Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Cycling and residual effects of phosphorus in the rainfed lowland rice ecosystem of Cambodia

Pheav, Sovuthy (2002) Cycling and residual effects of phosphorus in the rainfed lowland rice ecosystem of Cambodia. PhD thesis, Murdoch University.

PDF - Whole Thesis
Available Upon Request


Sandy textured, low fertility soils which experience intermittent loss of soilwater saturation are typical of the rainfed lowland rice ecosystems. Low P supply in the rainfed lowlands is one of the key constraints to rice yield, but little is known about the key processes of P cycling in these rice ecosystems. The objectives of the study were: 1) to determine the long-term P sorption-desorption behaviour in aerobic conditions, and P release characteristics of three main lowland rice-growing soils of Cambodia under alternating reduced and oxidised conditions; 2) to quantify the fate of fertiliser and residual P in the soil, and mass balances of P for rice in an acid sandy lowland soil; 3) to determine the potential for increased P uptake efficiencies by quantifying the turnover process of residual P fertiliser and crop residues during the fallow period in rainfed lowland conditions; and 4) to develop a general schematic of P cycling in lowland soils in which rice is grown under rainfed ecosystems.

To achieve these aims, a series of experiments in the field, glasshouse, and laboratory were undertaken on the major lowland rice soil of Cambodia: Prateah Lang, a strongly acid Plinthustalf with low productivity potential (White et al 1997a,b; Soil Survey Staff, 1994). Koktrap and Toul Samroung soils were also used to determine variation in P sorption and release on other major lowland rice soils of Cambodia. Like Prateah Lang, the Koktrap soil (Plinthaquult) was moderately to strongly acidic and had low to medium productivity potential, but it had a higher clay content. Toul Samroung soil (Endoaqualf) was clayey, near neutral in pH and had high productivity potential for rainfed lowland rice.

The long-term P release characteristics of the three soil groups were examined in pots by adding three P rates and successively growing five rice crops. Largest increases in shoot and root dry matter (DM), P concentrations and total P uptake of the first rice crop were obtained on the sandy Prateah Lang soil compared to the Toul Samroung and Koktrap soils. This was attributed to increased available P in the resin-extractable P fraction because of the reduced P sorption capacity of the sandy Prateah Lang soil. Conversely a higher residual P value was obtained in succeeding crops on the clayey Koktrap and Toul Samroung soils than the sandy Prateah Lang soil, since the high P sorption on clayey soils allowed prolonged release of P into the soil solution.

Field results suggested that P fertiliser application at about 17 kg P/ha on the sandy Prateah Lang soil corrected P deficiency and had significant residual value to maintain yields in the following crop, but thereafter yield declined due to P deficiency. The initial use of 17 kg P/ha, followed by 8-10 kg P/ha per crop was adequate to maintain grain yields of 2.5 to 3.0 t/ha, a positive P balance in the soil, and a satisfactory available soil P level for rice crops. Application rates of P fertiliser (16.5 to 33 kg/ha) higher than the recommended levels resulted in: 1) low relative P recoveries in plant uptake; 2) losses through leaching in a sandy soil when the ground water-table dropped below the soil surface; and 3) increased residual P associated with recalcitrant organic P and/or occluded P that is unavailable for crop uptake.

In the unfertilised soils, the amount of P extracted by resin was equivalent to less than 0.3 kg P/ha, and when sampled in air-dried soil samples at harvest was little influenced by fertiliser P application. Most of the added P fertiliser was retained in soil as residual P in the NaOH-Pi, NaOH-Po, and Residual-P pools. However, it was shown the shoot DM, grain yield, and P uptake of rice were highly correlated with the resin-extractable P pool in the soil, suggesting that the resin extracted P represents a good predictor of the plant P uptake.

Soil inorganic P (Resin-P and NaOH-Pi) fractions were dominant amongst other P fractions when soil was collected from continuously flooded conditions, and analysed without drying. By contrast, labile NaOH-Po and occluded Residual-P were major soil P fractions when flooded soils from the field were airdried before analysis. Fluctuations in most soil P fractions in the field, particularly of Resin-P and NaOH-Pi fractions over time, could be attributed to change from dry or moist to saturated soil conditions. This suggests the soil P pools were very dynamic and responsive to changing of soil-water regimes.

The residual value of the high-P (e.g. 16.5 and 33 kg P/ha) soils treated additionally with rice straw incorporation significantly increased biomass of volunteer pastures (legume and major non-legume) during the fallow after the wet season rice, particularly during the early wet season. The total biomass production and P uptake were significantly greater for non-leguminous compared to leguminous fallow species at all levels of residual P. This suggests that the non-legumes contribute much more to P turnover during the fallow period.

The application of crop residues either as rice straw or volunteer early wet season pastures marginally increased crop productivity and total P uptake, soil P reserves, and microbial biomass C, N and P. However, more striking increases of these parameters were obtained with the combination of crop residues with inorganic P fertiliser. Microbial biomass P increased rapidly and reached maximum at two and four weeks after straw or fallow crop residue incorporation for the field and glasshouse conditions, respectively. The greatest increase in microbial biomass P at these early growth stages suggests that uptake of inorganic P by microbial cells can be extremely rapid. However, soil microbial biomass P declined mostly during active growth stages of rice (maximum tillering and/or flowering). This suggests rapid turnover of microbial biomass and that its pool of sequestered P was released to the soil for P uptake by rice.

Soil labile organic P (NaOH-Po) extractable fraction and microbial biomass C both significantly increased over time in the field, and reached maximum at 40 weeks after rice straw incorporation, and their values were well correlated (r2 = 0.78). This suggests that the increase in the NaOH-Po fraction was closely related with extractable microbial biomass activity and its response to the additions of rice straw plus P fertiliser.

In conclusion, the rainfed lowlands are a complex environment for managing soil fertility because of the fluctuations in soil-water regimes, and these cause soil-water interactions with nutrients, especially with P, a major nutrient limiting rice yields. More information is needed to manage the effects of fluctuating soilwater conditions on P supply in order to increase productivity of rice over the range of soil groups including the soils reported in the thesis. This thesis has made a significant contribution to the understanding of P management and cycling in the rainfed lowland rice ecosystems of Cambodia, which would provide a basis for increasing output of rice production from the low P soils in Cambodia.

Item Type: Thesis (PhD)
Murdoch Affiliation(s): Division of Science and Engineering
School of Environmental Science
Notes: Note to the author: If you would like to make your thesis openly available on Murdoch University Library's Research Repository, please contact: Thank you.
Supervisor(s): Bell, Richard, White, P. and Kirk, G.
Item Control Page Item Control Page