SEARCH

SEARCH BY CITATION

Keywords:

  • microbial biomass carbon;
  • microbial biomass nitrogen;
  • microbial respiration;
  • soil–plant nitrogen dynamics;
  • tropical cropland

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Soil microbes are considered to be an important N pool in dry tropical croplands, which are nutrient poor. To evaluate the N contribution of soil microbes to plant growth in a dry tropical cropland, we conducted a maize cultivation experiment in Tanzania using different land management treatments (no input, plant residue application, fertilizer application, plant residue and fertilizer application, and non-cultivated plots). Over 104 experimental days, we periodically evaluated the microbial biomass N and C, plant N uptake, microbial respiration in situ and inorganic N in the soil. A significant amount of inorganic N was lost in all of the treatment plots as a result of leaching during the initial 60 days and inorganic N remained low thereafter (∼20–35 kg N ha−1 : 0–15 cm), whereas soil microbial respiration substantially decreased because of soil drying after 60 days (grain-forming stage). During the grain-forming stage (60–104 days), we found a distinct effect of plant N uptake on soil microbial dynamics, although we did not observe an obvious effect of plant residue and/or fertilizer application; microbial biomass N decreased drastically from 63–71 to 18–33 kg N ha−1 and the microbial biomass C : N ratio simultaneously increased (>10-fold) in all maize-cultivated plots; these features were not observed in the non-cultivated plot. Plant N uptake over the same period was 26.6–55.2 kg N ha−1, which was roughly consistent with the decrease in microbial biomass N. These results indicate that strong competition for N occurred between soil microbes and plants over this period and N uptake by plants prevented microbial growth. Thus, we concluded that soil microbes contribute to plant growth by serving as a N source during the grain-forming stage in dry tropical cropland.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Soil microbes act as an important N pool in soil–plant N dynamics in terrestrial ecosystems (Bardgett et al. 2002, 2007; Singh et al. 1989; Wardle et al. 2004), and recent studies have emphasized their importance for crop N uptake in dry tropical croplands developed on nutrient-poor soil (Cookson et al. 2006; Ghoshal and Sihgh 1995; Singh et al. 2007b). To improve the efficiency of N utilization in nutrient-poor agroecosystems, it is necessary to elucidate the time-course of N partitioning between the soil N pool (microbial biomass N [MBN] and inorganic N in soil) and plant N uptake during the crop growth period, and evaluate the influence of land management on the soil–plant N dynamics (Herai et al. 2006; Kushwaha et al. 2000).

Many studies that have evaluated seasonal microbial dynamics in dry tropical ecosystems have reported that MBN is depleted during the rainy season, but remains high during the dry season (Singh et al. 1989; Sugihara et al. 2010; Wardle 1992, 1998). This is because drastically increased soil moisture during the early rainy season depletes soil microbes as a result of microbial cell lysis (Fierer et al. 2003; Tripathi and Singh 2007) and enhanced grazing by soil macrofauna (Michelsen et al. 2004). The decrease in MBN leads to nutrient release from soil microbes during the rainy season, when plant activity is high, and the soil microbes are considered to be an N source for plant growth. In contrast, the decrease in MBN is also considered to result from strong competition for N among plants and soil microbes during the crop growth period (Kushwaha et al. 2000; Singh et al. 2007b; Srivastava and Lal 1994). Singh et al. (2007b) evaluated temporal variations in MBN with crop root biomass over 2 years in India and observed a negative relationship between MBN and crop root biomass and suggested that strong competition occurred between soil microbes and crops for available nutrients for the seedling and grain-forming stages during both rice and barley crop seasons. However, most studies have focused only on microbial biomass dynamics and, as far as we know, no study in dry tropical croplands has ever reported the time-course of MBN with the plant N uptake pattern. Therefore, both the relationship between MBN variation and plant N uptake pattern and the N contribution of soil microbes to plant growth remain unclear.

Furthermore, land management (e.g. organic matter application, fertilizer application, soil tillage) affects soil microbial dynamics (Joergensen and Emmerling 2006; Spedding et al. 2004). Many studies have reported that organic matter application increases the microbial biomass carbon (MBC) or MBN in tropical croplands, suggesting that increased MBC or MBN contributes to plant growth by promoting their nutrient source and sink function (Powlson et al. 2001; Singh et al. 2007a,b). To utilize the soil microbes as an N pool for crop growth, it is necessary to understand the effect of land management on the relationship between MBN variation and plant N uptake pattern in this agroecosystem.

In the present study, we conducted maize cultivation experiments using various land management treatments (plant residue and/or fertilizer application) in Tanzania and evaluated the dynamics between the soil N pool (MBN and inorganic N) and plant N uptake by considering the microbial decomposition rate and environmental factors. We also included a non-cultivated plot for comparison of MBN with cultivated/treated plots. Our objectives were: (1) to elucidate the effect of soil microbes as an N source for plant growth during the crop growth period, (2) to evaluate the effect of land management on the relationship between the soil N pool and plant N uptake in a dry tropical cropland.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Description of the study site

We conducted the field experiment at the agricultural experimental station of the Sokoine University of Agriculture, Morogoro, Tanzania (6°51′S, 37°40′E; 579 m a.s.l.). The mean annual temperature is 24.5°C (2000–2005) and annual rainfall ranges from 750 to 1000 mm. The rainy season usually exhibits a bimodal distribution pattern, that is, the long rainy season (from the middle of February to May) is more reliable and better distributed for planting and the duration and intensity of the short rainy season (from October to December) is less predictable. The study was conducted during the long rainy season in 2006. The soil at the experimental site is Kanhaplic Haplustults (Soil Survey Staff 2006); in brief, the soil characteristics of the plow layer (0–15 cm) are as follows: pH 5.8 (as determined in water; soil : water, 1:5 w/v); soil texture is sandy clay containing 56.0% sand, 10.9% silt and 33.1% clay; bulk density, 1.21 g cm−3; total organic C and N, 12.4 and 1.1 g kg−1 soil, respectively. Maize had been planted at the experimental site every year since 2003 without any fertilizer, and the soil characteristics in each plot were similar to those assessed in July 2005.

Experimental design

The experimental design included the following five treatments:

  • 1
     C plot: Control (no input);
  • 2
     F-plot: chemical fertilizer treated plot (urea equivalent to 100 kg N ha−1 and triple super phosphate equivalent to 50 kg P ha−1;
  • 3
     P-plot: plant residue treated plot (maize straw and leaf equivalent to 2.5 Mg C ha−1 and 40 kg N ha−1;
  • 4
     PF-plot: plant residue–chemical fertilizer treated plot;
  • 5
    B-plot: bare plot (without plants; the soil surface was kept bare of all plants, including weeds, during the experimental period).

Each experimental plot (8 m × 8 m) was laid down in a randomized block design using three replicate plots per treatment; an unplanted strip of >1 m separated each block.

In the P and PF plots, plant residue was applied at the end of July 2005, that is, 8 months prior to the experiment, as follows: after the harvest of maize, maize straw and leaf (6 Mg dry matter ha−1 equivalent to ∼2.5 Mg C ha−1) were chopped into 10-cm pieces and incorporated into the soil (to a depth of 15 cm) using hand hoes. The plant residue application did not significantly affect the soil characteristics (i.e. total organic C and N and bulk density in March 2006; data not shown).

According to fertilizer recommendations in Tanzania, chemical fertilizer in the F and PF plots was applied as follows: urea was broadcast in two applications (35 kg N ha−1 at 7 days after planting [DAP] and 65 kg N ha−1 at 35 DAP), and triple super phosphate was broadcast at the time of seeding (0 DAP).

Two maize (Zea maize L. var. Staha) seeds were planted per hole at a spacing of 80 cm × 30 cm on 16 March 2006 (i.e. 0 DAP) and were thinned to one plant per hole at 14 DAP. We weeded all treatment plots on 15 and 36 DAP; we also weeded only the B plot on 63 DAP to completely remove any grass. All weeded materials were removed from the plots. We harvested the maize at 104 DAP. No irrigation was applied and the experiment was maintained under rain-fed cultivation.

Environmental factors

Two replicate measurements of air temperature, soil temperature at a depth of 5 cm and volumetric moisture content in the surface soil (0–15 cm) were taken hourly from each C and PF plot using a datalogger system (107 thermistor probes for temperature and CS616 for soil volumetric moisture were connected to a CR-10X datalogger; Campbell Scientific (Logan, UT, USA). Rainfall was also monitored hourly at the experimental site using the same CR10X datalogger system and a TE525MM device (Campbell Scientific).

Soil sampling and analyses

Soil samples were collected seven times at 0, 14, 43, 61, 76, 92 and 104 DAP. For each sample, six soil cores (2-cm diameter × 15-cm depth) were taken within 10 cm of the plant, which was randomly selected inside the plot (7 m × 6 m; avoiding the edge of the plot). Each core was divided into two segments, 0–5 cm and 5–15 cm, and the six respective subsamples were composited and mixed per replication. The soil samples were immediately transported to the laboratory in a cooler at 4°C, sieved through a 4-mm mesh screen after removing visible plant debris and stored under field-moist conditions at 4°C. The samples were taken for determination of gravimetric soil moisture, inorganic N (NH4+ and NO3), MBC and MBN. All measurements, except for soil moisture, were conducted after transport to Japan within 3 months of sampling. As Mueller et al. (1996) reported that an 11-month period of storage at 4°C did not affect the MBC, our 3-month storage in the dark at 4°C was not expected to significantly change the MBC or the MBN.

After sieving, the soil samples (10.0 g) were weighed in aluminum dishes. These dishes were placed in a 105°C oven for 48 h and the dry weight was then recorded. The gravimetric soil moisture was the difference in soil weight before and after oven drying. Inorganic N was extracted from 10.0 g soil (dry base) with 30.0 mL of 1 mol L−1 KCl for 30 min on an orbital shaker, and the suspension was centrifuged (2000 g) and filtered through filter paper (Advantec No. 5C, Tokyo, Japan). The NH4+ in the extract was determined using the modified indophenol blue method (Rhine et al. 1998) and the NO3 in the extract was determined using the modified Greisess–Ilovay method (Mulvaney 1996). The MBC and MBN were measured using the fumigation–extraction method (Brookes et al. 1985; Vance et al. 1987). In brief, soil samples (8.0 g dry base) were fumigated with ethanol-free CHCl3 for 24 h at 25°C. After removal of the CHCl3, the soluble C and N were extracted from the fumigated and non-fumigated samples with 32.0 mL of 0.5 mol L−1 K2SO4 for 30 min on an orbital shaker. The total organic C and extractable N in the filtered extract were determined using a TOC-N Auto-analyzer (TOC-V carbon analyzer with an IN unit; Shimadzu, Kyoto, Japan). Microbial C flush (the difference between extractable C from fumigated and non-fumigated samples) was converted to MBC using a KEC factor of 0.45 (Vance et al. 1987). Microbial N flush was also converted to MBN using a KEN factor of 0.54 (Brookes et al. 1985). All measurements were done in triplicate.

Microbial respiration as microbial activity in situ

Microbial activity is closely linked to soil organic matter (SOM) decomposition, which results in N mineralization or immobilization; microbial respiration is one of several microbial activity indices. To evaluate the importance of soil microbes as decomposers of SOM during the plant growth period in this experiment, we measured soil respiration in all plots in the field using a closed-chamber system (Funakawa et al. 2006). As soil respiration consists of plant root respiration and microbial respiration, we excluded plant root respiration using the trenching method. Polyvinyl chloride (PVC) cylinders (diameter 13 cm, height 30 cm) were inserted into the soil to a depth of 15 cm and the enclosed soil was later covered with a fine mesh to support the soil in the core sample. The CO2 efflux rate was measured on 3, 10, 19, 26, 40, 54, 75, 89 and 102 DAP. For each measurement, to block the CO2 evolved from plant root respiration, we removed the PVC cylinder and covered the bottom of the cylinder with a plastic sheet, and then we placed the cylinder back into the hole again. We sampled the gases in the headspace of the PVC cylinder at 0 and 40 min after the top of the cylinder was covered with a plastic sheet, and the CO2 efflux rate was calculated based on the increase of CO2 concentration in the cylinder after 40 min. Gas samples were analyzed with an infrared CO2 controller (ZFP9AA11; Fuji Electric, Tokyo, Japan). Five replicate measurements were taken at each plot.

Plant sampling and analysis

Sampling of above-ground plants was conducted on 14, 27, 43, 61, 76, 92 and 104 DAP. Six plant samples were collected from 5 m × 0.3 m areas, which were chosen randomly in the respective plots. The plant material was divided into leaf, stem, cobs and grain, and dried in a glass greenhouse for 1 week and then in an 80°C oven for 48 h, and the dry matter was weighed. Each sample was ground and the N content was measured using a dry combustion method with an NC analyzer (Vario Max CHN; Elementar, Hanau, Germany). The data are expressed on an area basis (kg N ha−1) by multiplying the percentage of N by the plant biomass per unit area. To estimate crop yields, another 30 cobs of maize per replicate were collected and weighed on the last sampling day (104 DAP).

Statistical analyses

All statistical analyses were carried out using SYSTAT 11 (SYSTAT Software, Point Richmond, CA, USA). All data are expressed on a dry-weight basis. One-way ANOVAs were used to detect any significant differences in the variables (MBN, microbial biomass C : N ratio (MB C : N ratio), microbial respiration (CO2 efflux rate), NH4+, NO3, plant N and crop yields) between the sampling times in each treatment plot and soil layer and between the land management treatments (C, P, F, PF and B plots) in each sampling time and soil layer. When an ANOVA indicated a significant difference, mean comparisons were carried out using post-hoc Tukey’s Kramer multiple comparison tests. A repeated measures analysis of variance (MANOVA) was also used to detect any significant differences in variables between the soil layers in each treatment plot during the experimental period and between the land management treatments (crop cultivation, plant residue and/or fertilizer application) in each soil layer during the experimental period. A Pearson’s correlation coefficient test was conducted to investigate the relationships between the soil microbial factors (MBN, MB C : N ratio and microbial respiration) and environmental factors (soil moisture content, air and soil temperature). In all cases, P < 0.05 was considered to be significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Environmental factors

The total rainfall was 607 mm during the 104-day experimental period in 2006 (Fig. 1a). Approximately 85% of the total rainfall occurred before 60 DAP. As the historical average rainfall for this area of Tanzania over this period is ∼400–500 mm, the rainfall amount in 2006 was higher than the average. The volumetric soil moisture content was continuously high, at approximately −10 kPa, from 5 to 60 DAP, but it decreased steadily after 60 DAP and reached −1.5 MPa at 85 DAP (Fig. 1a). From the rainfall and soil moisture data, the experimental period could be divided into a wet period, from 0 to 60 DAP, and a relatively dry period, from 61 to 104 DAP. On 86–88 DAP, intense rainfall (56 mm in total) temporarily increased the soil moisture content from −1.5 MPa to −10 kPa, but the soil dried out rapidly thereafter and remained dry for most of the remaining experimental period. There was no clear difference in soil moisture contents between the plant residue application plots and the non-plant residue application plots from July 2005 to June 2006 (data not shown).

image

Figure 1.  Fluctuations in (a) volumetric moisture content (VMC) in the C plot as a function of daily rainfall and (b) daily average air temperature (AT) and soil (5-cm depth) temperature (ST) at the C and PF plots. The horizontal dotted lines in Fig. 1a indicate the value of −10 kPa (above) and −1.5 MPa (below).

Download figure to PowerPoint

The average air temperature during the experimental period was 23.3°C; the average daily air temperature gradually decreased throughout the experiment from 27.9 to 18.9°C (Fig. 1b). The average soil temperatures in the C and PF plots were 26.6 and 26.0°C, respectively. These values also gradually decreased with the air temperature over the course of the experiment.

Temporal dynamics of MBN and the MB C : N ratio in the surface and subsurface soils

Table 1 presents the variations in MBN, and in the MB C : N ratio in the C, P, F, PF and B plots. All soil microbial factors (MBN and MB C : N ratio) fluctuated significantly during the experimental period in all treatment plots and layers (P < 0.05), except for the MB C : N ratio in the B plot. The MBN in all plots and layers remained constant until 43 DAP, and then the MBN values in the surface layer increased on 61 DAP. After 61 DAP the MBN values decreased drastically until 92 DAP, particularly in the cultivated plots. The MBN in all plots finally increased again at 104 DAP. The MB C : N ratio gradually decreased until 61 DAP in all plots and layers. After 61 DAP, the MB C : N ratio significantly increased at 76 and 92 DAP in all cultivated plots and layers, but not in the B plots.

Table 1.   Microbial biomass N (MBN), microbial biomass C : N ratio (MB C : N ratio), NH4+ and NO3 at each treatment plot in each soil layer
DAPSurface soil (0–5 cm)Subsurface soil (5–15 cm)
C plotP plotF plotPF plotB plotC plotP plotF plotPF plotB plot
  1. Values followed horizontally by a different uppercase letter (A and B) indicate that the means are significantly different (P < 0.05) between the treatment plots (C, P, F, PF and B plots) within each sampling time; different lowercase letters (a– e) vertically indicate that the means are significantly different (P < 0.05) among sampling time within a plot in each soil layer. DAP, days after planting; C plot, Control (no input); F plot, chemical fertilizer treated plot (urea equivalent to 100 kg N ha−1 and triple super phosphate equivalent to 50 kg P ha−1; P plot, plant residue treated plot (maize straw and leaf equivalent to 2.5 Mg C ha−1 and 40 kg N ha−1; PF plot, plant residue–chemical fertilizer treated plot; B plot, bare plot (without plants; the soil surface was kept bare of all plants, including weeds, during the experimental period).

MBN (mg N kg−1 soil)
032.0 b32.3 ab26.2 bc29.9 ab31.1 b21.2 bc25.9 ab20.0 bc26.8 ab18.0 ab
1434.3 b30.8 ab33.4 ab41.7 a35.8 a26.8 bc24.3 ab33.3 ab31.7 ab20.6 ab
4330.4 b26.7 ab24.4 bc26.9 b25.9 b32.5 ab30.0 ab25.6 bc27.4 ab25.3 ab
6147.8 a38.2 a43.4 a38.3 ab36.3 a32.0 abAB31.3 aAB33.1 abAB40.5 aA21.3 abB
7622.5 b23.3 ab24.5 bc24.4 b31.2 b15.7 c15.4 b12.4 c15.6 b11.9 b
9224.7 bA15.1 bAB16.4 cAB10.2 cA21.1 bAB17.7 bcAB15.9 bAB23.4 bcA10.4 bB17.9 abA
10432.3 b29.2 ab32.3 ab28.0 ab30.5 ab41.2 a37.8 a40.4 a42.8 a26.8 a
MB C : N ratio
08.2 ab8.0 abc10.8 ab10.2 b10.810.2 ab8.0 bc11.1 ab9.0 ab10.4
1410.2 a10.5 ab10.0 abc7.7 b9.610.1 ab10.3 ab7.2 ab8.7 ab10.0
437.4 ab6.0 bc9.1 abc8.1 b7.25.0 b5.8 bc7.0 ab6.2 b5.6
615.6 b5.6 c6.2 c5.9 b6.86.2 b4.9 c5.8 b6.3 b8.2
7610.4 a9.3 abc9.9 abc8.6 b8.110.7 abAB8.7 bcAB11.5 aAB13.9 abA7.4 B
928.9 abB12.6 aAB12.9 aAB18.1 aA8.5 B13.0 aAB14.2 aAB7.4 abAB15.7 aA7.0 B
1048.1 ab7.1 bc8.6 bc9.1 b8.65.6 b6.1 bc6.2 ab6.8 b6.9
NH4+ (mg N kg−1 soil)
01.0 cB4.1 bA2.4 cAB1.9 cAB3.2 bAB1.0 b4.1 a2.4 b1.9 c3.2
142.0 bcB3.4 bB62.5 aA73.3 aA3.7 bB3.3 b6.3 ab6.3 b3.0 bc3.3
4310.9 a14.0 a25.1 b29.7 b8.6 a6.0 a7.6 ab7.8 a6.6 a5.3
613.7 bcA3.1 bAB2.6 cAB4.0 cA1.8 bB3.0 b3.6 ab2.7 b2.5 bc3.5
764.9 b1.9 b2.6 c3.0 c3.6 b3.5 b5.7 ab4.2 b3.2 bc4.1
924.2 bc5.2 b4.6 c4.9 c4.3 b3.6 b4.8 b4.5 b3.4 bc3.7
1044.4 bc3.4 b3.4 c3.4 c3.8 b3.0 b4.7 b3.1 b3.9 b3.3
NO3 (mg N kg−1 soil)
040.7 aAB49.0 aAB36.7 bB58.3 aA39.5 aB40.7 aAB49.0 aAB36.7 aB58.3 aA39.5 aB
1427.2 abB22.4 bB57.9 aA76.6 aA32.3 aB31.5 aB33.1 bB32.3 abB51.0 aA32.6 aB
432.1 cAB0.7 eB12 dAB6.0 cA12 cAB1.0 c6.1 c12.6 c9.3 c8.4 b
6120.1 b17.6 bc26.9 bc24.9 b18.1 b18.6 b14.7 c18.6 abc23.4 b16.5 b
7613.3 bcAB8.5 dB22.7 bcA19.2 bcA14.2 bAB13.0 b8.5 c17.9 abc17.0 b16.0 b
9219.9 bA16.2 cAB19.9 bcdA21.1 bcA14.1 bB13.3 b10.5 c14.8 bc14.8 b12.6 b
1048.0 cAB6.2 deB11.7 cdA10.6 bcA82 bcAB8.4 bcAB6.8 cB8.9 cAB10.6 bcA7.9 bAB

In the C, F and B plots, the soil layer had a significant effect on MBN during the experimental period, according to the MANOVA analysis (data not shown); the MBN of the surface soil was higher than that for the corresponding subsurface soil. In the P and PF plots, however, the soil layer did not have a significant effect on MBN. The MB C : N ratio did not differ between the soil layers in any of the treatment plots.

Crop cultivation affected variation in the MBN, but not in the MB C : N ratio, according to the MANOVA analysis (Table 2). In contrast, plant residue and/or fertilizer application did not clearly affect the variations in any soil microbial factors in any soil layer during the experimental period. Notably, the application of N fertilizer at 7 and 35 DAP did not increase MBN significantly as measured at 14 and 43 DAP in the F and PF plots (Table 1).

Table 2.   Summary of the MANOVA analysis for soil microbial factors, NH4+ and NO3 of crop cultivation, plant residue application and fertilizer application in each soil layer
SourceMBCMBNMB C : N ratioCO2 effluxNH4+NO3
0–5 cm5–15 cm0–5 cm5–15 cm0–5 cm5–15 cm0–5 cm5–15 cm0–5 cm5–15 cm
  1. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. MBC, microbial biomass carbon; MBN, microbial biomass nitrogen; MB C : N ratio, microbial biomass C : N ratio; PF, plant residue and fertilizer application.

Crop cultivation*ns**nsnsnsnsnsnsns
Sampling time (Time)*********************************
Time × crop cultivationnsns**nsnsnsnsnsnsnsns
Plant residue application (P)nsnsnsnsnsnsnsnsnsnsns
Time*********************************
Time × Pnsnsnsns**ns**ns**ns
Fertilizer application (F)nsnsnsnsnsnsns**ns*ns
Time*********************************
Time × Fnsnsnsns*nsns**ns**ns
PF**ns*nsns*ns**ns**
Time*********************************
Time × PFnsns*ns***ns**ns***

There was no clear relationship between soil microbial factors (MBN and MB C : N ratio) and environmental factors (soil moisture content and air and soil temperature) in any of the treatment plots or layers, according to correlation analysis (data not shown).

Temporal dynamics in microbial activity in situ

Microbial respiration, that is, the CO2 efflux rate, fluctuated substantially from 24.6 to 199.1 mg C m−2 h−1 over the experimental period (Fig. 2). The average CO2 efflux rate in the C, P, F, PF and B plots was 71.5, 73.6, 80.1, 76 and 61.5 mg C m−2 h−1, respectively. None of the treatments (crop cultivation, plant residue application or fertilizer application) significantly affected CO2 efflux; however, sampling time strongly influenced microbial respiration (Table 2). The CO2 efflux rates in all plots were mostly high up to 60 DAP, and gradually decreased after 60 DAP. For all plots, the CO2 efflux rate correlated significantly with soil moisture content (P < 0.01), but not with air or soil temperature (data not shown). These CO2 efflux data indicated that there was rapid decomposition of SOM during the wet period, but this was not the case during the dry period.

image

Figure 2.  Fluctuations in the CO2 efflux rate in all plots. The two upward-pointing arrows indicate urea application at 7 and 35 DAP. Error bars indicate the standard error.

Download figure to PowerPoint

Dynamics of inorganic N (NH4+ and NO3) in the surface and subsurface soils

Table 1 presents the variations in NH4+ and NO3 in the C, P, F, PF and B plots. In the C, P and B plots, the amount of NH4+ was relatively small and fluctuated little compared with the amount of NO3 and its fluctuation. The amount of NO3 in all plots and layers was large at 0 DAP. Because plant N uptake was quite small throughout the dry and short rainy season prior to our experiment, NO3 was expected to have accumulated at 0 DAP. At 43 DAP, the NO3 in all plots was essentially depleted. Even in the F and PF plots, NO3 was depleted at 43 DAP, despite the second fertilizer application (65 kg N ha−1) at 35 DAP, whereas NH4+ in the surface soil was higher than in the non-fertilized plots at 43 DAP. By 61 DAP, however, NO3 had increased significantly in all plots, and inorganic N (NH4+ + NO3) was depleted throughout the remaining experimental period (∼20–35 kg N ha−1 : 0–15 cm).

Plant residue application did not affect NH4+ and NO3 variations in this experiment (Table 2). In contrast, the application of fertilizer affected NH4+ and NO3 variations only in the surface soil; even the interactions between sampling time and fertilizer, and between sampling time and plant residue and fertilizer application, were significant in the surface soil. In general, the effect of the fertilizer application had disappeared by 43 DAP because there was little difference in NH4+ and NO3 between the treatments after 61 DAP (Table 1).

Plant N uptake and yield

The total plant N uptake in the C, P, F and PF plots was 106.4, 104.2, 118.3 and 136.2 kg N ha−1, respectively. Crop yields in the C, P, F and PF plots were 2.7, 2.6, 3.1 and 3.4 t ha−1, respectively. Plant N uptake in the PF plot and crop yields in the F and PF plots were significantly higher than in the C and P plots. Although plant N uptake at 27 DAP was less than 9% of the plant N uptake at harvest time, it increased substantially after 43 DAP (Fig. 3). Plant N uptake in the P and PF plots increased until 92 DAP, whereas plant N uptake in the C and F plots ceased at 76 DAP.

image

Figure 3.  Above-ground plant N uptake in all planted plots. Error bars indicate the standard error.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Soil microbes as decomposers: significance to plant growth during the wet period

It is likely that the SOM underwent substantial decomposition during the wet period (0–60 DAP), when heavy rainfall maintained high soil moisture content (−10 kPa). Shahandeh et al. (2004) showed that rapid N mineralization occurred in the early rainy season in Mali because easily decomposable SOM, which has a low C : N ratio, accumulates during the dry season. Singh et al. (2007a) also observed a high rate of N mineralization during the rainy season in India. In the current study, the cumulative soil respiration, which was estimated by the relationship between the CO2 efflux rate and the soil moisture content (according to Funakawa et al. 2006), was 1.42 Mg C ha−1 in the C plot, and more than 70% of the cumulative respiration was respired during the wet period. Thus, a substantial amount of N was presumably mineralized during the wet period.

In contrast, during the dry period, the CO2 efflux rate was consistently low. Because the CO2 efflux rate correlated strongly with soil moisture content in all plots in our study, soil moisture primarily limited microbial activity during the dry period. Chen et al. (2002) also observed a strong correlation between the CO2 efflux rate and soil moisture in the tropical savanna of northern Australia. Furthermore, depletion of easily decomposable SOM appeared to limit microbial respiration in this period because the CO2 efflux rate at 89 DAP was quite small, despite episodic heavy rainfall (56 mm) from 86 to 88 DAP. Garcia-Oliva et al. (2003) observed that easily decomposable SOM substantially decreased during the rainy season in Mexico. These results suggest that soil microbes contribute to plant growth mainly as decomposers during the wet period as a result of the high soil moisture content and easily decomposable SOM during this period.

Considering the substantial N mineralization during the wet period and the heavy rainfall at 41–43 DAP (91 mm in total), the depletion of NO3 at 43 DAP indicated severe N leaching beyond a soil depth of 15 cm in all plots. Chikowo et al. (2004) also reported substantial NO3 leaching as heavy rainfall (104 mm within 2 weeks) added to the already fully water-charged soil profile in Zimbabwe.

Competition for N between soil microbes and plants during the dry period (the grain-forming stage)

We found a clear effect of plant N uptake on soil microbial dynamics in the dry period (61–104 DAP) by comparing variations in MBN and in the MB C : N ratio between the cultivated plots (C, P, F and PF plots) and the non-cultivated plot (B plot). After 61 DAP (i.e. the grain-forming stage), plants continued to take up substantial amounts of N until 76 DAP in the C and F plots and until 92 DAP in the P and PF plots (Fig. 3). During this stage (61–92 DAP), we observed a distinct decrease in MBN (from 63–71 to 18–33 kg N ha−1) in all cultivated plots, but not in the B plot (Fig. 4); this MBN decrease was roughly consistent with the plant N uptake in each plot. Many studies have also reported MBN depletion during the crop growth period and have suggested that plants and soil microbes compete for N (Friedel et al. 2001; Kushwaha et al. 2000; Singh et al. 2007b). Singh et al. (2007b) found a negative correlation between crop (i.e. rice and barley) root biomass and MBN in India in control, fertilizer and organic matter application plots; crop root biomass increased from the seedling stage to reach a maximum at the grain-forming stage and declined sharply thereafter, whereas MBN decreased from the seedling stage to the grain-forming stage and increased again at the harvest stage. In our study, MBN continuously decreased until plant N uptake stopped in each plot (76 DAP in the C and F plots and 92 DAP in the P and PF plots), and MBN increased again thereafter (Fig. 4). This re-increase of MBN at 104 DAP was possibly caused by maize root deposition at the harvest stage (Singh et al. 2007b). In addition, we observed a significant increase in the MB C : N ratio at 76 and 92 DAP in all cultivated plots and layers, but not in the B plot. An increase in the MB C : N ratio (>10) indicates that N availability is greatly limited in soils (Joergensen and Emmerling 2006). The increased MB C : N ratio in all cultivated plots and layers decreased to a value similar to that of the B plot at 104 DAP, when plant N uptake had ceased, whereas the ratio remained steady in the B plot throughout the dry period (Table 1). These results also confirmed that plant N uptake caused severe competition for N with soil microbes within a soil depth of 0–15 cm during the grain-forming stage. Similar fluctuations in the MB C : N ratio during the crop growth period were also observed at a soil depth of 0–27 cm by Friedel et al. (2001) in Germany, and also at a soil depth of 0–20 cm by Petersen et al. (2003) in Denmark. Taking into consideration the drying soil and the depletion of easily decomposable SOM, microbial decomposition could supply only a small amount of mineralized N in the dry period. Therefore, the N pool derived from inorganic N and MBN must have accounted for the plant N uptake after 61 DAP, although we should also consider the contribution of soil layers below 15 cm to plant growth. On the basis of the above time-course of MBN, the MB C : N ratio and plant N uptake, we conclude that soil microbes contributed to plant growth mainly as an N source during the dry period (i.e. the grain-forming stage) in all cultivated plots.

image

Figure 4.  Above-ground plant N uptake and below-ground N pool (microbial biomass N [MBN] and inorganic N [Inorg N] in soil) in each plot throughout the experimental period. Error bars indicate the standard error. Different letters (a– e) indicate that the means of plant N are significantly different (P < 0.05) among the sampling times within a plot. All values are calculated for area base by multiplying the soil depth (0–15 cm) by the bulk density (1.21 g cm−3).

Download figure to PowerPoint

Effect of plant residue and fertilizer application on the relationship between the soil N pool and plant N uptake

Plant residue application did not have a clear effect on the soil microbial factors (Table 2) or environmental factors (i.e. soil moisture and temperature) in either soil layer. Ouedraogo et al. (2004) observed that soil macrofauna decomposed more than 90% of the added maize straw within 3 months during the rainy season in Burkina Faso. Considering the effect of the short rainy season at our Tanzanian site (the rainfall amount was approximately 170 mm from October to December 2005), the plant residues we applied in July 2005 may have been largely decomposed by soil macrofauna before our cultivation experiment was initiated in March 2006. In addition, the application of fertilizer had little effect on soil microbial factors (Table 2), although it significantly increased plant growth and crop yields. Because of severe N leaching during the wet period, fertilizer application affected only inorganic N in the surface soil, and its effect disappeared rapidly (∼1 month).

However, we observed different soil–plant N dynamics after 61 DAP that were presumably caused by plant residue application. Plant N uptake in the P and PF plots continued until 92 DAP, whereas plant N uptake in the C and F plots ceased at 76 DAP (Fig. 3). As a result, plant N uptake after 61 DAP was greater in the P and PF plots (43.7 and 55.2 kg N ha−1, respectively) than in the C and F plots (30.4 and 26.6 kg N ha−1, respectively). In addition, the decrease in MBN and the high MB C : N ratio in the P and PF plots in both layers continued until 92 DAP, whereas these changes were mostly finished by 76 DAP in the C and F plots (Table 1, Fig. 4). Because total plant N uptake in the C and P plots was similar, plant residue application may lead to a delay in plant N uptake, and may prolong the period of plant N uptake. The plant N uptake rate from 43 to 61 DAP was lower in the P and PF plots than in the C and F plots. Over this period, inorganic N should be produced because of high soil moisture content, N leaching should not occur owing to little rainfall, and plant N demand should be high in all cultivated plots. These results indicated that plant residue application may increase the N immobilization by soil microbes from 43 to 61 DAP and cause a deficiency in N in the soil, resulting in a lower rate and a longer period of N uptake by plants in the P and PF plots. However, further study is necessary to support this hypothesis.

Many studies have indicated that reducing NO3 leaching is the greatest challenge to improving the efficiency of N use in dry tropical croplands (Chikowo et al. 2006; Hartemink et al. 2000). If we can use MBN as a temporal N sink during the rainy season by applying organic matter as observed by Herai et al. (2006), we should be able to decrease the severe N leaching and improve the efficiency of N use in the soil (Mtambanengwe and Mapfumo 2006). Because the total plant N uptake in the PF plot was greater than that in the F plot, plant residue application with chemical fertilizer appeared to improve the efficiency of N use in the soil compared with chemical fertilizer alone. As we could not observe a clear effect of plant residue application on the dynamics of MBN and inorganic N, future studies should focus on the effect of plant residue and/or fertilizer application on temporal MBN variations in relation to N loss from soil and plant N uptake.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank Professor J. J. T. Msaky and the staff of the Sokoine University of Agriculture for their kind technical support in the field and with the laboratory experiments in Tanzania. Our work was financially supported by the Kyoto University Foundation and by a Research Fellowship from the Japan Society for the Promotion of Science of Young Scientists.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • Bardgett RD, Streeter TC, Cole L, Hartley IR 2002: Linkages between soil biota, nitrogen availability, and plant nitrogen uptake in a mountain ecosystem in the Scottish Highlands. Appl. Soil Ecol., 19, 121134.
  • Bardgett RD, Van Der Wal R, Jonsdottir IS, Quirk H, Dutton S 2007: Temporal variability in plant and soil nitrogen pools in a high-Arctic ecosystem. Soil Biol. Biochem., 39, 21292137.
  • Brookes PC, Landman A, Pruden G, Jenkinson DS 1985: Chloroform fumigation and release of soil N: a rapid direct extraction method to measure microbial biomass N in soil. Soil Biol. Biochem., 17, 837842.
  • Chen X, Eamus D, Hutley LB 2002: Seasonal patterns of soil carbon dioxide efflux from a wet-dry tropical savanna of northern Australia. Aust. J. Bot., 50, 4351.
  • Chikowo R, Mapfumo P, Leffelaar PA, Giller KE 2006: Integrating legumes to improve N cycling on smallholder farms in sub-humid Zimbabwe: resource quality, biophysical and environmental limitations. Nutr. Cycl. Agroecosys., 76, 219231.
  • Chikowo R, Mapfumo P, Nyamugafata P, Giller KE 2004: Mineral N dynamics, leaching and nitrous oxide losses under maize following two-year improved fallows on a sandy loam soil in Zimbabwe. Plant Soil, 295, 315330.
  • Cookson WR, Marschner P, Clark IM et al. 2006: The influence of season, agricultural management, and soil properties on gross nitrogen transformations and bacterial community structure. Aust. J. Soil Res., 44, 453465.
  • Fierer N, Schimel JP, Holden PA 2003: Influence of drying-rewetting frequency on soil bacterial community structure. Microb. Ecol., 45, 6371.
  • Friedel JK, Gabel D, Stahr K 2001: Nitrogen pools and turnover in arable soils under different durations of organic farming: II: source-and-sink function of the soil microbial biomass or competition with growing plants? J. Plant Nutr. Soil Sci., 164, 421429.
  • Funakawa S, Yanai J, Hayashi Y et al. 2006: Soil organic matter dynamics in a sloped sandy cropland of Northeast Thailand with special reference to the spatial distribution of soil properties. Jpn. J. Trop. Agric., 50 (4), 199207.
  • Garcia-Oliva F, Sveshtarova B, Oliva M 2003: Seasonal effects on soil organic carbon dynamics in a tropical deciduous forest ecosystem in western Mexico. J. Trop. Ecol., 19, 179188.
  • Ghoshal N, Sihgh KP 1995: Effect of farmyard manure and inorganic fertilizer on the dynamics of soil microbial biomass in a tropical dryland agroecosystem. Biol. Fertil. Soils, 19, 231238.
  • Hartemink AE, Buresh RJ, Van Bodegom PM, Braun AR, Jama B, Janssen BH 2000: Inorganic nitrogen dynamics in fallows and maize on an Oxisol and Alfisol in the highlands of Kenya. Geoderma, 98, 1133.
  • Herai Y, Kouno K, Hashimoto M, Nagaoka T 2006: Relationships between microbial biomass nitrogen, nitrate leaching and nitrogen uptake by corn in a compost and chemical fertilizer-amended regosol. Soil Sci. Plant Nutr., 52, 186194.
  • Joergensen RG, Emmerling C 2006: Methods for evaluating human impact on soil microorganisms based on their activity, biomass, and diversity in agricultural soils. J. Plant Nutr. Soil Sci., 169, 295309.
  • Kushwaha CP, Tripathi SK, Singh KP 2000: Variations in soil microbial biomass and N availability due to residue and tillage management in a dryland rice agroecosystem. Soil Till. Res., 56, 153166.
  • Michelsen A, Anderson M, Jensen M, Kjoller A, Gashew M 2004: Carbon stocks, soil respiration and microbial biomass in fine-prone tropical grassland, woodland and forest ecosystems. Soil Biol. Biochem., 36, 17071717.
  • Mtambanengwe F, Mapfumo P 2006: Effects of organic resource quality on soil profile N dynamics and maize yields on sandy soils in Zimbabwe. Plant Soil, 281, 173191.
  • Mueller T, Lavahun MFE, Joergensen RG, Meyer B 1996: The problem of pretreatment and unintentional variations in the fumigation-extraction method for time-course measurements in the field. Biol. Fertil. Soils, 22, 167170.
  • Mulvaney RL 1996: Nitrogen-Inorganic forms. In Methods of Soil Analysis Ed DLSparks., pp. 11231184, Soil Science Society of America, Madison.
  • Ouedraogo E, Mando A, Brussaard L 2004: Soil macrofaunal-mediated organic resource disappearance in semi-arid West Africa. Appl. Soil Ecol., 27, 259267.
  • Petersen SO, Petersen J, Rubæk GH 2003: Dynamics and plant uptake of nitrogen and phosphorus in soil amended with sewage sludge. Appl. Soil Ecol., 24, 187195.
  • Powlson DS, Hirsch PR, Brookes PC 2001: The role of soil microorganisms in soil organic matter conservation in the tropics. Nutr. Cycl. Agroecosys., 61, 4151.
  • Rhine ED, Sims GK, Mulvaney RL, Pratt EJ 1998: Improving the Berthelot reaction for determining ammonium in soil extracts and water. Soil Sci. Soc. Am., 62(2), 473480.
  • Shahandeh H, Blanton-Knewtson SJ, Doumbia M, Hons FM, Hossner LR 2004: Nitrogen dynamics in tropical soils of Mali, West Africa. Biol. Fertil. Soils, 39, 258268.
  • Singh S, Ghoshal N, Singh KP 2007a: Synchronizing nitrogen availability through application of organic inputs of varying resource quality in a tropical dryland agroecosystem. Appl. Soil Ecol., 36, 164175.
  • Singh S, Ghoshal N, Singh KP 2007b: Variation in soil microbial biomass and crop roots due to differing resource quality inputs in a tropical dryland agroecosystem. Soil Biol. Biochem., 39, 7686.
  • Singh JS, Raghubanshi AS, Singh RS, Srivastava SC 1989: Microbial biomass act as a source of plant nutrients in dry tropical forest and savanna. Nature, 338, 499500.
  • Soil Survey Staff 2006: Keys to Soil Taxonomy, 10th edn. United States Department of Agriculture Natural Resources Conservation Service, Washington.
  • Spedding TA, Hamel C, Mehuys GR, Madramootoo CA 2004: Soil microbial dynamics in maize-growing soil under different tillage and residue management systems. Soil Biol. Biochem., 36, 499512.
  • Srivastava SC, Lal JP 1994: Effect of crop growth and soil treatments on microbial C, N, and P in dry tropical arable land. Biol. Fertil. Soils, 17, 108114.
  • Sugihara S, Funakawa S, Kilasara M, Kosaki T 2010: Effect of land management and soil texture on seasonal variations in soil microbial biomass in dry tropical agroecosystems in Tanzania. Appl. Soil Ecol., 44, 8088.
  • Tripathi N, Singh RS 2007: Cultivation impacts nitrogen transformation in Indian forest ecosystems. Nutr. Cycl. Agroecosys., 77, 233243.
  • Vance ED, Brookes PC, Jenkinson DS 1987: An extraction method for measuring soil microbial biomass carbon. Soil Biol. Biochem., 19, 703707.
  • Wardle DA 1992: A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soils. Biol. Rev., 67, 321358.
  • Wardle DA 1998: Controls of temporal variability of the soil microbial biomass: a global-scale synthesis. Soil Biol. Biochem., 30, 16271637.
  • Wardle DA, Bardgett RD, Klironomos JN, Setala H, Van Der Putten W, Wall DH 2004: Ecological linkages between aboveground and belowground biota. Science, 304, 16291633.