Biogeosciences Increased phosphorus availability mitigates the inhibition of nitrogen deposition on CH 4 uptake in an old-growth tropical forest , southern China

It is well established that tropical forest ecosystems are often limited by phosphorus (P) availability, and elevated atmospheric nitrogen (N) deposition may further enhance such P limitation. However, it is uncertain whether P availability would affect soil fluxes of greenhouse gases, such as methane (CH 4) uptake, and how P interacts with N deposition. We examine the effects of N and P additions on soil CH4 uptake in an N saturated old-growth tropical forest in southern China to test the following hypotheses: (1) P addition would increase CH 4 uptake; (2) N addition would decrease CH 4 uptake; and (3) P addition would mitigate the inhibitive effect of N addition on soil CH4 uptake. Four treatments were conducted at the following levels from February 2007 to October 2009: control, N-addition (150 kg N ha −1 yr−1), P-addition (150 kg P ha−1 yr−1), and NP-addition (150 kg N ha −1 yr−1 plus 150 kg N ha −1 yr−1 plus 150 kg P ha −1 yr−1). Static chamber and gas chromatography techniques were used to quantify soil CH4 uptake every month throughout the study period. Average CH4 uptake rate was 31.2 ± 1.1 μg CH4C m−2 h−1 in the control plots. The mean CH 4 uptake rate in the N-addition plots was 23.6 ± 0.9 μg CH4C m−2 h−1, significantly lower than that in the controls. Paddition however, significantly increased CH 4 uptake by 24 % (38.8± 1.3 μg CH4-C m−2 h−1), whereas NP-addition (33.6± 1.0 μg CH4-C m−2 h−1) was not statistically different from the control. Our results suggest that increased P availability may enhance soil mathanotrophic activity and root growth, resulting in potentially mitigating the inhibitive effect of N deposition on CH4 uptake in tropical forests. Correspondence to: J. Mo (mojm@scib.ac.cn)


Introduction
Methane (CH 4 ) is considered the second most important greenhouse gas after carbon dioxide and with a global warming potential 25 times compare to carbon dioxide (CO 2 ) over a 100 year horizon (IPCC, 2007).The global atmospheric concentration of CH 4 has increased from a pre-industrial value of about 0.715 ppm to 1.732 ppm in the early 1990s, and 1.803 ppm in 2009 (IPCC, 2007;WMO, 2010), due primarily to the anthropogenic emissions from energy production, rice cultivation, ruminant animals, biomass burning and landfills (IPCC, 2007).
Atmospheric methane originates mainly from a biological process.In forest soils, it is produced in anoxic layer by methanogenic bacteria during the anaerobic digesting of organic matter.Methane is also eliminated into forest soils by microbial oxidation (methanotrophy) in the aerobic zone and which oxidises atmospheric methane.Methanotrophs in forest soil use CH 4 as only a carbon (C) and energy source and oxygen availability is the main factor limiting their activity (Le Mer and Roger, 2001).Therefore, methanotrophy in forest soils is primarily dependent on physical factors controlling soil diffusion, e.g., water content and soil texture (Butterbach-Bahl and Papen, 2002;Templeton et al., T. Zhang et al.: Increased phosphorus availability 2006).Teh et al. (2005) indicate that the effective diffusion at soil surface controls the overall rate of methane consumption.However, forest soils as biological sinks of CH 4 are also subjected to other abiotic and biotic controls.It is well known that methanotrophs have different temperature and pH optima.Effect soil temperature on methanotrophic community has been well studied (Mohanty et al., 2007).Soil pH seems to be a less important control for atmospheric methane oxidation (Kolb, 2009).Tree species would affect atmospheric CH 4 oxidation without altering community composition of soil methanotrophs (Menyailo et al., 2010), Jamali et al. (2011) found that termites offset 21 % of CH 4 consumed by soil in a tropical savanna woodland.Furthermore, it is believed that nitrogen (N) availability controls the activity and affects the community structure of methanotrophs in soil (Bodelier and Laanbroek, 2004).
Among the above factors, the effects of nitrogen deposition on CH 4 oxidation have received increasing attention (Steudler et al., 1989;Bodelier and Laanbroek, 2004;Tate et al., 2007;Zhang et al., 2008).Nitrogen addition alters the fluxes of greenhouse gases (GHGs, including CH 4 ) through regulating plant and microbial activities that are directly associated with GHGs production and consumption (Liu and Greaver, 2009).Steudler et al. (1989) first reported in a temperate forest that N fertilization reduced soil CH 4 uptake by 33 %.Extensive research has been conducted to investigate the relationship between CH 4 consumption and N input and it has been generally accepted that CH 4 uptake is inhibited by nitrogenous fertilization (Bodelier and Laanbroek, 2004;Chan et al., 2005).Laboratory studies indicate that oxidation of CH 4 by a variety of methanotrophs is competitively inhibited by N (Ferenci et al., 1975).In soils, the inhibition of CH 4 oxidation by ammonium is attributed to a competition at the level of the methane mono-oxygenase, a transfer of the methanotrophy activity towards nitrification (Castro et al., 1994) and the toxicity of NO 2 produced.Aluminium toxicity after extensive N input may also inhibit soil CH 4 uptake (Bradford et al., 2001a;Zhang et al., 2008).On the other hand, positive effects of increased N availability on CH 4 uptake rates were found in severely N-limited forests (Börjesson and Nohrstedt, 2000;Steinkamp et al., 2001).In addition, soil available phosphorus would also affect CH 4 flux by regulating changes in soil physico-chemical properties, plant root activities and soil microbial activities that are involved in CH 4 consumption.Increased P availability in forest soil would stimulate plant root growth and lead to higher water uptake, and then consequently lower soil water content which led to higher gas diffusion and, thus, higher CH 4 oxidation.Methanotrophs also needs phosphorus for community structure growth.
While N is often the primary limiting nutrient in temperate and boreal forests, phosphorus usually limits ecological processes in tropical and subtropical forests (Vitousek and Sanford, 1986;Cleveland et al., 2002;Vitousek et al., 2010).We studied the relationships between soil P availability and CH 4 flux in an old-growth tropical forest located in southern China where elevated N deposition has been well documented (Mo et al., 2006;Fang et al., 2008Fang et al., , 2011)).Earlier studies in this old-growth forest showed that no N retention occurred, but rather a net loss of 8-16 kg N ha −1 yr −1 from the soil was estimated in the old-growth forest.In total, up to 60 kg N ha −1 yr −1 was leached from the old-growth forest, indicating that this forest was completely N saturated (Fang et al., 2008(Fang et al., , 2011) ) from the chronic elevated N deposition in southern China.This interpretation is also supported by the results on litterfall production, which revealed no significant effects of N additions on total litterfall production in the old-growth forest (Mo et al., 2008).Studying forest CH 4 uptake and its relationship with soil available P and elevated N deposition is very important for evaluating the contribution of tropical forests to global climate change.Due to limited research in tropical forests, it is not clear how P availability would affect soil CH 4 uptake, and how P addition may interact with N deposition.This 33 months study experimentally tested the effects of P and N availabilities on soil CH 4 uptake.We hypothesized that: (1) N addition would inhibit soil CH 4 uptake as we found in a previous study in the same forest (Zhang et al., 2008); (2) P addition would increase soil CH 4 uptake due to that P addition world stimulate plant root growth and lead to higher water uptake, and then consequently lower WFPS which lead to higher gas diffusion and, thus, higher CH 4 uptake; (3) NP addition (interactive effect of N and P) would have less effect on soil CH 4 uptake comparing with N or P addition alone.This means that P addition would mitigate the inhibitive effect of N addition on soil CH 4 uptake.

Site description
This study was conducted in the 1200 ha Dinghushan Biosphere Reserve (DHSBR), which is located in the middle of Guangdong Province, southern China (112 There is an old-growth evergreen broadleaf forest (mature forest) in this reserve.The old-growth forest has been well protected from human activity for more than 400 years (Wang et al., 1982;Mo et al., 2003;Tang et al., 2006).The average annual precipitation of 1927 mm in the reserve has a distinct seasonal pattern, with 75 % falling from March to August and only 6 % falling from December to February (Huang and Fan, 1982).The mean annual temperature is 21 • C with the January mean temperature of 12.6 • C and July mean temperature of 28.0 • C (Huang and Fan, 1982).Annual mean relative humidity is 80 % (Huang and Fan, 1982).The wet N deposition was 36-38 kg N ha −1 in the 1990s (Zhou and Yan, 2001).Precipitation N deposition in this region was 34.1 kg N ha −1 yr −1 , with roughly 1.5:1 NH + 4 to NO − shale formation (Wu et al., 1982).The soil depth in the oldgrowth forest is more than 60 cm to the top of the C horizon (Mo et al., 2003).The forest in this experiment is situated on mountain slopes about 30 • -35 • .
A total of 20 plots of 5 m × 5 m were established and each plot was surrounded by a 5-m-wide buffer strip.Plots size and fertilizer level were referenced to the experiment in Costa Rica by Cleveland and Townsend (2006).Field plots and treatments were laid out randomly.NH 4 NO 3 and NaH 2 PO 4 solutions were sprayed once every other month to the forest floor with a backpack sprayer starting from February 2007 and continued through October 2009.Fertilizer was weighed and mixed with 5 l of water for each plot.Each control plot received 5 l of water without fertilizer.

Field sampling and measurements
CH 4 , CO 2 and nitrous oxide (N 2 O) flux were measured from January 2007 before the first fertilizer application.Static collars were installed in each plot in November 2006, two months before the gas sampling.Gas fluxes were monitored once every month using the static chamber and a gas chromatograph (Agilent 4890D).The static chamber was a 25cm-diameter by 16-cm-tall PVC pipe permanently anchored 8 cm into the soil.During gas collection, a 30-cm-tall removable cover chamber was attached tightly to the anchor ring with a rubber band.Gas samples were collected from each chamber from 09:00-10:00 LT.Diurnal studies in the adjacent forests found that greenhouse gas fluxes measured during the mid-morning (09:00-10:00 LT) were closer to the daily mean (Tang et al., 2006).The GHG concentrations remained linear for up to 100 min after the chamber was closed in our study.Gas samples were taken with a 60 ml plastic syringe at 0 and 30 min after the chamber closure.Before each sampling, syringes were flushed three times with chamber gas to mix the headspace.Laboratory tests showed that chambers and syringes were inert to N 2 O, CO 2 and CH 4 (Steudler et al., 1989;Bowden et al., 1990).Gas samples were analysed within 12 h in a gas chromatograph (Agilent 4890D) equipped with a flame ionization detector (FID) for CH 4 and CO 2 , and an electron capture detector (ECD) for N 2 O. CO 2 was transformed into CH 4 via (Ni)H 2 before the FID analysis.Calibration gases (CH 4 at 1.87 ppm, CO 2 at 418 ppm, N 2 O at 0.321 ppm, bottle's No. 070811) were obtained from the Institute of Atmospheric Physics, Chinese Academy of Sciences.In this paper, we report only data of soil CH 4 in the old-growth forest.
The calculation of GHG flux followed that described in Zhang et al. (2008), based on a linear regression of chamber gas concentration versus time (IAEA, 1992;Holland et al., 1999).Atmospheric pressure was measured at the sampling site using an air pressure gauge (Model THOMMEN 2000, Switzerland).Air temperature (enclosure), soil temperature (at 5 cm depth) and moisture (0-10 cm depth) were measured during each sampling.Soil moisture content was detected using a TDR-probe (Time Domain Reflectometry, Model Top TZS-I, China).Soil moisture (0-10 cm depth) values were converted to WFPS (Water Filled Pore Space) according to the following formula: where SBD is soil bulk density, Vol is volumetric water moisture and 2.65 is the density of quartz.
Soil samples were collected in February 2007 (before the first fertilizer application) and February 2009 (after two years of fertilization) for the physical and chemical properties.Five soil cores (2.5 cm inner diameter) were collected randomly from each of the 20 plots at 0-10 cm soil depths and combined to one composite sample.The litter layer was carefully removed before the soil sampling.The pH of the soil sample was measured in a 1:2.5 soil/water suspension.Total N concentration was determined by the micro-Kjeldahl digestion followed by the analysis of ammonium on a Wescan ammonia analyser, while total P concentration was analysed colorimetrically after acidified ammonium persulfate digestion (Anderson and Ingram, 1989).Available P was extracted with 0.03 M ammonium fluoride and 0.025 M hydrochloric acid and analysed colorimetrically (Anderson and Ingram, 1989).

Statistical analysis
Repeated measures of Analysis of Variance (ANOVA, PROC MIXED with AR(1) from SAS -SAS Institute Inc., Cary NC, USA) was used to examine the effect of fertilizer treatments on soil GHG fluxes from February 2007 to October 2009.Two-way ANOVA (PROC GLM from SAS) was used to examine the effects of N and P addition.One-way ANOVA was used to examine the difference in soil pH, NH + 4 , NO − 3 and available P among treatments.Linear regression analysis was performed by Origin 8.0 (OriginLab Corporation, Northampton, MA USA) to examine the relationship between CH 4 fluxes and soil WFPS contents and soil temperature.Out of 680 observations, three were identified as outliers, which were probably caused by chamber leaks, abnormally high WFPS, and other unknown factors, and were removed from the data analyses.

Soil temperature and WFPS
Soil temperature (at 5 cm depth) followed the air temperature in all plots, with temperatures increased from spring to summer and decreased from fall to winter (Fig. 1a).There was no treatment effect on soil temperature during the study period.Soil WFPS (0-10 cm depth) rose following the increased precipitation from dry winters to wet springs but decreased in summer, possibly due to plant uptake and higher evaporation, despite the high amount of precipitation in summer (Fig. 1b).Repeated measurement ANOVA showed that soil WFPS was significantly lower in the P-addition and NPaddition plots (p = 0.011 and p = 0.018, respectively) than in the control plots.However, there was no treatment effect in the N-addition plots (p = 0.165).

Other soil properties
Nitrogen and P treatments changed soil nutrient conditions (Fig. 2).Soil pH increased significantly in the P-addition plots (p = 0.021) and was a little higher in the NP-addition plots (p = 0.062), while no change in the N-addition plots (p = 0.933) (Fig. 2a) compared with the control plots.After 24 months of the treatment, a six-fold of soil available P was observed in P-addition plots (p = 0.0003) and four-fold in NP-addition plots (p = 0.012), compared to the controls (Fig. 2c).Furthermore, total P was significantly lower in Naddition plots compared to control plots (p = 0.001).This pattern is the same as pre-treatment (Fig. 2d).Soil NH + 4 concentrations did not show any significant change in any treat-  ment plots (N-addition, P-addition and NP-addition) compared to the controls (p = 0.163, 0.152 and 0.368, respectively) (Fig. 2b).

Soil CH 4 fluxes
Pre-treatment gas measurement (January 2007) showed no difference in different treatment plots compared to control plots (p = 0.807, 0.559 and 0.961, respectively, in N, P, NP treatment plots).After fertilization treatment, repeated measurement ANOVA showed that P addition significantly increased soil CH 4 uptake while N addition significantly decreased CH 4 uptake (Fig. 3).The mean soil CH 4 uptake rate was 31.2 ± 1.1 µg CH 4 -C m −2 h −1 in the control plots during the 33 months study period.Soil CH 4 uptake in the P-addition plots was significantly higher (mean CH 4 uptake rate was 38.8 ± 1.3 µg CH 4 -C m −2 h −1 , p = 0.0068).On the contrary, N-addition significantly inhibited soil CH 4 uptake (mean CH 4 uptake rate was 23.6 ± 0.9 µg CH 4 -C m −2 h −1 , p = 0.007).No significant difference was observed between NP-addition plots (33.6 ± 1.0,µg CH 4 -C m −2 h −1 ) and the controls (p = 0.177).CH 4 uptake was higher in summer and fall (when the soil was low in water content) and was lower in spring (when the soil was wet) and winter (when the soil was cold).Twoway ANOVA showed a significant positive P effect on CH 4 uptake in the summers of 2007, 2008and 2009, fall 2007, and the springs of 2008and 2009 (Fig. 4 (Fig. 4), and a negative N effect in the springs of 2008and 2009, and summer 2009 (Fig. 4 (Fig. 4).In summer 2007, CH 4 uptake in the P-addition and NP-addition plots was 26.6 % and 27.3 % higher than in the control plots (p = 0.049 and 0.051, respectively).CH 4 uptake in the N-addition plots was reduced by 30.CH 4 fluxes and soil WFPS were positively correlated in the control plots and N and P addition plots (Fig. 5).Under similar WFPS conditions, CH 4 uptake was the highest in Paddition plots and lowest in N-addition plots.CH 4 flux was not correlated to soil temperature.

Compare with other tropical forests
The annual CH 4 uptake rates in our old-growth tropical forest site in southern China ranged from 2.3 to 3.4 kg CH 4 -C ha −1 yr −1 (11 months in 2007, 12 months in 2008, 10 months in 2009), and were higher in the dry year ( 2007) than in the wet year (2008).The CH 4 fluxes, quantified in this study, are similar to the previous reported CH 4 fluxes measured at an adjacent forest (Tang et al., 2006;Zhang et al., 2008) and other parts of tropical Southwest China (Werner et al., 2006;Yan et al., 2008;Wang et al., 2010).The rates of CH 4 uptake in other tropical forests are also similar, ranging from 0.8 to 4.73 kg CH 4 -C ha −1 yr −1 (Steudler et al., 1991;Kiese et al., 2003;Davidson and Nepstad, 2004;Davidson et al., 2008).

Effects of soil temperature and pH
Temperature variation has a minor impact on soil CH 4 uptake in our studied forest, which is consistent with previous results from the adjacent forest (Tang et al., 2006).In temperate or boreal regions, in contrast, soil CH 4 oxidation was generally positively correlated with soil temperature (Crill, 1991;King, 1997).Castro et al. (1995) observed that methanotrophy was affected by soil temperature between −5 and 10 • C, but in our site, almost all of soil temperatures were between 10 and 30 • C (Fig. 1a).The higher pH after P addition might contribute to the increase in CH 4 uptake rate in this forest.Most of the soils oxidised CH 4 over a pH range of 3 ∼ 7.5 supporting in situ observations (Born et al., 1990).Bacteria extracted from boreal acidic forest soils had their highest CH 4 oxidation activity at pH 5.8 (Amaral et al., 1998).Soil pH in this acidic old-growth forest almost below 4 or even lower.Bradford et al. (2001b) observed that the drop in pH in a beech forest in the UK contributed to a decrease in soil CH 4 uptake.However, in some acidic peat soils, the rates of methane uptake were only slightly influenced by pH at values between 4.0 and 6.0 and decreased sharply at values below and above this range, although oxidation still occurred below pH 4.0 (Hanson and Hanson, 1996).Therefore, P addition increased soil pH in our study which may lead to increased methane oxidation.

Effects of soil WFPS
Soil CH 4 uptake was negatively correlated with the soil WFPS (a better measurement of soil moisture) in this study (Fig. 5), which was consistent with previous publications (Born et al., 1990;Castro et al., 1995;Kiese et al., 2003).Steinkamp et al. (2001) reported similarly that soil moisture was the dominant factor controlling CH 4 uptake when soil temperature was >10 • C. Methanotrophs use CH 4 as only a C and energy source and oxygen availability is the main factor limiting their activity (Le Mer and Roger, 2001)  WFPS after P addition in our study might lead to higher soil aeration and more available oxygen and atmospheric methane to methanotrophs.

Effects of P and N fertilization
We found P addition significantly increased CH 4 uptake, N addition decreased CH 4 uptake, and P addition mitigated the negative N effect in this 33 months field experiment, as hypothesized.We believe this is the first experimental testing of N and P limitation on soil CH 4 flux in tropical and subtropical forests.The treatment effects were particularly strong in the summers and falls, when soil uptakes of CH 4 were the highest (Fig. 4).Phosphorus fertilization was often conducted in agricultural ecosystems.On planted rice soils, phosphorus addition significantly decreased CH 4 emission (Lu et al., 1999) probably by increasing methanotrophic potential (Joulian et al., 1998).We infer that this response of CH 4 uptake to the experimental P additions could be resulted from increased soil diffusion due to plant root growth.
In our study, we found that soil WFPS was decreased after phosphorus addition, particularly in the summers (growing season) with a higher methane oxidation.This decreasing of soil WFPS was likely due to increased plant water consumption after P addition.Since phosphorus is a common limiting nutrient in tropical forests (Herbert et al., 2003;Wardle et al., 2004), low phosphorus availability due to strong sorption is often the main constraint on plant growth on highly weathered tropical soils (Vitousek, 1984).Ostertag (2001) also observed P-fertilizer leads to a greater root turnover in 4.1 million year-old forests in Hawaii.In these forests, two years fertilization of 150 kg ha −1 yr −1 did not increase total soil phosphorus probably due to serious plant uptake under high N deposition.Phosphorous deficiency in plant growth in this old-growth forest is also partially supported by the results on litterfall study, which showed that annual total litterfall fluxes were significantly increased after P addition (Liu et al., 2011).Thus, P addition might stimulate plant roots sorption of phosphorus and water uptake which lead to lower soil WFPS in this forest.Higher water uptake by plant roots would subsequently result in higher soil diffusion.Therefore, higher diffusion for oxygen and atmospheric methane would increase methanotrophy activities.In addition, since phosphorus is a common limiting nutrient in tropical forests (Herbert et al., 2003;Wardle et al., 2004), microbial activity could be strongly P limited (Cleveland et al., 2002).Increased soil CH 4 uptake in this study may be due to stimulated methanotrophy activities after increased P availability in soil.In a companion study, we found that microbial biomass increased significantly by P addition (Liu et al., 2011).However, this could be a minor contribution to the increased CH 4 uptake due to methane oxidizing bacteria representing only a very small portion of the total soil microbial biomass.On the other hand, phosphorus fertilizer may increase the mineralization of organic P (Ofori-Frimpong and Rowell, 1999) and reduce Al 3+ toxicity.Increased soil P levels might stabilize soil aluminum and iron through geochemical reaction of adsorption (Frossard et al., 1995).This coupling effect decreased aluminum toxicity to methanotrophs (Nanba and King, 2000).However, in our experiment, we found no treatment effect on soil exchangeable Al 3+ (data was not shown here).The inhibitive effect of N input on CH 4 uptake has been reported extensively.Zhang et al. (2008) suggested that the response of forest soil CH 4 uptake on N fertilization possibly depends on soil N status.They compared the CH 4 uptake under various N additions in three tropical forests with very different N conditions and found that N addition in Nsaturated old-growth forest significantly decreased soil CH 4 uptake while in N-limited young forests N addition had a limited effect.Singh et al. (1997) observed in a natural gradient of tropical forests, sites with higher soil mineral nitrogen concentrations had lower CH 4 uptake.In a Puerto Rican wet forest, Steudler et al. (1991) found that maximum reduction in CH 4 uptake coincided with the highest soil NH + 4 , suggesting that nitrogen-CH 4 linkage observed in temperate forests (Steudler et al., 1989) may also function in tropical forests.NH + 4 can reduce the growth rate of many methanotrophs by inhibiting CH 4 oxidation (Whittenbury et al., 1970).Laboratory studies indicate that oxidation of CH 4 by a variety of methanotrophs is competitively inhibited by nitrogen (Ferenci et al., 1975).In soils, the inhibition of CH 4 oxidation by ammonium is attributed to a competition at the level of the methane mono-oxygenase, a transfer of the methanotrophy activity towards nitrification (Castro et al., 1994) and the toxicity of NO 2 produced.Schnell and King (1994) observed that nitrite, the end product of methanotrophic ammonia oxidation, was a more effective inhibitor of CH 4 oxidation than ammonium.This inhibition becomes irreversible and can only be released at CH 4 concentrations higher than 100 ppm (King and Schnell, 1994).In this oldgrowth forest, we observed that atmospheric CH 4 concentrations was around 2 ppm and even in the soil can not come up to 100 ppm.We suggest that the inhibition of CH 4 uptake by ammonium has occurred and persisted.However, we did not observe soil NH + 4 concentration change after treatments in this study, which might be due to the rapid volatile and leaching of the fertilization before our soil sampling.

Conclusions
Nitrogen deposition would weaken the function of upland forest soils as CH 4 sinks by its inhibitive effect on soil CH 4 uptake and by decreasing availability of phosphorus due to continuous nitrogen deposition and phosphorus limitation in the tropical forests of southern China and other parts of the world.Galloway et al. (1994) forecasted a large increase in nitrogen deposition from fossil-fuel, growing population and agricultural activity for year 2020 in those regions.However, P addition could reverse such trends by stimulating forest soil CH 4 uptake.In this 33 months field experiment, we showed that increased P availability might mitigate the inhibitive effect of N addition on soil CH 4 uptake in this N-saturated oldgrowth tropical forest.While average annual consumption of CH 4 was enhanced after P addition and decreased after N addition, there was no difference between the NP-addition plots and the controls (i.e., the negative effect of N addition on CH 4 uptake was reduced by P addition).As far as we know, our study is among the first field study experimentally testing the responses of soil CH 4 uptake to N and P additions in tropical forests.Results from our study suggest that phosphorus fertilization in tropical forests could be one of the future choices for mitigating the inhibitive effect of N deposition on soil CH 4 uptake.

Figure 1 .
Figure 1.Soil temperature at 5cm depth (a) and soil WFPS (b) during the study period.Error bars represent standard error of means (N = 5).Asterisk (*) and double asterisk (**) indicates significant differences between control and at least one of the experiment treatments at p < 0.1 and p < 0.05, respectively.

Fig. 1 .
Fig. 1.Soil temperature at 5cm depth (a) and soil WFPS (b) during the study period.Error bars represent standard error of means (N = 5).Asterisk ( * ) and double asterisk ( * * ) indicates significant differences between control and at least one of the experiment treatments at p < 0.1 and p < 0.05, respectively.

Figure 2 .
Figure 2. Soil pH (a), ammonia nitrogen (b), available P (c) and total P (d) at 5cm depth in Feb. 2007 and Feb. 2009.Error bars represent standard error of means (N = 5).Different letters denote significant difference (p < 0.05) between treatments by the Repeated Measurement ANOVA.

Fig. 2 .
Fig. 2. Soil pH (a) , ammonia nitrogen (b) , available P (c) and total P (d) at 5 cm depth in February 2007 and February 2009.Error bars represent standard error of means (N = 5).Different letters denote significant difference (p < 0.05) between treatments by the Repeated Measurement ANOVA.

Figure 3 Fig. 3 .Figure 4 .Fig. 4 .
Figure 3 Comparisons of mean soil CH4 fluxes between treatments from 2007 to 2009 after N and P additions.Bars indicate ± 1 SE, N = 5.Different letters denote significant difference (p < 0.05) between treatments by the Repeated Measurement ANOVA.Fig.3. Comparisons of mean soil CH 4 fluxes between treatments from 2007 to 2009 after N and P additions.Bars indicate ±1 SE, N = 5.Different letters denote significant difference (p < 0.05) between treatments by the Repeated Measurement ANOVA.

Figure 5 .
Figure 5. Correlations between CH4 flux and soil WFPS under different fertilization treatments.Under similar soil WFPS conditions, CH4 uptake was higher in P-addition plots and lower in N-addition plots.

Fig. 5 .
Fig.5.Correlations between CH 4 flux and soil WFPS under different fertilization treatments.Under similar soil WFPS conditions, CH 4 uptake was higher in P-addition plots and lower in N-addition plots.