Spatio-temporal variation and eect of urea fertilization on
methanotrophs in a tropical dryland rice ®eld
S.K. Dubey, J.S. Singh*
Ecosystems Analysis Laboratory, Department of Botany, Banaras Hindu University, Varanasi 221 005, India
Accepted 26 September 1999
Abstract
Population size of methanotrophs in a dryland ®eld planted to Oryza sativa L. variety Narendra-118 was quanti®ed over a period of 13 weeks. Methanotroph numbers were higher in control plots (52.9±736.6105 cells gÿ1 dry soil) than in plots treated with urea (43.8±676.0105cells gÿ1dry soil), and were highest in the rhizosphere soil (499.8±736.6105cells gÿ1 dry soil) followed by bulk (451.4±684.1105cells gÿ1dry soil) and bare (43.8±67.5105cells gÿ1dry soil) soil. The concentrations
of NH4+-N were signi®cantly (P< 0.001) lower in the rhizosphere (3.1±6.4mg gÿ1soil) than in bulk (4.1±8.3 mg gÿ1 soil) and
bare soils (5.1±10.7mg gÿ1soil). The study suggests that the development of the rice rhizosphere brings about a spatial pattern in the distribution of methanotrophic bacteria which increases in size, over time, within the rhizosphere and adjoining bulk soil, and that the rhizosphere is a potential microsite of intense CH4 oxidation activity. 7 2000 Elsevier Science Ltd. All rights
reserved.
Keywords:Dryland rice; Methanotroph population; Rhizosphere; Urea fertilization
1. Introduction
A major anthropogenic source of atmospheric methane, a greenhouse gas, is wetland rice agriculture. The atmospheric concentration of CH4 is expected to
increase further due to expansion of rice cultivation (Singh and Singh, 1995). Although the chemical reac-tion of CH4 with hydroxyl radicals in the atmosphere
is considered to be the major sink (Crutzen, 1991), uptake of ambient methane by soils could be an ad-ditional sink, representing 1±15% of that oxidized by reaction with hydroxyl radicals (Born et al., 1990). Methane oxidation by soils in temperate ecosystems, tropical forest, desert, savanna and dryland rice ®elds has been reported (for review see King, 1992; Dubey et al., 1996; Topp and Pattey, 1997). Methane oxidiz-ing bacteria (gram-negative, aerobic bacteria belongoxidiz-ing
to the family Methylococcaceae) are ubiquitous in nature and are believed to reduce CH4¯ux to the
at-mosphere from sediments and soils (Oremland and Culbertson, 1992).
A recent investigation on the dynamics of CH4 ¯ux
demonstrated that tropical dryland rice ®elds are po-tential net CH4 sinks, and urea application reduced
their capacity for CH4 oxidation (Singh et al., 1999).
Singh et al. (1998a) reported that the rice plant was a part of this sink. Our goal was to investigate whether the methane oxidizing bacteria (MOB) in the rhizo-sphere are responsible for the CH4 sink strength of
dryland rice and whether population changes are re-sponsible for the urea eect on the sink strength. In the present investigation, we have measured the popu-lation size of methanotrophs from the rhizosphere, and bulk soil in a dryland rice ®eld planted to the rice var-iety Narendra-118, which exhibited highest CH4
oxi-dation activity in the earlier investigation (Singh et al., 1999). The population of methanotrophs in bare soil was also estimated.
0038-0717/00/$ - see front matter72000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 3 8 - 0 7 1 7 ( 9 9 ) 0 0 1 8 1 - 9
www.elsevier.com/locate/soilbio
* Corresponding author. Tel.: +91-542-312-989; fax: +91-542-317-074.
2. Materials and methods
2.1. Experimental site
Our study was carried out on the dryland rice ®eld of the experimental Botanical Garden, Banaras Hindu University, at Varanasi, India (25818'N 8383'E, 129 m above mean sea level). The climate is dry tropical and monsoonal, marked by a cold winter (November±Feb-ruary), a hot summer (March±June) and a warm rainy season (July±September). During the experiment, mini-mum temperatures ranged from 14 to 278C and the maximum from 22 to 388C. The annual rainfall was 1208 mm of which 970 mm fell during the rainy mon-soon from the southwest. The soil is a well drained Inceptisol, pale brown, silty loam (sand 32%, silt 65% and clay 3%) with pH 7±7.8.
2.2. Experimental design and rice cultivation
The experimental ®eld consisted of 12 plots each measuring 53 m. The experiment was laid down in a completely randomized block design. A 0.5-m strip separated plots. Basal treatment of KCl+P2O5
+farm-yard manure was applied at a rate of 60:60:1000 kg haÿ1, to all plots during plowing. Six plots were ferti-lized with urea and the remaining served as control. In the fertilized plots, urea was applied in three split doses, at the time of tillering, ¯owering and grain ®ll-ing stage at the rates of 40, 30 and 30 kg N haÿ1, re-spectively. Among the 12 plots, six plots (three with and three without urea) were sown to rice while the other six (three with and three without urea) were maintained as bare soil. Thus the experiment had three plots each for bare control, bare fertilized, vegetated control and vegetated fertilized treatments. Seeds of rice (Oryza sativa L., cultivar Narendra-118) were sown by dibbling on 10 July 1997, at a spacing of 15 cm (hill-to-hill) by 20 cm (row-to-row) in the plots designated as vegetated plots. No irrigation was pro-vided throughout the experiment and the sole source of water was rainfall.
2.3. Soil sampling and analysis for NH4+-N
Samples of bulk (between the plant rows) and bare (bare plots) soil were collected, separately for each plot, from 0±10 cm depth using a 5-cm dia soil corer. The 0±10 cm soil depth was chosen because obser-vation indicated thatr92% roots are concentrated in this soil layer. The rhizospheric soil was collected by tapping the roots on a plastic sheet (Lee et al., 1997). The soil samples were sieved (2 mm), and ®ne roots were removed. One part of each sample was weighed and oven dried at 1058C to determine the moisture content. Field moist samples, stored at 48C, were used
for chemical analyses and methanotrophic population counts within 2 days after sampling. The soil sampling was carried at 40 and 90 days after sowing (DAS). Ammonium nitrogen (NH4+-N) was measured by the
phenate method (APHA, 1985) in an extract with 2 M KCl.
2.4. Plant biomass
On each sampling date, one rice hill was harvested from each experimental plot with soil as a block (15
2015 cm depth) using a rectangular open-top plastic chamber. Roots were washed with water. After count-ing the tillers, root and shoot materials were dried at 608C to constant weight.
2.5. Population of methanotrophs
The numbers of methanotrophic bacteria were enumerated by the MPN (most probable number) technique as described in Bender and Conrad (1992). In brief, fresh soil (5 g) was suspended in 15 ml of a modi®ed ammonium mineral salts medium (AMS) and shaken for 12 h at 48C in the dark. This suspension served as the inoculum. The modi®ed AMS medium (Heyer et al., 1984) contained lÿ1 distilled water (pH 6.9): 10 mmol NH4Cl, 0.4mmol MgSO47H2O, 4 mmol
K2HPO4, 0.1 mmol CaCl2 and 1 ml trace element
sol-ution (Widdel and Bak, 1992). Instead of microtiter plates, culture tubes were used (Espiritu et al., 1997).
Dilution was carried out from 10ÿ1 to 10ÿ9, as described by Espiritu et al. (1997). Each dilution, 1 ml was inoculated into tubes containing 3 ml AMS med-ium. There were six replicates for each dilution. After inoculation under aseptic conditions, the tops of the tubes were closed with sterilized cotton plugs. The tubes were incubated under 20% methane in air at 258C in the dark in atmosbags (Sigma, USA) for 3 weeks. For control, culture tubes were prepared with-out soil inoculum (Espiritu et al., 1997). In tests we had used control with sterilized soil and found that control without soil was as good as a control with ster-ilized soil. A more appropriate control would be cul-ture tubes with soil inoculum under CH4-free air.
Further, perhaps a more reliable method to enumerate cultivable MOB would be MPN in tubes (6±8 weeks incubation) with measurement of CH4 consumption
(Escoer et al., 1997).
3. Results and discussion
3.1. Soil moisture and NH4 +
-N
Soil moisture ranged from 70 to 187 mg gÿ1 soil. ANOVA indicated signi®cant dierences in soil moist-ure due to days F1, 2654:69,P0:000 but dier-ences due to treatment and soil origin (rhizosphere, bulk and bare) were not signi®cant. Nevertheless, moisture content was consistently higher for bare soil than for bulk or rhizosphere soil. The ammonium-N concentration in the soil was higher for the fertilized plots compared to control plots F1, 2653:61,P 0:000 and was higher on 40 DAS compared to 90 DAS F1, 265:56,P0:026 (Table 1). The NH4
+
-N concentration declined from bare to bulk or bare to rhizosphere soil, both in control and fertilized soil, with the dierences between soil positions being signi®-cant F2, 2613:64,P0:000). The range of NH4
+
-N concentration observed in the present investigation for bulk soil was similar to that reported earlier for the dryland rice ®elds in this region (Singh and Singh, 1994; Jha et al., 1996). Evidently, the low NH4+-N in
the rhizosphere resulted from the continuous uptake by rice plants.
3.2. Plant biomass
In the present study we measured number of tillers, root biomass and shoot biomass as aected by urea fertilization and growth stage (Table 2). There was a signi®cant eect of urea treatment on these growth
characteristics (P ranging from 0.000 to 0.002). The values observed for plant growth parameters were within the range reported earlier for the same rice var-iety (Singh et al., 1999).
3.3. Population of methanotrophs
The size of the methanotroph community ranged from 52.9 to 736.6105cells gÿ1in control and from 43.8 to 676.0105cells gÿ1in fertilized soils (Table 3). There is a lack of information on the population size of methanotrophs in dryland rice ®elds. Joulian et al. (1997) sampled 22 rice paddies at the end of the crop cycle from ®ve dierent countries and found that the methanotroph population ranged from 1.5103to 3.5
107 cells gÿ1. The overall range in the population size observed in this dryland rice ®eld is thus higher than the range reported for paddy soils. The present range of population size is also higher than the range (2.4105±3.6106cells gÿ1) reported for a variety of soils from cultivated and natural ecosystems (Bender and Conrad, 1992).
Our results indicated that the population size decreased signi®cantly in plots amended with urea
F1, 2640:33,P0:000). Bender and Conrad (1994) demonstrated that population size of methanotrophs was lower in rice paddy soils incubated with NH4+(1.1 105±2.9105 counts gÿ1 dw) than in control soil (2.2 105±8.0 105 counts gÿ1 dw), indicating that NH4
+
inhibits the soil methanotrophs. A decline in methanotroph population in fertilized plots is in accord with the observation that CH4 consumption
was lower in fertilized than in unfertilized dryland rice plots (Singh et al., 1998a; 1999). The inhibitory eect of NH4+ on methane oxidation has also been
demon-strated by Bronson and Mosier (1994), King and Schnell (1994), Hutsch et al. (1996) and Gulledge et al. (1997).
There was an inverse relationship between the size of methanotroph population and NH4+-N
concen-tration in soil (Table 4). However, according to this re-lationship, variability in NH4+-N concentration
accounted for only 38% of the variability in the popu-lation size of methanotrophs, indicating involvement Table 1
Soil moisture (mg gÿ1soil) and NH4+-N concentration (mg gÿ1soil) in rhizosphere, bulk and bare soils from control and fertilized (100 kg N haÿ1) planted to rice variety Narendra-118, determined on two sampling dates. (DAS=days after sowing) (mean21 S.E.)
Variables DAS Rhizosphere Bulk Bare
control fertilized control fertilized control fertilized
NH4+-N 40 3.320.7 6.420.9 4.720.6 8.320.6 5.820.8 10.720.7
90 3.120.5 4.420.6 4.120.8 6.721.0 5.120.8 8.720.9
Moisture 40 163.326.6 150.0211.5 176.728.8 161.726.0 186.723.3 176.728.8
90 81.727.3 70.025.8 90.726.4 80.025.8 116.727.3 93.326.7
Table 2
Plant growth variables of the dryland rice variety Narendra-118 as aected by urea application (100 kg N haÿ1) determined at two dierent growth stages (mean21 S.E.)
Growth variables 40 Days after sowing 90 Days after sowing
control fertilized control fertilized
of additional factors. Plant biomass was much higher in the urea fertilized treatment which can lead to higher carbon input and thus higher oxygen consump-tion by heterotrophs. Although there is no empirical evidence from this study, oxygen could well be a limit-ing factor, particularly durlimit-ing periods of high soil moisture.
Our data indicate that the population size of metha-notrophs was signi®cantly higher F1, 264:89,P 0:036on 90 DAS as compared to 40 DAS in both the control and fertilized soils. Such a dierence in popu-lation size may be related to soil moisture since it was higher on 40 DAS 8169:2mg g
ÿ1soil than on 90
DAS 888:7mg g
ÿ1soilwhich, in turn, aected the
amount of aeration. Negative exponential relationships between CH4 oxidation rates and soil moisture have
been reported for natural ecosystems (Singh et al., 1997, 1998b) as well as for dryland rice agriculture (Singh et al., 1999). The dierences in the rhizospheric methanotroph population size between 40 DAS and 90 DAS could be related to plant development, more speci®cally to aerenchyma development with age. Fren-zel and Gilbert (1998) also observed an increase in cell numbers of methanotrophs with time in the rice rhizo-sphere.
In this study there were signi®cant dierences
F2, 26550:63,P0:000 in methanotroph popu-lation size among the rhizosphere, bulk and bare soils (Table 3). The lowest population size occurred in bare soil (43.8±67.5105cells gÿ1) and the highest in rhizo-spheric soil (499.8±736.6105cells gÿ1). The bulk soil had an intermediate population size (451.4±684.1105 cells gÿ1). Dierences in methanotroph population size between bulk and rhizosphere soils were also reported
by Espiritu et al. (1997) for a wetland rice ®eld. Rice plant roots had a stimulatory eect on total soil bac-teria including methanotrophs, and the number of methanotrophs was one order of magnitude higher in the rhizosphere than in bulk soil (Gilbert and Frenzel, 1995). The association of methanotrophs with roots (5
104±8105CFU gÿ1dry matter) and stem bases (7
103±2 105 CFU gÿ1 dry matter) in two wetland rice varieties was observed by Watanabe et al. (1997), who concluded that methane oxidizing activity was mainly associated with roots. Evidently the increased number of methanotrophs in the vicinity of rice plant roots is most probably the result of a higher O2partial
pressure due to aeration of the rhizosphere. In an ele-gant microcosm study simulating ¯ooded rice ®eld, Frenzel and Gilbert (1998) found that active CH4
oxi-dizing bacteria (MOB) occurred near to the root mat, and that in the bulk soil no O2 was detected below 2
mm depth. In the present dryland rice ®eld the MOB population was much higher in the bulk soil compared to the bare soil, indicating that the bulk soil was not entirely free from the in¯uence of roots.
The soil of dryland rice ®eld also gets periodically saturated due to heavy rainfall events when CH4
emis-sion instead of net consumption occurs (Singh et al., 1998a, 1999). Under such a situation, the O2-supplying
potential of plant roots is a major factor for the multi-plication, growth and sustenance of methanotrophic bacteria in the rhizosphere. The aerenchymatous tissue of rice plant serves as a conduit to transport CH4
from the anoxic soils to the atmosphere (Mariko et al., 1991) and oxygen from the atmosphere to the rhizo-sphere (Frenzel et al., 1992). The supply of both CH4
and oxygen would thus be more favorable for the Table 3
Distribution of methanotroph population size (105cells gÿ1soil) in rhizosphere, bulk and bare soils from control and fertilized (100 kg N haÿ1) plots on dryland rice ®eld planted to the variety Narendra-118, determined on two sampling dates (DAS=days after sowing) (mean21 S.E.)
DAS Rhizosphere Bulk Bare
control fertilized control fertilized control fertilized
40 691.025.6 499.828.6 579.526.2 451.425.1 52.921.2 43.822.1
90 736.6219.3 676.025.5 684.124.8 537.323.3 67.522.1 52.121.9
Table 4
Relationship between population size of methanotrophs (Y, 105cells gÿ1soil) and NH4+-N concentration (X,mg gÿ1soil) in a dryland rice ®eld, according toYabX
DAS Regression parameters Signi®cant statistics
a b r p
40 n18) 771.8 ÿ59.1 ÿ0.616 0.006
90 n18) 891.8 ÿ80.8 ÿ0.614 0.007
methanotroph population to develop in rhizosphere than in the bulk or bare soil. The view that supply of both CH4and O2 is essential for methanotroph
popu-lation is supported by the ®ndings that the popupopu-lation size in paddy soils exposed to air enriched with 20% methane, increased to 2.3107cells gÿ1in comparison to control soil (4.2106 cells gÿ1) (Bender and Con-rad, 1992). Singh et al. (1998a, 1999) found that plant variables, specially the number of tillers, root volume and root porosity, representing the conduit and venti-lation eects were important for CH4oxidation in
dry-land rice agriculture.
In conclusion, the development of the rhizosphere brings about a spatial pattern in the distribution of methanotrophic population which increases in size during the vegetative period, and within the rhizo-sphere and adjoining bulk soil as compared to the bare soil. Greater O2 availability due to ventilation by
rice plants, lower concentrations of NH4+-N due to
continuous plant uptake and a larger methanotroph population make the rice rhizosphere a microsite for intense CH4 oxidation activity. We thus demonstrate
that plant, plant age and fertilization aect MOB in dryland rice ®eld.
Acknowledgements
We thank Professor A.K. Kashyap for advice and methanotrophic population counting. Funding from Department of Biotechnology, Government of India, New Delhi for the research work is gratefully acknowl-edged.
References
APHA (American Public Health Association), 1985. Standard methods for the examination of water and wastewater. American Public Health Association, Washington.
Bender, M., Conrad, R., 1992. Kinetics of CH4 oxidation in oxic soils exposed to ambient air or high CH4 mixing ratios. Federation of European Microbiological Societies, Microbiology Ecology 101, 261±270.
Bender, M., Conrad, R., 1994. Microbial oxidation of methane, am-monium and carbon monoxide, and turnover of nitrous oxide and nitric oxide in soils. Biogeochemistry 27, 97±112.
Born, M., Dorr, H., Levin, J., 1990. Methane consumption in aera-ted soils of the temperate zone. Tellus B 42, 2±8.
Bronson, K.F., Mosier, A.R., 1994. Suppression of methane oxi-dation in aerobic soil by nitrogen fertilizers, nitri®cation inhibi-tors and urease inhibiinhibi-tors. Biology and Fertility of Soils 17, 263± 268.
Crutzen, P.J., 1991. Methane sources and sinks. Nature 350, 380± 381.
Dubey, S.K., Kashyap, A.K., Singh, J.S., 1996. Methanotrophic bac-teria, methanotrophy and methane oxidation in soil and rhizo-sphere. Tropical Ecology 37, 167±182.
Escoer, S., Le Mer, J., Roger, P.A., 1997. Enumeration of
metha-notrophic bacteria in rice®eld soils by plating and MPN tech-niques: a critical approach. European Journal of Soil Biology 33, 41±51.
Espiritu, B.M., Adachi, K., Senboku, T., 1997. Eect of application of rice straw and cellulose on methane emission and biological N2 ®xation in a subtropical paddy ®eld. III. Populations of methane oxidizing bacteria in soil and rice rhizosphere. Soil Science and Plant Nutrition 43, 729±734.
Frenzel, P., Gilbert, B., 1998. Rice roots and CH4oxidation: the ac-tivity of bacteria, their distribution and the microenvironment. Soil Biology and Biochemistry 30, 1903±1916.
Frenzel, P., Rothfuss, F., Conrad, R., 1992. Oxygen pro®les and methane turnover in a ¯ooded rice microcosm. Biology and Fertility of Soils 14, 84±89.
Gilbert, B., Frenzel, P., 1995. Methanotrophic bacteria in the rhizo-sphere of rice microcosms and their eect on pore-water methane concentration and methane emission. Biology and Fertility of Soils 20, 93±100.
Gulledge, J., Doyle, A.P., Schimel, J.P., 1997. Dierent NH4+ inhi-bition patterns of soil CH4consumption: a result of distinct CH4 oxidizer populations across sites? Soil Biology and Biochemistry 29, 13±21.
Heyer, J., Malashenko, Y., Berger, U., Budkova, E., 1984. Verbreitung methanotropher Bakterien. Zeitschrift fuer Allgemeine Mikrobiologie 24, 725±744.
Hutsch, B.W., Russell, P., Mengel, K., 1996. CH4oxidation in two temperate arable soil as aected by nitrate and ammonium appli-cation. Biology and Fertility of Soils 23, 86±92.
Jha, P.B., Kashyap, A.K., Singh, J.S., 1996. Eect of fertilizer and organic matter inputs on nitri®er populations and N-mineraliz-ation rates in dry tropical region, India. Applied Soil Ecology 4, 231±241.
Joulian, C., Escoer, S., LeMer, J., Neue, H.-U., Roger, P.A., 1997. Populations and potential activities of methanogens and metha-notrophs in rice ®elds: relations with soil properties. European Journal of Soil Biology 33, 105±116.
King, G.M., 1992. Ecological aspect of methane oxidation, a key determinant of global methane dynamics. In: Marshall, K.C. (Ed.), Advances in Microbial Ecology. Plenum, New York, pp. 431±468.
King, G.M., Schnell, S., 1994. Eect of increasing atmospheric methane concentration on ammonium inhibition of soil methane consumption. Nature 370, 282±284.
Lee, K.K., Wani, S.P., Sahrawat, K.L., Trimurtulu, N., Ito, O., 1997. Nitrogen and/or phosphorus fertilization eects on organic carbon and mineral contents in the rhizosphere of ®eld grown Sorghum. Soil Science and Plant Nutrition 43, 117±126.
Mariko, S., Harazono, Y., Owa, N., Nouchi, I., 1991. Methane in ¯ooded soil water and the emission through rice plants to the at-mosphere. Environmental and Experimental Botany 31, 343±350. Oremland, R.S., Culbertson, C.W., 1992. Importance of methane
oxidizing bacteria in the methane budget as revealed by the use of a speci®c inhibitor. Nature 356, 421±423.
Singh, H., Singh, K.P., 1994. Nitrogen and phosphorus availability and mineralization in dryland reduced tillage cultivation: eects of residue placement and chemical fertilizer. Soil Biology and Biochemistry 26, 695±702.
Singh, J.S., Singh, S., 1995. Methanogenic bacteria, methanogenesis and methane emission from rice paddies. Tropical Ecology 36, 145±165.
Singh, J.S., Singh, S., Raghubanshi, A.S., Singh, S., Kashyap, A.K., Reddy, V.S., 1997. Eect of soil nitrogen, carbon and moisture on methane uptake by dry tropical forest soils. Plant and Soil 196, 115±121.
Singh, J.S., Raghubanshi, A.S., Reddy, V.S., Singh, S., 1998b. Methane ¯ux from irrigated paddy and dryland rice ®elds, and from seasonally dry tropical forest and savanna soils of India. Soil Biology and Biochemistry 30, 135±139.
Singh, S., Singh, J.S., Kashyap, A.K., 1999. Methane consumption by soils of dryland rice agriculture: in¯uence of varieties and N-fertilization. Chemosphere 38, 175±189.
Snedecor, G.W., Cochran, W.G., 1989. Statistical Methods. Iowa State University Press, Ames.
Topp, E., Pattey, E., 1997. Soils as sources and sinks for atmos-pheric methane. Canadian Journal of Soil Science 77, 167±178. Watanabe, I., Hashimoto, T., Shimoyama, A., 1997.
Methane-oxidiz-ing activities and methanotrophic populations associated with wetland rice plants. Biology and Fertility of Soils 24, 261±265. Widdel, F., Bak, F., 1992. Gram-negative mesophilic
