Effect of leaching and clay content on carbon and nitrogen mineralisation in maize and pasture soils.
Abstract: We conducted a 7-week laboratory incubation experiment to evaluate the effect of leaching on net C and N mineralisation in soils. The soils were collected from adjacent fields of long-term pasture and maize, where each field contained an Inceptisol and an Andisol. The concentration of clay mineral was 200 g/kg halloysite in the Inceptisol and 120 g/kg allophane in the Andisol. Half the samples were leached weekly with 0.002 M Ca[Cl.sub.2] at a suction of 20 kPa to remove soluble products, and half were not leached. Carbon mineralisation was determined from [CO.sub.2]-C evolved each week. Net N mineralisation was measured for the leached samples from the [NH.sub.4]-N and [NO.sub.3]-N in the Ca[Cl.sub.2] extracts, and for the batch of non-leached samples by extraction in 0.5 M [K.sub.2][SO.sub.4]. Carbon and net N mineralisation were greater in the soils under pasture than in soils under maize. The proportion of total C mineralised as [CO.sub.2]-C, and of total N mineralised as [NH.sub.4]-N and [NO.sub.3]-N, followed the order Inceptisol-pasture [is greater than] Inceptisol-maize [is greater than] Andisolpasture [is greater than] Andisol-maize, suggesting that allophane and Al ions reduced net mineralisation. Dissolved organic carbon (DOC) produced during incubation, as a proportion of total C, was greatest for the Inceptisol-maize sample and least for the Andisol-pasture sample. Non-leaching resulted in the accumulation of acids and solutes, and decreased C mineralisation for the Inceptisol samples.

Additional keywords: aerobic incubation, allophane, C mineralisation, New Zealand soils.
Article Type: Statistical Data Included
Subject: Soils (Leaching)
Clay (Environmental aspects)
Soil research (Analysis)
Carbon (Environmental aspects)
Nitrogen (Environmental aspects)
Agriculture (Environmental aspects)
Humus (Environmental aspects)
Authors: Parfitt, R. L.
Salt, G. J.
Saggar, S.
Pub Date: 05/01/2001
Publication: Name: Australian Journal of Soil Research Publisher: CSIRO Publishing Audience: Academic Format: Magazine/Journal Subject: Agricultural industry; Earth sciences Copyright: COPYRIGHT 2001 CSIRO Publishing ISSN: 0004-9573
Issue: Date: May, 2001 Source Volume: 39 Source Issue: 3
Geographic: Geographic Scope: New Zealand Geographic Code: 8NEWZ New Zealand
Accession Number: 75434491
Full Text: Introduction

Soil organic carbon (C) and nitrogen (N) are important both as indicators of soil quality (Doran et al. 1994) and for sustaining agricultural production. They can change as a result of modifications in land cover or land management, and may influence long-term ecosystem sustainability. Therefore, further information on key processes, such as mineralisation of C and N, and how they are influenced by different soils and crops is needed.

Measurement of C and N mineralisation during laboratory incubations has become a common method for assessing C and N mineralisation potential in different soils with different crops. The leaching incubation procedure of Stanford and Smith (1972) has been used widely to determine N mineralisation potential. This procedure has the advantages of simulating field conditions, and avoiding the artificial influences of ammonium or nitrate build up and decreases in soil pH. There do not, however, appear to be any studies on the effects of this leaching technique on C mineralisation during laboratory incubations. Therefore, in this paper, we compare C and N mineralisation results in a 7-week laboratory incubation both with and without leaching. Since clay surfaces reduce the turnover of soil organic matter (Saggar et al. 1999), we have chosen 2 soils with contrasting clays to also obtain information on the effect of clay mineral on mineralisation.

Materials and methods

Site descriptions

Mineral soil (0-10 cm) was sampled from a farm in New Zealand that had 2 soil types and 2 land uses, i.e. pasture and maize that existed on both soil types. The farm was located in the central North Island at 175 [degrees] 24'E, 37 [degrees] 55'S. Annual precipitation is approximately 1230 mm, and mean annual air temperature is 13.4 [degrees] C. The soils were Horotiu sandy loam, classified as a Typic Hapludand (Vitric Orthic Allophanic Soil in the New Zealand Soil Classification; Hewitt 1998), and Te Kowhai silt loam, an Aeric Endoaquept (Typic Orthic Gley Soil; Hewitt 1998). Both were formed in rhyolitic alluvium and tephra, and they had different particle size distribution (Table 1). The site had previously been under native broadleaf/podocarp forest that was cut in the 19th Century. The perennial ryegrass (Lolium perenne L.)-clover (Trifolium repens L.) pasture had not been ploughed within the last 50 years. The pasture was grazed regularly and fertilised (30 kg P and 20 kg K/ha.year). The maize (Zea mays L.) field was first planted in 1971 and had been cropped with maize each year since then, and left fallow in winter. Lime (0.5 t/ha) and 55 kg N, 63 kg P, and 60 kg K were applied before planting, and the maize was later side-dressed with 185 kg N/ha as liquid urea. The yield of grain was 13.6 t/ha in 1996.

Sample collection and preparation

Using a 150-mm-diameter corer, we sampled mineral soil (0-10 cm depth) from the 4 sites on 19 May 1997, before the grain was harvested, and when the soils were at field capacity. Three replicate samples were collected from random positions from under maize and under pasture on both the Andisol and Inceptisol, giving a total of 12 soil samples. All samples were then sieved moist through a 5-mm sieve and stored for several weeks at 4 [degrees] C before commencing the experiment. Moisture content was determined by drying at 105 [degrees] C.

Incubation with leaching

Duplicate samples of field-moist soil (20.0 g oven-dry equivalent) were weighed into 150-mL vessels fitted with 50-mm-diameter GFC (glass microfibre filters); a suction could be applied to the vessel and leachate could be collected. There were 2 soil types, under 2 crops, with 3 field replicates and 2 laboratory replicates, giving 24 vessels. Each sample was placed under suction at 20 kPa and deionised water was added dropwise until a few drops of leachate were collected. After no more water could be collected, the vessels were removed from the suction and each placed in a I-L sealed glass jar fitted with a septum for [CO.sub.2] production measurements, and incubated at 25 [degrees] C in the dark. After 7 days, [CO.sub.2] production was measured by sampling head-space [CO.sub.2] concentrations and injecting a 1-mL sample into a gas chromatograph equipped with a thermal conductivity detector. The jars were then opened and samples removed. The soil solution was displaced from the samples by washing with 20 mL 0.002 M Ca[Cl.sub.2] into a clean flask under-20 kPa suction, using a vacuum pump. After no more solute could be collected, the leachate was retained for analysis of [NH.sub.4]-N and [NO.sub.3]-N, and dissolved organic carbon (DOC). The 1-L glass jars were then flushed with ambient air, and the soil (at -20 kPa) and vessels replaced within the jars, which were resealed and returned to the incubator. The leaching procedure was repeated every 7 days.

Incubation without leaching

Field-moist soil (20.0 g oven-dry equivalent) was weighed into 125-mL polypropylene containers, and water was added to adjust all samples to -20 kPa. Both the leaching and non-leaching experiments, therefore, had the very similar initial conditions. For [CO.sub.2] production measurements, 2 replicates of each sample were placed in a 1-L sealed glass jar fitted with a septum. For N mineralisation measurements, sufficient containers were prepared to allow for destructive sampling each week over the 7-week incubation. The containers were covered with polyethylene and placed on plastic trays containing water, enclosed in large polyethylene bags (to maintain high humidity), and incubated at 25 [degrees] C in the dark. The samples were extracted weekly with 50 mL 0.5 M [K.sub.2][SO.sub.4] by end-over-end shaking for 30 min, and filtration through MFS 5C (Whatman 40 equivalent) papers. The filtrates were retained for analysis of [NH.sub.4]-N and [NO.sub.3]-N. The initial values for each replicate for Week 0 were subtracted from the weekly data to give the net N mineralisation values.

Analytical procedures

Soil pH was measured on a 1:2.5 w/v mixture of soil and water, available P by the Olsen method, total P by Kjeldahl digest, and pyrophosphate-Al ([Al.sub.py]) by extraction at pH 10 (Blakemore et al. 1987). Total C and N were measured on a LECO FP-2000 CNS analyser. Particle size was measured by dispersing the fieldmoist soil in water with an ultrasonic probe and separating the [is less than] 2 [micro]m, 2-60 [micro]m, and [is greater than] 60 [micro]m fractions by sedimentation. Clay minerals were estimated by acid-oxalate extraction and differential thermal analysis (Parfitt and Wilson 1985).

Ammonium-N ([NH.sub.4]-N) concentrations (by reaction with salicylate/nitroprusside) and [NO.sub.3]-N concentrations (by diazotization following hydrazine reduction) were measured on a Technicon AutoAnalyzer II system (Blakemore et al. 1987). Dissolved inorganic and dissolved organic C were measured by Shimadzu TOC analyser. One of the field replicates from the Andisol-maize plot had a low pH (5.3 compared with 6.2 for the other 2 replicates) and a very high initial N status, possibly arising from an excess of N fertiliser. The Ca[Cl.sub.2] data for this replicate have been corrected by subtracting the excess amount of N as indicated in the first [K.sub.2][SO.sub.4] extract.

Statistical analysis

As the relationship between the dependent variable (e.g. C mineralisation, N mineralisation, DOC, pH) and time was close to linear for each factor combination (leaching x soil sample), we used a linear model in addition to non-linear models to fit the data. The linear model provided the best fit (higher [R.sup.2] values). We then used this slope (rate) for each replicate, and a 2-factor analysis of variance, to test whether these rates differed between treatments. In estimating the rates for each replicate, the intercept was constrained to be equal to zero. Following ANOVA, contrasts of treatment means for significant effects were made, using a Bonferroni adjustment to calculate P-values.

Results

The Inceptisol contained more clay than the Andisol; the clay minerals were predominantly halloysite in the Inceptisol, and allophane in the Andisol (Table 1). Organic C and total N (measured in g/kg soil) followed the order: Andisol-pasture [is greater than] Andisol-maize [is greater than] Inceptisol-pasture [is greater than] Inceptisol-maize; the C:N ratios ranged from 10.8 to 11.6. The Olsen P values were higher for the maize soils than for the pasture soils.

The effects of leaching on C mineralisation rate, expressed as a fraction of total soil C, were not independent of soil, and therefore the soils were considered separately ([F.sub.3,40] = 20.8, P [is less than] 0.001). Fig. la shows that values of C mineralisation rate, expressed on this basis, were significantly different (P [is less than] 0.01) for all leached samples, and followed the order: Inceptisol-pasture [is greater than] Inceptisol-maize [is greater than] Andisol-pasture [is greater than] Andisol-maize. Mineralisation rate of C differed between the non-leached samples (P [is less than] 0.03) (Fig. 1b), except that the mineralisation rate for the Inceptisol-maize was not significantly different from that of the Inceptisol-pasture and the Andisol-pasture (P = 1 and P = 0.24). The mineralisation rate was significantly higher for the leached samples (Fig. 1 a) than for the non-leached (Fig. 1 b) samples of both Inceptisols (P [is less than] 0.001), but not of the Andisols (P = 1).

[ILLUSTRATION OMITTED]

The mineral-N produced during incubation was largely [NO.sub.3]-N, with [NH.sub.4]-N being less than 5% of the total mineral-N. There was strong evidence that the soils behaved differently and this depended on the degree of leaching ([F.sub.3,28] = 11.1, P [is less than] 0.001). With leaching, net N mineralised (per unit of total N) was significantly greater for the Inceptisol-pasture and Inceptisol-maize (P [is less than] 0.001) than for Andisol-pasture and Andisol-maize (Fig. 2a). Without leaching, net N mineralisation rate (Fig. 2b) was significantly greatest (P [is less than] 0.001) for the Inceptisol-pasture, and did not differ significantly in the Inceptisol-maize, Andisolpasture, and Andisol-maize.

There is no evidence for a difference in net N mineralisation rate with and without leaching for the Andisol-pasture (P = 1) and Andisol-maize (P = 0.844). There is some evidence that net N mineralisation rate in the Inceptisol-pasture was lower with leaching than without leaching (P = 0.015), and for the Inceptisol-maize was greater with leaching than without leaching (P = 0.004); however, the differences associated with the leaching treatment were not large (Fig. 2a, b).

[ILLUSTRATION OMITTED]

The ratio of C mineralised to net N mineralised increased from Week 2 to 7 for all soils, but was highest in the first week (except for Andisol-maize), when N was possibly retained in the microbial biomass (Table 2). The mean values of the ratios were close to 19, except for the Andisol-pasture where the ratio was close to 25, showing that there was less net N mineralisation per unit of C mineralisation for this sample.

The DOC leached with Ca[Cl.sub.2] was expressed as a fraction of total soil C, and followed the order: Inceptisol-maize [is greater than] Inceptisol-pasture = Andisol-maize [is greater than] Andisol-pasture (Fig. 3). The proportions of C leached were 1 or 2 orders of magnitude lower than the C mineralised to [CO.sub.2].

[ILLUSTRATION OMITTED]

During incubation the pH values generally decreased more for the non-leached samples (about 0.4 pH units) than the leached samples (about 0.1 pH unit) ([F.sub.1,16] = 28.7, P [is less than] 0.001). There was some evidence of a difference between soil samples ([F.sub.3,16] = 4.6, P = 0.016), but no evidence for an interaction between these factors ([F.sub.3,16] = 0.1, P = 0.9). Pairwise comparisons of soils showed that the pH change for the Andisol-pasture was greater than for the Andisol-maize (P = 0.036), and Andisol-maize less than Inceptisol-pasture (P = 0.030); the other differences were not significant.

Discussion

The C mineralisation rate, expressed on a C basis, was greater in the Inceptisol than in the Andisol (Fig. 1). This is consistent with the results of Ross et al. (1982) who also showed that C mineralisation rate (per unit of C) was lower in an Andisol than in a gley soil with halloysite as the predominant clay mineral. An explanation may be that Andisols contain high concentrations of allophane (Table 1) that sorbs substrate, enzymes, and microbial biomass, thus reducing C mineralisation rate and increasing the turnover time of soil C (Aomine and Kobayashi 1964; Saggar et al. 1994, 1996, 1999). Percival et al. (2000) suggested that free Al ions were also a major factor in stabilising total soil organic matter in New Zealand soils. Pyrophosphate-extractable Al was higher in the Andisol than in the Inceptisol (Table 1) and so is consistent with this hypothesis. Thomsen et al. (1999) have suggested an alternative hypothesis, namely, that clay content influences the decomposability of organic matter through the effect of texture on porosity and soil moisture. This hypothesis has merit in the field situation and requires further testing. Our data, however, were obtained at a constant water potential and demonstrate the effects of clay mineral on C and N mineralisation rate.

The greater C mineralisation rate on a C basis for pasture than for maize soils (Fig. 1a) probably resulted from differences in the composition of the organic matter. Analysis of unpublished [sup.13]C-NMR showed that the soil organic matter under pasture in New Zealand contained a greater proportion of O-alkyl groups (such as those that occur in simple sugars, hemicellulose, and cellulose) compared with that under maize. Some of these less crystalline O-alkyl compounds are readily available to the microbial biomass, thus explaining the greater proportion of C mineralised in pasture soils.

Leaching enhanced C mineralisation rate in the Inceptisol samples but had no effect in the Andisol samples (Fig. 1). It is possible that substrates that are sorbed by allophane were unaffected by leaching, whereas the more available substrates in the Inceptisol, which has halloysite mineralogy, were mobilised with the Ca[Cl.sub.2] solution, and made more available. This is consistent with the DOC leaching data (Fig. 3), which showed that the greatest proportion of C was leached from the Inceptisol-maize samples, and the least from the Andisol-pasture samples.

The pH of the non-leached samples decreased more than that of the leached samples during the incubation, probably as a result of leaching of the protons produced during nitrification. This also demonstrates the effectiveness of 0.002 M Ca[Cl.sub.2] in removing solute produced during incubation.

Leaching, however, had a negligible effect on net mineral-N production in the Andisols during the incubation but had a variable small effect in the Inceptisols (Fig. 2). This suggests that leaching with 0.002 M Ca[Cl.sub.2] was as effective in extracting mineral-N during a laboratory incubation as was shaking with 0.5 M [K.sub.2][SO.sub.4].

The net mineralisation rate of N in the leached samples was greater for the Inceptisols than for the Andisols (Fig. 2a), and was closely related to the C mineralisation rate (Fig. 1a). Possibly the N substrates were also more strongly sorbed by allophane than by halloysite, making them less available to the microbial biomass. This is consistent with the results of Ross et al. (1982) who showed that net N mineralisation rate (per unit of N) was lower in an Andisol than in a halloysitic gley soil.

This study suggests that, for some soils, C mineralisation rate during laboratory incubations may be underestimated unless solutes are leached from the samples. It also demonstrates that C and N mineralisation rate differ between Inceptisols and Andisols. Sorption of substrates may be important in soils that contain allophane and iron oxides (Sollins et al. 1996), and further work is required to separate the effects of sorption, of protection of substrates in small pores, and water availability on rates of C and N mineralisation.

Acknowledgments

We are grateful to N. Fisher of Cambridge for generous assistance and access to his farm, to R. Webster for statistical analysis, and to D. J. Ross and N. A. Scott for comments on the manuscript. The work was supported by the Foundation for Research, Science and Technology under contract CO9811.

References

Aomine S, Kobayashi Y (1964) Effects of allophane on the enzyme activity of a protease. Soil and Plant Nutrition 10, 28-32.

Blakemore LC, Searle PL, Daly BK (1987) Methods for chemical analysis of soils. Department of Scientific and Industrial Research, New Zealand Soil Bureau Scientific Report No. 80. Lower Hutt, NZ.

Doran JW, Coleman DC, Bezdicek DF, Stewart BA (1994) Defining soil quality for a sustainable environment. Soil Science Society of America, Special Publication No. 35. Madison, WI, USA.

Hewitt AE (1998) `New Zealand soil classification.' 2nd Edn (Manaaki Whenua-Landcare Research: Lincoln, NZ)

Parfitt RL, Wilson AD (1985) Estimation of allophane and halloysite in three sequences of volcanic soils, New Zealand. Catena Supplement 7, 1-8.

Percival H J, Parfitt RL, Scott NA (2000) Factors controlling soil carbon in New Zealand grasslands: is clay content important? Soil Science Society, of America Journal 64, 1623-1630.

Ross DJ, Tate KR, Cairns A (1982) Biochemical changes in a yellow-brown loam and a central gley soil converted from pasture to maize in the Waikato area. New Zealand Journal of Agricultural Research 25, 35-42.

Saggar S, Tate KR, Feltham CW, Childs CW, Parshotam A (1994) Carbon turn-over in a range of allophanic soils amended with [sup.14]C-labelled glucose. Soil Biology and Biochemistry 26, 1263-1271.

Saggar S, Parshotam A, Sparling GP, Feltham CW, Hart PBS (1996) [sup.14]C-Labelled ryegrass turnover and residence times in soils varying in clay content and mineralogy. Soil Biology and Biochemistry 28, 1677-1686.

Saggar S, Parshotam A, Hedley C, Salt G (1999) [sup.14]C-labelled glucose turnover in New Zealand soils. Soil Biology and Biochemistry 31, 2025-2037.

Sollins P, Howmann P, Caldwell BA (1966) Stabilization and destabilization of soil organic matter: mechanisms and controls. Geoderma 74, 65-105.

Stanford G, Smith SJ (1972) Nitrogen mineralisation potentials of soils. Soil Science Society of America Journal 36, 465-472.

Thomsen IK, Schjonning P, Jensen B, Kristensen K, Christensen BT (1999) Turnover of organic matter in differently textured soils. II Microbial activity as influenced by soil water regimes. Geoderma 89, 199-218.

Manuscript received 17 July 2000, accepted 6 October 2000

R. L. Parfitt, G. J. Salt, and S. Saggar

Landcare Research, Private Bag 11052, Palmerson North, New Zealand. Corresponding author; email: parfittr@landcare.cri.nz
Table 1. Organic C, total N, and P concentrations (g/kg), sand, silt,
clay, and clay mineral contents in soil samples (g/kg) (0-10 cm),
Olsen P (mg/kg), and Al extracted in pyrophosphate reagent ([A.sub.py)
(g/kg) in soil samples (0-10 cm)
Standard errors are in parentheses;  n = 3

           Organic C     Total N       Total P

                        Inceptisol

Maize      23.8 (2.1)    2.1 (0.1)     1.2 (0.1)
Pasture    50.7 (1.7)    4.6 (0.2)     1.1 (0.1)

                        Andisol

Maize      56.3 (1.0)    4.9 (0.1)     2.8 (0.1)
Pasture    92.1 (5.0)    8.5 (0.5)     2.2 (0.1)

           C : N         C : P         Olsen P

                          Inceptisol

Maize      11.6 (0.1)    19.9 (0.6)    85 (10)
Pasture    11.1 (0.1)    44.7 (1.8)    26 (2)

                            Andisol

Maize      11.5 (0.1)    20.3 (0.3)    43 (3)
Pasture    10.8 (0.1)    42.7 (0.7)    11 (1)

           Sand    Silt    Clay    Allophane    Halloysite

                           Inceptisol

Maize       470     290     240       20           200
Pasture     420     330     250       20           220

                             Andisol

Maize       560     330     110       100           0
Pasture     360     520     120       110           0

           Gibbsite    [Al.sub.py]

              Inceptisol

Maize       5             1.4
Pasture     6             4.0

                Andisol

Maize       0             5.5
Pasture     tr            7.9


Table 2. Ratio of C mineralised to N mineralised for leached soil
samples

Standard errors are in parentheses; n = 3

  Week       Incept.-maize    Incept.-pasture

    1        22.2 (3.2)       27.7 (2.3)
    2        15.8 (1.7)       15.2 (2.0)
    3        18.3 (1.7)       17.5 (1.9)
    4        19.6 (1.4)       20.3 (2.0)
    5        20.0 (1.2)       21.5 (2.3)
    6        20.5 (1.3)       21.1 (1.9)
    7        19.8 (1.1)       21.9 (1.8)
  Mean
Weeks 2-7        19.0              19.6

  Week       Andisol-maize    Andisol-pasture

    1        14.6 (1.5)       36.2 (1.8)
    2        14.4 (0.5)       19.9 (1.7)
    3        17.7 (0.5)       23.9 (1.8)
    4        19.6 (1.1)       26.6 (1.7)
    5        19.8 (1.2)       26.5 (1.2)
    6        20.0 (1.5)       26.4 (1.1)
    7        20.2 (1.5)       25.8 (0.9)
  Mean
Weeks 2-7        18.6              24.8
Gale Copyright: Copyright 2001 Gale, Cengage Learning. All rights reserved.