each of the following: . Alkaline permanganate nitrogen (O. 95), percentage organic matter (0. 99), nitrification rate (0. 65) and the amount of ammonia nitrogen extractedby various concentrations of sulfuric and hydrochloric acids (0. 59' to 0. 71). However, they found that the tests 14 were not significantly correlated with nitrogen uptake by a second crop. Kamerman and Klintworth (39) made chemical determinations on soil for total and hydrolyzable nitrogen and for total and now-hydrolyzable carbon. Nonhydrolyzable nitrogen and hydrolyzable carbon were calcu- lated by difference. They observed that nitrifiable nitrogen was inversely related to the ratio of hydrolyzable carbon/hydrolyzable nitrogen and also inversely related to the ratio of total nitrogen/hydrolyzable nitrogen. ' Estimation of total nitrogen: The fourth type of procedure involves the direct or indirect measurement of total nitrogen. Gainey (27) observed a very close and direct relationship between nitrogen content of soils and their nitrate accumulating ability. He obtained a correlation of 0. 990 :1: 0. 012 fer a "non-fertile" series of soils and 0. 988 :t 0. 0006 for a "fertile" series. Allison 31: a_._l. (3) showed that a positive corre- lation existed between total soil nitrogen and nitrate formed at all incubation periods in limed and unlimed soils. Woodruff (89) estimated the rate of nitrogen delivery to crops from a chemical determination of soil organic matter. If all organic matter were alike, he concluded that the liberation of nitrogen from organic matter in a form available to plants would be proportional to the amount of organic matter present in the soil. Bartholomew (7) has expressed the relationship in mathematical terms as follows: dx E;- = kx Where x = the organic nitrogen content k = a constant for a. particular soil under specific management dx ‘ . . . w = the rate of liberation of nitrogen. If "t" is in yearly unit, k is a measure of the annual net mineralization of nitrogen. 15 In criticism of this concept, Bartholomew pointed out that the fraction of soil organic matter that decomposes to liberate nitrogen for a particular crop depends upon other factors also such as cropping pattern, texture, structure and moisture condition during the growing period of the crop. Woodruff (89) found that the delivery of nitrogen from soil to crop was a function of crOps. He showed that on the average the annual rate of delivery of nitrogen was 2 percent for corn, 1 percent for small grain, and 3/4 percent for a crOp of leguminous nature such as soybeans. Smith (68) observed that the amount of nitrogen that a soil will deliver will depend very much on texture. A clay or clay loam will release 1 1/4 to 2 1/2 percent of its total nitrogen in one season, a silt loam 1 1/2 to 3 percent and a sand or a sandy loam 4 to 6 percent. He also emphasized the importance of temperature and moisture con— ditions during the growing season, as well as the nature of any organic matter recently turned under. Thus, it is necessary to know how much nitrogen is released from the organic matter in a given soil under specific environmental conditions. Allison (2), discussing the merits and limitations of the various procedures, has stressed the need for further research to evolve a method that is simple, rapid and inexpensive. MATERIALS AND METHODS Field Treatments and CroppiggHistories Soil samples for this study were collected from five established field experiments.* Three of these were fertilizer experiments where widely divergent levels of N, P and K had been applied for one crop prior to the taking of soil samples. These experiments provided an opportunity for measuring the accumulation of nitrogen during a single cropping season in various fractions following applications of nitrogen ranging from 20 to 320 pounds per acre. In a fourth experiment, various residues had been applied with and without supplemental nitrogen at the beginning of a five-year rotation;.and soil samples were taken at the end of the first cycle of the rotation. A fifth experiment involved widely divergent cropping systems, fertility levels and supplemental nitrogen treatments imposed for four cycles of a five year rotation prior to sampling. ' Fertilizer experiments: A group of three extensive field experiments were established in 1955 to provide data for economic optima studies on fertilizer usage (Sundquist and Robertson). Numerous levels and combinations of N, P and K were employed with minimal replication, the original objective being to establish response surfaces rather than discrete incremental response points. From each of these experiments soil samples for the present study were taken from selected treatments covering the full range of nitrogen. treatments at each of several combinations of P and K. This was done to compensate for the fact that only two field replications Locations of field experiments and soil types are described in Appendix Ill. 16 17 were available for each treatment. Assuming norinteraction between N levels and P K combinations in their effects on soil nitrogen levels, the P K combinations would provide additional replication for evaluating the effects of applied nitrogen, levels on soil nitrogen. The Fick farmr and the Campbell farm experiments were both located on Kalamazoo sandy loam. - At the Fick farm, the first appli- cations of fertilizer were made in 1955 for corn. Soil samples were taken in the spring of 1956, just before the same fertilizer treatments were repeated on the same plots for oats. At the Campbell farm, the fertilizer treatments were applied for oats in the Spring of 1955 and again for winter wheat in September, 1955. Soil samples were taken just prior to fertilizing for wheat. The experiment at the Thompson farm was located on Sims loam. Here the fertilizer treatments were first applied for corn in 1955 and again in the spring of 1956 for white pea beans. Soil samples for analysis were taken during the winter of 1955-56. The actual levels of N, P, and K used in the treatments selected for this study are shown in Tables 1 to 6.* Ammonium nitrate, super- phosphate and muriate of potash were the) carriers used. All of the fertilizer was broadcast before plowing, withthe exception of 40 pounds P20, which was applied as a starter fertilizer at planting time on all plots which received superphosphate. Residue experiment: Beginningin 1951, an exPeriment was set up at the Ferden farm to determine the effects on crop yields of the addition of large amounts of sawdust in comparison with more normal quantities and types of residues. The objective was to determine if alfalfa-brome that is removed from the rotation as hay could be replaced by sawdust or straw to >1: Pp. 24—29. 18 maintain yields and soil building qualities in the rotation. The present study deals with the comparison of residual effects of residue treatment and supplemental nitrogen on mineral and organic forms of nitrogen and on yields of corn in plots establised in 1953. The crapping sequence was corn, beans, barley, followed by two years of alfalfa-brome. Four residue treatments were employed: 1. Two cuttings of alfalfa-brome hay removed each year. 2. Same as treatment one, except that neither cutting of the second year of alfalfa-brome was removed. This treatment is repeated each cycle of the rotation, the second repetition of the treatment having been made during the summer of 1958, prior to the taking of soil samples in September. 3. Same as treatment one, except that 35 tons of sawdust per acre was applied after the removal of the second cutting of hay on the second year of alfalfa-brome, at the beginning of the experiment only. This application was made in the fall of 1953, five years prior to the taking of soil samples for this investigation. 4. Same as treatment one, except that 3 to 4 tons of wheat straw was applied after the removal of the second cutting of hay during the second year of alfalfa—brome. This treatment is repeated each cycle of the rotation. The second application of straw was made in the fall of 1958 at about the time soil samples were taken. Fertilizer has been applied at the rate of 100 pounds per acre of 5-20-10 for corn, 2.00 pounds 0-20-10 for, beans and 240 pounds 5-20-10 for barley. No fertilizer has been applied during either hay year. Supplemental nitrogen has been applied on one half of each residue plot at the rate of 40 pounds per acre of nitrogen for corn, beans, and barley. 19 All treatments were replicated five times and composite soil samples were collected from each replicate in September, 1958. ' Chemical determinations on these soil samples were compared with yields of corn on the same plots in 1959. Rotation experiment: A series of rotation experiments was initiated in 1941 by the Soil Science Department of the Michigan Agricultural Experiment Station on the Ferden Farm in Saginaw County. These rotations were established to study the effects of cropping sequence, fertility level and supplemental nitrogen on cr0p yields and soil prOperties. For the present study, two rotations were selected which represented extremes of crop response to supplemental nitrogen. Rotation No. 1 was a livestock rotation with corn, sugar beets, barley and two years of alfalfa-brome. Ten tons of livestock manure has been applied for corn during each of four completed cycles of the rotation. No significant response to supplemental nitrogen has been obtained with any of the crops in this rotation. Rotation No. 6 was a cash crop rotation comprised of corn, sugar beets, barley, beans and wheat. Yields of corn on this rotation have averaged less than half those on Rotation l, and there have been consistently large increases in yields of corn, barley and wheat forapplied supple- mental nitrogen (61, 62). Two levels of fertilization have been maintained on each rotation. The low fertilizer rate from 1940 to 1950 was 400 pounds per acre 2-16v8 applied over the five-year rotationperiod. In 1951, it was raised to 800 pounds 4-16-8. The high fertilizer rate from 1940 to 1950 was 800 pounds 2-16-8, in' 1951 it was raised to 1600 pounds 4-16-8. Half of the five-year rate was applied tossugar'beets and half to the rest of the crops in the rotation other than hay. 20- Supplemental nitrogen was applied on one half of each fertilizer level plot on corn, beets and small grains, at the rate of 40 pounds per acre. . The soil is classified as Sims clay loam. Soil samples were taken from each of the four replicate plots of each treatment at the time of planting corn in May 1959 and were analyzed in the laboratory. The laboratory data were correlated with corn yields taken in the fall of 1959. Laboratory Determinations Analytical procedures which were employed in these experiments involved the determination of exchangeable ammonium, nitrate, and hydrolyzable and nonhydrolyzable forms of nitrogen on the same air dry soil sample. Exchanjeable ammonium: Twenty-five gram air-dry soil samples screened through a 40- mesh sieve were shaken for 30 minutes in 100 ml of N-KzSO4 in N/lO H2504 (13). The samples were filtered with suction and washed twice with distilled water. The filtrate and washings were transferred quantitatively into 800 ml Kjeldahl flasks. Eighty ml of 40 percent NaOH was added to each flask to‘make the filtrate alkaline, and the ammonia nitrogen was distilled into 25 ml of 4 percent boric acid containing bromocresel green- ‘methyl red as indicator. The distillates were titrated with'O. 025 N HCl. Nitrate: The residues remaining in the distillation flasks after the determin- ation of exchangeable ammonia were diluted until the volume was 300 ml. One tea3poon of Devarda's alloy (Cu, 50 percent,‘ Al, 45 percent and Zn, 5 percent) was added to bring about the reduction of the nitrate. 21 The flasks were then immediately connected to the distillation apparatus and distillation was continued into 25 ml of 4 percent boric acid contain- ing bromocresol green—methyl red as indicator. The distillate was titrated with 0. 025 HCl. Hydrolyzable and nonhydrolyzable nitrogen: Hydrolysis of extracted soil: The extracted soil left on the Buchner funnel was transferred with the filter paper into the original 500 ml Erlemneyer flask. Eighty ml of 80 percent H2504 (this amount contains 35 ml of concentrated H3504) was added to the soil and filter paper in the flask. The flask was shaken at intervals for two hours at room temperature. The volume was made to 350 ml by adding distilled water. The flask was then stoppered with a one-hole rubber stopper fitted with a Bunsen valve and was autoclaved at 15' lbs. pressure for four hours. The flask was allowed to cool to room temperature. The supernatant was decanted through a Coors No. '0' Buchner funnel and washed twice with distilled water. The filtrate and washings were transferred quantitatively into an 800 ml Kjeldahl flask for the determination of hydrolyzable nitrogen. The residue after hydrolysis was transferred into a second 800 m1 Kjeldahl flask for the determination of nonhydrolyzable nitrogen. Hydrolyzable nitrogen: To the first Kjeldahl flask which contained 350 mlof hydrolysate (35 ml of concentrated H3504), were added a few glass beads, 19. 8 g of K1804, 0.82 g of HgO and 0.16 g of CuSOg. The flask was then placed on the digestion rack. Surplus water was driVen off by slow heating. Heating was increased gradually to maintain a temperature above 360 degrees C but below 410 degrees C. - The heating was continued for onehour after the solution cleared. After digestion the flask was allowed to cool to the point where crystals started to form. Then 300 ml of distilled water was added cautiously with constant swirling, after 22 which 150 m1 of 40 percent NaOH plus '24 percent sodium thiosulphate solution was added down the side of the flask without mixing. The flask was connected to the distillation apparatus and swirled. Ammonia was then distilled into 50 m1 of 4 percent boric acid with bromocresol green- methyl red as indicator. The distillate was titrated with O. 1 N HCl. Nonhydrolyzable nitrogen: To the Kjeldahl flask containing the residue after hydrolysis were added 35 ml of concentrated H2804, l9. 8 g of K3504, 0. 82 g of HgO and 0.16 g of CuSO4. The rest of the procedure for digestion was the same as described for hydrolyzable nitrogen. After digestion was complete the flask was cooled. Ammonia was taken up in four 100 ml aliquots of distilled water and transferred quantitatively by decanting into a second Kjeldahl flask. Distillation and titration of ammonia were the same as for hydrolyzable nitrogen. Calculations: The amount of exchangeable ammonium, nitrate, and hydrolyzable and nonhydrolyzable nitrogen was expressed in pounds per acre according to the following expression: (T—B) x Nx0.014 x 2 x106 5 Where: T 2 m1 standard acid in sample titration B (ml standard acid in blank titration N = normality of acid S = weight of soil sample in grams. EXPERIMENTAL RESULTS Fertilizer Experiments The three nutrient level experiments all involved differential rates of N, P205 and K20 applied to one crop preceding sampling. Thus any differences in measured nitrogen fractions would represent one year's residual accumulation from applied fertilizer. The individual plot values for inorganic and organic forms of nitrogen, together with crop yields, arepresented in Tables 13 to 21 in the Appendix. Treatment means are given in Tables 1 to 6. Residual effects of fertilizer on ammonium and nitrate levels: On Sims clay at the Thompsontfarm (Tables 5 and 6) there were no significant differences in NH3-N or NO3-N following application of nitrogen from zero to 320 pounds per acre. There were no apparent trends in quantities of either form of nitrogen recovered in their relation to treat~ ment. On Kalamazoo sandy loam at two locations (Tables 1 to 4), both NH3-N and NO3-N increased with increasing levels of nitrogen applied one year prior to sampling. On the Fick farm NH3-N was higher at the three higher levels of P205 and KzO application than where X-40-80 was used. The values for NHa-N encountered were 5 to 10 times greater than levels normally found infield fresh soils. It would appear that much of this nitrogen was released from organic combination during air drying. These soils were stored in air dry condition for three to-four years. The levels of NH3-N were considerably higher in soils from the Fick and Thompson farms (Tables 1,. 2, 5, and 6) where the preceding crop had been corn than on the Campbell farm (Tables 3 and 4) where the preceding cr0p was oats. This suggests that the larger quantities of root residues 23 .. .a..-- -. -.........s.\ wsuneuuvmvn Hy M‘CW.$AJ-_~4~ Euwdvxu .VAv T.\.~.J~‘A Imc: HNU nihHoH.AJyu. Haj—.vmrrvlfl nuav VIN ”humid RUN. sz qhflv Huaflddflv.ul.fl~aflfildlfi Wang‘s-ii “inked ,U‘sv It.“ :vfwfii‘!‘ IQ ,Illi. II (II-l. C's... 24 .mumo HOV muouflfluom met»??? 0» uoflmsoxg one? usages flow .mumo new 0mm: mo wfiuam 5 Gamma new snob Ham mmwa mo magnum cw poflmmm 30530.3 souflmuuoh .. * wz mz wz wz m .3 m2 a. .n. soma— ~ .2. on: mom .22 3 mm 2. 3.903.- x N .3 $2 own «no 3 3 me om .23.. x o .3 $2 mam 33 mm mm om owiofi- x m .3 32 mam m2. No om em om - ow .. x Mmduom new oumaamoanm mo mcoflmcwnaou .m2 m2 m2 2: a .w mz a. .0 some o 2. 23 see 32 mm mm 3 x3 x .. 3m «.3 3.2 gm 2o 3 an cm x .. x - ow 4 am $2 3N ems so am . 3. x .. on. em mused somoufiz o .3 m2 m2 m2 mz m2 m2 some ~33 .82 sum was 2 on me x - a- z m mm 8.: Sm omw X an me o .. o .. o l MWNSSHPM m> usefiflnorm 02 $25 «>3 «)3 , «>34 ‘ «<34 «>54 $24 o~m-.o...m-z . Z .muO. 2 03.3: 2 33m Z assoc“; ZanOZ Zanmz 950 1308 nonpafioz nucfionptwm H.308 *udogmonh .omummmH .EMOH apnea ooudEstVm .mufiogumuwuu nouwfinuuom ogdm 33 mo Gowumoflmam 333m pcooom .m mnmgoHHoH sumo mo mp3?» don mo saucy Heapwmou so M was am .2 mo sofldoflmmd 3553 0:0 mo museum :1 .H 0.232. .653 sofa pad ammonia Ia Lil lllu.‘ snsf- Huh-{KI-firvwmflpwv'fi at IIIII din-fish :rbhuhviflfifi: Hwfiom vflfiv hunk-[NOV fidmvwthWerVlN dun...» V'N uyihnLdW n” .2 .VRJ mwpvuuflfiolm‘ll 9“! .v~e~.-nw.N. 25 .Mm @388. + M2 8888-. .. mz $888. + NM 5888. - M 2328 . + b .3388. + a 3338 . - N2 3828. - 2 2.323.... 23$ .2. n or aosofie asafioiaoa < M $288. + M m2 @388. .. a 2388. + a m2 823.3. 1.. z 3338. + 2 m2 3.82:: . + $333.4 u om. mod ”corona surgeons”... nofludposa 0.25 503 pofiduno ago?» “M was nm .2 mo munoaouonw vogue was was?» pm pouofluohm one mpuow> po>hom£o Scum mnflombadm tan— poflwmufio >uom£o on”. o» @033” mnoflocdw - - mz .02 m2 m2 m .2 m .s a .2 893 o .o e .m m .3 SE 3... 8.3 ms 3 S causes- Ba 5 .m a .o o as 3.: SN 9:: 2. an em. 3923... on aim- a .9. o .3 32 emu 32 am 3 em 3a.. 8.... 3 he in 9.2. 32 mom m5 mm on mm as .22... can ed 3 of. £2 own 32 em 8 mm as .23. cm H m a .T 9.3. 82 . mam mm... 3 MN 3. cm .22.. o... v.9 «to ode 3.: ms. 82 as as 3 ‘ as $2.. 3.... mum- 93- m1? 3.: SN 3m 3 s... 2. cm $2. on m m; a.2 92. :2 Sn o2. mm .2. 2. cm -02.. cm a... To - - .. - .. .. .. cm -313... mus 5.... 0.3 32 mum o2. S 3 mm om .. oe .. on m o m.m - 9mm 22 8... cos 3 mm an em - 2. - on how News: $25 «>54 «\an $23 «<23 $23 «>3 onM- “one... 2 .w - .w 3.5 23» z cammuo z ”33.: z 03... z ~8er ztoz 2..an Advasdpmmom persona—O “.308 noupcgsoz . 1530.53.“ H.308 uaogdoufi "n . Illlll 111%“ .em-mmsa ”.on sense cones—fig .858 Meg .359; 3333 usouusu £3 323683 sinuosusa soc mmmswamwh pH . use pad .930 mo spas?» persona—o demon»? 30m .«0 m§0w guano.” no M was nm .2 mo muuowwmnu N 3 -Nfl-(fundn 7. EC .Lthqfluikunhui! HdwflunNnhdw ano .Efiv [MayflqunhNIl om: flu NJ He r5 26 H03 335.. . .323, Sm 33:33“ Mo 8333mm." 3 Sim :oMMwMuoh .. ... /! Rama?» you” mmm; mo Saw GM memm mad memo Ham mmmL mo mflfihmm Gm domammd Hm .H. illl... . . .m m MM m2 m2 m2 M .8. m2 «.2 Wow“ M mncm 3: v.3 Kw wéo odm mdm owmuomT M ~33 No: 3.“. MS ~13 9% ~15 on -o T M . ~.m 3.: omm mom ~13. ogm ~.o~ om S .. M m 8: N3 mg m :3. odm méa om .. o... Mmlmuom wad mudfimmogm mo mnowuchn—EOU Mum m2 m2 m2 m2 m2 m2 . . .mcnmq- «m ~.S 5.: m3 2% 9.3 ogm 98 M- M-om .3 2: mm... 03 0.3 0.3 Tom M .. M. -o e S. 8: 9% mg 93 0.3 can M... M oN H9904 ammonuwz. . m .2 m2 m2 mz mz N .2 m2 .. .334 m .3 2: 0.3. mg a .3 m .3 «an M .. m. z on 33 8m 9.; m.m~ o3 m4: o .. o - o uuuflflhmh m> umuflfiumh OZ «<25 $34 $34 $34 $23 $34 $23 . ONM-...o~m.z z 3530 z @333 2 03.. z 122:2 z-.oz zkmz “NO; HNHOH. OHU>€OZ IN>HOH~U$TM HdH—OH. *uqvgflmhrfi .cmammmH .53 35am ooudgfimfi . .muaowufifi nonfifluow 086m mfi mo coflmoflmmd 3:58 vacuum m $53023 “awn? mo 30?» $0» mo mauow 353mm.” so M can am .2 mo Godudufimmd 1.553 23 no 300mmfi..- .m 03.3. .553 zonasu wad .aomouumd 1.1. I ‘llTW-uua.‘ll. EC VM Cid“. AH .2 -w 3 m. .. Etumwwumfi l I I WV .s. ~£ 3.? 27 M 3888. + M‘moewofifioo. + am 3:88. 3 + 2 $2.88. + 2 m2 333.8. + 233va u ,w moq - oooo . - mz 3288 + NM Nflmoooo. + M 3338. + ”M Unanmnmomvmbwoao . + NZ wwommooo . .. Zmovwomwo. + mnemwwmm .wm u 31% a 0 50¢ 35350 page; 0 07” w mpamgmnucw vmwvm vcm mgmfw Um>nmmno 23 on Umfifl mach m .5. view? .0 .. U u. .vmum 053 :3an Um>umm£o Scum wcwuomhundm >3 Umcwmuno mama—v.3 M V g cm 25m . gumH. :oflmum “Gwemhoaxmm Haudfinuflm/w GmmEowz .fiogumnom find an? w .. m m2 £338. a3 Eamoaomxm MA 2388. + M2 in; Mm 2388. 3030 ”noflona Hmflaosfiom :3 530363 oi .amS .mm .m m Eouh an. {If o .o m .7 .w2 m2 m2 m2 93 ~13. «$3 . 2mm”... 03 ed M; m.mm 52 SN 03 cam o.- o mm 3m. 2... cm 10 o.o mm: $2 mam omm m4; can mien owm o .. om To- To- cam 22 SN 02.. m4: oi. 0.3 ofiémw- SN o.” A: 0.3 8: SN mom :3 can 9% cm -23.. on To 0.7 o.$ m2: o3 mmm 93 oé~ man cm .52... N mic m5 own 3: «3. OS of. 9mm 9% ow $3. c To m; o.$ 3: o3. So 93 o .2. ed... om -03.. SN TN- ML- my... 3.: 3... mm. 93 0.2 m3 om -ofl- om m4- TN- 0.8 32 m: g mi o.m 0.3 cm .37 3 M .o- To own mp: 93 on. oz: 93 92 om .. ow .. 3” ,N .N- Tm- o.mm 33 SA. mom 93. 0.: o.m~ om .. ow .. om. If m S .w 2.2 now E 9.2. of mdm om .. 3 .. 3 .b o . . . . PM” WWW 33m $23 «>34 «>23 «>23 334 «<24 < 52:» .23» 2 3.830 2 @333 z 35 z 182:2 2-62 . Zamz AmwaBmoM 3238 :35. -ougfioz -538me :38. Eugene; .l'llr I'll. E. ooumnflm .m . . .omnmmS .EmoH__>vamm “mm H M and“ Svnhfimo .325“ 3073.2“ 30.735 fig nmudm00mmm radaoflogw uoc 3.2.630.“ 33> £3 van udmgk’ mo mafia?» Uv>ammno demonic 20m «0 mahow 353m?” no M cad nm .2 mo muuowmmau .w 3an Iqull nutr- Hfs fluff-III?! pahu-anIuiL. Cufiu Vi uvnhdfl. 1 .2 HO F~Ofludwwvw~fiuadk udunuauflnflw .Jaho 1N3 39.4..-N‘k3nlr. .9. my ~Q 4. r5 28 .cmummmL mo umudds msu wags w a a . . A . M mdmvn «an 333 new ~3de cam Guoo new mmmL mo 93.25 am v03 w G313 who? magma—Ham 20m and 30830..“ .0 u Honflmukoh .. * \I\\ . . . .8 o MM m2 m2 m2 m2 m2 m2 -me- M 53. cm; 32 mmom Z mm 3 92-3 .. M mg: 5.3. 83 2.3. «m 3 mm 0%. o¢ .. M 3: 32 3mm 8 mm 3 cm o gmduom. r93 oumammoam mo maofimflnaoo wuw m2 m2 m2 m2 m2. m2 , 2.334.. Sn o. m $3 82 3mm .3 «m 2. M .. M- of m .NN 3:. 32 3S. cm 3 3 M - M .. N M 23 5.: $3 mm S. mm M .. M o 3564 Gomonfiz » “w m2 m2 m2 m2 m2 mz . . .384 “.3. $3, 82 .35 $ mm mm M - m .. z 2 08m 3: mowm 3 MN 3 o .. o .. o Houfimtwh m> uonfiflumh OZ «\3m «\34 . $34 $25 «\34 . $35 «>34 O~M .. m09w .. z madam z .30. z 0333 z 2% z :233 2.62 2-62 H.308 nonmrwnaoz $30.33* H.309 *unogdonh. u "Hug “I4 DWIIIIIIIIIIII‘IICIIIIIIIIIIIIIII .om-mm2 .EmoM 35m . .848“ EEthOAH. .muavanfia umuauhmm 093m 05 mo aofldufimmm Amanda vacuom .m $33023 2309 mo ago?» and Gmwoufia gem mo mayo“ Madame.“ Go M 93 nm .2 mo newudoflmmd 3558. .38 mo muomwfimna .m 3nt 29 , AZ 3888. + nm 3288. .. m #353. + "Z 8888. .. 2 $338 . + 3138.: n a» 803834388304 m 2388. + m m3 2338 . + 2 $2.88 . + 2 m3 #338. + 322.3 .4 n a? wok 28383 4333324 HM mug.“ m .2 Ho WHGUEOHUflHM vmuvd MUG.“ mfiudmg vU>H0mfio @ . 43 0» mu”: 4833603 03» 30.3 465.340 2401 63345.3 23 2:0?» v3.3 mno 89: mcflumflndm >4 vmawunmmmfimwflwa 52.: .3. d J: 34.3. .03 fidm .488 :3m .4984 .nw< 533042 .comfionom van “.maqudm 5.0.4..” A3 .. .. p .: wz m2 m2 m2 m2 m2 . . .334 N .m- 97 0 .mm 3.3 82 02m 3 mm 3 on»... 2%. can m6 N .o- - .. - - - - - own. 03- of 0.? mg... 0.2 on; Sfi 8g 3 wm 3. 3.0.33- on o5- N 4.. 0.5. 2.2. 3.2 mmmm mm 2.. 3 3m- 3 - 85 ca N .N o.m~ 83 $3 $3 on mm mm 03. 2. - o3 9? m .o 98. $3. 8.3 3.3 mm mm mm own. 3. - om m4. m4. can 23 82 85 mm mm S cm .. o4. - 8m 3. mg- o 3. 03m 3.: £3 No on 3 om - 3. .. 34 m6. To o .3 3mm 2.: 33. em on S cm. - 3. - om dam .33 «>5 «\34 .134 «\24 «>54 $34 «\34 of- .63 - z 49 r or 38mm 32..» z 3530 z 033.: z mam 2 3.3.52 2:02 77an . ‘33:“:va vm>ummn0 H.309 nonviaqoz ourouwknm 4.30M. unogmmvuh. .om-mm94 .EHOH mg -4543 GOmAEOFH .3449: 94334.5 ”50.4.4440 £44? @3300me kflmnoflugm you 3343me 30.?» den. van .9309 mo “.64va Uo>ummno .nmmofizn 30m mo anyhow ddfiwwmmu no K van nm .2 mo muummflwfu .o .3nt 30 from corn may have been the source of NH3—N released during air drying. The quantities of residues from oats would have been less, due to the nature of crop itself and to the fact that a longer period of time inter- vened between harvesting of crop and sampling of soil. By. analysis of variance, these differences in inorganic forms of nitrogen attained only a low order of significance, since only two repli- cations were employed. More effective use of the number of observations available might have been possible through use of regression analysis, and some of the observed trends might have been shown to be significant. Regression analysis was not employed, however. Residual effects on organic forms of nitrogen: No significant differences in the organic forms of nitrogen were obtained in the three fertilizer experiments. This again may have been due toinadequate replication for analysis of variance. There was a consistent tendency at all three locations for total organic nitrogen to increase with increasing level of nitrogen application. Although there was a tendency for hydrolyzable and nonhydrolyzable nitrogen to vary directly with total nitrogen, this was not consistently true. Relation of treatment and nitrogn fractions tolield: The '1956 cr0p yields from these experiments were subjected to functional analysis by Sundquist (72, 73). He formulated an exponential production function involving inputs of N,‘ P and K which accounted for 37 percent of the variation in mean yield of beans at the Sims loam location. A five-variable polynomial function fitted to the same data accounted for 42 percent of the variability. These two functions are presented at the bottom of the Tables. 2 and 6. In the last two columns of these tables the differences between predicted and observed yields (residuals) are shown. 31 Graphical analysis failed to show any relationship between these residual unaccounted-for variations in yield and any of the nitrogen fractions measured, individually or in various combinations. In the case of oat yields on Kalamazoo sandy loam, Sundquist (73) was able to account for 48 percent of the variation with a nineeva‘riable polynomial fitted to the data and 58 percent by an exponential function. These functions and the unexplained residuals are presented in Table 2.. Similar polynomial and exponential functions fitted to wheat yields on Kalamazoo sandy loam accounted for 44 and 43 percent of the variance, respectively. These functions and corre8ponding residuals are given in Table 4. Again no relationship could be established between the residuals and any of the forms of soil nitrogen measured. In all three experiments the factors of major significance to yield in both types of production function was the input of nitrogen. As noted above, there was a tendency for the various forms of nitrogen to increase with nitrogen applied. To the extent that this occurred, the current year's input of nitrogen and the residual effects of the previous year‘s equivalent inputs were confounded. The failure to find any relationship between the forms of nitrogen measured and yield residuals not explained by nutrient input functions was probably due, in part, to relatively large errors in sampling and chemical determination, and, in part, to the inappropriate- ness of the experimental design. Organic Residue Experiment In this experiment soil samples were taken at the end of the first cycle of a five-year rotation in which various types of residue had been applied with and without supplemental nitrogen. There were five repli- cations of each treatment in this experiment. Individual plot determin- ations for nitrogen and corn yields are given in Tables 22 to 28 in the 32 Appendix. Mean values for residues, supplemental nitrogen and individual treatments are given in Table 7. Inorganic forms of nitrogen: Ammonium levels found in these soils were high. Much of this ammonia was probably released from organic combination during air drying and storing. The highest amount of ammonium was found where the last two cuttings of alfalfa-brome hay were left as the residue treat- ment and where supplemental nitrogen had been used. These two treat- ments also gave rise to significantly higher levels of nitrate nitrogen. Total inorganic nitrogen (ammonia plus nitrate) was, of course, signifi- cantly greater with the alfalfa-brome treatment and where supplemental nitrogen was used. Organic forms of nitrogen: All residue treatments tended to increase both hydrolyzable and nonhydrolyzable nitrogen over the check. The increase in hydrolyzable nitrogen was significant for the sawdust treatment and the increase in non- hydrolyzable nitrogen was significant for alfalfa-brome. Total organic nitrogen was significantly higher than the check for both of these treat- ments. As an average for all residue treatments the addition of supple- mental nitrogen had resulted in a 470-pound increase in total organic nitrogen. This is of the order of 3 to 4 times the total supplemental nitrogen applied during the first three years of the rotation (40 pounds each on corn, beans and barley). It is also 3 times greater than the expected variation between treatments (LSDOS = 165 pounds per acre). Sixty-three percent of this increase in total organic nitrogen was due to a 300-pound increase in nonhydrolyzable nitrogen. ,hofi.\c1 unhhdmaavd \Afiwuhv .rfinnnflvfin udHuiHflwuu Fume-UH wvrufi -~.-A.v.,-u NC nth.~Ht-.~\A ~.,v-.~mv -uvv4..«.v-d.~.1n~ NA.» .n..#--~.\v.fi 33 N .e SN of mm: ms. . l . v m . n v m: 2.2. N2: 33 m3 3 o Md onwd m .9: $3 N3 .8; m3 mm 2. z + Babe w .m: 33. a2: 33 t: we 32...” v .m: so 2 + 35953 acne mas sown mm“ Hm a .m: 33 3.: 3mm 5 seaward w 4;“ N wma on. 2: Z + MEOHQumfidmz . s? :3 33 MS 3 so oEoS-§&2 Wm: :3 2.2 8?. am: am 2: z + v3.28 2 33 5 So... 2; mm 8 v3.25 waoufiZ m> @53me mmnm. fig aim» TE m.m as o.m 393 N .e: 33. SS 3m... t: 8 2: z Hmsqofioaesm o a: 33 own 33 S; cm 8 2 oz Ho>m1~ ammouuwz .mz mi: 15 52: Ms. A}. as sows a .o: 2:. So SS 34 mm so 53m a e: 2.3. 2:. 3mm x: 3 8 333% m a: 32. $2 $3 a: E. NS ofiounézma. m a: 83 smw «Sn :2 mm so x026 363mm 33m $34 «\34 «\34 334 «\mnq «\34 «use CHOU Z oEdwhO Z manmnha Z wand Z amped“: anOZ ZunZZ «dog 3309 tenvkfioz nstfioumrwm 3309 mo mag“ .33 .Emofi an? mesa .Esmw quench .58 mo 33.2» was ”awoken .s «33. oficmmno was Handgun“ Go munmgummuu ammofiza Hmucmaodmmdm find ofipmmoh mo moommmm fiddpfinmm..- I... 34 The effects of supplemental nitrogen on the residual level of total organic nitrogen was greater on the check plot to which no residue was applied. Here the increase was 648 pounds of nitrogen per acre as compared with 352, 488, and 389 pounds where supplemental nitrogen was applied following alfalfa-brome, sawdust and straw re5pectively. Again 60 percent or more of the increase in total nitrogen occurred in the nonhydrolyzable fraction. All of the above increases in total organic nitrogen were signifi- cantly greater at one percent than the 120 pounds total supplemental nitrogen applied during the first three years of the rotation. This would imply that the fixation of atmospheric nitrogen by two years' growth of alfalfa was materially enhanced where supplemental nitrogen was used on preceding crops. Corn yields: Corn yield in 1955 was not influenced significantly by residue treatment. The addition of 100 pounds of supplemental nitrogen resulted in a significant average increase of 4. 3 bushels per acre. Within residue treatments, the increase for supplemental nitrogen was significant only in the case of the straw, where yields were depressed in the absence of nitrogen. This reflects the immobilizing influence of straw applied in the fall of 1958 Graphical analysis revealed only a slight overall tendency for yields to be higher at the higher levels of total nitrogen. Within residue treat- ments there was great variation in the relation between yields and total nitrogen, ranging from a distinct positive relationship for the check where no supplemental nitrogen was used to a negative relationship where supplemental nitrogen was applied on the check treatment. The scattered points for other residue treatments reflected varying degrees of ambiguity 35 and there appeared to be no basis for attempting a regression analysis of yields and total nitrogen. Graphical analysis also failed to show any useful relationship between corn yields and any of the nitrogen fractions measured. Ferden Farm Rotation Experiment I In this experiment soil samples were taken at the beginning of the fifth cycle of two five-year rotations inwhich two. fertility levels had been maintained and in which half of the area had received no nitrogen other than that applied at planting time and one half had received supplemental nitrogen applied on corn, sugar beets and small grain. The fertility - and supplemental nitrogen treatments were imposed on a split plot basis on all five replications of each rotation. The individual plot values for the various nitrogen determination and corn yields are presented in Tables 29 to 32 inthe Appendix. Mean values for the different treatments are given in Table 8. Inorganic forms of nitrogen: As was true in the case of the residue experiment, ammonium nitrogen was several times higher than normal .for fresh mineral soils in the field, indicating that some conversion of organic nitrogen to ammonium nitrogen may have occurred during drying and storage. .The levels of ammonium, nitrate, and total inorganic nitrogen were completely unrelated to treatment. Organic forms of nitro en: In Table 8 it appears that the only significant effect on total organic nitrogen was that of rotation. Twenty years after establishment there was a difference of 422 pounds in favour of Rotation 1. Only 24 percent of :Mufbfisd :n.~h.flv...~ \nd...~Pv Tunuufim nflPhuhduwfi FuwQquH-Q-‘ infinivntflavuflflflnh 13 .U!~..~\J1 -41..1JJ£~1-I ,,,,, ‘5'. P,Vl-I.l.~L.I. ‘DIJ [.lllwd‘!l.l‘»y,r I r l) II»! r Table 8. —-Residual effects of rotation, fertility level and supplemental nitrogen treatments on mineral and organic forms of nitrogen. Ferden farm. , Sims clay loam. 1958., . Total Hydrolyz- Nonhydro= ~ Total Treatment NH3—N NO3—N Mineral N able N lyzable N Organic N Corn Lbs/A Lbs/A Lbs/A Lbs/A Lbs/A Lbs/A Bus/A Rotation 1 vs 6 Rotation 1 74 ' 43 117 3252 874 4126 103. 6 'Rotation 6 75 43 118 2933 772 3705 69.. 0 LSD05 NS NS NS NS NS 302. 3 .7 16° 6 Low vs High Fertility Low fertility 70 47 117 3150 779 3929 82.3 High fertility 78 38 116 3035 868 3903 90. 3 LSDOS 7. 6 NS NS NS NS NS NS No Nitrogen vs Nitrogen No nitrogen 73 43 116 3156 ' 738 3894 78. 1 Suppl. N 75 43 118 3029 909 3938 94.5 LSD05 NS NS NS NS NS NS 8. Z Rotation vs Fertility Level Rotation 1 L F 71 49 120 3285 789 4074 97.0 Rotation 1 H F 76 37 113 3220 960 4180 110., 0 Rotation 6 L F 69 45 114 3015 770 3785 68. 0 Rotation 6 H F 80 40 120 2850 775 3625 70. 0 LSD05 NS NS NS NS NS NS 31. 2 Rotation vs Nitrogen Rotation l 73 40 113 3365 778 4143 101.0 Rotation 1 + N 75 43 118 3140 971 4111 106.0 Rotation 6 73 43 116 2948 698 3645 55. 0 Rotation 6 + N 76 43 119 2918 847 3765 83. 0 LSDOS NS NS NS NS NS NS 11., 5 Fertilizer vs Nitrogen Low F 65 50 115 3198 760 3958 78. 0 LowF+ N 75 44 119 3103 798 3901 87.0 HighF 81 36 117 3115 716 ‘ 3831 79.0 High F + N 75 41 116 2955 1019 3974 102.0 LSDO5 NS NS NS NS NS NS 11. 5 Rotation vs Fertilizer vs Nitrogen Rotation 1 L F NO 67 52 119 3400 782 4182 96. 0 Rotation 1 L F N1 76 47 123 3170 795 3965 97.4 Rotation 1 H F No 80 36 116 3330 774 4104 106.4 Rotation 1 H F N1 73 38 111 3109 1146 4255 114.4 Rotation 6 L F No 65 49 114 2995 738 3733 59. O Rotation 6 L F N1 73 41 114 3035 801 3836 76. 9 Rotation 6 H F No 82 36 118 2900 657 3557 50. 9 Rotation 6 H F N1 77 44 121 2800 893 3693 89. 3 LSDO5 NS NS NS NS NS NS 16. 3 9E Fertility level and supplemental nitrogen had no significant effect on either organic fractions or their total. However, there was a marked tendency for a reciprocal variation in the hydrolyzable and nonhydrolyzable fractions. Quite consistently hydrolyzable nitrogen was lower at the high fertility level than at the low and it was lower where supplemental nitrogen had been added than where it had not. Nonhydrolyzable nitrogen, on the other hand, was higher at the high fertility level and with supplemental nitrogen. The highest levels of nonhydrolyzable nitrogen were found where high fertility level and supplemental nitrogen were combined. This effect was greatest in Rotation 1. Conversely the lowest level of nonhydrolyzable nitrogen was found at the low fertility level without supplemental nitrogen. These trends, although not statistically significant by analysis of variance, appear to be significantly related to the observation of Mattson and Kouttler-Anderson (46), that the prOportion of acid resistant nitrogen in decomposing litter and humus increased with base status and nitrogen content of the original material. To determine whether a. similar dis- pr0portionate rate of increase in nonhydrolyzable nitrogen with total organic nitrogen existed in the soils studied here, linear regression analyw sis was resorted to. Table 9 presents the results of such analyses for these two rotations (Rotation 1 and Rotation 6) and for the residue experi- ment (Michigan rotation). Highly significant correlations ranging from 0. 632 to 0. 741 were obtained between hydrolyzable and total organic nitrogen for the three rotations. In the case of the nonhydrolyzable nitrogen, a relatively low though significant correlation of 0. 400 was obtained for the cash cr0p rotation (Rotation 6). This rotation contained no forage legume. E 38 —_ —_ Correlation Prediction equation Rotation Dependent variable with total organic X = Total organic nitrogen nitrogen (Y) (1") _ *4: /\ Rotation 1 Hydrolyzable N 0.632 Y = 1405 + 0.448'X ** A Nonhydrolyzable N 0. 714 Y = 1405 + 0. 552 X ** \ Rotation 6 Hydrolyzable N O. 741 Y = 263 + O. 721 X 3M: A Nonhydrolyzable N 0.400 Y = -263 + 0. 279 X Residue M, A experiment Hydrolyzable N 0. 720 Y = 1483 + 0.431 X *2}: .r\ Nonhydrolyzable N 0. 812 Y = -1483 + 0. 569 X *3}: /'\ Average Hydrolyzable N 0. 720 Y = 1142 + 0. 507' X - ** A Nonhydrolyzable N 0. 670 Y = 1132 + 0.493 X Both Rotation 1 and the Michigan rotation included two years of alfalfa- brome. In these two rotations highly significant correlations of 0. 714 and 0. 812 were obtained between nonhydrolyzable and total nitrogen. In these rotations there was a larger regression coefficient (0. 552 and 0. 569) for nonhydrolyzable nitrogen than for hydrolyzable nitrogen (0. 448 and 0. 431). In other words, over the range of values included in these rotations more than half of the increase in total nitrogen occurred in the nonhydrolyzable fraction. On the other hand, in the cash crop rotation only about one-fourth of each incremental increase in total nitrogen occurred in the nonhydrolyzable fraction (b = 0. 279). Johnston (37) found that the decomposibility of organic residues in soil decreased sharply withincreasing content of nonhydrolyzable nitrogen. The relationships observed in Table 9 suggest that there is a qualitative difference in the nature of soil organic matter formed under forage- legumes such as alfalfa (Rotation 1 and the Michigan rotation) than under non-legumes (Rotation 6). Organic matter formed under alfalfa would appear to be more resistant in nature because of its higher nonhydrolyzable nitrogen content. Millar it a}. (51) showed a greater residual accumu- lation of organic carbon and nitrogen after prolonged decomposition of leguminous materials than of nonlegumes. The marked tendency for nonhydrolyzable nitrogen to accmnulate at a more rapid rate than hydrolyzable nitrogen in the two legume rotations is shown graphically in Figure 1. The ratio of nonhydrolyzable to hydrolyzable nitrogenincreased from 0. 16 to 0. 36 as total nitrogen increased from 3300 to 4900 pounds per acre. ' Corn yields: Corn yields for the livestock rotation and the cash creprotation (Rotations 1 and 6) are shown in the last column of Table 8. Analysis of 4O 0.80.3133“? magma? .33 03 amohnm 03“ have? good can? macaw E ammonfid UEMMHO dBOu so ammo Cu somobws ofiouioupfasgu mo 03m." 05 mo sommmoumou can #8320 Go 33330." flu>~0kfl>£ p.80..- 4 mun—mam <\.mm... zwcomuLz o_z0 LS. moo. 13m o \\\ x 8.000.... mm... L s w x\ .. om. < b 5 em. 0 n on. O D O 1 NM. 0 D 4 0m. *\ o n 0?. C . 3. .. me. o p z moon > o D 2 Sort; KQRM Wmemt \ K00. .i .wk N39081|N 318V ZA‘IOBOKH NBSOHM N EWBVZA'IOHOAHNON 41 variance revealed that the significant main effects on yields were those of rotation and supplemental nitrogen. The average increase for the high fertility level was not significant. Significant interactions occurred such that there was a 20-bushel increase for supplemental nitrogen on the cash crap rotation (Rotation 6) and a nonsignificant Sobushel increase on the livestock rotation (Rotation 1). There was also a significant 23-bushel increase for supplemental nitrogen at the high level of fertility, but a 9—bushe1 increase at the low level of fertility was not significant at 5 percent. These differences observed in 1958 were similar to the long term yield results reported by Robertson (61, 62). Scatter diagrams of all observations revealed nothing but a very general trend between corn yields and hydrolyzable nitrogen, non- hydrolyzable nitrogen or total soil nitrogen. Graphical analysis of treat- ment means showed that consistently higher yields for Rotation l were associated with consistently higher'levels of total organic nitrogen (Figure Z-A). The effect of supplemental nitrogen was to raise the‘level of yield at a given level of soil organic nitrogen. There was a similar effect of high fertility level at each level of nitrogen. These effects of supple- 'mental nitrogen and fertility level on the level of reSponse to rotation were even more striking) when corn yields were plotted against hydrolyz- able nitrogen (Figure Z-B). In the case of nonhydrolyzable nitrogen, on the other hand, the effect of supplemental nitrogen was to displace the level of yield response to rotation horizontally in the direction of higher values for nonhydrolyzable nitrogen (Figure Z-C). In other words yield increases for Rotation 1 over Rotation 6 were smaller» as the level of 1 nonhydrolyzable nitrogenincreased. Where supplemental nitrogen was used this effect was exaggerated at the high fertility level. Thus there was a tendency for treatment means to approach a diminishing returns pattern. BUSHELS /A CORN YIELD KEY: nor/v.6 HOT/VJ NO NITROGEN O---0 LOW FERTILITY O ------- 0 HIGH FERTILITY o- ----- o PLUS NITROGEN o—-0 LOW FERTILITY o————o HIGH FERTILITY 0—-. I20- IIO- I00- 90- 80" 70- 60- 50-0“ 4o ' ' 3500 3700 TOTAL SOIL ORGANIC NITROGEN Lbs./A I I I I J I 3900 4IOO 4300 Figure 2. --Basic response of corn to rotation treatments asdrelated to supplemental nitrogen treatment, fertility level, an total organic nitrogen (A), hydrolyzable nitrogen (B). and nonhydrolyzable nitrogen (C) in Sims clay loam as reflected in treatment means. CORN YIELD BUSHELS IA ’ CORN YIELD BUSHELS IA 42 I20!" 8 IIO“ - I00- 90- soL 70- 60 T 50‘- 40 l 1 l L l l _l 2700 2900 3l00 3300 HYDROLYZABLE NITROGEN LBS./A IZOF C IIOF .’ IOOF '1’. I" 0". l':' so- . - [1:6 70F I': I": 0' ' 50" I (5 am <5 40 l I L l l J 600 800 I000 I200 NONHYDROLYZABLE NITROGEN 43 Production Function Analysis of Corn Yields for Rotations 1 and 6 Graphical analysis of treatment means described above suggested that it might be possible to fit a multifactor production function to the individual observations which would associate significant portions of the corn yield variance with known variables. A polynomial equation of - * the following type was formulated and fitted to the unit observations. 3‘! = a + 101xl + bzxfi + bax, + b4x3 + bsx, Where: . X1 = Soil nitrogen fraction X2 = Fertility level (low = a1, high = + 1) X3 = Rotation (Rotation 1 involving a forage legume = +1, T and Rotation 6 involving no legume = -1) X4 = Applied nitrogen (No nitrogen = -1, and supple- ~mental nitrogen = +1). In fitting the above function to the data, observed quantitative values for‘yield and soil nitrogen fractions were used. Due to the quali- tative nature of the rotation variable it was necessary to assign to the two rotations arbitrary numerical values the means of which would be zero. Because of the unknown cumulative residual effects of fertility treat- ments and supplemental nitrogen treatments over a period of four rota- tions, it appeared feasible to treat these also as qualitative variables. A numerical index of - 1 was assigned to the low fertility level, the cash crop rotation which contained no forage legume and to the no-nitrogen treatment. An index of +1 was assigned to the complementary'level of each of these variables. *Computational formulae used in these analyses are given in Appendix II. 4.4 Rotations 1 and 6: . Total nitrogen as soil nitrogen variable: The formulated production. function was first applied to the total nitrogen data for Rotations l and 6 combined. ‘ Equation I shows the fitted function. The standard errors of the estimated factor coefficients are shown in parentheses below each estimated value. Additional statistical measures shown include R2, the coefficient of multiple determination, and R, the coefficient of multiple correlation. Adjustment of R for the number of variables gives the , adjusted coefficient of multiple correlation, R, the square of which, R, is the adjusted coefficient of multiple determination. This last statistic provides an estimate of the percentage of total yield variance which was associated with the regression. 3 Also shown is the standard error, S, of the predicted yield. A Equation 1: Y = 2365.569728 + 1. 554255 x1 - .000184 x,2 + 3.012.541 x2 (1.153748) (.000148) (2.9093221) + 15.076648 x3 '+ 7.601438 x, M (3.198854) (2.784790) M = .674 = .821 R2 __2 R .611 .782 R E = 156.499 The adjusted coefficient of multiple correlation R, for this re- gression was . 782. The coefficient of multiple determination, Rf was . 611. This indicates that 61 percent of the variance in corn yield was explained by the relationship expressed in the regression equation. In this equation the estimated regression coefficients for rotation and applied nitrogen were significant at the one percent level, which indicates that corn yield was highly related to rotation and applied nitrogen. 45 None of the coefficients for other variables were significant at the 10 percent level of probability. The yields and independent variables were intercorrelated as follows: ryxl = . 512 ryxz = . 461 ryx3 == . 698 ryx‘ = . 332 Coefficients of determination estimated from these values indicate that approximately 48 percent of yield variance was associated linearly with rotation.,(X3), 26 percent with total soil nitrogen.(X1), and 10 percent with supplemental nitrogen treatment (X4); less than 2 percent was associ- ated with fertility level (X2). ' Analysis of variance had shown that dife- ferences in total soil nitrogen were mainly due to rotation. The above calculations suggest that as much as one-half of the rotational effect on corn yields may have been associated with differences in quantities of residual organic nitrogen. remaining in the two rotations. The individual coefficients of determination estimated in the last paragraph assume a linear relationship between independent and dependent variables. (If itcould be assumed that, in fact, the relationship between yield and total soil nitrogen were curvilinear, then it might be inferred that an even larger proportion of the variance associated with rotation might have been due to differences in the soil nitrogen. The non-significant coefficients for nitrogen in Equation I do not provide any strong evidence for sucha curvilinear relationship, although the negative sign for the coefficient of X12 does reveal a tendency towards curvilinearity. 46 Hydrolyzable nitrogen as soil nitrogen variable: The substitution of hydrolyzable nitrogen values for X! in the formulated function produced Equation II as follows: Equation 11: Q: —43. 163556 + .069063'xll - .000009 X12 + 4. 673082 x2 (.121159) (.000020) (2.928480) + 14.600104 x3 + 9.057379 xx, **(3.185036) (2.916283)** RZ=.671 R=.819 -2 .. = .607 R: .779 s = 15.734 The adjusted coefficient of correlation for the equation was . 779. The percentage variance in corn yield associated with regression was the same as in Equation I (61 percent). The standard error of estimate was less, however, by a factor of 10. Rotations and applied nitrogen were again found to be highly correlated with yield. The negative sign for the coefficient of X12 suggests a basic curvilinear relationship. However, neither of the coefficients for nitrogen were significant at 10 percent. The simple correlation between yields and hydrolyzable nitrogen accounted. for somewhat less of the total yield variance (15 percent) than did the simple correlation between yields and total nitrogen (26 percent). Nonhydrolyzable nitrogen as soil nitrogen variable: A third equation was calculated by using the nonhydrolyzable fraction as the soil nitrogen variable. The result of this fit is shown in Equation III: Equation 111: i? = 81. 289131 + .014340 x, - .000009 x,"- + 3. 988935'Xz , (.046781) (.000030) (3.019390) + 17. 390515'x3 + 8. 1805871:4 **(3. 985495) (3.134721)»: 47 R — .623 R: .790 .._.7- ._ R = .553 R: .743 S: 16.798 The adjusted coefficient of multiple correlation was . 743. The adjusted coefficient of multiple determination pointed out that 55 percent of the variance in corn yield was associated with regression. The co- efficient for rotation(X3) was significant at the one percent probability level, that for applied nitrogen (X4) at five percent. The rest of the coefficients were not statistically significant. Only about 5 percent of the total yield variance was accounted for by the simple correlation between yield and nonhydrolyzable nitrogen. Rotation 1 In fitting the above equations to the data for both Rotations 1 and 6, p it was observed that the major portion of the variance in corn yields was associated with rotation, and that at least a portion of the rotation- associated variance may have been due to differences in soil nitrogen which were inseparably associated with rotation. In point of fact, about 25 percent of the variation in total nitrogen was associated linearly with rotation and about 20 percent of the variation in hydrolyzable nitrogen was associated with rotation. Only about 3 percent of the variance in nonhydrolyzable nitrogen was associated with rotation. In order to elimiu nate the possible confounding of rotational and soil nitrogen effects in the fitted function, the formulated function was modified by eliminating the rotation variable. The modified function was fitted to the data for each rotation separately. Total nitrogen as soil nitrogen variable: In Equation IV yield is exPressed as a function of total nitrogen.(X1), fertility level (X2) and applied nitrogeni(X4) for Rotation 1 only. Equation 1v: T? = -350. 996089 + .225023 X, - .oooozsxf + 3.737259xz (.192736) (.000024) (4.722668) + 1.858591 x, (3.899593) R’- = .308, R = .555, R2 = .0558, R a .236, 5: 15.486. The adjusted coefficient of multiple correlation for this equation was . 236, and only about 5 percent of the yield variance was associated with regression. None of the individual coefficients was significant at 10 percent. It is apparent that none of the measured variables were associated significantly with yield in accordance with the postulated function. Hydrolyzable nitrogen as soil nitrogen variable: Hydrolyzable nitrogen values for Rotation 1 when substituted in the modified expression produced Equation V: A Equation v: Y = 1159. 352782 + .777978 xl -. 000119 x,z (.364969) (.000056) + 5. 673704 x2 + 4. 266871 x4 (3.526206) (3.808620) .2 _ R2: .458, R = .677, R: .261, R = .511, 5: 13.700. Here‘both the coefficients for nitrogep were significant at the 10 percent level of probability. However, the R indicated that only about 26 percent of the variance in yield was associated with the regression. There was a slight reduction in the standard error of estimate as compared with Equation IV. The larger negative coefficient for X12, provides more evidence of a curvilinear relationship between yields and this nitrogen 48 49 fraction than was apparent in any of the previous equations. This does suggest an approach towards a postulated diminishing returns type of relation between yield and hydrolyzable nitrogen. However, the low proportion of total variance associated with the over-all regression makes this observed relationship with hydrolyzable nitrogen of little practical significance. The lack of statistical significance in the applied nitrogen variable would have been expected in this rotation where no significant response to applied nitrogen was revealed by the analysis of variance. ' Rotation 6 Total nitrogen as soil nitrogen variable: In Equation VI are shown I the results of fitting the modified function to the data for Rotation 6, with total nitrogen as the soil nitrogen variable. . A 2 Equation VI: Y = 431. 373400 - . 216364 X, + . 000032 X1 (. 249506) (. 000034) + 4. 354247 x2 + 12. 635942 x, (4.049600) (3. 569171)“ R2 = .652 R: .807 -2 ._ R = .525 R: .725, 3: 14.024. The adjusted coefficient of multiple correlation was . 725 and about 52 percent of the total variance in yield was associated with the regression. However, only the coefficient for applied nitrogen (X4) was significant at. 1 percent. Coefficients of determination calculated from simple corre- lation coefficients indicated that about 50 percent of the yield variance was associated linearly with applied nitrogen, about 17 percent with the soil nitrogen variable and essentially none with fertility level. 50 Hydrolyzable nitrogen as nitrogen variable: When hydrolyzable nitrogen was substituted in this equation for Rotation 6 the estimated parameters were quantitatively similar to those in Equation VI (compare Equation VII). A Equation v11: Y = 321. 668840 - .202849 x1 + .000039 x3 (.342295) (.00C06l) + 5.131337-X, + 14.828507 x, (5.624063) (3.796944) Rz=.611 R=.782 R2 = .469 R = .685, 5: 14.822 The percentage variance associated with regression in Equations VI and VII was quantitatively of the same order (52 percent vs 47 percent). The quantitative similarities in form and effectiveness of Equations VI and VII would have been expected since there was little evidence of a change in the proportion of hydrolyzable nitrogen to total nitrogen with increasing total nitrogen in this rotation. This is in contrast with Rotation 1 where the proportion of hydrolyzable nitrogen was found to decrease with increasing total nitrogen. As a result the form and effectiveness of Equations IV and V were rather dissimilar. No attempt was made to fit production functions involving nonhydro- lyzable nitrogen to the individual rotations. DISCUSSION Significance of Mineral Forms of Nitrogen Nitrate: Nitrate may be considered the most available form of nitrogenfor many crop plants. The levels of nitrate found in the soil at any given time will reflect a balance between rate of mineralization of organic nitrogen and rate of removal by crop uptake, leaching or denitrification. Quantities of nitrate present at the time a crop is planted will reflect the initial supply and may show some relation to final crop yield (53, 85). In the case of nutrient level exPeriments on the Caznpbell farm (Table 3) nitrate levels were significantly related to previous treatments. The quantitative differences would have been enough to influence yields of the following wheat crop. Such effects were masked to a large extent by the current application of fertilizer. However, Sundquist (73), using data from the same experiment, found an increase in the correlation between wheat yield residuals from plant nutrient input-output functions and the nitrogen variable when nitrate present at planting time was added to nitrifiable nitrogen determined by an incubation procedure. Anderson (6) had found very little correlation with nitrifiable nitrogen alone. Sundquist's experience may have arisen from fortuitous circum— stances related to the time when soil samples were taken. In the absence of leaching rains during the period between oat harvest and the time of soil sampling, it was possible for large differences in accumulated nitrate to develop. These differences existed at the time the wheat was planted, and it would be reasonable to expect that they might have influenced wheat yields. In humid areas leaching rains during the fall, winter and spring 51 52 tend to erase any differences in nitrate accumulating capacity between soils by planting in the Spring time. This effect may be observed in the nonsignificant differences in nitrate which were found at locations where soil samples were taken in spring (Tables 1, 5, and 8). For the same reason, significant differences in nitrate found in fall samples from the residue experiment (Table 7) would have had little bearing on the yields of corn planted the following spring. Research completed and in progress at the Michigan Agricultural Experiment Station indicates that soil nitrate level during the period of peak nitrogen requirement (about tasseling time) may be very significantly related to corn yields (18, 88). Ammonium: The normal range of exchangeable ammonium nitrogen expected in mineral soils in the field is from 3 to 30 pounds per acre (35). The quantities of ammonium found in the soils used in this investigation were of the order of 3 to 4 times greater than would have been expected in the same soils in the field. It is safe to assume that breakdown of organic compounds during air—drying and storage over a 3 to 4 year period con- tributed to the high levels of ammonium encountered. The effect of air- drying would be similar to that of steam sterilization (86), and probably accounts for the increasing nitrifying capacities associated with prolonged air dry storage, as reported by Harpstead and Brage (33). The quantities of ammonium recorded were related in a general way to total organic nitrogen, in that both were considerably higher in the heavy Sims soil (Tables 5, 7, 8) than in the Kalamazoo sandy loam (Tables 1, 3). ' To the extent that ammonium was released from organic combination during air dry storage, it might be assumed to represent a more labile, hence more readily available, fraction of soil organic 53 nitrogen. However, no consistent. trends were observed between this fraction and yields in any of the experiments. Significance of Organic Forms of Nitrogen A technical problem involved in studies of soil nitrogen derives from the fact that yearly or rotational changes are small relative to the total quantity present. Errors inherent in the chemical determination are of the order of 1 to 2 percent of the total. - This is in addition to sampling errors. Thus errors in the determination of total soil nitrogen are of the same order of magnitude as the annual increments of ferti- lizer nitrogen used in normal management. Annual changes in total organic nitrogen due to residual from applied nitrogen would be even less. For these reasons, rather extensive replication is necessary to establish the reliability by analysis of variance of differences in total soil nitrogen associated with treatment at a given location on» a given soil type. ‘Differences which can be shown, even with adequate repli- cation, will normally be greater than year to year changes. Accumulative residual effects from several years of treatment are necessary to give differences that can be detected with any degree of reliability. Kamerman and Klintworth (39) found that, while total organic nitrogen remained unchanged for annual periods after addition of ferti- lizer nitrogen, the proportion of hydrolyzable nitrogen varied over a considerable range. Similarly Johnston (37) found more extreme variation in the hydrolyzable and nonhydrolyzable fractions than in their total following the addition of various residues with and without supple- mental nitrogen. These considerations led to the hope at the outset of the present studies that changes in one or the other of these fractions might be detected with greater precision than changes in total organic nitrogen. These hopes were not justified by the data. In the rotation and residue experiments there was evidence that nonhydrolyzable nitrogen increased with total nitrogen at a more rapid rate than did the hydrolyzable fraction. However, this behavior could not be related to crop performance. ‘ Attempts to correlate these two fractions with crOp yields were plagued by two major difficulties: 1. Inadequate knowledge of the normal range or the effect of soil and management factors on distribution of nitrogen between these two fractions. 2. The lack of anything more than fragmentary data regarding the significance of nitrogen in these two fractions to mineraliZn ability or crop availability. Factors affecting the proportion of nonhydrolyzable to total nitrogen: Mattson and Kouttler—Anderson (46) and Johnston (37) found that the acid resistant nitrogen fraction in decomposing plant residues and . humus increased with nitrogen content of the original residues or with the level of supplemental nitrogen. Proximate analyses of composting plant materials reported by Tenney and Waksman (74) reflect increasing levels of residual acid-resistant nitrogen with increasing original nitrogen content or added mineral nutrients. The effect of added nutrients reported by Tenneyand Waksman parallel the observation by Mattson and Kouttleru- Anderson (46) that acid-resistant nitrogen increased with the base status (cation content) of the soil or of the original litter. Johnston (37) and Tenney and Waksman (74) also associated high levels of acid-resistant nitrogen at advanced stages of decomposition with high original lignin content of plant residues. Trends observed in the present studies were consistent with the experience of these investigators. In the rotation experiment (Table 10), supplemental nitrogen applied to corn, sugar beets and cereal grains over four rotational cycles increased the prOportion of nitrogen not .5 5 hydrolyzable by sulfuric acid in both rotations and at both fertiiity levels. The effect of nitrogen was greater in the legume rotation than in the cash crop rotation, and it was strikingly enhanced when combined with a high level of addition of fertilizer salts. In the residue experiment (Table 11), the return of'high nitrogen residues in the form of two cuttings of alfalfa—brome hay increased the proportion of nonhydrolyzable nitrogen from 17 percent of total soil nitrogen to 23 percent. The use of supplemental nitrogen on the first three crops in the rotation raised the prOportion to 26 percent when combined with alfalfa-brome residues. The effect of sawdust and straw treatments were less marked, but supplemental nitrogen had a similar effect with these two materials as with the alfalfa-bromehay. Trends following a single year's fertilization in the nutrient level experiments were somewhat erratic. However, the combination of large applications of nitrogen with large additions “of phosphate and potash at the Fick farm resulted in large increases in the pr0portion of non- hydrolyzable nitrogen in Kalamazoo sandy loam (Table 12). 'At the other two locations, the overall effect of added N, P and K was in the same direction. At the Campbell farm on Kalamazoo sandy loam the proportion of nonhydrolyzable nitrogen was 19 percent where (no fertilizer was used and 22 percent as an average for all fertilized plots. Corresponding figures for Sims clay at the Thompson farm were 29 percent for unferti- lized soil and 32 for fertilized. An unexpected phenomenon encountered in these studies was an apparent synergistic effect of fertilizer nitrogen applied for preceding I crops on nitrogen fixation by alfalfa in the residue experiment (Table 7). Increases in total organic nitrogen recovered exceeded by three or four times the sum of the supplemental nitrogen added on the first three crops of the rotation. By statistical criteria, the net gain in nitrogen would 56 Table 10. --Percentage nonhydrolyzable nitrogen as related to supple- mental nitrogen treatment and levels of applied nutrients. Sims loam. Ferden farm. :======." Treatment Nonhydrolyzable N Treatment Nonhydrolyzable N Percent of total * Percent of total Cash crop rotation 19 Low fertility 19 Cash crop rotation + N 22 Low fertility + N 20 Legume rotation 19‘ High fertility l9 Legume rotation + N 24 High fertility + N 26 5): Low fertility = 800 pounds of 5-20-10 over 5-year period. High fertility = 1600 pounds of 5-20-10 over 5-year period. Table 11. --Percentage nonhydrolyzable nitrogen as related to supple- mental nitrogen and residue treatment. Sims loam. Ferden farm. Treatment Nonhydrolyzable N Treatment Nonhydrolyzable N Percent of total Percent of total No residue 17 Sawdust 19 No residue + N 24 Sawdust + N 22 Alfalfa-brome 23 Straw 20 Alfalfa-brome + N 26 Straw + N 24 57 Table 12. --Percentage nonhydrolyzable nitrogen as related to level of applied nutrients. Kalamazoo sandy loam. Fick farm. Treatment Nonhydrolyzable N Treatment Nonhydrolyzable N Percent of total Percent of total 0-0-0 ' 26 20-240—80 23 20 ~- 40 - 80 26 240 - 240 - 80 30 20 -160 - 80 25 20 - 240 -240 22 240 -160 - 80 27 240 240 ~240 28 58 appear to have been real. However, a similar effect was not observed in the legume rotation (Rotation l) of the rotation experiment. Significance of nonhydrolyzable nitrogen to nitrogen availability: Very few investigators have concerned themselves with the relationship between the acid-hydrolyzable or acid-resistent nitrogen fractions and the mineralizability or availability to crops of soil organic nitrogen. In respiration experiments, Johnston (37) found that resistance to microbial decomposition of organic matter in soil increased with increasn ing nonhydrolyzable content. Data by Waksman and Iyer (82) for their "ligno—protein" complexes may be similarly interpreted. This would suggest that changes in the resistant fraction may reflect qualitative changes in the total organic complex. These changes in turn may affect the mineralizability of nitrogen combined in the organic complex and its availability to plants. On the other hand, Johnston found that the proportion of nonhydro- lyzable nitrogen was drastically reduced in the presence of a growing nonleguminous crop, wheat. This would indicate that the organic combinations reflected in this fraction are not completely resistant to decomposition. Rather, it would appear that a reduced rate of release might result in continued release over a longer period of time. This is further suggested by seasonal patterns of nitrate release obtained by Brock (18) for the cash cr0p and livestock rotations studied here (Rotations 1 and 6). Soil nitrate levels were much higher all season in the livestock rotation than in the cash crop rotation. This was related primarily to the higher total nitrogen in the soil (Table 8). During the period of peak nitrogen requirement for corn, soil nitrate levels declined sharply on both rotations, both where supplemental nitrogen was used and where it was not. After the period of maximum uptake by corn, nitrate tended to accumulate again only on plots with a history of 5 9 supplemental nitrogen fertilization. This sustained capacity for nitrate release was associated with the high proportion of nonhydrolyzable nitrogen in soil from plots receiving supplemental nitrogen (Table 10). Significance of hydrolyzable nitrogen to nitrogen availability: Investigators elsewhere have attempted to measure labile portions of soil organic nitrogen on the premise that organic nitrogen combinations with a low degree of stability to mild chemical reagents would also be those most readily attacked by soil microorganisms. The nitrogen in such fractions would be exPected to be more closely related to seasonal availability to crOps than would the total nitrogen present. The alkaline permanganate distillation procedure is representative of this approach (76). Correlation studies involving this measurement have not given any strong support for the basic premise (6, 53). The acid-hydrolyzable nitrogen measured in the present study represents an analogous "labile" fraction. However, the chemical treat- ment was drastic, and 2/3 to 4/5 of total soil nitrogen was involved. Organic nitrogen compounds and complexes reflected in this fraction would represent a wide range of biological activity in terms of decomposibility. Also, carbonaceous constituents (sugars, uronic acids, etc.) present in labile combination would provide readily available energy for microbial synthesis and immobilization of associated nitrogen. These factors tend to undermine any theoretical postulation that crop performance might be directly related to the quantity of nitrogen released by acid hydrolysis. Johnston (37) found no relationship between the hydrolyzable fraction and mineralizability of soil organic nitrogen or its availability to wheat. On the other hand, Kamerman and Klintworth (39) found that nitrate released during 30 days'incubation varied inversely with C:N ratio of the hydrolyzable fraction of soil organic matter. Their results would suggest that both carbon and nitrogen contents should be 60 taken into account in attempts to correlate any labile fraction of soil organic matter with net mineralization or crop availability. A further noteworthy observation of Kamerman and Klintworth was their finding that the C:N ratio of the hydrolyzable fraction was very sensitive to recent additions of fertilizer nitrogen. Seasonal patterns appeared which covered a wide range of C:N ratios, and these patterns were related to treatment. In the present work, lack of sensitivity to recent fertilizer treatment in the nutrient level experiments was found to be a major liability of the chemical nitrogen determinations. Unless such sensitivity can be achieved, there‘can be little hope that chemical determinations can be used to explain the residual benefits which experi- ence has shown do accrue to applied fertilizer nitrogen. The relationship between C:N ratio of fresh plant materials and the balance between net mineralization or net immobilization of nitrogen during decomposition is well known. Theoretically the C:N ratio of recently returned crop residues should be reflected in the labile fractions of soil organic matter. A fruitful course for future fundamental research would appear to involve investi- gation of the relationship between mineralizability of soil nitrogen and the C:N ratio of various "labile" fractions. Such fundamental research is essential before mathematical models for functional correlation of soil organic nitrogen with cr0p response can be postulated with any approxi- mation to reality. Evaluation of Experimental Design The field experiments used in this investigation were not designed specifically for the correlation studies which were attempted. The three nutrient level experiments were designed for continuous function analysis of fertilizer input-crop yield output data. Minimal replication was employed with a large number of treatments to facilitate delineation of a yield 61 response surface. Precise determination of discrete yield points or residual nutrient levels as affected by treatment were irrelevant to this objective. Inadequate replication made it impossible to establish by the methods of analysis of variance whether significant variations in soil nitrogen had been achieved. In these three experiments, observed increases in residual soil organic nitrogen resulting from one year's application of nitrogen fertilizer ranged from 10 to 220 pounds per acre in the nonhydrolyzable fraction and from net losses to an increase of 140 pounds in the hydro- lyzable fraction. A maximum variation between duplicate chemical determinations of 40 pounds per acre was allowed in the nonhydrolyzable fraction and 80 pounds in the hydrolyzable. This variability inherent in the chemical determinations is large relative to the variance encountered in the soils studied. The additional replication afforded by sampling the applied nitrogen variable at several levels of P and K application proved inaffective because of apparent interactions between applied nitrogen and the P K combination. The two replications available at this level of subdivision were inadequate to establish the statistical validity of the apparent interactions or of the differences between treatment means. These difficulties might have been overcome to a degree by taking advantage of the internal replication available in regression analysis. This was not done. It would appear that future studies should make provision in experi- mental design for effective use of regression analysis to relate measured soil parameters with soil treatment variables. The failure to establish the reliability of specific soil nitrogen levels seriously weakened any inferences to be drawn regarding relationships between fertilizer treatment and changes in level or quality of soil nitrogen. An intercorrelation of soil tests and fertilizer inputs was inherent in the experimental designs used in this study. The historical treatments 62 which were expected to produce the required variance in soil nitrogen level were repeated as current inputs for the cr0p with which correlation was attempted. To an indeterminate degree, yield variance associated with differences in soil test arising from prior treatment would have been included in the variance associated with current treatment, since prior and current treatments were the same on any given plot. In effect, the portion of yield variance available for segregating the effects of soil test from those of current treatment was reduced to that portion associated with the differential availability or effectiveness of soil forms of nitrogen as compared with fertilizer forms. Considering the relatively large errors involved in sampling and chemical analysis, this loss of avail- able yield variance would appear to be prohibitive. Similar criticisms may be directed to the rotation and residue experiments. Here more extensive replication and longer periods of prior treatment gave rise to statistically significant differences in the level of total soil nitrogen and, in the residue experiment, of the two individual fractions. However, current treatments and treatments prior to sampling were the same, so that the soil nitrogen variable and the treatment variables tended to be intercorrelated. Also, the total number of measured points in each experiment was small relative to the require- ments for effective functional analysis. Evaluation of Functional Analyses The prediction functions formulated here represented an empirical incorporation of variables, some of which were in themselves hybrid combinations of input factors and levels of management. Rotation and supplemental nitrogen treatments were significantly reflected in the soil test values themselves. Rotational levels of applied nutrients (N, P and K) tended to influence the soil nitrogen values in a way which was consistent 63 with theoretical expectation, even though the observed effects were not significant statistically. When the data for the cash crop and livestock rotations were com- bined, a maximum of 61 per cent of yield variance was associated with regression when total nitrogen or the hydrolyzable fraction was used as the soil nitrogen variable. Only 55 percent was associated with regres- sion when nonhydrolyzable nitrogen was used. However, the major portion of the explained variance was associated with rotations and supplemental nitrogen. Since these two variables were responsible for significant variations in soil nitrogen, significant portions of yield variance associated with them would have accrued to the soil nitrogen variable had rotation and supplemental nitrogen treatment been left out of the formulated function. It would have been logically invalid to eliminate the supplemental nitrogen variable from the function, since a major part of the corn response would have been to the current year's input rather than to residual accumulation from prior treatment. Some argument might be offered for eliminating the rotation variable from the function and allow- ing intercorrelated variance to associate itself solely with the soil nitrogen variable. Rotations represent management levels, and, as such, are not subject to direct quantitation. To the extent that management levels give rise to measurable increments of identificable soil parameters, these parameters may be used to characterize management level quantitatively and may be substituted for it in production functions of the form used here. Conceivably, not all the chemical or physical changes associated with management level will be identifiable. In the combined :analysis for Rotations l and 6, about 26 percent of yield variance was associated with total soil nitrogen and 48 percent with rotation. Thus, it might be esti- mated that of the order of one-half of the effect of rotations on yield was due to total soil nitrogen and half to other unidentified factors associated with rotation. 64 Elimination of rotation from the function might have increased the significance of the regression coefficients for soil nitrogen, but it would also have increased the portion of variance unaccounted for. Due to shortage of time and logical reservations regarding the suitability of the data and the essential validity of the formulated functions, no attempt was made to fit such a function or other alternative functions to the combined data for both rotations. The attempts made here to eliminate rotation by restricting analy- sis to within-rotation variance was not very informative. In the livestock rotation (Rotation 1), coefficients approaching significance for hydro- lyzable nitrogen were obtained. However, the overall effectiveness of . the regression was low; only about 26 percent of the yield variance was accounted for. This was not unexpected, since the available yield variance was small and analysis of variance had revealed no significant differences in yield for the nitrogen and fertility subtreatments in this rotation. In the cash crop rotation (Rotation 6), greater yield variance was available for analysis. ‘ Functions involving either total or hydrolyzable nitrogen accounted for approximately 50mpercent of the variance in yield. Intercorrelations with supplemental nitrogen and fertility level may have contributed to lack of significance of the soil nitrogen coefficients. Highly significant coefficients for applied nitrogen were consistent with the results of analysis of variance. The degree of intercorrelation of independent variables inherent in the exPerimental designs would have reduced the effectiveness of any production function that might have been employed. However, a major difficulty was thelack of any logical or theoretical concepts to use as a basis for formulating a regression expression. Two fractions of organic nitrogen were measured directly, a third might have been estimated from the ammonium nitrogen found in air dry 6.5 soil. Attempts to postulate appropriate production ftmctions based on these measurements at once raised a number of questions. Does any measured fraction by itself represent an independent estimate of the potential availability of total soil nitrogen to the crop? Or does each represent a distinct level of chemical stability or remstance to microbial decomposition, making it necessary to include all measured fractions as separate variables in the same production function? Beyond these were even more fundamental questions. Is this fractionation potentially as useful as some others that might have been used? Few comparative studies of the significance of various nitrogen fractionation schemes to availability or mineralizability of soil organic nitrogen have been made. To what extent do carbonaceous materials associated with each fraction promote net immobilization or net minerali— zation? How might these be estimated and expressed in quantitative terms relevant to functional analysis ? These problems can be investigated most effectively by fundamental studies at the laboratory and greenhouse level. Until such studies have been made and tentative working hypotheses established, any attempt, such as that made here, to apply functional analysis to field data must. be considered premature. On the other hand, the requirements of functional. production analy- sis should be kept clearly in mind in fundamental studies. The results of agronomic research are translated by agricultural economists into farm management terms by application of formal deductive principles which rely on the manipulation of functional relationships. For this reason, fundamental agronomic research should be concerned with un- covering functional relationships between soil parameters and plant response. For this purpose, continuity of measurement along the surface of the function is of greater utility than precise measurement of discrete points. Functional mathematical analysis together with approPriate 66 measures of reliability should be included among the statistical tools for evaluating fundamental agronomic data. Agro-Economic Considerations Some of the more obvious relationships in the data are of interest. In the case of the rotation experiments it was found that corn r65ponded to supplemental nitrogen with significant yield increases in the cash crop rotation (Rotation 6) but not in the livestock rotation (Rotation 1). Accord- ing to Michigan Extension Bulletin E- 159, Fertilizer Recommendations for Michigan Crops (49), no response to nitrogen in addition to that in the starter fertilizer would be expected where 8 tons of manure and a legume sod are plowed down for corn. Ten tons of manure and a legume sod had been plowed down for corn in Rotation 1 over four cycles of the rotation. Thus the general validity of the current basis for fertilizer recommend- ' ations was borne out by these data. On the other hand, 100 pounds of supplemental nitrogen on the cash crop rotation failed by a significant increment to achieve the loo-bushel yields which might have been expected on this soil type. Obviously, the fundamental relationship between corn yields and supplemental nitrogen is different under these two systems of management. However, the data provide no clues as to the differences in form between these two functions. Speaking in general terms, available agronomic data is characterized by a large proportion of unexplained variance. The field experiments from which such data are derived are frequently not designed to facilitate the eXposition of functional relationships. Where there is a large amount of unexPlained variance it becomes difficult to select appr0priate mathematical functions. The combination of fair to poor agronomic data with empirically selected mathematical functions which may or may not express real relationships contributes to relatively low precision in making fertilizer recommendations . 67 The chemical data for these two rotations show that significant differences in residual organic nitrogen have resulted from the two systems of management. The reduced response to supplemental nitrogen in the livestock rotation was associated with a residual supply of organic nitrogen which was 420 pounds greater than in the cash crop rotation. There was no obvious basis for evaluating this difference in terms of its contribution to yield differences. Independent estimates made at different fertility levels and supplemental nitrogen treatments ranged from 6 to 22 pounds increase in soil nitrogen per bushel increase in yield in the livestock rotation as compared with the cash crOp rotation. Such variations cannot be interpreted in terms of a uniform seasonal rate of decomposition and release of nitrogen in soils of varying organic nitrogen contents. If a uniform seasonal rate of decomposition equal to 2 percent of total soil nitrogen were assumed, the. difference in total nitrogen between the two rotations would represent a difference in nitro- gen released to corn of only 8 to 10 pounds per acre. This is hardly enough to account for the 46 bushel difference in yield where no supple- mental nitrogen was used. These observations strongly suggest that differences in quality as well as quantity of soil nitrogen were involved. The need for significant measures of quality which will reflect both durability and the immobili- zation potential of associated carbonaceous material is apparent. The search for such measures should involve experimental designs which will permit functional interpretation. In the case of the residue experiments, the return of two cuttings of alfalfa hay resulted in a significant 300 pound increase in total soil nitrogen after five years. There was, however, no effect on yields of corn. There is no basis for placing a value on the residual nitrogen increment in terms of capital gain or of estimating its recovery in succeeding crops over a period of years. Collateral effects on soil 68 physical properties which may influence yields in later years are also unknown. For the farm manager these data present the alternatives of a tangible loss in feed and income from two unharvested cuttings of hay against a measurable increase in soil organic matter the benefits of which are intangible. Of course, it is possible that a psychic satis- faction might be derived, based on the traditional conviction that soil organic matter is a good thing and every effort should be made to build it up. Similar significant increases in total soil nitrogen were obtained five years after a 35-ton application of sawdust. Corn yields were not affected in 1959. In two previous seasons, corn yields planted six years after similar sawdust applications were significantly higher than for any other treatment. However, these residual benefits must be weighed against the high cost of transportation and application of such large amounts of sawdust, as well as large reductions in yield of crops planted during the first three years after application. Large investments in supplemental nitrogen during these first three years would have reduced the injurious effects of the sawdust and would probably have enhanced the residual accumulation of organic nitrogen. Here again the value of this residual nitrogen could not be assessed at the present state of our knowledge. The chemical data and yield results reported here are of agronomic interest. The nitrogen fractions studied did not reflect previous treatment with the hoped for degree of sensitivity. However, the differences in total nitrogen between rotations, residue treatments and supplemental nitrogen treatments do help to characterize the effects of these treat- ments on soils in a general way. It is obvious that more sensitive-measures are needed. Also that the observed relationships must be defined with much more specificity before soil nitrogen determinations can provide information useful for purposes of farm management. SUMMARY AND CONCLUSION The principal objective of the present study was to evaluate several chemically derived fractions of soil nitrogen in terms of their sensitivity as measures of residual nitrogen from previous treatment. In this regard, the findings may be summarized as follows: 1. The level of ammonium nitrogen in air dried samples was higher than expected in field fresh soils. Ammonium was probably released by breakdown of organic compounds during air drying and storage. The quantities of ammonium found were related in a general way to the content of organic nitrogen but were not closely related to treatment. Nitrate nitrogen was related to treatment in soil samples taken in the fall of the year. ‘Leaching during the fall and winter apparently obliterated any significant differences in soil samples taken in the winter or Spring. The two organic fractions measured were not measured precisely enough to reflect significant differences in residual nitrogen from a single year's application ranging up to 320 pounds of N per acre. Cumulative effects of rotation, residue treatment and supplemental nitrogen treatments over a period of years were reflected in the hydrolyzable fraction, the nonhydrolyzable fraction or in the sum of the two. There was a relatively more rapid increase in the nonhydrolyzable fraction than in the hydro- lyzable fraction where alfalfa (was included in the rotation. Supplemental nitrogen treatment and the level of'application of other fertilizer nutrients also promoted a disproportionately 69 rapid increase in the nonhydrolyzable fraction. These results are consistent with those reported by other investigators who have studied similar nitrogen fractions in composting organic residues . A secondary objective in this investigation was the application of functional analysis as a statistical tool for evaluating the significance of the measured nitrogen fractions to crOp performance. The attempts made here to formulate and fit prediction equations to field data appear to have been premature. Two principal obstacles were encountered: l. The lack of fundamental data bearing on the mineralizability or availability of nitrogen in the two organic fractions. Two aspects of this problem need investigation: (a) the relationship between the proportion of labile to resistant nitrogenous fractions and their mineralizability, and (b) the effect of associ- ated carbonaceous constituents on net mineralization. Z. The experimental design of field experimentsxwere inappropriate for correlating soil tests with crop performance. In all experi- ments there was excessive intercorrelation between treatment variables and soil tests. In the rotation and residue experiments, the number of levels of any treatment factor was inadequate to permit the application of regression analysis. Future research in this area should consider the use of experimental designs which will permit the application of regression analysis to uncover functional relationships between treatment variables and measured nitrogen fractions on the one hand, and between the measured fractions and crop response to fertilization on the other. Increased precision in defining such functional relationships is essential if soil tests are to contribute to in— creased precision of fertilizer recommendations based on current principles of economic analysis. 10. 11. 12. 13. BIBLIOGRAPHY . Adams, A. P., Bartholomew, W. V., and Clark, F. E. Measurement of nucleic acid compounds in soil. Soil Sci. Soc. Amer. Proc. 18:40-46, 1954. . Allison, F. E. Estimating the ability of soils to supply nitrogen. Agric. Chem. 11 (No. 4):46-48, 1956. . Allison, F. E., and Sterling, L. P. Nitrate formation from soil organic matter in relation to total nitrogen and cropping practices. Soil Sci. 67:239-252, 1941. . Allison, F. E. Nitrate assimilation of soil microorganisms in relation to available energy supply. Soil Sci. 24:79-92, 1927. . Anderson, C. Identification of derivatives of deoxyribonucleic acid in humic acid. Soil Sci. 86:169-174, 1958. . Anderson, G. R. An economic evaluation of three soil nitrogen tests. M. S. thesis, Department of Agricultural Economics, Michigan State University, 1958. Bartholomew, W. V. Availability of organic nitrogen and phOSphorus A from plant residues, manures and soil organic matter. Papers of the Soil Microbiology Conference. Purdue University, 1954. Bartholomew, W. V., and A. E. Hiltbold. Recovery of fertilizer nitrogen by oats in the greenhouse. Soil Sci. 73:193-201, 1952. Bollen, W. B. Soil reSpiration studies on the decomposition of native organic matter. Iowa State Jour. Sci. 15:353-374, 1941. Bremner, J. M. Studies on soil organic matter. Part I. The chemi- cal nature of soil organic nitrogen. Jour. Agric. Sci. 39:182-193, 1949. Bremner, J. M. The amino acid composition of protein material in soil. Biochem. J. 49:538-542, 1950. Bremner, J. M. A review of recent work on soil organic matter. Part I. Jour. Soil Sci. 2:67, 1950. Bremner, J. M. , and Shaw, K. Studies on the estimation and decomposition of amino sugars in soil. Jour. Agric. Sci. 44: 152-159, 1954. 71 14! 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 72 Bremner, J. M. -Amino-sugars in soil. J. Sci. Fd.vAgric. 9:528- 532, 1958. Bremner, J. M. Determination of fixed ammonium in soil. J. Agric. Sci. 52:147-160, 1959. 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. Broadbent, F. E. , and Bartholomew, W. V. The effects of quantity of plant materials added to the soil on its rate of decomposition. Soil Sci. Soc.- Amer. Proc- 13:271-274, 1948. Brock, W. Unpublished data. VDept. of Soil Science, Michigan Agr. Exp. Sta. 1958-1959. Brown, P. E. Bacteriological studies of field soil. II. The effects of continuous cropping and various rotations. Iowa Agr. Exp. Sta. Res. Bull. 6:214-245, 1912. - Burris, R. H. Nitrogen nutrition. Annual Review ovalant Physiology. Vol. 10:301-328. 1959. . Chapman,» H. D. , and Liebig, George F. , Jr; Field and laboratory studies of nitrite accumulation in soil. Soil Sci. Soc. Amer. Proc. 16:276-282, 1952. Dunn, L. E. ,v and Wheeting, L. C. Utilization of barnyard manure for Washington soils. Washington Agr. Exp. Sta. Bull.. 395, 1941. Ensminger, L. E., and‘Pearson, R. W. Soil nitrogen. Advances _ in Agronomy Vol. II. - Edited by Norman, A. G. Acad. Press, Inc. : 81-109, 1950. Fathi, O. M., and Bartholomew, W. V. Influence of oxygen concen- tration in soil air on nitrification. Soil Sci. 71:215-219, 1951. Fitts, J. W., Bartholomew, W. V. and Heidel, H. Predicting nitrogen ' fertilizer needs of Iowa soil. Soil Sci.- Soc. Amer. Proc. 17: 119-122, 1953. Foth, H., Carter, C., Guttay,'J. R., Robertson, L. 8., Shepherd, L. N., Wolcott, A. R., Davis, J. F., and Sundquist, W. Nitrogen Research 1958. ' Soil Science Department, Michigan State University, 1959. 27. Gainey, P. L. Total nitrogen as a factor influencing nitrate accumu- lation in soils. Soil Sci. 42:157-163, 1936. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 73 Goring, C. A. I. , and Clark, F. E. Influence of crOp growth on mineralization of nitrogen in soil. Soil Sci. Soc. Amer. Proc. 13:261-266, 1948. Gortner, R. A. , and Murrow, C. A. The organic matter of the soil V. A study of nitrogen distribution in different soils. Soil Sci. 3:297-331, 1917. Grogan, R. G. , and Zink, F. W. Fertilizer injury and its relation- ship to several previously described diseases of lettuce. Phyto- path. 46:416-422, 1956. Halvorson, A. R., and Caldwell, A. C. Factors affecting the nitrate producing power of some Minnesota soils. Soil Sci. Soc. Amer. Proc. 13:258-260, 1948. Harmsen, G. W. , and Van Schreven, D. A. Mineralization of organic nitrogen in soil. Adv. in Agronomy VII:293-398, 1955. Harpstead, M. I. , and Brage, B. L. Storage of soil samples and its effect upon the subsequent accumulation of nitrate nitrogen during controlled incubation. Soil Sci. Soc. Amer. Proc. 22:326-328, 1958. Hobson, R. P., and Page, H. J. Studies on the carbon-nitrogen cycles in soils. Jour. of Agr. Sci. 22:297, 497, 1932. Jackson, M. L. Soil Chemical Analysis. Prentice-Hall, Inc., Englewood Cliffs, N. J.: 194, 1958. Jans son, S. L. , and Clark, F. E. Losses of nitrogen during decompo- sition of plant materials in the presence of inorganic nitrogen. Soil Sci. Soc. Amer. Proc. 16:330-334, 1952. 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. Kaila, Armi. Biological absorption of phOSphorus. Soil Sci. 68: 279-289, 1949. Kamerman, P. , and Klintworth, H. Influence of fertilizer on the nitrogen and carbon cycles in soils. Union of South Africa, Department of Agri. Sci. Bull. 137:1-26, 1934. Kojima, R. T. - Soil organic nitrogen I. Nature of the organic nitrogen in muck soil from Geneva, New York. Soil Sci. 64:157-165, 1947. Kresge, C. B. , and Merkle, F. G. A Study of the validity of laboratory techniques in approaching the available nitrogen producing capacity of soil. Soil Sci. Soc. Amer. Proc. 21:516-521, 1957. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 74 Lohnis, F. Nitrogen availability of green manures. Soil Sci. 22: 253-290, 1926. Lyon, T. L., Bizzel, J. A., and Wilson, B. D. Depressive influ- ence of certain higher plants on the accumulation of nitrate in Soil. J. Am. Soc. Agr. 15:457-467, 1923. ‘ Lyon, T. L., Buckman, O. H., and Brady, N. C. The nature and properties of soils. New York, the Macmillan Company, 451, 1952. Marchal, E. The production of ammonia in the soil by microbes. Bull. acad. roy. Belge 25, 727-771, 1893. (Cited by Harmsen and Van Schreven, Adv. in Agronomy, V01. V11, 1955). Mattson, S. , and KouttlereAnderson,gE. . The acid-base conditionin vegetable litter and humus: VI. Ammonia fixation and humus nitrogen. Annals Agr. College Sweden 11:107-134, 1943. McCalla, T. M. The decomposition of Carex fil‘ifolia. Soil Sci. Soc. Amer. Proc. 13:284-286, 1948. Mehlich, Adolf, and Drake, Mack. Chemistry of the soil. Edited by Firman E. Bear. Reinhold Publishing Corporation, 430 Park Avenue, New York, pp. 286-327, 1958. Mevius, W. , and Dikussar, I. Nitrite als Stickstoffquellen fiir hbhere Pflanzen. Ein Beitrag zur Frage nach der Assimilation des Nitratstickstoffs, Jahrb, Wiss. Bot. , 73:633-703, 1930. (Cited by Miller, E. C. , Plant Physiology, MCGraw-Hill Book Company, Inc., New York, pp. 649-650, 1928). Michigan Agr. Exp. Sta. Fertilizer recommendations for Michigan crops. Extension Bull. E-159, 1959. Millar, H. C., Smith, F. B., and Brown, P. E. The rate of decompo- sition of various plant materials in soils. Jour. Amer. Soc. Agric. 28:914-923, 1936. Millar, H. C., Smith, ‘F‘. B. , .and Brown, P. E. The influence of organic matter on nitrate accumulation and the base exchange capacity of Dickinson fine sandy loam. J. Amer. Soc. Agr. 28: 856-866, 1936. Munson, R. D. and Stanford, ,G. Predicting nitrogen fertilizer needs for Iowa soils II. Evaluation of nitrate production as criteria of nitrogen availability. Jour. Soil Sci. Soc. Amer. Proc. 19: 464-468, 1955. Nightingale, G. T. The nitrOgen nutrition of green plants. Bot. Rev. 3:85-174, 1937. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 7.5 Norman, A. G. The biological decomposition of plant materials. VIII. The availability of the nitrogen of fungal tissues. Ann. Appl. Biol. 20:146-164, 1933. Panganiban, E. H. Temperature as a factor in nitrogen changes in soil. Jour. Amer. Soc. Agron. 17:1-31, 1925. Patrick, A. L. Influence of cr0p residue decay on soil nitrates. Soil Sci. 35:335-354, 1933. 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. 24:205-209, 1960. Pinck, L. A., Allison, F. E., and Gaddy'. V. L. The nitrogen requirement in the utilization of Carbonaceous residues in soil. J. Amer. Soc. Agr. 38:410-420, 1946. Plutnmer, J. K. Some effects of oxygen and carbon dioxide on nitrification and ammonification. New York, (Cornell) Agr. Exp. Sta. Bull. 384:305-330, 1916. Robertson, L. 3. ,Jr. A study of the effects of seven systems of cropping upon yields and soil structure. Amer. Soc. Sugar Beet. Tech.-Seventh Gen. Meet. 255264, 1952. ' Robertson, L. S. , Jr. A study of some effects of farming upon crop yields and soil structure (Ph. D. thesis, Department of Soil Science, Michigan State University, 1955). Rodrigues, G. Fixed ammonia in tropical soils. J. Soil Sci. 5:264-274, 1954. Rothwell, D. F., and Frederick, L. R. The decomposition of alfalfa and corn stover in soil: The effect of nitrogen on the loss of carbon at various temperatures. Agron. Abstr., pp. 21, 1956. Sabey, B. R., Bartholomew, W. V., and Shaw, R., and Pesek, J. Influence of temperature on nitrification in soils. Soil Sci. Soc. Amer. Proc. 20:357-360, 1956. Salter, R. M., and Green, T. C. Factors affecting the accumulation and loss of nitrOgen and organic carbon in cr0pped soils. J. Amer. Soc. Agron. 25:622-630, 1933. Saunder, D. H., Ellis, B. S., and Hall, A. Estimation of available nitrogen for advisory purposes in southern Rhodesia. Jour. Soil Sci. 8:301-312, 1957., Smith, G. E. Soil fertility and corn production. Missouri Agr. Exp. Sta. Bull. 583, 1952. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 76 .Sowden, F. J. Investigation on the amounts of hexoamines found in various soils and methods for their determination. Soil Sci. 88: 138-143, 1959. Starkey, R. L. Some observations on the decomposition of organic matter in soils. Soil Sci. 17:293-314, 1924. Stojanovic, B. J., and Alexander, M. Effect of inorganic nitrogen on nitrification. - Soil Sci. 86:208-215, 1958. Sundquist, W. B. An economic analysis of some controlled fertilizer input-output experiments in Michigan. (Ph. D. thesis, Department of Agricultural Economics, Michigan State University, 1957.) Sundquist, W. B., and Robertson, L. 8., Jr. An economic analysis of some controlled fertilizer input-output experiments in Michigan. Michigan Agr. Exp. Sta. Tech. Bull. 269, 1959. Tenney, F. G. , and Waksman, S. A. Composition of natural organic materials and their decomposition in soil. IV. The nature and rapidity of decomposition of the various organic complexes in different materials under aerobic condition. Soil Sci. 28:55-84, 1929. Tiedjens, V. A. , and Robbins, W. R. ’ The use of ammonia and nitrate nitrogen by certain crop plants. New Jersey Agr. Exp. Sta. Bull. 526:: 1-46, 1931. Truog, E. Technical information directions No. 697-18 for the Hellige-Truog Combination Soil Tester. Hellige, Inc. , New York. Turk, L. M. The effects of sawdust on plant growth. Michigan Agr. Exp. Sta. Bull. 26:10-22, 1943. Wallace, A. , and Mueller, R. T. Ammonium and nitrate nitrogen absorption from sand culture by rough lemon cuttings. Proc. Amer. Soc. Hort. Sci. 69:183-188, 1957. Waksman, S. A. , and Hutchings, I. J. The role of plant constituents in the preservation of nitrogen in the soil. Soil Sci. 40:487-497, 1936. Waksman, S. A., and Iyer, K. R. N. Contribution to our knowledge of the chemical nature and origin of humus: II. The influence of ”synthesized" humus compounds and of ”natural humus" upon microbiological processes. Soil Sci. 34:71-79, 1932. Waksman, S. A., and Iyer, K. R. N. III. The base exchange capacity of synthesized humus (ligno-protein) and natural humus complexes. Soil Sci. 36:57-67, 1933. 82. 83. 84. 85. 86. 87. 88. 89. 90. Waksman, S. A., and Iyer, K. R. N. IV. Fixation of proteins by lignins and formation of complexes resistant to microbial decompo- sition. Soil Sci. 36:69-82, 1933. Waksman, S. A. , and Tenney, F. G. The composition of natural organic materials and their decomposition in the soil: 11. Influence of age of plant upon the rapidity and nature of its decomposition-- rye plants. Soil Sci. 24:317-—333, 1927. Waksman, S. A. , and Tenney, F. G. Composition of natural organic materials and their decomposition in soils: I. Methods of quanti- tative analysis of plant materials. Soil Sci. 24:275—283, 1927. White, W. C. , and John Pesek. Nature of residual nitrogen in Iowa soils. Soil Sci. Soc. Amer. Proc. 23:39n42, 1959. Wiebe, J. Phytotoxicity as a result of heat treatment of soils. Proc. Amer. Soc. Hort. Sci. 72:331-338. 1958. Wolcott, A. R., and Johnston, H. H. Immobilization and mineralization in soil following organic amendment. Agron Abstr. p. 17., 1958. Wolcott, A. R. Soil and tissue tests for nitrate as guides for the use of fertilizer nitrogen for corn. Michigan Fert. Conf. Proc. Soil Sci. Department, Michigan State University, 1959. Woodruff, C. M. Estimating the nitrogen delivery of soil from the organic matter determination as reflected by Sanborn field. Soil Sci. Soc. AIner. Proc. 14:208-212,’ 1949. Wright, R. C. The influence of certain organic materials upon the transformation of soil nitrogen. J. Amer. Soc. Agr. 7:192-208, 1915. APPENDIX I Individual Obs e rvations 78 Table 13. --Residual ammonia and nitrate nitrogen one year after fertiu- lizer treatment and yield following a repeated application of. N, P and K. Fick farm, Kalamazoo sandy loam, 1955~56. Treatment Replication I Replication 11 Yield of Cats N P205 K20 NH3~N NO3-N NH3-N N03~»N Rep.I Rep. 11 Lbs/A Lbs/A Lbs/A Lbs/A Bus/A Bus/A 0 0 0 54 26 42 25 30.0 17.3 20 40 80 43 38 35 20 44. 3 65. 9 80 40 8O 31 24 35 23 65.4 78.8 320 40 80 - — - .. - - 20 160 80 42 45 44 39 75.2 65.6 80 160 80 43 31 37 20 47.5 51.3 320 160 80 81 36 51 28 59.2 73.1 20 640 80 42 25 45 19 42.8 59.1 80 640 80 40 22 36 20 68.3 76.2 320 640 80 57 31 49 28 78.6 66.0 20 640 320 33 26 38 19 65.6 55.6 80 640 320 36 40 35 36 86.4 82.2 320 640 320 51 45 51 42 81.2 80.0 80 Table 14. --Residual hydrolyzable and nonhydrolyzable nitrogen one year after fertilizer treatment. Fick farm, Kalamazoo sandy loam, 1955-56. Treatment Replication I Replication II Hydrolyz- Nonhydro~ Hydrolyz— Nonhydro- N P205 K20 able N lyzable N able N lyzable N Lbs/A Lbs/A Lbs/A Lbs/A 0 0 0 920 330 860 290 20 40 80 950 330 830 310 80 4O 80 870 300 1010 350 320 40 80 - - - .. 20 160 80 950 334 990 280 80 160 80 850 300 870 260 320 160 80 1620 640 780 270 20 640 80 880 280 970 270 80 640 80 1010 330 1000 310 320 640 80 810 470 936 260 20 640 320 1136 344 920 228 80 640 320 940 288 1150 234 320 640 320 990 320 1050 474 Table 15. --Residual inorganic and organic‘nitrogen one year after fertilizer treatment. - Fick farm, Kalamazoo sandy loam, 1955-56. Treatment Replication I Replication II N P205 KZO Inorganic Organic Inorganic Organic N N N N Lbs /A Lbs /A Lbs /A Lbs /A 0 0 0 80 1250 68 1150 20 40 80 81 1280 56 1140 80 40 80 55 1170 58 1360 320 40 80 - - - - 20 160 80 87 1284 83 1270 80 160 80 73 1150 58 1130 320 160 80 117 2260 79 1050 20 640 80 68 1160 64 1240 80 640 80 62 1340 57 1310 320 640 80 87 1280 77. 1196 20 640 320 60 1480 57 1148 80 640 320 76 1168 71 1384 320 640 320 96 1310 94 1524 82 treatment and yield follOW1ng a repeated application of N, P and K. Campbell farm, Kalamazoo sandy loam, 1955-56. Treatment Replication I Replication II Yield of Wheat N P205 K20 NH3"'N NO3‘N NH3'N NO3"N REP. I Rep. 11 Lbs/A Lbs/A Lbs/A Lbs/A Bus/A Bus/A 0 () 0 14 6 25 6 28.5 31.8 20 40 80 36 6 25 17 32.9 25.7 80 4O 80 25 25 25 17 38.0 31.8 320 40 80 22 28 14 28 35.0 35.2 20 160 80 29 8 l7 1 29.4 31.4 80 160 80 25 8 20 17 36.9 36.4 320 160 80 34 56 50 36 36.4 42.2 20 640 80 39 25 39 45 37.1 26.6 80 640 80 36 25 31 22 39.6 38.4 320 640 80 39 20 39 39 40.4 35.9 20 640 320 48 59 28 22 33.3 36.5 80 640 320 22 20 31 45 37.4 39.8 320 640 320 20 22 36 22 40.4 37.1 83 Table 17. --Residual hydrolyzable and nonhydrolyzable nitrogen one year after fertilizer treatment. Campbell farm, Kalamazoo sandy loam, 1955-56. Treatment Replication I Replication II N P205 KzO Hydrolyz- Nonhydro- Hydrolyz- Nonhydro- able N lvzable N able N lyzable N Lbs/A Lbs/A Lbs/A Lbs/A 0 0 0 730 196 970‘ -' 206 20 40 80 850 186 890 220 80 40 80 900 270 710 228 320 40 80 870 230 990 260 20 160 80 800 206 850 220 80 160 80 1000 310 890 260 320 160 80 870 228 950 310 20 640 80 950 248 790 220 80 640 80 780 260 930 220 320 640 80 930 218 860 206 20 640 320 900 230 1000 300 80 640 320 720 228 920 250 320 640 320 740 206 980 248 84 Table 18. --Residual total inorganic and organic nitrogen one year after fertilizer treatment. Campbell farm, Kalamazoo sandy loam, 1955-56. _1 Treatment Replication I Replication II N P205 KZO Inorganic N Organic N Inorganic N Organic N Lbs/A. Lbs/A Lbs/A Lbs/A 0 0 0 20 926 31 1176 20 40 80 42 1036 42 1110 80 40 80 50 1170. 42 938 320 40 80 50 1100 42 1250 20 160 80 37 1006 18 1070 80 160 80 33 1310 37 1150 320 160 80 90 1098 86 1260 20 640 80 64 1198 84 1010 80 640 80 61 1040 53 1150 320 640 80 59 1148 78 1066 20 640 320 107 1130 50 1300 80 640 320 42 948 76 1170 320 640 320 42 946 58 1228 85 Table 19. --Residual ammonia and nitrate nitrogen one year after fertilizer treatment and yield following a repeated application of N, P and K. Thompson farm, Sims loam, 1955-56. _:__ :- Treatment Replication Ifi Replication 11 Yield of Beans N P205 K20 NH3-N NO3-N NH3-N NO3-N Rep. 1 I Rep. 11 Lbs/A Lbs/A Lbs/A Lbs/A Bus/A Bus/A 0 0 0 2 12 89 32 7. 2 16. 3 20 40 20 73 31 70 26 18.6 29. 3 160 40 20 67 31 57 29 20. 0 27. 5 320 40 20 50 34 51 41 34. 0 30.7. 20 40 320 57 31 52 26 24.1 15.0 160 40 320 55 24 60 19 23.6 32. 0 320 40 320 47 58 45 26 20. 0 32.8 20 640 320 55 21 43 26 15. 3 14. 3 160 640 320 - - - - - - 320 640 320 22 33 70 12 29.4 34.0 Table 20. -—Residual hydrolyzable and nonhydrolyzable nitrogen one year after fertilizer treatment. Thompson farm, Sims loam, 1955-56. Treatment Replication I Replication II N P305 KZO Hydrolyz4 Nonhydroe Hy‘drolyz 1- Nonhydro- able N lyzable N able N lyzable N Lbs /A Lbs /A Lbs /A Lbs /A 0 0 0 3190 1070 2420 1180 20 4O 20 2290 960 2520 1320 160 40 20 2320 1280 2530 1250 320 40 20 3210 1510 2370 1050 20 40 320 2960 1440 2330 1400 160 40 320 2940 1380 3350 1510 320 40 320 2690 1280 2980 1400 20 640 320 2620 1130 3220 1290 160 640 320 - - - r 320 640 320 2760 1190 3100 1410 87 Table 21. --Residual inorganic and organic nitrogen one year after fertilizer treatment. Thompson farm, Sims loam, 1955-56. ‘1 t 1 it h Yield of Treatment Replication I Replication II Inorganic Organic Inorganic Organic Beans N P205 K30 N N N N Rep. I Rep. 11 Lbs/A Lbs/A Lbs/A Lbs/A Bus/A Bus/A 0 0 0 14 4260 121 3600 7. 2 16. 3 20 40 20 103 3250 96 3840 18.6 29.3 160 40 20 98 3600 86 3780 20. 0 27. 5 320 40 20 83 4720 92 3420 34. 0 30. 7 20 40 320 87 4400 78 3730 24.1 15. 0 160 40 320 78 4320 79 4860 23.6 32. 0 320 40 .320 105 3970 70 4380 19. 9 32. 8 20 640 320 76 3750 69 4510 15. 3 14. 3 160 640 320 - - - - - " 320 640 320 55 3950 82 4510 29.4 34. 0 88 Table 22. --Ammonia nitrogen in Sims clay loam as affected by organic amendments, and fertilizer nitrogen. Treatment _ ,7- Replication fl 1 21- 3 4 5 Average Pounds of NH3~N per acre Check ........ 83 95 89 ' 92 98 91 Check plus N . . . . 98 106 94 103 104 101 Alfalfa-brome . . . 100 100 91 98 92 96 Alfalfa-brome + N . 115 109 112 100 103 108 Sawdust. . . . . . . 89 80 89 86 89 87 Sawdust plus N . . . 92 100 103 92 94 96 Straw. . . . . . . . 92 92 92 81 92 90 Straw plus N . . . . 98 100' 101 89 98 97 Main effects of residues: Check ..................... 96 Alfalfa- brome . ............ 102 Sawdust . . . . .9 ............... 91 Straw ..................... 94 Main effects of nitrogen: No nitrogen ....... . . . . 91 101 Table 23. --Nitrate nitrogen in Sims clay loam as affecte amendments, and fertilizer nitrogen. d by organic 89 Treatment _ Replication fi 1 2 3 4 5 Average Pounds of NO3wN per acre Check ........ 51 48 63 51 45 52 Check plus N . . . . 54 54 74 55 55 58 Alfalfa-brome. . . . 68 68 71 63 66 67 Alfalfa-brome + N . 74 77 79 73 76 76 Sawdust ...... . 49 51 42 .57 57 51 Sawdust plus N . . . 71 71 66 71 63 68 Straw ....... . . 57 54 48 6O 54 55 Straw plus N . . . . 61 61 57 66 60 61 Main effects of residues: Check ................... 55 Alfalfa-brome ............... 72 Sawdust .................. 60 Straw .................... 58 Main effects of nitrogen: No nitrogen ................ 56 66 Supplemental nitrogen .......... Table 24. ---Inorganic nitrogen in Sims clay loam as affected by organic amendments, and fertilizer nitrogen." ' Treatment ’ Replication 1 2 3 4 5 Average Pounds of inorganic nitrogen per acre Check ...... 134 143 152 143 143 143 Check plus N . . 152 160’ i 168 158 159 159 Alfalfa-brome. . 168 t 160 162 161 ' ' 158 173 Alfalfa-brome + N 189 186 191 173 179 184 Sawdust. . . . . 138 131 131 143 146 138 Sawdust+ N. . . 163 .171 169 163 157 165 Straw . ..... 149 146 140 141 146 144 Straw plus N . . 159 161 ' 9158 155 158 158 Main effects of residues: Check . .................. 151 Alfalfa- brome .............. I 17 3 Sawdust ......... i ......... 151 Straw ................... 151 ' Main effects of nitrogen: No nitrogen ................ 147 Supplemental nitrogen . . ........ 166 Table 25. --Hydrolyzab1e nitrogen in Sims clay loam as affected by organic amendments, and fértilizer nitrogen.‘ Treatment Replication l 2 3 4 5 Average Pounds of hydrolyzable nitrogen per acre. Check . ....... 3100 3156 2712 3276 3192 3087 Check plus N . . . . [3348 3456 3.312 3216 3168 3300 Alfalfa—brome 3144 3216 3168 3184 3192 3221 Alfalfa-brome + N 3384 3240 3288 3456 3336 3341 Sawdust ..... 3456 3288 3216 3432 3576 3394 Sawdust plus N . 3552 3528 3624 3744 3624 3614 Straw . . . . . 3192 3144 3168 3216 3240 3192 Straw plus N . 3240 3264 3288 3312 3552 3331 Main effects of residues: 1 Check .................. 3194 Alfalfa-brome ............. 3281 Sawdust ................. 3504 Straw ........ . ...... 3262 Main effects of nitrogen: No nitrogen ............... 3223 Supplemental nitrogen ........ 3397 Table 26. --Nonhydrolyzab1e nitrogen in Sims clay loam as affected by organic amendments, and fertilizer nitrogen. Treatment Replication l 2 3 4 5 Average ' Pounds of nonhydrolyzable nitrogen per acre Check ....... 756 600 780 492 565 638 Check plus N . . . 888 1068 1188 1044 1176 1073 Alfalfa-brome. . . 888 996 1056 984 780 941 Alfalfa-brome + N 1212 1224 1368 1248 816 1174 Sawdust ...... 600 816 960 960 528 773 Sawdust+ N . . .. 816 960 1152‘ 1140 1128 1039 Straw . ...... 768 864 984 696 648 792 Straw + N ..... 1008 1128 1080 1152 840 1042 Main effects of residues: Check ................. 8.56 Alfalfa-brome ............. 1057 Sawdust ................ 906 Straw ................. 917 Main effects of nitrogen: ' No nitrogen .............. 786 1082 Supplemental nitrogen . ' ....... Table 27. --Organic nitrogen in Sims clay loam as affected bv organic Treatment 4 Replication ‘ ‘ 1 2 ' 3 4 5 Average Pounds of organic nitrogen per acre Check ...... 3856. 3756 3492 3768 3756 3726 Check + N 4236 4524 4500 4260 4344 4373 Alfalfa-brome. . 4032 4212 4224 4368 3972 4162 Alfalfa4brome+N 4596 4464 4656 4704 4152 4514 Sawdust ..... 4056 4104 4176 - 4392 4104 4166 Sawdust + N w 4368 4488 4776 4884 4752 4654 Straw ...... 3960 4008 4152 3912 3888 3984 Straw + N 4248 4392 4368 4464 4392 4392 Maln effects of re81due Check .................. 4049 Alfalfa- brome .............. 4338 Sawdust. . ............. 4410 Straw........ ........ 4178 Main effectsof nitrogen: No nitrogen. . . ...... A ...... 4009 4478 Q/ «.4: Na: «.9: 4.4: 9m: 2 634 385.... Tom ~.m: 0.82 ~64: Na: 383m ad: N43 To: 4.4: m6: 2 63a 853% ACS: ed: TM: “in: So: peeve/mm 1m: 4.4:, 66: m8: To: z 63.2 6865823324 0.22 4.4: 0.2: ad: 903 6839-823? Tm: 0.32 50:. Na: Na: 2 .34 026666 meg: we: 0.52, Mia: ad: .325 <\.mdm <\.mdm <\.mdm <\.msm <\.msm > soflmoflmom >H sofldofimom EH dogwoflmom E coEmUSmom H definefimom «soaumouh. .Gowoufid Houflflhow paw .muaogpsmgm oficwmho .3. pooconflafi mm Emofl >30 mcfim a“ mgofin CHOU-.. .mm 3nt a... 4m 06 em mm 3 mm 3 2 firm 6 em 8 2 pm mm 68 2. S e .2me e S. 2. mm K op 66 i. E z 364 e S. pm pm we we 3 3 me o 33 6 m4 .3 N... mm 3 E S. me 2 swam a me am 6m 6w MN em 3 S o swam H S . 2 e... 2: an E 3 am 2 364 2 en 3. B we Na. K E B o 364 2 «<23 «<34 «<85 «\24 «<83 $.83 «\34 $63 z-.oz z-.mz zkoz ztmz z-.oz z...mz z-.oz 77.52 eemogz 53.8.2 EH 283.83%wa HH noflmUanm H :oflmoflaom Eofifldonh soflmuom >H normenmem 1! $208305 Comofiwm HmunoaofimQSm was 35.3 333.8“ .oocosvom mEQQOMU >9 poososdcw mm £33 330 mcfim 5 somoufis 3.933 was 32208515...- .om 03.68 96 8.: 8mm moi: 3o... 3m $3 $2 omen z swam 6 6mm omam m3 3.; see 3.3 we» 83 o :3: e .3; cam... 0% 8mm 63 3% «we em: 2 364 6 4mm 83 3m 3.3 3.2 gm... ooe ommm o 364 e 82 62m 3: 33 2:: spam 2.3 83. z :3: 2 NE 33 eme comm «we 88 43 onm o swam H 2:. 38 46m 3mm 82 omen $2 ommm z 364 2 «cm 82 um: 33 com ommm new 3.3 o 364 a .135 $63 «\BA «\63 «\24 «\BA «\63 <83 zmz zm zmz zm zmz zm zmz zm eemoeaz 33:6,.“ >H coflmoHHm0m , HHH GOEMUHHQEH HH :0300HHQ0MH H :oflmoHHm0m 020580008 son—30m _ #2208300: :0moufic H0HG0S0HanHdm H05... H0>0H 433393 030.300 mcamouo .3 H00oc03fi: mm 503 >30 08$ 5 G0woufin 0H£0N>Honc>€oc Head 0H£0N>HOHH0>H+1 .om 030M. 97 83 £9 meow $9 $2 9% ~89 $9 2 @999 o ommm 099 momm 999 omwm 999 moom «N9 0 awn» e 399 $9 83 o: 2?. £9 3:7 9:9 z 264 o vamm w~9 7079 cm «omw ~99 omwm 799 o \76,9. 0 «ovw 099 wwmv «m9 wwom N7 ow~¢ No 79 gmnm 9 Nm9¢ 9m9 om9w N99 «77m oo9 779* mm o 9mm9 9 comm NNH vmwm QMH N93» m9: vmov wNH Z 304 9 womm ~99 mmwv o99 ow9¢ M99 N977 mm9 o 9504” 9 1991\83 «\319 $34 «\24 $83 $319 «>34 3319 79 79 79 79 79 79 79 79 UHGGMHO HNHOGHE UEdeO HNHQCME UfiGMWHO HGHQGME UEMMHO HNHQQME SQwOHHwZ knumeuhmnH . >H qoflmoflmmmw 14H: 990393399979 3 coflmoflmom H noflmofimmm pamfiummufi 9903339 .9988”.de ammonia Hmucmamfimmfim mam H973 9» .89 .mommswmm mcamouo 99b, wmoqmsmg mm 8809 >30 mawm 999 ammonia Umnmmno v9.8 Haumfizun Am 039mm. 99999 . 98 ox: 9.29 cam 99.3. 2 99399 o o.mm m.om numv H.Nm o 33$ 0 o.mm o.ooH oil. «1:. Z 304 o ¢.¢m m.wo m.9> o.mv o 304 o $.09: «.02 New: m.moH 7H fimHHH‘ H. m.m© ogHNH m.NHH Memo o :33 H 9..on @391. wxwo m6: Z 304 H NINOH mums néoH w.ooH o 304 H <\ .mde <\ .mfim <\ .mfimH <\ .man Gmwoufiz 3HH390h >H :oBmUHHnHmm HHH coHumUHHmmm HH 99039539va H nofidoHHmmm 9990839799. Coflmuom muflm. H®>®H knunfivh ow . vamfifimmha Gmwoufic HdquEmHmmdm “mocmwvmm muamouu 93 Uvudwsmfi mm EmoH >30 mEHm 99H mHuHmTH 99.80:: .mm mHndH APPENDIX 11 Computational Formulae Used in Functional Analyses 99 100 Computational Formulae Used in Functional Analyses N = Number of observations N' = Number of observations EX = Arithmetic sum of observations x 3 number Of variables or M-M (where M = number 2X2 2 Sum of squares of x of variables) 3(- = Mean of x's i2 = Mean of x2 7‘2 = Variance = (i2 - i'X) 7" = Standard deviation = m ZXIXZ = Sum of products of xlxz (cross products) XIX; = Mean of cross products R 2' Coefficient of multiple correlation E = Corrected R (also c R) _____ __ __ r1 2 = Correlation coefficient : XIX; - XIX; or NE X1X2 - XXL—EX} “'1 ”7 JINlez-(lefflNzxzz-(EXM Formulae for simple correlation (2 variables) A Y: a + bx y 2 dependent variable x 1’ independent variable ZXY - N32"? — - ‘37 - :57 b = I a = y - xb r = -—,.———-— 23le - N‘fi? d-x .0; S = standarderror of estimate 2 0'3, 1 - r 6— r r\/ N—Z (I: = standard error of b = ?'_l l - 1‘ t1. = or -'-———- X N a; N} 1 - r2 1 - r'2 b 0?: standard error of r = L— tb — E N} N - 2 Multiple correlation by formulae (3 variables - with x1 dependent) "1 (1'1") b-0(071 : 2' 113.23 .. 12" 2 Beta ([3) COfoiCients ‘32 1 _ r231 b coeff1c1ents (I? ' b3=03 < ”T 1 £3 = 1:13 ' (@1253) UT 3 1 " 1‘232 ~ 8.3321 - 132;; ' b3§3 2 _ N[—7- —"‘ 2 N-l R-Bzrxz+f33r13 R: R Rzl-(l-R)(W (E; :433 : stan. error of 0 coeff. : J_ 1 _ R2 . (ME...) (N-M) £1732 : stan. error Ofb : J/EZ ( fl ) 3 G3 :JE (1T);tb= same as above 0’3 f3” Residuals (after computing §r\= a + b; x, + b2 x2) Substitute computed values of a, b; and b; in equation along with original x1, x2 for each sample observation and get the predicted value of y, say )3). Subtract this from original V for that observation, to get residual. Suggested form below - - A ‘ AI Predic. equa: y = 5.0260 + .0738 x, + 1. 2867 x; , y - y = Residual :5: X1 X2 Sample A bxl b X; 9 y Residual' No. 3 10 1 5.0260 .2214 12.8670 18.1144 16.7345 4.3799 5 8 2 5.0260 .3690 10.2936 13.6886 17.4497 1.7611 1 5 3 5.0260 .0738 6.4335 11.5333 11.1521 - .3812 a): Residuals should total 0. If logs were used, antilog of y must be found before getting residuals. 102 Computation of Prediction Equation by "Least Squares" The inverse is computed omitting the dependent variable from the Identity and from the Inversion. The Raw Moments are first Augmented and Adjusted. Xi = Iidependgrg variables ii M321 of Xi Y = prendeit variable {1'- Mi of Y Ki = Adjustment factor Ki' (Deadjustment factor) is Iii;— Y Cii are the diagonal elements of the inverse. -1 bi = (Mix) Miy Where (Mix)"1 is inverse of the i row of the moments of all the x's. a=y-Zbi;i R2: Zbimiy E‘y R2 VRZ Where miy are the moments of all the Xi on Yi° 1 xi. H 1 - (l - R2) (FBI-LIV?) Where N is the number of observations and .. M the total number of variables. i = G?“ -7- ?- _ . . S = Z I}; jblmly— In computing R2, either adjusted or - deadjusted figures may be used as long as both the Equation and the b's are either S = «j -—Z adjusted or deadjusted. For S the adjusted figures must be used. For "a" the deadjusted figures are used with the means, which have not been adjusted. 0731 (Standard error of bi) = NT) (S) Prediction Equation: y = a+bx2+bx3+bx4+bx5 t- bi 1 FE Example ‘9==86.6182525764—9.87479703x2+-6.7076962x3- 14.2685815x,- 13.7574111x5 (6.86534778) (12.0923398) (12.2988143) (8.3785081) t: 1.438353503 .554706228 1.160159113 1.641988160 APPENDIX III Locations of Field Experiments and Soil Type Descriptions 103 104 Locations of Field Experiments Fe rtilizer experiments: Ewald Fick farm: Section 34, T Z S, R 7 W; Calhoun County. (Kalamazoo sandy loam). John Campbell farm: Section 4, T 2 S, R 10 W; Kalamazoo County. (Kalamazoo sandy loam). Kenneth Thompson farm: Section 4, T 10 N, R 2 W; Gratiot County. (Sims loam). Residue and rotation experiments: Lee Ferden farm: Section 33, T 9 N, R 3 E; Saginaw County. (Sims clay loam). Soil Type Descriptions Kalamazoo soils: The Kalamazoo series includes well-drained Grey-Brown Podzolic soils deve10ped on acid loam, sandy loam, and loamy sand materials over calcareous or neutral sands and gravel at depths of 42 to 66 inches. They have a sandy clay loam to clay loam subsoil between 10 and 20 inches thick. Sims soils: The Sims series includes naturally poorly drained soils developed in calcareous clay loam or silty clay loam till. They need tile drainage for c rop production.