EFFECTS OF APPLIED POTASSIUM AND PHOSPHORUS ON BRONZING IN RICE GROWN IN IRON TOXIC SOILS

Bronzing or iron tqxicity is a physiological disorder of rice caused by high concentrations of soluble iron and deficiencies in P and K i n wetland soils. ~ G l d experiments were conducted a t Mellewa and Dodangoda to evaluate the effects of applying P and K and varietal tolerance in iron toxicity. Soil analysis showed that available P and exchangeable K were deficient while exchangeable Fe was very high in iron-toxic soils. Application of 85.8 kg K/ha reduced the bronzingsgmptams and Fe content in the leaves, and increased growth and yield of rice. Very severe bronzing symptoms, corresponding to a tissue iron content > 700 rnglkg, were obtained when no fertilizers were applied. Symptoms of iron toxicity were mild after application of 57.2 kgWha to the tolerant variety (Bw 267-3) and after application of 85.8 kg Wha to the susceptible variety (Bg 94-1). Growth and grain yield of rice also increased with increased application of phosphorus fertilizer. Application of 85.8 kg K/ha and 32.1 kg Plha and varietal tolerance can significantly increase the growth and grain yield of rice grown on iron toxic mineral soil of the low country wet zone.


INTRODUCTION
Bronzing or iron toxicity is a physiological disorder of rice in Sri Lanka. It is the major soil constraint on rice cultivation in the low country wet zone. Low country wet zone(LCW2) comprises land below 300m in elevation and receivinga rainfall > 2500 mm per annum. Of the 90000 hectares of cultivated rice lands in LCWZ, approiimately 113 may be considered as potentially iron t0xic.l However, this disorder is seen in localized spots rather than.over continuous areas. Therefore the actual area affected by iron toxicity is about 5000 ha. Many rice varieties that are apparently tolerant to iron toxicity are affected nevertheless as shown by their low yield. This disorder has been associated with excess soluble iron, low pH, poor drainage, low levels of phosphorus and potassium and presence of respiratory inhibitors such as hydrogen sulphide. [2][3][4] In LCWZ it occurs generally in the Srst and second order valleys due to a combination of factors mentioned above. The low silica, phosphorus and potassium concentration of sandy soils in these valleys predispose the plant to bronzihg. Therefore iron toxicity is ascribed to a multiple nutritional soil stress rather than to a high level of active iron under acidcondition. 596 This disorder is characterized by a reddish brown discoloration of the leaves, 7*8 inhibition of tillering, retardation ofroot growth and coarseness, discoloration and death of root^.^^^ Tissue concentration of Fe for the occurrence of iron toxicity varies widely, depending on the variety, stage of crop and nutrient stat~s.~JO Varietal tolerance has been utilized to d e v i a t e this disorder but is effective only a t moderate levels of toxicity. Ameliorative measures such as liming, periodic surface drainage, interception of interflow and good fertilizer management are considered effective.'," However most of these methods are ei*er expensive or impractical under prevailing conditions. An integration of varietal tolerance and good fertilizer management may be the most effective and economical approach to alleviate this problem. This paper deals with two experiments on potentially iron toxic sandy @oils to determine the effectiveness of applied potassium and phosphorus inalleviating brc-lzing in a susceptible and tolerant rice variety.

METHODS AND MATERIALS
The experiments were carried out a t Millewa and Dodangoda in Kalutara district in LCWZ of Sri Lanka. The soils a t both sites were loamy sand with medium levels of organic matter, acidic in reaction and deficient inexchangeable K and available P ( Table 1). The soil contained comparatively high levels of total iron with 1N ammonium acetate extract ( Table 1). The rice. va'rieties used in this study were Bg 94-1, susceptible to iron toxicity and Bw 267-3, tolerant to iron toxicity.

Effect of applied potassium on iron toxicity in rice:
The first experiment w a~ camed out to determine the effect of applied K on bronzing in rice grown in iron toxic mineral soils at Millewa. Two variable factors considered were rice variety and potassium level. Two rice varieties mentioned earlier were used. Three ' weeks old seedlings of both varieties were transplanted a t a spacing of 20x15 cm with 3 seedlings per hill. Graded levels of potassium at 0, 28.6, 57.2, 85.8 and 114.4 kg Wha were applied in 3 equal split doses as basal, 3 and 6 weeks after transplanting. Treatments were replicated 3 times in a randomized complete block design. Urea was applied at a rate of 72.0 kg N/ha in 3 split doses (15.4 kg N as basal, 28.3 kg h ha each at 3 and 6 weeks after planting). Concentrated super phosphate was applied at the rate of 21.4 kg P h a just before planting. Hand weeding was the only method of weed control adopted and was done a t each fertilizer application. Insecticides were applied when required. Severity of bronzing was observed periodically and rating done according to IRRI standard evaluation system for rice.12 Plant growth was observed by tiller counts, plant height and yield components. Grain yield of each treatment was recorded. The experiment was repeated for two seasons, Maha 1982183 and Maha 1983/84 in the same field.
Soil was analysed to determine the fertility level of the experimental field. Soil was extracted with 1N ammonium acetate a t pH 4.8 and plant samples were digested using sulfuric~perchloric acid mixture. K and Fe in extract solution were determined by atomic absorption spectrophotometry.

Effect of appliedphosphorus on iron toxicity in rice :
A second field experiment was conducted to determine the effect of applied phosphorus on iron toxicity in rice. The same location as in experiment 1 was used in Yala 1984 and Maha 1984185. In Yala 1985, the experiment was conducted in.Dodangoda in the same district.
A two factor factorial experiment was laid out in a randomized complete . block design with 3 replicates. The two variable parameters considered were rice varieties and phosphorus levels. The same varieties with same plant spacing were used as in experiment 1. Five gradedlevels of P a t 0.0,10.7,21.4,32.1 and 42.8 kgP/ha were applied only as basal at transplanting. Potassium was applied for all treatments a t 114.4 kg Wha to ensure K was not limiting. Nitrogen fehilizer levels, weed and pest control methods and data recorded were similar to those of experiment 1.
With increased applications of potassium, potassium content in leaves increased whereas the Fe content decreased ( Table 2). This effect was more pronounced in the susceptible variety (Bg 94-1) which had significantly higher Fe content at all levels of applied potassium ( Table 2). A significantly negative correlation was found between Fe and K content ofleaves in both varieties (Fig. 1).     tillera ns -not significant ISD -least significant difference At low levels of K, very severe bronzing symptoms corresponding to the tissue iron content of more than 700 mgkg were observed in the susceptible variety. The symptoms were rated to be mild at tissue iron content less than 450 mgkg. This level was obtained in the tolerant variety at 57.2 kg K h a and in the susceptible variety at 85.8 kg Wha.

Effect of Phosphorus
In 1984 Yala season, phosphorus levels did not significantly increase the yield in spite of decreased iron content in leaves (Table 4). However, experiments carried out in the following two seasons showed that plant growth and grain yield increased with levels ofapplied P. Yieldincrease was 19% for Bg94-1 whereas it was 26% for Bw 267-3. The highest yield level was obtained after application of 32.1 kg Ptha (Table 4). Visual observation also showed that bronzing symptoms slightly decreased at higher doses of P fertilizer (Table 4). Also at higher doses of P, the concentration of Fe in tissue was found to decrease while the concentration of P in tissue increased ( Table 4).  (Table 5). However, table 5 also shows that yield components were not significantly increased by P application except in 1884185 Maha season in which P application increased the number of tillers per square meter, and number of filled grains per panicle. But it was observed in Yala 1985 season, contribution of P levels to changes in the yield components was very low ( Table 5).
Application of P is not however, as effective as application of potassium in increasing the yield of rice. Under submerged conditions a large amount' of P is fixed as Fe -P.lS The need for more Pin these soils is therefore, related to P-fixation.
Singh & Singh14 have reported that P content increased with K application while Fe concenkation was drastically reduced indicating K-P synergism and K-Fe antagonism. Gupta et a1. l6 showed that application of P decreased the available and reducible Fe and uptake of Fe while increasing the dry matter yield of IR-8 rice variety. It has also been proved that excessive Fe build up can be reduced if iron toxic soils are fertilized with I?, K and Ca + Mg.6 Resulk obtained fiom both experiments showed that application of potassium and phosphorus increased yields of both varieties showing beneficial effects in potentially iron toxic soils. Such beneficial effects of potassium have been reported earlier using rice grown in solution culture.16 However, tolerant variety Bw 267-3 recorded higher yield than Bg 941 at all levels of potassium indicating tolerant variety will perform better with or without K.
Applied potassium and phosphorus increased the K and P content in both varieties but Bg 9 4 1 had higher K content than Bw 267-3 (Tables 2 & 4). Thus   Bg 94-1 has a higher potassium requir'ement under these conditions. Adequate tissue K and P content do not impart bronzing tolerance to a variety.
Applied potassium and phosphorus depressed tissue Fe content significantly in both varieties (Tables 2 & 4). However, Bg94-1 had significantly higher Fe content than Bw 267-3 ateeach level of potassium applied. The tulerant variety showed its ability to keep the tissue Fe content low in comparison to susceptible variety. However, it is not clearly known as to whether this is due to exclusion of iron in growth medium or reduced transl~cation.'~~~~ The significantly negative correlation between K and Fe content inleaves shows that by increasing the K content through enhanced K application, the Fe content could be lowered substantially. Since the tolerant variety has low levels of Fe in leaf tissues, less potassium is required to depress it to non toxic level.
It was observed that toxicity symptoms are reduced to a mild level at about 450 mgkg in both varieties in the K experiment. Thus for Bg 94-1, it corresponded to an application of 85.8 kg Wha whereas for Bw 267-3 it was at 57.2 kg Wha. But P experiment shows that 32.1 kg P/ha is sufficient for tolerant variety whereas 42.8 kg P/ha is required for susceptible variety to give moderate yield of rice when grown in iron toxic mineral soils.
The results of these two experiments demonstrate the beneficial effect of adding higher dosages of K and P fertilizer to the iron toxic soils. This is due to the amelioration of the deficiencies that increase the oxidizing capacity of roote. This is in conformity with the findings of Trolldenier,18 that low P and Kenhance iron toxicity through decreased oxidizing capacity of roots.
The overall results of this series of experiments have shown that a yield increase could be obtained by usipg the tolerant variety alone. Alao application of 85.8 kg K and 32.1 kg P per hectare has given significant yield increase of rice irrespective of the variety. The use of the tolerant variety and the higher dosage of P and K fertilizer gave highest yield increase compared with each treatment alone.