Plant Soil (2009) 316:241–255 DOI 10.1007/s11104-008-9776-0
REGULAR ARTICLE
Quantification of proton budgets in soils of cropland and adjacent forest in Thailand and Indonesia Kazumichi Fujii & Shinya Funakawa & Chie Hayakawa & Sukartiningsih & Takashi Kosaki
Received: 30 April 2008 / Accepted: 3 September 2008 / Published online: 26 September 2008 # Springer Science + Business Media B.V. 2008
Abstract Continuous cultivation has the potential to accelerate soil acidification. The influence of cultivation on soil acidification was evaluated by calculating proton budget in a soil–vegetation system including solute leaching, vegetation uptake and organic matter decomposition in cropland and adjacent forest in Thailand and Indonesia. In the forests, excess cation accumulation in wood (2.1–3.8 kmolc ha−1 year−1) has contributed to soil acidification at the rate of 0.004 molc for production of 1 mol carbon. In the croplands, soil organic matter loss (2.2–3.9 Mg C ha−1 year−1) has contributed to both proton generation owing to nitrification (1.5–5.0 kmolc ha−1 year−1) and proton consumption owing to mineralization of organic anions (3.6–8.8 kmolc ha−1 year−1) at the rates of 0.008–0.015 and 0.019–0.026 molc for the loss of 1
Responsible Editor: N. Jim Barrow. K. Fujii (*) : S. Funakawa : C. Hayakawa Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan e-mail:
[email protected] Sukartiningsih Department of Forestry, Mulawarman University, Samarinda 75123, Indonesia T. Kosaki Graduate School of Global Environmental Studies, Kyoto University, Kyoto 606-8501, Japan
mol soil organic carbon, respectively. Although the influence of cultivation on proton budget is different depending on the budget of organic matter and soil types (soil pH and texture), cultivation results in soil organic matter loss and soil alkalinization at least during the initial stage of cultivation in tropical regions. Keywords Continuous cultivation . Dissolved organic carbon . Organic matter decomposition . Proton budget . Soil acidification . Soil organic matter
Introduction In the humid tropics, shifting cultivation (slash-andburn agriculture) is an extensive farming system on highly weathered and leached upland soils (Nye and Greenland 1960; Kyuma and Pairintra 1983). Owing to population growth, traditional shifting cultivation with an adequately long fallow period has been replaced with more intensive cropping systems with shorter fallow periods or continuous cultivation (Kyuma and Pairintra 1983). Continuous cultivation has the potential to result in unsustainable land use owing to the depletion of soil organic matter (SOM) and accelerating soil acidification and erosion. Soil acidification is a natural process, accelerated by agriculture, in humid regions under a climate where precipitation exceeds evapotranspiration (Helyar and Porter 1989; Juo et al. 1996). In the croplands, proton
242
generation associated with nitrification was reported to accelerate soil acidification owing to the enhanced mineralization of soil organic nitrogen (Tanaka et al. 1997; Funakawa et al. 2006), limited vegetation uptake at the beginning of the cropping season (Poss et al. 1995) and nitrogen fertilization (Bouman et al. 1995). However, soil acidification is also contributed by proton sources other than nitrification, i.e. acidic deposition, dissociation of organic anions and carbonic acid and excess uptake of cations over anions by vegetation. Since most of proton-generating processes are associated with the organic matter cycles, they are influenced by the cultivation-induced changes in organic matter cycles, typically the loss of SOM owing to the increased organic matter decomposition and product removal. However, the effects of cultivation on individual processes are various and thus, their influence on soil acidification is still unclear. Cultivation generally results in an increase in proton generation associated with nitrification and product removal (Bolan et al. 1991), but it also results in decrease proton generation associated with the dissociation of organic anions. In croplands, the concentrations of dissolved organic carbon (DOC), which is the source of organic acids, in surface soil solution are lower than in adjacent forests (Quideau and Bockheim 1996; Möller et al. 2005). Further, in contrast to forests, where excess cation accumulation in organic matter (biomass and humus) contributes to soil acidification (Hallbäcken and Tamm 1986), the loss of SOM has the potential to increase proton consumption owing to the mineralization of organic anions in croplands. To evaluate cultivation-induced soil acidification in comparison with natural soil acidification in forests, proton generation and consumption owing to these processes should be quantified. By application of the theory of proton budget, proton budget associated with the solutes leaching from the system and that associated with vegetation uptake, can be quantified in forests (van Breemen et al. 1983; Fujii et al. 2008). Besides, by quantifying proton consumption associated with the loss of SOM, proton budget can also be quantified for croplands (Poss et al. 1995). The objective of the present study was to evaluate the influence of cultivation on soil acidification by quantifying proton budget in a soil–vegetation system including solute leaching, vegetation uptake and organic matter decomposition in croplands and adjacent forests.
Plant Soil (2009) 316:241–255
Materials and methods Descriptions of experimental plots Experimental plots consisted of one forest and one cropland plot in both Thailand and Indonesia. The forest and cropland plots in Thailand (RPf and RPc, respectively) were located in Ban Rakpaendin, Chiang Rai Province (Fig. 1), where the mean annual air temperature and annual precipitation were recorded as 25.0°C and 2,084 mm year−1, respectively. There were distinct dry and wet seasons: the dry season was from November to March and the wet season with most of the rainfall was from April to October. The vegetation was dominated by Lithocarpus sp. and Eugenia sp. in RPf, while corn (Zea mays L.) had been cultivated during the wet season without fertilization in RPc for 3 years since the conversion of forest to cropland. Soils were derived from sedimentary rocks (RPf) and sedimentary rocks associated with granite intrusion (RPc) and classified as typic Haplustults (Soil Survey Staff 2006). The forest and cropland plots in Indonesia (BSf and BSc, respectively) were located in the Experimental Forest of the Tropical Rainforest Research Center, Mulawarman University, Bukit Soeharto, East Kalimantan Province (Fig. 1), where the mean annual air temperature and annual precipitation were
RP
Thailand
BS
0°
Indonesia
Fig. 1 Locations of the experimental plots
Plant Soil (2009) 316:241–255
recorded as 26.8°C and 1,977 mm year−1, respectively. The vegetation was dominated by Shorea laevis and Dipterocarpus cornutus in BSf, while chili (Capsicum sp.) had been cultivated for 2 years after deforestation in BSc. Soils were derived from sedimentary rocks and classified as typic Paleudults (Soil Survey Staff 2006). 0.71 Mg DW ha−1 year−1 of poultry manure was applied at the beginning of cropping season (October 2004–October 2005). Monitoring temperature and volumetric water content in soils
243
and K contents in plant samples were determined by flame photometry, Mg and Ca contents by atomic adsorption spectroscopy (A-A-640-01, Shimadzu), Fe and Al contents by inductively coupled plasma atomic emission spectrometry (ICP-AES), and P contents using the colorimetric method (UV–vis spectrophotometer UV-1200, Shimadzu) after nitric–sulfuric acid wet digestion. The Cl and S contents were determined by high performance liquid chromatography (HPLC, Shimadzu, ion chromatograph HIC-6A) after combustion according to Busman et al. (1983). Organic matter decomposition
The soil temperature at 5 cm depth was measured with a thermister probe (107 Temperature Probe, Campbell Scientific, Inc.) in two replications, while volumetric water contents in soils at depths of 5, 15 and 45 cm (30 cm in BS) were measured with time domain reflectometer probes (CS615 Water Content Reflectometer, Campbell Scientific, Inc.) in three replications. Data were recorded using dataloggers at 30-min intervals (Campbell Scientific, Inc., CR-10X) during the study. Plant materials In the forest plots, the aboveground biomass in RPf and BSf were estimated by applying the diameters of stems at breast height to the regression equations obtained by Tsutsumi et al. (1983) and Yamakura et al. (1986), respectively. The diameters of stems at breast height were measured by census in 0.25 ha quadrats. Litterfall was collected by circular litter traps of 60 cm diameter in five replications. Wood increment was estimated by stem analysis (tree ring analysis) and using the regression equation for estimating tree biomass according to Johnson and Risser (1974). Wood samples were collected using an increment borer. In the cropland plots, plant biomass (leaves, stems and fruits) was collected at the end of the wet season. Fine roots in the O horizon were collected in 30×30 cm quadrats in five replications, while those in mineral soil were collected at 5 cm intervals by taking 0.1 L cores in five replications. Roots were rinsed in distilled water to remove soil materials. Plant samples were oven-dried at 70°C for 48 h, weighed and milled. The C and N contents in plant materials were determined using an NC analyzer (NC-800-13N, Sumika Chem. Anal. Service). The Na
Soil respiration consists of organic matter decomposition and root respiration. Organic matter decomposition can be quantified by excluding root respiration using the trenching method. Firstly, rates of CO2 emission were measured monthly in five replications using a closedchamber method in the trenched plots, for which collars with a diameter of 15 cm and a height of 40 cm were installed in the soil to a depth of 20 cm. Gases in the headspace of the soil collars were sampled 0, 10, 20 and 30 min after the tops of the collars were covered with plastic sheets. Sampled gases were analyzed with an infrared CO2 controller (ZFP9, Fuji Electric Instruments Co., Ltd.). A detailed description for the procedure was given by Shinjo et al. (2006). Secondly, the annual rates of organic matter decomposition were calculated according to Funakawa et al. (2006), using the measured CO2 emission rates, the monitored soil temperature and the volumetric water content in soil. After a stepwise regression analysis of the soil temperature or moisture dependency of the CO2 emission rates, the CO2 emission rates were simulated for a given period (i.e. 1 year) using the regression equations and monitored data of soil temperature and volumetric water content in soil. The annual rates of organic matter decomposition were then calculated by summing the simulated rates of CO2 emission for one complete year. Precipitation, throughfall and soil solution Soil solutions were collected using a tension-free lysimeter, draining a surface area of 200 cm2 in five replications beneath the A, BA and Bt horizons (7, 20 and 45 cm depths) in RPf and RPc, the O, A and B1 horizons (0, 5 and 30 cm depths) in BSf and the Ap1,
244
Plant Soil (2009) 316:241–255
Ap2 and B1 horizons (5, 20 and 30 cm depths) in BSc. Precipitation and throughfall were separately collected using a precipitation collector in five replications. Sample solutions were filtered through 0.45 μm cellulose acetate membrane filters before each analysis. The concentrations of H+ in solution were determined with a glass electrode. The concentrations of Na+, K+, NH4+, Mg2+, Ca2+, Cl−, NO3− and SO42− in solution were determined by high performance liquid chromatography (HPLC, Shimadzu, ion chromatograph HIC-6A, shim-pack IC-C3 for cations, shim-pack IC-A1 for anions, conductivity detector CDD-6A). The concentrations of total Fe and Al in solution were determined by ICP-AES (SPS1500, Seiko Instruments Inc.). The total charge equivalent of Al ions was calculated as the equivalent sum of Al3+, AlOH2+ and Al(OH)2+. The concentrations of DOC and inorganic carbon (IC) were determined using a total organic carbon analyzer (TOC-VCSH, Shimadzu). The concentrations of HCO3− in solution were determined from the solution pH and IC concentration, based on pKa =6.3, although it should be mentioned that they might be underestimated in soil solutions collected by tension-free lysimeter owing to problems associated with the decreased partial pressure of CO2. The anion deficit, if any, was assigned to the negative charge of organic acids.
where C (m−1) is differential water capacity, t (day) is time, Z (m) is height, h (m) is soil water pressure head, K (m day−1) is unsaturated hydraulic conductivity and S (day−1) is sink term accounting for water uptake by vegetation and lateral water flow. The unsaturated hydraulic conductivity and soil water pressure heads at depths of 5, 15 and 45 cm (5, 15 and 30 cm in BSf and BSc) were estimated using the saturated hydraulic conductivity, water retention curves of soil and volumetric water content monitored at each depth at 30-min intervals (Mualem and Dagan 1978; van Gunuchten 1980). In this process, the Mualem–van Genuchten parameters describing the unsaturated hydraulic conductivity and soil water pressure heads were adjusted to meet water budget according to Klinge (2001). A detailed description for the calculation of water fluxes was presented in our previous report (Fujii et al. 2008). The water fluxes for each month were calculated by summing the halfhourly water fluxes thus estimated. The fluxes of ion and DOC leaching for each month were calculated by multiplying the water fluxes by the concentrations of ion and DOC in precipitation, throughfall and soil solution during each month. The annual fluxes of ion and DOC leaching were each calculated by summation.
Fluxes of ion and DOC leaching
Proton budget associated with the solutes leaching from the system and that associated with vegetation uptake can be quantified successfully in the forests at a steady state (van Breemen et al. 1983, 1984). In croplands, where cultivation results in the loss of SOM, proton consumption owing to mineralization of organic anions associated with the loss of SOM should also be quantified (Helyar and Porter 1989; Poss et al. 1995). As in the case of the deforested land (van Breemen et al. 1984), proton budget in cropland can be evaluated by quantifying proton budget associated with solutes leaching, vegetation uptake and SOM loss. Net proton generation (NPG) resulting from excess cation uptake by vegetation, nitrification, dissociation of organic acids, dissociation of carbonic acid and net proton influx from the overlying horizon were calculated in each soil horizon compartment based on the input–output budget of ions in soil–vegetation systems including solute leaching and vegetation uptake (Bredemeier et al. 1990). Net proton generation
The fluxes of ion and DOC leaching from each horizon were calculated by multiplying the water fluxes by the concentrations of ion and DOC in precipitation, throughfall and soil solution during the period. The water fluxes of throughfall were measured using a precipitation collector, while the halfhourly fluxes of soil water percolating from the surface (5–15 cm depth) and subsurface soil horizons (15–45 and 15–30 cm depths in RP and BS, respectively) were estimated by applying Darcy’s law to the unsaturated hydraulic conductivity and the gradient of the hydraulic heads in the surface and subsurface soil horizons. The one dimensional, vertical flow equation (Richards equation) in the unsaturated soil zone is written as
C ð hÞ
dh d dh ¼ K ð hÞ þ 1 S ð hÞ dt dz dz
ð1Þ
Proton budget
Plant Soil (2009) 316:241–255
245
associated with excess cation uptake by vegetation (NPGBio), which consists of wood increment (forest), product removal (cropland) and litterfall, is calculated as
carbon (SOC) loss (Mg C ha−1 year−1), the SOM to SOC ratio of 1.8, and the cation exchange capacity of SOM (CECSOM) (kmolc kg−1 SOM; Poss et al. 1995). Thus, NPCMin is calculated as
NPGBio ¼ ðCationÞbio ðAnionÞbio
NPCMin ¼ ðCationÞLFþM ðAnionÞLFþM
ð2Þ
where (Cation) and (Anion) represent the equivalent sum of cations and anions, respectively. The ionic species counted in the present study were Na+, K+, Mg2+, Ca2+, Fe3+ and Aln+ for cations and Cl−, H2PO3− and SO42− for anions. The suffix “bio” represents ion fluxes caused by vegetation uptake (kmolc ha−1 year−1). Net proton generation associated with the transformation of nitrogen (NPGNtr) is n þ o NPGNtr ¼ NHþ NH4 out ð3Þ 4 in n o þ NO 3 out NO3 in The suffix “in” represents ion fluxes percolating into the soil compartment by precipitation, throughfall and soil solution from the overlying horizon (e.g. precipitation for the Ap horizon), while the suffix “out” represents ion fluxes leaching out of the soil compartment by soil solution. Net proton generation associated with the dissociation of organic acids (NPGOrg) is NPGOrg ¼ ðOrgn Þout ðOrgn Þin
ð4Þ
where Orgn− represents the negative charge of organic anions. Net proton generation associated with carbonic acid dissociation (NPGCar) is NPGCar ¼ HCO ð5Þ 3 out HCO3 in
Net proton influx from the overlying horizon is ð6Þ ðHþ Þin ðHþ Þout
On the other hand, mineralization of organic anions contributes to proton consumption. Net proton consumption owing to mineralization of organic anions (NPCMin) is caused by decomposition of litterfall, manure and SOM. NPC Min from litter and manure is calculated from the difference between cations and anions of litterfall and manure entering into soil, while NPCMin associated with the loss of SOM is calculated using the rates of soil organic
ð7Þ
þ DSOC 1:8 CECSOM where the suffixes “LF” and “M” represent litterfall and manure, respectively. ΔSOC represents the rate of SOC loss (kg C ha−1 year−1). CECSOM was estimated based on soil pH-CECSOM equation proposed by Helyar and Porter (1989).
Results Physico-chemical properties of soils The physico-chemical properties of soils are presented in Table 1. Soil pH was low throughout the soil profiles in BSf (3.8–4.3) and BSc (4.2–4.3) than in RPf (4.6–5.0) and RPc (5.4–5.5). Clay contents in soils were higher in RPf (70–75%) and RPc (40– 51%) than in BSf (23–31%) and BSc (19–26%), although the differences of clay contents between RPf and RPc might be due to the influence of granite in RPc. Total carbon contents in the A horizons in the cropland plots were lower (27 and 14 g kg−1 in RPc and BSc, respectively) than in the adjacent forest plots (63 and 23 g kg−1 in RPf and BSf, respectively). Carbon stock Carbon stock is presented in Table 2. In the forest plots, organic carbon was stored as the aboveground biomass (169.1 and 292.6 Mg C ha−1 in RPf and BSf, respectively) as well as SOM (55.9 and 21.2 Mg C ha−1 in RPf and BSf, respectively). The higher aboveground biomass and the lower stock of organic matter in the mineral soil in the present study, as compared to temperate forests, were consistent with a previous report (Nakane 1980). In the cropland plots, C stock in the mineral soil was lower than in the adjacent forest plots. C stock in the mineral soil in RPc and BSc (43.5 and 18.9 Mg C ha−1, respectively) was 12.4 and 2.3 Mg C ha−1 less than in RPf and BSf, respectively (Table 2).
246
Plant Soil (2009) 316:241–255
Table 1 Physico-chemical properties of soils Horizon
Depth (cm) pH Total Ca,b Total Na,b Exchangeable basic cationa,c (cmolc kg−1) (g kg
RPf
O A BA Bt RPc Ap BA Bt BSf O A BA(E) B1 Bt BSc Ap1 Ap2 B1 Bt
+1–0 0–7 7–20 20–45+ 0–7 7–20 20–45+ +2–0 0–5 5–25 25–40 40+ 0–5 5–20 20–40 40–60+
5.0 4.9 4.6 5.4 5.5 5.5
−1
)
Ca
Mg
K
Na
CECa,d Particle size distributiona,c (%)
Clay
Silt
Sand
62.6 19.8 8.9 26.8 15.0 10.8
3.8 1.5 1.0 2.0 1.3 1.0
2.98 0.97 0.66 5.66 4.32 3.26
2.23 0.72 0.47 2.03 1.40 0.97
0.54 0.22 0.15 0.23 0.08 0.05
0.02 0.04 0.09 0.03 0.11 0.00
27.6 19.9 20.1 11.5 14.4 14.4
70 73 75 40 55 51
25 23 19 18 13 10
5 4 6 42 32 39
4.0 22.9 3.8 4.2 4.0 3.5 4.3 2.5 4.3 14.3 4.2 4.8 4.2 3.7 4.3 3.2
1.4 0.6 0.5 0.5 1.6 0.5 0.5 0.4
0.60 0.63 0.59 0.60 1.54 1.26 0.85 0.68
0.11 0.13 0.19 0.14 0.12 0.41 0.22 0.16
0.41 0.04 0.04 0.09 0.15 0.05 0.06 0.06
1.11 0.00 0.03 0.14 0.29 0.04 0.10 0.00
8.5 6.2 5.0 5.0 6.8 8.7 10.3 11.3
23 24 27 31 19 28 33 26
25 27 30 35 16 19 21 19
52 49 43 34 64 54 46 54
a
Oven dried basis
b
Total carbon contents were determined using an NC analyzer (Sumika Chemical Analytical Service, SUMIGRAPH NC-800-13N)
c
A batch extraction with ammonium acetate (1 M and pH 7.0), exchangeable basic cations were determined using atomic adsorption for Ca and Mg, and flame photometry for Na and K
d
Residual soil after ammonium acetate extraction was washed with deionized water and ethanol, and the remaining NH4+ was extracted with 10% NaCl and then determined by the Kjeldahl distillation method
e
Clay (
