CHANGES IN THE ORGANO-MINERAL COMPLEX 0F SOILS DUE TO MANURE APPLICATIONS ‘ Dissertation for the Degree of PILD. MICHIGAN STA’FE UNIVERSIIY KENNETH WAYNE .LINVILLE 1974? . . 9“,». 'V‘P‘I‘. ;. .3 3 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII L 3 3:131 :33; 3 {f3 3V 3. . A ABSTRACT CHANGES IN THE ORGANO—MINERAL COMPLEX OF SOILS DUE TO MANURE APPLICATIONS By Kenneth Wayne Linvillel Profile distributions of N and P were examined in soils which had received nine annual applications of beef cattle manure at moderate rates or single applications atomassive rates. Surface and subsoil samples from selected treatments were subjected to two different extraction sequences in— volving 1.7% KCl, 0.1N H280“, and 0.1N NaOH or 0.1M NauP207. The extracts were analyzed for total N and, after perchloric acid digestion, for Ca, Mg, Fe and Al. After nine annual applications at 10, 20 and 30 T/a, more organic N was retained in the plow layer where continu- ous corn was harvested for grain and stover was returned than where silage was removed. Downward movement of organic N was detected to depths of A to 6 ft and was greater where silage was removed. Ammonium in subsoils was greater with silage harvest, whereas nitrate in the plow layer was greater where only grain was removed. Extractable P in the plow layer was less with manure treatments than where high rates of fertilizer P had been used, but downward movement was greater with manure and was detectable to 15 inches. Kenneth Wayne Linville Eight months after single applications of 100, 200 and 300 T/a, essentially all of the input N could be accounted for in the plow layer, although some downward movement to depths up to 18 inches was detected. About 16% of input P could be accounted for as extractable in the plow layer, with no evidence for downward movement. Fifteen percent or less of input Ca and Mg could be accounted for in exchangeable forms in the plow layer; whereas nonexchangeable but extractable forms were equivalent to 90% of input or more. Most of the nonexchangeable Ca and Mg appeared to be present in mineral salts or complexes and was associated with Fe and Al in H280“ extracts. Neverthe- less, all cations were present in organo-mineral complexes which contained 27 to 58% of total soil N and were extract- able with NaOH or pyrophosphate. The quantities present in complexes with N were, in general, related directly to valence and inversely to ionic radius of the metal cation: Ca < Mg < Fe < Al. Wide fulvic/humic recovery ratios for Ca in NaOH extracts of the plow layer indicate that Ca was the cation mainly responsible for precipitating fulvic acids and stabiliz— ing N in the fulvic acid of the upper profile. Total non— exchangeable Ca and N in the plow layer were highly corre- lated (r = .84, P = .01). In the subsoil, Al appeared to be the principal stabilizing cation. Iron in NaOH extracts was associated mainly with the humic acid fraction. To a lesser extent, this was true also for Mg. CHANGES IN THE ORGANO-MINERAL COMPLEX OF SOILS DUE TO MANURE APPLICATIONS By Kenneth Wayne Linville A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences 197“ ACKNOWLEDGMENTS This investigation was supported by research funds from The Michigan State University Agriculture Experiment Station, to which the author is indebted. Gratitude is extended, by the author, to Dr. Boyd G. Ellis and Dr. Arthur R. Wolcott for suggesting the problem, for guidance, for encouragement, and for the personal interest shown. Appreciation is expressed to Dr. Donald R. Christenson, Dr. Henry D. Foth, Dr. Maurice L. Vitosh, Dr. Stephen L. Yelon, and members of the Crop and Soil Sciences Depart— ment for their assistance. A special Thank You is conveyed to Mrs. Elizabeth Shields, Dr. Thomas Schueneman, Mrs. Susie Barkyoumb, Mr. Randy Shantz, A. D. Ruedemann, M.D., and Rev. Lawrence Delaney. ii TABLE OF CONTENTS CHAPTER Page I. INTRODUCTION................................... 1 II. LITERATURE REVIEW.............................. A Surface and Ground Water..................... A Manure Additions to Soils.................... 6 Organo—Mineral Complexes..................... 9 Methods for Extracting Soil Organic Matter... 11 III. EXPERIMENTAL PROCEDURES AND METHODS............ 1“ Selection and Description of Research Sites.. 1A Experiment I.............................. 1A Experiment II............................. 17 Description of soils...................... 1? Sampling Procedures.......................... 18 Laboratory Methods of Analysis............... 19 Sequential extractions.................... l9 1. KCl (Seq. I-a and II-a)............. 21 2. H280“ (Seq. I-b).................... 23 3. Nauon7 (Seq. I-c).................. 23 N. NaOH (Seq. II—b).................... 2A 5. Fulvic acid (Seq. I—d and II-c)..... 2A 6. Humic acid (Seq. I-e and II—d)...... 25 7. Extended series, H SO“ extract (seq. I-g and II-f 000.0000000000000 25 8. Extended series, NaOH extract (seQ0 1-8 and II-f)00000000000000000 26 Perchloric acid digestions................ 26 iii CHAPTER Page 1. Procedure for Ca, Mg and Fe...... 27 2. Procedure for A1................. 27 Chemical Analysis...................... 28 l. Nitrogen......................... 28 2. Carbon...... ..... ................ 28 3. Phosphorus....................... 28 A. Calcium and magnesium............ 29 . Iron............................. 29 AluminumOOOOOOOOOOOOOOOO000...... 29 N ON U1 0 0 pH and buffer I)HO O O O O O O O I I O O O O O O 0 30 Statistical Methods....................... 30 V. RESULTS AND DISCUSSION. 0 O O O O O O C O O O O O O O O O O O O O 32 Influence of Manure and Fertilizers on the 8011 Environment................... 32 Experiment I.............................. 32 Distribution of Soil N................. 33 Distribution of Soil P................. 36 Experiment II............................. 35 Distribution of Soil N................. 36 Distribution of Soil P................. A0 Sequential Extraction of Organic Matter and Associated Metals.............. AO Exchangeable Forms of N and Metals...... A2 Nonexchangeable Forms of N and Metals... AA 1. Overall recoveries................ AA 2. Fractional recoveries............. A7 1v CHAPTER Page 3. Interactions among N and metals.. 53 A. Regression of N on metals........ 61 5. Fulvic acid/humic acid separa- t10n800000000.0000000000000000000 65 6. Fulvic/humic ratios.............. 68 7. Carbon/nitrogen relationships.... 73 V. SUMMARY AND CONCLUSIONS..................... 80 BIBLIOGRAPHY...................................... 86 APPENDIXOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.00.0.0... 91 Table LIST OF TABLES Page Manure analysis and approximate loading..... 16 Hodunk sandy loam profile exposed at the site of Experiments I (Rep 1) and II.... 19 Summary of nitrogen forms recovered in the plow layer and in subsoils to 10 feet after nine annual applications of manure or commercial fertilizers for corn........................................ 33 Distribution of extractable phosphorus (Bray P-l) with depth in the profile after nine annual applications of manure or com- mercial fertilizers for corn................ 37 Summary of nitrogen forms recovered in the plow layer and in subsoils to five feet eight months after a single applica- tion of manure at high rates for corn harvested as grain.......................... 38 Distribution of extractable phosphorus (Bray P-l) with depth in the profile eight months after a single application of manure at high rates for corn...................... Al Recoveries of exchangeable N and metals in 1.7% KCl.................................... A3 vi Table 10 ll l2 13 1A 15 Page Total recoveries of nonexchangeable N and metals by sequential extraction...... A5 Simple correlations (r) between totals for nonexchangeable N and metals re- covered in extraction sequences I and II... 51 Simple correlations (r) between N and metals in sequential extracts (through humic acid) in the plow layer (0-9 in.).... SA Simple correlations (r) between N and metals in extended sequential extracts of plow layer (0-9 in.).................... 55 Simple correlations (r) between N and metals in sequential extracts (through humic acid) in the subsoil layer (18- 21 in.).................................... 56 Simple correlations (r) between N and metals in extended extracts of the subsoil layer (18-21 in.).................. 57 Optimal solution's for linear regres- sions of N on metals in extracts of Sequence 1................................. 62 Optimal solution's for linear regres- sions of N on metals in extracts of Sequence II.O.00....O....OOOOOOOOOOCCOOOOOO 63 vii Table Page 16 Comparisons of summed recoveries in fulvic and humic fractions with analysis made directly on pyrophosphate or sodium hydroxide extracts.......................... 66 17 Fulvic/humic recovery ratios for extract c (0.1M NauP2O7 + 0.1N NaOH), Sequence I... 7O 18 Fulvic/humic recovery ratios for extract b (0.1N NaOH), Sequence II................. 71 19 Nitrogen recoveries in original soil and in residual soil after final NaOH extraction................................. 7A 20 Carbon recoveries in original soil and in residual soil before and after final NaOH extraction............................ 75 21 Percent of original soil nitrogen and 0 carbon removed by sequential extractions... 76 22 C/N ratios in original soil, extracted fractions, and in residual soil............ 78 23 Distribution of organic-N with depth in the profile nine years after annual applications of manure or commercial fertilizers for corn harvested as grain.... 91 24 Distribution of organic-N with depth in the profile nine years after annual ap- plications of manure or commercial ferti- lizer for corn harvested as silage.......... 92 viii Table 25 26 27 28 29 30 Page Distribution of ammonium-N with depth in the profile nine years after annual applica- tions of manure or commercial fertilizer for corn harvested as grain................ 93 Distribution of ammonium-N with depth in the profile nine years after annual applica- tions of manure or commercial fertilizer for corn harvested as silage.................... 9A Distribution of nitrate-N with depth in the profile nine years after annual applica- tions of manure or commercial fertilizers for corn harvested as grain................. 95 Distribution of nitrate-N with depth in the profile nine years after annual applica- tions of manure or commercial fertilizers for corn harvested as silage............... 96 Distribution of total mineral-N (NHu-N+NO3-N) with depth in the profile nine years after annual applications of manure or commercial fertilizers for corn harvested as grain.... 97 Distribution of total mineral N (NHu—N+NO3-N) with depth in the profile nine years after annual applications of manure or commercial fertilizers for corn harvested as silage.................... 98 ix Table Page 31 Distribution of organic-N with depth in the profile eight months after a single application of manure at high rates for corn....................................... 99 32 Distribution of ammonium-N with depth in the profile eight months after a single application of manure at high rates for corn....................................... 100 33 Distribution of nitrate-N with depth in the profile eight months after a single application of manure at high rates for corn....................................... 101 3A Distribution of total mineral-N ‘(NHu-N + NO3-N) with depth in the profile eight months after a single application of manure at high rates for corn.................................. 102 35 Recoveries of N and metals in extract b (0.1N H280”), Sequence I................ 103 36 Recoveries of N and metals in extract c (0.1M NauP2O7 + 0.1N NaOH), Sequence I......................................... 10“ 37 Recoveries of N and metals in extract b (0.1N NaOH) Sequence II................. 105 38 Recoveries of N and metals in extract f (001E H280“), sequence I0000000000000000 106 X Table 39 A0 A1 A2 “3 AA “5 A6 Page Recoveries of N and metals in extract e (0.1N H280“), Sequence II............... 107 Recoveries of N and metals in extract g (0.1N NaOH), Sequence I................. 108 Recoveries of N and metals in extract f (0.1N NaOH), Sequence II................ 109 Recoveries of N and metals in extract d (Fulvic Acid), Sequence I............... 110 Recoveries of N and metals in extract e (Humic Acid), Sequence I................ 111 Recoveries of N and metals in extract c (Fulvic Acid), Sequence II.............. 112 Recoveries of N and metals in extract d (Humic acid), Sequence II............... 113 Total recoveries of nonexchangeable N and metals in two parallel sequential extraction sequences...................... 11” xi Figure LIST OF FIGURES Page Flow diagram for extraction Sequences I and II....................... 22 Distributions of the total re- coveries of nonexchangeable N and metals by sequential extraction.......... A8 Distributions of the recoveries of N and metals in sequential extractions of the plow layer (0-9 in.)..................................... A9 Distributions of the recoveries of N and metals in sequential extractions of the subsoil layer (18-21 in.)......... 50 Flow diagram for extraction Sequences.... 85 xii CHAPTER I INTRODUCTION Soil organic-N and N fixed by legumes or recycled through crop residues and barnyard manures were the major sources of nitrogen for crop production until the end of - World War II. After the war, industrially fixed N became available in large quantities at a relatively cheap price. At present, higher levels of crop production are being maintained by liberal use of N from industrial sources, balanced with other fertilizer nutrients. Numerous agronomists have estimated that commercial fertilizers now account for one third of the total food production in the United States and predict that the quantity of fertilizer used will increase as population grows. Increased human populations have also affected the livestock industry. The demand for more meat has resulted in the development of large feedlots. With this increase in animal populations, concentrated in small areas, have come the problems of disposing of large quantities of animal waste and increased potential for pollution of streams and ground water. Metropolitan areas are also faced with problems of waste disposal. Under recent Federal legislation, state and local agencies are imposing increasingly rigorous requirements for renovation of water in waste effluents before discharge into surface waterways. They too are now looking to the soil as a possible medium for treating waste water and for disposing of solid wastes. In the context of water treatment or waste disposal, the tendency is to go to very high application rates because of high costs for land and distribution facilities. It is recognized that excessive rates can exceed the capacity of soils and crops to remove N and that the excess N will appear very quickly as nitrate in drainage or ground water. Most soils have a large capacity for removing soluble P from per- colating water, but this capacity can be saturated over a period of years at high rates of input, thus reducing the useful life of a waste disposal site. Potentially toxic industrial metals are retained effectively in many soils. Some of these metals are essential nutrients but, at high rates of input, available concentrations can build up over time to the point where plants are injured or the concentra- tions taken up into vegetation become toxic to livestock. Some sort of compromise must be made between exces- sively high rates of waste application, with emphasis on disposal, and lower rates which will permit much of the nitrogen to be removed into standing or harvested vegetation or by denitrification. With many wastes, even if application rates are regulated to control N, the inputs of P and other nutrients will still exceed, sometimes greatly, the quantities that can be removed by vegetation. As yet, there is insufficient data from long—term experiments to predidt reliably how long soils can remove P or metals before becoming saturated. Mechanisms whereby P and metals are retained in soils (exchange, adsorption, precipitation, complexation) have been studied extensively, but seldom under conditions where their effectiveness could be related to long-term inputs in the field. These considerations have prompted this investigation to study (1) the profile distributions of N and P and (2) the fractional distribution of N and cations in soils from field experiments involving long-term differential applica- tions of manure and commercial fertilizers or recent heavy applications of manure. CHAPTER II LITERATURE REVIEW Traditionally, industry has looked to agriculture as a potential market for waste products, many of which are worthless or harmful to plants. Fortunately, the waste from animal feeding can be utilized in crop production. As large amounts of feedlot manures or fluid materials are applied to valuable crop land, two important considerations must be taken into account: (1) the impact of these additions on surface and ground waters and (2) the composition of waste materials and their soil reactions, since yields and crop composition may be adversely affected by improper use. Surface and Ground Water High concentrations of NO in water consumed by infants 3 or livestock is of direct concern because of methemoglobinemia. Nitrate may be harmful to ruminants. When consumed, this ion may be reduced to nitrite, enter the blood stream (Garmer 1973, Wright 196A), and tie up the oxidative mechanism of the blood (Winter 1962). Drinking water standards of the U. S. Public Health Service (1962) list ten ppm NO3-N (A5 ppm N03) as the level that should not be exceeded for infants. Investigation by Smith (1965) of 6,000 rural water supplies in Missouri showed that animal wastes, improperly constructed shallow wells and septic tank drainage are the main sources of contaminating N03. His investigation revealed that livestock was a more important source of contamination in improperly constructed wells than N fertilizers. These findings are substantiated by Stewart gt a1. (1967) by work done in Colorado. To determine the effect of fertilizer N on contamina- tion of ground water Linville and Smith (1971) studied five soil types in Missouri treated with various rates of nitro- genous fertilizers and under continuous culture of corn. ’ITheir results indicated that N additions which exceeded N removal increased the potential for downward movement of ‘ No3. These results are substantiated by Adriano 33 a1- (1972) in a recent study of nine row-crop sites in California. In another study in California, Lund gt_al. (197A) found significant correlation between soil profile characteristics and average N03 concentration below the root zone. He recom- mends that soil profile characteristics be considered in selecting land for high N input through fertilization or waste disposal. A detailed discussion of the fate ofIN in soils, water and plants has been given by Viets and Hageman (1971). Eutrophication is a term applied to the situation where an increase in nutrients, notably P or N03,in surface waters stimulates the growth of aquatic plants. This leads to excessive production followed by depletion of 02 due to decomposition of large masses of dead plant materials. In lakes, ponds or slow moving streams, a bog or swamp will develop eventually. Engineers responsible for producing potable water for domestic use are concerned about N and P concentrations in water that stimulate this aquatic growth. The most likely processes that can lead to P contamination from the soil-plant system are surface runoff and erosion. Any P-containing particles of soil, organic matter or manure present on the surface are subject to being washed away unless adequate erosion control practices are followed. Organic or mineral forms of N are also subject to runoff and erosion. Nitrate can move into surface waters by seepage or through agricultural drainage systems. Concentrations of P in the water are usually more crit- ical than N03 in eutrophication. Phosphate content in fresh water that will limit the growth of aquatic plants is about 0.02 ppm and, for N03, it is from 0.05 to 1.0 ppm (Taylor 1967, Smith 1968). Therefore, the levels of nutrients that will restrict eutrophication present more critical limits for agricultural and waste disposal operations than do the higher levels of N03 that can be tolerated in water for animal or infant consumption. Manure Additions to Soils Numerous investigations have been conducted on the effects of manure additions on soils. Massachusetts research- ers report no economic advantage from manure applications in excess of 20 tons per acre (Weeks gt gt, 1972). A Michigan study indicates ten tons per acre was the most favorable rate for corn grain on a sandy loam soil (Vitosh gt gt. 1973). High salt content of the soil solution or toxic concentrations of NH3-N in the soil atmosphere, or both, affected germination on barley, sudangrass, spinach and radish in a greenhouse study (Adriano gt gt. 1973). Murphy reports depressed yields due to salt injury to corn grown on soils receiving large additions of solid beef feedlot manure. He found significant accumulations of N03-N, total N, and available P in the soil. The high soil concentrations of P were reflected in high P values in corn silage (Murphy gt gt. 1972). An extensive review of the literature and analysis of research needs has been conducted (Powers gt gt, 197A). They concluded that there is a large volume of research data available on short- term effects of applying animal waste to land, but there is little information available on long-term effects of animal wastes on physical, chemical and biological properties of soils. There are few guidelines for calculating animal waste loading rates. Results from a recently completed study suggest that housing type in feedlots can modify climatic influence and conserve nutrients from loss by evaporation, runoff and leaching (Adriano gt 1. 197A), thereby increas— ing the quantities which must be managed in ways that will give more positive control over their movement through the environment. It is known that N can be removed by denitrification in the presence of plant residues and root exudates under conditions of low oxygen tension which occur during water- logging or even in microhabitats in well-drained soils (Allison 1973, Broadbent 1973). N can also be immobilized in organic forms if carbonaceous residues are returned in quantities sufficient to give actual increases in soil organic matter content (Bartholomew and Clark 1965, Broadbent and Stevenson 1966). How soils can be managed to maximize such removals over indefinite periods of time is not known (Jacobs 197“)- From 15 to 80% of the P in soils has accumulated in organic forms over geological periods of time (Allison 1973). Most soils have large capacities to adsorb or precipitate P0“ in mineral fractions, but this capacity can be saturated over a period of years where inputs are high (Ellis gt gt; 1972, Ellis 1973). ‘ Metal cations are retained in soils by exchange re- actions, precipitation or complexation with mineral and organic colloids (Ellis gt gt. 1972, Ellis and Knezek 1972). On the other hand, most metals can form soluble salts or complexes with mineral acids, such as N03, Cl, SO“, or with organic acids and chelating compounds which are formed during decomposition of wastes in soils. How these competing pro— cesses will balance out over long periods of time at high levels of input is not known. Organo—Mineral Complexes Numerous studies and extensive reviews have been published on the nature of complexes formed between metals and organic compounds found in soils (Mortenson 1963, Ellis and Knezek 1972, Schnitzer and Khan 1972, Khan 1969, Stevenson 1960, Stevenson and Gascho 1968, Stevenson and Ardakani 1972). There is general agreement that the interacting organic species include simple monomeric organic acids, phenols, etc., microbial polysaccharides and polyuronides, and a more or less continuous series of complex humus substances related to each other but increasing in molecular weight from less than 2,000 to more than 300,000, in the order: fulvic acid < humic acid < humin. Most surface functional groups which can interact with metals contain oxygen: carboxyl (-COOH), acidic hydroxyl (:DOH), carbonyl (>C=O). The ratio of active functional groups attains a maximum in the more highly oxidized and lower molecular weight fulvic acids, some of which are soluble. In larger fulvic acid and humic acid molecules, an increasing proportion of potentially active sites is bound up by various bridging structures which hold together the basic skeleton of aromatic rings derived mainly from lignin or plant and microbial pigments (Stevenson 1969). Important bridging structures include: -O-, -CH2-, -NH-, ==N-, -S-. Bridges involving N arise by condensation 10 reactions between NH3 or amino acids and intermediates which arise during the oxidation of lignin or aromatic-based pig- ments. Nitrogen bridges apparently involve heterocyclic rings which are very resistant to decomposition. The acidic functional groups in microbial polyuronides and humus substances give to these materials a cation exchange capacity which greatly exceeds that of clay minerals like vermiculite or montmorillonite (Ahlrichs 1972). Because of the diversity of oxygen-containing groups, they also enter into coordinate-covalent complexes with polyvalent cations. Some of these complexes are extremely stable, as with Cu and Zn (Allison 1973). Low molecular weight chelates are soluble and may be responsible for the downward displace- ment of Fe and Al during podzolization. 0n the other hand, the formation of complexes with Ca, Fe, Mn and Al is an impor- tant factor in stabilizing and retaining fulvic acids, humic acids and humin in soils (Kononova 1966). The reverse, of course, is also true: metals are stabilized in these com- plexes with humus substances; their mobility in soils and their availability to plants is regulated to an important degree in this way (Allison 1973). The importance of organo-metal interactions in soils used for waste treatment has been pointed out by numerous workers. However, few systematic attempts have been made to assess their significance for soil solution—solid phase equilibria among competing cations (Lindsay 1973). No reports were encountered where the distribution of cations associated 11 with different soil organic fractions has been studied in relation to applications of manure or other wastes. Methods for Extracting Soil Organic Matter A wide variety of methods has been used to extract and fractionate organic matter from soils (Ahrlichs 1972, Felbeck 1971, Stevenson 1965, 1969, Mortenson and Himes 196A, Kononova 1966, Kononova and Dokuchayev 1967, and Aleksandrova 1967). The quantity of water soluble organic compounds in soils is very low, except during the early stages of decom- position of added organic materials (plant trash, animal remains). A "bitumen" fraction, including fats, waxes and resins, is soluble in organic solvents but is normally a small proportion (one to six percent) of the total soil organic matter. About ten to 20% of soil organic matter can be shown to be carbohydrate in nature because constituted sugars and sugar acids are released by acid hydrolysis. Similarly, 30 to 50% of the organic N in soils can be hydrolyzed to amino acids. Much of the carbohydrate is present in the form of polysaccharides and polyuronides. Only a very small propor— tion of the amino nitrogen is present as protein, the remainder being intimately associated with or incorporated in humus substances (fulvic acids, humic acids, and humin). Methods for extracting humus substances are all based on the principle of breaking complexing bonds with minerals through use of alkali, chelating agents or reducing agents 12 (to reduce ferric iron). The most frequently used procedure involves extraction with dilute NaOH (0.1 to 0.5 N). The largest yields of organic matter are obtained with 0.1N NaOH. This concentra— tion also coextracts the largest quantities of complexing minerals (Al, Fe, Si, and P). For studies of organic matter composition and properties, Levesque and Schnitzer (1966) found this high ash content to be undesirable. In the present study, however, the principal objective was to identify metals associated with organic fractions, so 0.13 NaOH was used. This is also the concentration recommended by Russian workers (Kononova 1966). Workers outside of Russia have been concerned with the possibility that oxidation in alkaline extracts would alter the properties of organic matter extracted. Milder agents include sodium pyrophosphate adjusted into the lower alkaline range (pH 7.0 to 8.0). The Russian workers have examined this possibility and conclude that negligible alteration occurs during alkaline extraction. But they have found that sodium pyrophosphate (0.1M NauP2O7 in 0.1N NaOH) gives ef- ficient extraction without the need for prior decalcification with acid. Alkaline extracts are further fractionated into fulvic acid and humic acid by acidification to pH 2 or 3, at which point the humic acids are precipitated. The ratio of fulvic acid to humic acid has a diagnostic value in the Russian scheme for soil organic matter studies (Kononova 1966). 13 Additional quantities of organic matter can be ex- tracted by repeated sequences of acid and alkaline extrac— tion (Kononova 1966). The above extractants and fractiona- tions were used in sequential extraction schemes which are described in the next chapters. CHAPTER III EXPERIMENTAL PROCEDURES AND METHODS Selection and Description of Research Sites The proposed research undertook to utilize existing research plots of Michigan State University, Department of Crop and Soil Sciences. The plots selected were established in 1963 at the soils farm at East Lansing, and were designed to study the long-term effects of fertilizer and manure on continuous corn. In addition, a new study was established in 1972 adjacent to the existing plots where very high rates of manure were applied and corn was grown. Harvesting pro- cedures were the same for both studies. These two research sites were worthy of study in regard to the influence of long-term lower rates versus a single large application of manure on the chemical properties of the soil. The research plots selected for study are described in detail below. Experiment I The research plots selected were part of a long-term study to determine the optimum rate of manure to be applied annually to a sandy loam soil which was intensively cropped with continuous corn and to compare the efficiency of cattle manure Kg commercial fertilizer with respect to corn 1A 15 production for grain and silage (Vitosh gt _t. 1973). The treatments were replicated three times. The field experiment was a split-block design with each replication divided into grain and silage areas. The size of the whole plots was A2 x 128 feet. The sub-plot size was A2 x 6A feet. The study consisted of five treatments: (1) 10 T/a manure, (2) 20 T/a manure, (3) 30 T/a manure, (A) 160-A0-A0 (lb/a N-P205-K20), and (5) 160-190—190. In the spring of 1971, two treatments were added to the experiment by the application of three tons of lime per acre to one-half of the plots receiving 160-A0-A0 and 160-190-190 lb/a (treat- ments 6 and 7). The N for treatments A, 5, 6 and 7 was supplied as ammonium sulfate and the P and K as 0-20-20, plowed down prior to planting. The manure used came from the loose—housing beef barns at Michigan State University and was applied annually in the fall. This material contained considerable amounts of straw bedding. Manure obtained in 1971 showed the following chemical analysis: 73% water, .55% N, .15% P, .A6% K, .18% Ca, .12% Mg, .06% Na, 179 ppm Fe, 70 ppm Al, 3A ppm Mn and 2A ppm Zn (wet basis) (Vitosh gt gt. 1973). Totals applied for elements and treatments of special interest to this report are given in Table 1. All treatments received 200 lb/a 5-20-20 starter fertilizer each year, banded two inches to the side and two inches below the seed at planting. Minimum tillage practices 16 Table 1. Manure analysis and approximate loadingf. Total applied Quantity/IOT 3OT 300T Element Analysis application (9years) (lyear) lbs - - — - lb/a - - - - N 0.55% 110 2970 3300 P 0.15% 30 810 900 Ca 0.18% 36 972 1080 Mg 0.12% 2A 6A8 720 Fe 179 ppm 3.58 97 107 A1 70 ppm l.A0 38 A2 1'Chemical analysis on wet basis (73% water) 1971. 17 were employed. Weeds were controlled with herbicides. Plant populations varied from 13,600 plants per acre in 1963 to 28,000 in 1970. Row spacings were A2 inches, except in 1967- 70 when two row spacings (28 and A2 inches) were used and in 1971, when only the 28-inch spacing was used. Corn hybrids were different each year. Experiment II A new study was started in the spring of 1972 across the alley from Experiment I. The purpose of the new study was to determine the effect on the soil and crop of a single application of manure at very high rates on a sandy loam soil. The field experiment was a randomized block design with four replications. The plot size was 1A x 115 feet. The study consisted of four manure treatments: (1) no manure, (2) 100 T/a, (3) 200 T/a, (A) 300 T/a. The manure used came from the same source as in Experi- ment I. It was applied in February of 1972 and plowed under. All treatments received 200 lb/a 5-20-20 starter ferti- lizer, applied as in Experiment I. The no manure treatment received 150 lb/a N as NHu80u just prior to planting. Description of soils Replication l of Experiment I and the four replications of Experiment II were classified as a Hodunk sandy loam (Ochreptic Fragiudalfs) with 0-2 percent slope. This soil series consists of moderately well drained gray-brown l8 podzolic soils with fragipans, developed on calcareous sandy loam glacial till. The typical profile consists of an Ap 0-7 in, sandy loam; A2 7-16 in, sandy loam; le 16-25 in, sandy loam to light sandy clay loam; B2t 25-A6 in, sandy clay loam; and C A6 in+, sandy loam. The profile observed on the site is described in Table 2. The predominent soil type in replications 2 and 3 of Experiment I vuns classified as a Metea loamy sand (Arenic Hapludalfs) with 0-2 percent slope. This soil series consists of well drained to moder- ately well drained soils with 20 to A0 in of sands or loamy sands over loam to clay loam till materials. The typical profile consists of an Ap 0-8 in, loamy sand; A21 8-15 in, loamy sand or sand; A22-Bl 15—28 in, loamy sand; II B2t 28-A0 in, clay loam; II C A0 in+, loam or clay loam till. Sampling Procedures The soil samples were taken with hydraulic probe mounted on a truck. Samples were taken in the fall of 1971 to a depth of 10 feet from three replications of each treatment in Experiment I. Experiment II was sampled in the fall of 1972; samples were taken to a depth of five feet from three replica- tions of each treatment. The surface cores were taken with a 1 1/2 inch by A8 inch probe. The cutting point was 1/8 inch smaller than the inside diameter of the tube. This reduced friction, and prevented compaction of the cores. Compaction of samples was slight, seldom being more than one inch for a A8-inch 19 Table 2. Hodunk sandy loam profile exposed at the site of Experiments I (Rep 1) and II. Depth Horizons (in) Ap 0-9 A21 9-12 A22 12-15 BC 39-A2 C A2-120 20 depth core. For the lower depths a smaller one-inch tube was substituted. This reduced side friction, and made possible deeper sampling. After the probe was inserted to the desired depth, it was brought to the surface and disengaged from the hydraulic equipment. A wooden rod was used to push the core of soil from the probe. The core was then cut into the following increments: Ap horizon 0—9 inches, three-inch increments from nine inches to five feet, and six-inch increments from five to ten feet. Two cores were taken from each plot. One core was taken to a depth of ten feet and a second adjacent core was taken to five feet. Sampling increments from the two cores to five feet were combined. Each soil sample was placed in a paper bag and stored in an ice-cooled, insulated chest while being transported to the laboratory. The results of three replica- tions were averaged for the values reported. Laboratory Methods of Analysis Soil samples were refrigerated in the field and were held below 0 C until dried for analysis. Soil samples to be tested were dried in a forced air oven at a temperature not exceeding 55 C. These samples were then ground with a mortar and pestle to pass through a two mm sieve and stored in tightly stoppered glass bottles. Sequential extractions Two sequential extracting procedures were employed (Figure l) to study the distribution of N, C, and metals 21 (Ca, Mg, Fe, and Al) in soluble and exchangeable forms (KCl extraction) and in various nonexchangeable fractions. Se- quences of acid and alkaline extraction were used to effect differential recoveries from mineral and organic complexes in the soil solid phase (Kononova 1966). Forty grams of selected soil samples from Experiments I and II were reground in a Spex Mixer/Miller (This grinding rendered the soil to the texture of face powder.) The pul— verized soil was thoroughly mixed and stored in plastic air- tight bags. Three 1.00 g samples were removed and total N was determined. Two 0.100 g samples were used for determina- tion of total C. Two 12.5 g samples were then weighed into two 200 ml Nalge centrifuge bottles, one for each of the two extraction sequences in Figure 1. 1. KCl (Seq. I-a and II-a) Thirty ml of 1.7% KCl was added to each of the two 12.5 g samples of soil in the Nalge centrifuge bottles. The bottles were covered and shaken on a rotary shaker for 15 minutes. The bottles were then removed and centrifuged in a Sorvall Refrigerated Centrifuge at -7 C and at 10,000 RPM for 20 minutes. The supernatants were decanted and the soil freed of C1 by resuspending and centrifuging in four successive 50 ml aliquots of cold deionized water. Extract II-a and washings were discarded. Extract I—a and washings were combined, allowed to come to room temperature, and then made up to 250 m1. Three 25 m1 aliquots were analyzed for Kjeldahl N immediately._ The remaining solution was stored 22 SOIL Sequence I Sequence II I I-a. 1.7% KCl I I-b. 0.1g H280” I I-c. 0.1M NauP207 in 0.13 NaOH II-c. _‘ 1 II—d. I-d. Fulvic Acid I-e. Humic Acid A I-f. 0.15 H I I—g. 0.1g NaOH I Residual Soil (b) 230” W II-b. 0.1g NaOH r Fulvic Acid Humic Acid I I II-f. 0.1 N NaOH Residual Soil (b) Figure 1. Flow diagram for extraction Sequences I and II. 23 in a plastic bottle for metal analysis. 2. H2§Qu (Seq. I-b) To the Nalge centrifuge bottle containing the soil from extract I-a, 200 ml of 0.1N H280“ was added. The bottle was capped and shaken by hand for a few minutes to insure mixing and left to stand for 2A hr. At the end of the 2A-hr period, the bottle was centri- fuged and the liquid decanted into a 500 m1 volumetric flask. The remaining soil was then washed twice with 100 ml ali- quots of cold 0.1N H280“ acid. The washings were recovered by decanting after centrifugation and combined with the 2A hr extract. The combined solution was allowed to come to room temperature before making up to 500 ml with deionized water. Three 25 ml aliquots were then analyzed for Kjeldahl N and the remaining solution was stored in a plastic bottle for analysis of metals. 3. Nau§2g7 (Seq. I-c) To the soil residue from extract I-b, 200 ml of 0.1M NaIIP207 in 0.1N NaOH was added in the evening. The bottle was capped and mixed thoroughly by hand. The next morning the bottle was shaken and centrifuged. The liquid was decanted into a 250 ml volumetric flask and allowed to come to room temperature before making up to volume. Three 25 ml aliquots of this extract were analyzed for Kjeldahl N. A portion was stored in plastic for metal analysis, and the remainder was separated into fulvic and humic acid 2A fractions (extracts I-d and I-e). A. NaOH (Sgg. II-b) To the soil residue from extract II-a, 200 m1 of 0.1N NaOH was added. After capping and mixing thoroughly by hand, the bottle was allowed to stand for 15-16 hours before resuspending and centrifuging. The extract was decanted into a 250 m1 volumetric flask and allowed to come to room temperature before making up to volume. Three 25 ml aliquots were analyzed for Kjeldahl N. Part of the remaining solution was separated into fulvic and humic acid fractions (extracts II-c and II-d). The remainder was stored in a plastic bottle for analysis of metals. 5. Fulvic acid (Seq. I-d and II-c) Fulvic acid/humic acid separations were performed on 100 ml aliquots of the pyrophosphate extract (I-c) and of the NaOH extract (II-b). After acidifying from pH 13 to 2.5 with concentrated sulfuric acid, the solutions (in covered 200 ml beakers) were heated for 30 minutes on a boiling water bath and allowed to stand overnight at room temperature. The next day, the precipitate (humic acid) was removed on Whatman No. A2 filter paper. Cold 0.053 H280“ was used to effect complete transfer and separation of acid-soluble fulvic acids from the precipitate. The fulvic acid frac- tions and washings were collected in 200 m1 volumetric flasks and allowed to come to room temperature before making 25 up to volume. Aliquots were analyzed for Kjeldahl N and the remaining solution stored in plastic bottles for metal analysis. 6. Humic acid (Seg, I—e and II—d) The precipitated humic acid fractions were taken up again in warm 0.05N NaOH. The filters were washed several times to assure complete recovery. The recovered solutions were made up to 200 m1 at room temperature. Aliquots were analyzed for Kjeldahl N and the balance stored in plastic for metal analysis. 7. Extended series, H2§9A extract (Seq, I—f and II-e) Organic matter remaining in soil residues from NauP2O7 extraction (I-c) and from NaOH extraction (II-b) were sub- jected to a further succession of extractions in acid and alkali to get some indication of the extent to which metals may have been stabilized in complexes associated with more resistant organic or mineral fractions in soils. The residues from extracts I-c and II—b were extracted for 2A hr in 0.1N H280“, as has been described for extract I-b. Aliquots of the combined extract and washings were analyzed for Kjeldahl N and the remainder stored in plastic for metal analysis. The soil after this acid extraction was dried in a forced air oven at 20 C and pulverized again in the Spex Mixer/Miller. A 10.0 g aliquot was returned to the Nalge centrifuge bottles for alkali extraction (I-g and II-f) and the remainder set aside for organic C analysis. 26 8. Extended series, NaOH extract (Seq. Iigtand II-f) The 10.0 g aliquots of dried soil residue from extracts I—f and II-e were extracted for 15—16 hr in 190 ml of 0.1N NaOH as has been described for extract II-b. Aliquots were analyzed for Kjeldahl N and the remainder stored in plastic for metal analysis. The soil residue after this extraction was dried and pulverized again, then transferred into 0.1N HCl for about 2A hr to destroy carbonates. After thorough washing with deionized water, they were dried at room tempera- ture, ground with mortar and pestle and set aside for deter- mination of Kjeldahl N and organic C. Perchloric acid digestions Perchloric acid digestion was used to destroy complex- ing organic matter in all extracts after KCl so that a quantitative measurement of the metals could be made (Leves- que and Schnitzer 1966). The basic procedure followed was developed by Johnson and Ulrich (1959). A 50 m1 aliquot of the extraction liquid was pipetted into a 100 m1 Kjeldahl flask. To this flask five m1 of concentrated nitric acid was added and digestion was carried out on an electric digesting unit until about ten m1 of solution remained. At that point the digestion was stopped; the flask was removed and allowed to cool. Then ten ml of concentrated nitric acid and two m1 of 70% perchloric acid were added. The digestion was resumed until about two ml of solution was left and the dense white per- chloric acid fumes had subsided. (Sometimes this last 27 digestion with nitric and perchloric had to be repeated due to high organic matter content of the liquid or because of high concentrations of phosphate.) The flask was removed from the digesting unit and allowed to cool. At this point two different procedures were developed to promote higher quantitative measurement of the metals to be analyzed. The need for two procedures arose when A1 could not be determined in the presence of diethylenetriamine- pentaacetic acid (DTPAl). 1. Procedure for Ca, Mg and Fe After the digestion mixture had cooled, a few m1 of deionized water was used to wash down the side of the long neck flask. Then ten ml of 0.5M DTPA was added to complex all metal cations and to keep them in solution. After heating to the boiling point, the contents were swirled to insure that nothing adhered to the walls of the flask. After cool- ing overnight, the contents of the flask were transferred quantitatively to a 50 ml volumetric flask and brought up to a concentration of 0.1M DTPA. The solution was then analyzed for Ca, Mg and Fe. 2. Procedure for A1 After the flask had cooled, a few ml of deionized water was added to wash down the neck and sides. Then five m1 of 93 HCl was added. After heating to the boiling point, 1Compliments of Ciba-Geigy Chemical Company 28 the contents were swirled to insure that nothing adhered to the walls of the flask. After cooling, four drops of 2, A—dinitrophenol indicator and about three ml of lON NaOH were added to bring the solution to a light yellow (pH 3.5). The contents were then transferred quantitatively to a 50 m1 volumetric flask, made to volume, and analyzed for A1. Chemical analysis 1. Nitrogen Ammonium and N03 in the K01 soil extracts were recovered by steam distillation in the presence of MgO, using Devarda's alloy to reduce N03 to NH3 (Bremner 1965a). Kjeldahl N (excluding N03) was determined in soils, extracts and ex- tracted soil residues by a semimicro-Kjeldahl procedure, using CuSOu, K280“ and Se as catalysts (Bremner 1965b). Ammonia in distillates was collected in two percent H3B0u containing two drops of methyl purple mixed indicator (Fleisher Chemical Co.) and titrated against standard acid. 2. Carbon A 0.100 g sample was used to determine organic C in selected soils and extracted residues, using a Leco Carbon Analyzer. Where necessary, the sample was pretreated with 0.13 HCl to remove carbonate. 3. Phosphorus Available P in soil was recovered in Bray's P—1 ex- tracting solution (0.025N HCl and 0.03N NHuF). P was 29 determined by Jackson's Method I (Jackson 1958), except 1- amino-2-naphthol-A-sulfonic acid (Eastman 360) was used as the reducing reagent instead of chlorostannous acid (Fiske and Subbarow 1925). A. Calcium and magnesium Ca and Mg were determined directly in KCl extracts or, in the case of other extracts, in the presence of 0.1M DTPA after perchloric acid digestion. A Perkin-Elmer 303 Atomic Absorption Spectrophotometer was used. Solutions for analysis contained five m1 of sample and five m1 of a 16,000 ppm solution of LaCl3, made up to 25 ml with deionized water. Standard solutions of Ca and Mg were made to be .1M DTPA plus 20% of the LaCl3 stock solution in a perchloric acid matrix for the perchloric acid extracts or 1.7% KCl plus 20% LaCl3 solution for the K01 extracts. 5. Iron Fe was determined directly in the 0.1M DTPA perchloric acid digests and the 1.7% KCl extracts using the Atomic Absorp- tion Spectrophotometer. Some of the samples required dilu- tion. The standards were made to be 0.1M DTPA with a per- chloric acid matrix the same as the samples. Lanthanum chloride was not used. 6. Aluminum The procedure developed by McLean (1965) was used, with the following alterations. (l) A 50 ml volumetric 30 flask was used instead of a graduated test tube. (2) Due to extreme difficulty in adjusting pH in the perchloric acid digest, 15 ml of pH A.0 ammonium acetate buffer was added to perchloric acid samples and standards before making to final volume (50 ml). Additions of buffer kept the pH at 3.8. (3) A stock solution of perchloric acid was made. This solution served as a blank and was added to the stand- ards to provide the same matrix as in the unknown samples. (A) The 1.7% KCl extracts were analyzed as stated by McLean, except 1.7% KCl was used in the standards and blank instead of 1N KCl. 7. pH and buffer pH Soil water and buffer pH was determined, on selected soils samples, according to Schoemaker gt gt. (1961). Statistical Methods Data for Experiment I were analyzed as described by Cochran and Cox (1957) for a split plot, randomized block in three replications, with systematic main plots (corn or silage harvest) and randomized sub-plots (nitrogen treat- ments). Experiment II was treated as a randomized block with three replications. The data for sequential extractions were analyzed as for a split-split plot in three replications, with manure treatments as randomized main plots, depths as sub-units and extracts or soil and soil residues as sub-sub-units. Simple 31 and multiple correlations among N and metals within depths and extracts were calculated. Facilities and programs of the Michigan State University Computer Laboratory were used. CHAPTER IV RESULTS AND DISCUSSION Influence of Manure and Fertilizers on the Soil Environment Experiment I Experiment I was designed to study the long-term ef— fects of fertilizers and manure on continuous corn. After nine years of annual treatment, soil samples were obtained for analysis. To compare soil horizons to sampling depths, see Table 2. Distribution of Soil N The results of the 1971 sampling are summarized in Table 3. Tables 23 to 30 in the Appendix give the N content of the soil by incremental depths. Soil variation was great and only a few differences associated with treatment were statistically significant. The significant differences which were found, however, serve to identify trends which appear to be important and which were expressed more generally through the data. Organic N in the plow layer was higher at the 30T level of manure application than for other treatments (Table 3). It tended to be higher where more organic residues for im— mobilizing N were returned as stover after grain harvest than where all above ground production was removed as 32 323 Table 3. -Summary of nitrogen forms recovered in the plow layer and in subsoils to 10 feet after nine annual applications of manure or commercial fertilizers for corn. Depth Corn Treatments (in) harvested as Manure 160 lb N 160 lb N NPK 160 lb N + + + 10T 20T 30T P a K lime lime ppm a) Organic-N 0-9I Grain 8331 7659 107A5 7572 7816 6098 73A1 Silage 7812 3810 9702 7217 6231 6569 5772 9-120 Grain A607 A312 A829 3960 5152 A870 5A38 Silage A600 A902 5A78 A781 5000 A666 A1A7 0-9 Grain 38 A0 SA 55 A1 22 A1 Silage 22 30 61 17 37 25 23 9-1205 Grain 17 A6 55 2A 26 50 70 Silage 82 81 82 62 A5 28 56 0—9' Grain 60 56 95 88 3A 32 A5 Silage Al 27 37 26 18 18 35 9-120 Grain 116 126 180 119 188 161 235 Silageq 103 153 213 250 220 290 159 d) Total mineral-N 0-9' Grain 98 96 1A9 1A3 75 SA 85 Silage 63 57 98 A3 55 A3 58 9-120 Grain 133 172 235 1A3 21A 211 305 Silage 185 23A 299 312 265 318 215 +Mean of three replications. *Organic-N (0-9 in. only): Treatments within crop LSD (.05) - 2A81. §NHu-N (9-120 in. only): Treatments within crop LSD (.05) - 35 LSD (.05) ‘ 51 ‘NO3-N (0-9 in. only): Treatments within crop LSD (.05) ' 57 ”Total mineral-N (0-9 in. only): Crop within treatments LSD (.05) - 72. 3A silage. There is evidence in Tables 23 and 2A that mobile forms of organic N have been displaced to considerable depths into parent materials below A2 inches. Thus, the average for the bottom five increments (90 to 120 in) at the 10T rate is about 80 ppm organic N, whereas the values for incre- ments between A2 and 60 inches are significantly greater for most treatments. The indicated depth of penetration is greater for the 20 and 30T rates than for 10T, particularly where silage was removed. The addition of lime to plots harvested for silage tended to reduce the depth to which organic N was displaced, but the reverse was true for plots where only grain was harvested. These differences are reflected in organic N totals for 9-120 inches in Table 3, although not with statistical significance. Ammonium is the first mineral form of N released when plant materials decompose. Since it is a cation, it is not very mobile in soil. It is also quickly nitrified. Twenty to 70% of the total NHu-N encountered was found in the plow layer where most of the decomposition of manure and corn residues would have taken place (Table 3). Concentrations found at greater depths were very much lower (Tables 25 and. 26). Nevertheless, when subsoil quantities were totaled to 120 inches, significant relationships to treatment and cropping system were expressed. Since much of this NH“ was probably produced t3 gttg or released from organic matter during air drying of samples, these data provide indirect evidence 35 to support the data for organic N which indicate that the nature and properties of soil organic matter were influenced to considerable depths in the soil by treatment and the quantity of corn residues returned. Nitrate and total mineral N in the plow layer tended to be lower where silage was harvested than where only grain was removed (Table 3). This reflects the more efficient harvest removal of N and other nutrients. Over the nine- year period, silage yields have responded to increasing manure rate and additional P and K, whereas grain yields have not responded to more than ten tons of manure and were frequently depressed by the addition of P and K with fer- tilizer N (Vitosh gt gt. 1973). On the other hand, NO3 and total mineral N in subsoils tended to be lower where grain was harvested (Table 3). Of the total NO3-N recovered from plots under grain harvest, 57 to 85% was found below the plow layer, whereas 72 to 9A% was recovered from subsoils under silage harvest. Thus, it seems that, over the years, the additional organic matter in returned stover has contributed to closer cycling of N by immobilization and/or greater losses by denitrification. Nitrate which passes beyond the influence of plant roots and associated microflora can be expected to remain stable in the percolation stream at greater depths (Viets and Hageman 1971). Concentrations below A8 inches ranged from two to 12 ppm, on a dry soil basis (Tables 27 and 28). Concentrations in percolating solution above field capacity 36 would be several-fold greater. Distribution of Soil P The soil analyses for extractable P (Bray P1) for all N treatments (see Table A) showed no significant differences between corn harvested as grain or silage. The Ap horizon shows higher accumulations of P where 30T of manure, 160 1b N + P & K, and NPK + lime were applied. The values were 121, 1A3 and 153 lb/a respectively. Downward movement of P was greater for the manure treatments, even though the Ap horizon had less extractable P than for mineral treatments. Experiment II Experiment II was designed to study the effects of manure at high rates on soils and continuous corn. The soils were sampled eight months after the first applications of 100, 200 and 300 T/a of beef cattle manure. Distribution of 8011 N The 1972 sampling results are summarized in Table 5. Tables 31 to 3A in Appendix give the N concentrations by increments of depth. Organic N in the Ap horizon (0-9 in) increased with rate of manure application (Table 5). The increase for 300T (3,500 1b/a) would account totally for the estimated 3,300 lb N applied (cf. Table l). The quantities found, however, were substantially less than in Experiment I where essentially similar total applications of manure had been made over a 37 .omeHm go cfimgw mm oopmo>gm£ cgoo goozpon cocoaommao pcmoHMchHm oz+ .msoaquHHoog oops» no cmoz+ m.~ m.u N.m m.~ m.m m.> N.~ Amo.v omq beospoohp eases; spoon a: CH a ma OH NH ma ea mHImH a: ma OH ea NH mm mm om mHImH no mm as mm as mm o: m: NHIa m.em mmfi co mas mm HNH no em mIo IIIIIIIIIIIII IIIIIIIIIIIIIIIIIIII IIIIsoaIIII IIIII IIIIIIIIIIIIIIIIII IIIII III oEHH oEHH x w m Bom Bom Boa + + + 2 ha 0 H Amo.v emu xaz 2 pg oefl 2 ha ooH o oases: AcHV spoon canes: nocossoose spoon pCoEpmogB + +.:poo you mpoufiafiuaom HmaogoEEOo no ogscme mo mcofiooOHHoom amazed mafia pound oHHmogd on» Ca spawn Spa: AHId zmgmv msgcnomozd manmpomauxo no soapsnanpmfim .z canoe 38 Table 5. Summary of nitrogen forms recovered in the plow layer and in subsoils to five feet eight months after a single application 0; manure at high rates for corn harvested as grain. Depth Tons of manure per acre (in) None 100T 200T 300T --------------- lb/a--—------------- a) Organic-N " 0-91 A710 597A 7110 8239 9-60 2639 2259 2887 2787 b) NHu-N 0-9 A1 37 36 58 9—60 5A 50 30 20 c) NO3-N 0-9 235 151 580 A06 9—60 176 222 367 2A1 d) Total mineral-N 0-9 276 188 616 A6A 9-60 230 272 397 261 1'Mean of three replications. IOrganic-N (0-9 in. only): Treatment LSD (.05) = 2280. 39 nine-year period at one-tenth these annual rates. Since there were no control plots in Experiment I, the increase due to manure cannot be estimated to resolve this apparent discrepancy. Organic N in the subsoil was increased, but only to depths of 15 to 18 inches (Table 31). The subsoil concentra— tions were less than for the same depth increments under manure treatments in Experiment I (of. Tables 23 and 2A). Ammonium in the plow layer was not related to treat- ment (Table 5). However, subsoil levels were significantly influenced over most depth increments to 60 inches (Table 32). The extractability of subsoil NH“ tended to decrease with increasing manure application. This is the reverse of that observed in Experiment I (cf. Tables 3 and 25). It was difficult to incorporate these very heavy applica- tions of manure. The surface soil lost its infiltration capacity for a time, becoming very wet, with ponded water observed on several occasions. The potential low oxygen tensions are the most likely influence which could have been transmitted through 60 inches of profile over the short period of eight months to reduce the extractability of NH“. The mechanism of the effect is not clear, however. Variable oxygen status may have contributed also to great variability in the N03 analyses. Nitrate increased generally with increasing manure rate, both in surface and subsoils (Table 5). Variations in total mineral N were due mainly to variations in N03. A0 The concentrations of NO3 in the plow layer were several-fold greater than in Experiment I (of. Table 3). Significant differences for manure rate were encountered in subsoils to 27 inches (Table 33). Below 27 inches, N03 concentrations were the same as for comparable depths in Experiment I (Tables 27, 28). Distribution of Soil P The results of soil analysis for extractable P is tabulated in Table 6. Phosphate content increased in the Ap horizon with additions of manure. The ranges were 57 ppm for no manure to 105 ppm for the 300T. There was no evidence for any downward movement. Sequential Extraction of Organic Matter and Associated Metals Sequential extractions were used to study exchangeable and nonexchangeable N and metals, and their interactions. For this study, the no manure, 300T (1 year) and 30T (9 years) manure treatments (stover returned to the soil plots) were used. The essential treatment variable was time, since the total manure applications were essentially the same (Table l). A further source of variation must be recognized and that is that the 30T/year manure treatment was in a separate experiment (Experiment I) which was, however, immediately adjacent to Experiment II from which samples were taken for no manure and 300T/a in one application eight months before sampling. A1 Table 6. Distribution of extractable phosphorus (Bray P-l) with depth in the profile eight months after a singlg application of manure at high rates for corn. Depth Tons of manure per acre Treatment (in) within depth None 100T 200T 300T LSD (.05) -------------------- ppm—-------—----------—- 0-9 58 69 97 105 18.0 9-12 19 12 21 20 ns 12-15 10 7 l2 8 ns 15-18 8 7 13 A ns Depth within treatment LSD (.05) 11.5 11.5 11.5 11.5 1. Mean of three replications. A2 Surface (0-9 in) and subsoil (18-21 in) samples from each of the three field replications of each treatment were carried through the extraction schemes in Figure 1. (The 18-21 in depth was used because it was a sufficient distance from the plow layer and preliminary C/N ratios increased with manure applications.) All chemical analyses were in duplicate or triplicate and the average value reported. Statistical analyses were based on the three field replications. Exchangeable Forms of N and Metals It appeared that the level of exchangeable N in the plow layer increased with time and with increasing total organic N rather than with the annual rate of manure application (cf. Tables 7, 23 and 31). Calcium was the dominant exchangeable cation (Table 7). In the plow layer, it tended to increase with manure applica- tion over time, and exchangeable Ca and N were positively correlated (r = .89, P = 0.01). Exchangeable Ca was signifi- cantly higher in the subsoil than in the plow layer in no manure plots. There was a marked tendency for it to decline in the subsoil with manure application over time. Magnesium showed a similar trend in the subsoil, and Ca and Mg were positively correlated in the subsoil (r = .85, P = 0.01). Exchangeable Fe and A1 were very low, the values for Al being at the lower limit for detection. “3 Table 7. Recoverigs of exchangeable N and metals in 1.7% KCl . Treatments Treatment Elements within depth No 300T 30T LSD (.05) manure (1 year) (9 years) ------------- ppm-----—----—-- Plow layer N 2A A6 59 28.9 Ca 321 376 389 ns Mg 116 132 107 ns Fe A 10 A0 ns A1 20 20 20 ns Subsoil layer N 8 9 5 ns Ca 728 622 387 ns Mg . 130 115 110 ns Fe 8 A 18 ns A1 20 20 20 ns Depth within treatment LSD (.05) N ns 23 23 Ca 3A0.9 ns ns Mg ns ns ns Fe ns ns ns A1 ns ns ns IMean of three replications. AA Nonexchangeable Forms of N and Metals 1. Overall recoveries Vitosh gt gt. (1973), in studies which included Ex- periment I of this study, observed that a major portion of nutrients applied in manure could not be accounted for by crop removals or by standard soil tests for available P or exchangeable cations. It was, therefore, of interest to compare the capacity of soils to retain cations by exchange mechanisms with their capacity for retaining them by precipi- tation or complexation in soil fractions extractable with relatively mild reagents which have been used for isolating important organic constituents of soils. Total recoveries of N and metals in nonexchangeable fractions, averaged over both extraction sequences, are presented in Table 8 for comparison with exchangeable quanti- ties in Table 7. The N in Table 8 is organic and provides an estimate of extracted organic matter. The increase in extractable organic N in the plow layer (0-9 in) for the 300T manure treatment over no manure was 171 ppm or about 500 1b/a. This is equivalent to about 15% of input N (cf. Table l). The indicated increase for 30T annually over nine years would be 700 1b/a, or 2A% of input; however, this estimate must be questioned, since the 30T plots were in a separate, though immediately adjacent experiment (Experiment I). Extractable Ca in the plow layer (Table 8) increased A5 Table 8. Total recoveries of nonexghangeable N and metals by sequential extraction. Treatments Treatment Elements within depth Maggre (13322r) (9 gggrs) LSD (.05) ------------- ppm---——-—-—----- Plow layer (0-9 in.) N 388 559 625 1A0.2 Ca A89 993 1A70 900.9 M8 A36 657 A61 ns Fe 3555 323A 21A2 107A. A1 2202 2526 1863 ns Subsoil layer (18-21 in.) N 81 76 89 ns Ca 228 3A3 1301 900.9 MB 91“ 879 735 ns Fe 6879 A635 3028 107A.8 A1 3806 3AA3 2260 969.8 Depth within treatment LSD (.05) N 106.2 106.2 106.2 Ca ns ns ns Mg 3A7.0 ns ns Fe 127A.9 127A.9 ns Al 9A1.A ns ns 'Mean for two extraction sequenCes and three replications. A6 as organic N increased. The extractable concentrations were much greater than the exchangeable concentrations in Table 7. The increase in exchangeable Ca in the plow layer for the 300T treatment over no manure was 165 lb/a, or 15% of input, whereas the increase in extractable but nonexchange- able Ca was 1,500 lb/a, or 139% of input. The indicated increase for the 30T (9 years) treatment would have been 2,900 lb/a, or 300% of input. This is clearly not a treatment effect and is due to unsuspected differences in soils between Experiments I and II. Extractable Ca was related directly to pH; 6.0, 6.1, 6.9 in the plow layer for no manure, 300T and 30T, respectively, and 5.6, 5.6, 6.9 in the subsoil. In the subsoil layer (18-21 in), exchangeable Ca (Table 7) was greater than nonexchangeable (Table 8) except for the 3OT (9 years) treatment. There was no relation between extractable N and Ca in the subsoil. Nevertheless, the very high level of extractable Ca in the subsoil under the 3OT treatment (Table 8) was associated with much higher levels of organic N at this depth in Experiment I (Table 23) than in Experiment II (Table 31). Larger quantities of Mg were recovered by sequential extraction (Table 8) than were present in exchangeable form (Table 7). The indicated increases over no manure in the plow layer for the 300T manure treatment were A8 lb/a ex- changeable Mg (seven percent of input) and 660 lb/a extract— able (92% of input). Very much larger quantities of Fe and Al were extracted A7 than were present exchangeably. In both surface and subsoil layers, extractable Fe was related inversely to extractable Ca. This is the most striking relationship to be seen in Figure 2. In the subsoil this was true also for Al. 2. Fractional recoveries The summed recoveries of nonexchangeable N and metals for Extraction Sequences I and II are presented separately for comparison in Table A6. Both sequences removed very similar quantities of organic N. In general larger quantities of metals were removed by Sequence I. By reference to Figures 3 and A and to Appendix Tables 35 through Al, it can be seen that most of the organic N came out in the Nauon7 extract (I-c) in Sequence I and in the first NaOH extract (II-b) in Sequence II. In the case of plow layer samples, additional quantities of organic N were removed in the final NaOH extraction I-g and II-f (Figure 3); only small amounts appeared in this final extract from sub- soils (Figure A). Total extractable Ca was correlated positively with total extractable N (Table 9). Nevertheless, relatively little Ca was in fact associated with organic N in the same extracts (Figures 3 and A). Most of the non-exchangeable Ca came out in the initial 0.1N H280” extract (I-b) in Se- quence I and in the 0.1N H280" extract (II—e) after NaOH in Sequence II. A8 3000 3234 fig” Mg \ E Fe .5“ A/ 9'3 2000 - Q \ r § q - q l000 6879 3806 4535 3000 * 34“ '\. L S \ $1 2000 - Q I— ‘~. § R |000 ~ 1 L None 300 T 30 T (I Year) (9 years) AMA/UH" TREATMENTS Figure 2. Distributions of the total recoveries of non-exchangeable N and metals by sequential extraction. 2000- fi l000 L I000[ PPM 3000- 2000L I000“ |000 Figure 3. LLB..— A9 H2504 (I'D) KCI (1'0 00d H‘O) CJAI [111111160 aarar EH? HE!” y. '_ L um" [MIL JIM... MWMWWMWWWWWWW _— — — - — _— — I ‘— - _— .— — — _— — — — — —. _- — _— _— _- _— .- — — _— — .- — .- .- — — — _- .— _— .— .- c‘ _— _- _ _ _— _ _- — _— ‘ — — == 5 - , —' ' -'—' = = = = H2 $04 (1.” HWMWMWWWWWMWWWMWMWWWMWNWM _ __ .— _ _— _— _— .._. _— _— _ _.... _— _. _— __ __ _. _ _— .— * _. .— .._. _. _r _ _. _, .— _ _— _ _ _. E _, I— Na 0H (I-g) Na 0H (II-f) [ HM None 300 T I ' 30 1 None 300 r 30 r (I year) (9 years) (I year) (9 years) )MMAMMHF FREM7IH5N73 Distributions of the recoveries of N and metals in sequential extractions of the plow layer (0-9 in.). 50 5' 9 H2304 (I‘b) KCI (1’0 and H-0) 3000 . * III/V " IIIIIIIlé‘a .Mg 6 $69 .4/ :MML WNa 0H (Il'b) O 2000 r I000 ~ _— — .— .— _— — — _— _— .- —. _— _— .— _- c...— = .— _— _— .— _— _— _— .— _— _ _. _- =‘ = .— _— __.. — E gun—- —- .—~ = _— g _- —.- _— .— r— 2000 IOOO ~ - PPM I N U) C b 7.: .1. :‘(IIIIIIIIIIIIIII 2000 r - l 000 1 MWWMWWWWWWWWWWWWM IIImmmmunmmuummuumuuumlmlummuu mmmummImumunuunmnmuuuummummmn: WWWWWWWWM W \ ‘§§ //, / Na 0H (II-f) IOOO .— _— _— .- — _— —— — _- .— _- a. _— .— ~— — .— _- _— — _- _— _ — .— — .— — _— .1.- _- _— - THIIIIIIIIIIIIIIIIIIII nmlmuummmmm ’ ’/» None 300 T 30 T None 300 T 30 T (I year) (9 years) (I year) (9 years) AMA/0R5 TREATMENTS Figure A. Distributions of the recoveries of N and metals in sequential extractions of the subsoil layer (18-21 in.). 51 Table 9. Simple correlations (r) between totals for non— exchangeable N and metals recovered in extraction sequences I and II. N Ca Mg Fe N Ca Mg Fe Sequence I Sequence II Plow layer (0-9 in.) Ca .8A" .85** Mg .16 .12 .57 .58 Fe -.65 -.66 .u7 —.60 -.72* -.07 Al .02 -.11 .69* .63 -.12 -.38 .38 .79** Subsoil layer (18-21 in.) Ca .87'* -.07 Mg .36 .us .03 .80** Fe -.29 —.A7 .3A .39 -.57 -.07 Al -.20 -.38 .57 .71* .09 -.u2 .13 .7u* *Significant at P(.05); r > .666. **Significant at P(.01); r > .798. 52 In the Russian scheme, an initial acid extraction is employed before alkali extraction to "decalcify" the soil and permit larger quantities of organic matter to be re- moved (Kononova 1966). Some fulvic acid materials will be removed also in this initial acid extraction. In the soils examined here, Fe and Al were removed in quantities equal to or greater than Ca, and substantial quantities of Mg were removed also. Due to prior removal in acid, and probably also to chelation and precipitation by pyrophosphate, cations were very low in extract I-c as compared with the parallel NaOH extract II-b. The quantities of organic N recovered by I-c and II-b were very similar (cf. Tables 36 and 37). The dominant cation removed in the first NaOH extract in Sequence II was A1 (Figures 3 and A). The effect of this alkali extraction on the subsequent acid extraction (II-e) was to increase the quantities of Fe removed as compared with the first acid extraction (I—b) in Sequence I which was not preceded by an alkaline extraction (cf. Tables 35 and 39). The Al extracted in II—e was much less than in I-b, but the total for the alkali followed by acid extractions (II-b plus II-e in Tables 37 and 39) was greater than for the single acid extraction I-b in Table 35. This data is in line with the principle that organo-mineral complexes in soils are mutually protective and that fractional removal of one component can increase the extractability of another (Kononova 1966, Schnitzer and Khan 1972). 53 The second H280“ extraction (I—f) in Sequence I removed additional quantities of Mg, Fe and Al, together with much smaller quantities of Ca (Figures 3 and A, Table 38). This second acid extraction is largely responsible for the larger total recoveries of cations in Sequence I than in Sequence II (Table A6). The two acid extractions in Sequence I may have increased the extractability in the final NaOH extract of Al in the plow layer (Figure 3) and of Fe in the subsoil (Figure A), but they had little effect on the residual extractability of organic N, Ca or Mg (cf. Tables A0 and Al). 3. Interactions among_N and metals The positive correlations between Ca and N in total recoveries from the plow layer (Table 9) were reflected in similar positive correlations between Ca and N exchangeable to KCl and between Ca and N in the first NaOH extract of Sequence II (Table 10). There was no strong relationship between Ca and N in any other fractional extract from the plow layer (Tables 10 and 11) or in any of the extracts from the subsoil (Tables 12 and 13). Overall recoveries of Ca and N from the subsoil were positively correlated in Sequence I but not in Sequence II (Table 9). Thus, there is little evidence that Ca was extensively complexed with extractable organic compounds containing N. The same conclusion can be drawn for Mg. The divalent cations were recovered mainly in acid extracts (Figures 3 5A Table 10. Simple correlations (r) between N and metals in sequential extracts (through humic acid) in the plow layer (0-9 in.) N Ca Mg Fe N Ca Mg Fe §2§9u_11121 KCl (I-a and II-a) Ca -.30 .89H Mg -.08 .66 .12 .A3 Fe .09 -.58 .09 .02 .10 .27 A1 .21 -.29 .25 .65 .00 .00 .00 .00 figug2g7_1;:gl_ NaOH (II-b) Ca -.33 .77* Mg .32 .A2 .39 .A3 Fe .A9 -.57 -.A8 -.79* -.81** -.39 A1 -.29 -.26 .0A -.07 -.28 -.A7 .19 .1A Fulvic (I-d) Fulvic (II-c) Ca .31 .5A Mg .3u .90** -.3A .55 Fe .07 -.78‘ -.80** -.29 -.71’ -.A0 A1 .AA -.36 -.20 .A8 -.30 -.11 .A6 .18 Humic (I-e) Humic (II-d) Ca .61 ' .11 Mg .A9 .95** -.2A .75* Fe —.65 -.83** -.79* -.27 .AA .8A** Al .21 -.22 -.31 -.ll -.53 .27 .0A -.0A *Significant at P(.05); r > .666. **Significant at P(.Ol); r > .798. 55 Table 11. Simple correlations (r) between N and metals in extended sequential extracts of plow layer (0-9 in.) N Ca Mg Fe N Ca Mg Fe ega1___ fizéQu—(II-e) Ca .66 -.25 Mg .29 .71* -.0A .69* Fe -009 all 071* 033 -070” -018 A1 .OA .27 .75* .93** .A9 .3A .A9 .2A NaOH (I-g) NaOH (II-f) Ca -.1A -.0A Mg -076* .142 -060 076* Fe -.72* -.2A .75* -.11 .A8 .A5 A1 .77* -.06 -.50 -.37 -.A3 .77“ .85** .83** *Significant at P(.05); r > .666. *“Significant at P(.Ol); r > .798. 56 Table 12. Simple correlations (r) between N and metals in sequential extracts (through humic acid) in the subsoil layer (18-21 in.). N Ca Mg Fe N Ca Mg Fe E2804 (I—b) KCl (I-a and II-a) Ca -.37 .AO Mg .09 .87** .03 .85** Fe .68* -/A5 -.09 .09 -.50 -.36 A1 .87** -.32 .05 .59 .00 .00 .00 .00 E212297_L£:21 NaOH (II—b) Ca -.59 .13 Mg -.30 .7A* .03 .A2 Fe .A5 -.97** -.67* -.02 -.A3 .60 Al —.06 .Al .35 -.A9 -.02 .09 .57 .58 Fulvic (I-d) \ Fulvic (II—c) Ca -.26 -.33 Mg -.15 .A9 .00 .00 Fe .30 -.76* -.39 -.16 .75* .00 Al .60 -.1A -.28 -.21 -.16 .13 .00 .Al Humic (I—e) Humic (II-d) Ca -.59 .18 Mg .00 .00 .22 .95** Fe .82** -.3A .00 -.55 .12 .26 Al -.6A .91** .00 -.27 -.AA -.10 .08 .6A *Significant at P(.05); r > .666. **Significant at P(.Ol); r > .798. 57 Table 13. Simple correlations (r) between N and metals in extended extracts of the subsoil layer (18-21 in.). N Ca Mg Fe N Ca Mg Fe 52§9n_£l:£1 52§9A_£l£:§l Ca -.30 -.12 Mg -.29 .81** -.O9 .89** Fe -.2A .39 .71* .6A -.58 —.29 Al -.A6 -.01 .A7 .A7 .27 -.37 .06 .65 NaOH (I-g) NaOH (II-f) Ca -.30 -.12 Mg -.39 .50 -.1A .13 Fe -.37 .08 .89** .01 -.26 .90** A1 -.A5 .80** .72* .A8 .A8 .18 .06 .1A “Significant at P(.05); r > .666. **Significant at P(.01); r > .798. 58 and A), in which they were positively correlated with each other (Tables 10 through 13). However, the small amounts of Ca and Mg which did ap- pear in pyrophosphate extracts (Table 36, Figures 3 and A) were negatively correlated with Fe (Tables 10 and 12), either in the extract itself (I-c) or in its fulvic (I-d) or humic (I-e) sub-fractions. Calcium was negatively correlated with Fe in the first NaOH extract (II-b) of the plow layer and its fulvic sub-fraction (II-c). In the NaOH extract, Fe was also correlated negatively with N. Iron and aluminum are characteristically complexed with organic matter extracted from soils, and the complexes are referred to as Fe- and Al-humates (Schnitzer and Khan 1972). The negative correlations noted for Fe in the preced- ing paragraph would be expected, since the cations would be competing for complexing sites in organic matter. Incorporation of N can also be competitive for the same sites (Stevenson and Butler, 1969). However, where incorpora- tion or release of N and metals occurs in association with substantial increases or decreases in quantity of humus, positive correlations can be expected since N becomes a mea— sure of the quantity of complexing organic matter. Thus, in Sequence I extracts of subsoils (Table 12), Fe and A1 are correlated positively with N in the first H280“ extract (I—b), and Fe and N are positively correlated in the humic sub- fraction (I-e). Iron and aluminum were the dominant cations extracted 59 (Table A6). Overall, they tended to be correlated with each other (Table 9). However, this inter-correlation did not carry over into many of the fractions in Tables 10 through 13, and many of their interactions with Ca and Mg were distinctly different. The pattern of interaction varied with extraction sequence and with position in the profile. Thus, in Table 11, Fe and Al were both positively cor- related with Mg and with each other in the second H280“ extract (I-f) of Sequence 1. In the following NaOH extract (I-g), the positive Fe vs Mg relationship remained, but the positive relationships between A1 and Fe, A1 and Mg were lost; A1 became positively related to N, and Mg and Fe became negatively related to N. In the parallel NaOH extract of Sequence II, Al is positively correlated with all three cations but not with N. Significant correlations between Fe and Mg in extracts I-f and I-g in Table 11 appear again in the same extracts from subsoils in Table 13, as well as in II—f. Relation— ships to Al in extracts I—g and II-f are very different in subsoils than in the plow layer. It is to be expected that interactions of metals with soil organic matter will vary with specific prOperties of individual cations and of organic species varying in their content of oxygen, nitrogen and aliphatic side chains (Kononova 1966, Stevenson and Butler 1969). Environmental conditions at the interface with the soil solution will also have a profound effect. 60 Schnitzer and Khan (1972) note that humic substances can be considered as true solutions of macro-ions or nega- tively charged hydrophilic colloids which vary in their susceptibility to coagulation by different cations. Coagu- lation depends also on pH and the chemical potential of the solution. In general, cations of the same valence with the largest ionic radius are the most effective coagulants. This rule does not apply to trivalent cations which, because of higher charge density, do not occur as simple cation species in solution and do form stable complexes with humic substances. They note also that complexes of Fe3+ and Al3+ with fulvic acids and humic acids are more susceptible to coagulation by Ca2+ as .99, Mg2+ as .66, Fe2+ as .7A, Fe3+ as .6A and Al3+ as .51 K. It is of interest to note that Ca, Mg, Fe and Al were associated with N in NaOH extracts in quantities related directly to valence and inversely to ionic radius (Tables 37, A0, A1). There were deviations for Fe in the no manure treatment (Tables A0, A1) and for Ca in the plow layer (Table 37). The relationship was less clearly expressed in pyrophosphate extracts (I-c, Table 36) but appeared again in the following H280“ extract (I-f) OTable 38), except for Al which appeared in smaller quanti- ties than Fe. It may be postulated from these observations that cations may have been complexed with humic substances in quantities related directly to charge density and inversely to ionic radius. Extracted complexes were stable in NaOH extracts 61 but were subject to breakdown through chelation of the metals by perphosphate. The chelated metals may have been pre- cipitated and contributed to metal recoveries in the succeed— ing H280“ extract (cf. Figures 3, A). Aluminum appears to have been the dominant complexed cation in the initial NaOH extract (II-b). There was an increase in the proportion of Fe in the final NaOH extracts (I-g, II—f), specifically in the more acid soils of Experi- ment II (no manure and 300T). Since most of the extracted Ca and Mg appeared in H280” extracts, it seems likely that these cations were present mainly in precipitated salts or colloidal mineral complexes. The highly significant positive overall correlations between total extracted Ca and organic N (Table 9) indicate that the acid-extracted Ca (Ca salts) served as the major coagulant for precipitating and stabilizing the portion of soil organic N recovered in these two extraction sequences. Thus, some of the extractable Ca may have been present as humate salts from which, however, it was more readily displaced by H+ than were Fe or A1. A. Regression of N on metals The overall extractability of N was influenced mainly by Ca. However, it appeared reasonable to expect that other cations would be influential at different stages of extrac- tion. Accordingly, Tables 1A and 15 give the results of step- wise least squares regression which was used to estimate 622 Table 1A. Optimal solution's for linear regressions of N on metals in extracts of Sequence I. Depth Regressions (b) of N on Coefficients Extracts (in) of Ca Mg Fe A1 determination a. 1.7: KCl 0-9 0.118012" -0.31906 Delete Delete R2 - .88!”I 18-21 0.02335!I -o.08892 Delete Delete n9 . .51 b. 0.15 H280“ 0-9 Delete Delete Delete Delete —-- 18-21 —0.00952* -0.03o37' Delete -a.00595 R2 = .92!“l c. 0.1! NauP207 + 0.15 NaOH 0-9 Delete 15.20115' 1.377u5' Delete R2 - .65! 18-21 -0.85518 )eiete Delete Delete r2 - .35 d. Fulvic Acid 0-9 Delete Delete Delete Delete --- 18-21 Delete Delete Delete 0.26320 r2 - .36 e. Humic 2 Acid 0-9 0.78170 Delete Delete Delete r . .37 18-21 Delete Delete 1.02561 -0.21906 R2 - .86" r. O-IN H280“ 0-9 1.15096H -a.53961' 0.07969“ Delete R2 - .89' 18-21 Delete Delete Delete Delete --- g. 0.15 NaOH 0-9 Delete Delete 0.123u9* 0.19213ll R2 - .81" 18-21 Delete Delete Delete Delete --- “Significant at P(.05). "Significant at P(.01). Delete - deleted from optional solution of P(.lo) 623 Aoa.vm De soapsflom Henceueo gone eeeeaeo I eeeflen AHo.0m ee uceeeefiewemee .Amo.vm we unmeaeacwam. III opmfioo opoaoo mpoaoa opoaoa HmImH :emm. u mm eeflmoam.0I uemamam.0 ouoaoo namasom.= mIo momz mH.0 .0 Hz. a mp ouoamo 23500.0 opoaoa opmaoo HmeH III ouoaoo opoHoQ opoaoo ouoamo mlo 20mm: mH.0 .0 0m. u mm muoaoo m-u0.0I opmaoa ouoamo HmImH new. I mm :mmumm.0I opoaoa commzm.~I .050H0.H mIo 0Ho< oassm .0 III ouoaoo muofloo opoaoo ouoaoo HmImH cenm. u mm c.mmm0a.0 ouoaoo eewsmmm.mar auoawmm.0 mlo 0H0< oa>Hsm .0 III ouoaoo ouoaoa opoaoa opoaoo HmImH .Hm. . we epeaeo .mommm.ar epofleo apogee mro :oezmfl.o .6 am. I mm opoaoo opoaoo mmmmo.0I .mmmm0.0 HmImH .000. I mm ouoaoa ouoaoo momam.0I acmflom=.0 mIo aux u~.H .w coaumcaspouoo H< mm m: do no Acfiv muomnpxm ucofioauuooo co 2 no ADV aconmopwom suave no muowppxo Ca madame co 2 ho mcofimmopwoa Leonaa Lou m.coausaom HaEaqu .HH cocosdom .mfi wands 6A the best relationship between N, the dependent variable, and metals, the independent variables. In this analysis, all of the metals were in the initial regression equation. They then were deleted, one by one, if the significance probability of the F statistic for the least squares co- efficient for the metal was higher than the significance level of P(.lO). The above process was repeated at each step. The stepwise procedure was terminated when only metals meeting the deletion criterion were left. The cause and effect relationships implied in these regressions are subject to question. Nevertheless, the approach was used as a tool to see which metals strongly interacted with N. The results of multiple regression analyses within extracts are presented in Tables 1A and 15. Several sig- nificant to highly significant regressions were found, accounting for 65 to 95% of the variation in fractional recoveries of N. The partial regression coefficients in these two tables may be compared as to sign and level of significance with the simple correlation coefficients in Tables 10 to 13. In the case of the K01 extractions for exchangeable N and metals, only Ca and Mg were retained in the multiple regressions for both sampling depths. The relationship of N to Ca remained positive, as it had been in the simple correlations (Tables 10 and 12), and there was an increase in significance for the partial over the simple correlation 65 in the subsoil. There were several surface soil fractions where sig- nificant positive regressions of organic N on Ca were express- ed in the multiple functions but not in the simple correla- tions (extracts I-f, II-c, d and f). In each case, there was also an increase in significance for the coefficients of two other cations, usually accompanied by a change in sign for one of them. Thus, the overall positive correlation between organic N and Ca in surface soils (Table 9) was borne out in important soil fractions when the more variable interactions with Mg, Fe and A1 were also taken into account. Again this supports the inference made earlier that Ca was the dominant cation involved in precipitating and stabilizing organic N in these soils. Partial regression coefficients for organic N in sub- soil fractions were rarely significant at P(.05). Negative regressions on Ca and Mg were indicated for extract I-b. Regression coefficients significant at P(.lO) involved Ca in extract I-c and Fe and/or Al in a few other fractions. 5. Fulvic acid/humic acid separations The quantities of N and metals found in fulvic and humic sub-fractions are given in Appendix Tables A2 and A3 for the pyrophosphate extracts and in Tables AA and A5 for the NaOH extracts. In Table 16, the sums for the fulvic plus humic sub- fractions are compared with the parent extracts. Summed 66 Table 16. Comparisons of summed recoveries in fulvic and humic fractions with analysis made directly on pyrophosphate or sodium hydroxide extracts. Fulvic plus humic acid recoveries Element Depth for extracts and treatments (in) No manure 300T 3OT N;;;;0;_—N;OH NauP207 NaOH NauP207 NaOH ppm N 0-9 Extract 282 293 387 A29 A12 A73 F'+ H 289 323 39A 393 361 A59 Recovery 1021 1101 1011 921 881 971 18-21 Extract 36 56 28 60 AA BA F + H 59 69 37 66 70 78 Recovery 1631 1231 1321 1101 1591 1AA1 Ca 0-9 Extract 28 5 28 78 5 133 F + H 20 63 167 203 26 128 Recovery 711 1261 5961 2601 5201 961 18-21 Extract 5 5 19 27 5 9 F + H 20 20 96 133 20 23 Recovery A001 A001 5051 A921 A001 2551 Mg 0-9 Extract 5 23 12 35 5 29 F + H 20 27 AS 38 20 20 Recovery A001 1171 3751 1081 A001 691 18-21 Extract 6 32 11 3A 5 1A F + H 20 27 27 5A 20 22 Recovery 331 8A1 2A51 1591 A001 1571 Fe 0-9 Extract A7 287 A1 190 122 191 F + H 110 263 50 2A8 138 185 Recovery 23A1 921 1211 1301 1131 971 18-21 Extract 27 31A 18 158 26 128 F + a ~ 38 250 31 223 6A 139 Recovery 1A01 801 1721 1A11 2A61 1081 A1 0-9 Extract 71 880 59 1028 56 655 F + H 87 781 96 811 1A6 581 Recovery 1221 891 1621 791 2601 891 18-21 Extract AA 16A3 69 1223 61 775 F + H 102 13A6 101 1010 53 608 Recovery 2311 821 1A61 821 901 781 67 recoveries of N in the plow layer were in fairly good quanti- tative agreement with the analyses made on the parent extracts. The sums for the subsoil were greater than in the parent solutions. The lack of agreement for metals was great. In particular, the sums for Ca and Mg were frequently four or five times greater than in the parent extract. The source of these analytical difficulties was not resolved. The discrepancies for pyrophosphate were generally greater than for NaOH. About 80% of the time, the percentage recovery in pyrophosphate sub-fractions was greater than for NaOH; only about 13% of the time was the reverse true. There is the possibility that cations complexed with pyrophosphate or with soil P were less completely solubilized by DTPA after perchloric acid digestion of the parent extracts than in the case of the sub-fractions. DTPA was not used with A1, and excess recoveries for sub-fractions were much less than for Ca or Mg. On the other hand, DTPA was used with Fe, but recoveries were generally similar to A1. If the analytical procedures were, in fact, more sensitive to metals in sub-fractions than in parent extracts, it would have served to accentuate rather than detract from interactions already described between cations and organic N. At least, this inference is supported by the increased information and increased significance of regression coef- ficients in multiple regressions for fulvic and humic frac- tions as compared with the function for the parent NaOH 68 extract of surface soil (Table 15). 6. Fulvic/humic ratios In recent reviews, Stevenson and Butler (1969) and Schnitzer and Khan (1972) note that fulvic acids are more highly oxidized than humic acids, contain more oxygen, and much of the oxygen is present in surface functional groups of an acidic nature (-COOH, ,:%OH). Humic acids have a higher proportion of their oxygen content in interiorized carbonyl groups (C=O) and structural linkages (C-O-C). It appears that substantial quantities of polyvalent cations are trapped or bonded in the interior structure of humic acid molecules, since ash constituents, such as Al, Fe, Si, are extremely difficult to eliminate in preparing highly purified humic acids (Stevenson and Gascho 1968). This is true also of fulvic acids although to a lesser extent because of their smaller molecular size (Levesque and Schnitzer 1966). Due to their lower molecular weight and greater surface concentration of acidic groups, fulvic acids are more soluble and tend to be more mobile in soils than humic acids. Nevertheless, both form insoluble salts and complexes with polyvalent cations through electrovalent and chelate bonding to surface functional groups. Their salts with monovalent cations, such as Na and NHu,_are soluble, and this is in part the basis for their extraction from soils using alkaline reagents. Elimination of precipitating polyvalent cations through replacement with H+ in acid solutions and by removal 69 as insoluble hydroxides, salts or chelates at alkaline pH is also involved in extraction procedures such as those used here. The ratio of fulvic acid to humic acid provides an estimate of the degree of oxidation of soil organic matter (Kononova 1966), Schnitzer and Khan 1972). The fulvic/humic recovery ratios in Tables 17 and 18 show the distribution of N and cations between these two fractions. This sort of comparison for cations does not appear to have been reported in the literature. The actual quantities of cations in the parent pyro- phosphate extracts were very low because of the preceding extraction with H280” and additional removals by precipita- tion during the pyrophosphate extraction (cf. Figures 3, A). Thus, the fulvic/humic ratios in Table 17 are for the more tightly complexed or interiorly trapped cations. The range of ratios is smaller than for the NaOH sub-fractions in Table 18. In both cases, more of the extracted N was present in potentially mobile fulvic acid fractions in the subsoil than in the plow layer. A significantly larger proportion of tightly bound Ca in Table 17 was associated with fulvic acids in the plow layer than in the subsoil, whereas the reverse was true for Al. These relationships are even more striking in Table 18, where 8A to 92% of the extracted Ca in the plow layer and similar proportions of A1 in the subsoil appeared in the fulvic fraction for no manure and the 30T treatment. The 70 Table 17. Fulvic/humic recovery ratios for extract c (0.1M NauP2O7 + 0.1N NaOH), Sequence I. Elements Treatments Treatment No 300T 3OT within depth Manure (1 year) (9 years) LSD ('05) Plow layer (0-9 in.) N 1.3 1.0 1.1 ns Ca 1.0 1.3 1.6 ns Mg 1.0 1.0 1.0 ns Fe 1.5 1.7 2.9 ns Al 2.A 1.3 1.0 ns Subsoil layer (18-21 in.) N 2.2 2.2 l.A ns Ca 1.0 0.9 1.0 ns Mg 1.0 1.7 1.0 ns Fe 1.7 0.9 l.A ns Al A.l 0.6 1.7 1.5 Depth within treatment LSD (.05) N ns 1.1 ns Ca ns 0.3 0.3 Mg ns ns ns Fe ns 0.7 0.7 A1 0.9 ns ns +Mean of three replications. 71 Table 18. Fulvic/humic recovery ratios for extract b (0.1N NaOH), Sequence II. Treatments Elements Treatment No 300T 30T within depth Manure (1 year) (9 years) LSD (.05) Plow layer (0-9 in.) N 0.7 0.6 0.6 ns Ca 5.3 1.9 11.8 2.A Mg 0.7 0.5 1.0 0.3 Fe 0.5 0.3 0.5 ns A1 2.7 2.A 2.8 ns Subsoil layer (18-21 in.) N 1.2 1.0 1.0 ns Ca 1.0 1.5 1.3 ns Mg 0.6 0.2 0.7 0.3 Fe 0.2 0.3 O.A 0.2 A1 5.2 3.8 12.2 ns Depth within treatment LSD (.05) N ns ns ns Ca 2.0 ns 2.0 Mg ns ns ns Fe 0.1 ns ns Al ns ns ns +Mean of three replications. 72 proportion was very much lower for the 300T treatment, reflecting the much lower oxidation status of organic matter due to the massive manure addition and periodic waterlogging associated with it. This effect appears also for the more tightly bound pyrophosphate-extracted Al in the subsoil (Table 17) and for NaOH-extracted Mg in the plow layer and in the subsoil (Table 18). More of the tightly bound Fe in Table 17 was associated with fulvic than humic acids, whereas much of the additional Fe extracted with NaOH was apparently associated with humic acids (Table 18). The same trend was shown for Mg. It may be visualized that the tightly bound cations in pyrophosphate extracts are not associated with surface functional groups and, therefore, have little influence on surface properties of the fulvic or humic acid molecule. Those complexed in this way with fulvic acids will move with the fulvic acid if surface acidic groups are occupied by monovalent cations or if conditions of pH and chemical potential in soil solution are favorable for mobilization (Schnitzer and Khan 1972). On the other hand, the additional polyvalent cations found here in NaOH extracts are more likely to have been associated with surface functional groups and would act as coagulants to precipitate and stabilize fulvic and humic acids at soil pH's commonly associated with crop production. The data in Table 18 indicate that Ca was the dominant pre- cipitating cation responsible for stabilizing otherwise 73 mobile fulvic acids in the plow layer. In the subsoil, it was A1. 7. Carbon-nitrogen relationships Nitrogen was used as the measure of organic matter in all extracts. Due to time restrictions and lack of suitable procedures and equipment, carbon was determined only in the original soil and in residual soil before and after the final NaOH extraction (Residues a and b in Table 20). Nitrogen was determined in the original soil and in Residue (b) (Table 19). In the surface soil, both N and C were related signi- ficantly to treatment, both in the original soil and in resi- dues from both extraction sequences. The lowest values were for no manure and the highest for 30T annually over nine years. Simple correlations between N and C were significant at P(.Ol): r = 0.9A7 for the original soil, r = 0.928 for Residue (a) and r = 0.7A6 for Residue (b). Values in the subsoil were significantly lower than in the plow layer, but there were no significant differences for treatment and simple correlations between N and C were not significant at P(.05). Percent removals in Table 21 were calculated from the differences between original soil and Residue (b) in Tables 19 and 20. Percentage removals of N and C from the plow layer were similar and ranged from 55 to 6A%, with no rela- tion to treatment. These are consistent with commonly 7A .coapompuxo momz Hanan pound coxmp ADV oscamom+ .mpcoEpmoap :anfiz HH 0cm H .Uom coozuon moocoaoumau unonMchHm 02+ 0.0HH Ne. w: m: am we m.Hma h.wma >2: m.Hmm m.mm 0.0HH m.mm m: m.mm A20 osofimom Ham ~.HHN ~.HHN Haom Am0.v 0mg oocosuom canvas acmEpmoap manna: canon 50 mm 00H mmH 0ma +Anv osofimom 00H mmH mma Haom A.ca Hmumav gamma Haomeam was mmm NH: amm mmm +180 machete mmmH nooa Haw Hfiom A.CH mlov comma scam HH .eem H .eem HH .eem Amo.0 0mg Ameee mocoscom Canvas spate canes: I-IIIIIIIIIIIIIIIIIIIIIIII IIIIII I--- ..... Irena ..... I- IIIIII IIIIIIIIIIIIIIIIIIIIII +H .eem HH .eem +H .eem HH .eem +H .eem a m0 90m Apmomrav 900m madame oz ospamoa .HO deem acmEpmopB osnfiwop no Haom CH cowoaufiz .cofipomapxo momz Hmcfim gouge HHow Hmsofiwop :H 0cm HHow Hmcfiwfipo CH mmfiao>oooa cowoppfiz .mH magma 75 .CoHpomCuxo mowz HMCHM Couum memp A00 osuflmomw .Coapomhuxo momz Hmcfim opomon Coxmu Amy osvfimom+ .mquEuwon Canvas HH 0C0 H .dom Coozpon mooCoaomuHC quonandm oz+ WC mC mC WC mC NH.0 5H.0 mm.0 NN.0 mm.0 NH.0 0H.0 NH.0 0H.0 NH.0 0H.0 ADV osvfimom 0H.0 :m.0 0H.0 am.0 0H.0 m: Amy oSUHmmm sm.0 nm.0 hm.0 Haom Am0.0 000 ooCosuom Canfiz uCoEpmoCu CHCuas Cuooo 0H.0 0H.0 m0.0 :H.0 HH.0 00.0 ADV osvammm NH.0 ma.0 mm.0 0H.0 m0.0 no.0 Amv osnfimmm m:.0 Hm.0 NH.0 Hfiom A.CH HmImHV Comma HHomnsm m~.0 05.0 mm.0 mm.0 0m.0 :m.0 m ADV oscfimmm no.0 mm.H Hm.0 00.0 mm.0 0m.0 *Amv oscfimom NN.H nm.a 00.0 HHom A.CH mIov Comma 30am lil' II' I-“l".ll.l' l- I '. HH .uem H .eem Amo.0 0mg ooCosaom CfiCuaz Cuaov CHCuas uCoEuwoCB HH .eem H .eem HH .eem H .eem HH .eem H .eem e + + Awpmom 00 90m Anew» H0 900m oascme oz oscfimop Co ospfimoC so Hfiom CH Conpmo Haom oCOHPOMhuKO mowz HmCHu nouum Ccm oaomon HHom Hmsofimoa Ca ucm Haom HmCHmHCo CH moaao>ooop Connoo .0m manme 76 .muCoEumon CHCpH3 HH 0cm H .000 Coozpon ooCoaommao uCMoHMHCme oz+ mC mC mC mC :0 HH HH mm mC mC mm m: mC m0 mC mC mC 0m: mC 0 m: :H 3H 2H :H z Am0.v 0mg ooCosdom eHequ pcaepeeae CHCeHz spate :0 mH mm mm me o Hm mm em mm mm 2 A.CH HmImHv CeHeH HHOmnam mm mm H0 :0 H0 0 H0 :0 mm mm :0 z A.CH 0I00 Cozmfl 30am IIIIIIIIIIII R I'Il'l|.ll.l..l'.ll.lnl|lnll'|l|l.l.|lu.lnl HH .eem H .eem HH .eem eH .eem HH .eem +H .eem HH .eem _H .eem Amo.0 0mg “memes m0 90m Anew» H0 900m DHSCmE oz ooCosvom CHCuH3 Cease eHCeHz uCpEumohB mpCoEon mpCoEpmoCB .mCOHpomexo HprCosdom an vo>oEoC conpmo 0C0 CowopuHC HHom HmCHwHuo mo uCooHom .Hm magma 77 reported values (Kononova 1966). In the subsoil, a significantly lower proportion of soil organic N was removed by both extraction sequences for the no manure and 300T treatments than for 30T. A similar (although not significant) result was obtained for carbon, except for what may be an erroneously high value for 300T in Sequence II. In Table 22, C/N ratios have been calculated for original soil, extracted fractions and final residue. The unique feature in these data is the very wide C/N ratios for soil, extracted fractions and residues in subsoil under the 30T manure treatment. The soil and extracted ratios are significantly higher than in the plow layer. Soil samples from the 30T treatment were taken from plots harvested for grain in Experiment I. It is possible that the wide C/N ratios in subsoil reflect effects of carbonaceous organic matter returned in the form of stover. This inference is consistent with the larger retention of organic N in the surface and lesser movement into subsoils indicated in Table 3 for grain vs silage. If these subsoil C/N ratios were in fact greater than under silage harvest, they would also explain the significantly lower levels of NH“ recovered from subsoils under grain harvest (cf. Table 3 and text on page 3A). A similar wide C/N ratio was calculated for fractions extracted in Sequence II from subsoils under the 300T treat— ment. This value derives from the unusually low C analysis 78 .mpCoEumoCu .COHpoprxo momz HmCHu Locum Come A00 onwamom+ cHeSHa HH ece H .000 Coozpon ooCoCoumHU quoHuficme oz+ Hm.m Hm.w m: m: m0.0 mC mC mC mC mC WC mC Hm.mH Hm.mH m0.~ mm.0m 05.0m 0m.mm mn.mm ms.mm He.HH a:.0H 0m.mH mo.MH mm.:H mH.e me Hm.mH me mc m=.m HH.HH 0H.0m 05.ma 0m.mH 0H.0H ~0.NH mm.ma 00.:H mm.ma mC mC m8 WC mC H60 maeHmem mCOHCQMHm wouomppxm Hwom Am0.v 0mg ooCosdom eHceHz newscasts cHeeHz spate Hm.m :m.0 0:.0H 0m.0H mm.0 *Hev eaeHmem mCOHuompm omuomapxm afiow H.eH HmImH0 CeseH HHomnam >0.ma 03.:H 0m.ma mH.NH 0m.ma +Hev eseHmem mCoHpomau wopompuxm HHom A.CH mIov Cozwa zoam HH .aem H .eem Hmo.0 0mg moCosdom CHCqu spate :HLHH: pCoEpmoHe HH .eem +H .eem Anamoz my 90m HH .dom +H .Umm AaeeH H0 Boom HH .emm +H .eem ounces oz oschoC Ho HHOm CH moprH 2\o osonoH Co Coapompm couomexo .HHom .HHoa Hospawoa CH 0:0 .mcoHuomHm ooquHuxm .HHOm HCCHmHHo CH modumm 2\0 .Nm magma 79 for Residue (b) in Table 20 and is to be questioned. There is little basis for speculating what the inter- actions of C with cations may have been. The wide C/N ratios in subsoil for the 30T treatment were associated with substantially higher levels of extractable Ca and lower levels of Mg, Fe and Al than in subsoils from the other two treatments (cf. Figure 2, Table A6). The functional groups which can interact with cations are associated with C rather than N in fulvic acid and humic acid molecules. It is, therefore, essential that accurate C determinations be used as the measure of organic matter in future studies of this sort. CHAPTER V SUMMARY AND CONCLUSIONS Profile distributions of N and P were examined in sandy loam soils to a depth of 10 feet after nine annual applications of 10, 20 and 30 T/a beef cattle manure or variable inputs of N and P in commercial fertilizer for continuous corn (Experiment I). In a second closely adjacent experiment, profile distributions were examined, to a depth of 5 feet, eight months after additions of 100, 200 and 300 T/a manure for corn. Fractional distributions of N, Ca, Mg, Fe and A1 were examined in the plow layer and in the 18-21 in subsoil increment from plots which received no manure and 300T in a single application in Experiment II and 30T annually over nine years in Experiment I. Two fractionation sequences were used: Sequence I (1.7% KCl + 0.1N H280“ + 0.1M NauP207 in 0.1N NaOH + 0.1N H280“ + 0.1N NaOH); Sequence II (1.7% KCl + 0.13 NaOH + O-lfl H280“ + 0.191 NaOH). The alkaline pyrophosphate extract of Sequence I and the first NaOH extract of Sequence II were separated into fulvic acid and humic acid sub-fractions, Except for the exchangeable frac- tion (1.7% KCl), fractional extracts were subjected to per- chloric acid digestion to remove organic matter before analyzing for the metal cations. In the long-term Experiment I, more organic N was retained in the plow layer (0-9 in.) where corn was 80 81 harvested for grain and stover was returned than where silage was removed. Downward movement of organic N (detectable to depths of A to 6 feet) was greater under silage harvest and increased with increased additions of manure; it was also greater where fertilizer N, P and K were used without lime. Nitrate and total mineral N in the plow layer were higher with grain harvest, whereas NH“ and total mineral N were lower in the subsoil than with silage harvest. This suggests closer cycling of N due to carbonaceous stover residues and perhaps greater losses of N by denitrification over the years. Extractable P (Bray Rél)in.the plow layer increased with rate of manure addition, but not to levels as great as where com- mercial fertilizer supplying 190 lb/a P205 annually was used. However, downward movement of P, detectable to 15 in, was greater with manure than with commercial fertilizer. There were no control plots in Experiment I, so re- coveries from soil of N and P could not be compared with inputs. In Experiment II, however, the increase in organic N in the plow layer (3,500 lb/a) for 300T/a manure applied eight months earlier would account totally for the estimated 3,300 lb N applied. There was some evidence for downward movement of organic N to depths up to 15 or 18 in. Ammonium in the plow layer was similar in the two experiments, with no relation to treatment; nitrate was much higher in Experi- ment II, ranging from 50 to 19A ppm NO3-N for manure treat- ments, as compared with 9 to 32 ppm for applications of 10 to 30T in Experiment I. Nitrate concentrations below about 82 27 inches were again similar at comparable depths to 5 feet in both experiments (7 to 17 ppm NO3-N for 100 to 300T manure in Experiment II vs 7 to 11 ppm for 30T with silage harvest in Experiment I). In the subsoil, NH“ concentrations in 3- inch increments to 5 feet were significantly lower for 200 and 300T applications than for no manure; this appeared to be due to a lower oxidation status due to temporary loss of infiltration capacity in the plow layer and periodic ponding associated with these massive manure additions. Extractable P in the plow layer increased with manure application in Experiment II, but there was no evidence for downward move- ment out of the plow layer; the increase for 300T was 1A0 lb/a P, or about 16% of estimated input. Only about 15% of the Ca and seven percent of the Mg applied in 300T of manure could be accounted for in exchange- able forms in the plow layer. Nonexchangeable but extract- able forms were equivalent to 139% of input Ca and 92% of input Mg. Most of the nonexchangeable Ca and Mg appeared to be present in minearal salts or complexes and was found in H280“ extracts. Nevertheless, Ca and Mg were present, together with larger quantities of Fe and much larger quantities of Al, in organo-mineral complexes which contained 27 to 58% of the total soil organic N and were removed in pyrophosphate or NaOH extracts. There was a strong tendency for the quantities present in these complexes to be related directly to valence and inversely to ionic radius of the metal cation in the order: Ca < Mg < Fe < A1. 83 Significant interactions among the cations themselves and with N appeared in simple and multiple correlation analyses of data for individual fractions and for their sums in the two extraction sequences. These correlations indi- cated that Ca was the dominant cation responsible for pre— cipitating and stabilizing organic N in the plow layer. This inference was supported by fulvic/humic recovery ratios which showed that 8A to 92% of the Ca in NaOH extracts was associated with the fulvic fraction. In the subsoil similar percentages of the NaOH-extracted Al appeared in the fulvic fraction, indicating that Al was principally responsible for precipitating and stabilizing fulvic acids in the subsoil. Fulvic/humic recovery ratios for all cations were sharply reduced at the 300T level of manure addition as compared with no manure or 30T. This effect was expressed either in surface or subsoil and, in the case of Mg, in both. As with NH“ levels in subsoils of Experiment II, this re- flects the lower oxidation status of organic matter due to reduced infiltration and periodic waterlogging which occurred with this massive manure application. Although only a small proportion of the total Ca ex- tracted appeared in the same fractions with organic N, total extracted N and Ca in both sequences were positively cor- related at P(.Ol) in the surface soil. This indicates that Ca in mineral fractions extractable with H28014 were also influential in stabilizing N. Calcium, overall and in extracted organo-mineral 8A fractions, was negatively correlated with Fe, both in sur- face and subsoils, and tended to be negatively correlated with A1, particularly in subsoils. Interactions involving Mg were more variable but were also consistent with the inference that cations compete for similar sites in organic matter. Nitrogen undoubtedly competes for some of the same sites but no strong evidence for this was found. Much smaller quantities of cations remained in associa— tion with N in pyrophosphate extracts than in NaOH. It was inferred that these are tightly complexed or trapped inside organic molecules and would be subject to movement with fulvic acids under appropriate conditions. The additional quantities of cations in NaOH extracts are more likely to be associated with surface functional groups and to act as coagulants for precipitating organic matter. This postula- tion appears to be promising to pursue in further research. For this purpose, the four extraction schemes in Figure 5 are proposed. To avoid problems of interpretation encountered in this study, soil samples should be taken from a single experi- ment and from the same soil type. Analytical difficulties in quantitating cations in the presence of phosphate in extracts will need to be resolved. Both C and N should be determined in all fractions and in original and residual soil, using procedures tested for reliability on similar preparations. 85 ACHssmv osnfimom Ham Com CoCHmuno ohm moumCCodsm Cmoao HHHCJ momz mH.0 _ moez mH.o 0 eommm mH.o «z I TIImomz 2H.0 <0 % 0mm 0 om Hex HH.H >H mocoscom _ .mooCosdom COHHoMCuxo Com Ewpwmfio BOHm .m oCstm ACHECmv osvfimmm l _ moez mH.o 0 sommm mH.o Ar em I +IImoez 2H.o «a 20mm: mH.o 0mm 0 cm > Hoe HH.H xi HHH moCoscom _ ma m.asma omom comm smhm oaam swwm zamm m.o~=a o.omma. o.omma whom sea: mosm msom mean wmom m.mza m: a: mum mpm mow Hos mmHH mmHH an n: a: amaa Nam zmm woza mmm «mm m: n: on mu mm 5m moa um wh H< on w: wo z A.Ca Hmlmav gamma HHOmnsm a: an: on wmma mmmm omow anam oowm ammm H< a: a: a: omow mzmm smmm mmam mamm mzmm cm as n: a: mu: cam mmm was omm «mm m: a: a: a: mmaa mom 2m: HHmH omoa :mm do a: ~.mmH ~.mmH mmw wmm mam mam mmm wmm z A.:« mnov gamma roam Eon Amo.v omq HH .oom H .com Amado» av Anew» av caucus Amado» mv Ranch av chance canoe can eom Boom oz 90m Boom oz uncapaeu» Amo.v emu canon Casual new oocosaou oocosuom cane“: upcoEpaops HH mucosaom H mucosaom ucosoam +.nooco:oon coauoapuxo Haapcuaaou Hoaaapaa oz» ea madame can 2 manaowcmnoxococ uo uoaao>ooou Hmuoe .m: manna "I11111111111111111“