This is a draft manuscript prepared for peer-review and pre-review purposes only. No publication in any format is authorized by the authors. All rights reserved. Dec '96

Effects of Long-term Soil Acidification due to Agricultural Inputs in Wisconsin

Phillip Barak^*1, Babou O. Jobe², Armand Krueger³, Lloyd A. Peterson³, and David A. Laird⁴.

* Corresponding author (barak@calshp.cals.wisc.edu).
¹Dep. of Soil Science, Univ. of Wisconsin­Madison, 1525 Observatory Dr., Madison, WI 53706-1299;
²National Agricultural Research Institute, Yundum, The Gambia;
³Dep. of Horiculture, Univ. of Wisconsin­Madison, 1575 Linden Dr., Madison, WI 53706-1299;
⁴USDA-National Soil Tilth Lab, Ames, IA.

Abstract

Agroecosystems are domesticated ecosystems intermediate between natural ecosystems and fabricated ecosystems, and occupy nearly one-third of the land areas of the earth. Chemical perturbations as a result of human activity are particularly likely in agroecosystems because of the intensity of that activity, which include nutrient inputs intended to supplement native nutrient pools and to support greater biomass production and removal. At a long-term fertility trial in South-Central Wisconsin, significant depletion of exchangeable Ca²⁺ and Mg²⁺, increases in exchangeable acidity, decline in CEC, and decline in base saturation have been noted in association with application of ammoniacal N fertilizer. Plant analysis shows that a considerable portion of the alkalinity generated by assimilation of N (and to a lesser extent by S) is sequestered in the above-ground plant parts as organic anions and is not returned to the soil if harvested. Elemental analysis of Ca-saturated soil clays indicates an irreversible loss of 16% of the CEC of the soil clay and minor increases in Fe and Al. The reversibility of these changes due to prolonged acidification is doubtful if the changes are due to soil weathering.

Introduction

Agricultural activities have become the dominant ecological force over nearly one-third of the land areas of the earth (Cox and Atkins, 1979). Agroecosystems differ from natural systems in the auxiliary energy sources to enhance productivity, leading to high fluxes of inputs and outputs, and in the external, goal-oriented control rather than internal control via subsystem feedback (Odum, 1984). Key to the modern agroecosystem is nutrient inputs intended to supplement native nutrient pools and to support greater biomass production and removal.

Almost none of the fertilizer materials in common usage are acidic, with the exception of sulfuric and phosphoric acids, which are a very small part of total fertilizer consumption in the U.S. (Meister, 1995). Many nutrient inputs, however, are themselves acid-forming, a fact that was recognized by the pioneering work of the soil scientist W.H. Pierre in 1933 and codified in the Association of Official Analytical Chemists (AOAC) Method 936.01 for determining the acid-forming or non-acid forming quality of fertilizers. The most important acid-forming reaction for fertilizers is microbial oxidation of ammoniacal fertilizers, which may themselves be strong bases, by the following reactions:

1) NH₃ + 2O₂ = H⁺ + NO₃^­ + H₂O (nitrification of ammonia)

2) NH₄NO₃ + 2O₂ = 2H⁺ + 2NO₃^­ + H₂O (nitrification of ammoniacal nitrate)

3) CO(NH₂)₂ + 4O₂ = 2H⁺ + 2NO₃^­ + H₂O + CO₂ (hydrolysis of urea and nitrification of products)

The materials above are the most common synthetic N inputs to agroecosystems, either as single or mixed NPK fertilizers, and will oxidize to the equivalent of nitric acid under the well-drained, aerobic conditions of most dryland agricultural soils.

Interestingly, the assimilation of nitrate to N_org and of sulfate to S_org both consume protons, i.e., generate alkalinity, by the reactions:

4) R^.OH + NO₃^­ + H⁺ = R^.NH₂ + 2O₂

5) R^.OH + SO₄^2­ + 2 H⁺ = R^.SH + 2O₂ + H₂O

Although uptake and assimilation of nitrate and sulfate causes release of alkalinity from the root, some of the alkalinity generated is retained in the shoot as organic anions. The total concentration of organic anions in plant tissue may either be measured as ash alkalinity or excess base (Pierre and Banwart, 1973) or calculated from elemental analysis of plant tissue, with either measurements or assumptions regarding the concentrations of NO₃-N and SO₄-S in the plant tissue (van Beusichem et al., 1988).

Uptake and assimilation of nitrate by biota--both plants and microorganisms--is, from a redox and proton production point of view, opposite in direction and is the reverse of the process of nitrification, although separated temporally and spatially. The net reaction of ammoniacal N addition, nitrification, and subsequent nitrate uptake and assimilation by biota is acid/base neutral:

5) [NH₄NO₃, 2NH₃, CO(NH₂)₂] + 2R^.OH = 2R^.NH₂ + H₂O (+2O₂) (+CO₂)

From this perspective, soil acidification due to nitrification of N inputs is not directly caused by N inputs themselves but rather: i) N inputs greater than those assimilated by biota and stored either in biota or soil organic matter, and ii) incomplete return to soil of the alkalinity of organic anions.

Other acid/base reactions involving the N cycle are denitrification, which consumes protons, and ammonia volatilization, which generates one proton for each cation reacting. It should be noted that reactions causing soil acidification and alkalinization other than nitrification and assimilation of N, are known (Van Breeman et al., 1983). The most significant of these are redox changes, most notably with metals such as Fe and Mn. However, these processes will generally be of small magnitude in well-drained, well-aerated agricultural/prairie soils compared to redox reactions of N added as fertilizers.

Materials and method

Site and experimental design

A N-P-K long-term fertility experiment was laid out in 1962 by Dr. Lloyd A. Peterson at the Agricultural Research Station, located in south central Wisconsin (43° 18' N, 89° 21' W, 791 mm annual precipitation). The soil is classified as Plano silt loam (formally known as Parr silt loam), a fine, silty, mixed, mesic Typic Argiuoll. The soil formed under prairie grasses in loess and is level and well-drained to moderately well-drained at the experimental site. The mineralogy of the Plano silt loam closely resembles that of the Wapun silt loam, in which the fine silt fraction contains moderate amounts of feldspar, quartz, and mica. The clay fraction is predominantly interstratified smectite-illite in the fine <0.06 µm fraction, with interstratified smectite-illite, kaolinite, quartz and illite in the 0.06µm to 2 µm fraction (Liu, 1995). In 1962, the soils at the site had pH values (in water) from 6.5 to 7.0 and 3.0 to 3.5% organic matter (Peterson and Krueger, 1980).

The design was a 4x4x4 (NxPxK) duplicated split-strip factorial arrangement with main effects confounded. The original intent was to evaluate the response of local crops to different soil fertility levels established and maintained by various fertilizer application rates. The layout contained four plot sizes, the smallest of which was 6.4m x 10.8m. The fertilizer treatments were: 1) four levels of N (0, 56, 112, and 168 kg N ha^-1) as either urea or ammonium nitrate, assigned to vertical strips, 2) four levels of K (0, 84, 168, and 252 kg K ha^-1) as either KCl or K₂SO₄, assigned to horizontal strips, with N and K interacting in the intersection plot, and 3) four levels of P (0, 45, 90, and 135 kg P ha^-1) as triple superphosphate (20% P), assigned to the sub-sub plot. The fertilizer treatments were maintained annually from 1962 until 1982, with plots planted to various horticultural crops. After 1982, P and K fertilization was terminated and only N was applied at the original rates. Between 1982 and 1987, the plots were planted to tobacco and soybeans, and from 1987 to 1993, sweet corn was the test crop. For the 30-yr duration of the experiment, conventional tillage to a depth of 20 cm and removal of above-ground crop residues after harvest was continued.

Soil and clay analyses

In 1993, soil samples from each of the 128 6.4m x 10.8m plots in the experiment were collected as nine 2.5-cm soil cores to a depth of 20 cm, composited, and homogenized. One set of subsamples was maintained at field-moist conditions under 4 °C storage for analysis. The chemical properties measured included pH in a 1:1 soil:water slurry, exchangeable Ca, Mg, Na, and K by displacement with 1 M ammonium acetate (pH 7) and measurement by atomic absorption spectroscopy (Ca and Mg) and by flame photometry (K and Na), exchangeable acidity by displacement by 1 M KCl (McLean, 1982) followed by automatic titration to pH 8.2 by the procedure of Barak et al. (1996), and exchangeable ammonium by displacement by 2 M KCl and Kjeldahl distillation (Liu et al., 1996). In addition, pH was measured in the 1 M KCl extracts before titration. Soil concentrations of C and N and isotopic abundances of ¹³C and ¹⁵N were measured with a Carlo Erba NA 1500 series II carbon/nitrogen analyzer and Europa mass spectrometer. Effective cation exchange capacity (CEC) was calculated as the sum of exchangeable base cations (including ammonium, i.e., Ca²⁺, Mg²⁺, Na⁺, and K⁺, and NH₄⁺) and exchangeable acidity. Soil chemistry data was subjected to analysis of variance by SAS using a strip-split model (Steel and Torrie, 1980). Following mean comparison tests, differences greater than least significant differences (LSD) at the 5% level were reported as significant.

A detailed chemical analysis of the clay fraction (<2 µm) of two adjacent plots, one of which received no N fertilizer and the other 168 kg N ha^-1 yr^-1 for 30 years, was conducted. The clay fraction was dispersed by the method of Edwards and Bremner (1965) using a Na-saturated cation exchange membrane (CR 67 HMR-412, Ionics, Inc; Watertown, Mass.) and separated by sedimentation and decantation. The clay fractions were rinsed free of excess salt with deionized water and centrifugation, saturated three times with 0.1 M Ca(NO₃)₂ for a total fiftyfold excess of Ca²⁺, freeze-dried, and analyzed using suspension nebulization ICP-AES (Laird et al., 1991b). Calcium concentrations by this method are interpreted as equivalent to CEC. Clay extractions were performed in duplicate and reported elemental analyses were replicates of four determinations.

Plant analyses

In 1991, the above-ground biomass of a sweet corn crop in the center row of each plot was harvested 14 days past horticultural maturity, leaving 2.3-m margins at either end. The plant material was chopped, homogenized, subsampled, dried at 60 °C, ground to pass 40-mesh, and analyzed for N by Kjeldahl digestion and distillation and for Ca, Mg, K, P, and S by ICP-AES. Organic anion accumulation was calculated from plant analysis by the method of van Beusichem et al. (1988), with the assumptions that all plant N was organic N and the molar ratio of S_org/N_org was 0.032 (Dijkshoorn and Wijk, 1967).

Results

Effect of N fertilizer on pH and exchangeable acidity

A significant long-term effect of application of ammonium nitrate and urea N fertilizers at the fertility trials at Arlington, Wisconsin, was soil acidification. Values of pH measured in water and 1 M KCl (Fig. 1) are correlated with each other and with the N fertilizer rates applied for 30 yr. Mean treatment values of pH ranged from 5.6 to 4.8 when measured in water and from 5.5 to 4.1 when measured in 1 M KCl for 0 to 156 kg N ha^-1 yr^-1 treatments, respectively.

Exchangeable acidity of the Plano soil plots was strongly dependent upon the rate of N fertilization applied during the course of the 30-yr fertility trials (Fig. 2) but in a nonlinear manner. In the absence of applied N fertilizer, exchangeable acidity ranged from 0.05 to 0.30 cmol_c kg^-1, averaging 0.12 cmol_c kg^-1. The difference between the control plots (0 N) and 56 kg N ha^-1 yr^-1 is not statistically significant but exchangeable acidity of the 112 and 168 kg N ha^-1 yr^-1 treatments averaged 0.48 and 1.58 cmol_c kg^-1 greater than the control , respectively.

Assuming an average bulk density of 1.25 T m^-3, the top 20 cm of soil of the 112 and 168 kg N ha^-1 yr^-1 treatments contained 12 and 40 kmol(+) ha^-1 exchangeable acidity, respectively, which may be compared to the 240 and 360 kmol N ha^-1 added over the course of 30 years; this comparison indicates that 5 and 11% of the potential acidity added as ammoniacal N fertilizers remained in the top 20-cm as exchangeable acidity.

Effect of N fertilizer on exchangeable base cations, cation exchange capacity and base saturation

The increase in exchangeable acidity associated with N fertilization is not unexpectedly accompanied by a decline in exchangeable base cations (Fig. 3). The decline is most significant in exchangeable Ca²⁺, which ranges between 7.40 and 5.11 cmol_c kg^-1, a decline of 31% associated with the high N fertilization rate. A similar decline in exchangeable Mg²⁺, from 3.01 to 1.94 cmol_c kg^-1, or 36%, was significant. Individual monovalent cations show statistically significant trends in response to N fertilizer rates, such as higher exchangeable K in the 0 N treatment, presumably due to reduced biomass removal, and statistically higher NH₄⁺ in the 168 kg N ha^-1 yr^-1 treatments, probably due to inhibition of nitrification in the more acidified plots (Liu et al., 1996). However, for the purposes of this discussion of exchangeable bases, exchangeable K⁺, Na⁺ and NH₄⁺, have been summed and together represent a relatively small fraction of the total exchangeable bases, with no statistically significant response by their sum to N fertilizer levels.

Although exchangeable acidity increased most in those treatments in which exchangeable base cations were most depleted, the losses of exchangeable bases were not entirely offset by the addition of exchangeable acidity. The cation exchange capacity (CEC) of the long-term fertility plots, calculated as the sum of exchangeable base cations and exchangeable acidity and often termed effective CEC, showed a strong correlation to the rate of N fertilizer (Fig. 4). Base saturation fell from nearly 100% to 80% (Fig. 5), and would have fallen to yet smaller values but for the concomitant decline in CEC.

The mean CEC reduction in this study was 0.015 cmol_c kg^-1 per (kg N ha^-1 yr^-1 x 30 yr), or 5^.10^-4 cmol_c kg^-1 per kg N ha^-1 applied. The rate of CEC reduction is therefore below practical detection in a single-year field study, but other long-term fertility plots have similar findings. For example, Blevins et al. (1977) found that a Kentucky silt loam (Paleudalf) planted to no-till and conventional tillage corn lost as much as 4.3 cmol_c kg^-1 exchangeable bases, and gained 1.0 cmol_c kg^-1 exchangeable acidity (and 0.7 cmol_c kg^-1 Mn) in the 0 to 5 cm soil depth, upon fertilization with 336 kg N ha^-1 yr^-1 as NH₄NO₃ for 5 yr. The effective CEC was therefore reduced from 8.9 to 6.13 cmol_c kg^-1. At a fertilizer application rate of 168 kg N ha^-1 yr^-1, CEC was reduced to 7.1 cmol_c kg^-1 over 5 yr. This is a 31% and 20% reduction in CEC, respectively, over 5 years, similar to the results obtained in this study.

In other work, Schwab et al. (1989) reported that for a Kansas silt loam (Pachic Argiustoll) in bromegrass with up to 224 kg N ha^-1 yr^-1 for 40 yr, CEC measured by neutral salt (CaCl₂) decreased from 19.1 to 12.8 cmol_c kg^-1, compared to 0 N treatments. This is a 33% reduction in CEC over 40 years. The results of Bouman et al. (1995) show that a loamy Saskatchewan soil fertilized with up to 180 kg N ha^-1 yr^-1 as urea for 9 yr lost 1.8 cmol_c kg^-1 exchangeable cations (bases and acids) in the 0 to 15 cm soil depth, compared to a 0 N control. The increase in exchangeable acidity and decline in cation exchange capacity due to application of acid-forming N fertilizers observed in the Plano silt loam at the long-term fertility plots must therefore be seen as fitting into the context of long-term fertility studies in the United States that demonstrate the same phenomena and similar rates of change.

Effect of N fertilizer inputs on clay analyses

Clay analysis shows a 16% decline in exchangeable cations measured as Ca²⁺ concentration associated with long term N application, not reversed by Ca²⁺ saturation under laboratory conditions (Table 1). Analyses of Si, Mg, Ti, P, and Mn are almost identical among the duplicate extractions and the two samples, indicating that the clays were likely identical before alteration by acidity. A small increase in Al and Fe concentrations may indicate with accumulation of Al and Fe from other size fractions, either as oxides or hydroxy-interlayered nonexchangeable cations. The significance of the difference in K concentrations between the two samples is uncertain but may represent some weathering of illitic or micaeous minerals in the clay due to prolonged acidification.

Effect of plant growth and crop removal on exchangeable base cations and acidity

Based on dry matter yields and elemental concentrations of the corn crop in 1991, plants played a major role in the disposition of the acidity generated from the N fertilizers (Table 2). Native N fertility was high due to mineralization of soil N, despite the absence of added N for 30 years. With increasing N fertilizer rates, N fertilizer efficiency decreased from ~100% to 38%. Between one-half to three-quarters of the alkalinity produced by assimilation of N and S was sequestered in the above-ground portion of the plant and removed during harvest as organic anions and was therefore not available for return to soil for neutralization of acidity caused by nitrification. Relative to the control, the amount of residual acidity remaining in the soil increased with higher rates of N application and the associated decline in fertilizer usage efficiency. The maximum rate of acidification due to acid-forming fertilizers in 1991 is associated with the 168 kg N ha¹ treatment, which potentially left 8.9 kmol H⁺ ha^-1, or 75% of the potential acidity of 12 kmol H⁺ ha^-1 added as N fertilizer; it is of particular significance that 168 kg N ha^-1 rate is a commonly recommended and applied rate for cash grain production in this region. The long-term effect of cropping without addition of acid-forming fertilizers cannot be determined from this experiment because of the lack of archived samples for comparison.

The potential rates of acidification calculated from the Arlington long-term fertility experiment compare well with other studies in which rates of net acidification in various agricultural systems on the order of 3-5 kmol H⁺ ha^-1, and as much as 10-20 kmol H⁺ ha^-1 in exploitative systems, have been previously noted (Poss et al., 1995.) Rates of removal of assimilatory alkalinity as harvested organic anions depends upon both crop species and plant part; for example, grain of corn, sorghum, barley and oats will contain almost no organic anions while leaves of corn, barley, soybean, cotton, beets, and carrots will contain organic anions stoichiometrically equal to or in excess of the N content (Pierre and Banwart, 1973). Consequently, differences in crop management will affect acidification rates not only because of differences in N fertilizer usage efficiency but also because of differences in return of assimilatory alkalinity stored as organic anions: Crops harvested for grain, with residues either returned as litter or burned in the field, would be expected to return more alkalinity to soil and therefore have less acidifying effect from acid-forming nutrients than crops removed in toto for forage or fuel.

Changes in soil C and N associated with N fertilizer inputs

Evidence of subtle changes in the soil organic matter was sought, even though no above-ground crop residues had been returned for 30 yr. No significant trend was discernable individually for either quantity or isotopic ratio of C (1.64±0.20%C, ¹³C/¹²C =1.0924±0.0005 ) or N (0.15±0.02%N, ¹⁵N/¹⁴N = 0.3692±0.0006) in response to N treatment.

Discussion

Changes in CEC

The fate of the missing cation exchange capacity in this study and others (Blevins et al., 1977; Schwab et al., 1989; Bouman et al., 1995) is not known based on any direct evidence for these soils. Many soils have mixed constant and variable charge components (McBride, 1989) and CEC is pH-dependent. However, the many soils that are recognized as variable charge soils usually have a high sesquioxide content (free oxides of Fe, Al, and Si) or are kaolinitic, and are typically more weathered soils from more humid climates than this fertility study site (Uehara and Gillman, 1981). The Plano silt loam at the Arlington, Wisconsin trials is an ex-prairie soil and has a smectitic/illitic mineralogy that is quite unlike that of typical variable charge soils. Humic substances in the soil organic matter contain weakly acidic and very weakly acidic moieties and may therefore be regarded as variable charge components of soil that also exhibit charge dependence in response to pH and to electrolyte concentration (Barak and Chen, 1992) and cannot be eliminated out of hand as a cause of pH-dependent CEC. A rough calculation of the expected amount of organic charge reduction can be made using values from the often-cited study of 60 Wisconsin soils by Helling et al. (1964) which apportioned pH-dependent CEC by statistical correlation into clay and organic matter fractions. For an average difference of 0.86 pH units in water between 0 and highest N treatments at the Arlington, WI site, containing 1.6% C and 12% clay, the expected charge reduction is 0.71 cmol_c kg^-1 attributable to 1.6% C and 0.64 cmol_c kg^-1 attributable to 17% clay content. The total predicted pH-dependence of CEC is therefore 1.35 cmol_c kg^-1, which is less than half of the observed drop in CEC between these N treatments.

Other possible explanations for CEC reductions due to prolong soil acidification include both mineral weathering, including weathering of the clay minerals themselves, and formation of nonexchangeable hydroxy-Al complexes in the interlayer region of the clays. Acid dissolution studies conducted under laboratory conditions have shown that the rate of dissolution of montmorillonite clay minerals is correlated to the degree of substitution in the clay lattice (Shainberg et al., 1974). Work by Laird et al. (1991) has shown that the smectitic mineral in the fine clay fraction from a Minnesota soil (similar in parent material, climate, and vegetation to that at Arlington, WI) is a high-charge, Fe-rich montmorillonite with octahedral substitution in the soil clay exceeding that of the reference clays analyzed by Shainberg et al. (1974). This suggests that the soil clays at the Arlington site and elsewhere in the loess-derived soils of the Upper Midwest may be particularly prone to dissolution upon soil acidification. On the other hand, although all permanent negative charge in the phyllosilicate clays must be balanced with cations, not all of the cationic charge is necessary exchangeable. Formation of polynuclear hydroxy-Al interlayers in the 2:1 expansible clay minerals, either smectite or vermiculite, is associated with reduction in cation exchange capacity because the polynuclear hydroxy-Al material forms nonexchangeable "atolls" or "pillars" in the interlayer (Barnhisel and Bertsch, 1989). The process of formation of hydroxy-Al interlayers is a recognized form of soil weathering, often termed "chloritization" because of its tendency to form aluminous chlorite as an end product (Sposito, 1989).

The extent of reversibility of the charge reduction associated with soil acidity at the Arlington site is unknown. Of the three explanations for CEC reduction, clay mineral dissolution and chloritization are almost certainly not reversible processes within a soil environment, and certainly not reversible with aglime amendment. If the cation exchange capacity reduction is due to variable charge considerations, then the reduction may be reversible by pH amelioration and base supply with agliming, although hysteresis cannot be ruled out. Without studying the mechanism of charge reduction in this soil and the extent of the reversibility of the phenomenon, the extent to which soils have been permanently altered by common agronomic fertilization practices will be unresolved.

Geographic scope of CEC reductions

Evidence of CEC reduction due to N inputs, and its potential for reversibility, might be more widespread if repeated measurements of CEC of agricultural fields were made routinely or well-documented soil samples were more commonly archived. Instead, CEC is rarely measured in evaluation of soil fertility, let alone measured repeatedly on the same field or plot. Instead, evidence for CEC reduction comes only from long-term fertility studies, as noted in the studies above, and then only by side-by-side comparison with controls receiving no fertilizer N input rather than comparison with archived soil samples.

The consequences of this and related studies of acidification due to agricultural nutrient inputs are greater than agroecosystems since N-saturation, i.e., acidifying N inputs greater than the ability of the biota and soil organic matter to sequester, is implicated as a cause of acidification and continued decline of natural ecosystem saturation of natural ecosystems (Aber et al., 1989; Driscoll and van Dreason, 1992; Stoddard, 1994). Processes occurring in agricultural settings due to inefficiencies in N utilization of fertilizer N inputs may therefore present an accelerated view of the consequences of atmospheric N deposition under conditions of N saturation. For example, at the Arlington, Wisconsin site, precipitation-weighted mean atmospheric deposition includes 2.61 kg NH₄-N and 0.19 kg H⁺ (pH 4.63) ha^-1 yr^-1 (NADP, 1993), indicating an acid input of ~0.5 kmol ha^-1 yr^-1. Compared to agricultural inputs of 12 kmol ha^-1 yr^-1 in this study and cash-grain agriculture in this region, the rate of acidification due to fertilizer is underway at a rate that is ~25 times faster than the rate induced by acid rain. In such a case, the 30-year records established by long-term fertility plots might represent the effects of 750 years of soil weathering by acid rain and perhaps thousands of years by pristine rain on fertile soils of the Upper Midwest region. If the soils affected are part of relatively unmanaged ecosystems, alteration of soils from N deposition alone, although slower than acidification due to N fertilizers, will inexorably alter soil properties without any ameliorative steps likely to have been taken. It is fitting to at least learn from agroecosystems what subtle changes are taking place in natural ecosystems, even if there is no general management strategy to reverse those changes.

Conclusions

Side-by-side comparisons of soils that received various rates acid-forming fertilizer inputs for 30 years reveals reduction of pH measured in water and strong salt, accumulation of exchangeable acidity, decrease in exchangeable Ca²⁺ and Mg²⁺, and reduction of CEC associated with the rate of N fertilization. Clay analysis showed that the CEC losses of 15% were not recoverable by saturating with 50-fold excess of Ca²⁺ under laboratory conditions. Analysis of above-ground plant material shows that increasing N fertilization rates cause decreasing N fertilizer efficiency, leaving a larger portion of the acidity generated by nitrification in the soil. Plants sequestered between one-half and three-quarters of the alkalinity generated by N and S assimilation in the harvestable plant material, although harvesting grain only and retaining plant residue to soil would decreased the acidifying effect of fertilization.

Acknowledgments

Thanks are due to Mauricio Avila for assistance in preparing yield data and clay analyses. Financial support by the National Agricultural Research Institute of the Gambia and the Graduate School of the University of Wisconsin-Madison is gratefully acknowledged.

References

Aber J D, Nadelhoffer K J, Steudler P and Melillo J M 1989 Nitrogen saturation in Northern forest ecosystems. BioSci. 39, 378-386.

Barak P and Chen Y 1992 Equivalent radii of humic macromolecules from acid-base titration. Soil Sci. 152, 184-195.

Barak P, Sherman L A and Jobe B O 1996 Comments on "Design and construction of a personal-computer-based automatic titrator". Soil Sci. Soc. Am. J. 60, 630.

Barnhisel R I and Bertsch P M 1989 Chlorites and hydroxy-interlayered vermiculite and smectites. In Minerals in Soil Environments. 2nd Ed. J B Dixon and S B Weed. pp. 729-788. Soil Sci. Soc. Am., Madison, WI.

Blevins R L, Thomas G W and Cornelius P L 1977 Influence of no-tillage and nitrogen fertilization on certain soil properties after 5 years of continuous corn. Agron. J. 69, 383-386.

Bouman O T, Curtin D, Campbell C A, Biederbeck V O and Ukrainetz H 1995 Soil acidification from long-term use of anhydrous ammonia and urea. Soil Sci. Soc. Am. J. 59, 1488-1494.

Brimhall G H, Lewis C J, Ford C, Bratt J, Taylor G, Warin O 1991 Quantitative geochemical approach to pedogenesis: importance of parent material reduction, volumetric expansion, and eolian influx in laterization. Geoderma 51, 51-91.

Cox G W and Atkins M D 1979 Agricultural ecology. W H Freeman, San Francisco.

Dijkshoorn W and van Wijk A L 1967 The sulphur requirement of plants as evidenced by the sulphur-nitrogen ratio in the organic matter. A review of the published data. Plant and Soil 26, 129-157.

Dise N B and Wright R F 1995 Nitrogen leaching from European forests in relation to nitrogen deposition. For. Ecol. Manag. 71, 153-161.

Driscoll C T and van Dreason R 1992 Seasonal and long-term temporal patterns in the chemistry of Adirondack lakes. Water Air Soil Pollution, 319-344.

Edwards A P and Bremner J M 1965 Dispersion of mineral colloids in soils using cation exchange resins. Nature (London) 205, 208-209.

Helling C S, Chesters G and Corey R B 1964 Contribution of organic matter and clay to soil cation-exchange capacity as affected by the pH of the saturating solution. Soil Sci. Soc. Am. Proc. 28, 517-520.

Laird D A, Barak P, Nater E A and Dowdy R H 1991a Chemistry of smectitic and illitic phases in interlayered soil smectite. Soil Sci. Soc. Am. J. 55, 1499-1504.

Laird D A, Dowdy R H and Munter R C 1991b Suspension nebulization analysis of clays by inductively coupled plasma­atomic emission spectroscopy. Soil Sci. Soc. Am. J. 55, 274-278.

Liu Y 1995 Fixation and release of NH₄ and K in long-term fertility plots. MSc. thesis, Iowa St. Univ. Ames, IA.

Liu Y, Laird D A and Barak P 1996 Dynamics of fixed and exchangeable NH₄ and K in soils under long-term fertility management. Soil Sci. Soc. Am. J. 60, (in press).

McBride M B 1989 Surface chemistry of soil minerals. In Minerals in Soil Environments. 2nd ed. J B Dixon and S B Weed. pp. 35-88. Soil Sci. Soc. Am., Madison, WI.

Meister R T (Ed.) 1995 Farm chemicals handbook '95. Vol. 81. Meister Publ. Co., Willowby, OH.

National Atmospheric Deposition Program 1993 NADP/NTN Annual Data Summary. Precipitation Chemistry in the United States. 1992. National Resource Ecology Laboratory, Colorado State Univ., Fort Collins, CO. 480 pp.

Odum E P 1984 Properties of agroecosystems. In Agricultural Ecosystems: Unifying Concepts. Ed. R Lowrance, B R Stinner and G J House. pp. J Wiley, New York.

Peterson LA and Krueger A R 1980 Variation in content of available P and K (Bray I) in soil samples from a cropped N, P, and K fertility experiment over 8 years. Commun. Soil Sci. Plant Anal. 11, 993-1004.

Pierre W H and Banwart W L 1973 Excess-base and excess-base/nitrogen ratio of various crop species and parts of plants. Agron. J. 65, 91-96.

Poss R, Smith C J, Dunin F X and Angus J F 1995 Rate of soil acidification under wheat in a semi-arid environment. Plant and Soil 177, 85-100.

Schwab A P, Ransom M D and Owensby C E 1989 Exchange properties of an argiustoll: effects of long-term ammonium nitrate fertilization. Soil Sci. Soc. Am. J. 53, 1412-1417.

Shainberg I, Low P F and Kafkafi U 1974 Electrochemistry of sodium-montmorillonite suspensions: I. Chemical stability of montmorillonite. Soil Sci. Soc. Am. Proc. 38, 751-756.

Sposito G 1989 The chemistry of soils. Oxford Univ. Press, New York.

Steel R G D and Torrie J H 1980 Principles and procedures of statistics: A biometrical approach. McGraw-Hill Inc., New York.

Stoddard J 1994 Long-term changes in watershed retention of nitrogen: Its causes and consequences. In Environmental Chemistry of Lakes and Reservoirs. Ed. LA Baker. pp. 223-284. Am. Chem. Soc.

Uehara G and Gillman G 1981 The mineralogy, chemistry, and physics of tropical soils with variable charge clays. Westview Press, Boulder, CO.

Van Breemen N, Mulder J and Driscoll C T 1983 Acidification and alkalinization of soils. Plant and Soil 75, 283-308.

Van Beusichem M L, Kirkby E A and Baas R 1988 Influence of nitrate and ammonium nutrition on the uptake, assimilation, and distribution of nutrients in Ricinus communis. Plant Physiol. 86, 914-921.

Tables

Table 1. Elemental analysis of Ca²⁺-saturated <2 µm clay from adjacent plots receiving either 0 or 168 kg ammoniacal N ha^-1 yr^-1 for 30 years. Oxygen completes analysis to 100%. Element-by-element relative standard deviations (RSD) are pooled for 4 measurements of each of four samples.

Table 2. Effect of plant-generated alkalinity on residual soil acidity from acid-forming N fertilizer treatments. All values are means of 18 plots; plots receiving no P or K fertilizers for 30 years were omitted from analysis. Standard deviations of the mean are enclosed in parentheses.

Figures

Fig. 1 Correlation between pH measured in water (1:1) and in 1 M KCl for samples from 30-yr fertility plots and mean pH values for N fertilizer treatments. Ellipse shows standard deviations of the means.

Fig. 2 Relationship between exchangeable acidity and N fertilizer rates in long-term fertility plots. Error bars indicate standard deviation of the mean. Means indicated by the same letter are not significantly different at the 5% level by ANOVA.

Fig. 3 Relationship between exchangeable Ca2+, Mg2+, and K++Na++NH4+ and N fertilizer rate in long-term fertility plots. Mean comparisons are valid for the same measure across treatments. Means indicated by the same letter are not significantly different at the 5% level by ANOVA.

Fig. 4 Correlation between effective CEC, the sum of exchangeable bases and exchangeable acidity, and N fertilizer rates in 30-yr fertility plots. Error bars indicate standard deviation of the mean. Means indicated by the same letter are not significantly different at the 5% level by ANOVA.

Fig. 5 Relationship between base saturation and N fertilizer rates in long-term fertility plots. Error bars indicate standard deviation of the mean. Means indicated by the same letter are not significantly different at the 5% level by ANOVA.

This is a draft manuscript prepared for peer-review and pre-review purposes only. No publication in any format is authorized by the authors. All rights reserved. Dec '96