Published in Agron. J. 96:91-99 (2004).
© American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
RICE
Grain Yield and Kernel Smut of Rice as Affected by Preflood and Midseason Nitrogen Fertilization in Arkansas
Nathan A. Slaton*,a,
Edward E. Gbur, Jr.b,
Richard D. Cartwrightc,
Russell E. DeLonga,
Richard J. Normand and
Kristofor R. Bryed
a Dep. of Crop, Soil, and Environ. Sci., 1366 W. Altheimer Drive, Fayetteville, AR 72704
b Agric. Stat. Lab., Univ. of Arkansas, Fayetteville, AR 72701
c Univ. of Arkansas Coop. Ext. Serv., 2301 S. University, Little Rock, AR 72203
d Dep. of Crop, Soil, and Environ. Sci., 115 Plant Science Bldg., Fayetteville, AR 72701
* Corresponding author (nslaton{at}uark.edu).
Received for publication March 24, 2003.
 |
ABSTRACT
|
|---|
Kernel smut, Neovossia horrida (Tilletia barclayana), of rice (Oryza sativa L.) has been a persistent disease in Arkansas for more than 50 yr. Recently, epidemic levels of kernel smut have occurred on highly susceptible cultivars, which have increased awareness within the rice industry of this disease. Nitrogen fertilization treatments consisting of four preflood (50–152 kg N ha–1) and three midseason (0–100 kg N ha–1) N rates were arranged in a factorial design and used to delineate the effects of N rate and application timing on rice yield and disease incidence and severity in five site-year–cultivar studies (environments). Significant environment x preflood N rate interactions occurred for yield, incidence, and severity. The minimum preflood N rates, averaged across midseason N rates, that produced maximum yields varied among the environments and ranged from 84 to 152 kg N ha–1. Depending on the environment, disease incidence ranged from 2 to 93% and severity ranged from <0.1 to 4.8% among preflood N rates. Preflood N rate had no significant effect on smut incidence and severity for three environments receiving optimum-to-excessive N and a fourth environment receiving inadequate-to-optimum N. For the remaining environment that received optimum-to-excessive N, incidence increased linearly (43–93%), and severity increased nonlinearly (0.5–4.8%) as preflood N rate increased. Midseason N rate did not affect severity but caused a positive, linear increase for incidence. Data suggest that excessive preflood N has the greater potential to increase kernel smut but only when environmental conditions are favorable for kernel smut.
Abbreviations: PTBS, Pine Tree Branch Station • RREC, Rice Research and Extension Center • SPF, single preflood • 2WS, two-way split • 3WS, three-way split
|
INTRODUCTION
|
|---|
KERNEL SMUT OF RICE is caused by the fungus Neovossia horrida (Takah.) Padwick and Khan [=Tilletia barclayana (Bref.) Sacc. and Syd.] and has been a persistent disease in Arkansas for more than 50 yr (Templeton et al., 1960). Templeton et al. (1960) indicated that 1 to 10 rice grains per panicle may show partial or complete smut infection. This range of infection likely represents from <1 to 15% direct grain yield losses. Although kernel smut has historically been regarded as a minor disease with regard to grain yield, it can cause substantial loss of rice grain quality (Webster and Gunnell, 1992; Sharma et al., 1999). The undesirable effects of kernel smut on both grain quality and yield have increased the concern of rice farmers, millers, and buyers because of recent kernel smut epidemics on recently released and highly susceptible, long-grain cultivars like LaGrue, Cypress, Priscilla, and Cocodrie (Cartwright et al., 1999). The black teliospores from a few smutted grains are distributed to other healthy grains during rice harvest and processing, giving milled and parboiled rice an undesirable grayish color rather than the desirable white color (Gravois and Bernhardt, 2000). Rough rice containing more than 3% smutted kernels results in a special grade designation of "Smutty" (USDA, 1995).
Some fungicides provide partial control of kernel smut (Cartwright and Lee, 2001; Sharma et al., 1999). Application of propiconazole {cis-trans-1-[2-(2,4-dichlorophenyl)-4-propyl-1,3-dioxolan-2-ylmethyl]-1H-1,2,4-triazole} fungicide to rice in the late-boot stage is recommended for highly susceptible cultivars grown in fields with a history of severe kernel smut (Cartwright and Lee, 2001). Fungicide applications are strictly preventative since the visual symptoms of kernel smut do not appear until near maturity. Thus, the use of fungicides to control this disease is expensive and a short-term remedy. Growers are encouraged to use best cultural management practices, such as seeding less-susceptible cultivars, to reduce yield and economic losses attributed to kernel smut. However, the more resistant cultivars often have lower overall yield potential. Development of disease-resistant cultivars has been a primary objective of rice breeding programs in the Midsouth rice-producing area of the USA, but efforts have been focused primarily on major diseases, such as blast (Pyricularia grisea Sacc.) and sheath blight (Rhizoctonia solani Kuhn), that can cause catastrophic grain yield losses, rather than kernel smut resistance (Adair et al., 1973). Thus, cultivars with excellent overall agronomic traits that are susceptible to kernel smut will continue to be released. Gravois and Bernhardt (2000) concluded that kernel smut resistance, as well as other factors that contribute to pecky rice (kernels discolored or damaged by insects, disease, or various other means), is inherited quantitatively and the selection process for such an endeavor would be possible, but slow.
Nitrogen fertilizer recommendations have changed in response to phenotypic characteristics of rice cultivars over the past 30 yr. Compared to obsolete tall cultivars, the currently grown, stiff-strawed cultivars require higher total N rates to produce their yield potential. The excellent straw strength of modern cultivars allows growers to apply higher-than-recommended N fertilizer rates without a major concern for lodging. The use of excessive N fertilizer has been shown to increase the incidence and severity of kernel smut (Templeton, 1963; Kumar et al., 1978; Sharma et al., 1999; Slaton et al., 2001) as well as other rice diseases for many years (Groth and Lee, 2002). Atkins (1973) suggested, but provided no data, that late-season N applications increased kernel smut incidence. Data from a preliminary study conducted by Slaton et al. (2001) with modern cultivars contradict the observations of Atkins (1973).
Nitrogen is generally applied in a single, large application before flooding that accounts for 80 to 100% of the total N requirement, with a second split application applied, if needed, between panicle initiation and differentiation (Wilson et al., 1998, 2001). A review of rice fertilization practices in the USA by Norman et al. (2003) showed that studies to delineate rice yield response to N fertilization are routinely conducted, but few studies have examined how N fertilization practices affect kernel smut (Kumar et al., 1978; Sharma et al., 1999).
A better understanding of the interaction between N fertilization practices and disease development would better define the best management practices that could be implemented by growers to manage kernel smut without the use of fungicides. Therefore, the primary objective of this study was to determine the effects of N fertilizer rate and application timing on kernel smut incidence and severity and rice grain yield. We hypothesized that the application of N rates in excess of the rate required to produce optimum grain yield would increase both the incidence and severity of kernel smut, regardless of the application time or number of applications.
|
MATERIALS AND METHODS
|
|---|
Studies were conducted at the Rice Research and Extension Center (RREC) near Stuttgart, AR (34°18' N lat), and the Pine Tree Branch Station (PTBS), near Colt, AR (35°5' N lat), during 2001 and 2002. A total of five studies were conducted during this 2-yr period. The five studies will be referred to by their location (PTBS or RREC), year (2001 or 2002), and cultivar (Cocodrie or LaGrue) and were considered the environmental effect. The soil at the RREC was a Dewitt silt loam (fine, smectitic, thermic Typic Albaqualf) with an average soil water pH (1:2 soil-water suspension) of 6.0, Mehlich-3 extractable P of 14 mg P kg–1, and Mehlich-3 extractable K of 91 mg K kg–1. The soil at the PTBS was a Calhoun silt loam (fine-silty, mixed, active, thermic Typic Glossaqualf) with an average soil water pH of 7.5, Mehlich-3 extractable P of 9 mg P kg–1, and Mehlich-3 extractable K of 77 mg K kg–1. Soybean, Glycine max (L.) Merr., was the previous crop in all studies except the PTBS01 (Cocodrie) study, which followed rice in the rotation (Table 1). The research areas at both locations had been cropped in a soybean–rice rotation for at least 25 yr, and teliospores of the kernel smut fungus were assumed to be highly prevalent in the soil. Before seeding each study, 20 kg P ha–1 as triple superphosphate and 50 kg K ha–1 as KCl were broadcast-applied and incorporated by tillage. Due to high soil pH, 10 kg Zn ha–1 as ZnSO4 was also applied to each study conducted at the PTBS.
View this table:
[in this window]
[in a new window]
|
Table 1. Rice management information for five studies investigating the effect of N fertilization on rice grain yield and the incidence and severity of kernel smut at the Pine Tree Branch Station (PTBS) and Rice Research Extension Center (RREC) in 2001 (01) and 2002 (02).
|
|
The semidwarf Cocodrie (Linscombe et al., 2000) and short-statured LaGrue (Moldenhauer et al., 1994) cultivars were selected for these trials because they are both rated highly susceptible to kernel smut (Cartwright and Lee, 2001). Both cultivars were seeded at a rate of 110 kg ha–1 into plots consisting of nine rows that were 4.9 m long with 18-cm row spacings. Seeding, N application, and harvest dates for each study are summarized in Table 1.
Nitrogen fertilization treatments consisted of urea applied at four preflood N rates of 50, 84, 118, and 152 kg N ha–1 and three midseason N rates of 0, 50, and 100 kg N ha–1. The 12 treatment combinations of preflood and midseason N rates represent a wide range of total N applied (50–252 kg N ha–1) with three different strategies, which can be described as a single preflood (SPF), two-way split (2WS), or a three-way split (3WS) application. For the SPF fertilization strategy, N fertilizer was applied only at the four- to five-leaf stage immediately before establishing the permanent flood (preflood). The 2WS and 3WS strategies used identical preflood N rates as the SPF but included one (2WS) or two (3WS) midseason split applications of 50 kg N ha–1. For the 2WS and 3WS strategies, the first midseason N was applied when the topmost internode had elongated 1.0 cm, which generally corresponds to the beginning of reproductive growth (i.e., between panicle initiation and differentiation). A second 50 kg N ha–1 midseason application was made 7 d later for the 3WS application strategy (Table 1). Crop management practices with respect to irrigation and weed control were similar to guidelines recommended by the Cooperative Extension Service for the dry-seeded, delayed-flood rice cultural system (Slaton, 2001).
At physiological maturity, 20 panicles were randomly collected from the center seven rows of each plot and stored in brown paper bags until evaluated for kernel smut. For kernel smut evaluation, panicles were submerged in 0.27 M KOH overnight, rinsed three times in water, and then inspected over a light box to identify smutted kernels (Lee et al., 1992). Soaking panicles overnight in KOH made the rice hulls translucent, allowing for the identification of partially and completely smutted kernels. The numbers of smutted and total kernels on each panicle were counted, and the average number of kernels per panicle was calculated. The total number of partially and completely smutted kernels on the 20 panicles was summed, divided by the total kernel number, multiplied by 100, and reported as severity, or the percentage of the total kernels infected with kernel smut. The incidence of kernel smut in each plot was calculated by adding the number of panicles with at least one smutted kernel, dividing by 20 (number of panicles evaluated), and multiplying by 100.
At maturity, a 2.6-m2 area from the center four rows of each plot was harvested for grain yield using a small-plot combine. The reported grain yields were adjusted to a uniform moisture content of 120 g H2O kg–1 before statistical analysis was performed.
Each individual study was a randomized complete block design with four replications and a 4 (preflood N rate) x 3 (midseason N rate) factorial treatment structure. The five studies were combined and analyzed as a split plot, in which environment was the whole-plot factor and preflood and midseason N rates were the split-plot factors. Significant treatment effects and interactions were decomposed into linear (preflood and midseason N rates)- and quadratic (preflood N rate)-trend contrasts as appropriate (Table 2). Single degree-of-freedom contrasts were used to compare trends for specific pairs of environments (Table 3). The Fishers Protected Least Significant Difference (LSD) procedure (p = 0.05) was used to compare treatment means when appropriate. All statistical analyses were performed using SAS version 8.2 (SAS Inst., Cary, NC).
View this table:
[in this window]
[in a new window]
|
Table 2. Analysis-of-variance p values for rice grain yield, kernel smut incidence, and kernel smut severity as affected by preflood N rate, midseason N rate, and environment for five studies conducted during 2001 and 2002 at the Pine Tree Branch Station and Rice Research Extension Center.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Single-degree-of-freedom contrast results comparing individual test environments for linear and/or quadratic trends for grain yield, kernel smut incidence, and kernel smut severity response to preflood N (PFN) and midseason N (MSN) rates of five test environments at the Pine Tree Branch Station (PTBS) and the Rice Research Extension Center (RREC) in 2001 (01) and 2002 (02).
|
|
|
RESULTS
|
|---|
Grain Yield
Rice grain yield was significantly affected by the environment x preflood N (quadratic) rate (Table 2 and Fig. 1A) and environment x midseason N (linear) rate interactions (Table 2 and Fig. 1B). Single degree-of-freedom contrasts used to compare trends between individual tests indicated that yield response trends attributed to preflood N rate varied for each environment (Table 3 and Fig. 1A). The minimum preflood N rates, averaged across midseason N rates, that produced the statistically greatest grain yields were 84 kg N ha–1 at the RREC01 (Cocodrie), RREC02 (Cocodrie), and RREC02 (LaGrue); 118 kg N ha–1 at the PTBS02 (LaGrue); and 152 kg N ha–1 at the PTBS01 (Cocodrie). At the PTBS01 (Cocodrie), grain yield increased with the addition of increasing preflood N rates and may not have reached maximum yield despite the application of relatively high total N fertilizer rates (152–252 kg N ha–1). A higher N requirement was expected at this location since rice followed rice in the rotation with the previous years stubble incorporated immediately before seeding (Norman et al., 1990).
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 1. Effect of (A) preflood and (B) midseason N rate on Cocodrie or LaGrue rice grain yields in five studies conducted at the Pine Tree Branch Station (PTBS) and Rice Research Extension Center (RREC) during 2001 (01) and 2002 (02). For the environment x preflood N rate interaction (A), the LSD(0.05) values to compare grain yield among preflood N rates within an environment = 372 kg ha–1 and between environments = 382 kg ha–1. For the environment x midseason N rate interaction (B), the LSD(0.05) values to compare grain yield among preflood N rates within an environment = 322 kg ha–1 and between environments = 333 kg ha–1.
|
|
Total season N rates of 152 to 190 kg N ha–1 would have been recommended for the cultivars and previous crops used in these studies (Wilson et al., 2001). Based on rice grain yield responses, four of the five environments received preflood N rates in excess of the N rates required to maximize rice grain yield. The application of excessive N resulted in significant yield declines for only RREC01 (Cocodrie), indicating that maximum yields can be produced with a relatively wide range of preflood N rates that generally fall within University of Arkansas recommendations. In general, rice grown at the RREC produced higher maximum grain yields than rice grown at the PTBS.
Rice grain yields were also affected by a significant environment x midseason N rate interaction (Table 2 and Fig. 1B). Single degree-of-freedom contrasts demonstrated that the RREC01 (Cocodrie), RREC02 (Cocodrie), PTBS02 (LaGrue), and RREC02 (LaGrue) environments had similar slopes (Table 3) that were not significantly different from zero, indicating that midseason N rate had little effect on rice grain yields for these environments (Fig. 1B). Based on the rice yield response to preflood N rate at these four environments (Fig. 1A), a significant response to midseason N was not expected since low to intermediate preflood N rates were generally adequate to produce maximum grain yields. Previous research has clearly shown that midseason N is needed for maximum grain yield production only when the preflood N rate is inadequate or preflood N is mismanaged, resulting in excessive N loss (Norman et al., 1999; Ntamatungiro et al., 1999). The grain yield response trend to midseason N rate for PTBS01 (Cocodrie) was significantly different from the other four environments showing that grain yield increased when midseason N was applied (Table 3 and Fig. 1B).
The five test environments represent rice that received inadequate-to-optimum N fertilizer rates [PTBS01 (Cocodrie) and PTBS02 (LaGrue)] or optimum-to-excessive N fertilizer rates [RREC01 (Cocodrie), RREC02 (Cocodrie), and RREC02 (LaGrue)]. Although the grain yield responses varied among environments, sufficient ranges of yield responses to N fertilization were measured to examine the effect of N fertilization rate and application time on the incidence and severity of kernel smut.
Kernel Smut Incidence
Kernel smut incidence, defined as the percentage of panicles with at least one smutted kernel, was significantly affected by the main effect of midseason N rate and the environment x preflood N rate (linear) interaction (Table 2 and Fig. 2A). The incidence of kernel smut increased linearly as preflood N rate, averaged across midseason N rates, increased only for RREC01 (Cocodrie) and was significantly different than the trends for the other four environments (Table 3 and Fig. 2A). Single degree-of-freedom contrasts showed that the trends for PTBS01 (Cocodrie), RREC02 (Cocodrie), PTBS02 (LaGrue), and RREC02 (LaGrue) were similar (Table 3), with preflood N rate having little or no effect on kernel smut incidence within each environment (Fig. 2A). Despite the application of high amounts of total N at the PTBS01 (Cocodrie) and excessive N at the RREC02 (Cocodrie), PTBS02 (LaGrue), and RREC02 (LaGrue), kernel smut incidence did not differ among preflood N rates (Fig. 2A). However, the numerically highest kernel smut incidence for these environments generally occurred at the highest preflood N rate and tended to decrease as preflood N rate decreased.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2. Effect of preflood N rate, averaged across midseason N rates, on the (A) incidence and (B) severity of kernel smut of Cocodrie or LaGrue rice in five studies conducted at the Pine Tree Branch Station (PTBS) and Rice Research Extension Center (RREC) during 2001 (01) and 2002 (02). For the environment x preflood N rate interaction on incidence (A), the LSD(0.05) values to compare incidence among preflood N rates within an environment = 8.5% and between environments = 10.5%. For the environment x preflood N rate interaction on severity (B), the LSD(0.05) values to compare severity among preflood N rates within an environment = 0.5% and between environments = 0.7%.
|
|
Data from the RREC01 (Cocodrie) environment clearly showed that excessive preflood N increased kernel smut incidence (Fig. 2A). For the RREC01 (Cocodrie) environment, near-maximum grain yields were produced with 50 kg preflood N ha–1 (Fig. 1A), and disease incidence increased with each incremental increase in preflood N rate (Fig. 2A). The PTBS01 (Cocodrie) followed rice in the rotation and had relatively uniform, but high, kernel smut incidence across all preflood N rates despite the absence of excessive N. This suggests that the previous crop may affect the level of kernel smut in the subsequent crop, weather conditions have more influence than N management in determining the level of kernel smut, or both previous crop and weather influence kernel smut incidence. Data from four of the five site-year–cultivar environments suggest that N management plays a secondary role in determining kernel smut incidence. Kernel smut incidence was relatively low in both cultivars at the RREC in 2002 and although this study was not designed to compare cultivar susceptibility, incidence data from the RREC in 2002 suggest LaGrue is slightly more susceptible to kernel smut than Cocodrie.
The average kernel smut incidence differed among some of these environments, with the highest incidence occurring in 2001. The lowest range of kernel smut incidence (2–8%) occurred for PTBS02 (LaGrue), with no smutted kernels (i.e., 0% incidence) reported on panicles in 48% of the plots. The minimum kernel smut incidence values in 2001 were greater than the maximum incidence values in 2002, regardless of the location and cultivar (Fig. 2A). Although climatic data (Fig. 3) for July and August show slight differences among locations and years and incidence data suggest weather conditions were more favorable for kernel smut in 2001, the weather conditions during July and August alone do not provide conclusive evidence to explain the differences in kernel smut incidence between test environments. Weather conditions during July and August were presumed to be the most influential on kernel smut because most Arkansas-grown rice heads and matures during July and August, but perhaps the weather conditions at other times influence the presence and/or viability of kernel smut teliospores and affect the level of kernel smut in the subsequent rice crop.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 3. Minimum and maximum air temperatures and precipitation from 1 July to 31 August at the Pine Tree Branch Station (PTBS) and Rice Research Extension Center (RREC) during 2001 and 2002. Precipitation amounts >40 mm were entered as 40 mm. Temperature data were obtained from the NCDC (2003) for weather stations located at Wynne, AR (temperature data only), and Stuttgart 9ESE, AR (temperature and precipitation). At the PTBS, precipitation was recorded on site during 2001 and 2002.
|
|
During 2001, maximum and minimum daily air temperatures at 50% heading were higher at the RREC than at the PTBS, with both locations receiving frequent but small amounts of precipitation after heading (Fig. 3). However, the RREC received several additional rainfall events before heading. In 2002, precipitation was frequent both shortly before and after heading at both sites but in greater amounts than in 2001 when kernel smut incidence (Fig. 2A) and severity (Fig. 2B) were highest. This concurs with the observations of Singh (1975) that frequent, small amounts of precipitation provide a highly conducive environment for kernel smut infection. These data do not prove that relatively frequent and heavy rainfall provide a poor environment for kernel smut infection, but they do provide a direction for future research to investigate the specific weather conditions that favor kernel smut infection. Perhaps heavy rainfall physically rinses away the kernel smut sporidia and reduces the chance of infection.
Kernel smut incidence was not affected by the environment x midseason N interaction (Table 2), but it increased linearly as midseason N rate increased when averaged across preflood N rates and environments. Application of 50 and 100 kg N ha–1 at midseason, averaged across environments and preflood N rates, significantly increased kernel smut incidence by 5% compared with the 0 kg N ha–1 rate (34%). Although significant, the magnitude of increased kernel smut incidence from midseason N rate was generally less than the increase from increasing preflood N rates (Fig. 2A).
Kernel Smut Severity
Unlike kernel smut incidence, kernel smut severity, defined as the percentage of smutted kernels, was not affected by midseason N rate but was significantly affected by the environment x preflood N (quadratic) rate interaction (Table 2; Fig. 2B). Severity increased in a nonlinear fashion (i.e., quadratic response) as preflood N rate increased only for RREC01 (Cocodrie), which also showed a significant linear increase for incidence (Fig. 2). At the 152 kg N ha–1 preflood N rate, the percentage of smutted kernels per panicle (4.8%) may explain the majority of yield loss (5.5%) compared with the yield produced with the 84 kg N ha–1 preflood N rate for RREC01 (Cocodrie). On average, each smutted panicle (93% incidence) receiving a preflood N rate of 152 kg N ha–1 contained seven partially or fully smutted kernels compared with two smutted kernels per smutted panicle at 50 kg N ha–1 (43% incidence).
The other four environments had similar and nonsignificant trends that were different from RREC01 (Cocodrie) (Table 3; Fig. 2B). The number of smutted kernels per smutted panicle, averaged across N rates, was three for PTBS01 (Cocodrie), one for RREC02 (LaGrue), one for RREC02 (Cocodrie), and less than one for PTBS02 (LaGrue). The average number of total grains per panicle by environment was 137 for RREC01 (Cocodrie), 104 for PTBS01 (Cocodrie), 127 for RREC02 (Cocodrie), 127 for PTBS02 (LaGrue), and 132 for RREC02 (LaGrue).
|
DISCUSSION
|
|---|
The published literature indicates that increasing total N rate increases kernel smut incidence (Grewal et al., 1996; Kumar et al., 1978; Sharma et al., 1999) and severity (Kumar et al., 1978), but P and K fertilization rates have no effect on kernel smut (Sharma et al., 1999). However, these reports do not provide specific details describing how various N application rates and times influence both kernel smut and grain yield so that N fertilizer could be managed in a manner that would possibly minimize yield and quality losses from kernel smut while maximizing grain yield potential.
The data from RREC01 (Cocodrie) suggest that the preflood N rate and subsequent management practices that maximize rice growth and uptake of the preflood N have the greatest effect on kernel smut incidence and severity in Arkansas but only when weather conditions are favorable for kernel smut. Likewise, the application rate and subsequent management of preflood N have been shown to be the most critical factors determining rice grain yield potential in the dry-seeded, delayed-flood production system practiced in Arkansas and other Midsouth rice-producing states (Norman et al., 2003). The PTBS01 (Cocodrie) data hint that when a susceptible cultivar is grown under environmental conditions that are favorable for kernel smut, kernel smut incidence may be at relatively high levels even when preflood N rates are insufficient for the production of near-maximum grain yields. Three of the four environments that received excessive N fertilization also had low levels of kernel smut, indicating that weather or other environmental factors were primarily responsible for determining the incidence and severity of kernel smut.
Increasing the preflood N rate has also been shown to increase early-season sheath blight levels in rice (Cartwright et al., 2000). Savary et al. (1995) reported that high N rates increased both sheath blight contact with rice foliage and moisture retention below the canopy to create an environment favorable for its development and spread. Sharma et al. (1999) demonstrated that removal of flag leaves before heading reduced kernel smut incidence, suggesting that the lush growth associated with high grain yields and high N fertilizer rates may also create a more humid, favorable environment for kernel smut infection. Alternatively, Sharma et al. (1999) suggested that the flag leaf facilitated kernel smut infection by serving as a pathway for transporting moisture containing kernel smut sporidia directly to the boot and emerging panicle. If true, flag-leaf orientation (i.e., horizontal or vertical) could play a role in cultivar susceptibility.
Grain yield response data from the stiff-strawed cultivars, Cocodrie and LaGrue, show that rice generally has a broad range of total N rates that produce maximum yields before significant yield declines occur. Since lodging is not a major problem with most currently grown cultivars, growers often apply relatively high preflood N rates followed by topdress N applications at midseason to rice that, in many cases, does not need additional N to achieve its maximum yield potential. Soil test methods that could be used to refine the preflood or early-season N requirement are currently unable to accurately account for differences in native soil N release among soils; thus, uniform N application rates are recommended for similar soil textures in Arkansas (Wilson et al., 2001).
Based on grain yield data from the five environments of this study, current soil texture–based N fertilizer recommendations seldom result in underfertilization with N but more often result in maximum grain yield production from the application of excessive N on silt-loam soils, which can lead to increased levels of kernel smut in commercial rice fields. As the amount of excessive N increases, so does the potential for increasing levels of kernel smut. When a susceptible cultivar is grown under environmental conditions favorable for kernel smut, use of the minimum N rate required to produce maximum grain yields is essential to minimize grain and quality yield losses associated with kernel smut without the use of efficacious fungicides. Research efforts to develop a soil test method that can accurately predict native soil N availability for field-grown, flood-irrigated rice have not been successful. However, a method described by Kahn et al. (2001) holds promise that such a method can be developed in the near future and enable refinement of preflood N fertilizer rates on a field-by-field basis.
Since rice grain yield is most affected by preflood N management (Norman et al., 2003), current N fertilization guidelines in Arkansas recommend that growers apply a larger percentage of the total-season N rate preflood and reduce or potentially omit the midseason N applications. This strategy of fertilization has been shown to reduce the total N fertilizer requirement when used appropriately (Bollich et al., 1994; Norman et al., 1999). However, this N fertilization method also appears to have the greatest potential to increase kernel smut.
While our range of midseason N rates had no effect on kernel smut severity, they significantly increased disease incidence but not to the same magnitude as the range of preflood N rates. Therefore, accurately assessing whether rice requires topdressed applications of N fertilizer at midseason to produce maximum grain yields is also necessary to minimize rice yield and quality losses due to kernel smut. The plant area method described by Ntamatungiro et al. (1999) was developed to refine midseason N fertilizer rates but has not been readily adopted by Arkansas rice farmers. Other methods (e.g., chlorophyll meter, tissue N concentration, or total N uptake) of assessing the need for or the rate of midseason N applications have not been verified as reliable or are not practical to use in large-scale commercial production (Ntamatungiro et al., 1999). Research has yet to develop another method, which might be suitable for refining in-season N fertilizer needs in Arkansas.
Production of maximum grain yields, without increasing diseases like kernel smut, can only be accomplished by integrating pest management with nutrient management practices. Growers should use the lowest total N rates and application strategies required to produce grain yields that result in the greatest net returns. These data also suggest that the potential for increasing kernel smut incidence and severity may be minimized by use of the 2WS or 3WS application strategies, which use lower preflood N rates that are supplemented with N at midseason. Development of accurate methods for predicting early-season and assessing in-season N requirements that will be readily embraced by the rice industry is essential to integrate management practices, maximize economic returns to the grower, and prevent the use of excessive N. Because the effects of N management on kernel smut apparently depend on a host of other factors, perhaps future research should focus on identifying these environmental factors. The ability to accurately predict the conditions that lead to kernel smut epidemics and allow for the timely application of preventative fungicides would be most beneficial to growers.
|
ACKNOWLEDGMENTS
|
|---|
Appreciation is extended to the Arkansas Rice Research and Promotion Board for financial support of this research. Special thanks is extended to Marvin Bennett, Danny Boothe, Shawn Clark, Sixte Ntamatungiro, Charlie Parsons, Chuck Pipkens, and Elizabeth Sutton for their assistance in maintaining research plots and collecting data.
|
NOTES
|
|---|
Published with the approval of the Director, Arkansas Agric. Exp. Stn.
|
REFERENCES
|
|---|
- Adair, C.R., C.N. Bollich, D.H. Bowman, N.E. Joson, T.H. Johnston, B.D. Webb, and J.G. Atkins. 1973. Rice breeding and testing methods in the United States. p. 22–75. In Rice in the United States: Varieties and production. USDA-ARS Agric. Handb. 289. U.S. Gov. Print. Office, Washington, DC.
- Atkins, J.G. 1973. Rice diseases. p. 141–150. In Rice in the United States: Varieties and production. USDA-ARS Agric. Handb. 289. U.S. Gov. Print. Office, Washington, DC.
- Bollich, P.K., C.W. Lindau, and R.J. Norman. 1994. Management of fertiliser nitrogen in dry-seeded, delayed-flooded rice. Aust. J. Exp. Agric. 34:1007–1012.
- Cartwright, R.D., and F.N. Lee. 2001. Rice diseases. p. 87–100. In N.A. Slaton (ed.) Arkansas rice production handbook. Misc. Publ. 192. Univ. of Arkansas Coop. Ext. Serv., Little Rock.
- Cartwright, R.D., F.N. Lee, C.E. Parsons, W.J. Ross, S.R. Vann, and R. Overton. 1999. Monitoring of rice disease and on-farm evaluation of rice varieties in Arkansas. p. 148–156. In R.J. Norman and T.H. Johnston (ed.). B.R. Wells rice research studies 1998. Res. Ser. 468. Arkansas Agric. Exp. Stn., Fayetteville.
- Cartwright, R.D., N.A. Slaton, and R.J. Norman. 2000. Effect of nitrogen rate and application timing on sheath blight of rice in Arkansas. p. 74. In Proc. Rice Tech. Working Group, 28th, Mississippi State Univ., Biloxi, MS. 27 Feb.–1 Mar. 2000. Louisiana State Univ. Agric. Cent., Louisiana State Univ. Agric Exp. Stn., Rice Res. Stn., Crowley.
- Gravois, K.A., and J.L. Bernhardt. 2000. Heritability and genotype x environment interactions for discolored rice kernels. Crop Sci. 40:314–318.[Abstract/Free Full Text]
- Grewal, R.K., S.S. Chahal, and K.S. Aulakh. 1996. Effectiveness of triazole fungicides against kernel smut of rice. Indian Phytopathol. 49:404–405.
- Groth, D., and F. Lee. 2002. Rice diseases. p. 413–436. In W.E. Smith and R.H. Dilday (ed.) Rice: Origin, history, technology, and production. John Wiley & Sons, Hoboken, NJ.
- Kumar, I., M.S. Kang, and S.S. Saini. 1978. Fertilizer levels and incidence of bunt disease in rice in India. Int. Rice Res. Newsl. 3:5.
- Kahn, S.A., R.L. Mulvaney, and R.G. Hoeft. 2001. A simple soil test for detecting sites that are nonresponsive to nitrogen fertilization. Soil Sci. Soc. Am. J. 65:1751–1760.[Abstract/Free Full Text]
- Lee, F.N., G.E. Templeton, K.A.K. Moldenhauer, K.A. Gravois, R.H. Dilday, G.S. Jones, and D.H. Long. 1992. Management strategies for rice pests: Kernel smut research. p. 29–33. In B.R. Wells (ed.) Arkansas rice research studies 1991. Res. Ser. 422. Arkansas Agric. Exp. Stn., Fayetteville.
- Linscombe, S.D., F. Jodari, P.K. Bollich, D.E. Groth, L.M. White, Q.R. Chu, R.T. Dunand, and D.E. Sanders. 2000. Registration of Cocodrie. Crop Sci. 40:294.
- Moldenhauer, K.A.K., K.A. Gravois, F.N. Lee, R.J. Norman, J.L. Bernhardt, B.R. Wells, R.S. Helms, R.H. Dilday, P.C. Rohman, and M.M. Blocker. 1994. Registration of ‘LaGrue’ rice. Crop Sci. 34:1123–1124.[Free Full Text]
- [NCDC] National Climatic Data Center. 2003. On-line climate weather inventories [Online]. Available at http://lwf.ncdc.noaa.gov/oa/climate/stationlocator.html (accessed 7 Feb. 2003; verified 26 Oct. 2003). NCDC, Asheville, NC.
- Norman, R.J., J.T. Gilmour, and B.R. Wells. 1990. Mineralization of nitrogen from nitrogen-15 labeled crop residues and utilization by rice. Soil Sci. Soc. Am. J. 54:1351–1356.[Abstract/Free Full Text]
- Norman, R.J., C.E. Wilson, Jr., and N.A. Slaton. 2003. Soil fertilization and mineral nutrition in U.S. mechanized rice culture. p. 331–412. In C.W. Smith and R.H. Dilday (ed.) Rice: Origin, history, technology, and production. John Wiley & Sons, Hoboken, NJ.
- Norman, R.J., C.E. Wilson, Jr., N.A. Slaton, K.A.K. Moldenhauer, and A.D. Cox. 1999. Grain yield response of new rice cultivars to nitrogen fertilization. p. 257–267. In R.J. Norman and T.H. Johnston (ed.). B.R. Wells rice research studies 1998. Res. Ser. 468. Arkansas Agric. Exp. Stn., Fayetteville.
- Ntamatungiro, S., R.J. Norman, R.W. McNew, and B.R. Wells. 1999. Comparison of plant measurements for estimating nitrogen accumulation and grain yield by flooded rice. Soil Sci. Soc. Am. J. 91:676–685.
- Savary, S., P. Castilla, F.A. Elazegui, C.G. McLaren, M.A. Ynalvez, and P.S. Teng. 1995. Direct and indirect effects of nitrogen supply and disease source structure on rice sheath blight spread. Phytopathology 85:959–965.
- Sharma, R.C., S.S. Gill, D.P. Joshi, A. Rang, G. Bassi, and T.S. Bharaj. 1999. Kernel smut—a major constraint in hybrid seed production of rice and its remedial measures. Seed Res. 27:82–90.
- Singh, R.A. 1975. Source of inoculum and the epidemiology of bunt of rice. Riso 24:77–80.
- Slaton, N.A. (ed.). 2001. Rice production handbook. Misc. Publ. 192. Univ. of Arkansas Coop. Ext. Serv., Little Rock.
- Slaton, N.A., S. Ntamatungiro, E.A. Sutton, C.E. Parsons, R.D. Cartwright, R.J. Norman, and D.L. Boothe. 2001. Effect of nitrogen fertilizer rate and method of application on yield and kernel smut of ‘LaGrue’ rice. p. 210–218. In R.J. Norman and J.-F. Meullenet (ed.). B.R. Wells rice research studies 2000. Res. Ser. 485. Arkansas Agric. Exp. Stn., Fayetteville.
- Templeton, G.E. 1963. Kernel smut of rice as affected by nitrogen. Arkansas Farm Res. 12:12.
- Templeton, G.E., T.H. Johnston, and S.E. Henry. 1960. Kernel smut of rice. Arkansas Farm Res. 9(6):10.
- USDA. 1995. Official United States standards for rice. Available at http://www.usda.gov/gipsa/reference-library/standards/rice.htm (verified 26 Oct. 2003). USDA, Washington, DC.
- Webster, R.K., and P.S. Gunnell (ed.). 1992. Compendium of rice diseases. Am. Phytopathological Soc., Minneapolis, MN.
- Wilson, C.E., Jr., P.K. Bollich, and R.J. Norman. 1998. Nitrogen application timing effects on nitrogen efficiency of dry-seeded rice. Soil Sci. Soc. Am. J. 62:959–964.[Abstract/Free Full Text]
- Wilson, C.E., Jr., N.A. Slaton, R.J. Norman, C.A. Beyrouty, and D.M. Miller. 2001. Efficient use of fertilizer. p. 51–74. In N.A. Slaton (ed.) Arkansas rice production handbook. Misc. Publ. 192. Univ. of Arkansas Coop. Ext. Serv., Little Rock.
TOP
ABSTRACT
INTRODUCTION
