PROHiE DISEREBUTION 0F MINERAL NROGEN {N RELATION TO RATES OF FERFEUZER APPLICATION Thesis for the Degree of Ph. D. MECEIGAN SEATE UNIVERSWY MEHMEF RIFAT DERICt 1975 v ,.' "_, ' 1n . '~ -n .12 .t.’\ 4' a " D p f f L _- ,.. . ‘-2 ~ 939831 33:4. ". -. .7 ‘ r7 » . . \.- “,7 '2.. .768. .' This is to certify that the thesis entitled PROFILE DISTRIBUTION OF MINERAL NITROGEN IN RELATION TO RATES OF FERTILIZER APPLICATION presented by Mehmet Rifat Derici has been accepted towards fulfillment of the requirements for Ph.D. dpgvpin Crop & Soil Sciences A. R. Wolcott Major professor Signed in his absence gafiemfi l7. 14»?ng Date August 8, 1975 0-7639 ABSTRACT PROFILE DISTRIBUTION OF MINERAL NITROGEN IN RELATION TO RATES OF FERTILIZER APPLICATION BY Mehmet Rifat Derici A one year field experiment was conducted to investi— - + 3 and NH4 varying rates of nitrogen application. Urea was applied in gate the profile distribution of NO under corn, with early spring at rates of 0, 84, 168 and 336 kg N/ha. Chloride was applied uniformly to all soils as a reference anion. Soils were sampled at incremental depths to 150 cm before fertilizers were applied, twice during the first season after application and again the following spring. Significant increases in corn yield were obtained only with the first increment (84 kg) of N. The next two incre- ments resulted in progressively smaller yield increases. It appeared that NH: released by hydrolysis of urea had been taken up by the crop, immobilized or nitrified within two months of application. The profile distributions of NH: were essentially the same as background for all treat- ments throughout the experiment and paralleled the probable distribution of organic matter in soil. Mehmet R. Derici Large quantities of Cl- were found within 15 to 60 cm of the soil surface 2 and 5 months after the application of fertilizers. By the following spring, Cl- had disappeared from 150 cm of profile in quantities approaching input. Changes in NOS/Cl- ratios at the end of the growing season and in the following spring indicated contributions of soil derived N0; in the upper part of the profile. There was a general correspondence between NOS/Cl- ratios and NH: dis- tribution in soil before and one year after the application of fertilizers. It appeared that soils had the capacity to retain low 3 the leaching action of water. Minimal concentrations for concentrations of both Cl- and NO against crop depletion or Cl- varied over a range of 7 to 12 ppm over the profile. For NOS—N, minimal concentrations ranged from 2 to 4 ppm in the lower profile, increasing to 8 or 9 ppm in the surface 15 cm. At harvest time, NOE-N in the upper profile had been drawn down to levels approaching these minimal concentrations by corn which had received no fertilizer N or 84 kg or 168 kg N/ha. These concentrations, and totals for NOS-N to 150 cm, were essentially unchanged after a winter and spring when precipitation exceeded normal precipitation by 70%. 3 of input was retained in the upper profile at harvest time At the 336 kg N rate, surplus NO equivalent to 40% but had disappeared by the following spring, leaving minimal residual concentrations similar to the other treatments. Thus it appeared that a 40% reduction in input (from 336 to 200 Mehmet R. Derici kg N/ha) would have given the same degree of control over No; in the soil system as the use of no fertilizer N at all. The resulting sacrifice from maximum yield would have been much less than experimental error in field experiments such as this. PROFILE DISTRIBUTION OF MINERAL NITROGEN IN RELATION TO RATES OF FERTILIZER APPLICATION BY Mehmet Rifat Derici A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences 1975 ACKNOWLEDGEMENTS I sincerely thank Dr. A. R. Wolcott for his valuable guidance and suggestions in the course of this work. I also thank Dr. E. C. Doll for his suggestions in the initiation of this research, Dr. M. L. Vitosh for his financial support in the computer use, and the consultants of the Michigan State University Computer Laboratory for their help in statistical analyses. I extend my thanks to my wife, Emel, for her moral support and for her help in the laboratory. Finally, I am deeply grateful to the people of Turkey for their financial support. ii TABLE OF CONTENTS ACKNOWLEDGEMENTS. . . . . . . . . . . . . . . . . LIST OF TABLES. . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . INTRODUCTION. . . . . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . . . . . General Concepts of Movement . . . . . . . Movement of Nitrogen in Soils. . . . . . . Movement of Nitrate in Soils . . . . . . . Factors Affecting the Movement of Nitrate. Soil Factors. . . . . . . . . . . . Plant Factors . . . . . . . . . . . Climatic Factors. . . . . . . . . . Fertilizer Practices and N Sources. METHODS AND MATERIALS . . . . . . . . . . . . . . Field Procedures . . . . . . . . . . . . . Laboratory Procedures. . . . . . . . . . . Statistical Procedures . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . Chemical Analyses. . . . . . . . . . . . . Background. . . . . . . . . . . . . Nitrate . . . . . . . . . . . . . . Ammonium. . . . . . . . . . . . . . iii Page ii vi 10 11 13 16 16 20 21 23 23 23 26 Chloride. . . . NOS/Cl- Ratios. Crop Yields. . . . . . DISCUSSION. . . . . . . . . . Movement of Chloride . Nitrogen Relationships Ammonium. . . . Nitrate . . . . NOS/Cl- Ratios. Practical Implications CONCLUSION. . . . . . . . . . LIST OF REFERENCES. . . . . . iv Page 29 33 35 37 37 40 40 44 45 46 49 50 LIST OF TABLES Table Page 1. Distribution of rainfall and irrigation during 1973 growing season. . . . . . . . . . . . . . . . 18 2. Monthly distribution of rainfall and irrigation from April 1973 to April 1974. . . . . . . . . . . 19 3. Partitioning of degrees of freedom and mean squares, F ratios and probabiliti s for F in analyses of variance for NO3 and NH4 . . . . . . . . . 22 4. Background analyses. Means and standard dev- iations for 10 cores sampled April 21 and 22 before fertilizer applications on April 24, 1973 . . . . . . . . . . . . . . . . . . . . . . . 24 5. Vertical distribution of NOS-N in soils under four N treatments on three sampling dates. . . . . 25 6. Quantities of NO-, NH+ and total mineral N recovered in soiI pro iles to 150 cm . . . . . . . 27 7. Vertical distribution of NHZ-N in soils under four N treatments on three sampling dates. . . . . 23 8. Vertical distribution of C1. in soils under four N treatments on three sampling dates. . . . . 30 9. Vertical distribution of NOS/Cl— ratios in soils under four N treatments on three sampling dates . . . . . . . . . . . . . . . . . . 34 LIST OF FIGURES Figure Page 1. Yield of corn as influenced by rates of N application. . . . . . . . . . . . . . . . . . . . 36 2. Depth distribution of c1" in soil with varying rates of N application at three sampling dates . . . . . . . . . . . . . . . . . . 38 3. Depth distribution of NO--N as influenced by rates of N application at three sampling dates. 0 O O O O O O O O O O O O O O I O I O O O O 41 4. Depth distribution of NHz-N in soil as influenced by rates of N application at three sampling dates . . . . . . . . . . . . . . . 42 5. Depth distribution of NO-/Cl- ratios in soil as influenced by rates of N application at three sampling dates . . . . . . . . . . . . . . 43 vi INTRODUCTION The criticism of the use of N fertilizers has increased as environmentalists and health authorities have become more and more concerned with N0; pollution. Undeniably, the use of N fertilizers can contribute to N0; pollution. Under different cropping systems, soils and climatol- ogical conditions, the crOp utilization efficiency varies from less than 50% to about 70% (Owens 1960, Broadbent and Nakashima 1968). There is always a quantity of nitrogen which cannot be accounted for in N balance studies. Part of this loss is due to leaching of the N03 out of the soil profile. Bingham gt El“ (1971) reported that annual leaching loss was 60 lb N/a in a citrus orchard. This loss represent- ed about 45% of the input fertilizer. 3 soil for 15 years. Considerable quantities of NOS were found McGregor gt 31. (1974) followed NO movement into sub-V at 10 m depths following the continued annual fertilization. Although most of the fertilizer N was in the first 8 m of soil, they assumed a continued downward movement with time. According to the Environmental Fact Sheet No. 10 (September 1970) released by the American Fertilizer Insti- tute, N fertilizers represented only 13% of the total N sources which might contribute to environmental pollution in the United States in 1970. This small percentage reduces the share of the responsibility of N fertilization for the pollu- tion problem. However, the goal of soil scientists is to reduce the pollution hazard due to N fertilization. The objectives of this research were to monitor the profile distribution of mineral nitrogen in soil after vary- ing rates of N fertilization and to relate observed distributions to those for C1- as a reference anion and to yields as an index of crop utilization. LITERATURE REVIEW General Concepts of Movement Movement of any chemical substance in soil can be described as its lateral or vertical displacement in time. This displacement then results in a new concentration distri- bution throughout the soil profile. The mechanisms by which the movement takes place are numerous. However, many of these mechanisms either take place under very special conditions, or they do not affect the overall movement significantly. Therefore only three major mechanisms will be discussed here. The first mechanism is the relocation of any substance due to the mechanical stirring of soil by cultivation. This type of movement is confined entirely to the plow layer and has little immediate importance, except that it may provide more uniform distribution of substances in the tilled volume of soil. However, it does affect the nature of the subse— quent movement by other processes. Secondly, and probably the most important mechanism of movement, is mass flow. When the soil solution is moved by mass flow it carries the substances dissolved in it. In this process the amount of a non-interacting substance passing through a unit area per unit time is a function of its concentration and of water flow. The third mechanism of movement is the diffusion of ions or molecules of a substance in soil solution or soil atmos- phere due to a concentration gradient. In a soil system both liquid and gaseous phase diffusion can be expected. The rate and extent of each is partly determined by the soil water content and the volatility of the substances. In the case of the more important liquid phase diffusion, Fick's law states that the flux of a substance is a function of its concentra- tion distribution along the path of diffusion, and it depends on the water content and diffusion coefficients of soil. Mass flow and diffusion processes are simple only when uniform transport media and non-reactive substances are considered. In a soil system the above processes are highly complicated by the heterogeneity of the soil and its reactive components. A detailed and rather mathematical discussion of diffusion, mass flow and reaction with soil components, and the simultaneous occurrence of these processes, can be found in Gardner (1965). From the standpoint of soil fertility, the most import- ant consequence of the above mentioned movement processes is the leaching of nutrients and other agrochemicals. Movement of Nitrogen in Soils Almost all forms of soil N are reactive with the biological and the chemical components of soil. A dynamic equilibrium between the various forms of N in soil exists at any moment. The equilibrium concentrations of these forms are greatly affected by a vast number of factors. Thus, it is not easy to follow the movement of individual forms of N in soil. . . . . - + Nitrogen IS present in 3011 as NO3 and NH4, NH and 3: 'soluble organic forms, such as free amino acids, amino sugars and amides, and bound N in organic matter. Some NH: may be fixed in interlamellar spaces of expanding layer silicates or in insoluble complex salts with Mg++ and P. Except for nitrogen bound in immobile organic forms and a much smaller quantity fixed in mineral complexes, both mineral and organic forms of soil nitrogen are susceptible to movement. However, the reversible or irreversible reactions of most forms slow their movements. Furthermore, some of these forms are either intermittent in existence or are present at very low concentrations. So far, NO- has been considered as the most mobile form 3 of soil N. Movement of Nitrate in Soils 3 any physically similar anion. However, the concentration of In many respects the movement of NO resembles that of NO3 at a particular place in soil is determined to an impor- tant extent by the reactions of the N cycle. Therefore, any factor that affects these reactions also affects the movement of nitrate, at least indirectly. Several attempts have been made to establish mathemati- cal models for N0; movement in soil (Levin, 1964; Gardner, 1965; Kirda £3 31., 1974; Starr gt 31., 1974). These models cannot yet be applied to field soils with great certainty. However, along with the experimental evidence, they reveal certain patterns of NO3 movement. Movements of N0; and of another physically similar but not so reactive anion, such as C1-, are often compared. 3 jected mostly to the physical processes, while N0; is, in Applied simultaneously with NO to the same soil, Cl- is sub- addition, subjected to transformations into other forms of N. The comparative study of the concentration distributions of the two can provide useful information on the behavior of N in soils. This technique was used extensively by Wetselaar (1961, 1962). Experimenting with columns of soil, Corey gt El' (1967b) found that N03 moved slower than C1- in surface soils contain- ing organic matter. In subsoils the velocities of the two anions were similar. They did not attribute the different movement of the two anions in the presence of organic matter entirely to microbial activity. They stated that dispersion may be an important factor. The degree of dispersion was 3 than for C1. in the presence of probably different for N0 organic matter. Factors Affecting the Movement of Nitrate The numerical values assigned to the rate and extent of N0; movement are usually meaningless, per se, under field conditions. The diversity of the results obtained by differ- ent researchers on similar soils reveals this fact. Perhaps one good approach to the study of N0; movement in field soils is to begin with the study of the factors acting upon it. However, the individual effects of these factors cannot be easily studied except under laboratory conditions. Soil Factors The rate and extent of major mechanisms of movement, namely, mass flow and diffusion are determined to a great extent by the physical properties of soil. Most important of these properties are soil texture, structure and water content. In the simplest sense, soil structure,affects the hydraulic conductivity and thus the mass flow movement of solutions. Such a way of thinking may be valid for homogeneous systems. In field soils, however, one needs to be concerned with the entire geometry of the system. This geometry is the result of the simultaneous interaction of the above properties. Gardner and Brooks (1956) described the role of pore size distribution in the leaching process. According to their discussion, the homogeneity of the soil system is determined by the narrowness of this distribution. Heterogeneity is introduced into this system by the presence of large pores or pore combinations and blind pores. Such heterogeneity gives rise to velocity differences and thus to mixing of the displacing and the displaced solutions. The end result is dispersion. Wetselaar (1962) refers to this process as the dilution, rather than complete displacement of soil solution, by the water moving through the profile. Terry and McCants (1974) summarized the findings of numerous investigators who studied the movement of ions leached from applied fertilizers. These findings indicate that such movement is similar to a normal distribution when the soil does not possess significant granulation. Terry and McCants also observed a similar distribution of NO3 ions from 5 cm wide fertilizer bands. The total lateral and below-band distributions of N0; were normal when the fertilizer was banded on the soil surface. When the bands were placed 10 cm below the soil surface, the distribution patterns were compli- 3. Working with Cl_ and tritiated water, McMahon and cated by the upward movement of NO Thomas (1974) found that the solute movements were different in disturbed and undisturbed soil columns. The disturbed columns were packed to resemble the natural density and stratification of the undisturbed columns. They detected some by-passing of soil by the percolating water or Cl- solu- tion. This by—passing was due to the presence of natural cracks or large pores in the undisturbed columns. The work done by Thomas gt 31. (1973) demonstrates the effect of one important soil management practice which changes the physical properties of soil. They found that substan- 3 cm of a soil which was mulched with killed sod than from a tially higher amounts of NO were being lost from the top 90 conventionally tilled soil. Denitrification or nitrification was insignificant in both soils. Therefore, they attribute 3 tions. They suggested, first, that in the mulched soil the the higher leaching rate for NO to the changed physical condi- upward capillary movement was restricted due to the lowered 3 capillary action. Second the mulched soil was moist and evaporation. Therefore NO could not be brought up by well aggregated. The rainwater, carrying the N0; from topsoil by-passed the aggregates through the pores and moved deeper. On the other hand the conventionally tilled soil was drier and not aggregated as much. Thus, both water and dissolved NO3 moved into the aggregates instead of moving downwards. Wild (1972) found that the leaching of NO; formed by mineralization was slower than expected in a bare fallow, even in a rainy season. The rainwater was moving rapidly through the large cracks and channels without entering the aggregates in which NO3 was forming. Soil moisture content affects the movement of N0; and other solutes by setting up the conditions for flow. Flow under saturated and unsaturated conditions is discussed in detail by Biggar and Nielsen (1965). Ghuman et 31. (1975) found that, in initially dry soils, salt and water fronts 10 coincided when displacing a salt solution with water. But, in moist soils, the salt front lagged behind the water front. Increasing the initial water content resulted in a broader peak for distribution of the salt in the soil profile. The type and amount of soil colloids and the presence of energy rich compounds also affect the movement of NO 3 Anion exclusion, or negative adsorption (Thomas and Swoboda, F5 1970; Smith, 1972; McMahon and Thomas, 1974) and anion adsorp- 9 tion (Thomas, 1963) may occur in soil to some extent. These V. processes slow down or speed up the movement of NOS. Corey gt gt. (1967a) studied the effect of sucrose H content and the flow velocity on No; movement in soil columns. Increasing the sucrose content and/or decreasing the flow velocity of NO3 solution resulted in an increased dis- appearance of N0; from soils. This was due to either denitri- fication or immobilization of N03. Plant Factors Rooting habits of plants and their gross or seasonal N requirements have a marked effect on the fate of applied N. The competition between plant uptake and the leaching process, and the depletion of soil water by evapotranspiration reduce the N0; leaching rates. Bolton gt gt. (1970) measured the nutrient losses through tile drains under three cropping systems for seven years. With 336 kg N/ha/year applications, the annual losses of N varied from 0.7 kg/ha for continuous bluegrass to 6.6 kg/ha for continuous corn. Although the corn plots had 11 received an additional 112 kg N/ha as sidedressing, the leach- ing losses under these plots were still considered to be both absolutely and relatively higher. Power gt gt. (1973) grew corn and bromegrass for four years, with annual applications of 100 kg/ha as Ca(NO3)2 or urea. They then grew barley until there was no residual effect on yield of the previously applied N. They noted a marked accumulation of NC; at 150 to 250 cm depths. No significant accumulation was found under bromegrass plots. They pointed out the importance of planting dates in relation to the periods of maximum precipitation and maximum water use. Climatic Factors Intensity, duration and the seasonal distribution of rainfall are important factors in leaching of NOS. Wetselaar (1962) measured 5.5, 14.5 and 27.6 in of anion movement for 7.4, 15.4 and 23.7 in of rainfall, re- spectively, in a clay loam soil. The mean movement of N0; was 1.075 in/in of rainfall for this soil. The mean movement of C1— in two sandy soils was 0.99 and 2.12 inches. Since the former soil had high silt content in the surface, he assumed that surface runoff reduced the amount of water infiltrating the soil. Owens (1960) used the amount of water actually passing through the soil profile in his rate estimations. He reported that 5 to 20 percent of the N was leached in one season. The leaching rates were directly proportional to the amount of water percolating through the soil. The leaching rates 12 ranged from 1 to 1.19 lb/a per inch of rainfall in excess of 12 inches. 3 tions must be based, not on the amount of rainfall, but on Jones (1975) suggests that NO leaching rate estima- the amount of water moving through the soil. 'In his experi— ments the average leaching rate was 0.52 cm/cm rainfall. But he estimates an upper limit of 2 cm/cm of percolating water for his soils. The direct effect of temperature on the movement of any anion in soil is thermodynamic, and the discussion is be- yond the scope of this text. The indirect effects, however, are significant. Temperature along with other environmental factors, controls the rate of N0; production or utilization. It is generally accepted that the biological transfor- mations are at a minimum during winter months. Furthermore, in temperate regions, the topsoil is frequently frozen in winter, allowing no movement of NO3 in surface horizons. In the unfrozen part of the profile, the movement is at a minimum because of restricted water flow from upper horizons. During the transition time when the soil is freezing or thawing, an interesting.mode of N0; movement is observed. Campbell gt_gt. (1970) detected an increase in No; content in the 0 to 30 cm depth during early winter months. Since nitrification was depressed by the added N—serve and by the low temperatures, they suspected an upward movement of NOE. Their suggestion was that, during freezing of tepsoil, a suction gradient was created. This 13 suction resulted in an upward movement of the soil solution, 3 movement of water to a frozen zone under field conditions carrying NO with it. Ferguson gt gt. observed upward when the soil-water tension was less than 5 atm. It appeared that suction gradients of about 5 atm were created as a result of freezing. Fertilizer Practices and N Sources E} The type of application of N fertilizers to soil has at least an indirect effect on the movement of NOS. First, it determines the initial concentration of N carriers, thus if influencing the rate of N0; production. Second, it determines the initial surface distribution of the fertilizer and its subsequent vertical or lateral movement and depth distribu- tion. The first point is well illustrated by the experiments of Wetselaar gt gt. (1972). The equivalent of 750 kg N/ha as 4)ZSO4 was applied in 1 cm wide bands, 15 cm below the soil surface. Following the fertilizer placement, urea or (NH a "normal" concentration distribution of NO2 and NH2, peaking at 15 cm depth, was observed with (NH4)ZSO4. With urea, the distribution of NH: was similar but the concentration of NO2 was less than 1 ppm throughout the profile. Nitrate distribution, however, was quite different. 4’2 3 weeks after the application. After eight weeks, a concentra- With (NH 804, NO concentration remained very low for four tion peak had moved to within 5 cm of the surface. There was a much smaller secondary peak at 15 cm. With urea the peak 14 moved upward gradually, and there was a secondary broad peak around 20 cm depth. The authors concluded that the increased concentration of fertilizers due to banding had a regulatory effect on the activity of nitrifying organisms. This regula- tion was due to osmotic effects and pH changes, and was dependent on the source of fertilizers. The nitrification of NH: was taking place at a distance from the point of placement. The source of NO3 was moving rather than the N03 itself. Felizardo gt gt. (1972) applied 112 kg N/ha as various fertilizers broadcast over the entire surface or over 0.5 m bands or over 0.125 m hands. This way they created a 1 to 10 to 100-fold increase in local concentration of fertilizer in the soil surface. Under the total area broadcast applica- tion, NO3 leached to 30 cm depth with (NH 804, to 60 cm 4’2 depth with NH4NO3 and to 90 cm depth with Ca(N03)2 months. This carrier dependent movement pattern was more in five marked with band placements. Nitrate concentrations for a given carrier were considerably higher at given depths, especially with narrow band placement. However, the original 1 to 10 to 100-fold concentration relationship was not apparent in the final depth distribution of NOS. Plant uptake and lateral movement was assumed to be the cause of this. Burns and Dean (1964) developed and tested the hypothe- sis that, with soil moisture contents above field capacity, water is pulled into the fertilizer hands by osmotic suction, 15 creating a highly concentrated solution. This solution then moves downwards due to its higher density. They called this phenomenon "drop out" and discussed its significance in N03 leaching. Time of application, source, and rate of application of fertilizers with respect to crop type, planting dates and the rainfall regime are of great concern in N efficiency studies. Herron gt gt. (1971) observed that the delayed application of N fertilizers increased the crop use efficiency, and resulted in more residual N in soil. Fertilization in the spring caused substantial losses through leaching or denitri- fication. Chalk and Keeney (1975) compared the effectiveness 15 of fall and spring applied N labelled aqueous NH With 3. fall applications, some fertilizer derived NC; was found 60 to 75 cm below the point of placement in the following spring. With respect to application rates, Linville and Smith (1971) reported that rates which exceed crop needs increased the possibility of downward movement of NO3 in soil. Protasov and Korostoleva (1972) studied the losses from various N fertilizers. They found that more No; leached into deeper layers from N0; than from ammoniacal sources. Source of N fertilizers may influence the physical properties of the soil and thus the movement of the produced N03. The N fertilizers with monovalent cations such as NaNO3 and NH: fertilizers change the permeability of the soil due to their deflocculation effects (Fireman gt gt., 1945; Fox gt 2.1.- , 1952). METHODS AND MATERIALS A one year field experiment was initiated in May 1973 to study the movement of N in soil under corn, with varying “I rates of N application. Chloride was also applied to the soils for purposes of comparison. E Soil samples were taken at incremental depths three 51 times during the experiment. The samples were analyzed for + 4 of statistical tests and graphical analysis. NOS, NH and C1_. The results were evaluated with the aid Field Procedures Experimental plots measuring 15 ft x 25 ft (4.6 x 7.6 m) were laid down on the Michigan State University Soil Science Research Farm in East Lansing. Four treatments were replicated four times in randomized complete block design. The treatments were: Check, 84, 168 and 336 kg/ha of N as urea. Regardless of the treatment, each plot also received 560 kg/ha Cl-. Of this amount 224 kg was in the form of KCl and 336 kg was as CaCl Following the broadcast application 2. of all fertilizers on April 24, the plots were plowed and then disced to provide uniform mixing in the topsoil. In May 1973, after the fertilizers were applied, corn (Zea mays L.) was planted on 32 in (81 cm) rows along with 16 17 200 1b/a (224 kg/ha) of 0-26-26 (0-11.4-21.6 N-P-K) fertili- zer, banded 5 cm to the side and 5 cm below the seed. This planting time fertilizer provided an additional 44 kg of Cl-, making the total C1- input 604 kg/ha. The plots were sprinkler irrigated at times of low soil moisture. The daily rainfall data, and the dates and rates of irrigation during the growing season are presented in Table 1. Monthly totals of precipitation and irrigation waters in the experimental period are listed in Table 2. Soil samples were taken July 6 to 13 and September 24 to October 2 in 1973, and May 6 to 12 in 1974. Five random borings were made on each plot using a 3-in hand auger. Depth samples were taken from each boring at 15 cm intervals from 0 to 60 cm and at 30 cm intervals from 60 to 150 cm. The bulk of soil obtained from each increment was passed through 0.5 cm mesh screen, thoroughly mixed and sub- sampled once. All depth samples from individual borings were taken separately in sealed plastic bags. Immediately after collection a few drops of toluene were added, and the samples were stored in a freezer. The experimental soil was a Hodunk sandy loam (Ochrep— tic Fragiudalfs). This soil series consists of moderately well-drained, gray—brown podzolic soils, developed on calcareous sandy loam glacial till, with a fragipan typically at 40 to 64 cm. At this location, soil materials encountered (including the fragipan) were sandy loam in texture to a depth of 75 cm, changing to a sandy clay loam below this depth. 18 Table 1. Distribution of rainfall and irrigation during 1973 growing season. Date April May June July August September ___________ cm _ _ _ _ _ - _ _ _ - _ _ _ 1 2.79 1.71 .13 2 .23 .38 a 3 .23 1.27 (2.54) 4 .23 1.14 (2.54) 5 .25 6 .25 7 8 1.78 1.91 9 .66 .64 1.14 .64 10 .66 11 12 .13 13 14 (2.54) (2.54) 15 16 .41 2.39 17 .58 18 .28 3.23 19 2.21 .09 20 21 .29 22 2.31 2.23 .25 23 .25 .05 24 25 1.26 26 .38 .20 27 2.72 2.11 .80 28 .89 2.10 29 .10 1.27 (2.54) 30 .30 .25 .03 31 .64 .41 .13 Total 8.20 11.87 7.79 3.20 13.09 12.88 a . . . . The values in parentheses are irrigation data. 19 Table 2. Monthly distribution of rainfall and irrigation from April 1973 to April 1974. Rainfall Total water Date Deviation Deviation Monthly from Irrigation Monthly from normal normal 1973 ----------- cm ------------- April 8.20 -2.40 8.20 -2.40 May 11.87 .16 11.87 .16 June 7.79 .60 7.79 .60 July 3.20 - .85 3.20 - .85 August 5.47 -3.96 7.62 13.09 3.66 September 7.80 .58 5.08 12.88 5.66 October 6.52 -l.36 6.52 -l.36 November 13.82 6.58 13.82 6.58 December 7.70 2.44 7.70 2.44 1974 January 6.71 2.44 6.71 2.44 February 4.78 1.78 4.78 1.78 March 10.92 5.01 10.92 5.01 April 4.17 -1.67 4.17 -1.67 Total 98.95 9.35 12.70 111.65 22.05 20 Pebbles and stones were encountered occasionally at various depths in the profile, sometimes constituting lenses. The corn crop was harvested in October 1973. Ears were collected from 30 ft of row from each plot and weighed. Moisture determinations were made on kernels and the yields were corrected to 15.5 percent moisture (Marlin and Leonard, 1967). Laboratory Procedures Ammonium and N0; contents of soil samples were determined with a semi-micro Kjeldahl distillation method (Bremner, 1965). Approximately 3 g moist soil was distilled in the presence of 10 m1 of 2N KCl solution. Ammonium was 3 Devarda's alloy. Both distillates were collected in two liberated first with MgO. Then NO was reduced to NH: with percent H BO and titrated with standard H SO solution in 3 3 2 4 the presence of methyl purple indicator (Fleisher Chemical Company). Ammonium and N0; were determined on each of the 5 samples taken for each depth increment on each plot. Chloride was determined for each depth on a composite of the 5 samples taken per plot. Chloride determinations were made with a Cl- specific electrode. Oven dry soil (20 g) was shaken with 50 ml of saturated CaSO4 solution for 30 minutes. Supernatant solu- tions were decanted into 50 m1 beakers and millivolt (mv) readings were taken with a Cl- specific electrode, using a 21 double junction calomel reference electrode (A. H. Thomas 4092-H10). The mv readings were converted to ppm Cl-, using previously prepared standard curves. Moisture determinations were made on soil samples at the time of chemical analysis, and the results are expressed on the basis of oven dry soil. Statistical Procedures Data for C1. and NOS/C1- ratios were analyzed in accordance with the randomized complete block design of the field experiment. Data for NH: and N03 design for sub-units in strips (Cochran and Cox, 1957). were analyzed as a split-plot Nitrogen treatments were taken as units and depths as sub- units. The partitioning of degrees of freedom is shown in Table 3, together with mean squares, F ratios and probabili- ties for F. 22 Table 3. Partitioning of degrees of freedom and scan equates, F ratloe and probabilities for 9 In analyses of variance for no; and NH’. July 1973 September 1973 May 1974 Source ”3 I P 998 P P 93 I P IITRATI 91och 3 53.55 .54 .669 109.65 .69 .593 19.96 2.04 .179 Treatment 3 1793.21 17.94 (0.0005 2677.09 16.90 (0.0005 33.76 3.63 .059 Error (Dr?) 9 99.99 259.40 9.29 Depth 6 5490.11 109.32 (0.0005 2795.99 32.09 (0.0005 420.66 97.35 (0.0005 Error (9x0) 19 50.66 96.95 4.32 Treatment 8 Depth 19 974.66 10.30 < 0.0005 575.06 7.79 (0.0005 5.57 1.46 .142 Error (DxTrD) 54 94.63 73.96 3.91 Sanpunq errorb «a 61. as 32.13 2.36 Tota1 559 ANNDIIUI Ilock 3 6.06 .39 .766 .95 .06 .991 .65 .09 .969 Treat-out 3 49.26 3.14 .090 9.34 .56 .656 69.67 9.56 .004 Error (9:?) 9 15.70 16.76 7.19 m 6 422.34 54.09 (0.0005 424.79 63.35 (0.0005 542.46 93.39 (0.0005 Irror (9x0) 19 7.91 6.71 6.50 Treateent x Depth 19 4.65 .70 .793 6.22 .96 .514 5.01 1.27 .242 Error (DxTxD) 54 6.61 6.46 3.93 Sanling error5 449 3.92 2.99 2.50 Total 559 I’Componente of sampling error: 4» (Sample 3: Block x Treamont) 9 (Sample a Block a Treatment x Depth). (Sample a: flock) o (Sample x Depth) 9 (mph a 910:: x Depth) RESULTS Chemical Analyses Background Analyses for samples taken before fertilizers were applied are given in Table 4. These are means for 10 cores taken randomly over the experimental area and analyzed sep- arately. Nitrate In July 1973, N03 in check profiles was unchanged from background except in topsoil (Table 5). Significant increases for N applications were found only in the 0-15 cm layer. Con- centrations in this layer were substantially higher than at 15-30 cm, which suggests that upward capillary movement may have helped to maintain N0; levels near the surface. The plow layer extended to 25 cm. Thus, there may have been some dilution of NO- in the 15-30 cm sample by 5 cm of subsoil which 3 was included in the sample. There was no evidence that significant quantities of NOS had moved out of the plow layer to greater depths. Just prior to corn harvest, in September, significant differences for treatment and for depth were associated only with the 336 kg application of N (Table 5). At this rate, 23 24 Table 4. Background analyses. Means and standard deviations for 10 cores sampled April 21 and 22 before fertili- zer applications on April 24, 1973. Depth NOS-N NHz-N c1" NOS/C1- ---------- ppmisd - - — - - RatioIsd 0-15 7.714.36 6.214.36 10.314.39 .69$.480 15-30 5.413.49 6.4;3.68 10.514.08 .60;.512 30-45 3.4;2.61 3.8;3.19 9.8;3.29 .41;.355 45-60 2.4;1.28 3.4;3.67 9.812.46 .27$.185 60-90 1.8; .91 1.41 .95 10.6;3.04 .17;.089 90-120 2.1; .89 1.3; .88 12.314.35 .17$.071 120-150 3.0;1.18 1.311.09 17.2;3.57 .18;.072 Profile total, 66 56 243 kg/ha 25 a.~ m.~H H.ma .uwu segue: spoon we o.mH H.ea snoop canvas .uwa Amo.v own as em as as mam mm me me was and om we nexus Hmuou daemons m.~ ~.m o.~ e.H o.~ H.~ e.H o.H m.~ e.~ m.H m.~ omauoma m.~ H.m o.~ m.~ m.~ m.H m.H o.~ m.~ o.m e.~ e.m omauom m.~ a.m m.m m.~ m.e H.~ o.~ e.~ m.~ ~.m m.~ e.~ canoe o.m o.e m.m m.m m.aa m.m m.~ o.~ o.~ e.m m.~ m.~ omume m.a m.o m.e m.a o.e~ o.e e.m o.m m.e m.e m.m ~.m meIom m.m m.m ~.e e.m o.mm a.oa e.e o.o H.oa H.oa o.o m.e omImH e.m o.m m.» m.m o.eH m.~H H.mH ~.ea m.me m.om o.ma e.m mauo I I I I I I I I I I I Ema .ZImoz I I I I I I I I I I I I I I I I omm moa em xomno omm mos em xomeo omm was em xoono seas as: mead Monsoudom mean sass .nudwm mn\z ox .mucwfiummuu one mmumo mcHHmEMm Aflom .mmumo mcmeEMm owns» co mucoEummuu z HSOM Moos: mawom Ge zImoz mo cofiusofiuumfio HMOfluHm> .m manna 26 concentrations at 15 to 45 cm were significantly higher than at greater depths or in the surface 0-15 cm. Totals recovered to 60 cm (Table 6), corrected for the check, represent -4, 4 and 35% of input urea N for the 84, 168 and 336 kg applica- tions respectively. 3 tions for all treatments (Table 5) were essentially the same 57 In May 1974, NO concentrations and profile distribu- as in the background samples taken one year earlier (Table 4). Concentrations below 30 cm tended to be higher at the 168 kg rate than for other treatments, and the profile total was 3’; greater (Table 5). The treatment main effect, however, fell just short of significance (P=.058). Ammonium The urea applied in May 1973 would have been hydrolyzed quickly. The data in Table 7 indicate that NH: released by hydrolysis had been completely nitrified or assimilated by the crop or by the microbial population by July. Concentra- tions found were essentially the same as background (Table 4). Ammonium concentrations in September were generally higher than in July 1973, and concentrations in May 1974 were even higher. These changes~over time were not great and probably reflect differences in climatic and soil conditions at the time of sampling. Ammonium varied significantly with depth on all samp- ling dates. The observed distributions are undoubtedly related to variations in quantity and quality of organic matter throughout the profile. 27 Table 6. Quantities of N0;, NH: and total mineral N recovered in soil profiles to 150 cu. Sampling dates and treatments, kg N/ha 3::th, July 1973 September 1973 May 1974 c. Check 94 169 336 Check 94 169 336 Check 94 169 336 - --------- - - - - - - - kg N/ha - - - - - - - - - - - - - - - - Nitrate 0-30 23 41 92 131 49 46 47 100 29 30 34 29 30-60 12 13 17 15 13 12 21 77 15 17 21 16 60.150 31 27 35 30 24 20 24 39 27 32 39 34 Sun 66 81 134 176 85 78 92 216 71 79 94 79 9 of total mineral N 54 66 69 83 57 56 61 82 47 49 53 59 AI-oniun 0-30 24 19 27 22 26 25 29 28 34 34 33 26 30-60 10 7 11 5 12 13 12 9 16 18 16 10 60-150 23 16 22 8 25 24 19 12 31 30 34 20 Sun 57 42 60 35 63 62 60 49 81 82 83 56 8 of total mineral N 46 34 31 17 43 44 39 18 53 51 47 41 Total mineral N 0-30 47 60 109 153 74 71 76 129 63 64 67 55 30-60 22 20 29 20 25 25 33 96 31 35 37 26 60-150 54 43 57 39 49 44 43 51 59 62 73 54 Total 123 123 194 211 149 140 152 265 152 161 177 135 28 H.m e.m m.m .uwu dashes tempo we m: we cameo Cannes .uue Amo.e own om mm mm as me so me no mm as me em hs\ox Hmuou wawmoum m.H m.~ H.m m.~ m. v.H m.a o.H m. m.a H.H m.H omaloma m.a m.m v.~ m.~ ~.H m.H ~.H H.~ m. o.~ ¢.H H.~ oNHIom m.H ~.m H.m m.m H.H m.H m.~ e.~ m. m.H o.H m.a omIom N.~ m.m ~.m v.m h.H >.~ n.~ m.~ ~.H m.~ h.H ~.m ooImv n.m H.¢ m.m m.v m.~ a.m m.m m.m m.H o.m h.H m.~ mVIom o.m v.n n.h a.» v.h v.m b.m m.m m.m H.m m.~ w.v oMImH H.m e.m o.m H.0H m.w m.h m.m m.> m.m v.m m.m «.5 mHIo I I I I I I I I I I I I m . Iv I I I I I I I I I I I I I s m z +mz mmm mod em xomno mmm mod em xomso wmm mod em xoocu Eo «baa he: mnma umnfimumom muma wash .numoo Hfiom mn\z ox .mucmfiumouu one mmuoo mcflamenm .mmueo mewamfiem woman no mucofiummuu z snow Moos: mHHom cw zIwmz mo cofluooeuumwo Hmowuuo> .5 manna 29 No significant differences for treatment were found at any depth. However, a highly significant (P=.004) main effect of treatment was expressed in May 1974. At the 336 kg N rate, NH: was consistently lower than for other treatments at depths below 15 cm (Table 6), and the profile total (56 kg/ha) in Table 7 is significantly less than for other treatments. On all three sampling dates, NH: as a percentage of total mineral N decreased with increasing N rate (Table 7). The increasing ratio of N03 to NH: suggests that the size of the nitrifying population was enhanced early in the season by. + 4 for NH: by this larger population would have reduced the size increasing levels of NH released from urea. Increased demand of the NH: pool which could be maintained later by mineraliza- tion from soil sources. Chloride In July 1973, major concentrations of chloride were encountered in the surface 30 cm (Table 8). Comparison with the background data in Table 4 suggests that some Cl- may have moved downward as far as the 45-60 cm layer. Concentrations below 60 cm were similar to background. Although the data indicate that all of the applied C1- was retained in the surface 60 cm, total recoveries to 60 cm represented only 46, 61, 54 and 46% of the input 604 kg/ha for the check and increasing N rates, respectively. These low recoveries may be related to the fact that soil samples were taken between rows of corn at a distance greater than 8 or 10 inches to avoid concentrations of banded 30 Né oém o.o~ .uwu can»? spoon me new we nudes 5.5a: .98. Amo.v omq mom NNN mam mmm mom hum new mmb «we moo mmv «He m3\mx Hmuou mafimoum m.ma o.mH m.e~ >.m~ nama a.m~ «.mm m.e~ m.HH m.NH m.aa H.¢H omanma m.mH m.ma H.N~ H.m~ o.mH m.mH h.h~ m.om m.~H o.HH m.m H.HH omanm m.vH H.m m.mH m.HH H.HN m.m~ o.¢e m.mm m.oa m.oa w.m v.m omIoo m.oH m.h h.>. m.m m.om m.hm m.mb m.Hv m.va ¢.ba m.HH m.a~ omImv m.h m.m m.m 0.5 «.mm e.mm ¢.ob H.wh H.¢H m.mH w.ma H.0H vaom m.m H.h H.m H.HH o.mv H.0m H.mm v.~e m.~o «.me m.em H.mh omImH h.oa H.m n.m v.m v.ma m.b~ H.5H h.o~ m.>m «.mm m.mm ~.mm mHIo I I I I I I I I I I I Ema .IHU I I I I I I I I I I I mmm mod em xowno mmm mod em xowno mmm and em xomno u whoa an: eema uwnEmummm mead wash sawmo mn\z mx .mucmEummuu one mmumo mcwamsmm Haom .mmueo mcwamaem owns» :0 mucosummuu z snow noon: maflom ca Iao mo coausnfluumflo Heowuum> .m manna 31 fertilizer. Chloride concentrations in the surface layers at the point of sampling may have been reduced by capillary mi- gration into ridges left by the planter near the rows of corn. The pattern of scattered light rains in June through July, prior to and into the July 6-13 sampling period, would have been conclusive to this sort of capillary migration (Table 1). The problem of assessing Cl- movement is complicated further by the fact that extractable quantities increased from July to September. Total Cl- extracted to 150 cm in September ranged from 609 to 877 kg/ha (101 to 145% of input), as com- pared with totals in July of 414 to 493 kg/ha (68 to 82% of input) (Table 8). There had been considerable apparent move- ment of C1- out of the 0-15 cm layer by September. This displaced Cl- may have been distributed more uniformly so that more representative samples were obtained. Displaced Cl- had accumulated to significantly higher concentrations at 15 to 60 cm than in the surface 15 cm or at depths greater than 60 cm. Significant differences for treatment occurred only at 15-30 cm. It appeared that lower concentrations at this depth for the check and 84 kg N/ha may have been due to displacement of C1- to greater depths. These apparent differences in downward displacement would be consistent with observed differences in volume of growth and density of the canopy for these three treatments. A larger percentage of precipitation and irrigation water would have been intercepted by the more dense canopy which 32 developed at the 168 kg rate. Consumptive use would have been greater because of the larger volume of growth, and evapotranspiration may have been greater because of the more extensive, deeper root development which may be inferred from the greater volume of top growth. Thus, net percolation and leaching would have been reduced relative to the check or the 84 kg rate. A similar argument would explain the low concentrations in the lower profile under the 336 kg N rate where a vigorous, dense canopy developed also. It would not account for the generally lower levels in the upper profile compared with the 168 kg rate. By May 1974, Cl- concentrations to 90 cm (Table 8) had declined, with all treatments, to background levels (Table 4). There were no significant differences for treatment. Never- theless, significantly higher concentrations for depths below 90 cm under the check and the 84 and 168 kg N rates indicate that Cl- levels in the lower profile were still being influ- enced by the fertilizer applications made a year earlier. Comparing profile totals in May 1974 with those in July 1973, net disappearance from the profile represented 12, 30, 40 and 36% of the input Cl- for the check and increasing N rates, respectively. Disappearance from September 1973 to May 1974 represented 66, 94, 100 and 67% of input for the same treatments.r Since recoveries in July 1973 were obviously low, the September-based estimates are more valid. They indicate that Cl. disappeared from these profiles in quantities 33 approaching input. It may be inferred that most of this Cl- had been displaced to depths greater than 150 cm. NOS/Cl- Ratios The ratio of N applied as urea to Cl- applied as KCl (including the 44 kg Cl-/ha in 0-26-26 fertilizer) and CaCl2 was .14, .28 and .57 for the 84, 168 and 336 kg rates of N, respectively. The ratios of NOS-N to C1- for these treatments (Table 9) were very similar to input in the 0-15 cm surface layer in July 1973. At greater depths, these ratios tended to converge toward a common value of .2 to .3 for all treat- ments. It may be that these ratios were characteristic for percolating soil solution in these soils under previous management which involved annual applications of fertilizer. By September, ratios in the 0-15 cm surface layer had increased, reflecting the movement of Cl- out of this layer. At greater depths, sharply narrower ratios reflect greater downward movement of Cl- than of NO- for the check and the 3 84 and 168 kg N rates. Ratios remained wider through most 3 higher (Table 5) and C1- lower (Table 8) than for other of the profile at the 336 kg N rate because NO was generally treatments. In May 1974, NOS/Cl- ratios had widened markedly, reflecting further removals of Cl- (Table 8) while moderate concentrations of NO3 were maintained by release of N from soil sources (Table 5). ”I I6 34 ~m. aw. an. .uuu ensues spoon no em. om. sumac ensues .une Amo.v own ma. ma. mo. mo. ma. mo. mo. no. om. mu. 5H. om. omHIONH ma. mm. ma. Ha. mm. ma. no. mo. Hm. om. hm. om. omanm Hm. mo. mm. mm. mm. ca. no. mo. mm. AM. ha. ov. omIom hm. hm. mm. Ne. mm. wo. wo. mo. Hm. om. mm. mm. omIme hm. mm. he. no. mm. no. mo. mo. hm. mm. om. Hm. vaom on. mN.H em. as. me. ma. mm. 5H. me. am. NH. ca. omImH Ho.H mv.H m~.H mo.H mm.H mm. mm. mm. on. mm. ma. no. mHIo I I I I I I I I I I I I I I I oases Inoxmoz I I I I I I I I I I I I I I I mmm and em xoenu mmm mod om xomno mmm me em xoesu . «Add an: mean nonsouoom mean mass .tuMmc en\z ox .munefiueeuu one meueo onwamnem Haom .meueo mnaamneu woman no muneEueeuu z unom neonn waaOm nw mowueu IHU\moz mo nowunnwuuuwo He0fluue> .m eaoea 3S Crop Yields Corn yields followed a typical N response curve (Figure 1). Corn on check plots was severely N deficient throughout the season. The major yield response was to the first increment of N. Additional small increases for 168 and 336 kg were not significant. The yield increase over the check for the first 84 kg of N was 3.76 T/ha. The additional increase for the second 84 kg was only 0.81 T/ha. In spite of the relatively much smaller increase for the second increment of N, No; levels at the end of the growing season for these two treatments were essentially the same as for the check (Tables 4 and 5). Only with the 336 kg treatment were the quantities of unused N0; in September significantly greater than the check. The excess to 150 cm for this treatment was 131 kg, or 39% of input. 36 YIELD, T/I'Io l.LsotosI I I I 1 O 84 l68 336 N APPLIED, kq/ho Figure 1. Yield of corn as influenced by rates of N application. . DISCUSSION Movement of Chloride The major objective of this research was to study the relationship between rates of fertilizer application and potential contamination of groundwaters or subsurface drain- 3. used as a stable reference anion. Unfortunately, there are age with NO To follow movement of soil solution, Cl- was some ambiguities in the total quantities of Cl' recovered from sampling to sampling and among treatments. Nevertheless, changes in profile distribution in Figure 2 provide the basis for useful estimates of leaching intensity. In July 1973, major concentrations of Cl- were found in the surface 0 to 30 cm. Upward capillary movement appears to have contributed to peak concentrations in the 0-15 cm layer. Small secondary peaks appeared for the check and the two higher N rates at 45-60 cm. These were not statistically significant (Table 8), but they had not appeared in the back- ground data (Table 4). These small peaks do suggest that some plowed down Cl- may have leached with early rains and was isolated at this depth from upward capillary movement which led later to concentration in the upper part of the plow layer. Such downward movement would have been most likely to 37 M" ”d 40. 904 IZO - SOIL DEPTH (cm) |20~ .nl Figure 2. l69 kg N/M cI" coarser (pp-II o aI «I «I a: no I 04 II. am o 20 no so so no 536 9. RIM ONLY” D SEPTJD'IS A MAY [’74 Depth distribution of C1- in soil with varying rates of N application at three sampling dates. 39 occur when the soil was bare and before the corn had developed to the point where it could exert a significant demand for water. Rainfall from April 24, when the Cl- was applied, to June 6, when corn was beginning to grow rapidly, totalled 14.04 cm (Table l). Plowed down el', at a mean depth of 20 cm, would have had to move 32.5 cm to the mean depth of the July peaks at 52.5 cm (Figure 2). The indicated rate of leaching is 2.31 cm/cm rain. This is not unreasonable when compared with similar estimates for sandy soils in the lit- erature (Wetselaar, 1962; Jones, 1975). The peak concentrations which were found at 0-15 cm in July had been displaced downward by September to an average depth of 37.5 cm for the four treatments. The average displacement from the mean peak depth of 7.5 cm in July was 30 cm. Rainfall plus irrigation between the July and Septem- ber samplings totalled 26.63 cm. The average leaching rate was, therefore, 1.13 cm/cm of water. By the following May, major concentrations of C1- had been displaced to depths greater than 150 cm. The mean displacement from September to May was, therefore, greater than 120 cm. Rainfall between the two samplings totalled 54.62 cm. Thus, the average rate of leaching for the fall, winter and early spring periods was in excess of 2.2 cm/cm of rain. Precipitation normally exceeds evapotranspiration in Michigan, except during the active growing season. Thus, 40 leaching rates in April and May and for fall, winter and spring periods would be expected to be greater than in July, August and September. The leaching encountered in this study would have been greater than with normal rainfall because of irrigations in August and September and because rainfall from November through March was 70% greater than normal (Table 2). Nitrogen Relationships Only the mineral forms of N (NH: and N03) were followed by chemical analysis. These data provide no direct evidence as to the extent to which N was removed into the crop or immobilized in microbial tissues or may have been removed from the soil by denitrification. Some general inferences can be made from the sequence of changes in profile distributions of NH: and N0; (Figures 3 and 4), their rela- tionship to the development of the crop (reflected in corn yields, Figure l), and their relationship to changes in profile distribution of Cl- (Figures 2 and 5). Ammonium As the substrate for nitrification, NH: would have been supplied by hydrolysis of urea and by mineralization of soil organic matter, of root exudates and root debris from the growing crop, or of residues from a previous crop. The profile distributions in Figure 4 follow the expected variation with depth in the quantity and the quality of organ- ic materials which might decompose to release NHZ. 0 D 20 30 O D 20 I) 40 so I l l l L l l I I5 - - 30 - - 45 q - .0 - CHECK 34 kg III/ho ” - m - f: I ‘0 J .- It a O D 20 30 O D 20 I) 40 so _I I l l l l l l J 5 6" l5 - . w "' el 45 -‘ .. 60 -I I“ kg N/lIa .. 338 to Nine 0 JULY l973 so J . D SEPT. I973 A HAY |974 92° "‘ u so . 1 Figure 3. Depth distribution of NOS-N as influenced by rates of N application at three sampling dates. SOIL DEPTH (cm) 42 N04: CONTENI (no) O 4 6 D O 2 4 6 8 IO 1 9 n a a a n s 1 n J 1 a a a 1 a J 40 I‘ n w .I -I 49 s .. Q -I .. cases 04 II. an. n "' .I 020 - - ml 1 O 2 6 D O 2 6 9 l0 9 j 1 1 l l a a l l 1 1 l l a I 4 1 l5 - - 30 q q 46 - .. ao- - l69 IIgII/IIII 333 l. who 90 . -I 0 JULY I973 D SEPT. D73 l20 -I -I A HAY 074 Inc J J + . . . Figure 4. Depth distribution of NH4-N in 8011 as influenced by rates of N application at three sampling dates. 451 ‘01 420 -I '50" SOIL DEPTH (on) 30-: «at 90" I20- I50 J Figure 5. I69 kg HIM 43 HGVd'sano n4 a4 II. II/IIII 858 kg Nike 0 MY 7073 D SEPT.” A HAY l974 Depth distribution of NOS/C1- ratios in soil as influenced by rates of N application at three sampling dates. 44 There was no evidence that NH: levels in July 1973 were still being influenced by additions to the NH: pool from hydrolysis of urea. Thus, the concentrations found represent equilibrium levels maintained by release of NH: from decom- posing organic matter and removals by microbial immobiliza- tion, nitrification and crop uptake. Because it is a cation, NH: is not very mobile in soils. Some downward movement undoubtedly does occur in company with anions in the soil solution. Such movement is likely a significant factor in maintaining low levels of exchangeable NH: at depths in the profile where the quantity and decomposability of soil organic matter is low. Levels of NH: increased generally from sampling to sampling, perhaps reflecting inputs of organic matter from the 1973 corn crop. Differences in soil temperature, moisture and aeration at the time of sampling may have been involved also. Nitrate The distribution curves for N0; in Figure 3 provide a sharp contrast to those for C1- in Figure 2. With the check and the first two increments of N, there was no downward displacement in September comparable to the bell shaped curves for C1-, but there was at the 336 kg rate. The peak above 60 cm for this treatment contained surplus N0; (in excess of the check) equivalent to about 40% of input N (Table 7). By the following May, all of this NO3 had disappeared. Pre- sumably it had been displaced to depths greater than 150 cm, 45 as in the case of Cl-. Also, some may have been immobilized or denitrified in the fall before soils became too cold to suppress microbial activity. 3 N rate occurred in the 15-30cm layer, whereas the peak for The September peak of NO concentration for the high C1- was centered at 30-45 cm. This retardation of the peak for NO; can be ascribed, at least in part, to the dynamic nature of nitrogen transformations in soils. Some of the N0; which was present in surface layers in July would have been taken up later in the season by the crop or intercepted by biological systems in the soil, whereas some of the N0; appearing in September would have come from nitrification of + . . . NH4 released by decomposition of organic matter. NOS/Cl- Ratios The contributions of soil-derived N03 to the upper portion of the September peak for the high N treatment is evident in the sharply wider NOS/Cl- ratios in the two surface soil layers (Figure 5). This widening of the NOS/Cl- ratio in surface soils in September is apparent also for the lower N rates and the check. Much wider ratios in May 1974 parallel roughly the distribution of NH: (Figure 4) and the probably distribution of organic matter. This general correspondence between NOE/Cl- ratios and NH: was apparent also in background samples taken in April 1973 (Table 4). The relationship suggests that some N0; had accumulated again prior to sampling due to resumption of microbial activity in early spring. On the other hand, NO3 46 may be influenced to a greater extent than C1. by interactions with organic colloids. The slower movement of N0; in soils high in organic matter has been ascribed to differences in disperion coefficients for the two anions (Corey gt gt. 1967a). Differences in anion exchange behavior may well be involved, although such interactions have not been studied extensively (Thomas, 1963). Such differences would have a significant bearing on studies such as this where Cl- is used as a 3 The data do suggest that soils may have the capacity reference anion for studying NO movement in soils. to retain low concentrations of both C1- and N03 against the leaching action of water. This retention capacity might be estimated for C1- from minimal concentrations in Tables 4 and 8 at 10 to 12 ppm for subsoil materials below 60 cm and 7 to 10 ppm in the upper profile. Minimal retention capacities 3- order of 2 to 4 ppm in the lower profile, increasing to 8 or for NO N, estimated from Tables 4 and 5, would be of the 9 ppm in the surface 0 to 15 cm. Practical Implications Since no samples were taken from September to May, the data obtained in this study are not entirely satisfactory for inferring what may have been the fate of residual No; or how 3 been over the winter. Additional removals from the N0; pool extensive actual movement of NO through the soil may have may have occurred during the fall by immobilization or denitrification in the presence of decomposing corn residues. 47 On the other hand, net mineralization may have led to addi- tional accumulations of NO3 before microbial activity ceased due to cold winter temperatures. This limitation is not damaging, however, to implica- tions for practical management which the data clearly support. With the 84 and 168 kg/ha applications of N, residual NOS at harvest time had been drawn down to the same level as in check profiles which received no N. Although the yield in- crease for the second increment of N was not statistically significant, it was well within the response portion of the yield response curve (Figure l). 3 30 cm had been reduced for these three treatments to mean By the time of the July sampling, NO in the surface 3- (Table 5). The critical concentration in the plow layer for concentrations of 6, 10 and 20 ppm NO N, respectively corn at silking time is considered to be 20 ppm (Lucas, 1969). The mean concentration to 30 cm for all three treatments in September was about 12 ppm. Thus, yields for all three treatments were, in fact, limited by an insufficiency of available N. With the 336 kg application, the mean concentration of NOS-N for the surface 30 cm was 32 ppm in July and 25 ppm in September. Therefore, N cannot be considered to have been limiting. In fact, residual unused NC} at harvest exceeded that for the check by 131 kg N/ha. A similar quantity (137 kg) disappeared from the profile from September to May. Presumably this loss was due to leaching, although immobiliza- tion and denitrification may have been involved also. 48 There was essentially no change in profile totals for the other three treatments from September to May. If there was, in fact, a significant release of N0; from soil sources during the fall after the September sampling, then an equiva- lent quantity may have been leached to depths greater than 150 cm. Nevertheless, the potential for such movement was very much less than at the 336 kg rate of application. It would appear that there is a limiting concentration to which the N0; pool can be depleted by a cultivated crop such as corn, even under stress. The similarity in profile totals for the check and the first two increments of N in September and the totals for all four treatments in May (Table 7) suggests that this limiting concentration is related to the 5011's capacity to retain NO; against the leaching action of water. In other words, NOS at these concentrations does not appear to be freely mobile in the soil solution. If 3 growing season, the leaching hazard during winter and spring residual NO can be depleted to these levels by the end of the can be greatly minimized. The 336 kg application exceeded the capacity of the corn crop and associated biological systems in the soil to 3 ence may be made that, if this input had been reduced by 40% deplete NO to this limiting level by about 40%. The infer- (to 200 kg N/ha), the desired depletion of NOS could have been achieved, together with a slight increase in yield over the 168 kg rate (Figure l). The yield would still have been some- what less than a maximum response to N. CONCLUS ION Data obtained in this study support the view that the agronomic objective of fertilizing for maximum yields is incompatible with the farmer's responsibility to society for minimizing the environmental impact of fertilizer use. On the other hand, the degree of adjustment in fertilizer practice necessary to fulfill this responsibility may not be great. In the field experiment described here, a slight sacrifice in yield (statistically non-significant) due to reducing the highest N application from 336 kg/ha to 200 kg would have given the same degree of control over NO3 in the soil system as the use of no fertilizer N at all. 49 LIST OF REFERENCES LIST OF REFERENCES American Fertilizer Institute. 1970. Environmental Fact Sheet No. 10, September 1970. Biggar, J. W. and D. R. Nielsen. 1965. Miscible displacement and leaching phenomenon. tg_Irrigation of agricultural lands. Agron. Mon. 11:254-274. Bingham, F. T., S. Davis and E. Shade. 1971. Water relations, salt balance, and nitrate leaching losses of a 900-acre citruc watershed. Soil Sci. 112:410-418. Bolton, E. F., J. W. Aylesworth and F. R. Hore. 1970. Nutrient losses through tile drains under three cropping systems and two fertility levels. Can. J. Soil Sci. 50:275-279. Bremner, J. M. 1965. Inorganic forms of nitrogen. t2 Methods of soil-analysis. Agron. Mon. 9:1179-1237. Broadbent, F. E. and T. Nakashima. 1968. Plant uptake and residual value of six tagged nitrogen fertilizers. Soil Sci. Soc. Am. Proc. 32:338-392. Burns, G. R. and L. A. Dean. 1964. The movement of water and nitrate in soils around bands of sodium nitrate in soils and glass beads. Soil Sci. Soc. Am. Proc. 28:470-474. Campbell, C. A., W. S. Ferguson and F. G. Warder. 1970. Winter changes in soil nitrate and exchangeable ammonium. Can. J. Soil Sci. 50:151-162. Chalk, P. M. and D. R. Keeney. 1975. Fate of fall and spring applied 15N-labeled aqueous ammonia in porous buried bags. Agron. J. 67:37-41. Cochran, W. G. and G. M. Cox. 1957. Experimental design. Wiley, New York, p. 306. Corey, J. C., D. R. Nielsen and D. Kirkham. 1967a. Miscible displacement of nitrate through soil columns. Soil Sci. Soc. Am. Proc. 31:497-501. 50 51 Corey, J. C., D. Kirkham and D. R. Nielsen. 1967b. The movement of chloride and nitrate through certain Iowa soils. Iowa Acad. Sci. 74: 130- 141. Ferguson, H., P. L. Brown and D. D. Dickey. 1964. Water movement and loss under frozen soil conditions. Soil Sci. Soc. Am. Proc. 28:700-703. Felizardo, B. C., N. R. Benson and H. H. Cheng. 1972. Nitrogen, salinity, and acidity distribution in an irrigated orchard soil as affected by placement of nitrogen fertilizers. Soil Sci. Soc. Am. Proc. 36:803-808. Fireman, M., O. C. Magistad and L. V. Wilcox. 1945. Effect of sodium nitrate and ammonium fertilizers on the permeability of western soils. J. Am. Soc. Agron. 37:888-901. Fox, R. L., R. A. Olson and A. P. Mazurak. 1952. Persis- tence of ammonium ion and its effect upon physical and chemical properties of soil. Agron. J. 44:509-513. Gardner, W. R. and R. H. Brooks. 1956. A descriptive theory of leaching. Soil Sci. 83:295-304. Gardner, W. R. 1965. Movement of nitrogen in soil. In W. Bartholomew and F. Clark (ed). Soil nitrogen. Agron. Mon. 10:550-572. Amer. Soc. of Agron., Madison, Wis. Ghuman, D. S., S. M. Verma and S. S. Prihar. 1975. Effect of application rate, initial soil wetness, and redistribution time on salt displacement by water. Soil Sci. Soc. Am. Proc. 39:7-10. Herron, G. M., A. F. Dreier, A. D. Flowerday, W. L. Colville and R. A. Olson. 1971. Residual mineral N accumula- tion in soil and its utilization by irrigated corn. Agron. J. 63:322-327. Jones, M. J. 1975. Leaching of nitrate under maize at Samaru, Nigeria. Trop. Agric. (Trinidad) 52:1-10. Levin, L. 1964. Movement of added nitrates through soil columns and undisturbed soil profiles. Trans. 8th Intern. Congr. of Soil Sci. (Bucharest) IV:1011-1022. Linville, K. W. and G. H. Smith. 1971. Nitrate content of soil cores from corn plots after repeated nitrogen fertilization. Soil Sci. 112:249-255. Lucas, R. E. 1969. Cultural trials on irrigated corn. Mich- igan State Univ. Agric. Expt. Sta. Research Report 84. 52 Kirda, C., J. L. Starr, C. Misra, J. W. Biggar and D. R. Nielsen. 1974. Nitrification and denitrification during miscible displacement in unsaturated soil. Soil Sci. Soc. Am. Proc. 38:772-776. Marlin, J. H. and N. H. Leonard. 1967. Principles of field crop production. The Macmillan Co., New York. McGregor, J. M., G. R. Blake and S. D. Evans. 1974. Mineral nitrogen movement into subsoils following continued annual fertilization for corn. Soil Sci. Soc. Am. Proc. 38:110-113. McMahon, M. A. and J. W. Thomas. 1974. Chloride and tritiated water flow in disturbed and undisturbed soil cores. Soil Sci. Soc. Am. Proc. 38:727-732. Owens, L. D. 1960. Nitrogen movement and transformations in soils as evaluated by a lysimeter study utilizing isotopic nitrogen. Soil Sci. Soc. Am. Proc. 24:372-376. Power, J. F., J. Alessi, G. A. Reichman and D. L. Grunes. 1973. Recovery, residual effects, and fate of nitrogen fertilizer sources in a semiarid region. Agron. J. 65:765-768. Protasov, P. V. and G. D. Korostoleva. 1972. Loss of nitrogen from various nitrogen fertilizers on irrigated lands. Soviet Soil Sci. 4:566-570. Starr, J. L., F. E. Broadbent and D. R. Nielsen. 1974. Nitrogen transformations during leaching. Soil Sci. Soc. Am. Proc. 37:283-289. Smith, S. J. 1972. Relative rate of chloride movement in leaching of surface soils. Soil Sci. 114:259-263. Terry, D. L. and C. D. McCants. 1974. Distribution patterns in sandy soils of ions leached from fertilizer bands. Soil Sci. 118:412-420. Thomas, G. W. 1963. Kinetics of chloride desorption from soils. J. Agr. Food Chem. 11:201-203. Thomas, G. w., R. L. Blevins, R. E. Phillips and M. A. McMahon. 1973. Effect of a killed sod mulch on nitrate movement and corn yield. Agron. J. 65:733-736. Thomas, G. W. and A. R. Swoboda. 1970. Anion exclusion effect on chloride movement in soils. Soil Sci. 110: 163-166. 53 Wetselaar, R. 1961. Nitrate distribution in tropical soils. I. Possible causes of nitrate accumulation near the surface after a long dry period. Plant Soil 15:110- 120. Wetselaar, R. 1962. Nitrate distribution in tropical soils. III. Downward movement and accumulation of nitrate in the subsoil. Plant Soil 16:19-31. Wetselaar, R., J. B. Passioura and B. R. Singh. 1972. Consequences of banding nitrogen fertilizers in soil. I. Effects on nitrification. Plant Soil 36:159-175. Wild, A. 1972. Nitrate leaching under bare fallow at a site in Northern Nigeria. J. Soil Sci. 23:315-324.