PHOSPHORUS TEST METHODS CALIBRATION BY FIELD EXPERIMENTAL DATA FOR THE SOILS OF LOW COUNTRY INTERMEDIATE ZONE OF SRI LANKA

Green house studies showed that the 2.5 % Acetic acid (HAc), Olsen's bicarbonate (0.5 M NaHC0 3 ) , Anion Exchange Resin (AER) and Anion Exchange Resin + Cation Exchange Resin (AER + CER) methods were satisfactory for the estimation of plant available phosphorus (P) of soils of the Low Country Intermediate Zone of Sri Lanka. The objective of this study was to calibrate the four extraction methods by field experiments. Relative yield of Pueraria phaseoloides grown in field on five soil series treated with and without P fertilizer, but with basal treatment of nitrogen, potassium and magnesium fertilizer were employed a s the plant index for soil P status. Soil P status of each soil prior to treatment was determined by the four extraction methods. The 2.5 % HAc P values showed the best correlation with relative dry matter yield of Pueraria by fitting significantly to the Cate and Nelson model with R 2 of 0.628 (P<0.01). Both AER P and AER + CERP values showed the poor correlation with R 2 of 0.460 (P<0.05) compared to HAc P. NaHC0 3 -P showed poorest correlation with R 2 of 0.449 (P<0.05). The 2.5 % HAc P was found to be superior to others a s it accounted for 62.8 % of the variation of relative yield against 45 46 % b y the other extractants. Threshold P levels a s determined by the correlation data were 7.8, 3.84, 5.34 and 8.47 mg/kg for 2.5 % HAc P, AER P, (AER + CER) P and N a H C 0 3 P respectively. The 2.5 % HAc method was found to be the most suitable P extraction method for Low Country Intermediate Zone of Sri Lanka.


INTRODUCTION
Soil phosphorus (P) estimation by laboratory methods is of major importance in an effective phosphate fertilizer recommendation programme.A large number of soil P tests for plant available soil P estimation are found in literature.
But none of the methods can be recommended for any soil, unless such methods are evaluated and properly calibrated with respect to a particular soil -plant combinations.
Evaluation of methods for the determination of available P of some coconut growing soils in Low Country Intermediate Zone of Sri Lanka by green house pot experiments revealed that 2.5 % Acetic acid (HAc) (Anon, 1985), Olsen's Bicarbonate (0.5 M NaHC0 3 ) (Olsen et al., 1954), Anion Exchange Resin (AER) and Anion Exchange Resin + Cation Exchange Resin (AER + CER) (Somasiri and Edwards, 1992) extraction were suitable (Wijebandara and Somasiri, 1994).But the P values by the above methods cannot be interpreted meaningfully unless each of them are calibrated by field experiment.The objective of the present study was to calibrate the 2.5 % HAc method, 0.5 M NaHC0 3 method, AER method and AER + CER methods by field experiments with respect to Pueraria phaseoloides which is an important cover crop grown under coconut.

MATERIALS AND METHODS
A field experiment was design to study the plant response to the applied P in ten coconut growing soils.

2.5% Acetic acid extraction: (Anon, 1985)
A sample of 2.5 g soil was shaken with 100 ml of 2.5 % (v/v) acetic acid solution for 2h.The mixture was filtered through Whatman no 42 filter paper and the filtrate was analyzed for P.

1
Olsen's Bicarbonate extraction: (Olsen et. al., 1954) Five grams of soil was shaken with 100 ml of 0.5 m sodium bicarbonate solution (pH 8.5) (with 5 ml of 0.5 % polyacrylamide in one liter of NaHC0 3 ) for 30 minutes.The mixture was filtered through Whatman 42 filter paper and the pH of the filtrate was adjusted to 5 using paranitrophenol indicator, and the filtrate was analyzed for P.
Anion Exchange Resin extraction: (Sibbesen, 1977 and1978) Five grams of soil was weighed into a 250 ml polypropylene bottle, and 100 ml distilled water and 4 ml of anion exchange resin (IRA 400 CI") in a mesh bag was added to each bottle.The contents were shaken in an end-overend shaker for 4 h.The resin bags were recovered, washed with distilled water to make them free of soil residues and absorbed phosphate onto resin was eluted by shaking each resin bag in 50 ml of 1M NH 4 CI solution for 30 minutes, The elute was analyzed for P.

RESULTS AND DISCUSSION
According to pH measurement in 1:5 soil-water suspension (Table 1) seven soils (including A and B horizon) prior to P treatment showed strongly acidic reaction (pH 5.0 -5.5) and three soils showed very strongly acidic reaction (pH 4.5 -5.0) (Troug, 1948).Laboratory estimates of P in different soils obtained by individual methods were quite variable (Table 2).provides a method for determining the dividing line, or the critical level between two categories of soils; viz., high probability of response (low soil P status) and low probability of response (high soil P status) to applied P. Each soil P data set (Table 2) was arranged in the ascending order with corresponding RY% values for Pueraria (Table 3), and by a simple iterative process a series of R 2 values were obtained for arbitrary divisions made into high and low P status made at various levels of soil P. The R 2 was calculated for each arbitrary division and the division at which R 2 becomes maximum is taken as the critical level of soil P for the particular soil-plant combination, with respect to the method tested.The value of R 2 is also an indication of the degree of fitness of the data set to the model.The calculated R 2 values, and the critical level for each soil P test are given in Table 4.   4).There was no clear geometrical relationship between soil P values by different extraction methods and leaf P concentration and P uptake of Pueraria.
The results of the field experiment showed that, out of the soil P test methods examined, 2.5 % HAc -P gave the best correlation with plant yield under different soil conditions.The correlation was poorest with 0.5 M NaHC0 3 -P values.Threshold P levels as determined by the correlation data were 7.8, 3.84, 5.34 and 8.47 mg/kg for HAc -P, AER -P, (AER + CER) -P and NaHC0 3 -P respectively.The 2.5 % HAc method is simple and much economical for soil phosphorus estimation compared to other three methods.The 2.5 % HAc can also be used as a multi -element extractant for the estimation of soil potassium and magnesium (Anonymous, 1985).Therefore, it is very attractive as a routine soil testing method, compared to single element estimation methods.

CONCLUSION
Results showed that 2.5 % HAc-P accounted for 63 % of the variation in relative dry matter weight and therefore considered the best extractant for the determination of available P of soil.As other methods also accounted for only 45 -46 % of the variation of relative dry matter weight, they are inferior to HAc method.The threshold level for 2.5 % HAc -P was found to be 7.8 mg/kg serves as an index for separating soils on the basis of P fertility in respect of leguminous creeping cover crops.As 2.5 % HAc method is a multi-element, rapid and economical extraction method, it is more attractive than other methods for routine work in soils of Low Country Intermediate Zone of Sri Lanka.
series, Sudu series and Lunuwila series soils.In each soil, beds were prepared including two depths; viz., the top soil ranging from 0-20 cm from the surface (A horizon) and the sub soil obtained after completely removing the top soil.The surface gap occurred due to removal of top soils were filled with the sub soil taken from the surrounding areas.The beds were arranged in the randomized block design with three replicates in each of which there were two treatment plots, +P and -P.For each depth, beds were prepared separately.Pueraria phaseoloides seedlings were planted at distances of 30 cm between rows and 15 cm between plants in each bed.Three months after planting, vegetative part of Pueraria were harvested within an area of 0.5 m 2 and fresh and dry weights of harvested materials were recorded.Percent Relative yield (RY%) of the crop in each soil of vegetative parts was taken as the yield of the Pueraria.Total Phosphorus content of vegetative parts of Pueraria was estimated by wet digestion with HN0 3 : HCI0 4 mixed at 1 : 4 ratio (Manual of analytical methods, 2000) Soils were sampled prior to treatment application up to a depth of 30 cm from each replicate and bulked.Air dried 2 mm sieved soils were analyzed for available P by 2.5 % HAc method, 0.5 M NaHC0 3 , AER method and AEP + CER method.Phosphorus in solution resulting in each extraction was estimated by the molybdenum-blue method(Murphy and Riley, 1962).
Exchange resin (IRA 400, CI") filled into separate mesh bags were added to each bottle.The contents were shaken in an end -over-end shaker for 4 h.The resin bags were recovered and washed with distilled water to make them free of soil residues.Phosphate sorbed onto resins were eluted by shaking each pair of resin bags in 50 ml 1 M NH4CI solution for 30 minutes, and the elute was analyzed for P.Analytical data were analyzed according to Cate and Nelson statistical procedure(Cate and Nelson, 1971) and threshold levels for soil P were determined.
reaction in the soil / solution / resin system, anion exchange resin in the presence of cation exchange resin extracted a greater quantity of soil P than anion exchange resin alone.This tendency is very similar to previous observations byVaidyanathan and Talibudeen (1970).The Anion exchange resin extractable P values in soils of the field experiment obtained by employing 4h shaking time were much lower than that of other extractants (2.5 % HAc, 0.5 M NaHC0 3 and AER + CER -4h) ranging from 2.37 mg/kg to 7.00 mg/kg.But the values by all four methods showed highly significant linear correlations between each other (fit of soil test data and corresponding plant indices to Cate and Nelson statistical model can be used to evaluate the suitability of soil P testing methods.Such a soil test can also be calibrated to get the critical soil P levels.Cate and Nelson statistical model(Cate and Nelson, 1971) showed the highest correlation with relative yields of Pueraria fitting significantly to the Cate and Nelson model with R 2 of 0.628 (P < 0.01) for dry weight of vegetative parts.Thus the 2.5 % HAc-P values could account for the variation of 63 % of the production of vegetative parts of Pueraria in different soil condition in the field.The correlation coefficients R 2 between AER -P at 4h extraction and relative yields of dry weight of Pueraria was 0.460 (P<0.05) which explained only 46 % of the variation in the production of dry weight of Pueraria.The situation was much the same with