Rice Response to the Time and Rate of Potassium Fertilization

Transcription

1 RICE CULTURE Rice Response to the Time and Rate of Potassium Fertilization N.A. Slaton, B.C. Pugh, R.E. DeLong, S.D. Clark, R.J. Norman, and C.E. Wilson, Jr. ABSTRACT Rice (Oryza sativa L.) requires adequate potassium (K) nutrition for optimal growth, grain yield, and disease resistance. Studies were conducted in 2001 and 2002 at the Pine Tree Branch Experiment Station (PTBS) to characterize K uptake by rice and investigate the effect of both K-fertilizer application rate and time on rice growth, K uptake, and grain yield. Potassium fertilizer was applied at rates of 0, 30, 60, 90, and 120 lb K 2 O/acre before rice emergence, preflood, panicle differentiation, and at the lateboot stage (LB). Potassium uptake was monitored through plant sampling every 14 d after flooding during the growing season. Potassium fertilization increased rice yields only in Rice yields tended to decrease as the time of K application was delayed, but small numerical yield increases occurred when K was applied shortly before heading. Rice uptake of K was greatest during vegetative growth and declined with time. Between 40 and 60% of fertilizer K applied before flooding was recovered by the earlyboot stage and then declined as rice progressed to heading and maturity. Potassium fertilizer applications made before flooding at the 5-leaf stage are the most efficient. INTRODUCTION Rice requires adequate K nutrition for optimal growth, grain yield, and disease resistance (Slaton et al., 1995). Because the rice plant is relatively efficient in absorbing available soil K, K has not been a major yield-limiting factor in most Arkansas rice fields. For this reason, few studies have focused on the K nutrition of rice. We lack a good understanding of the seasonal uptake patterns of K and a general knowledge of how to manage K deficiency when it does occur. In the past 15 years, K deficiency of 329

2 AAES Research Series 504 rice has become more common not only in Arkansas fields but also in California and other rice-growing regions of the world. Most of our previous research efforts with K have focused on rice yield response to K fertilization rate. The results of many of these fertilization trials have documented that rice is a relatively efficient user of soil K. Because of this efficiency, K-deficient soils available for research are difficult to find. Although this information is useful for developing K-fertilizer recommendations it has not aided our fundamental understanding of how to manage K-deficient rice or characterized the trends for plant uptake of K during the growing season. Therefore, the objectives of this project were to investigate the effect of both K-fertilizer application rate and time on rice growth, K uptake and nutrition, and grain yield. MATERIALS AND METHODS A study was conducted at the PTBS in 2001 and 2002 on a Calhoun silt loam soil to evaluate the effect of K-fertilization rate and application time on rice grain yields. Before seeding, composite soil samples were taken from each plot to a depth of 4 in. to characterize the initial soil-chemical properties, including soil test-k. Potassium fertilizer (muriate of potash or KCl) was applied at rates of 0, 30, 60, 90, and 120 lb K 2 O/acre (0, 25, 50, 75, and 100 lb K/acre) before rice emergence (PE), preflood (PF), panicle differentiation or midseason (MS), and at the late-boot stage (LB). Each year the K fertilization treatments were identical, but the plots were established in different, adjacent locations. In 2001, rice followed soybean [Glycine max (L.) Merr.] in the rotation, but in 2002, rice followed rice with the grain and straw produced by the 2001 crop being removed immediately after harvest to lower the soil K. Potassium fertilizer was not applied to the crop grown previous to each year s study. Each year, Wells rice was drill-seeded (100 lb/acre) into a conventionally tilled seedbed in plots that were nine rows wide by 16 ft long with 7-in. row spacings. Whole-plant samples were harvested from a 3-ft row section every ~14 d after the preflood K application was made until heading; then a final sample was taken at maturity with panicles removed from the straw at the top node to determine K partitioning between the straw and grain. At each sample date only the control and the K treatments that had received K fertilizer were taken. Plant samples were dried at 60 C to a constant weight, weighed, ground, and a 0.25-g subsample was digested for elemental analysis. Total K uptake and K-fertilizer recovery efficiency were calculated from dry matter and tissue-k concentration at each sample time for only the 2001 data since analysis of 2002 data is not yet complete. The center four rows of each plot were harvested with a small plot combine at maturity for grain yield. Grain yields were adjusted to a uniform moisture content of 12% for statistical analysis. Each experiment was arranged as a randomized complete block design split-plot treatment arrangement with four (2002) or five replications (2001). Potassium application rate was the main plot factor with application time as the subplot factor. For drymatter and tissue-k concentration data, each sample date was analyzed separately. 330

3 B.R. Wells Rice Research Studies 2002 Mean separations were performed by Fisher s protected least significant difference (LSD) at a significance level of RESULTS In 2001, rice grain yield was not significantly affected by K application rate, time, or the rate time interaction (Table 2). The lack of grain yield response was not surprising since soil samples taken immediately before seeding averaged 113 mg K/kg (226 lb K/acre) and ranged from 99 to 132 mg K/kg (198 to 262 lb K/acre). In 2002, application of K, regardless of application rate and time, increased rice grain yield (Table 2). Grain yield tended to decrease as application time was delayed, but even the LB K application showed a numerical yield increase compared to the unfertilized control (Table 2). Based on current recommendations, K fertilizer would have been recommended in 2002 but not in Although K-deficiency symptoms were not observed, the data suggest that when K deficiency occurs, application of K before heading can nominally increase rice grain yields. In 2001, the main effects of K application rate and application time (None, PE, and PF; MS and LB had not yet been applied) significantly affected straw-k concentration at 47 and 61 DAS. Straw-K concentration increased as K rate increased (data not shown) and increased as the time of K application, averaged across K rates, was delayed (Fig. 1). Potassium fertilizer applied immediately before flooding resulted in higher tissue-k concentrations (Fig. 1) and total K uptake (Fig. 2) than PE applications during early vegetative growth. Total K uptake also increased as K rate increased (data not shown). At 75 DAS, after K fertilizer was applied at MS (66 DAS), tissue-k concentration (Fig. 1) and total K uptake (Fig. 2) were greatest when K was applied either PE or PF. Potassium fertilizer applied at MS, 6 d before samples were taken, numerically increased tissue-k concentrations and total K uptake, but was not significantly different than None (0 lb K/acre, Fig. 2). This suggests that K fertilizer applied at MS (66 DAS) was absorbed, but adequate time had not passed for complete uptake. Potassium application rate also affected straw-k concentration and total K uptake at 75 and 89 DAS (data not shown) but did not influence the panicle-k concentration or content at 89 and 122 DAS (data not shown). At 89 DAS, after K fertilizer was applied at LB (81 DAS), tissue K concentrations among PE, PF, and MS K applications were equal but greater than LB and None (Fig. 1). Again, data suggest that K applied at LB was being absorbed since the LB tissue-k concentration was significantly greater than None. Total K uptake 89 DAS showed that PE and PF were not different but were greater than the MS K applications (Fig. 2). Likewise, K applied at MS resulted in significantly greater K uptake than None but was not different than LB. Maximum K uptake rate by rice occurred between flooding (33 DAS) and midtillering (47 DAS) and midtillering and midseason (panicle differentiation, 61 DAS) (Fig. 3). This is of interest because K-deficiency symptoms of rice are normally first 331

4 AAES Research Series 504 expressed between panicle differentiation and heading, which corresponds to the time when K-uptake rate is very low and crop growth rate is at its highest (Fig. 3). The rapid crop growth rate, along with depletion of soil K, results in decreased tissue-k concentration during the growing season. Although K fertilizer was applied at MS and LB, it did not appreciably increase K-uptake rate during reproductive growth. The fact that most K was absorbed by rice roots during early-vegetative growth shows the importance of soil testing and fertilizer application before tillering begins. In this 2001 study, maximum plant uptake of K occurred between panicle differentiation and the mid-boot stage (between 65 and 89 DAS, Fig. 2). Potassium uptake trends and response to K application during reproductive growth may be different for K-deficient rice than as was shown in this 2001 study since K was not deficient. The average K-fertilizer recovery efficiency for K applied at PE and PF, averaged across the K application rates, ranged from 35 to 45% at 47 DAS, 33 to 53% at 61 DAS, 49 to 55% at 75 DAS, and 25 to 40 % 89 DAS (Fig. 4). In contrast, K applied at MS, averaged across K application rates, resulted in K recovery efficiency of only 10% at 75 DAS and 21 to 22% from 89 to 112 DAS. Although only 10% of fertilizer K applied at LB was recovered, most of the K was absorbed within 1 week after application. The data show that a large proportion of soil and fertilizer K is absorbed by rice during vegetative growth when the root-growth rate and K availability are high (Slaton et al., 1990). However, maximum root length is not reached until early reproductive growth, shortly after MS, which suggests that the soil becomes depleted of available K, K is lost via leaching, and/or K is somehow fixed in the soil. SUMMARY Data from the 2002 study suggest that nominal rice yield increases may occur when K fertilizer is applied to rice low in K after the onset of reproductive growth, but early application of K fertilizer is required to maximize yields on K-deficient soils. Additional studies, especially with severely K-deficient rice, are needed to confirm this observation. Data also suggest that K fertilizer recovery is approximately 40 to 60% when measured shortly before or after panicle differentiation and decreases with time. The decreased fertilizer recovery with time may be associated with K losses from older leaves and tillers. The K-uptake rate data from this study provide some insight on the reasons why K deficiencies are expressed at the onset of reproductive growth. LITERATURE CITED Slaton, N.A., C.A. Beyrouty, B.R. Wells, R.J. Norman, and E.E. Gbur Root growth and distribution of two short season genotypes. Plant Soil 121: Slaton, N.A., R.D. Cartwright, and C.E. Wilson, Jr Potassium deficiency and plant diseases observed in rice fields. Better Crops 79:

5 B.R. Wells Rice Research Studies 2002 Table 1. Selected soil chemical properties from K fertilization studies conducted at the Pine Tree Branch Experiment Station (PTBS) during 2001 and Mehlich 3-extractable nutrients Year Soil ph P K Ca Mg Zn (mg/kg) Table 2. Rice grain yield as affected by the main treatment effects of K application time and rate for studies conducted at the Pine Tree Branch Experiment Station in 2001 and K application rate (lb K 2 O/acre) (grain yield, bu/acre) LSD (0.05) NS z (P = ) 5 (P = ) K application time None Preemergence Preflood Midseason Late boot LSD (0.05) NS (P = ) NS (P = ) z NS = not signficant. 333

6 AAES Research Series 504 Fig. 1. Rice straw tissue-k concentration response to K application time, averaged across K application rates, at five sample times at the Pine Tree Branch Experiment Station during the 2001 growing season. The LSD(0.05) values or NS (not significant at 0.05) are shown below data points for each sample time. The symbols *, **, and *** indicate significance at the 0.05, 0.01, and probability levels, respectively. 334

7 B.R. Wells Rice Research Studies 2002 Fig. 2. Total and panicle-k content of rice as affected by K application time, averaged across K application rates, at five sample times at the Pine Tree Branch Experiment Station during the 2001 growing season. The LSD(0.05) values or NS (not significant at 0.05) are shown below data points for each sample time. The symbols * and *** indicate significance at the 0.05 and probability levels, respectively. 335

8 AAES Research Series 504 Fig. 3. The average seasonal rice-crop growth rate, averaged across all K-fertilizer treatments, and K uptake rate as affected by K application time, averaged across K fertilizer rates, of rice grown at the Pine Tree Branch Experiment Station during Values represent growth or uptake rates between the 5-leaf stage (33 DAS) and maturity (122 DAS). 336

9 B.R. Wells Rice Research Studies 2002 Fig. 4. The average K-fertilizer recovery percentages by K application time, averaged across K fertilizer rates, at five sample times during 2001 at the Pine Tree Branch Experiment Station. Fertilizer recovery efficiency values were calculated by difference. 337