CAUSES OF CHANGES IN pH IN ACIDIC MINERAL ... - Science Direct

13 downloads 0 Views 1MB Size Report
We have determined and measured some processes which are likely to contribute to pH changes within these acidic, mineral soils. The studies were undertaken ...
Soil Biol. Biochem. Vol. 27, No. 1I, pp. 1383-1392, 1995

Pergamoln

Copyright 0 1995Elsevier Science Ltd Printed in Great Britain. All rights reserved

0038-0717(95)00083-6

0038-0717/95 $9.50 + 0.00

CAUSES OF CHANGES IN pH IN ACIDIC MINERAL

SOILS

M. K. CONYERS,‘* N. C. UREN* and K. R. HELYAR’ lAgricultural Research Institute, Pine Gully Road, Wagga Wagga, NSW 2650, Australia and 2School of Agriculture, La Trobe University, Bundoora, Vie. 3083, Australia (Accepted

5 June 1995)

Summary-Seasonal variation in soil pH and long-term decreases in pH have been observed for the soils of the south-western slopes of New South Wales. We have determined and measured some processes which are likely to contribute to pH changes within these acidic, mineral soils. The studies were undertaken on closed systems in the absence of plants. Changes in soil pH, in the absence of plants, were controlled by the relative rates of alkali-producing (acid-consuming) reactions and of acid-producing (alkaii-consuming) reactions. The alkali-producing reactions studied were ammonification, reduction of Mn-oxides, oxidation of organic anions and SOf - adsorption. The acid-producing reactions were nitrification, oxidation of Mn*+ and oxidation of organic S. Measurements of changes in the concentration of the chemical species involved in these reactions were converted to the corresponding changes in pmol H+ or OH - g-l soil. The sum of the H+ budg:et for each soil was transformed to change in pH by using the soil’s titration curve. With this procedure 96% of the variance of the observed changes in soil pH was explained. On the basis of our results we propose 1:hatchanges in temperature and water potential determine changes in microbiological activity which in turn determine changes in the H+ budget and its physicochemical consequences.

INTRODUCTION

on causes of pH change in soil tends to be focused on long term acidification, due either to acidic deposition (Kennedy, 1992) or to agricultural practice (Helyar and Porter, 1989). There is, additionally, a body of literature on seasonal variation in soil pH (Friesen et al., 1985; Skyllberg, 1991) which indicates that soil p.H undergoes cycles of decrease and increase. There are therefore internal reactions which neutralize H+ as well as produce H+. While a concept of the proton cycle has been put forward (Verstraten et al., 1990) we have regarded H+ consumption or production as being primarily a result of the metabolic processes involved in the traditional elemental cycles. Deposition studies have focused on the N and S cycles (Kennedy, 1992) and agricultural acidification studies have concentrated on the C and N cycles (Helyar and Porter, 1989). Studies on seasonal variation in soil pH either give inference only to the likely mechanisms or explain the variation by processes appropriate to a particular environment. For example, Brinkman (1977) coined the term “ferrolysis” for a p:rocess which explained the effects of seasonal water inundations on soil acidity, while Topark-Ngarm et a!. (1990) noted the effects on pH of large changes in the concentration of soluble salts in soils of north-eastern Thailand. Under less extreme climatic variation, seasonal variation in soil pH of the order of 0.2-0.5 pH units has been recorded by us on the south-western sl.opes of New South Wales. Field observations indicated that there were large decreases The literature

in soil pH during autumn, prior to significant pasture growth or to sowing of crops. These changes were associated with increasing concentrations of NOT and decreasing concentrations of extractable MI?+. Soil pH increased again during late spring and early summer. Our purpose was to determine and measure processes which may be responsible for changes in pH within some acidic, mineral soils of the south-western slopes of New South Wales. Theoretical

The equations given in this section summarize the biological and chemical reactions we used to account for the observed changes in soil pH. The list of reactions is not comprehensive but covers the major reactions which are likely to occur in these acidic, mineral soils and under the experimental conditions. In the case of microbiologically-mediated reactions the equations represent the net reaction with respect to the external environment of the organism. Nitrogen. The first step in the decomposition of organic N is usually described by the enzymatic deamination process: RI-CH(NHJ-R

+ ;o+R’-CO-R

+ NH~,

(1)

where R, R’ represent organic groups. The hydrolysis of NH3 is an alkali-producing reaction: NH3 + HlO+NH: Chemoautotrophic

*Author for correspondence. 1383

+ OH - .

(2)

bacteria may use the oxidation

M. K.

1384

Conyers et al.

of NH: as an energy source. For simplicity the oxidation (nitrosification) is given as the net reaction: NH: + ;O+NOc

+ H20 + 2H’.

The second oxidation step (nitrification) NO,

(3)

is:

+ ;Oz-+NO, .

(4)

Unless there are adverse conditions, such as the concentration of NH, substrate being high, little NOT accumulates in soil. The net overall reaction of NH4+ to NO< can be expressed as NH: + 202+NO,

+ 2H+ + HZO.

(5)

The concentrations of (NHt + NH,) and NO, are readily measurable quantities. Hence, depending on the relative rates of mineralization of N and nitrification, this sequence of reactions of the N cycle can cause the soil to undergo an increase in pH, no change or a decrease in pH. Heterotrophic nitrification is an additional consideration to the simple sequence of mineralization outlined here. Our associated studies suggest that heterotrophs make only a very small contribution to nitrification in these mineral soils. Denitrification was not measured during our aerobic incubation experiments. Recoveries of added NO< exceeded 95%. Carbon. Since no plant material was added to or removed from the soil during our experiments, the task was to measure changes in the H+ budget due to transformations of native soil organic C. Organic compounds with high relative charge tend to be of lower molecular weight and more soluble in acidic and alkaline solutions (Hayes and Swift, 1978). Therefore, we assumed that change in soil pH within the conditions of our experiments is likely to have principally involved the oxidation of water-soluble organic anions: C,Ht-

10; + H’ + xO+xCOz

+ xHz0.

(6)

Unlike other oxidation reactions this is an alkali-producing reaction. The distinction between the association of H+ with organic anions and the oxidation of organic anions is not always clear. Experimentally, the quantification of this reaction has been elusive. However, Oliver et al. (1983) demonstrated that polyprotic humic acids can be operationally described by an empirical equation relating pK and the pH of the aqueous suspension. Early studies on the pK values for the “carboxyl” functional range of dissolved organic C gave values in the range 3.5-5.0 (Cronan and Aiken, 1985; Eshleman and Hemond, 1985). Since dissolved organic C is polyprotic there is a range of pK in any given sample. As the pH of the sample changes so does the apparently dominant pK. The advance of the study by Oliver et al. (1983) was to confirm that the range of pK for dissolved organic C was 3.5-5.0 but more importantly that the pK varied within that range

according to sample pH: average pK = 0.96 + 0.90pH - 0.039(~H)~. (7) The more complex model proposed by Bonn and Fish (199 1) was a bimodal continuum of functional groups centred on a mean pK of 4.5 for the “carboxyl” group and 10.0 for the “phenolic” group. This same empirical approach however, philosophically underlies the simpler approach of Oliver et al. (1983) and of Thurman (1985). The merit of the relationship between pK and pH proposed by Oliver et al. (1983) was confirmed for water samples by Tipping et al. (1991) and Bisogni and Arroyo (1991). The calculation procedure of Bisogni and Arroyo (1991) allows the concentration of organic anions, A-, to be obtained: [A-l = RGo,l/U?

+ [H’I),

(8)

where K = 10 -pK and CToT= dissolved organic C (mg C 1-l) x 6.5 (pequiv mg-’ C); I? and concentrations are in gequiv 1-l. Note that the measured dissolved organic C includes A- and HA (i.e. dissociated and associated acid forms) so that K appears in the denominator as well as in the numerator. The value of 6.5 pequiv mg-’ C is based on results for the average “carboxyl” content of dissolved organic C. The “carboxyl” content is an operational definition and includes carboxyl, enolic, sulphonyl and hydroxyquinone functional groups (Thurman, 1985). We have used the value of 6.5 pequiv of “carboxyl” groups per mg C from Tipping et al. (1991), which agrees well with the range of 6-7 mequiv H+ g-’ C of Cronan and Aiken (1985) and lies between the 7.5 mequiv H+ g-’ C of Eshleman and Hemond (1985) and the 5 pmol H+ mg-’ C of Driscoll and Bisogni (1984). Sulphur. S-containing amino acids such as cysteine and methionine are microbially degraded to elemental S, and in the presence of sufficient 02, to SO:(Kennedy, 1992). The initial steps are: R-SH + H>O+R-OH

+ H&S =2H+ + S2- + R-OH

(9)

and S? - + ;Opz+ 2H+ -S + HZO.

(10)

The hydrolysis and oxidation steps are neutral to this point. Chemoautotrophs may then oxidize the S: 2S+02+OH-+S@-

+H+

(11)

and SzO:- + 202 + OH--+2SOf-

+ H+.

(12)

In these latter two steps, 2 mol of base are consumed and 2 mol of acid are produced per 2 mol of S. Hence the overall oxidation of S2- to S6+ produces 2 mol H+/mol S. A potential subsequent reaction is the adsorption of SOi- onto the surfaces of iron oxides. This may be an

Causes of soil pH change

alkali-producing reaction but the stoichiometry varies (Turner and Kramer, 1991). Manganese. The reduction of Mn oxides and the oxidation of Mn2+ consumes and produces protons, respectively: Mn02 + 4H+ + 2e- +Mn2+ + 2H20

(13)

and Mn2+ + ;(I2 + H20+Mn02

+ 2H+.

(14)

The electrons for the reduction reaction equation (13) come from the oxidation of organic C: ~C,H2,0.Y + ixH+3+2xH+

+ 2xe- + ixC02.

(15)

When x = 1 for simplicity, Mn02 + fCH20 + 2H++Mn2+

+ $02 + iH20. (16)

Manganese oxides also occur in soils with Mn in the 2 + and 3 + oxidation states, sometimes in mixtures of oxidation states (Gilkes and McKenzie, 1988; McKenzie, 1989). The following equations are given as examples to demonstrate that dissolution of an Mn-solid consumes 2H+ mol-’ of Mn, regardless of the oxidation state(s) of Mn in the solid phase. The generalized organic C molecule of C.VH2.V0, has been simplified here by making x = 1: MnOOH+3H+

Mn’+

$CH20 + $H20+H+

The sum of these latter two equations is measurable as the effective cation exchange capacity of the soil. Relationships between processes: the model. It is necessary to draw a distinction between processes which alter the H+ budget and those which buffer soil pH. The first and most intuitive distinction is that processes which alter the H+ budget are active, biological processes while the other processes, such as equations (19)-(2 l), are passive, chemical responses to the activity of protons (H+). This distinction is not always clear since the reduction of Mn oxides to Mn2+ can be an abiotic or a biological process. A second distinction is that some processes consume or produce H+ while other processes simply “store” and release protons by reversible adsorption-desorption or association-dissociation processes. The distinction between H+ adsorption and clay dissolution (H+ consumption) is probably not clear and is largely a kinetic matter. Despite the non-definitive nature of these distinctions, a 2 x 2 classification of soil processes involving H+ can be made in the following way:

PROCESSES

Production/ consumption

+ e- +Mn2+ + 2H20 + e- + $0,

MnOOH -I- 2H+ + $ZH20

Mn(OH)2 -t- 2H+ +Mn2+ + 2H20 .

(18)

Si(OH)4Al(OH)j + 3H+ +A13+ + 3H20 + H4Si04.

(19)

of Al and Si are

Other reactions kvolving protons. Equations (20) and (21) represent reversible dissociation-association and desorption-adsorption reactions:

R-COOH=R-COO

- + H+

(20)

and -M-OH+-M-O where M = (Al, Fe or Si). SBB 27/l

I-c

CHEMICAL

(active) equations (I-5)

(passive) equation (19) equations ( 16)-( 18)

equation (6) equations (11) and (12) equations (14), (16)-(18)

(17)

Hence both the reduction and oxidation of Mn involves a change of 2H+ mol-’ of Mn. The reduction process may be abiotic (chemical) or enzymatic while the oxidation of lMn2+ under acidic conditions is undertaken by some heterotrophs and photoautotrophs(Bromfield, 1958, 1976, 1978; Ghiorse, 1988). Aluminium. Equation (19) represents a generalized reaction for the dissolution (forward) or precipitation (reverse) of an aluminosilicate mineral. When the concentration of {Si is relatively low the reverse reaction represents the hydrolysis of Al:

The changes in the concentrations readily measurable.

BIOLOGICAL

1

+Mn2+ + :C02 + iH20; Mn2+

1385

- + H+,

(21)

3

2 Storagerelease

Root CEC

Zquation equation

(20) (21)

The pH buffering intensity, pHt3, can be defined as the set of processes which respond passively to the addition of H+ and OH and are thermodynamically reversible under the range of pH and pe conditions and in the time frame of an experiment. The chemical processes in boxes 3 and 4, with the exception of the chemical reduction-oxidation of Mn, are encompassed by this definition. The chemical oxidation of Mn’+ does not readily take place at pH < 7, and so it does not fit the criterion of being thermodynamically reversible in acidic soils. The acknowledged ambiguity which remains is that MnO? reduction can be an abiotic process and so could be within box 3, or a biological process within box 1. A further potential problem with the criterion of thermodynamic reversibility is that as the time frame extends, mineral dissolutions may become more significant and some changes may not be thermodynamically reversible. For example, a hysteresis in Al solubility may occur if the liming of an acidified gibbsite-free soil results in the precipitation of Al(OHb. This could result in the pHD being different for alkali addition to that for acid addition. The root CEC is offered as a possible process

M. K. Conyers et al.

1386

influencing (H+) in the biological storage-release box 2, although the effect of plants is not dealt with in this study. In this paper we examine the set of processes in box 1 and their interaction with the overall summation, except for Mn02+Mn2+, of boxes 3 and 4 as the pHl3. The simplest expression of these concepts (22)

is: Ajbiological H+} x pHJ = estimated ApH, where A {biological A(pmol H+)

H+} = equations

(l)-( 18) =

and pHl3 = equations (19-21) = ApH/@mol H+). The pHD is empirically derived as the slope of the titration curve of a soil. Therefore, in contrast to current concepts such as acid and base neutralizing capacity (Ulrich, 1991), we have separated biochemical and physical chemical processes in our model. In the experiments which we describe here the measured changes in soil pH were compared with the estimated ApH [equation (22)] based on the measurement of changes in the concentrations of the chemical species involved in biological acid-producing and acid-consuming reactions. MATERIALS

AND METHODS

Soils

The four principal soils used are described in Table 1. The remaining soils, which are taken from eastern New South Wales, are described in Conyers et al. (1991). The clay fractions of all four of the principal soils were dominated by illite, kaolinite and quartz. Chemical methods

Soil pH,, was routinely measured at constant ionic strength using 10 mM CaClr at a 1:5 soil-to-solution ratio. Soil pH, was also measured in a 1:5 aqueous extract in Experiment 4. Samples were mixed end-over-end for 1 h at 5 rev min-I. A combination glass-calomel electrode was calibrated with buffers of pH 6.88 (25 mM KH,PO, + 25 mM Na2HP04) and 4.00 (50 mM K-phthalate) at 20°C. The soils were titrated batchwise, with varying ratios of 10 mM CaClr and standardized 10 mM Ca(OHh to give a total volume of 25 ml for each 5 g

soil sample. The titration:; of the soil samples were therefore conducted at constant ionic strength and a constant soil-to-solution ratio. The concentration of Mn’+ in the 10 mM CaClr extracts, [MnJ. was measured by atomic absorption spectroscopy using a Varian 875 spectrophotometer. The extractable [Mn,,] was generally 8@-100% of that determined to be exchangeable by the Gillman and Sumpter (1986) method. The subscript “ex” refers to an exchangeable cation and the effective cation exchange capacity is denoted as ECEC. Mineral N (NH: and NO, ) was extracted from soil with 1 M KC1 at a 1:5 soil-to-solution ratio. Samples were extracted by mixing end-over-end for 1 h at 5 rev min-‘. The [NH,‘] was determined as the emerald-green NHrsalicylate complex at 660 nm (Crooke and Simpson, 1971) and [NO