Dry matter production and digestibility of Centrosema pubescens and
Pueraria phaseoloides with rock phosphate fertilization and mycorrhizae
inoculation in Latosolic soil

Lukiwati, D.R.¹

Keywords: Centrosema pubescens, Pueraria phaseoloides, rock phosphate, mycorrhizae

Introduction

Centro (Centrosema pubescens) and puero (Pueraria phaseoloides) are important forage legumes as protein and mineral sources for ruminant livestock in the tropics. Centro and puero have also been used as cover crops in forest plantations or in agroforestry systems. However, most land used for forage production in Indonesia is characterized by a low phosphorus (P) content and a low soil pH (latosolic soil). The application of P fertilizer during the periods of active growth increases forage legumes production and quality (Coates et al.,1990).

Superphosphate (SP) fertilizer has been widely used to improve agricultural production. However, its high cost makes the use of rock phosphate (RP) attractive. Rock phosphate is a slow release source of phosphorus, thus the inoculation with vesicular-arbuscular mycorrhizae (VAM) fungus is a promising technique to increase P bio-availability. According to Jones (1990), the response obtained to applications of P fertilizers is a function of many factors i.e. the initial availability of soil P, the form of fertilizer applied, and the presence or absence of effective mycorrhizae in the soil.

Most research on VAM inoculation has been done on forest-trees and agricultural crops, but rarely on forage crops. The VAM fungus of Glomus mosseae, for instance, is the most common species associated with agricultural crops or forests (Chen et al., 1998). Centro and puero are suitable host plants for VA M fungi culture (Lukiwati & Supriyanto, 1995). Two species of VAM (Glomus fasciculatum, Entrophosphora columbiana) proved to increase dry matter production and nutrient uptake of Pueraria phaseoloides similarly (Lukiwati & Simanungkalit, 2004). The results showed that SP can be replaced by RP when combined with VAM fungi inoculation (Lukiwati & Simanungkalit, 2001). Success of VA M inoculation in the field affected by effectiveness of indigenous-VAM fungi or depending upon VAM inoculum potential (Mitiku-Habte & Fox, 1993). The symbiosis between VAM fungi and legumes has been less studied in unsterilized soils than in sterilized soils. The objective of this work was to investigate under field conditions the effect of RP fertilization, VAM (Glomus sp) inoculation, and their combination on dry matter production and digestibility of centro and puero in a latosolic soil low in available P.

Materials and Methods

A completely randomized field experiment with three blocks was conducted for 7 months on an acid (pH (H₂O) 5.1 to 5.3) latosolic soil low in available P (Bray II extractable between 4.0 and 5.7 g/kg^-1). The experiment was conducted on 4 × 5 m plots.

The design consisted in three factors as follows 1) legume species (centro, puero), 2) VA M inoculation (with, and without VAM inoculation), and 3) five levels of rock phosphate fertilization (0, 44, 87, 131, and 175 kg P ha^-1). The periods of defoliation were used as sub factors. The defoliation of the plants was done three times. Standard fertilizers, i.e. 50 kg N ha^-1 as urea and 83 kg K ha^-1 as KCl, were applied to each plot. Legume seed of centro and puero were dibbled into small holes made with a wooden stick at the rate of two seeds per hole, spaced 100 × 50 cm. Each plot was inoculated with 100 gram of VAM inoculum/hole at sowing. The inoculum of VAM contained approxi­mately 820 spores/100 gram. The parameters measured were dry matter (DM) production at the three times of defoliation, and in vitro DM digestibility on the second and third defoliation only, because of the limited biomass at the time of the first defoliation.

The first defoliation was conducted three months after sowing and subsequent defoliations were conducted every two months. The plants were cut close to the soil surface to determine dry matter production. Dry matter production of each replicate was calculated from 1 m² subplots. To measure DM production the defoliated forage legumes was chopped, subsampled, and oven-dried to constant weight at 70ºC for 48 hours. The samples of the second and the third defoliation were finely ground and analysed to determine in vitro digestibility by the Terry and Tilley method (1963).

The analysis of variance on DM production and DM digestibility was done using the general linear model procedure of SAS. The significant differences among the treatments were tested using Duncan’s Multiple Range Test (DMRT).

Results and Discussion

Results

The effect of RP fertilization was not significant on the DM production of uninoculated legume (Table 1). Rock phosphate fertilization increased DM production on inoculated legume. The DM production of inoculated legume was not significantly higher than of the uninoculated one, at the same level of RP except on the unfertilized treatment.

Table 1. Dry matter production (kg ha^-1) of forage legumes with rock phosphate fertilization and mycorrhiza inoculation

Dry matter production of centro and puero increased after the first and after the second defoliation (Table 2). Dry matter production was higher for puero than for centro at the second period of defoliation (120.8 kg ha^-1 against 90.6 kg ha^-1).

For the same level of RP, dry matter production of the second and third defoliation was significantly higher compared to the first defoliation. Dry matter production of the third defoliation was significantly higher compared to the second defoliation for 0, 131, and 175 kg P ha^-1. RP fertilization did not significantly increase DM production at the first defoliation. However, RP fertilization increased DM production on the second and the third defoliation (Table 2).

Table 2. Dry matter production (kg ha^-1) of forage legumes on three periods of defoliation

Dry matter digestibility of puero was signi­ficantly higher compared to centro, with or without VAM inoculation on the second and third defoliation (Table 3). Dry matter digestibility of centro and puero inoculated with VAM was not significantly different compared to uninoculated one on the second defoliation. However, DM digestibility of centro was significantly lower compared to uninoculated one on the third defoliation. Contrasting this, DM digestibility of puero inoculated by VAM was significantly higher compared to uninoculated one on the same defoliation.

Table 3. Dry matter digestibility (%) of forage legumes on the second and third period of defoliation with mycorrhiza inoculation

Discussion

Dry matter production of VAM inoculated legumes was increased by rock phosphate fertilizer. However, both VAM inoculated and uninoculated legumes gave a similar DM production at the same level of rock phosphate (Table 1). Previous field experiments have shown that the response to VAM inoculation in the field greatly varies, and sometimes inoculation did not increase the production (Lin and Hao, 1991). This is because most of agricultural soils already contain indigenous populations of VAM fungi. Mycorrhizal inoculation would be successful in the field only if the native population was low and low effectiveness of indigenous-VAM fungi or depending upon VAM inoculum potential (Mitiku-Habte and Fox, 1993). Field experiment was carried out on the unsterilized soils. A year before, cassava had been harvested from the field, and since then the field was underfallow. Cassava is a VAM-obligate type. Cassava rhizosphere could increase the effectiveness of indigenous-VAM fungi (Potty, 1988). Spore isolation in the beginning of the field experiment showed that, there was an indigenous population of VA M fungi with a density up to 496 spores/100 g. During an experiment aiming at isolating spores from the soil, some spores infected by pathogen fungi were found. These pathogens could have decreased the effectiveness of VAM inoculum. Mycorrhizal spores could have been infected by pathogen fungi during the storage of the soil inoculum (Bagyaraj, 1988).

Rock phosphate fertilization did not signi­ficantly increase DM production of forage legumes at the first defoliation (Table 2). That was because the plants were still at the initial growth, while rock phosphate belongs to the group of slow release source of phosphorus (Jones, 1990).

The DM production of puero was higher compared to centro after first and second defoliation (Table 2). Performance of plant growth and root geometry (number of roots and distribution in the soil) of each plant species are different, as well as their response to the treatments (Kerridge and Ratcliff, 1982). Dry matter production stimulated after the first and the second defoliation promoted new stolon growth. Defoliation promoted vegetative regrowth of legume as shown in the DM production. At the earliest period, the growth of puero was slower compared to centro, therefore, DM production of puero was lower compared to centro. However, the regrowth of puero was faster than centro, therefore, the DM production of puero was higher than centro after each defoliation which in turn decreased the copper (Cu) content of puero (data not showed). This is so because of Cu as co-factor phenoloxidase enzyme which influenced the lignification process (Dell et al, 1995). Increase in phenoloxidase activity tend to increase the lignification process. As a result, in vitro DM digestibility of centro was lower than puero on the second and the third defoliation with or without VAM inoculation (Table 3).

Conclusion

Dry matter production on the second defoliation and DM digestibility of puero was higher than centro. Rock phosphate fertilization could increased the DM production of VAM inoculated legumes. Success of VA M inoculation in fields is affected by the effectiveness of indigenous-VAM fungi or depending upon VAM inoculum potential.

References

Bagyaraj, D.J. 1988. Technology for inoculum production of VA M fungi. In: Mahadevan, A., Raman, N., and Natarajan, K., eds., 1988. Proc. of the First Asian Conference on Mycorrhizae. Madras, 29-31 January, p. 326-328.

Chen Yinglong, Gong Mingqin, Wang Fengzhen, Chen Yu, Zhang Meiqing, Dell, B., and Malajczuk, N. 1998. Diversity putative ectomycorrhizal fungi and arbuscular mycorrhizal fungi in Eucalyptus plantations in Southern China. In: Gong Mingqin, Xu Daping, Zhong Chonglu, Chen Yinglong, Dell, B., and Brundrett, M., eds., 1998. Proc. of Guangzhou ACIAR International Workshop. Guangzhou, 31 August – 6 September, p. 21-28.

Coates, D.B., Kerridge, P.C., Miller, C.P., and Winter, W.H. 1990. Phosphorus and beef production in Northern Australia. 7. The effect of phosphorus on the composition, yield and quality of legume-based pasture and their relation to animal production. Tropical Grasslands. 24: 209-220.

Dell, B., Malajczuk, N., and Grove, T.S. 1995. Nutrient Disorders in Plantation Eucalyptus. ACIAR Monograph 31: 110 p.

Jones, R.J. 1990. Phosphorus and beef production in Northern Australia. 1. Phosphorus and pasture productivity. Tropical Grasslands. 24: 131-139.

Kerridge, P.C., and Ratcliff, D. 1982. Comparative growth of four tropical pasture legumes and guinea grass with different phosphorus sources. Tropical Grasslands. 16: 33-40.

Lin Xian-Gui, and Hao Wen-Yin. 1991. Occurrence and efficiency of VA Mycorrhizaein Fluvo-Aquic soils in North China. BIOTROP Special Publication. 42: 111-121.

Lukiwati, D.R. and Simanungkalit, R.D.M. 2001. Improvement of maize productivity with combination of phosphorus fertilizer from difference sources and vesicular-arbuscular mycorrhizae inoculation. In: Rajan, S.S.S, and Chien, S.H., eds. 2001. Proc. of International Meeting “Direct Application of Phos­phate Rock and Related Appropriate Technology Latest Developments and Practical Experiences”. Kuala Lumpur, 16-20 July. p. 329-333.

Lukiwati, D.R., and Simanungkalit, R.D.M. 2004. Production and nutritive value of Pueraria phaseoloides with vesicular-arbuscular mycorrhizae inoculation and hosphorus fertilization, abstr.p.90, Abstract of the 4^th International Symposium of the Working Group MO “Environmental Significance of Mineral-Organic Component-Microorganism Interactions in Terrestrial Systems”. Wuhan-China, 20-23 September.

Lukiwati, D.R. and Supriyanto. 1995. Performance of three VAM species from India for inoculum production in centro and puero. In: Prana, S.S., ed. 1995. International Workshop on Biotechnology and Development of Species for Industrial Timber Estates. Proc. LIPI Bogor. 27-29 June. Pp. 257-265.

Mitiku-Habte and Fox, R.L. 1993. Effectiveness of VA M fungi in non sterile soils before and after optimization of P in soil solution. Plant and Soil. 151(2): 219-226.

Potty, V.P. 1988. Response of cassava (Manihot esculenta) to VA M inoculation in acid lateritic soil. In: Mahadevan, A., Raman, N, and Natarajan, K. 1988. Proc. of the First ACOM. Madras, 29-31 January. pp. 246-249.

Terry, R.A. and Tilley, J.M.A. 1963. A two state technique for the in vitro digestion of forage crops. Journal British Grassland. F.O.C. 18: 104-111.

¹ Faculty of Animal Agriculture, Diponegoro University, Campus UNDIP Tembalang, Semarang-Indonesia

Short-term dynamics of soil organic matter and microbial biomass after
simulated rainfall on tropical sandy soils

Sugihara, S.¹; S. Funakawa¹; H. Shinjo¹ and T. Kosaki² 

Keywords: Soil organic matter, microbial biomass, microbial activity, soil organic matter management

Introduction

It is necessary both to understand the dynamics of the soil organic matter (SOM) from an environmental perspective (for example, whether soils are possible sinks or sources of atmospheric CO₂), and also from an agricultural perspective, as it is widely recognized that SOM is closely related to soil fertility. Microbial biomass (MB), which is defined as the living microbial component of the soil, is the primary agent of the soil ecosystem that is responsible for the decomposition of SOM, nutrient cycling and energy flow (Wardle, 1992). Above all, many studies have been carried out to evaluate the effects of drying/ rewetting on the decomposition of SOM and microbial dynamics due to its importance in the overall dynamics of SOM and atmospheric CO₂ (Kieft et al., 1987; Van Gestel et al., 1993a, b; Pulleman and Tietema, 1999; Franzluebbers et al., 2000; Fierer and Schimel, 2002; Fierer and Schimel, 2003; Wu and Brooks, 2005; Mihka et al., 2005). However, most of these experiments were conducted under laboratory conditions in which soil moisture contents were mostly fixed at high values after rewetting, although soils actually redry rapidly after rewetting under field conditions. There are few reports studying the effect of the dry-wet cycle on SOM and/or soil microbes under field conditions (Murphy et al., 1998; McNeil et al., 1998). To evaluate the actual dry-rewetting effect, responses and sensitivity of soil microbes to dry-rewetting changes should be investigated in more detail under field conditions. In addition, there is little information in arid/semi-arid ecosystems (Schwinning et al., 2004; Saetre and Stark, 2005), as most studies were carried out in temperate regions. Austin et al. (2004) report that pulsed-water events have a key role in a number of below-ground processes in arid/semi-arid ecosystems, and that changes in the nature of pulsed events due to human impact may be more important than larger-scale changes in total rainfall or average temperature in affecting biogeochemical cycling in water-limited ecosystems. As our understanding of the MB in tropical climates remains poor, it is important to identify whether general trends generated for mostly temperate ecosystems can also apply to tropical ecosystems (Wardle, 1992).

In tropical sandy soil, it is expected that the short-term dynamics of SOM and MB are crucial in the overall dynamics of SOM, because sandy soils are inherently poor in SOM retention and high temperature usually accelerates rapid SOM decomposition. It is possible, therefore, that the effect of drying/rewetting on tropical sandy soil is different from the results obtained in temperate ecosystems. In this study, simulated rainfall experiments were conducted in Thailand and Niger. Both soils are classified as tropical sandy soil. In Thailand, relatively high amounts of crop residues (sugarcane), composed of leaf and root biomass (1.1 and 1.0 Mg C ha^-1 y^-1, respectively), were incorporated into the soils at harvest, in addition to a considerable amount of litter-fall occurring during the cropping period and, according to Funakawa et al. (2005), approximately 4 Mg C ha^-1 y^-1 of SOM was annually decomposed. On the other hand, most of the crop residues were removed from the cropland after harvest either by humans or by cattle, so that only below-ground crop residues were incorporated into the soil in Niger. The Niger plots had been left fallow for more than 10 years and were converted to millet field 2 years ago. Annual SOM decomposition rates were about 1.0 Mg C ha^-1 y^-1 (Shinjo 2005; data not shown). Therefore, the dynamics of SOM are very different in the two regions.

The objectives of this study were (1) to evaluate the possible effects of rapid wetting/drying on the dynamics of soil microbes and the rate of SOM decomposition on tropical sandy soils by applying simulated-rainfall experiments under field conditions on the two tropical sandy soils, and (2) to compare the effect of different farming systems on the short term dynamics of SOM and microbial biomass. The influence of the residue management was also analysed in relation to microbial responses to water/substrate additions.

Materials and methods

Description of the study sites

The experiments were conducted on two tropical sandy soils at the end of dry season; the soils had different farming systems. One was Sam Jan Village near Khon Kaen City, Thailand (SJ) and the other was ICRISAT (International Crops Research Institute for the Semi-Arid Tropics) Sahelian Center (ISC) near Niamey City, Niger (SD). The soils of SJ and SD are classified into Typic Ustipsamments and Psammentic Paleustalfs, respectively, according to Soil Taxonomy (Soil Survey Staff, 2003). The soil properties of SJ and SD were, respectively: pH determined in water, 6.0 and 5.2; soil texture, sand with 90.7 and 95.1%; and organic C, 1.2 and 1.5 g kg^-1, and total N contents, 0.1 and 0.1 g kg^-1. Mean annual precipitation and average temperature in SJ and SD were 1,189 and 560 mm, and 27.1 and 29.1ºC, respectively.

Experimental design

The experiments were carried out at the end of the dry season (13-28 March 2004 in SJ and 26 May – 8 June 2004 in SD, respectively). We installed each of the three experimental plots (C, W and G plots) in the cropping field of SJ and SD (Figure 1-a). Each cropping field was bared and there were just plant residues.

1. The C plots (SJ-C and SD-C) were left without simulated rainfall as a control.
2. The W plots (SJ-W and SD-W) were treated with 10 mm of simulated rainfall at the start of the experiments.
3. The G plots (SJ-G and SD-G) were sprayed with a glucose solution (at a rate of 75 g C m^-2) as a substrate, together with the 10 mm of rainfall treatment. The G plots were designed to compare the possible activities of soil microbes with the effect of substrate addition at each site.

Each plot was further divided into subplots A and B (Figure 1-b) to measure the CO₂ efflux rate and MB, respectively, after removing visible plant residues from the soil surface.

The rainfall treatment was then applied, and it took 0.5 h for the irrigation procedure; following this, the start time of the experiment was set to 07:00 h to observe the effect of the drastic evaporation in the initial stage of the experiment. The W and G plots were left dry after the simulated rainfall.

Monitoring of air and soil temperature and moisture content

The air and soil temperature at a depth of 5 cm and the volumetric soil moisture content in the surface 0-15 cm were continuously monitored at the C and W plots using a data logger system (107 thermistor probe for soil temperature and CS616 for soil volumetric moisture connected to CR10X data logger; Campbell Scientific, Inc., Logan, USA).

Figure 1. Design of experimental plots (a), and plot description (b) 

Measurement of CO₂ efflux rate from soil surface

The CO₂ efflux rate from the soil surface was measured using a closed-chamber system in the subplot A in triplicate. For each measurement, a polyvinyl chloride (PVC) cylinder (13 cm in diameter × 15 cm height) was inserted into the soil to a depth of 5 cm immediately after the simulated rainfall. For each measurement, after the top of the cylinder was covered tightly with a plastic sheet and left for 40 min, a 50-mL gas sample was collected using a syringe, and then kept in a 30-mL glass vial until measurement, which was previously evacuated. At the same site, an air sample was collected to determine a baseline atmospheric CO₂ concentration. The concentration of CO₂ was measured using an infrared CO₂ controller (ZFP9AA11; Fuji Electric, Tokyo, Japan) and the increase in CO₂ concentration during the 40 min relative to the control sample was assigned as the CO₂ efflux from the soil surface. Sample collection was carried out at 4, 6.5, 9, 12, 14.5, 26, 32, 38, 50,5, 56, 61,5, 75,5, 86,5, 99.5, 346.5 and 363.5 h in the SJ plots, and 3.5, 5.5, 8, 11, 14, 26, 32, 37.5, 50, 55, 60.5, 99, 122, 317 and 320 h in the SD plots after the simulated rainfall in triplicate.

Measurement of MB and its activity

The microbial biomass was measured in the subplot B every 4-6 h using an applied substrate-induced respiration (SIR) method (Anderson and Domsch 1978; Lin and Brooks 1999), as follows:

1. As the SIR is strongly influenced by soil water content, a sufficient amount of water was added to the surface soil prior to each measurement to adjust water content to above 50% of the field-holding capacity. The amount of water added was determined based on the amount of evaporation expected. After the upper 5-cm layer of soil was well mixed, water was sprayed onto soil surface.
2. Glucose solution was sprinkled onto the soils using a syringe at a rate of 60 g C m^-2. Then, the upper soil was mixed uniformly.
3. After 30 min, a cylinder (the same one as for the CO₂ measurement described above) was placed into the soil to a depth of 5 cm, and then the top was covered tightly with a plastic sheet.
4. After 1 h, a 50-mL gas sample was collected using a syringe, and then kept in a 30-mL glass vial, which was previously evacuated. The concentration of CO₂ was measured using the infrared CO₂ controller (ZFP9AA11; Fuji Electric).
5. At the same time, an air sample was also collected to determine the atmospheric CO₂ concentration.

SIR method is commonly used for laboratory experiment, but we applied the method for field experiments, in order to evaluate the short term dynamics (4-6 h) of microbial biomass.

Sample collection was carried out at 5.5, 9.5, 15, 26.5, 38.5, 51, 62, 100, 124, 346.5 and 364 h in the SJ plots, and 3.5, 8, 13, 25.5, 36.5, 49.5, 59.5, 98.5, 121.5 and 316 h in the SD plots after the simulated rainfall in triplicate.

To calculate the microbial biomass based on the SIR measured, we used the following equation (Anderson and Domsch, 1978): where B is the microbial biomass C (g) and R is the soil respiration rate (mL h^-1). Since this equation is established on temperate soils and on laboratory condition, we tentatively used this equation for comparison.

It is widely recognized that temperature is one of the controlling factors for soil CO₂ efflux. To eliminate the effect of temperature fluctuations over a day, we corrected the soil respiration rate based on the Q₁₀ relationship, where the Q₁₀ factor is the ratio of respiration rates observed at temperatures differing by 10ºC (Fang and Moncrieff, 2001): where R₂ and R₁ are the respiration rates observed at temperatures T₂ and T₁, respectively. There are many reports on the relationship between temperature and CO₂ efflux from both a short-term and long-term perspective (Fang and Moncrieff, 2001; Parkin and Kaspar, 2003). We used Q₁₀ values of 2.2 and 1.7 for SJ and SD, respectively, which was preliminary calculated by our laboratory incubations. Using these parameters, we converted the respiration rate measured (R₁) to that expected at 22ºC (R₂) using MB determined by SIR.

It has been suggested that microbial activity reflects the effect of environmental conditions (Anderson and Joergensen, 1997; Mamilov and Dilly, 2002), and that the activity can be measured through the efficiency of substrate utilization (Dilly and Munch, 1998). In this study, we evaluated the contribution of microbial activity to the decomposition of SOM after drying/rewetting, using the following equation:

In this equation, both the CO₂ efflux rate and the microbial biomass C were expressed on an area basis (g m^-2 h^-1 and g m^-2, respectively). The CO₂ efflux rate was corrected to 22ºC based on the Q₁₀ relationship, as in SIR methods, in order to eliminate the effect of temperature fluctuation over a day.

Results

Fluctuation of air and soil temperature and soil moisture during the experiments

The air temperature at the SJ and SD plots fluctuated from 21ºC and 23ºC to 40ºC and 39ºC, respectively. Similarly the soil temperature at the SJ and SD plots fluctuated from 23ºC and 29ºC to 46ºC and 49ºC, respectively (Figure 2).

Figure 2. Fluctuation of air and soil (5 cm) temperature at the SJ plot (a) and the SD plot (b)

Figure 3. Fluctuation of volumetric moisture contents of soil (0-15 cm) (%) at the SJ plot (a) and the SD plot (b)

The daily temperature fluctuation was therefore substantial. On the other hand, before the experiment the soil was very dry – that is, the volumetric moisture content (VMC) in the SJ and SD plots was 1.5 and 3.5%, respectively, which is equivalent to -4.3 and -1.9 MPa, respectively. The simulated rainfall increased VMC in the SJ and SD plots up to 9.2 and 9.7%, respectively, which is equivalent to -7.7 and -3.5 kPa, respectively. Thereafter, the surface soil started to dry and VMC decreased continuously (Figure 3). Water loss through evaporation was high, especially in the daytime compared with the night-time. VMC in the SJ and SD plots decreased to 5.7 and 7.0% after 1 day, 3.7 and 3.9% after 3 days, and 2.0 and 2.9% after 2 weeks, respectively. The final VMC in the SJ and SD plots was equivalent to -3.8 and -2.2 MPa, respectively.

CO₂ efflux from the soil surface

The soil surface CO₂ efflux, which was mostly due to microbial respiration in our experiments, was small in the C plot compared with the W and G plots. After the simulated rainfall, the CO₂ efflux rate in the SJ-W, SJ-G and SD-G plots increased immediately, whereas that in the SJ-C, SD-C and SD-W plots did not change appreciably (Figure 4). In the SJ-W, SJ-G and SD-G plots, the CO₂ efflux rate continued to increase up to 6 h, and finally reached 2.6, 8.0 and 2.8 times higher than that of the control plots, respectively. After the peak of CO₂ efflux, it gradually decreased with fluctuation for fluctuated temperature. After correction of CO₂ data by soil temperature, however, there is no fluctuation of the CO₂ efflux (data not shown). The CO₂ efflux rate in the SJ-G plot was still 1.4 times higher than in the SJ-C plot at the end of the experiment, whereas CO₂ effluxes of the SJ-W and SD-G plots were almost equal to that of the control plots after 100 and 122 h, respectively.

The cumulative CO₂ efflux for the SJ-C, SJ-W, SJ-G, SD-C, SD-W and SD-G plots reached 4.25, 7.43, 13.68, 2.46, 2.68 and 7.16 g C m^-2, respectively, up to the 13^th day of the experiment. The difference in the cumulative CO₂ efflux between each treated plot and the control plots was 3.18, 9.43, 0.22 and 4.70 g C m^-2 in the SJ-W, SJ-G, SD-W and SD-G plots, respectively. These differences were primarily contributed to by the CO₂ flush in the initial 100 h.

Microbial biomass and activity

In the SJ-W, SJ-G and SD-G plots, MB was increased rapidly after the simulated rainfall (Figure 5). MB of the SJ-W plot was increased up to 10 h and then stabilized without decreasing in spite of rapid drying. After 100 h, it started to decrease to the level of the SJ-C plot. The MB of the SD-W plot fluctuated similarly to that of the SD-C plot, keeping a low level throughout the experiment. The MB of the SJ-G and SD-G plots was 5 and 1.5 times, respectively, higher than that of the control plots after 10 h and then gradually decreased with drying, but was twice that of the control plot by the end of the experiment (Figure 5).

Figure 4. Fluctuation of CO₂ efflux rates at the SJ plot (a) and the SD plot (b) Bar indicates standard deviation

Figure 5. Fluctuation of microbial biomass at the SJ plot (a) and the SD plot (b) Bar indicates standard deviation

Figure 6. Fluctuation of microbial activity (CO₂ efflux rate per microbial biomass)

Generally, the microbial activity reached the highest a few hours after the simulated rainfall and then decreased with water depletion (Figure 6). The duration of high microbial activity of the G plots was longer than that of the W plots.

Discussion

CO₂ flush and multiplication of soil microbes

The simulated-rainfall treatment significantly increased the MB and the microbial activity in the SJ-W, SJ-G and SD-G plots, resulting in large differences in the cumulative CO₂ efflux compared with the C plots. Many studies reported that the CO₂ flush after drying/rewetting could be caused by the increase of microbial activity and finished within a few days (Cui and Caldwell, 1997; Franzluebbers et al., 2000; Fierer and Schimel, 2003). Mihka et al. (2005) also reported that the flush of CO₂ was mainly related to microbial activity and microbial turnover (or microbial origin), and it finished in a day under laboratory conditions. In these studies, a rapid multiplication of MB after the drying/rewetting was not postulated. In our field experiments, however, not only microbial activity but also MB was rapidly increased in the SJ-W, SJ-G and SD-G plots after the rainfall treatment and the CO₂ flush lasted for 4 days, which were longer than the studies cited above. It is possible to assume that the longer CO₂ flush in our experiment was caused mainly by the multiplication (and subsequent stabilization) of MB, in addition to the increased microbial activity, although its reasons are still unclear as there are few reports on the effect of drying/rewetting on tropical sandy soils (Murphy et al., 1998).

Response of soil microbes to rewetting under different cropping systems in terms of residue management

The MB in the SD-W plot did not increase appreciably, unlike that in the SJ-W plot. This is one of the reasons why the CO₂ efflux rate in the SD-W plot did not increase significantly after water addition. As smaller amounts of visible plant residues were observed in the SD plots than in the SJ plots, readily decomposable SOM may also be less in the former in spite of similar total carbon contents in the soils. Such differences in the amount of available substrates, which were primarily caused by residue management in the respective farming systems, may have resulted in the differences in the response of soil microbes at the W plots in the present study. The cumulative CO₂ efflux in the SJ-G plot after simultaneous addition of water and glucose reached twice that of the SD-G plot and the difference was mainly caused by the CO₂ efflux rates in the initial 100 h. To demonstrate the multiplication of MB at the W and G plots, the ratios of MB of the W or G plot to the comparable C plots were plotted in Figure 7. It is clearly shown that the multiplication of MB in the SJ-G plot was significantly larger than the SD-G plot. There were small differences in environmental conditions between plots – that is, the much carbon substrate, the soil temperature and moisture contents condition, and soil texture. Therefore, the large differences observed for cumulative CO₂ efflux between the SJ-G and SD-G plots might be caused by the responses of soil microbial communities to substrate addition – namely, some species could increase rapidly using substrates added in the SJ plots, unlike in the case of SD. It might be one of the possible explanations for this; namely, each farming system in SJ and SD makes different amounts of crop residues, which were put into the soil, resulting in different decomposition dynamics. Another factor, which affects the microbial communities, still remains, for examples, climate and soil.

Figure 7. The ratio of microbial biomass of the W or G plot to the comparable C plot

Generally, sandy soils cannot retain SOM in soils compared with clay soils, as sandy soils are more aerated and SOM is scarcely protected from decomposition by being bound in clay-humus complexes or sequestered inside soil aggregates (Brady and Weil, 2002). Actually, most plant residues in SJ and SD would be decomposed within 1 or 2 years because of high temperature, so that there were few visible plant residues after 2 years from the last addition of plant residue in the SD plots (Funakawa and Shinjo 2005, data not shown). In addition, the turnover rates of MB and SOM in sandy soils and/or tropical regions are higher than in clay soils and/or temperate regions (Gregorich et al., 1991; Sakamoto and Hodono, 2000; Wardle, 1992). Therefore, it is possible to assume that the change of microbial composition is induced by rapid depletion of SOM (within 1 or 2 years) in the tropical sandy soils. We therefore suggest that in tropical sandy soils the farming histories strongly affect the decomposition dynamics of SOM through inducing different microbial responses, such as rapid multiplication on rainfall events. In general, established models for simulating SOM dynamics – for example, the Century model (Parton et al., 1987) and the Roth-C model (Jenkinson et al., 1991) – do not take into account the direct influence of the farming history on the composition of the soil microbial community. Further studies are needed to elucidate the influence of land-use histories, such as past-residue incorporation, on the response of soil microbes under fluctuating environments on tropical sandy soils.

Conclusion

The effects of rapid wetting/drying on SOM dynamics lasted for a limited number of days and depended on both the dynamics of the MB and the microbial activity in the tropical sandy soils. The multiplication of MB largely contributed to the initial CO₂ flush.

The acceleration of the CO₂ flush after addition of substrates and/or water was more pronounced in the SJ plots, which had been receiving higher amounts of residue input in recent years. Hence, it is necessary to take account of such historical factors of land management to appropriately simulate SOM dynamics in tropical sandy soils.

References

Anderson, J.P.E. and Domsch, K.H. 1978. A physiological method for the quantitative measurement of microbial biomass in soil. Soil Biology and Biochemistry, 10, 215-221.

Anderson, T.-H. and Joergensen, R.G. 1997. Relationship between SIR and FE estimates of microbial biomass C in deciduous forest soils at different pH. Soil Biology and Biochemistry, 29, 1033-1042.

Austin, A.T., Yahdjina, L., Stark, J.M., Belnap, J., Porporato, A., Norton, U., Ravetta, D.A., and Schaeffer, S.M.

2004. Water pulses and biogeochemical cycles in arid and semi-arid ecosystems. Oecologia, 141, 221-235.

Brady, N.C. and Weil, R.R. 2002. The Nature and Properties of Soils. Person Education, New Jersey, USA, pp. 521-532.

Cui, M. and Caldwell, M. 1997. A large ephemeral release of nitrogen upon wetting of dry soil and corresponding root responses in the field. Plant and Soil, 191, 291-299.

Dilly, O. and Munch, J.C. 1998. Ratios between estimates of microbial biomass content and microbial activity in soils. Biology and Fertility of Soils, 27, 374-379.

Fang, C. and Moncrieff, J.B. 2001. The dependence of soil CO₂ efflux on temperature. Soil Biology and Biochemistry, 33, 155-165.

Fierer, N and Schimel, J.P. 2002. Effect of drying-rewetting frequency on soil carbon and nitrogen transformation. Soil Biology and Biochemistry, 34, 777-787.

Fierer, N and Schimel, J.P. 2003. A proposed mechanism for the pulse in carbon dioxide production commonly observed following the rapid rewetting of a dry soil. Soil Science Society of American Journal, 67, 798-805.

Franzluebbers, A.J., Haney, R.L., Honeycutt, C.W., Schomberg, H.H., and Hons, F.M. 2000. Flush of carbon dioxide following rewetting of dried soil relates to active organic pools. Soil Science Society of American Journal, 64, 613-623.

Funakawa, S., Yanai, J., Hayashi, Y., Hayashi, T., Noichana, C., Panitkasate, T., Katawatin, R., and Nawata, E. 2005. Analysis of spatial distribution patterns of soil properties and their determining factors on a sloped sandy cropland in Northeast Thailand. Proceedings of Management of Tropical Sandy Soils for Sustainable Agriculture. In printing.

Gregorich, E.G., Voroney, R.P., and Kachanoske, R.G. 1991. Turnover of carbon through the microbial biomass in soils with different textures. Soil Biology and Biochemistry, 23, 799-805.

Jenkinson, D.S., Adams, D.E., and Wild, A. 1991. Model estimates of CO₂ emissions from soil in response to global warming. Nature, 351, 304-306.

Kieft, T.L., Soroker, E., and Fierestone, M.K. 1987. Microbial biomass response to a rapid increase in water potential when dry soil is wetted. Soil Biology and Biochemistry, 19, 119-126.

Lin, Q. and Brooks, P.C. 1999. An evaluation of the substrate-induced respiration method. Soil Biology and Biochemistry, 31, 1969-1983.

Mamilov, A.Sh. and Dilly, O. 2002. Soil microbial eco-physiology as affected by short-term variations in environmental conditions. Soil Biology and Biochemistry, 34, 1283-1290.

McNeil, A.M., Sparling, G.P., Murphy, D.V., Braunberger, P., and Fillery, I.R.P. 1998. Changes in extractable and microbial C, N, and Pin a Western Australia wheatbelt soil following simulated summer rainfall. Australian Journal of Soil Research, 36, 841-854.

Mihka, M.M., Rice, C.W., and Milliken, G.A. 2005. Carbon and nitrogen mineralization as affected by drying and wetting cycles. Soil Biology and Biochemistry, 37, 339-347.

Murphy, D.V., Sparling, G.P., Fillery, I.R.P., McNeil, A.M., and Braunberger, P. 1998. Mineralization of soil organic nitrogen and microbial respiration after simulated summer rainfall events in an agricultural soil. Australian Journal of Soil Research, 36, 231-246.

Parkin, T.B. and Kaspar, T.C. 2003. Temperature controls on diurnal carbon dioxide flux: Implications for estimating soil carbon loss. Soil Science Society of American Journal, 67, 1763-1772.

Parton, W.J., Stewart, J.W.B., Cole, C.V., and Ojima, D.S. 1987. Analysis of factors controlling soil organic matter levels in great plains grasslands. Soil Science Society of American Journal, 58, 530-536.

Pulleman, M. and Tietema, A. 1999. Microbial C and N transformations during drying and rewetting of coniferous forest floor materials. Soil Biology and Biochemistry, 31, 275-285.

Saetre, P. and Stark, J.M. 2005. Microbial dynamics and carbon and nitrogen cycling following re-wetting of soils beneath two semi-arid plant species. Oecologia, 142, 247-260.

Sakamoto, K. and Hodono, N. 2000. Turnover time of microbial biomass carbon in Japanese upland soils with different textures. Soil Science and Plant Nutrition, 46, 483-490.

Schwinning, S., Sala O.E., Loik M.E., and Ehleringer J.R. 2004. Thresholds, memory, and seasonality: understanding pulse dynamics in arid/semi-arid ecosystems. Oecologia, 141, 191-193.

Soil Survey Staff 2003. Keys to Soil Taxonomy. Ninth Edition. U.S. Department of Agriculture and National Resources Conservation Service, Washington.

Van Gestel, M., Merckx, R., and Vlassak, K. 1993a. Microbial biomass responses to soil drying and rewetting: The fate of fast- and slow-growing microorganisms in soils from different climates. Soil Biology and Biochemistry, 25, 109-123.

Van Gestel, M., Merckx, R., and Vlassak, K. 1993b. Microbial biomass and activity in soils with fluctuating water contents. Geoderma, 56, 617-626.

Wardle, D.A. 1992. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biological Reviews, 67, 321-358.

Wu, J. and Brooks, P.C. 2005. The proportional mineralization of microbial biomass and organic matter caused by air-drying and rewetting of a grassland soil. Soil Biology and Biochemistry, 37, 507-515.

¹ Graduate School of Agriculture, Kyoto University, Japan, [email protected]
² Graduate School of Global Environmental Studies, Kyoto University, Japan

Eucalypt litter quality and sandy soils: addressing two cumulative effects on
topsoil organic-matter and soil faunal activity in African plantations

Bernhard-Reversat, F.¹; I. Mboukou-Kimbatsa² and J.J. Loumeto³

Keywords: phenolics, soluble organic matter, soil organic matter fractionation, eucalypt, acacia

Introduction

Eucalypts are extensively grown in the tropics, either as farmer forestry or as industrial plantations mainly for paper pulp production. In Africa, many eucalypt plantations are grown on sandy soils, usually nutrient poor, which however are able to support tree growth. Eucalypt plantations were tried in the semi-arid Senegal in the years 1970-80s, on sandy soil and on loamy sand soil. In the Congo, commercial plantations are grown since 1978 near Pointe Noire, covering now more than 40,000 ha on sandy soils, under a wet climate, and a few experimental plantations are also grown on clay soils.

The main constraint of African sandy soils is their low soil organic matter (SOM) content and its decrease with cultivation (Feller et al. 1991, Walker and Desanker 2004). However SOM provides exchange sites, which contribute to nutrient conservation. Besides its importance for soil fertility and structure, SOM is the primary feeding resource for most soil living organisms and especially soil invertebrates, which in turn are essential for soil functioning (Lavelle and Spain 2001). So for sustainability of fast growing tree plantations, it is of significant importance to manage sandy soils in order to increase SOM.

In a forest environment, SOM originates mainly from litterfall and litter decomposition. It is well known that the litter quality of eucalypts is low (Woods 1974). Eucalypt leaf litter contains large amounts of phenolics (Bernhard-Reversat et al. 2001) which are antibiotic and anti-feeding agent for invertebrate and vertebrate fauna (Waterman and Mole 1994; Harborne 1997). Eucalypt litter quality leads to low biological activity and results in a low litter decomposition rate in eucalypt plantations.

Whether this low decomposition rate results in SOM accumulation in the topsoil is not clearly understood. The observation of particulate SOM fractions at the soil litter interface could bring some useful information. SOM fraction distribution was shown to be dependent on soil texture and on vegetation (Feller et al. 1991). The aim of the present paper is to study the effect of eucalypts and soil texture on organic matter incorporation to soil, and on macro-invertebrate density which is known to be dependant on vegetation (Lavelle and Spain 2001). Planted tree species of contrasting litter quality were compared to eucalypts.

Sites and methods

The climate of the Senegalese sites is semi-arid with nine dry months, and the study was carried out over several-years that included periods of lower rainfall than the average. One site was on a loamy sand soil (mean annual rainfall 500 mm) and the other on a sandy soil (mean annual rainfall 800 mm), both being tropical ferrugineous soils. The clay content of the topsoil is 5 and 15% respectively. In both situation the native vegetation was a dry forest dominated by Acacia seyal. The study was conducted in experimental plantations where several tree species were grown, and Azadiracta indica is compared here to Eucalyptus camaldulensis (Table 1). Eucalypts were finally shown not to be suitable for the climatic region. A. indica was previously extensively planted in villages, on roadsides, and in small farmer plantations.

In the Congo, large commercial plantations are grown around Pointe Noire on a ferralic arenosols (sandy soil), and in experimental plantations in the Niari valley on a ferralic clay soil. The clay content of topsoil is 5 and 50% respectively. Both sites have a seasonal equatorial climate (mean annual rainfall 1,250 mm) with four dry months, although atmospheric humidity remains high throughout the year. Eucalypts were hybrid clones resulting from the 1950-70s forestry research, Eucalyptus PF1 and Eucalyptus 12 ABL x saligna, here called HS2 (Delwaulle and Laplace 1988). Australian Acacia mangium and Acacia auriculiformis (Table 1) were established experimentally and are not usually planted in Congo. The Congolese eucalypt plantations area was large and comprised many plots among which many samples were collected. Only a few plots were available for the clay soil situations where the sampling was less extensive.

Table 1. Vegetation of the studied sites and number of samples for soil fractionation of SOM

Litter analyses were carried out on freshly fallen leaves from various sites, and the results were averaged regardless to the soil texture, which had little influence compared to species. Litter samples were air dried before being ground at 1.5 mm. Chemical analysis methods were previously described (Bernhard-Reversat 1998). Soil sampling for particle size fractionation was made in the 0 to 1-2 cm layer. The particle size fractionation of organic matter was adapted from Feller (1979), by sieving under water and floating. Carbon analysis was performed on each fraction. In the present paper, the data of the SOM fractions were collected in order to give three fractions, a coarse particulate fraction from 0.5 to 4 mm (Senegal) or 0.5 to 2 mm (Congo), a fine particulate fraction from 0.05 mm to 0.5 mm, and the remaining organo-mineral fraction less than 0.05 mm. In Senegal samples, the coarse particulate fraction under A. seyal included part of the small leaflets before their decomposition. It was observed in the Congolese plantations that soil organic matter content increased with the age of plantations, regardless to logging, and the planted plots older than 12 years old were not taken into account in order to have comparable age ranges in all planted species. Carbon was analysed with a Carmhograph® (Senegalese soils) or with the Ahn method (Congolese soils). Soil C mineralisation was measured in some experiments by CO₂ release during in vitro incubations of humid soil at 30ºC, by the NaOH method (Bernhard-Reversat 1993).

Soil invertebrate density was estimated in the Congolese plantations, through the TSBF method (Tropical Soil Biology and Fertility Program, Anderson and Ingram, 1993) in ten soil monoliths per studied plot. On clay soils, only one plot was investigated for each species, and the results should be checked with further studies.

Comparisons were made with the non-parametric tests of Kruskal-Wallis and Mann-Whitney, with the software Statview®.

Results and discussion

Litter quality

Annual litterfall from Eucalyptus accounted for 2.9 to 3 t.ha^-1 in the Senegalese plantations and approximately 5 to 7 t.ha^-1 in the Congolese plantations (Bernhard-Reversat 1993, Laclau et al., 2003). Australian Acacia litterfall ranged from 9 to 10 t ha^-1, whereas A. seyal litterfall was only 1.4 to 1.9 t ha^-1. Freshly fallen leaves of eucalypt species, compared to acacia species or other species had a low N content and had a high water-soluble organic matter and phenolic content (Table 2). However A. seyal seems to be an exception among leguminous plants, with also a very high phenolic content, also observed in green leaves (Breman and Kessler). Water-soluble compounds in eucalypts ranged from 20 to 40% of litter dry weight. Soluble organic matter included a great amount of water-soluble phenolics which ranged from 9 to 15% of litter dry weight. The total water soluble and methanol soluble phenolics reached 15 to 20% of litter dry weight in eucalytpt leaf litter compared to 2 to 5% in the other species studied but A. seyal.

Table 2. Chemical composition of fresh leaf litter from various tree species, in mg g^-1 of litter dry weight, with standard error in brackets. Solu OM: soluble organic matter, Solu phen: soluble phenolics, N.s. phen: non-soluble phenolics, Lign: lignin

Soil organic matter at the soil-litter interface

Organic matter at the soil-litter interface represents the first stage of litter incorporation into the soil. The SOM particle size distribution was significantly dependent on soil texture, as previously observed by Feller et al., (1991). The particulate organic fractions of SOM accounted for an average of 83% of total SOM in sandy soils, 75% in loamy sand soils, and 53% in the clay soils. Only the SOM in the organo-mineral fraction was significantly higher in the loamy sand soil than in the sandy soils. Unlike this, both fine particulate and organo-mineral fractions were significantly higher in the clay soil than in sandy soils (Figure 1). The adsorption of soluble organic compounds on the clay fraction and the formation of complexes with clay might be involved. In a laboratory experiment with Senegalese soils, 15% of added soluble C from E. camaldulensis litter was mineralized within 6 days in the sandy soil, and 4% in the loamy sand soil, showing the protection of the organic matter by clay.

Figure 1. Carbon in the SOM particle size fractions of the 0-2 cm layer of soil, in mg g^-1 of soil, according to soil texture, in Senegalese and Congolese tree plantations

SOM particle size distribution was also dependent on tree species. The eucalypt effect resulted in a lower amount of all SOM fractions, compared to the other tree species in the Senegalese sites (Figure 2). In the Congolese sandy soils, the fine particulate SOM fraction was lower under both eucalypt hybrids, whereas only E. HS2 showed this trend in the clay soil (Figure 3). Consequently, SOM accumulation was low in eucalypt plantation compared to other tree species, (comprising some species which are not presented here, as Prosopis juliflora, Acacia laeta, and except Pinus caribaea). This may be associated with litter quality. The lower accumulation of particulate SOM fraction under eucalypts might be related to the high non-phenolic soluble organic matter of eucalypt litter, as suggested by its significant relationship with the amount of fine particulate SOM in the species studied (Figure 4). The effect of phenolics on SOM is difficult to asses. Phenolics are known to exert an adverse effect on decomposition (Coûteaux et al., 1995) and thus might enhance organic matter accumulation at the soil-litter interface. Phenolics also prevent SOM accumulation (Inderjit and Mallik 1997) through the formation of soluble metal-organic complexes (Bernhard-Reversat 1999; Jansen et al., 2004). In the present study, the negative effect of soluble phenolics on SOM accumulation was obvious on all the component sizes of the particulate organic matter fractions (coarse and fine, Figure 3). No relationships were observed between litter quality and organo-mineral C fraction, the amount of which was highly related to soil texture.

Figure 2. Carbon in the SOM particle size fractions of the 0-2 cm layer of soil, in mg g^-1 of soil, in tree planta­tions and native vegetation in Senegal

Figure 3. Carbon in the SOM particle size fractions of the 0-2 cm layer of soil, in mg g^-1 of soil, in tree plantations and native vegetation in Congo

Soil macro-invertebrates

The positive effect of soil invertebrates on soil fertility is now recognized (Lavelle and Spain 2001). Acacia and eucalypt plantations were compared on the Congolese soils (Mboukou et al., 1998). The density of earthworms was higher on clay soil than on sandy soil in eucalypt plantations, unlike what was observed in A. mangium plantations, where earthworms were more numerous in sandy soil (Figure 5). It appeared that soil texture effect was added to eucalypt litter effect to keep earthworm density at a low level, although Lavelle and Spain (2001) reported the influence of vegetation having a significant influence. Termite and ant density was higher in sandy soil under the two tree species. According to Jones (1989), soil texture influences termites, and Meyer et al., (2000) reported the preference of some termite taxa for sandy soils. The other litter-dwelling taxa, taken together as “litter group”, were much more numerous under A. mangium, which has a higher quality litter, than under eucalypts, but the density was not influenced by the soil texture. Although the low number of species for which we recorded invertebrate density and SOM fraction data, (four) did not allow a statistical evaluation, relation­ships between particulate SOM amount and termite density (r = -0.984) and earthworm density (r = 0.931) could be expected significant with more data.

Figure 4. Relationships between the particulate organic matter of the 0-2 cm layer of soil and the litter quality of the corresponding vegetation, in Senegal and Congo. Data on A. seyal forests were removed from the total particulate C, because the coarse fraction contained a part of litterfall

Figure 5. Density, in number of individuals per m², of some invertebrate taxa in Congolese E. PF1 and A. mangium plantations

The negative effect of soluble phenolics on termites was suggested in previous researches on Congolese sandy soils, when eucalypt litterfall leaves were considered (Mboukou-Kimbatsa and Bernhard-Reversat 2001). A weak negative effect of the insoluble phenolics from the forest floor litter on earthworms (p = 0.06) and soil-feeding termites (p = 0.05) was also suggested in another study with 27-28 TSBF samples containing the taxon from 50 samples. The depressive effect of eucalypt plantations on soil invertebrates was reported previously (Zou 1993; Maity and Joy 1999), but its relationship with sandy soil texture should be assessed with more extensive studies.

Conclusions

Eucalypts and sandy soils present together some adverse effects on soil fertility, through both organic matter and biological activity. However more extensive sampling on clay soils and experimental studies on the role of litter chemistry should provide a better understanding of these relationships. Nevertheless, silvicultural practices that may increase SOM are essential to the sustainability of eucalypt plantations on sandy soils. The management of logging residues is currently adopted in Congolese commercial plantations in order to increase organic matter and nutrient conservation, and research in this area is being conducted (Nzila et al., 2004). Input of organic matter from non-eucalypt vegetation would also help increase SOM and invertebrate density. This is practiced by UR2PI (Unité de Recherche sur la Productivité des Plantations Commerciales) in studies on inter-planting leguminous trees with eucalypt, in order to improve soil nitrogen in Congolese plantations, but also expected to change litter quality and soil biology.

References

Anderson, J.M. and Ingram, J.S.I. 1993. Tropical soil biology and fertility. A handbook of methods. C.A.B/ International, Oxon, 221 pp.

Bernhard-Reversat, F. 1993. Dynamics of litter and organic matter at the soil-litter interface in fast-growing tree plantations on sandy ferralitic soils (Congo). Acta Oecologica, 14, 179-195. Bernhard-Reversat 1998.

Bernhard-Reversat, F. 1999. The leaching of leaf litter during early decomposition: laboratory experiments and ecological implications in Eucalyptus hybrids and Acacia auriculiformis plantations. Applied Soil Ecology. 12, 251-261.

Bernhard-Reversat, F., Loumeto, J.J. and Laclau J.P. 2001. Litterfall, litter quality and decomposition changes with Eucalyptus hybrids and plantation age. in Bernhard-Reversat F. (ed.) 2001 Effect of exotic tree plantations on plant diversity and biological soil fertility in the Congo savanna: a reference to eucalypts. CIFOR, Bogor, 23-30.

Breman H. & Kessler JJ. Le rôle des ligneux dans les agro-écosystèmes des régions semi-arides (avec un accent particulier sur les pays sahéliens) http://library.wur.nl/ way/catalogue/documents/Sahel/LIGNEUX/ INDEX.HTM

Coûteaux, M.M., Bottner, P. and Berg B. 1995. Litter decomposition, climate and litter quality, Trends in Ecology and Environment. 10, 63-66.

Delwaulle, J.C. and Laplace, Y. 1988. La culture industrielle de lEucalyptus en république populaire du Congo. Bois et Forêts des Tropiques n° 216: 35-42.

Feller, C, 1979. Une méthode de fractionnement granulometrique de la matière organique des sols. Application aux sols tropicaux à textures grossières très pauvres en humus. Cahiers ORSTOM, serie Pédologie. 17, 339-346.

Feller C. and Beare, M.H. 1991. Physical control of soil organic matter in the tropics. Geoderma 79, 69-116.

Harborne, J.B., 1997. Role of phenolic secondary metabolites in plants and their degradation in nature. In: Driven by nature: plant litter quality and decomposition, Cadisch G. & Giller K.E. eds, CAB International, Oxon, 67-74.

Inderjit and Mallik, A.U. 1997. Effect of phenolic compounds on selected soil properties. Forest Ecology and Management 92, 11-18.

Jansen, B. Nierop, K.G.J. and Verstraten, J.M. 2004. Mobilization of dissolved organic matter, aluminium and iron in podzol eluvial horizons as affected by formation of metal-organic complexesand interactions with solid soil material. European Journal of Soil Science 55, 287-297.

Jones, J.A., 1989 Environmental influences on soil chemistry in central semiarid Tanzania. Soil Science Society of America Journal 53, 1748-1758.

Laclau, J.P, Deleporte, P., Ranger, J. Bouillet, J.P. and Kazotti, G. 2003. Nutrient dynamics throughout the rotation of eucalyptus clonal stands in Congo. Annals of Botany 91, 879-892.

Lavelle, P. and Spain, A.V. 2001. Soil Ecology. Kluwer Acad. Publ, Dordrecht, 654 pp.

Maity, S.K. and Joy, V.C. 1999. Impact of antinutritional chemical compounds of leaf litter on detritivore soil arthropod fauna. Journal of Ecobiology 11, 193-202.

Mboukou-Kimbatsa, I.M.C., Bernhard-Reversat, F. and Loumeto, J.J. 1998. Change in soil macrofauna and vegetation when fast growing trees are planted on savanna soils. Forest Ecology and Management 110, 1-12.

Mboukou-Kimbatsa, M. and Bernhard-Reversat, F. 2001. Effect of exotic tree plantation on invertebrate soil fauna and poupulation changes with hybrid and plantation age. in Bernhard-Reversat F. (ed.) 2001 Effect of exotic tree plantations on plant diversity and biological soil fertility in the Congo savanna: a reference to eucalypts. CIFOR, Bogor, 49-56.

Meyer, V.W., Braack, L.E.O. and Biggs, H.C. 2000. Distribution and density of Cubitermes Wasmann (Isoptera: Termitidae) mounds in the Northern Kruger National Park. Koedoe 43, 57-65.

Nzila, J.D., Deleporte, P., Bouillet, J.P, Bouillet, J.P, Laclau, J.P. and Ranger J. 2004. Effects of Slash Management on Tree Growth and Nutrient Cycling in Second-rotation.

Eucalyptus Replanted Sites in the Congo. In Nambiar, E.K.S., Ranger, J.,Tiorks, A. and Toma T. (eds.) Site management and productivity in tropical plantation forests. Proceeding of of workshops in Congo, july 2001 and in China, february 2003. Cifor, Bogor. 15-30.

Walker, S.M. and Desanker, P.V. 2004. The impact of land use on soil carbon in Miombo Woodlands of Malawi. Forest Ecology and Management 203, 345-360.

Waterman, P.G. and Mole, S. 1994. Analysis of phenolic plant metabolites. Methods in Ecology. Blackwell Scientific Publications, London. 238 pp.

Wood, T.G. 1974. Field investigation on the decomposition of leaves of Eucalyptus delegatensis in relation to environmental factors. Pedobiologia 14, 343-371.

Zou, X. 1993. Species effect on earthworm density in tropical tree plantations in Hawaii. Biology and Fertility of soils 15, 35-38.

¹ IRD, Tropical Soil Ecology Laboratory, 32 rue Henri Varagnat, 93143 Bondy, France. France.Reversat@bondy. ird.fr
² IRD, Ecology Laboratory, BP 1286, Pointe Noire, Congo
³ University Marien Ngouabi, Department of Biology and Plant Physiology, BP 69, Brazzaville, Congo

Effect of fallowing on carbon sequestration in a humid tropical sandy soil
(Mangodara, Burkina Faso)

Nacro, H.B.¹; D. Masse ² and L. Abbadie³

Keywords: fallow, carbon, nitrogen, mineralization, soil particles

1. Introduction

When soil is brought into cultivation there is a progressive decline of organic matter content, and soil becomes quickly infertile (Piéri 1992). Therefore, particularly in Sudanian cropping systems, soil is left uncropped for several years (fallow) to improve soil fertility. The effect of fallow on soil rehabilitation processes, particularly on soil organic matter recovery has been extensively studied and benefits (Somé 1996; Manlay et al., 2002), as well as no significant changes (N’Dour et al. 2000) have been reported. The process of soil organic matter recovery following cessation of cultivation remains little understood, probably because it depends on physical, chemical and biological processes which are influenced by local environmental conditions.

Soil organic matter is made of a physically and chemically heterogeneous mixture of organic compounds at different stages of decomposition. Study of the relationship between the physical environment of soil microorganisms and their activity is useful for a better understanding of the dynamics of soil organic matter (Christensen 1992). Indeed microbial mineralization activities depend on soil organic matter quantity and quality, and may change in response to management practices. Although many studies focused on the effect of fallow on soil fertility, the simultaneous C and N net mineralization of soil fractions under this agricultural management has never been studied. This paper examines the relationships: (i) between the fallow duration (5, 10 and 30 years) and organic matter pools of soil size fractions and (ii) between these organic pools and their mineralization activities.

2. Materials and methods

2.1 Site characteristics and soil sampling

Soil samples were collected in the Mangodara area (9º54′N 4º25′W, Burkina Faso, West Africa). The average temperature is 27ºC and annual rainfall averages 1,100 mm yr^-1. The rainfall is characterized by high intensity and short duration (May to September). Soils are classified as tropical ferruginous soils (FAO-UNESCO 1990?: Acrisols). The vegetation is a typical open savannah dominated by Isorberlinia doka and Andropogonea.

Samples were collected in June 2001 from the 0-40 cm depth under three fallow sites from which cultivation had been excluded for 5, 10 or 30 years. In each of the nine sampling areas, ten samples were randomly collected, air-dried, thoroughly mixed, and gently sieved (<2 mm) to disrupt the macro-aggregates. The fraction >2,000 µm was discarded.

2.2 Determination of soil texture and isolation of organo-mineral particles

Soil texture was determined after the destruction of organic matter by H₂O_2, dispersion in hexameta-phosphate, and shaking for 16 hours, according to Balesdent et al. (1991). The organo-mineral particles of the soil were also separated as above, but without using H₂O_2, and hexametaphosphate; only shaking soil samples in water (20 g: 200 ml, ratio soil:water). The resulting 6 fractions were dried at 40ºC: 250 to 2,000 µm (coarse sand), 100 to 250 µm (fine sand), 100 to 50 µm (very fine sand), 20 to 50 µm (coarse silt), 2 to 20 µm (fine silt), and 0.05 to 2 µm (clay). The silt and the sand fractions were gathered; the following three fractions were thus obtained: 0.05 to 2 µm (clay), 20 to 50 µm (silt), and 50 to 2,000 µm (sand).

2.3 Analysis

Organic C and total N were determined using an automatic CHN analyser (NA 1500 Series 2, Fisons). The results were expressed as µg C g^-1 soil, and µg N g^-1 soil. Soil respiration was determined by placing 15 g soil in 130 ml closed flasks at 80% of their water holding capacity and incubated in the dark at 28ºC (±0.5ºC) for up to 3 days. The contribution of each size fraction to the total microbial activity in soil was assessed according to the method of Nacro et al. (1996), by comparing the activity of the whole soil to the activity of soils without a soil fraction (soil without coarse sand; soil without fine sand; etc.). Each incomplete soil was prepared by combining 5 fractions in the same proportions as in whole soil. The omitted fraction was replaced by chemical-free sand (particles >20 µm). Soil prepared by combining the 6 fractions was used as reference soil. The CO₂ concentration (µg C-CO₂ g^-1 dry soil) of each sample was measured on a gas chromatograph (Auto Analyser apparatus, Chrompack CP-2002 P Micro GC) after 1 and 3 days. Gas samples were automatically taken from the flasks with a 250 ml gas-tight syringe.

2.4 Statistical analysis

Analyses were replicated four times for organic C and total N, and three times for soil respiration. Data were subjected to an analysis of variance using SAS (Statistical Analysis System, SAS Institute Inc. 1990). Means that differed significantly were separated using the Scheffe’s test-; all tests were performed at the 95% level of probability.

3. Results

3.1 C and total N distribution

Organic C content of the whole soil increased slightly from the younger (3,005 µg C g^-1 soil) to the older (3,400 µg C g^-1 soil) fallow (Table 1). A reverse trend was observed for total N (Table 1): the lowest level was observed in the 30 year-old fallow (197 µg N g^-1 soil), and the highest in the 5 year-old fallow (246 µg N g^-1 soil). However, the effect of the lenght of fallow on organic C nor total N contents was not significnat. On the other hand, the C:N ratios of whole soils significantly increase with the fallows’ age (Table 1).

The C and N contents of the fractions are shown in Table 1. Most of soil organic matter (69 to 75% of total C) was found in the silt fractions (-2,248 to 2,264 µg C g^-1 soil) and this content was not significantly modified with the age of fallow. The lowest C content was found in the clay fractions and it decreased significantly with the fallow duration. The C content of the sandy fraction increased significantly with the fallow age. Regarding total N content, only the clay fractions were significantly affected by the fallow age. The C:N ratios of the sand fractions were 2 to 8 times higher than in the silt and clay fractions (Table 1), showing the recent origin of organic matter associated with the sand fractions.

3.2 Soil respiration

The CO₂ evolved by whole soils, and the contribution of each fraction are shown in Table 2. The potential contribution of a size class to C mineralization was calculated by subtracting the quantity of C-CO₂ produced by the incomplete soil lacking this class from the quantity of C-CO₂ produced by the completely recombined soil. The CO₂ production of the whole soils was low (11 µg g^-1 dry soil) and not influenced by the fallow age. The CO₂ production of soil fractions varied from 0.30 to 8.78 �g C g^-1 soil (Table 2). As observed for the C content, only the soil respiration of silt fractions was not affected by the fallow age, whereas that of clay increased after 30 years of fallow. On the other hand, soil respiration of sand fraction decreased with the fallow age (Table 2).

Table 1. Particles size distribution (percent of dry soil), and its relationship to organic carbon (µg C g^-1 soil), total nitrogen (µg N g^-1 soil) content, and C:N ratios of soils after fallows of 5, 10 and 30 years duration

Table 2. Calculated net contributions of separate size classes to C mineralization (µg C-CO₂ g^-1 soil), and ratio of carbon mineralized to initial organic C (% kc) after 3 days of incubation

 4. Discussions

Usually soil C content increases significantly during an extended fallow period (Somé 1996; Manlay et al., 2002). This was not observed here, probably because the C input to soil was reduced by fire in the dry season and grazing all year long, or because C “produced” was mostly added to the standing biomass rather than to the soil component (Pieri, 1992; Manlay et al. 2002). It could also be due to the nature of the soil, the crop history, and the land management. Feller et al. (1993) have shown that the effect of fallow on C content is particularly important on soils that are clayey. Soils studied here are very sandy (81 to 85%) with a low clay content (4 to 5%).

The C content of sand fractions increased significantly with the fallow age, indicating that dead plant matter enters gradually the soil organic matter pool: the greater the fallow age, the more organic matter enters the soil through the sand fractions. This clearly indicates the influence of the sand fractions on total organic C variation in sandy soils. However, the contribution of sand fractions to soil organic C was low (10 to 31%) probably because of intense microbial activities leading to high accumulation of by-products in the fine fractions especially in the silt.

Just as fallow age had no effect on soil organic C content, the CO₂ released from whole soils, and the C mineralization coefficients (Dommergues 1960) were also unchanged (Table 2). Since plant diversity and composition change considerably with time in fallow systems, an opposite result was expected. Probably in the short term, the overall soil respiration depends on the proportion of ready available C rather than on plant diversity and composition.

By contrast with its effects on organic C distribution, the contribution of sand fractions to C mineralization decreased with fallow age (from 38% to 28 and 23% respectively for the 10 and 30 year-old fallows). A reverse trend was observed for the silt (59 to 75%) and clay fractions (3 to 13%). This indicates that the quality of organic compounds associated with mineral fractions changed with time during the fallow (Ashman et al., 2003). In the same way, the C mineralization coefficient values (Table 2) showed that 0.3 to 1% of organic C associated with the sand fractions was mineralized, against 0.3 to 4% for the silt and clay fractions. The highest values were observed for the clay fractions from the 10 and 30 year-old fallow; hence, we can hypothesize that in the young fallow (5 year-old), C mineralization activity mainly depends on the availability of C associated with the sand fraction. But later, the C mineralization activity will be determined by the quality of organic compounds and microbial activity associated with the clay fraction. Such activity can lead to losses of soil organic matter, mitigating the effect of fallow on carbon sequestration in sandy soils. Further investigations are needed to understand that process.

5. Conclusion

While fallowing is generally expected to increase soil organic matter, long-term fallow (up to 30 years) did not significantly affected the soil organic C content of a sandy tropical soil. However, the C content of sand fractions was significantly affected by the age of fallow, suggesting that in sandy soil, soil organic matter storage is mainly controlled by soil particle larger than 50 µm.. But when considering soil microbial activity, most of the CO₂ was potentially produced by the particles smaller than 50 µm, particularly in old fallows. Therefore, if soil organic matter increases through plant residues entering the soil, its depletion occurs through fine organo-mineral particles. This paradox can be explained by the fact that most heterotrophic soil micro-organisms (Kandeler et al. 1999) and most of easily metabolisable compounds were associated with the fine particles. It seems that soil microbial activity is controlled by the coarse fractions in the young fallow, and by the fine fractions in the old fallow. Fallow is an efficient tool for improving soil fertility, but in line with the findings of Manlay et al. (2002), our results show that long fallows are potentially constrained in their accumulation of soil organic matter content on sandy soils.

Acknowledgments

The authors thank the International Foundation of Science (IFS) and the French Institute for Research and Development (I.R.D.) for providing a grant to HB Nacro. They also gratefully acknowledge the support of Mrs Danièle Benest from Ecole normale Supérieure (Laboratoire d’Ecologie, Paris, France) for her efficient help during laboratory experiments.

References

Ashman, M.R., Hallett P.D., and Brookes, P.C. 2003. Are the links between soil aggregate size class, soil organic matter and respiration rate artefacts of the fractiona­tion procedure? Soil Biology and Biochemistry, 35, 435-444.

Balesdent, J, Pétraud, J.P, and Feller, C. 1991. Effets des ultrasons sur la distribution granulométrique des matières organiques des sols. Science du sol, 29, 95-106.

Christensen, B.T. 1992. Physical fractionation of soil organic matter in primary particle-size and density separates. Advances in Soil Science, 20, 1-90.

Dommergues, Y. 1960. La notion de coefficient de minéralisation du carbone dans les sols. Agronomie Tropicale, 15, 54-60.

Feller, C, Lavelle, P., Albrecht, A., and Nicolardot, B. 1993. La jachère et le fonctionnement des sols tropicaux: rôle de l’activité biologique et des matières organiques. Quelques éléments de réflexion. In. Floret, Ch. and Serpentier, G., ed., La jachère en Afrique de l’Ouest. ORSTOM, Colloques et séminaires, Paris, 33-46.

Kandeler, E., Stemmer, M., and Klimanek, E.M. 1999. Response of soil microbial biomass, urease and xylanase within particle size fractions to long-term soil management. Soil Biology and Biochemistry, 31, 261-273.

Manlay, R.J., Masse, D., Chotte, J.L., Feller, C, Kairé, M., Fardoux, J., and Pontanier, R. 2002. Carbon, nitrogen and phosphorus allocation in agro-ecosystems of a West African savanna. II. The soil component under semi-permanent cultivation. Agriculture Ecosystems and Environment, 88, 233-248.

Nacro, H.B., Benest, D., and Abbadie, L. 1996. Distribution of microbial activities and organic matter according to particle size in a humid savanna soil (Lamto, Côte dIvoire). Soil Biology and Biochemistry, 28, 1687-1697.

N’Dour, Y.B., Fardoux, J., and Chotte, J.L. 2000. Statut organique et microbiologique de sols ferrugineux tropicaux en jachère naturelle (Sénégal). In. Floret, Ch. and Pontanier, R., ed., La jachère en Afrique tropicale. Paris, John Libbey Eurotext, 354-360.

Pieri, C. 1992. Fertility of soils. A future for farming in the West African savannah. Germany, Springer-Verlag, 348 p.

SAS Institute Inc., 1990. SAS System for Linear Models. SAS Institute. Cary, NC.

Somé, N.A. 1996. Les systèmes écologiques post-culturaux de la zone soudanienne (Burkina Faso). Structure Spatio-temporelle des communautés végétales et évolution des caractères pédologiques. Thèse de Doctorat, Université Paris VI, 212 p.

 ¹ Institut du Développement Rural, Université Polytechnique de Bobo-Dioulasso, 01 BP 1091 Bobo-Dioulasso, Burkina Faso, [email protected]
² Institut de Recherche sur le Développement, BP 162 Ouagadougou, Burkina Faso
³ Ecole Normale Supérieure, Laboratoire d’Ecologie, 46, rue d’Ulm 75230 Paris cedex 5, France