与小麦水稻有关的英文文献
Plant and Soil195:233–245,1997.233 c1997Kluwer Academic Publishers.Printed in the Netherlands.
Seasonal changes in?uxes of methane and volatile sulfur compounds from rice paddies and their concentrations in soil water
Isamu Nouchi,Tatsuo Hosono1and Kaori Sasaki2
National Institute of Agro-Environmental Sciences,3-1-1Kannondai,Tsukuba,Ibaraki305,Japan.Present addresses:1Department of Protected Cultivation,National Research Institute of Vegetables,Ornamental Plants and Tea,45Minaminakane,Taketoyo,Aichi Prefecture470-23,Japan and2Shikoku National Agricultural Experiment Station,1-3-1Senyu-cho,Zentsuuji city,Kagawa Prefecture765,Japan
Received7March1997.Accepted in revised form24June1997
Key words:dimethyl sul?de,?ux,methane,rice paddy,seasonal change,volatile sulfur compounds
Abstract
Rice paddies emit not only methane but also several volatile sulfur compounds such as dimethyl sul?de(DMS: CH3SCH3).However,little is known about DMS emission from rice paddies.Fluxes of methane and DMS,and the concentrations of methane and several volatile sulfur compounds including hydrogen sul?de(H2S),carbonyl disul?de(CS2),methyl mercaptan(CH3SH)and DMS in soil water and?ood water were measured in four lysimeter rice paddies(2.54m,depth2.0m)once per week throughout the entire cultivation period in1995in Tsukuba, Japan.The addition of exogenous organic matter(rice straw)was also examined for its in?uence on methane or DMS emissions.Methane?uxes greatly differed between treatments in which rice straw had been incorporated into the paddy soil(rice straw plot)and plots without rice straw(mineral fertilizer plot).The annual methane emission from the rice straw plots(37.7g m2)was approximately8times higher than that from the mineral fertilizer plots (4.8g m2).Application of rice straw had little in?uence on DMS?uxes.Signi?cant diurnal and seasonal changes in DMS?uxes were observed.Peak DMS?uxes were found around noon.DMS was emitted from the?ood water in the early growth stage of rice and began to be emitted from rice plants during the middle stage.DMS?uxes increased with the growth of rice plants and the highest?ux,15.1g m2h1,was recorded before heading.DMS in the soil water was negligible during the entire cultivation period.These facts indicate that the DMS emitted from rice paddies is produced by metabolic processes in rice plants.The total amount of DMS emitted from rice paddies over the cultivated period was estimated to be approximately5–6mg m2.CH3SH was emitted only from?ood water during the?rst month after?ooding.
Introduction
Rice is a plant adapted from a tropical region to a warm wet lowland area and has been continuously cul-tivated for several thousand years in various areas in Asia.Since almost two-thirds of the world’s population depend on rice as their primary food source,rice has been and will continue to be the most important cereal for sustaining the rapidly increasing world population. The total global harvested rice area is about147mil-lion hectares(FAO,1994).More than85%of this area is wetland or irrigated rice?elds.Flooded rice?elds, FAX No:+8129838-8199.E-mail:nouchi@niaes.affrc.go.jp however,have negative effects on the atmosphere such as global warming.
Rice paddies emit signi?cant quantities of methane, one of the greenhouse effect gases.Methane is pro-duced by microbial anaerobic decomposition of organ-ic matter.Current estimates suggest that rice pad-dies account for12%of the global methane emission (IPCC,1992,1995).Rice paddies also emit volatile sulfur compounds,which are ultimately oxidized to sulfate which in?uence cloud formation along with contributing to acid rain.Dimethyl sul?de(DMS: CH3SCH3)is the major sulfur gas emitted from rice paddies(Kanda et al.,1992)and wetland(Aneja et al.,
234
1979;Cooper et al.,1987;Steudler and Peterson,1985) in the terrestrial ecosystem.Although DMS is emitted from?ooded rice paddies,it is well known that the greater part of the global emission is emitted from the surface layer of the ocean by marine algae and plank-ton(Erickson et al.,1990;Rennenberg,1991).Present estimates of the contribution of marine environments and terrestrial ecosystems to the atmospheric sulfur burden(38–89Tg per year)are in the same range as man-made sulfur emissions(coal and oil combustion), but have a high degree of uncertainty(Rennenberg, 1991).
There are many reports concerning seasonal changes in methane?uxes and annual emissions from rice paddies in the world(e.g.,Sch¨u tz et al.,1989; Yagi and Minami,1990).The large differences in dai-ly,seasonal changes and in annual emission are due to the occurrence of many variables such as temperature, the use and mode of application of various fertiliz-ers,application of organic matter(rice straw,manure, etc.),soil type,irrigation management,and agricul-tural practices around the world(IPCC,1995).On the other hand,there is very little information con-cerning seasonal changes in volatile sulfur compound ?uxes including DMS(Kanda et al.,1992).Kanda et al.(1992)reported seasonal emission of DMS,sea-sonal absorption of carbonyl sul?de(COS)and some measurements of carbon disul?de(CS2)emission dur-ing cultivation and fallow periods.They reported that DMS?ux from rice paddies was in?uenced by the growth stage of rice plants and the highest?ux was observed immediately after heading.
There are three processes involved in the release of methane into the atmosphere from rice paddies:ebul-lition of gas bubbles,molecular diffusion across the water surface and plant mediated transport(Sch¨u tz et al.,1991).Of these,the most important is the?ow of methane through the aerenchyma system in rice plants(Cicerone et al.,1983;Nouchi et al.,1990). The release sites of methane in rice plants are the leaf sheaths(Nouchi et al.,1990).However,the mech-anism of emission of volatile sulfur compounds from wetlands is not well understood.It is uncertain whether the DMS emitted originates from microbial processes in the soil or from biosynthetic processes in the plant (Rennenberg,1991).
The main objective of our study was to gain bet-ter information about the seasonal changes in DMS and methyl mercaptan(CH3SH)?uxes from rice pad-dies.In this study,we measured the concentrations of volatile sulfur compounds including DMS,H2S,CS2and CH3SH in soil water and compared them with that
of methane which was measured simultaneously.The data allowed us to clarify the emission mechanism of
DMS from rice paddies.
Materials and methods
Experimental?eld site and treatment plots
Field measurements of methane,DMS and CH3SH
?uxes and the concentration of methane and S-compounds in soil water were carried out from May 10,1995to September18or October5,1995using
four lysimeters at the National Institute of Agro-Environmental Sciences,Tsukuba(3601N,14007 E).Each lysimeter had a surface area of10m2(2.5 4m)and contained a?ne textured Eutric Fluvisols with a depth of2.0m.The soil characteristics were as follows:soil texture:clay loam,total-C content:19g
kg1,total-N content:1.5g kg1,pH(H2O):6.0.Rice straw(total-C content:345g kg1,total-N content: 4.6g kg1)was incorporated into the soil of two plots at the rate of700g m2on March20,1995.Each lysimeter plot was plowed and fertilized with granular compound mineral fertilizers(N,P2O5and K2O,14% each)at71.4g m2as a basal application just before ?ooding on May8.Measurements of gaseous sulfur emission and dissolved sulfur gases were performed in two mineral fertilizer plots and two rice straw plots (mineral fertilizer+rice straw).
‘Nipponbare’,one of the leading rice cultivars in Japan,was pre-germinated in water on April12for three days and sown into seedling boxes.After incuba-tion,the seedlings were moved and grown in a green-house.On May11the seedlings were transplanted into hills,each hill had three seedlings,with a spacing of
0.21m between rows,and a spacing of0.21m within
a row.The planting density was22.7hills per square meter.The plots were permanently irrigated to keep the water level at3–15cm above the surface throughout the entire growing season.
Sampling procedure for methane and DMS?uxes
The closed chamber method described in detail by Nouchi et al.(1994)was used for measuring methane, CH3SH and DMS?uxes from the paddy?elds to the atmosphere.The chamber(length30cm;width30cm; height100cm,and open-bottomed)was constructed of colourless polycarbonate with a small fan for mixing
235
the air and a1L bag for regulating the air pressure in the box.The chamber enclosed two rice hills.To maintain stability against strong wind and to avoid dis-turbing the soil,the chamber was placed on an iron rectangular frame which had been previously placed on the soil surface.For measurement of?uxes from the?ood water,the chamber was placed on an unveg-etated area.Air samples(0.3L)for determining the methane concentration within the chamber were col-lected into a1L Tedlar(polyvinyl?uoride)bag?ve times every5min by an air pump(MP-2N,Toshiba Scienti?c Technology Ltd.,Tokyo,Japan).
Air samples for DMS were collected twice every 15or30min using a500mL glass syringe.The?rst air sampling was performed immediately after the cham-ber was set up,and a second air sampling was per-formed for30min at the early growth stage of rice and for15min at the middle or late growth stages. The second air sampling at the unvegetated area was performed30min after the?rst air sampling.
Methane and DMS?uxes were calculated from the temporal increase of concentration inside the cham-ber(Nouchi et al.,1994).All measurements were performed between1000h and1500h,except for measurements of the diurnal change in?ux.Methane, CH3SH and DMS?uxes were measured weekly during the entire cultivation period.
Collection of soil water
Soil water was collected according to the method of Nouchi et al.(1994).Soil water extractors were con-structed using hollow cylinders(100mm in length; 8mm in diameter)of highly permeable sintered polyethylene?lter cup.The polyethylene?lter cups were connected to Te?on tubes(1mm i.d.)and four ?lter cups(two per plot)were buried horizontally in the soil at a depth of5cm in each plot on May10. Soil water was extracted with a siphon and collected with a microsyringe in the case of methane or in the case of sulfur compounds stored in30mL glass tubes containing a small amount of powdered ascorbic acid (about40mg)to prevent oxidation.The exposed ends of the te?on tubes were capped with stoppers except at the time of sampling.
Measurements of methane
The methane concentration was determined by a gas chromatograph equipped with hydrogen?ame ioniza-tion detectors(GC-9A,Shimadzu Co.,Kyoto,Japan),a gas sampler with a5mL volume tube(MGS-4,Shi-madzu Co.)and an integrator(Chromatopac C-R6A, Shimadzu Co.).Methane gas was separated at90C using a glass column(3.2mm in diameter;2m in length)packed with molecular sieve5A(60–80mesh). The amount of methane dissolved in water was deter-mined by using a similar glass column packed with activated carbon(80–100mesh).The gas chromato-graph was calibrated by using a10mL CH4L1stan-dard gas from a pressurized cylinder. Measurements of sulfur compounds
A schematic diagram of the apparatus used to measure the sulfur compounds is shown in Figure1.First,the air in the500mL glass syringe was passed through a water trap(a glass midget impinger of5mL volume) maintained at18C in an ice-salt bath(Figure1A). With most of the water removed,the sulfur compounds were cryogenically trapped in a U-shaped glass tube (4mm in diameter;30cm in total length),which is packed with1,2,3-tris(2-cyanoethoxy)propane on60–80mesh Chromosorb W(AW DMCS treatment)and wrapped with a ribbon heater,submerged in liquid oxygen.The syringe,water trap,U-shaped tube and air pump were connected using te?on tubes(1mm in diameter),stainless steel needles and silicon septums. The air from the500mL glass syringe was completely trapped within8min by this procedure.
To concentrate sulfur compounds in the soil water or?ood water,a1mL sample of the water was poured into a plumbed glass tube which had a permeable glass ?lter at the bottom,and N2gas was bubbled from the bottom at a rate of40mL min1(Figure1B).The N2 gas including several volatile sulfur gases was intro-duced into the U-shaped glass tube trap immersed in liquid oxygen for3min.The U-shaped tube was then attached to a?ush sampling system(FLS-3,Shimadzu Co.)on a carrier gas line of a gas chromatograph(GC-8A,Shimadzu Co.)to measure the trapped sulfur com-pounds(Figure1C).The liquid oxygen was removed and the U-shaped tube was rapidly heated from183 C to80C within2min so that the contents reached a gaseous form.The trapped sulfur compounds were analysed with the gas chromatograph equipped with a?ame photometric detector(FPD)and an integra-tor(Chromatopac C-R6A,Shimadzu Co.).The sulfur compounds were separated at50C using a glass col-umn(3.2mm in diameter;3m in length)packed with 25%’oxydipropionitrile(ODPN)on60–80mesh Chromosorb W(AW DMCS treatment).Under these
236
Figure 1.A schematic diagram of the apparatus used to measure volatile sulfur compounds.A :Concentration method for sulfur compounds in an air,B :Concentration method for dissolved sulfur compounds in water,C :A cryogenic measuring system for sulfur compounds using a gas chromatograph with FPD.
conditions,the sulfur compounds,except for COS,were separated in the order of H 2S,CS 2,CH 3SH and DMS.COS could not be adequately separated from the air peak.
Standard H 2S,CS 2,CH 3SH and DMS gases were generated by a permeation method using a permeator
(P-1,Gastec Co.,Ltd.,Tokyo,Japan)at 35C from certi?ed permeation tubes for H 2S,CH 3SH and DMS (Gastec Co.,Ltd.)and a diffusion tube for CS 2.The permeation or diffusion rates at 35C were 382ng min 1cm 1for H 2S,36.4ng min 1cm 1for CH 3SH,
237 Figure2.Seasonal changes in methane?uxes from vegetated sites and the concentration of methane in soil water at a depth of5cm in the
mineral fertilizer and rice straw plots in Tsukuba during the1995cultivation period.The values of the?ux and concentration are the mean SD
of three samples from the two plots.
122ng min1cm1for DMS and580mg min1for
CS2.
Results
Seasonal changes in methane?ux from vegetated sites
and methane concentration in soil water in the
mineral fertilizer and rice straw plots
Seasonal changes in methane?ux from the vegetated
sites and methane concentration in the soil water in
the mineral fertilizer and rice straw plots are shown in
Figure2.Methane emission from the mineral fertilizer
plots was very low compared with the rice straw plots
and the highest?ux was4.56mg m2h1on Septem-
ber4(119days after?ooding).There were large spatial
variations in methane?uxes between the two rice straw
plots especially from early June through mid August.
The largest coef?cient of variation(CV)was102%for
the rice straw plots and57%for the mineral fertilizer
plots.The highest methane?ux from the rice straw
plots was26.6mg m2h1which was recorded on
July31(84days after?ooding).This was followed by
a gradual decrease during August and September.
The methane concentration in the soil water in the
mineral fertilizer plots was nearly zero until mid July,
after which it began to increase sharply until the end of
August(110days after?ooding),and maintained high
levels till harvest on September18.The concentration
in the soil water in the rice straw plots was detectable
from just after?ooding,but stayed at low levels until
early July(56days after?ooding).Thereafter,the
concentration in the soil water increased sharply and
continued rising till harvest(135days after?ooding).
Although the largest CVs during the cultivation period
in the mineral fertilizer and rice straw plots were123%
and64%,respectively,the highest concentration in the
soil water in the rice straw plots was6.6g CH4mL1.
From the early to the middle growth stages of rice,
an increase in methane?ux from the rice straw plots
coincided with an increase in the methane concentra-
tion in the soil water.The methane?ux from the min-
eral fertilizer plots was nearly zero because of the neg-
ligible methane concentration in the soil water.During
the late growth stage,however,methane?ux from the
mineral fertilizer plots didn’t increase much with the
increase in methane concentration in the soil water.
Moreover,the methane?ux from the rice straw plots
began to decrease,irrespective of whether the methane
concentration in the soil water increased.
Examination of methods for measuring the?ux of
sulfur compounds
To test a linear temporal increase in the concentra-
tion of several volatile compounds in the air within
the chamber,the chamber was successively placed on
rice grown in a pot on a water pan for30min and air
sampling was performed three times every15min.A
typical example obtained with the pot is presented in
Figure3.The concentrations of CH3SH and DMS in
the air within the chamber increased with time after
the setting of the chamber.Although there were linear
238
Figure 3.Temporal increase in the concentration of sulfur com-pounds in the air within the chamber.Data were obtained from pot grown rice plants.
relationships between the concentration and the time (CH 3SH:r =0.99and DMS:r =0.99),the ?rst slopes between 15min and 30min were lower than the second slopes between 0min and 15min.This result indicated that the ?uxes of CH 3SH and DMS from rice paddies decreased with the lapse of time under undesirable con-ditions for rice plants and their community.On the oth-er hand,the concentrations of H 2S and CS 2increased during the ?rst 15min and decreased thereafter.H 2S and CS 2may be decomposed by contact with the metal propelling fan,reabsorbed by rice plants or dissolved in waterdrops on the wall and rice leaves.From exam-ination of the chamber and the sampling methods,we determined that the ?ux data for this study were limited to CH 3SH and DMS and the air sampling was usually performed for 15min after the setting of the chamber.Diurnal change in DMS ?ux from the vegetated sites The relationships between the diurnal variations of DMS ?ux in the mineral fertilizer plots,solar radia-tion,air temperature and temperature in the chamber at the time just before the second air sampling on June 30(weather:lightly cloudy)and on July 11(weather:cloudless)are shown in Figures 4A and B,respectively.A large diurnal variation in DMS ?ux was observed.The DMS ?uxes were high in the daytime and low at night.The DMS ?ux at night was approximately 0.30times the maximum daytime ?ux.Although the DMS ?uxes were related to solar radiation,air temperature and temperature in the chamber,the DMS ?ux corre-lated most strongly with the solar radiation (r =0.83on June 30and r =0.90on July 11)compared with the
temperature in the chamber (r =0.73on June 30and r =0.91on July 11).
Seasonal changes in DMS ?uxes from the mineral fertilizer and rice straw plots
The DMS ?uxes from the mineral fertilizer and rice straw plots are shown in Figures 5and 6,respective-ly.The DMS ?uxes from the unvegetated sites in the mineral fertilizer and the rice straw plots were ?rst observed on May 18(10days after ?ooding),continued until July 19(72days after ?ooding)and disappeared thereafter.The DMS emission from the unvegetated sites in the paddy ?elds is considered to be associated with the DMS generation by algae,phytoplankton and microorganisms in the ?ood water.
A large seasonal change in DMS ?ux from the vegetated sites of the mineral fertilizer and rice straw plots was observed during the entire cultivation peri-od.Application of rice straw had little in?uence on DMS ?uxes.The trend of seasonal change in DMS ?ux from the vegetated sites at the rice straw plots was almost the same as that observed at the mineral fertil-izer plots.The DMS ?uxes from the mineral fertilizer and the rice straw plots increased with the growth of rice plants from 43days after ?ooding.The maximum values of 13.9g m 2h 1and 15.1g m 2h 1at the mineral fertilizer and rice straw plots,respective-ly,were observed on August 3,7days before head-ing (August 10).After heading,DMS ?uxes rapidly decreased.Since the DMS ?uxes from the unvegetat-ed sites in the mineral fertilizer and rice straw plots from May 18to July 19were almost the same com-pared to those from the vegetated sites in both plots,the DMS appears to be associated with growing rice plants,i.e.from early July to mid September.
CH 3SH ?ux from rice paddies
CH 3SH emission from the unvegetated and vegetated sites in the mineral fertilizer and rice straw plots was detected only in May (Figure 7).CH 3SH ?uxes from the vegetated sites were nearly equal to those from the corresponding unvegetated site in both plot types.CH 3SH ?uxes from the rice straw plots were very high compared with the mineral fertilizer plots.Maximum ?uxes from the rice straw and mineral fertilizer plots were 10.7g m 2h 1and 4.2g m 2h 1,respec-tively,which were recorded on May 24.
239
Figure4.Diurnal changes in DMS?uxes,global solar radiation,air temperature and temperature in the chamber measured during the second sampling at the vegetated site in the mineral fertilizer plot on June30(A)and July11(B),1995.June30was a slightly cloudy day whilst July 11was an almost clear day except for having light intermittent clouds at noon.
Figure5.Seasonal changes in DMS?uxes from unvegetated and vegetated sites in the mineral fertilizer plots in Tsukuba during the1995 cultivation period.The change in plant height is also shown.The values of the?ux from the vegetated sites are the mean SD of three samples from the two plots.Each?ux data point for the unvegetated site was obtained from only one site.Each point of the plant height data is the mean standard deviation of4rice hills.